Close
About
FAQ
Home
Collections
Login
USC Login
Register
0
Selected
Invert selection
Deselect all
Deselect all
Click here to refresh results
Click here to refresh results
USC
/
Digital Library
/
University of Southern California Dissertations and Theses
/
Solution processing of chalcogenide functional materials using thiol–amine “alkahest” solvent systems
(USC Thesis Other)
Solution processing of chalcogenide functional materials using thiol–amine “alkahest” solvent systems
PDF
Download
Share
Open document
Flip pages
Contact Us
Contact Us
Copy asset link
Request this asset
Transcript (if available)
Content
SOLUTION PROCESSING OF CHALCOGENIDE FUNCTIONAL
MATERIALS USING THIOL–AMINE “ALKAHEST” SOLVENT
SYSTEMS
by
Carrie L. McCarthy
__________________________________________________________________
A Dissertation Presented to the
FACULTY OF THE USC GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(CHEMISTRY)
May 2018
Copyright 2018 Carrie L. McCarthy
ii
Acknowledgements
The work presented in this dissertation would not have been possible without such an
empowering advisor. Prof. Richard Brutchey allowed me the freedom to explore my
interests, in both chemistry and sustainability, while providing guidance and motivation
to construct a meaningful dissertation. He allowed me to be curious and questioning,
without ever making me feel lesser for asking questions or needing clarification. I would
like to thank him for the time and thorough attention he gave my projects. Under his
guidance I applied for, and was granted an NSF Graduate Research Fellowship, which
allowed me to focus fulltime on research, and attend conferences around the world to
present my work; I will cherish these experiences always.
Thank you to my entire dissertation committee, Prof. Richard Brutchey, Prof. Mark
Thompson, and Prof. Josh West. You all have been great throughout the process, and I
am proud to have each of you on my committee. Thanks also to Prof. Sri Narayan and
Prof. Smaranda Marinescu for supporting me as members of my screening and qualifying
exam committee.
The Brutchey group has been an amazing work family during my time at USC. Together,
we worked hard and played hard, and I will never forget the times we shared, both inside
and outside of the lab. I have learned so much from each member, about science, about
beer, and about life. I could not have asked for better colleagues and friends to share my
days with.
I would like to thank USC, and the Chemistry Department in particular for providing an
excellent atmosphere to learn and grow in over these years. I specifically would like to
thank Prof. Mark Thompson’s group for allowing me to use their lab space and
equipment. I would also like to thank all of the collaborators that I worked with at USC:
Courtney Downes, Dr. David Webber, Dr. Jannise Buckley, Dr. Sean Culver, Dr. Patrick
Cottingham, Emily Schueller, and Karla Abuyen.
My dearest Mark, thank you for never doubting me a single time, and for giving me more
support than I could ever imagine. Your endless encouragement, unguarded honesty,
contagious laughter, and selfless willingness to be my metaphorical punching bag, have
made this journey seem doable through every trying moment. There is no one else I
would have rather had by my side through this, and there is no one I would rather be
taking on the future with.
I would like to thank my parents for always encouraging me to do my very best, and for
teaching me to do everything with integrity. They have always made me feel like I can do
anything I set my mind to, and they believe in me without fail. I am forever grateful that
they fostered in me a hard work ethic, a positive attitude, and a curious mind. I thank my
iii
sister, Mellissa, for being someone I could always look up to, especially regarding how to
work for what you want and how to be kind to everyone along the way. I thank my
brother, Daniel, for showing me what it looks like to know and be confident in who you
are, and to let that inspire the work you do.
I have made the truest of friends during my time at USC, and I cannot thank them enough
for their encouragement, support, and advice. Looking back at the adventures we went
on, the gallons of coffee we drank together, and the laughs we shared, makes this whole
experience all the more meaningful. Mark, Karol, Ola, Becky, Devang, Sean C., Arunika,
Emily, Sara, Sean M., Rahne, Shruti, Harsh, Helena, Ferran, Lena, Jimmy, and Nick, you
are the best friends I could ask for.
Finally, I would like to thank the National Science Foundation for the Graduate Research
Fellowship and the research grant that funded my work. The support I was provided gave
me confidence in taking ownership of my studies as well as my career.
iv
Table of Contents
Acknowledgements ii
List of Tables vii
List of Figures viii
List of Schemes xvi
Abstract xvii
Chapter 1. Solution Processing of Chalcogenide Materials Using Thiol–Amine
“Alkahest” Solvent Systems 1
1.1. Abstract 1
1.2. Introduction 2
1.3. The Alkahest Method 8
1.4. Alkahest-processed Photovoltaic Devices 16
1.5. Alkahest-processed Electrocatalysts 28
1.6. Alkahest-processed Flexible Substrates for Photodetectors 31
1.7. Alkahest-processed Thermoelectrics 32
1.8. Alkahest-processed Ligands for Colloidal Nanocrystals 33
1.9. Elucidation of Solvated Species and Dissolution Mechanisms 36
1.10. Conclusions 44
1.11. References 44
Chapter 2. Solution-Phase Conversion of Bulk Metal Oxides to Metal
Chalcogenides Using a Simple Thiol–Amine Solvent Mixture 51
2.1 Abstract 51
2.2 Introduction 51
2.3 Results and Discussion 54
2.4 Experimental 64
2.4.1 General Considerations 64
2.4.2 Preparation of Oxide Inks 65
2.4.3 Recovery of Products 65
2.4.4 Sb
2
Se
3
Thin Films and Photoelectrochemistry 65
2.4.5 Spectroscopic Characterization 66
2.4.6 Thermogravimetric Analysis 67
2.4.7 Powder X-ray Diffraction 67
2.4.8 Elemental Analysis 67
2.4.9 X-ray Photoelectron Spectroscopy 68
2.5 Conclusions 68
2.6 References 69
v
Chapter 3. Earth-Abundant CuSbS
2
Thin Films Solution Processed From
Thiol–Amine Mixtures 72
3.1 Abstract 72
3.2 Introduction 72
3.3 Results and Discussion 75
3.4 Experimental 84
3.4.1 General Considerations 84
3.4.2 Thin Film Characteriaztion 84
3.4.3 Elemental Analysis and Organic Content of Recovered
CuSbS
2
85
3.4.4 Structural and Optical Characterization 85
3.4.5 Preparation of CuSbS
2
Ink 86
3.4.6 Recovery of CuSbS
2
86
3.5 Conclusions 87
3.6 References 87
Chapter 4. Method for the Solution Deposition of Phase-Pure CoSe
2
as an
Efficient Hydrogen Evolution Reaction Electrocatalyst 91
4.1 Abstract 91
4.2 Introduction 91
4.3 Results and Discussion 93
4.4 Experimental 105
4.4.1 General Considerations 105
4.4.2 Experimental Methods 106
4.4.3 Microscopy 107
4.4.4 Recovered CoSe
2
Composition 107
4.4.5 Structural and Optical Characterization 108
4.4.6 Electrochemical Measurements 109
4.5 Conclusions 111
4.6 References 112
Chapter 5. Room Temperature Dissolution of Bulk Elemental Ni and
Se for Solution Deposition of a NiSe
2
HER Electrocatalysis 115
5.1 Abstract 115
5.2 Introduction 115
5.3 Results and Discussion 117
5.4 Experimental 128
5.4.1 General Considerations 128
5.4.2 Ink Preparation and NiSe
2
Recovery 128
5.4.3 Material Characterization 129
5.4.4 Electrochemical Measurements 130
5.5 Conclusions 132
5.6 References 133
vi
Chapter 6. Solution Deposited Cu
2
BaSnS
4-x
Se
x
from a Thiol-Amine
Solvent Mixture 138
6.1 Abstract 138
6.2 Introduction 138
6.3 Results and Discussion 141
6.4 Experimental 149
4.4.1 General Considerations 149
4.4.2 Content of Recovered Cu
2
BaSnS
4-x
Se
x
150
4.4.3 Structural and Optical Characterization 150
4.4.4 Preparation of Cu
2
BaSnS
4-x
Se
x
Inks 151
4.4.5 Recovery of Cu
2
BaSnS
4-x
Se
x
151
6.5 Conclusions 152
6.6 References 153
Appendix A. Dissolution of Nickel Precursors in Alkahest Solvent Mixtures
For Solution Deposition of Nickel Sulfides 157
A.1 Introduction 157
A.2 Results and Discussion 158
A.2.1 General Considerations 158
A.2.2 Dissoltuion and Recovery Experiments 158
A.2.3 UV-vis Experiments 167
A.2.4 Single Crystal X-ray Diffraction 168
A.3 Conclusions 170
A.4 Future Work 170
Bibliography 172
vii
List
of
Tables
Table 1.1 Precursor ink formulations, annealing conditions, and device
performances for PV devices using a thiol/amine-solution deposited
absorber layer. 15
Table 2.1 Oxides solubility limits and thermalized product characterization. 54
Table 3.1 Rietveld analysis of X-ray diffraction data of CuSbS
2
. 78
Table 4.1 Pertinent electrochemical values for three different CoSe
2
thin film
thicknesses. 101
Table 5.1 Rietveld analysis of X-ray diffraction data of NiSe
2
. 119
Table 5.2 Key values for electrocatalytic activity for various loadings of NiSe
2
on HOPG. 125
Table 6.1 Fitted peak positions and splitting values corresponding to the high-
resolution XPS spectra in Figure 6.5. 145
Table A.1 Examples of the appearance of inks prepared from five nickel
precursors in merc/en or EDT/en inks. The approximate time allowed
for dissolution at room temperature and all phases that have been
recovered from each ink are also listed. 160
viii
List
of
Figures
Figure 1.1 Nine V
2
VI
3
semiconductor inks prepared from bulk precursors. 8
Figure 1.2 (a) Photograph of the SnS/EDT/en ink (left) and a SnS thin film on
a glass substrate recovered from the SnS ink (right); (b) diffuse
reflectance UV-vis-NIR spectrum and corresponding Tauc plot
measured for the recovered SnS; (c) transient photocurrent
response of a spin coated SnS thin film. 11
Figure 1.3 (a) Lattice parameter and unit cell volume vs. composition for a
series of Sb
2
Se
3-x
S
x
alloys prepared by thermally annealing
Sb
2
O
3
/Se thiol/ amine inks; (b) Tauc plots for the same series of
alloys showing band gap tunability. 13
Figure 1.4 SEM micrographs of CZTS films prepared by the method in ref.
34. (a and c) Are top-down images before and after selenization,
respectively. (b and d) Are cross-sectional images before and after
selenization, respectively. 18
Figure 1.5 (a) XRD diffractograms of as-prepared (red) and selenized (black)
CZTSe thin films prepared using a Cu/Zn/Sn/Se-based ink on Mo
sub- strates; (b) Raman spectrum of selenized CZTSe thin film
showing an additional peak associated with a ZnSe impurity at 251
cm
-1
. 20
Figure 1.6 SEM micrographs of CIGSe thin films prepared from a
Cu/In/Ga/Se EDT/en ink, showing top-views of (a) an as-prepared
CIGSe film, and (b) a selenized CIGSe film, and cross-sectional
views of (c) an as-prepared CIGSe film and (d) a final CIGSe
device. 22
Figure 1.7 (a–d) SEM images of recovered (a) FeSe
2
, (b) Cu
1.8
S, (c) CuSe,
and (d) CoSe
2
materials, showing nanostructured morphologies;
(e) photographs of precursor inks; (f) CV curves for FeSe
2
and Pt
electrodes where Ox’ and Red’ correspond to the iodide/triiodide
redox species; (g) CV curves for Cu
1.8
S, CuSe, and Pt electrodes
in a polysulfide electrolyte; (h) polarization scans of CoSe
2
thin
films on HOPG substrates showing better surface utilization for
thinner films.
29
ix
Figure 1.8 (a) Ligand exchange of CdSe nanocrystals using various dissolved
bulk inorganic materials; (b) ligand exchange of various types of
nanocrystals using stibanate ligands from dissolved bulk Sb
2
S
3
; (c)
solution absorption spectra of CdSe nanocrystals before and after
ligand exchange with various dissolved bulk semiconductors; (d)
chopped photocurrent response of as-prepared (heated to 150 ˚C)
and ligand-exchanged (heated to 300 ˚C) CdSe films. 36
Figure 1.9 (a)
119
Sn NMR spectra of dissolved SnO, SnS, and Sn, showing
the same single peak at 217 ppm; (b) TGA/DSC data indicating
nearly identical decomposition traces for each of the three tin-
based inks. 41
Figure 2.1 Photograph of dilute (ca. 5 wt %) solutions of ten bulk oxides in a
1:4 vol/vol ethanethiol–1,2-ethylenediamine solvent mixture. 55
Figure 2.2 Powder XRD diffraction patterns of crystalline materials
recovered from annealing (a-f) metal oxide inks from three
different alkanethiol-en solutions, (g-k) selenium compound inks
from EtSH-en solutions, and (l-m) as-precipitated materials from
compound inks with tellurium. 57
Figure 2.3 (a) Dissolved Sb
2
O
3
and selenium in EtSH–en and a Sb
2
Se
3
film
on a glass substrate. (b) Cross-sectional SEM micrograph of a
Sb
2
Se
3
thin film on glass. (c) Linear voltage sweep of p-type
Sb
2
Se
3
with chopped 1-Sun simulated illumination in contact with
0.1m Eu(NO
3
)
3
(aq), and (d) transient photoresponse of Sb
2
Se
3
under potential control at -700 mV versus Ag wire. 58
Figure 2.4 TGA traces for five nominal compositions of Sb2Se
3–x
S
x
alloy inks
first dried to 125 ˚C under flowing nitrogen showing end point of
decomposition by ~300 ˚C. 59
Figure 2.5 FT-IR spectra of compound Sb
2
O
3
/Se inks from EtSH-en (a) dried
at room temperature under nitrogen flow and b) annealed under
nitrogen to 300 ˚C. After annealing to 300 ˚C, there is no evidence
for remaining organics from the solvent mixture.
59
Figure 2.6 Raman spectrum of crystalline Sb
2
Se
3
derived from a EtSH-en ink.
Bands could be assigned to Sb
2
Se
3
at 84 cm
-1
(translation), 118
cm
-1
(Se-Sb-Se bending) 190 cm
-1
(Se-Sb-Se bending), and 153
cm
-1
(Sb-Se Stretching). 60
x
Figure 2.7 XPS spectrum of Sb3d/O1s region of Sb
2
Se
3
recovered from an
EtSH-en ink showing a single Sb doublet and the absence of any
O1s signal. 60
Figure 2.8 (a) Powder XRD patterns of five compositions of
Sb
2
Se
3−x
S
x
alloys. (b) Lattice parameters and unit cell volumes of
the five solid solutions showing a linear dependence of unit cell
volume on composition. 62
Figure 2.9 Tauc plots showing band gap tunablity of five Sb
2
Se
3−x
S
x
alloy
compositions as measured by UV/vis diffuse reflectance
spectroscopy. The dotted traces correspond to as-bought bulk
Sb
2
Se
3
(E
g
= 1.2 eV) and Sb
2
S
3
(E
g
= 1.6 eV) powders. The alloys
possess direct band baps of 1.63, 1.52, 1.33, 1.27, 1.22 eV for
nominal x = 3, 2.25, 1.5, 0.75, and 0 compositions, respectively. 63
Figure 2.10 Band gap energy as a function of nominal alloy composition
showing a quadratic dependence E
g,dir.
(x) = 0.03175 x
2
+ 0.04743 x
+ 1.216 for Sb
2
Se
3–x
S
x
alloys. 64
Figure 3.1 (a) TGA traces of inks containing Sb
2
S
3
, Cu
2
S, and Sb
2
S
3
+ Cu
2
S
(CuSbS
2
precursor ink), showing end of mass loss by 300 °C in all
cases. (b) FT-IR absorption spectra of CuSbS
2
precursor ink dried
at 100 °C under vacuum and annealed to 300 °C under flowing
nitrogen. 75
Figure 3.2 Powder XRD patterns of alternative ink formulation (Cu:Sb 1:1) and
lower annealing temperature (350 ˚C), revealing impurity phases
when either Cu-poor conditions, or a lower annealing temperature, is
used. 76
Figure 3.3 (a) Powder XRD data from recovered phase-pure orthorhombic
CuSbS
2
and (b) Raman spectrum of recovered CuSbS
2
, showing a
single peak at 335 cm
−1
corresponding to the A
g
mode.
77
Figure 3.4 Rietveld analysis of the powder XRD pattern for recovered
CuSbS
2
shown as experimental data (circles), calculated pattern
(red trace), difference data (blue trace), and green tic marks
corresponding to the refined CuSbS
2
phase. 77
Figure 3.5 Space-filling polyhedra depiction of refined structure of recovered
CuSbS
2
, with CuS
4
tetrahedra, and SbS
5
square pyramids were Cu is
pink, Sb is red, and S is aqua. 79
xi
Figure 3.6 XPS spectra for the recovered CuSbS
2
. (a) Survey scan, and high-
resolution scans of the (b) Sb 3d, (c) Cu 2p, and (d) S 2p regions. 80
Figure 3.7 (a) Photograph of CuSbS
2
precursor ink and thin film of
CuSbS
2
on glass substrate and (b) cross-sectional SEM
micrograph of CuSbS
2
thin film deposited from precursor ink. (c
and d) Optical data derived from transmittance spectrum of
CuSbS
2
thin film. (c) Plot of absorption coefficient (α) as a
function of wavelength, and (d) Tauc plot extrapolated to estimate
a direct band gap of 1.6 eV. 81
Figure 3.8 SEM micrograph of a CuSbS
2
film made by depositing 9 layers of
precursor ink.
82
Figure 3.9 Plane view AFM micrographs of CuSbS
2
film surfaces made by
depositing (a) 2 layers and (b) 9 layers of CuSbS
2
precursor ink. 82
Figure 3.10 Transmittance spectrum of CuSbS
2
thin film on glass substrate. 83
Figure 4.1 (a) Picture of the filtered CoSe
2
precursor ink in merc-en solvent
system. (b) TGA trace of the dried CoSe
2
precursor ink (5 ˚C min
-1
under flowing nitrogen). (c) FT-IR spectra of the dried ink and the
ink annealed to 300 ˚C. (d) Rietveld refinement of XRD patterns
of CoSe
2
recovered from precursor ink after annealing to 350 ˚C. 94
Figure 4.2 Powder XRD patterns for CoSe
2
recovered by annealing precursor
inks to 350 ˚C in nitrogen atmosphere using (a) merc, (b) EDT,
and (c) EtSH.
95
Figure 4.3 (a) High-resolution TEM image of the CoSe
2
showing lattice
fringes corresponding to the (111) planes of marcasite-type CoSe
2
,
and (b) SEM micrograph of CoSe
2
recovered from the merc-en
solvent system. 96
Figure 4.4 XPS spectra of recovered CoSe
2
. (a) survey scan; (b) high-
resolution Co 2p; and (c) high-resolution Se 3d region. 97
Figure 4.5 Polarization curve of a 4-layer CoSe
2
film in 0.5 M H
2
SO
4
. 98
Figure 4.6 Tafel plot for a 4-layer sample where the Tafel slope is 55 mV
dec
–1
.
99
xii
Figure 4.7 Atomic force micrographs of (a) blank HOPG substrate, (b) 4L,
(c) 8L, and (d) 12L samples on HOPG. The rms roughnesses are
101 nm, 48.2 nm, 17.2 nm, and 35.3 nm, for blank HOPG, 4L, 8L,
and 12L samples, respectively. 100
Figure 4.8 Plot of average CoSe
2
loading on HOPG substrates (measured by
ICP-OES) and average film thickness (measured by SEM) as
functions of number of coats of precursor ink revealing a linear
increase of catalyst loading and film thickness with increased
number of ink coats. 101
Figure 4.9 Cyclic voltammograms at various scan rates used to derive double
layer capacitance values for (a) 4L, (b) 8L, and (c) 12L films. 102
Figure 4.10 Scan rate vs. current density difference at 0 mV vs. RHE from CV
curves for 4L, 8L, and 12L films. Where Cdl = slope/2 for y =
2.993x + 0.02366 (4L), y = 9.645x + 0.06266 (8L), and y = 13.64x
+ 0.08923 (12L). 102
Figure 4.11 Electrochemical characterization of the catalytic activity toward
HER. (a) polarization curves and (b) Tafel plots for different
CoSe
2
loadings. (c) 8 hour stability test measuring driving
potential required to produce a magnitude current density of 20
mA cm
-2
. 103
Figure 4.12 Polarization curves taken at 1 h intervals during extended stability
tests.
104
Figure 5.1 Image of unfiltered Ni and Se dissolved in the en/merc solvent
mixture. 117
Figure 5.2 (a) Filtered ink prepared from the dissolution of bulk Ni and Se in
a mixture of 4:1 (vol/vol) en/merc. (b) TGA trace of the filtered
ink. (c) XRD pattern of NiSe
2
recovered from decomposition of
the en/merc ink. (d) FT-IR spectra of the dried ink (blue) and after
annealing to 325 ˚C (red) under a N
2
atmosphere. 118
Figure 5.3 Rietveld refinement of XRD pattern of NiSe
2
recovered from the
precursor ink after annealing to 350 °C. Experimental (×) and
calculated (red) patterns are shown, along with the difference
curve (blue) and tick marks (green) corresponding to the refined
phase. 119
xiii
Figure 5.4 SEM-EDS spectrum of NiSe
2
recovered from en/merc ink. 120
Figure 5.5 XPS survey scan of recovered NiSe
2
. 121
Figure 5.6 High-resolution XPS spectra in the (a) Ni 2p and (b) Se 3d
regions. 122
Figure 5.7 (a) LSV traces for NiSe
2
and Pt/C electrodes on HOPG substrates.
(b) Tafel plot for an 8-layered NiSe
2
electrode on HOPG substrate.
(c) SEM micrograph of recovered NiSe
2
showing nanostructured
morphology. 123
Figure 5.8 Cyclic voltammograms at various scan rates used to derive C
dl
values for 4-, 6-, and 8-layered NiSe
2
electrodes. 124
Figure 5.9 Plot of non-Faradaic current density vs. voltage scan rate extracted
from the CV plots in Figure 5.8. C
dl
= slope/2 for y = 0.2402*x +
0.004956 (4 layers), y = 0.4691*x + 0.00514 (6 layers), y =
0.608*x + 0.006348 (8 layers). 125
Figure 5.10 HER electrocatalysis characterization of electrodes made with
various loadings (4L = 4 layers, 6L = 6 layers, and 8L = 8 layers)
of NiSe
2
on HOPG: (a) polarization curves of as-measured (solid)
samples and ECSA-corrected (dotted) data, and (b) Tafel plots
with linear regions extrapolated to show log(I
0
) at the x-intercept
for each of the three loadings. 126
Figure 5.11 EIS Nyquist plots (fitted data solid lines) for various loadings of
NiSe
2
at (a) η = 0.25 V and (b) η = 0.35 V. 127
Figure 5.12 Equivalent circuit model used to fit the EIS data. 128
Figure 6.1 Powder XRD pattern of material recovered from a precursor ink
that was only allowed to stir for 3 d at 50 ˚C; the BaS impurity is
likely due to incomplete dissolution of the bulk BaS precursor. 141
Figure 6.2 Ink decomposition as observed by (a) TGA trace of Cu
2
BaSnS
4
precursor ink and (b) FT-IR spectra for Cu
2
BaSnS
4
inks dried to
100 ˚C and annealed to 350 ˚C. 142
Figure 6.3 Powder XRD pattern of phase-pure Cu
2
BaSnS
4
recovered from an
EDT/en ink at 350 ˚C.
143
xiv
Figure 6.4 Tauc plot generated from the diffuse reflectance UV-vis spectrum
for Cu
2
BaSnS
4
recovered from EDT/en ink at 350 ˚C. 144
Figure 6.5 High-resolution XPS data for the recovered Cu
2
BaSnS
4
. (a) Cu 2p
region, (b) Ba 3d region, (c) Sn 3d region, and (d) S 2p region. 144
Figure 6.6 XPS survey scan of Cu
2
BaSnS
4
recovered from an EDT/en ink. 146
Figure 6.7 (a) Powder XRD patterns collected for each nominal composition
of Cu
2
BaSnS
4-x
Se
x
, the zoomed region to the right shows more
clearly the shifting of peaks to lower 2θ values with increasing
selenium content. (b) Plot of calculated a and c lattice parameters
and unit cell volume vs experimental composition showing a near
linear increase in all unit cell parameters with increased Se
content. 148
Figure 6.8 (a) Tauc plots of each solid solution composition with extrapolated
band gaps denoted and (b) plot of direct band gap vs experimental
composition in Cu
2
BaSnS
4-x
Se
x
showing a nearly linear decrease
in E
g,dir.
with increasing x. 149
Figure 6.9 SEM micrographs of solution deposited Cu
2
BaSnS
4
thin film. 152
Figure A.1 Ni-S phase diagram. 157
Figure A.2 XRD of as-purchased “Nickel Sulfide” from Alfa Aesar (black
trace) and of the recovered material from a merc/en ink made
using the bulk nickel sulfide as the precursor annealed to 350 ˚C. 160
Figure A.3 XRD patterns for various dissolution conditions for Ni + S in
en/merc inks. The dissolution reaction temperature and time are
given above each plot. 162
Figure A.4 XRD patterns of the bulk Ni(acac)
2
(bottom) and the [Ni(en)
3
]
2+
salt heterogeneous reaction product when Ni(acac)
2
is added to en
without stirring (top). 162
Figure A.5 XRD patterns for recovered materials from Ni(acac)
2
inks with
varying concentrations and annealing temperatures as denoted on
each of the patterns. (a) merc/en inks; (b) EDT/en inks. 163
Figure A.6 XRD patterns of the recovered materials from 100 mg mL
-1
Ni(acac)
2
inks with varying en/merc ratios as denoted above each
164
xv
pattern.
Figure A.7 XRD pattern of recovered material from 80 mg mL
-1
Ni(acac)
2
in
merc after centrifugation and annealing to 352 ˚C. 164
Figure A.8 XRD patterns from materials recovered by annealing Ni(COD)
2
in
merc/en and EDT/en, inks to 353 ˚C after centrifugation. 165
Figure A.9 Images of a Ni(COD)
2
ink at various stages of dissolution. (a) 10
min, (b) 4 h, and (c) 24 h. 165
Figure A.10 XRD patterns from materials recovered by annealing Ni(OH)
2
in
merc/en and EDT/en inks to 350 ˚C. 166
Figure A.11 XRD patterns from materials recovered by annealing NiCl
2
in
merc/en and EDT/en inks to 353 ˚C. 166
Figure A.12 (a) UV-vis spectra of increasing equivalents of en in an aqueous
Ni(acac)
2
solution and (b) plot of peak shifts for the spectra,
showing a plateau after ca. 3-4 eq of en. 167
Figure A.13 UV-vis spectra of nickel precursors in mixutres of (a) merc/en and
(b) EDT/en. 168
Figure A.14 Structures observed in the single crystal X-ray diffraction data for
crystals grown out of (a) merc/en Ni(acac)
2
and merc/en
Ni(COD)
2
, (b) merc/en Ni + S, and (c) EDT/en Ni(acac)
2
inks. For
(c) only the basic structures could be obtained from the data. 169
xvi
List
of
Schemes
Scheme 1.1 Proposed mechanism for the base-catalyzed thiol–sulfur reaction. 7
Scheme 1.2 Mechanism proposed for dissolution of Se in a thiol/amine
mixture. 42
xvii
Abstract
In the face of anthropogenic climate change we must begin to think and act based on
a systems approach rather than addressing problems as discrete events. Regarding the
design of products, we should not only design for sustainability in the direct use of the
product, but also for low-impact fabrication as well as possible reuse, upcycling, or
recycling. Specifically, for the design of functional inorganic materials, such as catalysts
and semiconductors, we must consider the various impacts, beyond functionality, of
choosing specific elements and sources to incorporate into materials. A second aspect of
a systems approach to functional materials is the design of versatile, low-impact, high-
throughput fabrication processes. Currently, the industry standards for fabricating
functional inorganic materials are energetically and financially burdensome, requiring
capital-intensive hardware and energy-intensive deposition processes. As such, there is
room for improvements toward more cost- and energy-effective manufacturing methods.
The deposition of molecular precursor inks prepared from bulk materials is an ideal
approach to material preparation, as it requires only a single dissolution step followed by
direct deposition onto a suitable substrate. Additionally, the composition of molecular
inks can be easily tuned by simply changing the formulation of the ink, and the very
nature of such precursor inks facilitates thorough mixing on the molecular level to yield
more compositionally homogeneous films.
With these advantages in mind, our group
developed a novel binary solvent system comprised of ethylenediamine (en) and a short
chain thiol that is capable of dissolving bulk inorganic precursors, including
chalcogenides, oxides, and elemental materials to give solution processable molecular
inks. Since its initial discovery, this “alkahest” solution processing method has been
employed for the deposition of a wide range of Earth-abundant functional inorganic thin
films including semiconductors for photovoltaic (PV) devices, electrocatalysts,
thermoelectrics, and photodetectors. This method is an example for which a systems
approach has been used to couple Earth-abundant compositions with sustainable
manufacturing and the possibility for reuse or recycling.
Since the seminal report on the solvent system, wherein the dissolution of nine bulk
V
2
VI
3
chalcogenide semiconductors were shown to dissolve at room temperature, under
ambient pressure, and over a short time, we and others have worked toward broadening
the scope of soluble precursors. We studied the solubility of ten bulk oxides, showing that
upon annealing their respective alkahest inks phase pure sulfides could be recovered. In
some cases, when elemental selenium or tellurium was added to the ink, selenides or
tellurides could be made. Using Sb
2
O
3
and selenium as an example system, we showed
how varying the nominal content of selenium in the ink formulation could lead to a series
of Sb
2
Se
3-x
S
x
alloys with tunable band gaps from 1.2 – 1.6 eV. Using Sb
2
S
3
and Cu
2
S as
precursors, we prepared high quality thin films of CuSbS
2
, an Earth-abundant alternative
to CuInGaS
2
as a PV absorber. We demonstrated that the alkahest-processed CuSbS
2
xviii
films had optoelectronic properties promising for application in PV devices.
In addition to absorber materials for PV, we applied the alkahest method to the
preparation of Earth-abundant alternatives to platinum group metal electrocatalysts for
the hydrogen evolution reaction (HER). We first prepared nano-structured marcasite-type
CoSe
2
by annealing an ink formulated using Co(OH)
2
and elemental selenium as
precursors. Thin films of CoSe
2
that were spin coated onto Highly Ordered Pyrolytic
Graphite (HOPG) facilitated HER with 100% Faradaic efficiency and key electrocatalysis
parameters on par with previous reports for similar materials prepared by other methods.
By comparing samples with various CoSe
2
loadings, we were able to demonstrate further
value of the alkahest method in the ability to easily vary the film thickness by simply
adjusting the number of ink coats, which lead to optimization of catalyst utilization. Next
we prepared pyrite-type NiSe
2
using elemental nickel and selenium as precursors.
Comparison of the HER performance for various NiSe
2
loadings showed good catalyst
utilization for all loadings, indicating the ability to use thicker films to achieve higher
hydrogen production per unit of geometric surface area.
In our most recent work, we reported on the first solution synthesis of
Cu
2
BaSn(S,Se)
4
(CBTS). This material has gained recent interest as an alternative to
Cu
2
ZnSn(S,Se)
4
(CZTS) because of its higher theoretical PV performance stemming
from the rendering of antisite defects energetically unfavorable upon replacement of zinc
with a larger atom, barium. We used Cu
2
S, BaS, and SnO as precursors in an
ethanedithiol/en-based formulation, which after several days of mixing gave a thick
solution that could be annealed to yield CBTS. Thorough material characterization
verified the phase purity, elemental composition, elemental oxidation states, and band
gap of the CBTS. Although the as-made ink was viscous and not able to be spin coated
directly, we were able to prepare uniform CBTS thin films by diluting the ink with 2-
methoxyethanol. We were also able to show that by incorporating elemental selenium
into the ink formulation a set of Cu
2
BaSnS
4-x
Se
x
alloys with experimentally determined
values from x = 0 – 2, could be prepared with tunable band gaps from 1.56-1.86 eV, thus
demonstrating a straightforward handle for engineering optimized photovoltaic absorbers.
1
Chapter 1. Solution Processing Of Chalcogenide Materials Using Thiol–
Amine “Alkahest” Solvent Systems *
*Published in Chem. Commun., 2017, 53, 4888-4902.
1.1. Abstract
Macroelectronics is a major focus in electronics research and is driven by large area
applications such as flat panel displays and thin film solar cells. Innovations for these
technologies, such as flexible substrates and mass production, will require efficient and
affordable semiconductor processing. Low-temperature solution processing offers mild
deposition methods, inexpensive processing equipment, and the possibility of high-
throughput processing. In recent years, the discovery that binary ‘‘alkahest’’ mixtures of
ethylenediamine and short chain thiols possess the ability to dissolve bulk inorganic
materials to yield molecular inks has lead to the wide study of such systems and the
straightforward recovery of phase pure crystalline chalcogenide thin films upon solution
processing and mild annealing of the inks. In this review, we recount the work that has
been done toward elucidating the scope of this method for the solution processing of
inorganic materials for use in applications such as photovoltaic devices, electrocatalysts,
photodetectors, thermoelectrics, and nanocrystal ligand exchange. We also take stock of
the wide range of bulk materials that can be used as soluble precursors, and discuss the
work that has been done to reveal the nature of the dissolved species. This method has
provided a vast toolbox of over 65 bulk precursors, which can be utilized to develop new
routes to functional chalcogenide materials. Future studies in this area should work
2
toward a better understanding of the mechanisms involved in the dissolution and recovery
of bulk materials, as well as broadening the scope of soluble precursors and recoverable
functional materials for innovative applications.
1.2. Introduction
Macroelectronics has emerged as a powerful industry in recent years with the widespread
deployment of technologies such as photovoltaic (PV) cells and flat panel displays. These
applications require new and innovative approaches to large area electronics, as the drivers are
not compatible with microelectronic architectures and fabrication facilities and costs.
1
High-
quality functional thin films (thickness ca. 10 nm for thin film transistors (TFTs) and up to 2
µm for PV) are the key component of these technologies, making inexpensive and light-
weight products a possibility.
1
For example, although the traditionally high processing costs of
conventional Si solar cells (single crystalline Si, commercial power conversion efficiency
(PCE) = 16–25%, lab = 26.3% single cell, and 24.4% module) have been greatly reduced with
the introduction of polycrystalline silicon modules (commercial PCE = 14–18%, lab PCE =
21.3% single cell, and 19.9% module), next generation thin film PV modules, using
alternative absorber materials such as CdTe or Cu(In
1−x
Ga
x
)(S
2−y
Se
y
) (CIGS), are making their
way back into the market as their efficiencies have risen to comparable values (commercial
PCEs = 7% for a-Si, 16.8% for CdTe, and single cell lab PCEs = 14% for a-Si, 22.1% for
CdTe, and 22.6% for CIGS) while their material processing costs remain significantly lower.
2–
5
Additionally, flat panel displays are a second leading thrust in commercially available
macroelectronic applications. This technology relies on uniform distributions of TFTs over a
3
substrate to govern the electronic switching of pixels. Previous efforts toward the
advancement of this industry have focused on the uniform deposition of TFT components
(e.g., 10 nm thin film semiconductors, <1 µm channel widths) over a wide substrate area while
retaining high electronic performance (e.g., high semiconductor mobilities and high switching
frequencies), with more recent research focusing on developing lower temperature (e.g. <300
°C) semiconductor deposition methods that are compatible with versatile substrates such as
plastics, foils, and paper for flexible display applications.
1,6
The prospective innovations pushing the macroelectronics industry forward reach beyond
displays and solar energy harvesting to include other large area technologies such as sensors,
imaging devices, electronic textiles, and radio frequency tags, which cannot feasibly be
brought to fruition via traditional semiconductor chip fabrication.
1
Therefore, leading research
in next-generation macroelectronics materials deposition and processing, in general, is
focusing on two main challenges whose combined value will continue to grow the market
share for thin film macroelectronics: (1) increased device efficiencies and (2) low-temperature,
high-throughput processing.
2,7
Using these parameters as a guiding framework, innovation in
the field of thin film functional materials requires working within the space of known
materials to develop techniques for optimizing thin film quality and composition, studying
thin film materials properties in order to realize the ideal functionality for each material,
searching for new material compositions using both theoretical and experimental approaches,
and improving upon the thin film deposition techniques with the goal of low-temperature,
inexpensive processing for scalable production.
4
Currently, some of the most widely used methods for depositing thin films include
physical vapor deposition (e.g., sputter coating and thermal evaporation) and chemical vapor
deposition; these two methods continue to dominate the thin film deposition market valued at
$7.35 billion in 2015, and forecasted to nearly double to $14.2 billion by 2020.
8
While these
methods are capable of depositing very high-quality thin films, they require a large amount of
energy input as a result of the ultra high vacuum and/or high temperature deposition
conditions. Additionally, these deposition methods necessitate processing equipment with
high capital costs, and the operational cost of the deposition methods is a major barrier for
companies looking to drive the overall cost of thin film technologies down to a competitive
level for the market.
8
Therefore, new, cost-effective processing strategies are needed.
Solution processing is an excellent approach being developed for thin film deposition, as it
is a low-energy, low-cost method. Effective solution processing would allow for high-
throughput manufacturing, using inexpensive deposition equipment such as roll-to-roll or ink
jet printers and spray coaters. Ideally, this could be performed under mild conditions (i.e., low
temperature and ambient pressure) leading to lower overall energy use and cost. Additionally,
solution processing enables the execution of new device architectures and versatile form
factors like non-planar or irregular coatings and flexible substrates. Two of the most
straightforward methods for solution processing involve the use of precursor inks: (1)
colloidally stable nanocrystal inks and (2) molecular precursor inks. While nanocrystal inks
require the initial preparation of nanocrystals, often including several washing steps prior to
formulation of the solution processable ink, molecular precursor inks offer a simpler
5
preparation requiring only the dissolution of precursors into an appropriately processable and
volatile solvent. Additionally, molecular inks allow for homogeneous mixing of the precursor
atoms such that a uniform thin film may be achieved.
One of the most versatile groups of functional inorganic materials available for thin film
applications, both in terms of composition and functionality, are chalcogenide compounds.
These materials are valuable for applications including, but not limited to, PV,
thermoelectrics, TFTs, and catalysts.
1,9,10
Certainly, the dissolution of bulk chalcogenide
materials to give molecular inks with the ability to subsequently recover phase-pure thin films
at low temperatures would be a valuable mechanism for the deposition of solution processed
functional thin films. In particular, the ideal system would have the capability to dissolve a
wide range of bulk materials quickly, and under benign conditions, thus building a versatile
toolbox of precursor options for depositing a wide variety of functional thin films.
Unfortunately, though, chalcogenides are generally insoluble in most traditional solvents, and
therefore, until recently, have not been studied as direct precursors for molecular inks for thin
film deposition.
To this end, Mitzi and coworkers first reported on the ability of hydrazine to dissolve bulk
SnS
2
and SnSe
2
in the presence of excess elemental chalcogen via the mechanism of
dimensional reduction. The process is facilitated by the in situ reduction of the elemental
chalcogen by hydrazine to produce chalcogenide anions that subsequently begin a series of
nucleophilic reactions with the bulk chalcogenide, breaking it down to eventually yield soluble
molecular metal chalcogenide species. This method yields molecular inks, soluble in
6
hydrazine and other polar solvents, that can be spin coated and thermally annealed to
temperatures <300 °C to yield phase-pure SnS
2−x
Se
x
thin films, the functionality of which was
demonstrated in high-mobility TFT devices.
11
Since that flagship report, a series of bulk
chalcogenide materials have been shown to dissolve in a similar way in hydrazine (e.g., SnS
2
,
SnSe
2
, In
2
Se
3
, ZnTe, In
2
Te
3
, GeS
2
, GeSe, GeSe
2
, Cu
2
S, Sb
2
Se
3
, Sb
2
S
3
, Sb
2
Te
3
, Bi
2
S
3
, HgSe),
and many useful chalcogenides have been solution processed from these precursors (e.g.,
SnSe
2−x
S
x
, In
2
Se
3
, GeS
2
, GeSe
2
, Cu
2
S, Sb
2
Se
3
, Sb
2
Te
3
, CuInSe
2
, Cu(In
1−x
Ga
x
)Se
2
, CuInTe
2
,
Ga
2
Se
3
, GeSbSe, KSb
5
S
8
, ZnTe, LiA
5
S
2
, and MoS
2
).
11–23
The advantage of the hydrazine method for recovering functional chalcogenide thin films
lies in the carbon-free and volatile nature of the solvent.
23
These characteristics allow for the
low-temperature volatilization of solvent molecules to proceed without the possibility of
carbonaceous impurities being left behind in the thin film, thus leading to high-quality thin
films. For example, thin film PV devices using hydrazine solution processed CIGS as an
absorber layer (ca. 1–2 µm thick) have reported remarkable gains in PCE, from the first report
in 2008 at ca. 10% PCE to the current hydrazine solution processed CIGS record of 17.3%
reported in 2016.
24,25
In addition, several groups have deposited high-quality semiconductor
thin films (<10 nm thick) for TFT devices using this hydrazine solution processing
method.
11,12,15,17
Unfortunately, hydrazine is toxic and explosive, making the direct scale-up of
this method for commercial applications impractical. Still, this gives an example of how a
simple dissolution and recovery technique for bulk chalcogenides can be advantageous toward
solution processing of functional thin films. Thus, the search has continued for solvent
7
systems with analogous solvation power and volatility, while additionally being less toxic and
safer to work with.
Our group wished to develop such an “alkahest”, or universal solvent, finding inspiration
in previous literature reports. For example, Vineyard et al. reported the mechanism for the
dissolution of bulk sulfur by thiol/amine mixtures, which was proposed to occur via an amine-
catalyzed reaction resulting in alkyl di-, tri-, and tetrasulfides (Scheme 1).
26
In this reaction,
the amine deprotonates the thiol and the resulting nucleophilic thiolate reacts with the bulk
sulfur (in the form of 8-membered rings) to create alkyl hydrogen polysulfide chains that
further iteratively react with thiolate molecules to produce a mixture of smaller polysulfide
products. Similarly, Liu et al. reported that bulk Se is soluble in a mixture of dodecanethiol
and oleylamine as a result of a redox reaction where dodecanethiol reduces selenium to yield
oleylamine–polyselenide complexes and didodecyldisulfide.
27
They showed that while slow
oxidation of the solution occurred in air to precipitate elemental selenium, this was reversible
when excess dodecanethiol was added back into the solution as an antioxidant.
Scheme 1.1. Proposed mechanism for the base-catalyzed thiol–sulfur reaction.
26
8
1.3. The Alkahest Method
In 2013, our group first reported on the remarkable ability of a binary solvent mixture of
ethanedithiol (EDT) and ethylenediamine (en) to dissolve nine bulk V 2VI 3 chalcogenide
materials at room temperature, ambient pressure, in air, and on the order of minutes (Figure
1.1).
28
In the following Feature Article, we will review the work that has been done to
investigate this alkahest solvent system. First, we discuss the early reports that revealed the
powerful and wide-ranging solvent ability of the alkahest. Next, we review the studies that
have demonstrated the utility of the alkahest as a powerful tool for solution processing of thin
films for applications including PV devices, electrocatalysts, flexible photodetectors, and
thermoelectrics. Then, we review the studies that have focused on the chemistry of alkahest
solutions, including mechanisms of dissolution and the identity of the resulting molecular
solutes. Finally, we conclude by discussing future research directions toward fully
understanding and utilizing this alkahest solvent system.
Figure 1.1. Nine V
2
VI
3
semiconductor inks prepared from bulk precursors.
28
9
In our first report on the alkahest system, saturated solutions of each chalcogenide
material were studied by thermogravimetric analysis (TGA), revealing impressively high
solubility limits for As
2
S
3
, As
2
Se
3
, As
2
Te
3
, Sb
2
S
3
, Sb
2
Se
3
, and Sb
2
Te
3
(21–32 wt%), a
moderate solubility limit for Bi
2
S
3
(9.8 wt%), and lower but still significant solubility
limits for the heavier Bi
2
Se
3
and Bi
2
Te
3
congeners (0.75 and 1.5 wt%,
respectively).
28
Furthermore, in the case of the sulfides and selenides (As
2
S
3
, As
2
Se
3
,
Sb
2
S
3
, Sb
2
Se
3
, Bi
2
S
3
, Bi
2
Se
3
), upon mild thermal annealing (270–350 °C) under a
nitrogen atmosphere the initial crystalline and phase-pure materials were recovered. Most
significantly for the purpose of thin film applications, we demonstrated how these
precursor inks could be used for solution processing by spin coating thin films from the
Sb
2
Se
3
and Bi
2
S
3
inks onto glass and silicon substrates followed by an annealing step to
yield uniform and high-quality crystalline thin films of phase-pure Sb
2
Se
3
and Bi
2
S
3
as
observed by scanning electron microscopy (SEM) and powder X-ray diffraction (XRD).
Following this initial account, we further reported on the alkahest showing the ability
to dissolve bulk gray selenium and tellurium.
29
The solubility limits of gray Se in a 1 : 4
(v/v) thiol/en mixture using ethanethiol, mercaptoethanol, or EDT were quite high (38,
40, 44 wt%, respectively) while the solubility limits for Te were a bit lower (9.3, 2.2, and
4.6 wt%, respectively). Interestingly, phase-pure Te (with no organic residues detected by
Fourier transform infrared (FT-IR) spectroscopy) was recovered by simply allowing the
ink to dry at room temperature. Upon mild annealing (250 °C), phase-pure gray Se was
also recoverable. We showed that in the presence of dissolved Se and gentle heating at 60
10
°C, elemental antimony could also be incorporated into the ink, which upon thermal
annealing under flowing nitrogen yielded phase-pure Sb
2
Se
3
. Under similar conditions,
elemental tin was combined with dissolved Te to give an ink from which SnTe could be
recovered. Thin films of both Sb
2
Se
3
and SnTe were easily deposited on glass substrates
by spin coating their respective inks followed by mild thermal annealing.
We next moved to the group IV monochalcogenide semiconductor SnS as a target for
alkahest solution processing because of its attractive optoelectronic properties for PV
applications. A typical ink was prepared by dissolving bulk SnS in a mixture of EDT/en
under a nitrogen atmosphere at 50 °C to yield a clear pale yellow solution as shown in the
image in Figure 1.2a (maximum room temperature solubility of 12 wt% SnS).
30
TGA and
FT-IR analysis indicated the end point of decomposition at 350 °C, and powder XRD
data confirmed that annealing to this temperature under flowing nitrogen facilitated
recovery of phase-pure orthorhombic SnS with a composition of 52 at% Sn and 48 at% S
by SEM-energy dispersive X-ray spectroscopy (SEM-EDX). Conversely, when annealed
in the presence of air, the ink gave a mixture of SnS, SnS
2
, and SnO. X-ray photoelectron
spectroscopy (XPS) data of the recovered SnS matched well with expected literature
values for SnS, and indicated a small contribution of surface oxidation (as evidenced by
the presence of Sn
4+
). Diffuse reflectance UV-vis spectroscopy was used to determine
both indirect (E
g,ind
= 1.1 eV) and direct (E
g,dir
= 1.3 eV) band gaps that are in good
agreement with literature values (Figure 1.2 b). The ink was ideal for solution processing
because the SnS films were not appreciably soluble in the alkahest at room temperature
11
on a short timescale, so multiple coats could be layered onto an FTO/glass substrate
(annealing to 350 °C after each coat) allowing for ca. 245 nm thick films of SnS to be
prepared (example film shown in Figure 1.2 a). A final treatment at 500 °C was added to
improve the film quality. The p-type transient photocurrent response of the films was
measured using a Eu(NO
3
)
3(aq)
electrolyte and 1 Sun chopped illumination to demonstrate
the potential use of the films for solar-energy conversion. As shown in Figure 1.2 c,
switchable photocurrent response was observed as the light was cycled on and off
(maximum photocurrent density of 170 µA cm
−2
at −700 mV vs. SCE), and the stability
of the films toward photodegradation was confirmed by a 3 h stability test.
Figure 1.2. (a) Photograph of the SnS/EDT/en ink (left) and a SnS thin film on a glass substrate
recovered from the SnS ink (right); (b) diffuse reflectance UV-vis-NIR spectrum and
corresponding Tauc plot measured for the recovered SnS; (c) transient photocurrent response of a
spin coated SnS thin film.
30
12
Further broadening the scope of alkahest chemistry, in 2015 we reported on the
dissolution of ten bulk oxide materials in thiol/en (where the thiol was ethanethiol,
mercaptoethanol, or EDT), all at room temperature, ambient pressure, in air, and all in
less than 24 h to give inks that were stable for up to several months.
31
The room
temperature solubility limits of the inks were found to be 25–30 wt% for PbO, 15–20
wt% for Ag
2
O, CdO, Sb
2
O
3
, and Bi
2
O
3
, 10–15 wt% for ZnO and SnO, and 5–10 wt% for
Cu
2
O, GeO
2
, and As
2
O
3
. Phase-pure sulfides (i.e., ZnS, Ag
2
S, CdS, SnS, Sb
2
S
3
, and PbS)
were recovered by annealing the inks under a nitrogen atmosphere (300–375 °C), where
the sulfur source was the thiol in the solvent system. In some cases, the phase of the
recovered material was dependent upon which thiol was used in the ink formulation. For
example, ethanethiol or mercaptoethanol/en inks of CdO and ZnO yielded phase-pure
zinc blende CdS and ZnS, respectively, while EDT/en inks gave a mixture of the zinc
blende and wurtzite phases when annealed to the same temperature, thus indicating that
the alkahest's composition may have a kinetic effect on the recovered phase. Compound
inks were prepared by mixing the metal(oid) oxide inks with dissolved selenium or
tellurium. In this way, Cu
2
Se, ZnSe, CdSe, and Sb
2
Se
3
were recovered after annealing an
ink obtained by mixing dissolved selenium with the corresponding oxide ink. Phase-pure
Ag
2
Se, Ag
2
Te, and PbTe immediately precipitated out of solution upon combining the
respective oxide and chalcogen inks.
As an example of the practicality of this method, alloys relevant for PV applications
with the composition Sb
2
Se
3−x
S
x
(confirmed by elemental analysis) were recovered by
13
simply adjusting the nominal amount of selenium in various Sb
2
O
3
/Se compound inks.
The recovered solid solutions followed Vegard's law for ideal alloys as confirmed by
powder XRD, and a systematic decrease in their direct band gap from 1.63–1.22 eV with
decreasing x value was observed (Figure 1.3). This highlights the utility of the alkahest
for facile tuning of relevant material properties. Finally, the functionality of the prepared
thin films was demonstrated by the cycled on/off p-type transient photocurrent response
observed for Sb
2
Se
3
thin films on FTO/glass using a standard 3-electrode electrochemical
cell and Eu(NO
3
)
3(aq)
redox mediator (maximum photocurrent density of ca. 130 µA cm
−2
under 1 Sun illumination with applied bias of −700 mV vs. Ag wire).
Figure 1.3. (a) Lattice parameter and unit cell volume vs. composition for a series of Sb
2
Se
3-x
S
x
alloys prepared by thermally annealing Sb
2
O
3
/Se thiol/ amine inks; (b) Tauc plots for the same
series of alloys showing band gap tunability.
31
With such a diverse toolbox of precursors, we sought to push our exploration toward
the possibility of using the alkahest method to synthesize more complex materials. We
reported on alkahest-solution processed CuSbS
2
, an Earth-abundant and non-toxic
14
alternative to CIGS or CdTe for PV applications.
32
Both Sb
2
S
3
and Cu
2
S were dissolved
in a mixture of mercaptoethanol/en to formulate an ink with slightly Cu-poor
stoichiometry that yielded phase-pure orthorhombic CuSbS
2
(by powder XRD) upon
annealing to 400 °C under nitrogen. The band gap and visible range absorption
coefficient of the recovered CuSbS
2
were determined by diffuse reflectance spectroscopy
to be 1.6 eV and ca. 1 × 10
5
cm
−1
, respectively, in agreement with previous reports on
this material and nearly ideal for a single junction PV device. Multiple coats could be
layered (with annealing at 350 °C between layers), allowing for thickness-tuned uniform
thin films to be deposited. The van der Pauw method was used to determine the free hole
concentration (n = 3.18 × 10
19
cm
−3
), and resistivity (ρ = 3.04 × 10
−3
Ω cm) of the films,
indicating the characteristics of a heavily doped p-type semiconductor. Additionally, the
hole mobility of the films was found to be 64.6 cm
2
V
−1
S
−1
, one of the highest reported
for thin films of CuSbS
2
.
Following our initial report on the alkahest system, there have been reports on similar
derivative solvent systems.
43
For example, Pan's group has reported on the thiolglycolic
acid/ethanolamine system, which they used to formulate inks from CuO, Ag
2
O, ZnO,
SnO, In(OH)
3
, Ga(acac)
3
, Cd(OH)
2
, B
2
O
3
, Sb
2
O
3
, Bi
2
O
3
, MnO, Fe(acac)
2
, Ni(acac)
2
,
Cu
2
O, CuCl
2
, Cu(Ac)
2
, and Cu(acac)
2
precursors, and have since further developed the
use of such thiolglycolic acid/amine solvent mixtures to formulate inks using various
other precursors and to prepare alloyed and doped materials.
33,44–47
In their initial report,
they demonstrated the utility of this system by depositing thin films of CdS, SnS, CuInS
2
,
15
CuSbS
2
, Cu(InGa)S
2
, and Cu
2
ZnSnS
4
. Luminescent Ag-doped Zn
x
Cd
1−x
S quantum dot
films were also prepared. Most notably, though, the authors reported on the first working
PV device where the absorber layer was solution processed using a thiol/amine molecular
ink. The specific details regarding the devices in this work will be described in detail in
the following section. This work was the beginning of a steady stream of PV device
reports using thiol/amine solution processing of absorber layers that continues to expand
upon the functionality of thin films derived this way. The following section on PV
devices will present key reports in chronological order of their publication, drawing
attention to the significant findings of each. Table 1.1 includes ink formulations,
annealing conditions, and corresponding device performances.
Table 1.1. Precursor ink formulations, annealing conditions, and device performances for
PV devices using a thiol/amine-solution deposited absorber layer.
Absorber
Material
Precursors
Ink
Formulation
a
Dissolution
Conditions
b
(temperature,
time,
atmosphere)
Recovery
Conditions
(˚C)
c,b
PCE
d
Voc (V)
d
Jsc
(mA cm
-2
)
d
FF
d
E
g
(eV)
e,b
Ref.
1 CZTSSe
CuO,
ZnO, SnO
0.6:1:2
TA/EA/MO
Et
65 ˚C, 30 min,
air
320/540 6.83% 0.424 31.5 51.1% 1.11 33
2 CZTSSe
Cu
2
O,
SnO, ZnO
1:1:2
EDT/EA/MO
Et, dilutied
with MOEt
to 0.66 M
total metal
RT, NA, air 320/540 7.34% 0.436 34.0 49.5% 1.12 34
3 CZTSe
Cu, Zn,
Sn, Se
1:2 TA/EA,
diluted with
MOEt to
0.35 M total
metal
55 ˚C, hours,
air
320/540 8.02% 0.408 33.4 58.8% 1.00 35
16
4 CZTSSe
CuCl,
ZnCl
2
,
SnCl
2
, Se,
S
1:4 or 1:2
propanethio/
HA
RT, 2 h,
nitrogen
250/500 7.86% 0.382 34.4 60.1% 1.08 36
5 CIGSe
Cu, In,
Ga, Se
1:10
EDT/en
60 ˚C, hours,
argon
350/550 9.50% 0.528 26.6 67.5% 1.17 37
6 CZTSSe
Cu
2
Se,
SnSe, Zn,
S
1:1
EDT/HA,
dilution with
HA to
[Sn]=0.1
RT, hours-
days, nitrogen
300/500 6.84% 0.360 35.5 53.5% 1.08 38
7 CZTSSe
Cu, Sn,
Zn, S, Se
1:1
EDT/HA,
dilution with
HA to
[Sn]=0.1
RT, hours-
days, nitrogen
300/500 7.02% 0.390 33.0 56.1% NA 38
8 CIGSe
Cu
2
Se,
In(OAc)
3
,
Ga(acac)
3
,
Se
1:10
EDT/HA
RT, overnight,
nitrogen
325/500 12.2% 0.560 33.3 65.4% 1.17 39
9 CISe
Cu
2
S,
In
2
S
3
1:10
EDT/en,
dilution with
ethyl acetate
RT, overnight,
nitrogen
310/550 8.0% 0.452 31.4 56.2% 1.03 40
10 CIGSe
In
2
S
3
,
Cu
2
S, Ga,
Se
1:10 EDT/en
,dilution
with ethyl
acetate
RT, overnight,
nitrogen
310/NA 9.80% 0.528 30.7 60.2%
1.15-
1.2
41
11 CdTe Te, CdCl
2
1:3
propanethiol/
en
RT, overnight,
nitrogen
500/550 0.50% 0.460 3.89 30% 1.44 42
a
All ratios are vol/vol. HA = hexylamine, TA = thioglycolic acid, EA = ethanolamine, MOEt = 2-methoxyethanol.
b
NA
= not reported.
c
Between coats temperature (˚C) / selenization temperature (˚C), in entry 11 the second value is a final
annealing step without excess chalcogen.
d
Champion device data. All devices had the typical structure,
glass/Mo/Absorber/CdS/i-ZnO/ITO/contact electrode, except entries 9 and 10 which did not have an ITO layer, entry 7
which included a MgF
2
antireflective layer, and entry 11 which had the structure FTO/CdS/CdTe/Cu/Au.
e
As
extrapolated from device EQE data.
1.4. Alkahest-processed Photovoltaic Devices
The predominant demonstration of functionality for alkahest-processed thin films has been
their use as absorber layers in PV devices. This is not surprising as many chalcogenide
17
materials are semiconductors with optimal band gaps for single junction devices, and are
especially fit for thin film devices as they have direct band gaps with visible region absorption
coefficients on the order of 10
4
cm
−1
.
48
Thin film PV arrays based on CIGS (lab record 22.6%
by ZSW), CdTe (lab record 22.1% by First Solar) and a-Si (lab record 14% by AIST) are
already commercially available with efficiencies up to 16% (CdTe, vide supra).
3
Kesterite
materials, with reported efficiencies up to 12.6% (lab record for CZTSSe, IBM),
3,49
such as
CZTS, CZTSe, and CZTSSe are very promising Earth-abundant alternatives to CIGS and are
less toxic than CdTe.
In the first devices reported by the Pan group, a CZTSSe absorber layer was prepared
using a 2-methoxyethanol/thioglycolic acid/ethanolamine mixture with CuO, ZnO, and SnO
as bulk precursors. After dissolution, the solution was spin coated as multiple layers on
Mo/glass substrates, to produce ca. 2 µm thick CZTS films that were then selenized to
incorporate selenium and facilitate grain growth. The devices were completed by chemical
bath depositing a CdS layer, RF sputtering intrinsic ZnO, DC-sputtering ITO, and thermally
evaporating an Al collection grid electrode to achieve a device with a PCE of 6.83%.
The Pan group followed their initial work with a second report on CZTSSe devices, this
time exchanging the thioglycolic acid for EDT, and using the ternary mixture of
EDT/ethanolamine/2-methoxyethanol to dissolve Cu
2
O, SnO, and ZnO precursors.
34
To
formulate the final CZTS precursor ink, the initial solution was further diluted with 2-
methoxyethanol to yield a total metal concentration of 0.66 M. TGA of this ink showed the
end of decomposition by 320 °C, at which temperature XRD confirmed that tetragonal CZTS
18
was obtained. Approximately 2 µm thick CZTS thin films on Mo coated glass were obtained
after eight cycles of spin coating and annealing. XRD analysis confirmed the phase purity of
kesterite CZTSSe films achieved via selenization of the as-prepared CZTS films, and Raman
spectroscopy further verified that no impurity phases were present. EDX analysis revealed
ratios of Cu/(Zn + Sn), Zn/Sn, and S/(S + Se) as 0.94, 1.22, and 0.047, respectively. XPS data
suggested the valence states for Cu, Zn, and Sn were +1, +2, and +4, respectively, for both as-
prepared and selenized films. Field emission-SEM (FE-SEM) images confirmed that upon
selenization, the films coarsened into large grains with sizes on the order of the film thickness,
as shown in Figure 1.4 c and d. Additionally, a smaller fine-grained layer was observed
directly under this large-grained layer (Figure 1.4 d). While their champion device (PCE =
7.34%) showed an increased PCE from their previous report, the authors attribute the lower
fill factor (FF) of 49.5% (values from literature devices possess FF ∼ 60%) to a high series
resistance, R = 2.1 Ω cm
2
, resulting from recombination and insufficient carrier transport at the
fine-grained CZTSSe sub-layer. Device EQE data indicated a band gap of 1.12 eV for the
CZTSSe material, while UV-vis diffuse reflectance data gave an optical band gap of 1.03 eV.
Figure 1.4. SEM micrographs of CZTS films prepared by the method in ref. 34. (a and c) Are top-
down images before and after selenization, respectively. (b and d) Are cross-sectional images before
and after selenization, respectively.
34
19
Bulk elemental metals were used to prepare precursor inks for the deposition of PV
absorber layers in a thiol/amine solution for the first time by the Pan group in late
2014.
35
Cu, Zn, and Sn metal were co-dissolved with Se in a solution of thioglycolic acid
and ethanolamine. To decrease the viscosity of the solution for solution processability,
they diluted the ink with 2-methoxyethanol to give a total metal concentration of ca. 0.35
M. Several coats were iteratively spin coated/annealed to yield CZTSe films with a
thickness of ca. 1.6 µm. XRD analysis showed that the as-deposited films were
comprised of ca. 8 nm CZTSe nanoparticles, while selenized films had increased
crystallinity and grain size (also seen by SEM). Although XRD indicated phase-pure
kesterite CZTSe (Figure 1.5 a), Raman spectroscopy was used to detect a small amount
of a ZnSe impurity (251 cm
−1
) in addition to kesterite CZTSe (172, 196, and 234 cm
−1
) in
the selenized films, as seen in Figure 1.5 b. The ink formulation targeted Cu/(Zn + Sn)
and Zn/Sn ratios of 0.77 and 1.2, respectively, and EDX analysis of pre- and post-
selenization films both showed increased ratios of Cu/(Zn + Sn) = 0.91 and Zn/Sn = 1.25.
The as-prepared films contained a small amount of sulfur from the decomposed
thioglycolic acid, but after selenization no sulfur remained in the material (within the
detection limits of EDX analysis), and XPS verified the expected valence states of Cu
1+
,
Zn
2+
, Sn
4+
, and Se
2−
. SEM/EDX line scans of the selenized films show a ca. 1 µm-thick
large-grained layer on top of an ca. 850 nm-thick carbon-rich fine-grained layer,
consistent with other reports on solution processed CZTS/Se films.
33,34,50,51
Devices with
a standard CZTS/Se structure were fabricated, with an average PCE of 7.7% and
champion PCE of 8.0%.
20
Figure 1.5. (a) XRD diffractograms of as-prepared (red) and selenized (black) CZTSe thin films
prepared using a Cu/Zn/Sn/Se-based ink on Mo sub- strates; (b) Raman spectrum of selenized
CZTSe thin film showing an additional peak associated with a ZnSe impurity at 251 cm
-1
.
35
As an extension of the original alkahest system, various mono- or dithiols with primary
monoamines and/or diamines can also be utilized in binary solvent mixtures to formulate
molecular inks, though it is important to keep in mind the need for a mixture with sufficient
volatility for low-temperature recovery of the desired material. The Agrawal group
demonstrated the solubility of several metal precursors (e.g., Zn, ZnO, Zn(OAc)
2
, Cu
2
O,
Cu(OAc)
2
, CuCl
2
, Cu(acac)
2
, and Sn(acac)
2
Cl
2
) in such solutions.
36
They used propanethiol
and hexylamine to make a CZTSSe precursor ink by combining separate solutions of CuCl +
21
ZnCl
2
+ SnCl
2
and Se + S. The resulting inks were spin cast in multiple coats followed by
selenization to give high quality ca. 1 µm thick CZTSSe films as observed by SEM. Similar to
our previous finding that the identity of the thiol in the alkahest formulation could affect the
recovered phase of the chalcogenide, when propanethiol and butylamine, instead of
hexylamine, were used in the ink formulation, wurtzite phase CZTS was obtained with low
temperature annealing. Selenization converted the material to the kesterite phase, however
these films gave lower device performances compared to the propanethiol/hexylamine-
processed films as a result of the observed pinholes formed during the wurtzite to kesterite
phase transition. Variation in Cu content in chalcogenide materials is known to affect defect
densities.
52
The effect was studied by varying the concentration of Cu in the precursor ink,
which allowed for simple tuning of the Cu content in the final selenized films such that an
optimum formulation of Cu/Sn = 1.53 was reached, leading to a CZTSSe film with a band gap
of 1.08 eV and a champion device efficiency of 7.9%. Such a simple yet effective ink
modification once again exemplifies the versatility of this method for fine-tuning material
properties to affect application optimization.
The first report of a solution processed purely selenide CIGSe device (and first alkahest-
processed chalcopyrite-type device) was published by the Pan group in 2015.
37
They utilized
the EDT/en system to simultaneously dissolve elemental Cu, In, Ga, and Se at 60 °C to
formulate a precursor ink that could be used at room temperature to spin coat CIGSe thin
films. While Cu and In were soluble on their own within ca. 20 min, it was found that
elemental Ga was not, and that it required the presence of Se for dissolution to occur. Nine
22
iterations of spin casting/annealing followed by selenization to improve film quality yielded
CIGSe films with a bilayer structure of ca. 600 nm-thick large grains on top of a fine-grained
layer (Figure 1.6 d). These absorber layers were integrated into a typical CIGS device
architecture to achieve a champion device efficiency of 9.5% with V
OC
= 0.53 V. These device
parameters are comparable to the reported PCE (10.3%) and V
OC
(0.61 V) for the highest
efficiency hydrazine-processed CIGSe device at the time (without Sb-doping), and in fact
the J
SC
and FF are higher for the thiol/amine processed device.
24
The lower overall
performance of the thiol/amine device may be attributed to charge recombination occurring at
the fine-grained carbon-containing CIGSe layer below the coarsened larger grains remaining
after selenization (which is not observed in the hydrazine based device). Subsequent work has
been done to optimize the selenization process in order to avoid such fine-grained layers.
53
Figure 1.6. SEM micrographs of CIGSe thin films prepared from a Cu/In/Ga/Se EDT/en ink, showing
top-views of (a) an as-prepared CIGSe film, and (b) a selenized CIGSe film, and cross-sectional views
of (c) an as-prepared CIGSe film and (d) a final CIGSe device.
37
In Agrawal and co-workers' second report using the thiol/amine dissolution
chemistry, they demonstrated the dissolution of several bulk chalcogenides (including,
23
for the first time CuS, CuSe, and SnSe), as well as elemental metal precursors (Cu, Zn,
Sn, and In) in binary mixtures of EDT and primary amines (butylamine and hexylamine)
at room temperature.
38
The inks were stable for months with solubility limits ranging
from 0.3–0.78 M. In this work, the authors present the first example of recovering a
binary chalcogenide material from an alkahest ink where the only precursor is an
elemental metal. ZnS (mixed phase zinc blende/wurtzite) films were recovered (300 °C)
from an ink containing Zn as the only precursor, using EDT as a sulfur source. Additional
thin films of phase-pure CuS and SnS (from their corresponding bulk powders) as well as
CTSSe, CZTS, and CZTSSe were deposited by spin coating their respective precursor
inks followed by thermal annealing to 300 °C. Two different precursor inks were used to
deposit CZTSSe films for PV devices, the first consisting of dissolved Cu
2
Se, SnSe, Zn,
S and Se, and the second consisting of purely elemental components of Cu, Sn, Zn, S,
and Se. STEM-HAADF imaging was used to investigate an as-deposited film made from
the first ink (annealed to 300 °C). The results clearly show layers formed by multiple
coats of ink, comprised of in situ formed nanoparticles (<5 nm). A post-annealing
selenization step was used to facilitate grain growth for both types of films, after which a
large grain/fine grain bilayer was observed as in previous examples (vide
supra).
34,35
Both inks resulted in devices with similar PCEs (i.e., 6.84% for the first ink
and 7.02% for the second, purely elemental precursor ink).
Next, the Agrawal group reported on the dissolution of non-chalcogenide precursors
(i.e., Cu(OAc)
2
, In(OAc)
3
, Cu(acac)
2
, In(acac)
3
, Ga(acac)
3
CuCl, InCl
3
, GaCl
3
, and GaI
3
)
24
in a mixture of EDT or propanethiol and hexylamine at room temperature.
39
They also
found that Cu
2
Se and In
2
Se
3
were only soluble if the dithiol was used in the mixture and
not when propanethiol was used. The authors set out to use these precursors to formulate
an ink to solution process high quality, sulfur-depleted CIGSe films for PV devices. XRD
and Raman spectroscopy confirmed the recovery of CIGSe, with no sign of a sulfide
CIGS phase, after thermal annealing and selenization of an ink using Cu
2
Se, In(OAc)
3
,
Ga(acac)
3
, and Se as precursors. They achieved a relatively high-efficiency ultra thin
(520 nm) CIGSe device (PCE = 10.3%); however, several observations lead the authors
to a final optimized preparation method for (thicker) absorber layer devices: (1) they
observed that while increased annealing temperature and time facilitated more complete
incorporation of Cu, these conditions also lead to higher crystallinity of the as-deposited
CIGSSe films that, in general, tend not to coarsen into large grains during the selenization
process. (2) The as-deposited film appeared to be fairly porous, so a soaking procedure
prior to selenization was developed to address this. The as-deposited films were soaked in
an EDT/hexylamine solution containing only the cation (i.e., Cu, In, Ga, and no Se)
precursors for 10 min, followed by drying at 300 °C. This helped prevent a sandwiched
fine-grained layer from forming within the CIGSSe films during selenization, however,
they could not eliminate a thinner, fine-grained layer toward the bottom of the films. The
authors postulated that the soaking step served to decrease the porosity of the as-
deposited layer, such that during selenization less liquid Se was able to intercalate to the
bottom of the CIGSSe layer to nucleate grain growth. (3) When chloride salts were used
as precursors for the inks, the annealed films were always Ga-poor due to the formation
25
and subsequent low-temperature volatilization of GaCl
3
. After optimization based on
these observations, their most successful method was to use an ink formulation with a
precursor ratio of Cu/In/Ga/Se = 0.95 : 0.7 : 0.3 : 2.6. Multiple coats (annealing to 325
°C between layers) of the ink followed by the cation solution soaking/drying step and
selenization gave a 1.1 µm-thick film (<100 nm fine grain layer) with a composition of
Cu
0.91
In
0.70
Ga
0.30
Se
2.05
by SEM-EDX. Such films gave devices with an average PCE of
11.8% (champion 12.2%).
Soon after, Bowers and coworkers also studied as-deposited chalcopyrite absorber
film morphology as a key factor in increasing device performance.
40
By utilizing spray
pyrolysis instead of spin coating to deposit CuIn(S,Se)
2
(CIS) precursor inks, they
showed that selenization of denser, as-deposited films is advantageous toward obtaining
large grained absorber films. They spray coated mixtures of Cu
2
S and In
2
S
3
dissolved in
EDT/en onto Mo-coated glass at 310 °C in air to yield CIS precursor films. They
compared two selenization techniques, the first in a tube furnace and the second in a rapid
thermal processing (RTP) oven, and determined that the RTP oven facilitated faster
temperature ramping leading to a thinner MoSe
2
layer, and thus a lower series resistance
and improved fill factor when compared to the tube furnace-selenized film. Upon
selenization using an RTP oven, the majority of the sulfur in the films was displaced by
selenium to give a single phase of CISe with a band gap of 1.03 eV. Device efficiencies
up to 8% PCE were achieved.
The Bowers group followed this work with a report in which they utilize the
26
straightforward compositional tunability of the alkahest method to study the effects of
varying Ga content in CIGSSe-based devices.
41
Individual solutions of In
2
S
3
, Cu
2
S, and
Ga + Se in EDT/en were prepared, combined in the desired Cu-poor stoichiometric ratio,
and diluted with ethyl acetate to form the precursor ink. The films were deposited by
spray coating the ink onto a Mo–glass substrate held at 310 °C with 90 s allotted for
drying in between layers. They varied the nominal precursor inks to aim for compositions
of Cu
0.9
In
1−x
Ga
x
Se
2
with x = 0, 0.1 (0.08), 0.2 (0.19), 0.3 (0.28), and 0.4 (0.39), where the
value in parenthesis is the measured composition in the selenized film by EDX.
Interestingly, the selenized films retained nearly the same Cu content as was initially used
in the precursor ink formulation, which is different than the loss of Cu observed by
Agrawal and coworkers in their report where spin coating was used to deposit thiol/amine
CIGSSe precursor inks.
39
The authors attribute the retention of Cu to a lack of layer re-
dissolution in the case of spray deposition as opposed to spin coating (in agreement with
the Agrawal group's findings). Thus, the effect of layer re-dissolution by the solvent
system is a key consideration when targeting specific compositions of spin-cast thin
films. Characterization by XRD and Raman indicated that phase-pure alloys are achieved
in all cases, except where x = 0, in which case there appeared to be phase segregation
(CuInS
2
, CuIn
5
S
8
, and Cu
2−x
S). The selenized CIGSe films were incorporated into
devices that gave EQE spectra with systematic blue-shifting onsets as the Ga content was
increased, as expected for the higher band gap materials. A champion PCE of 9.8% was
achieved for a device where the CIGSe absorber layer had a graded band gap prepared by
decreasing the Ga content in the ink as subsequent layers were deposited.
27
Recently, the Agrawal group has expanded upon the use of alkahest solution
processing of absorber layers to include CdTe.
42
In this work, an ink containing Te and
CdCl
2
in a mixture of propanethiol/en was spin coated onto a CdS-coated FTO–glass
substrate followed by thermal annealing on a hot plate at 500 °C. A total of 35–40 coats
were required to achieve a 1.5 µm-thick film. While these as-prepared films are mainly
CdTe, XRD verified that a final annealing step in an evacuated ampoule at 550 °C for 30
min is required to remove Te and other impurities. Notably, in contrast to sulfide and
selenide films that require the presence of their respective chalcogen atmosphere for grain
growth, the final CdTe films consist of micron-sized grains without a Te atmosphere in
the final annealing step. In fact, when a Te atmosphere was used, the authors observed
Te-rich films, whereas SEM-EDX analysis revealed a Te/Cd ratio of ca. 1 : 1 before and
after the final annealing step without a Te atmosphere. The final films contained a small
amount of S and Cl, with slight shifts of XRD peaks to higher 2θ values indicating that
the sulfur incorporation was uniform throughout the film. It was concluded that the
source of the sulfur was due to reaction with the thiol as well as partial dissolution of the
CdS layer. Carbon was also present in the annealed films, though a bilayer morphology
(as reported for the CZTS and CIGS/Se films, vide supra) was not observed.
34–40
For
device fabrication, the films were annealed in air after every five coats to incorporate
oxygen in the growth process (this has been beneficial for CdTe devices made by other
techniques),
54
and the final films were finished with a nitric-phosphoric acid etch. Poor
rectifying behavior and low device efficiencies (0.5% PCE) were observed and attributed
to several factors including high recombination and shunting stemming from
28
consumption of the CdS layer and the presence of C, Cl, and S impurities.
1.5. Alkahest-processed Electrocatalysts
While most examples using the alkahest chemistry toward fabrication of PV devices
involve the deposition of absorber layers, in 2015 Liu et al. utilized the method to deposit
efficient transition metal chalcogenide thin film counter electrodes (CEs) for sensitized
solar cells.
55
Tetragonal FeSe, monoclinic Cu
1.9
S, and monoclinic Cu
2
Se precursors were
dissolved in EDT/en solutions at 60 °C for ca. 5 h to yield the precursor inks. Notably,
after annealing the inks to 330 °C, new phases and/or stoichiometries (orthorhombic
FeSe
2
, hexagonal Cu
1.8
S, and hexagonal CuSe) were recovered, as observed by XRD,
which the authors suggest is either due to partial oxidation as the inks were prepared in
air, or due to partial evaporation of the metal during the annealing step. Counter
electrodes were made by drop casting the inks on FTO/glass substrates followed by
annealing to yield ≥25 µm thick microstructured films (Figure 1.7 a–c) with high specific
surface areas (22.5–29.3 m
2
g
−1
). Cyclic voltammetry (CV), electrochemical impedance
spectroscopy (EIS), and Tafel polarization studies were used to investigate the electrodes'
ability to facilitate the reduction of either triiodide (for FeSe
2
) or polysulfide (for Cu
1.8
S
and CuSe). CV (Figure 1.7 f) and EIS data showed that the FeSe
2
electrode had higher
activity than Pt toward reduction of triiodide with higher peak currents, lower charge
transfer resistance at the electrode/electrolyte interface, and a larger diffusion coefficient.
Accordingly, dye-sensitized solar cells (DSCs) utilizing the FeSe
2
CEs had higher
efficiencies (avg. 9.0%) than those with Pt CEs (avg. 8.4%). To investigate the catalytic
29
properties of the CuSe and Cu
1.8
S CEs compared to Pt, CdSe-quantum dot-sensitized
solar cells (QDSCs) with polysulfide electrolyte were prepared using CuSe, Cu
1.8
S, and
Pt CEs. EIS measurements on the devices indicated that those having CuSe and Cu
1.8
S
CEs generate much lower charge transfer resistances for reduction of polysulfide
compared to Pt. Additionally, as seen in the CV plot in Figure 1.7 g, current densities in
the reduction zone were greatest for the CuSe CEs (i.e., 8 to 10 times greater than for Pt),
indicating the impressive catalytic activity of the recovered CuSe for the reduction of
polysulfide. Reproducibility and stability of the CEs were verified by observation of
comparable results when comparing different batches (including recycled CE materials)
and over multiple cycles.
Figure 1.7. (a–d) SEM images of recovered (a) FeSe
2
, (b) Cu
1.8
S, (c) CuSe, and (d) CoSe
2
materials, showing nanostructured morphologies; (e) photographs of precursor inks; (f) CV
curves for FeSe
2
and Pt electrodes where Ox’ and Red’ correspond to the iodide/triiodide redox
species; (g) CV curves for Cu
1.8
S, CuSe, and Pt electrodes in a polysulfide electrolyte; (h)
polarization scans of CoSe
2
thin films on HOPG substrates showing better surface utilization for
thinner films.
55, 56
Broadening the scope of solution processing using the alkahest method has not only
30
lead to new solution-processable materials, but it has also inspired exploration into new
applications. For instance, in 2016 we used Co(OH)
2
and gray Se to prepare a precursor
ink for the deposition of marcasite-type CoSe
2
that could be used as a hydrogen evolution
reaction (HER) electrocatalyst.
56
Figure 1.7 d shows the nanostructured morphology of
the recovered phase-pure CoSe
2
material. We showed that multiple coats of the ink
(annealed to 350 °C between coats) could be used to control the thickness of CoSe
2
on
highly ordered pyrolytic graphite (HOPG) substrates. Electrocatalytic HER activity
studies were performed in 0.5 M aqueous H
2
SO
4
solutions for films of differing
thicknesses. The Faradaic efficiency for H
2
production for all films was >99%, and an 8 h
stability test showed that the films were indeed robust. Polarization curves (Figure 1.7 h,
as-measured traces are dotted) revealed that although all of the films had the same onset
potential as determined from Tafel plots (−117 mV vs. RHE), the overpotential required
to reach 10 mA cm
2
(η
10 mA/cm2
; e.g., −272 mV vs. RHE for a 391 nm thick film)
increased with film thickness as a result of mass transport-dominated HER at higher
overpotentials. Electrochemically active surface areas for the films (determined by CV
double-layer capacitance measurements) were used to normalize the polarization curves
and η
10 mA cm
−2
values (Figure 1.7 h, normalized traces are bolded), which revealed that
thicker films had poor surface utilization due to diffusion-limited current, and thus,
thinner films were most efficient. The practicality of the alkahest molecular ink method is
demonstrated in this work as it facilitates the straightforward tuning of film properties
(such as film thickness) that can be optimized to most effectively utilize a catalyst.
31
1.6. Alkahest-processed Flexible Substrates for Photodetectors
Flexible polyimide substrates were used in Hasan et al.'s recent work to construct
flexible photodetectors based on alkahest-processed Sb
2
Se
3
nanowire thin films.
57
While
previous reports on Sb
2
Se
3
nanowire photodetectors required separate steps for the
nanowire synthesis and deposition (separated by several work up steps), the morphology
of spin-cast Sb
2
Se
3
thin films could be tuned from nanowires to nano-grains by simply
changing the ratio of elemental Sb and Se in a mercaptoethanol/en ink.
58,59
A Sb/Se ratio
of 1 : 3 yielded 0.1–1.2 µm long nanowires with 100–300 nm diameters, while a lower
ratio (1 : 1.6) resulted in nano-grained morphology. The recovered nanowires were
single-crystalline orthorhombic Sb
2
Se
3
as observed by phase-contrast high-resolution
TEM and by selected-area electron diffraction, with further corroboration by grazing
incidence X-ray diffraction of the thin films. Elemental analysis and mapping indicated
that Sb and Se were uniformly distributed throughout the nanowires and were present in
the expected stoichiometric ratio for Sb
2
Se
3
. XPS and Raman spectroscopy data showed
the presence of Sb
2
O
3
in the films, which the authors attributed to storing the films in air.
UV-vis absorption spectroscopy was used to generate a Tauc plot from which the direct
band gap of the Sb
2
Se
3
films was extrapolated as 1.12 eV. Flexible photodetectors were
constructed by spin-coating the Sb
2
Se
3
precursor ink onto polyimide substrates and
annealing to 350 °C for 5 min. The flexible photodetectors showed response (24 ms) and
recovery (9 ms) times that were two orders of magnitude smaller than previously reported
Sb
2
Se
3
-nanowire photodetectors. Optical simulation results suggest that tuning the size of
32
the nanowires (by adjusting the precursor ratio or annealing conditions), could result in
optimization of broadband performance, or alternatively that spectrally selective
detection may be achievable. The mechanical stability of the photodetectors was
confirmed with no significant change in photoresponse observed after 40 bending cycles
over a radius of 10 mm. This report showcases a strong example of how the alkahest
method can be paired with versatile flexible substrates to fabricate highly functional thin
films with morphology control for optoelectronic applications.
1.7. Alkahest-processed Thermoelectric Materials
Many binary metal chalcogenides (e.g., Cu
2−x
E, Bi
2
E
3
, PbE, SnE, where E = S, Se, or
Te), as well as alloys thereof, are excellent thermoelectric materials.
60
Although several
of these materials were prepared using the alkahest method in early reports (e.g., PbTe,
Bi
2
Te
3
, SnS, etc., vide supra), it was not until recently that Ma et al. reported measuring
the thermoelectric properties of an alkahest-prepared material.
61
They investigated
Cu
2−x
Se
y
S
1−y
and Ag-doped Cu
2−x
Se
y
S
1−y
thin films (ca. 60–90 nm thick) prepared by spin
coating inks consisting of Cu
2
Se, Cu
2
S, and Ag
2
O as precursors in EDT/en mixtures,
followed by annealing for at least 30 min to various temperatures above 300 °C. Thermal
recovery of Cu
2
Se inks resulted in stoichiometries ranging from Cu
1.78
Se
0.69
S
0.31
to
Cu
1.94
Se
0.65
S
0.35
with increasing annealing temperature from 310 to 390 °C. The change in
composition was accompanied by an increase in Seebeck coefficient from 26 to 34 µV
K
−1
and decrease in electrical conductivity from 1380 to 890 Ω
−1
cm
−1
, while the thermal
conductivity remained constant at ca. 0.6 W m
−1
K
−1
. These systematic changes with
33
temperature were attributed a decrease in Cu vacancies (which leave holes), and thus a
decrease in electrical conductivity and corresponding increase in Seebeck coefficient.
Utilizing Cu
2
S or a mixture of Cu
2
Se + Cu
2
S as precursors allowed for the recovery of
Cu
2−x
Se
y
S
1−y
alloys with a constant near-zero value of x and tunable value of y. XRD
analysis of recovered bulk powder samples revealed that inks containing Cu
2
Se as the
only precursor gave a mixture of monoclinic and cubic Cu
2−x
Se
y
S
1−y
, while Cu
2
S inks
gave tetragonal Cu
2−x
S and inks utilizing mixtures of Cu
2
S + Cu
2
Se as precursors yielded
a mixture of Cu
2−x
S and Cu
2−x
Se phases with signs of alloying. With decreasing Se/S
ratio, the average Seebeck coefficient increased from 29 to 83 µV K
−1
and the average
electrical conductivity decreased from 1163 to 163 Ω
−1
cm
−1
. Ag-doped
Cu
2−x
Se
y
S
1−y
films prepared at 350 °C had an average Seebeck coefficient of 52 µV
K
−1
(80% increase over non-doped samples) despite the larger number of Cu vacancies
and higher Se/S ratio, both shown to decrease the Seebeck coefficient in the case of the
undoped samples. The Ag-doped films exhibited thermal conductivities 30% lower than
the undoped films. This work nicely presents an example of how different approaches can
be facilitated (utilizing different precursors or doping) within the alkahest solution
processing method to engineer electronic and thermal conductivities as well as overall
thermoelectric properties of a given material.
1.8. Alkahest-processed Ligands for Colloidal Nanocrystals
In addition to solution processing of functional chalcogenide thin films, the chemistry
of the molecular inks themselves can be an interesting platform for materials chemistry.
34
An example of this can be seen in Talapin's work using metal chalcogenide complexes
(MCCs), prepared by dissolution of bulk chalcogenides with excess chalcogen in
hydrazine, as strongly binding ligands for semiconductor nanocrystals. A simple phase-
transfer ligand exchange can be used to facilitate the exchange of native long-chain
organic ligands for MCCs on various colloidal nanocrystal platforms. Upon mild thermal
annealing, the MCC ligands are converted into amorphous or crystalline chalcogenide
intermingled with the nanocrystals. Since there is no source of nonvolatile carbon
species, this method allows for clean deposition of all-inorganic nanocrystalline thin
films within an electronic-grade chalcogenide network. Additionally, replacing the bulky
organic native ligands allows for closer nanocrystal spatial coupling and enhanced
electronic coupling between adjacent particles. For example, when dodecanethiol ligands
were replaced with Sn
2
S
6
4−
MCC ligands, the conductivity of corresponding Au
nanocrystal solids was increased by ca. 11 orders of magnitude.
62
This method has been
successfully used to cap many compositions and structures of nanocrystals with Sn
2
S
6
4−
ligands (e.g., CdSe, CdTe, CdS, Bi
2
S
3
, Au, Pd) and also works to cap CdSe nanocrystals
with various MCC ligands (e.g., Sn
2
Se
6
4−
, In
2
Se
4
2−
, In
2
Te
3
, Ga
2
Se
3
, CuInSe
3
, ZnTe,
HgSe
2
2−
, Sb
2
Se
3
). Unfortunately, as discussed previously, working with hydrazine
presents a safety and practicality issue that may make this method prohibitive to scale.
Drawing on the similarities we found between the hydrazine dissolution method and
the alkahest method regarding their wide ranging solvent power, we investigated the
nature of dissolved Sb
2
S
3
in mercaptoethanol/en as a model system for phase-transfer
35
ligand exchange to install the dissolved Sb
2
S
3
species onto CdSe nanocrystals.
63
By
combining several analytical techniques (inductively coupled plasma atomic emission
spectroscopy (ICP-AES), TGA, electrospray ionization mass spectroscopy (ESI-MS),
Raman and FT-IR spectroscopies), we were able to deduce the structure of the dissolved
species as a mixture of molecular stibanates containing either one or two antimony atoms
coordinated by deprotonated mercaptoethanol and counterbalanced in charge by
protonated en ligands. Ligand exchange of native stearate-bound CdSe nanocrystals was
complete within 5 min of stirring with a solution of molecular stibanates. Zeta potential
measurements indicated that the negatively charged surface ligands provided electrostatic
colloidal stability for extended periods (>1 month). Additionally, as shown in Figure 1.8
a and b, stibanate ligands could be installed on CdS/CdSe core/shell, and Pt nanocrystals
to yield colloidally stable suspensions, and other chalcogenide semiconductors (As
2
S
3
,
As
2
Se
3
, Sb
2
Se
3
, SnS, and ZnS) were shown to work in a similar fashion for facile ligand
exchange of CdSe nanocrystals. As shown in Figure 1.8 c, there is no blue shift in exciton
peak or scattering observable in the solution absorption spectra for the ligand-exchanged
nanocrystals, indicating that the nanocrystals were not etched during the ligand exchange
process and were indeed colloidally stable. Thin films of as-prepared and stibanate-
capped CdSe nanocrystals were prepared by spin coating colloidal suspensions onto
ITO/glass substrates. Figure 1.8 d shows that the stibanate-capped nanocrystal films gave
much higher photocurrent densities (ca. −45 µA cm
−2
) than the as-prepared nanocrystal
films (<2 µA cm
−2
) as a result of replacing the insulating native ligands. Additionally, the
stibanate-capped nanocrystal films showed signs of enhanced interparticle coupling as
36
observed by a broadened and red-shifted exciton peak in the absorption spectrum without
peak sharpening in the XRD patterns.
Figure 1.8. (a) Ligand exchange of CdSe nanocrystals using various dissolved bulk inorganic
materials; (b) ligand exchange of various types of nanocrystals using stibanate ligands from
dissolved bulk Sb
2
S
3
; (c) solution absorption spectra of CdSe nanocrystals before and after ligand
exchange with various dissolved bulk semiconductors; (d) chopped photocurrent response of as-
prepared (heated to 150 ˚C) and ligand-exchanged (heated to 300 ˚C) CdSe films.
63
1.9. Elucidation of Solvated Species and Dissolution Mechanisms
The majority of research involving alkahest chemistry has been aimed at the design
and recovery of useful chalcogenide materials in the form of thin films for relevant
applications. Perhaps due the straightforward dissolution/recovery methodology, the
specific chemistry of the dissolution process has not received as much attention. Granted,
such studies are quite complex and require thoughtful consideration of several nuanced
characteristics of the system; for example, the dissolution process is air-sensitive for
some solutes, and as we showed in the case of Sb
2
S
3
dissolution, there can be several
different species present in solution, possibly in dynamic equilibrium. Additionally,
37
ethylenediamine and most of the thiols used to formulate the solvent mixtures are
classified as toxic and corrosive, therefore care needs to be taken when working with the
solvent mixtures. Importantly though, in contrast to hydrazine ethylenediamine is not a
carcinogen and is far less toxic. It should also be noted that the thiol/amine solvent
mixtures are inherently malodorous as the short chain alkane thiol components produce
strong sulfurous odors. Thus, investigation of the dissolution process for bulk precursors
in this system by direct characterization of dissolved species is indeed a difficult task.
However, several groups have put efforts toward understanding the mechanisms at play.
In the following section we will summarize these studies (some of which are included in
the application articles discussed above) to illustrate what is known about the system, and
draw some connecting lines between the findings.
In our first report on the alkahest system, we studied the crystal structure of the binary
EDT/en solvent mixture and found it to consist of the addition product of EDT and en
with a stoichiometry of (enH
+
)
2
(EDT
2−
)(en), where the crystal lattice is stabilized by
extensive N–H⋯S hydrogen bonding.
28
The ability for the system to hydrogen bond is
evidently an important factor in its solvent power toward chalcogenides, as replacing en
with a bidentate tertiary amine of similar basicity (N,N,N′,N′-tetramethylethylenediamine)
results in negligible solvent power toward the V
2
VI
3
chalcogenides. The ionic nature of
the solvent system was confirmed by measuring an increase (ca. 15000×) in electrolytic
conductivity upon the addition of EDT to en, while further evidence from
1
H NMR and
Raman spectroscopy indicate at least partial deprotonation of the thiol.
29
Additionally, the
38
strong solvent power of the EDT/en mixture seems to be associated with the 1,2-chelating
ability of both the thiol and amine as binary mixtures containing a thiol/mono-amine or
1,3-chelating thiol had much poorer solvent power.
Regarding the dissolution of the V
2
VI
3
materials, we observed that none were soluble
in the absence of EDT (with the exception of As
2
S
3
and As
2
Se
3
, which were previously
known to dissolve in primary amines).
64
The dissolved species appeared to be molecular
rather than nanoparticulate, as the solutions were free of visible scattering and dynamic
light scattering did not give evidence of any clusters greater than 1 nm. At this point, our
working hypothesis for the dissolved chalcogenide species was an en-ligated
thiolatochalcogenometallate anion counterbalanced by enH
+
. The ionic nature of this
proposed species was supported by electrolytic conductivity data, where conductivity
increased by a factor of ca. 1.4–1.6 upon V
2
VI
3
dissolution relative to the solvent system
on its own.
In our study of the dissolution of bulk Se and Te, we found that although the
solubility limit of Se in the thiol/en mixture increases when moving from ethanethiol to
mercaptoethanol to EDT, that of Te drops in the inconclusive order of ethanethiol to EDT
to mercaptoethanol.
29
1
H NMR and Raman spectroscopy indicated that the Se ink was
comprised of a heterogeneous mixture of Se rings/chains and species containing Se–S
bonds created by a nucleophilic thiolate, whereupon addition of Se, the thiol S–H Raman
band disappears. Additionally, we found that the presence of dissolved Se or Te greatly
increases the solubility of elemental Sb and Sn, respectively, presumably due to increased
39
nucleophile concentration facilitated by the dissolution of the chalcogen, similar to the
mechanism functioning in the hydrazine method where excess chalcogen is reduced in
situ to yield an effective nucleophile.
In our investigation of molecular stibanates as ligands for CdSe nanocrystals (vide
supra), we first studied the nature of the dissolved chalcogenide species in the ink.
Negative-ion ESI-MS data indicated that in a dilute Sb
2
S
3
mercaptoethanol/en solution
there are four primary stibanate species having mercaptoethanol ligands and either one or
two Sb atoms. ICP-AES and combustion elemental analysis data indicated that en was
also associated with the dissolved species, as excess N and C were observed in a dried
sample of the ink. Further information about the dissolved stibanates gathered from FT-
IR and Raman spectroscopies was compared with literature reports on relevant model
species in order to support the following proposed stibanate structures: [S
Sb(SC
2
H
4
O)]
−
, [(SC
2
H
4
O)Sb(SC
2
H
4
O)]
−
, [(HOC
2
H
4
S)Sb(µ-S)
2
SbS]
−
, and
[(HOC
2
H
4
S)Sb(µ-S)
2
Sb(SC
2
H
4
O)]
−
where each molecular stibanate is counterbalanced
by enH
+
.
Further motivation to understand the complexities of the alkahest chemistry lead to
our investigation into the dissolution of Sn, SnO, and SnS, which allowed us to utilize
solution
119
Sn NMR toward identifying the dissolved species.
65
Added value for studying
these systems was recognized in the recovery of phase-pure SnS after mild thermal
annealing of all three precursor inks, which, as discussed above, is a narrow band gap
semiconductor candidate for single junction PV devices. EDT/en solutions of each
40
precursor were prepared by room temperature dissolution over 1–5 days. All three
precursors gave light yellow, optically transparent solutions with identical
119
Sn NMR
signals at δ
119
Sn = 217 ppm and no J
Sn–Sn
coupling (Figure 1.9 a), suggesting that all
precursors give the same four-coordinate mononuclear species upon dissolution; this is
supported by near identical TGA/DSC ink decomposition traces (Figure 1.9 b).
Comparing literature reports and the observed δ
119
Sn, our initial hypothesis for the nature
of the dissolved species was a complex akin to (EDTate)
2
Sn(IV). However,
119
Sn NMR
experiments indicated that the (EDTate)
2
Sn(IV) complex (as synthesized by literature
methods) is easily chelated by EDT
2−
or en to yield six-coordinate species. Specifically,
the
119
Sn NMR spectrum we obtained of (EDTate)
2
Sn(IV) in the EDT/en solvent system
gave a broad resonance at δ
119
Sn = −346 ppm indicative of a six-coordinate species. To
resolve the oxidation state of Sn in the dissolved species we studied the
119
Sn NMR
spectra of EDT/en solutions containing Sn
2+
and Sn
4+
sources, SnCl
2
·2H
2
O (singlet
at δ
119
Sn = 210 ppm) and SnCl
4
·5H
2
O (singlet at δ
119
Sn = −347 ppm), respectively. As
the signal of the Sn
2+
control sample had nearly the same chemical shift as the dissolved
Sn, SnO, and SnS, we concluded that the solvated complex contains only Sn
2+
. This was
further supported by the observation of a single resonance at δ
119
Sn = 216 ppm in the
spectrum of a solution containing a 1 : 1 wt/wt mixture of SnCl
2
·2H
2
O and SnO in
EDT/en. Lack of signals associated with ν(Sn–N) in the Raman spectra of the three tin
precursor inks indicated that en does not chelate to the Sn species. Therefore, the resting
hypothesis is that Sn, SnO, and SnS dissolved in EDT/en yield a four-coordinate
Sn
2+
species with en acting as a noncoordinating solvent and likely countercation. This
41
was corroborated by DFT calculations modeling bis(1,2-ethanedithiolate)tin(II) in the gas
phase that revealed a δ
119
Sn of 308 ppm, which is within the experimental range for four-
coordinate Sn–thiolate complexes.
Figure 1.9. (a)
119
Sn NMR spectra of dissolved SnO, SnS, and Sn, showing the same single peak
at 217 ppm; (b) TGA/DSC data indicating nearly identical decomposition traces for each of the
three tin-based inks.
65
A noteworthy characteristic of the alkahest method is the redox chemistry that allows
such a range of precursors to be utilized toward recovering a common chalcogenide
material. For example, in the EDT/en solvent system, Sn
0
is oxidized to the stable
oxidation state of Sn
2+
to yield the same solute as is given by dissolution of the SnO and
SnS precursors, such that all three inks yield the same phase-pure SnS upon annealing.
42
Other elemental precursors (e.g., Cu, Zn, In, Ga, Mg, Fe, Co, Ni, Mn and Sb, vide supra)
must also go through a similar oxidation process involving the binary solvent system, in
some cases requiring the presence of elemental chalcogen, to achieve appreciable
dissolution.
29,35,37
In the other direction, alkahest chemistry also facilitates the reduction
of elemental Se and Te in the case where a countercation precursor is present so that
chalcogenide materials may be recovered.
29
Walker and Agrawal studied the dissolution
of elemental Se in a mixture of ethanethiol and oleylamine, and using gas
chromatography-mass spectrometry (GC-MS) found diethyldisulfide and elemental
Se
1
and Se
8
in solution, leading them to propose the mechanism shown in Scheme 1.2 for
the dissolution of Se in a thiol/amine mixture.
66
Scheme 1.2. Mechanism proposed for dissolution of Se in a thiol/amine mixture.
66
As the major motivation in the pursuit of the alkahest system was to find a system
with similar wide-ranging applicability and low-temperature recovery properties as are
offered by the hydrazine method, it is useful to take stock of how the two methods
compare thus far. More than 65 bulk precursors have already been reported to be soluble
in a thiol/amine mixture, including: Mg,
35
Mn,
35
MnO,
33
Fe,
35
FeSe,
55
Fe(acac)
2
,
33
Co,
43
35
Co(OH)
2
,
56
Ni,
35
Ni(acac)
2
,
33
Cu,
35,37,38
CuO,
33
Cu
2
O,
31,33,34,36
CuS,
38
Cu
2
S,
32,38,40,41,53,67
Cu
1.9
S,
55
CuSe,
38
Cu
2
Se,
38,39,55,67
CuCl
2
,
33,36
CuCl,
36,39,42
Cu(Ac)
2
,
33
Cu(acac)
2
,
33,36,39
Cu(
OAc)
2
,
36,39
Ag
2
O,
31,33
Ag
2
S,
38
Ag
2
Se,
38
Zn,
35,36,38
ZnO,
31,33,34,36
ZnS,
63
ZnCl
2
,
36
Zn(OAc)
2
,
3
6
CdO,
31
Cd(OH)
2
,
33
CdCl
2
,
42
In,
35,37,38
In
2
S
3
,
38,40,41,67
In
2
Se
3
,
38,40,41,53,67
InCl
3
,
39
In(OH)
3
,
33
In(acac)
3
,
39
In(OAc)
3
,
39
Ga,
35,37,41,53
GaCl
3
,
39
GaI
3
,
39
Ga(acac)
3
,
33,39
PbO,
31
PbS, Sn,
35,38,6
5
SnO,
31,33,34,65
SnS,
30,38,63,65
SnSe,
38
SnCl
2
,
36
Sn(acac)
2
Cl
2
,
36
GeO
2
,
31
As
2
O
3
,
31
As
2
S
3
,
28,63
As
2
Se
3
,
28,63
As
2
Te
3
,
28
Sb,
29,57
Sb
2
O
3
,
31,33
Sb
2
S
3
,
28,32,63
Sb
2
Se
3
,
28,63
Sb
2
Te
3
,
28
Bi
2
O
3
,
31,33
Bi
2
S
3
,
28
Bi
2
Se
3
,
28
Bi
2
Te
3
,
28
S,
38
Se,
29,35,37,41,56,57
Te.
29,42
This list clearly shows how numerous
and wide-ranging the scope of viable precursors is for this method, though it is most
definitely not all-inclusive as ongoing research continues to grow the pool of suitable
soluble materials. Notably, several of the materials in this pool have been reported as
insoluble in hydrazine, e.g. CdS, In
2
S
3
, and ZnS.
23
The procedure for dissolution using
the alkahest system (i.e., dissolution on the order of minutes to hours in most cases) is in
general, less cumbersome than what is required for the hydrazine method (i.e., dissolution
up to days or longer). The annealing conditions required to recover phase-pure thin film
materials for devices are fairly similar for both methods with crystalline materials
recoverable at 300–350 °C and higher temperatures used to obtain high quality films. For
example, in the case of alkahest-processed CZTSSe devices (7–8% PCE) the absorber
layers were annealed to 250–320 °C for 1–5 min between coats of ink, with final
selenization steps at 500–550 °C for 10–50 min (Table 1.1) and the best devices using
hydrazine-processed CZTSSe or CIGS absorber layers also used a 5 min low-temperature
annealing step (280–340 °C) in between layers, and final annealing temperatures of 500–
44
600 °C, though they do not require a chalcogen atmosphere (e.g., selenization).
25,49
1.10. Conclusions
The necessary future work involving the alkahest chemistry will focus on studying
the dissolution mechanism and resulting solvated species for various precursors aiming to
gain a better understanding of the fundamental chemistry that is at play. We need to
understand how this remarkable binary thiol/amine solvent mixture works to dissolve
such an extensive library of bulk material precursors to yield phase-pure chalcogenides
upon solution deposition and relatively mild thermal annealing. Once we have a handle
on the chemistry of the system, better-focused work can be performed to solution process
high quality, technologically relevant materials that may not yet by achievable by other
solution processing methods over a range of applications. Additionally, with a strong
understanding of how this method works, we can conceivably begin working on
synthesizing new compositionally complex chalcogenides, as a practical alternative to the
high-energy techniques that are most often used in investigating new materials.
1.11. References
(1) Mitzi, D. B. Solution Processing of Inorganic Materials; John Wiley & Sons:
Hoboken, New Jersey, 2008.
(2) Trends 2016 In Photovoltaic Applications; International Energy Agency, 2016.
(3) “Best Research-Cell Efficiencies” Chart; National Renewable Energy Laboratory -
National Center for Photovoltaics, 2016.
45
(4) Green, M. A.; Emery, K.; Hishikawa, Y.; Warta, W.; Dunlop, E. D.; Levi, D. H.;
Ho-Baillie, A. W. Y. Solar Cell Efficiency Tables (Version 49). Prog. Photovolt. Res.
Appl. 2017, 25, 3–13.
(5) Jackson, P.; Wuerz, R.; Hariskos, D.; Lotter, E.; Witte, W.; Powalla, M. Effects of
Heavy Alkali Elements in Cu(In,Ga)Se
2
Solar Cells with Efficiencies up to 22.6%. Phys.
Status Solidi RRL 2016, 10, 583–586.
(6) Sun, Y.; Rogers, J. A. Inorganic Semiconductors for Flexible Electronics. Adv.
Mater. 2007, 19, 1897–1916.
(7) Global Flexible Electronics Market 2016-2020; Technavio, 2016.
(8) Global Thin-Film Semiconductor Deposition Market 2016-2020; Technavio, 2016.
(9) A. V. Powell and P. Vaqueiro, in Thermoelectric Materials and Devices, eds. I.
Nandhakumar, N. M. White and S. Beeby, The Royal Society of Chemistry, Cambridge,
UK, 2016, ch. 2, pp. 27–59.
(10) Kong, D.; Cha, J. J.; Wang, H.; Lee, H. R.; Cui, Y. First-Row Transition Metal
Dichalcogenide Catalysts for Hydrogen Evolution Reaction. Energy Environ. Sci. 2013,
6, 3553–3558.
(11) Mitzi, D. B.; Kosbar, L. L.; Murray, C. E.; Copel, M.; Afzali, A. High-Mobility
Ultrathin Semiconducting Films Prepared by Spin Coating. Nature 2004, 428, 299–303.
(12) Mitzi, D. B.; Copel, M.; Chey, S. J. Low-Voltage Transistor Employing a High-
Mobility Spin-Coated Chalcogenide Semiconductor. Adv. Mater. 2005, 17, 1285–1289.
(13) Mitzi, D. B. Polymorphic One-Dimensional (N
2
H
4
)
2
ZnTe: Soluble Precursors for
the Formation of Hexagonal or Cubic Zinc Telluride. Inorg. Chem. 2005, 44, 7078–7086.
(14) Mitzi, D. B. Synthesis, Structure, and Thermal Properties of Soluble Hydrazinium
Germanium(IV) and Tin(IV) Selenide Salts. Inorg. Chem. 2005, 44, 3755–3761.
(15) Mitzi, D. B.; Copel, M.; Murray, C. E. High-Mobility P-Type Transistor Based on
a Spin-Coated Metal Telluride Semiconductor. Adv. Mater. 2006, 18, 2448–2452.
(16) Mitzi, D. B. N
4
H
9
Cu
7
S
4
: A Hydrazinium-Based Salt with a Layered Cu
7
S
4
-
Framework. Inorg. Chem. 2007, 46, 926–931.
(17) Milliron, D. J.; Mitzi, D. B.; Copel, M.; Murray, C. E. Solution-Processed Metal
Chalcogenide Films for P-Type Transistors. Chem. Mater. 2006, 18, 587–590.
46
(18) Mitzi, D. B.; Raoux, S.; Schrott, A. G.; Copel, M.; Kellock, A.; Jordan-Sweet, J.
Solution-Based Processing of the Phase-Change Material KSb5S8. Chem. Mater. 2006,
18, 6278–6282.
(19) Milliron, D. J.; Raoux, S.; Shelby, R. M.; Jordan-Sweet, J. Solution-Phase
Deposition and Nanopatterning of GeSbSe Phase-Change Materials. Nat. Mater. 2007, 6,
352–356.
(20) Kovalenko, M. V.; Spokoyny, B.; Lee, J.; Scheele, M.; Weber, A.; Perera, S.;
Landry, D.; Talapin, D. V. Semiconductor Nanocrystals Functionalized with Antimony
Telluride Zintl Ions for Nanostructured Thermoelectrics. J. Am. Chem. Soc. 2010, 132,
6686–6695.
(21) Feser, J. P.; Chan, E. M.; Majumdar, A.; Segalman, R. A.; Urban, J. J. Ultralow
Thermal Conductivity in Polycrystalline CdSe Thin Films with Controlled Grain Size.
Nano Lett. 2013, 13, 2122–2127.
(22) Wang, R. Y.; Feser, J. P.; Gu, X.; Yu, K. M.; Segalman, R. A.; Majumdar, A.;
Milliron, D. J.; Urban, J. J. Universal and Solution-Processable Precursor to Bismuth
Chalcogenide Thermoelectrics. Chem. Mater. 2010, 22, 1943–1945.
(23) Mitzi, D. B. Solution Processing of Chalcogenide Semiconductors via
Dimensional Reduction. Adv. Mater. 2009, 21, 3141–3158.
(24) Mitzi, D. B.; Yuan, M.; Liu, W.; Kellock, A. J.; Chey, S. J.; Deline, V.; Schrott,
A. G. A High-Efficiency Solution-Deposited Thin-Film Photovoltaic Device. Adv. Mater.
2008, 20, 3657–3662.
(25) Zhang, T.; Yang, Y.; Liu, D.; Tse, S. C.; Cao, W.; Feng, Z.; Chen, S.; Qian, L.
High Efficiency Solution-Processed Thin-Film Cu(In,Ga)(Se,S)
2
Solar Cells. Energy
Environ. Sci. 2016, 9, 3674–3681.
(26) Vineyard, B. D. Versatility and the Mechanism of the N-Butyl-Amine-Catalyzed
Reaction of Thiols with Sulfur. J. Org. Chem. 1967, 32, 3833–3836.
(27) Liu, Y.; Yao, D.; Shen, L.; Zhang, H.; Zhang, X.; Yang, B. Alkylthiol-Enabled Se
Powder Dissolution in Oleylamine at Room Temperature for the Phosphine-Free
Synthesis of Copper-Based Quaternary Selenide Nanocrystals. J. Am. Chem. Soc. 2012,
134, 7207–7210.
(28) Webber, D. H.; Brutchey, R. L. Alkahest for V
2
VI
3
Chalcogenides: Dissolution of
Nine Bulk Semiconductors in a Diamine-Dithiol Solvent Mixture. J. Am. Chem. Soc.
2013, 135, 15722–15725.
47
(29) Webber, D. H.; Buckley, J. J.; Antunez, P. D.; Brutchey, R. L. Facile Dissolution
of Selenium and Tellurium in a Thiol–amine Solvent Mixture under Ambient Conditions.
Chem. Sci. 2014, 5, 2498–2502.
(30) Antunez, P. D.; Torelli, D. A.; Yang, F.; Rabuffetti, F. A.; Lewis, N. S.; Brutchey,
R. L. Low Temperature Solution-Phase Deposition of SnS Thin Films. Chem. Mater.
2014, 26, 5444–5446.
(31) McCarthy, C. L.; Webber, D. H.; Schueller, E. C.; Brutchey, R. L. Solution-Phase
Conversion of Bulk Metal Oxides to Metal Chalcogenides Using a Simple Thiol–Amine
Solvent Mixture. Angew. Chem. Int. Ed. 2015, 54, 8378–8381.
(32) McCarthy, C. L.; Cottingham, P.; Abuyen, K.; Schueller, E. C.; Culver, S. P.;
Brutchey, R. L. Earth Abundant CuSbS
2
Thin Films Solution Processed from Thiol–
amine Mixtures. J. Mater. Chem. C 2016, 4, 6230–6233.
(33) Tian, Q.; Wang, G.; Zhao, W.; Chen, Y.; Yang, Y.; Huang, L.; Pan, D. Versatile
and Low-Toxic Solution Approach to Binary, Ternary, and Quaternary Metal Sulfide
Thin Films and Its Application in Cu
2
ZnSn(S,Se)
4
Solar Cells. Chem. Mater. 2014, 26,
3098–3103.
(34) Tian, Q.; Cui, Y.; Wang, G.; Pan, D. A Robust and Low-Cost Strategy to Prepare
Cu
2
ZnSnS
4
Precursor Solution and Its Application in Cu2ZnSn(S,Se)
4
Solar Cells. RSC
Adv. 2015, 5, 4184–4190.
(35) Yang, Y.; Wang, G.; Zhao, W.; Tian, Q.; Huang, L.; Pan, D. Solution-Processed
Highly Efficient Cu
2
ZnSnSe
4
Thin Film Solar Cells by Dissolution of Elemental Cu, Zn,
Sn, and Se Powders. ACS Appl. Mater. Interfaces 2015, 7, 460–464.
(36) Zhang, R.; Szczepaniak, S. M.; Carter, N. J.; Handwerker, C. A.; Agrawal, R. A
Versatile Solution Route to Efficient Cu
2
ZnSn(S,Se)
4
Thin Film Solar Cells. Chem.
Mater. 2015, 27, 2114–2120.
(37) Zhao, D.; Tian, Q.; Zhou, Z.; Wang, G.; Meng, Y.; Kou, D.; Zhou, W.; Pan, D.;
Wu, S. Solution-Deposited Pure Selenide CIGSe Solar Cells from Elemental Cu, In, Ga,
and Se. J. Mater. Chem. A 2015, 3, 19263–19267.
(38) Zhang, R.; Cho, S.; Lim, D. G.; Hu, X.; Stach, E. A.; Handwerker, C. A.;
Agrawal, R. Metal–metal Chalcogenide Molecular Precursors to Binary, Ternary, and
Quaternary Metal Chalcogenide Thin Films for Electronic Devices. Chem. Commun.
2016, 52, 5007–5010.
48
(39) Zhao, X.; Lu, M.; Koeper, M. J.; Agrawal, R. Solution-Processed Sulfur Depleted
Cu(In, Ga)Se
2
Solar Cells Synthesized from a Monoamine–dithiol Solvent Mixture. J.
Mater. Chem. A 2016, 4, 7390–7397.
(40) Arnou, P.; van Hest, M. F. A. M.; Cooper, C. S.; Malkov, A. V.; Walls, J. M.;
Bowers, J. W. Hydrazine-Free Solution-Deposited CuIn(S,Se)
2
Solar Cells by Spray
Deposition of Metal Chalcogenides. ACS Appl. Mater. Interfaces 2016, 8, 11893–11897.
(41) Arnou, P.; Cooper, C. S.; Uličná, S.; Abbas, A.; Eeles, A.; Wright, L. D.; Malkov,
A. V.; Walls, J. M.; Bowers, J. W. Solution Processing of CuIn(S,Se)
2
and
Cu(In,Ga)(S,Se)
2
Thin Film Solar Cells Using Metal Chalcogenide Precursors. Thin Solid
Films 2016, DOI: 10.1016/j.tsf.2016.10.011.
(42) Miskin, C. K.; DuboisCamacho, A.; Reese, M. O.; Agrawal, R. A Direct Solution
Deposition Approach to CdTe Thin Films. J. Mater. Chem. C 2016, 4, 9167–9171.
(43) Cooper, C. S.; Arnou, P.; Wright, L. D.; Uličná, S.; Walls, J. M.; Malkov, A. V.;
Bowers, J. W. An Innovative Approach for Fabrication of Cu
2
ZnSnSe
4
Absorber Layers
Using Solutions of Elemental Metal Powders. Thin Solid Films 2016, DOI:
10.1016/j.tsf.2016.12.026.
(44) Tian, Q.; Huang, L.; Zhao, W.; Yang, Y.; Wang, G.; Pan, D. Metal Sulfide
Precursor Aqueous Solutions for Fabrication of Cu
2
ZnSn(S,Se)
4
Thin Film Solar Cells.
Green Chem. 2015, 17, 1269–1275.
(45) Yang, Y.; Kang, X.; Huang, L.; Wei, S.; Pan, D. A General Water-Based
Precursor Solution Approach to Deposit Earth Abundant Cu
2
ZnSn(S,Se)
4
Thin Film
Solar Cells. J. Power Sources 2016, 313, 15–20.
(46) Zhao, W.; Pan, D.; Liu, S. Kesterite Cu
2
Zn(Sn,Ge)(S,Se)
4
Thin Film with
Controlled Ge-Doping for Photovoltaic Application. Nanoscale 2016, 8, 10160–10165.
(47) Yang, Y.; Kang, X.; Huang, L.; Pan, D. Tuning the Band Gap of Cu
2
ZnSn(S,Se)
4
Thin Films via Lithium Alloying. ACS Appl. Mater. Interfaces 2016, 8, 5308–5313.
(48) Scheer, R.; Schock, H. Chalcogenide Photovoltaics: Physics, Technologies, and
Thin Film Devices; Wiley-VHC: Weinheim, Germany, 2011.
(49) Wang, W.; Winkler, M. T.; Gunawan, O.; Gokmen, T.; Todorov, T. K.; Zhu, Y.;
Mitzi, D. B. Device Characteristics of CZTSSe Thin-Film Solar Cells with 12.6%
Efficiency. Adv. Energy Mater. 2014, 4, 1301465.
49
(50) Miskin, C. K.; Yang, W.; Hages, C. J.; Carter, N. J.; Joglekar, C. S.; Stach, E. A.;
Agrawal, R. 9.0% Efficient Cu
2
ZnSn(S,Se)
4
Solar Cells from Selenized Nanoparticle
Inks. Prog. Photovolt. Res. Appl. 2015, 23, 654–659.
(51) Qu, Y.; Zoppi, G.; Beattie, N. S. The Role of Nanoparticle Inks in Determining
the Performance of Solution Processed Cu
2
ZnSn(S,Se)
4
Thin Film Solar Cells. Prog.
Photovolt. 2016, 24, 836–845.
(52) Chen, S.; Walsh, A.; Gong, X.; Wei, S. Classification of Lattice Defects in the
Kesterite Cu
2
ZnSnS
4
and Cu
2
ZnSnSe
4
Earth-Abundant Solar Cell Absorbers. Adv. Mater.
2013, 25, 1522–1539.
(53) Zhao, D.; Fan, Q.; Tian, Q.; Zhou, Z.; Meng, Y.; Kou, D.; Zhou, W.; Wu, S.
Eliminating Fine-Grained Layers in Cu(In,Ga)(S,Se)
2
Thin Films for Solution-Processed
High Efficiency Solar Cells. J. Mater. Chem. A 2016, 4, 13476–13481.
(54) Rose, D. H.; Hasoon, F. S.; Dhere, R. G.; Albin, D. S.; Ribelin, R. M.; Li, X. S.;
Mahathongdy, Y.; Gessert, T. A.; Sheldon, P. Fabrication Procedures and Process
Sensitivities for CdS/CdTe Solar Cells. Prog. Photovolt. Res. Appl. 1999, 7, 331–340.
(55) Liu, F.; Zhu, J.; Hu, L.; Zhang, B.; Yao, J.; Nazeeruddin, M. K.; Grätzel, M.; Dai,
S. Low-Temperature, Solution-Deposited Metal Chalcogenide Films as Highly Efficient
Counter Electrodes for Sensitized Solar Cells. J. Mater. Chem. A 2015, 3, 6315–6323.
(56) McCarthy, C. L.; Downes, C. A.; Schueller, E. C.; Abuyen, K.; Brutchey, R. L.
Method for the Solution Deposition of Phase-Pure CoSe
2
as an Efficient Hydrogen
Evolution Reaction Electrocatalyst. ACS Energy Lett. 2016, 1, 607–611.
(57) Hasan, M. R.; Arinze, E. S.; Singh, A. K.; Oleshko, V. P.; Guo, S.; Rani, A.;
Cheng, Y.; Kalish, I.; Zaghloul, M. E.; Rao, M. V.; et al. An Antimony Selenide
Molecular Ink for Flexible Broadband Photodetectors. Adv. Electron. Mater. 2016, 2,
1600182.
(58) Zhai, T.; Ye, M.; Li, L.; Fang, X.; Liao, M.; Li, Y.; Koide, Y.; Bando, Y.;
Golberg, D. Single-Crystalline Sb
2
Se
3
Nanowires for High-Performance Field Emitters
and Photodetectors. Adv. Mater. 2010, 22, 4530–4533.
(59) Chen, G.; Wang, W.; Wang, C.; Ding, T.; Yang, Q. Controlled Synthesis of
Ultrathin Sb
2
Se
3
Nanowires and Application for Flexible Photodetectors. Adv. Sci. 2015,
2, 1500109.
(60) Ge, Z.; Zhao, L.; Wu, D.; Liu, X.; Zhang, B.; Li, J.; He, J. Low-Cost, Abundant
Binary Sulfides as Promising Thermoelectric Materials. Mater. Today 2016, 19, 227–
239.
50
(61) Ma, Y.; Vartak, P. B.; Nagaraj, P.; Wang, R. Y. Thermoelectric Properties of
Copper Chalcogenide Alloys Deposited via the Solution-Phase Using a Thiol–amine
Solvent Mixture. RSC Adv. 2016, 6, 99905–99913.
(62) Kovalenko, M. V.; Scheele, M.; Talapin, D. V. Colloidal Nanocrystals with
Molecular Metal Chalcogenide Surface Ligands. Science 2009, 324, 1417–1420.
(63) Buckley, J. J.; Greaney, M. J.; Brutchey, R. L. Ligand Exchange of Colloidal
CdSe Nanocrystals with Stibanates Derived from Sb
2
S
3
Dissolved in a Thiol-Amine
Mixture. Chem. Mater. 2014, 26, 6311–6317.
(64) Chern, G. C.; Lauks, I. Spin-coated Amorphous Chalcogenide Films. J. Appl.
Phys. 1982, 53, 6979–6982.
(65) Buckley, J. J.; McCarthy, C. L.; Del Pilar-Albaladejo, J.; Rasul, G.; Brutchey, R.
L. Dissolution of Sn, SnO, and SnS in a Thiol–Amine Solvent Mixture: Insights into the
Identity of the Molecular Solutes for Solution-Processed SnS. Inorg. Chem. 2016, 55,
3175–3180.
(66) Walker, B. C.; Agrawal, R. Contamination-Free Solutions of Selenium in Amines
for Nanoparticle Synthesis. Chem. Commun. 2014, 50, 8331–8334.
(67) Arnou, P.; Cooper, C. S.; Malkov, A. V.; Bowers, J. W.; Walls, J. M. Solution-
Processed CuIn(S,Se)
2
Absorber Layers for Application in Thin Film Solar Cells. Thin
Solid Films 2015, 582, 31–34.
51
Chapter 2. Solution-Phase Conversion of Bulk Metal Oxides to Metal
Chalcogenides Using a Simple Thiol–Amine Solvent Mixture*
*Published in Angew. Chem. Int. Ed. 2015, 54, 8378 –8381.
2.1 Abstract
A thiol–amine solvent mixture is used to dissolve ten inexpensive bulk oxides (Cu
2
O,
ZnO, GeO
2
, As
2
O
3
, Ag
2
O, CdO, SnO, Sb
2
O
3
, PbO, and Bi
2
O
3
) under ambient conditions.
Dissolved oxides can be converted to the corresponding sulfides using the thiol as the
sulfur source, while selenides and tellurides can be accessed upon mixing with a
stoichiometric amount of dissolved selenium or tellurium. The practicality of this method
is illustrated by solution depositing Sb
2
Se
3
thin films from compound inks of dissolved
Sb
2
O
3
and selenium that give high photoelectrochemical current response. The direct band
gap of the resulting material can be tuned from 1.2–1.6 eV by modulating the ink
formulation to give compositionally controlled Sb
2
Se
3−x
S
x
alloys.
2.2 Introduction
Inexpensive solution processing and deposition of inorganic semiconductors is of
great interest for large-scale macroelectronics applications, especially low-temperature
methods that are compatible with a wide range of substrates. Solution processing and
deposition of inorganic semiconductors has important implications in the areas of
photovoltaics,
1
flat panel displays,
2
thermoelectrics,
3
phase change memory,
4
and thin-
52
film transistors;
5
however, direct solution processing of these bulk materials is typically
difficult because of their complete insolubility in “normal” solvents. As an exception to
this statement, the dissolution of bulk chalcogenide semiconductors to form molecular
inks has been most successfully and broadly realized using hydrazine as the solvent,
through a process that was developed by Mitzi and co-workers at IBM.
6
In this process,
molecular solutes are formed by the reaction of bulk metal chalcogenides with
E
2−
(E
2−
=S
2−
, Se
2−
, and Te
2−
), where E
2−
is formed by the in-situ reduction of a
stoichiometric amount of chalcogen with hydrazine.
While the dissolution of bulk chalcogenide semiconductors in hydrazine gives
molecular inks that can be solution processed into extremely high-quality and functional
thin films, the explosive, highly toxic, and carcinogenic properties of hydrazine make it
less attractive for scale. Thus, new solvents or solvent mixtures are needed that 1) can
dissolve a wide scope of bulk materials, 2) are less hazardous, and 3) have sufficient
volatility for solution processing. Toward this goal, we recently reported the remarkable
ability of binary alkanethiol–1,2-ethylenediamine (en) solvent mixtures to dissolve bulk
gray selenium, tellurium, SnS, and a series of nine V
2
VI
3
chalcogenides (where V=As,
Sb, Bi and VI=S, Se, Te) at room temperature and ambient pressure.
7
This chemistry is
inspired by the reaction of sulfur with mixtures of thiols and amines, whereby the thiol is
deprotonated and the resulting thiolate, through a series of nucleophilic attacks, opens
and comminutes the sulfur ring.
8
This new alkahest (or “universal” solvent) for
chalcogenide semiconductors gives stable inks that upon solution deposition and mild
53
thermal annealing regenerate crystalline chalcogenide thin films. Further, this solvent
system is also capable of making inks from the reaction of dissolved gray selenium and
tellurium with elemental metal(loid)s, such as tin and antimony. In this way, crystalline
Sb
2
Se
3
and SnTe films were prepared when elemental antimony or tin were dissolved in
the presence of selenium or tellurium, respectively, and the resulting ink was then
subsequently solution deposited and annealed to 250 °C.
7a
Building off this result, Pan et
al. used a derivative solvent mixture of thioglycolic acid and ethanolamine to dissolve
elemental copper, zinc, and tin together with selenium to form a precursor ink that was
used to generate CZTSe thin films for photovoltaics.
9
While the binary thiol–amine solvent mixture has proven effective for the solution
processing of bulk chalcogenide semiconductors, the bulk metal chalcogenide starting
materials can be expensive, and the elemental metal(loid) starting materials can have
little to no inherent solubility in the absence of dissolved selenium or tellurium.
7a
It is
therefore of interest to explore alternate metal containing starting materials that can
overcome these issues. Along these lines, the equivalent dissolution of cheap, abundant,
and sustainable metal oxide starting materials would be an attractive alternative entrée to
semiconductor inks if the corresponding metal chalcogenides could be produced. Herein,
we demonstrate the extensive capability of this thiol–en solvent mixture to dissolve ten
bulk oxides under ambient conditions. Specific focus is placed on the solution processing
of Sb
2
Se
3
thin films from compound inks of dissolved Sb
2
O
3
and selenium, and the
54
accessibility of crystalline Sb
2
Se
3−x
S
x
solid solutions from these inks, which can be used
as a valuable tool in photovoltaic absorber layer band gap engineering.
2.3 Results and Discussion
Inks of ten metal(loid) oxides (i.e., Cu
2
O, ZnO, GeO
2
, As
2
O
3
, Ag
2
O, CdO, SnO,
Sb
2
O
3
, PbO, and Bi
2
O
3
) were readily prepared by dissolving the bulk oxide powder in a
binary mixture of a thiol (ethanethiol (EtSH), 2-mercaptoethanol (Merc), or 1,2-
ethanedithiol (EDT)) and en in a 1:4 vol/vol ratio at room temperature and ambient
pressure. Solubility limits of the various oxides were determined gravimetrically after 24
h of stirring at room temperature (Table 2.1). Solubilities of these oxides, expressed as
wt % solute in the saturated solution (20–25 °C, 1 atm), range from 5-10 wt % for Cu
2
O,
GeO
2
, and As
2
O
3
, 10–15 wt % for ZnO and SnO, 15–20 wt % for Ag
2
O, CdO, Sb
2
O
3
, and
Bi
2
O
3
, and 25–30 wt % for PbO. The resulting oxide inks are all optically transparent
(Figure 2.1), and remain stable for periods ranging from days to months.
Table 2.1. Oxides solubility limits and thermalized product characterization. Ten metal
oxides were each dissolved in three solvent mixtures of en and one of three thiols (EtSH, Merc, or
EDT) to form molecular inks. The top half of each cell gives the room temperature solubility
limit for each ink and the bottom half of each cell gives the powder XRD characterization of the
material obtained upon thermal annealing to >300 ˚C.
55
Figure 2.1. Photograph of dilute (ca. 5 wt %) solutions of ten bulk oxides in a 1:4 vol/vol
ethanethiol–1,2-ethylenediamine solvent mixture.
Conversion of the metal(loid) oxide inks into the corresponding metal(loid) sulfides
was possible upon mild annealing to 300–375 °C in nearly all cases. Six of the oxide inks
studied are converted into their corresponding phase-pure, crystalline sulfide (i.e., ZnS,
Ag
2
S, CdS, SnS, Sb
2
S
3
, and PbS), whereby the alkanethiol in the solvent mixture acts as
the sole sulfur source (Figure 2.2 a–f). In some cases, the chemical identity of the
alkanethiol has an effect on the crystal structure of the recovered sulfide. For example,
while the EtSH-en and Merc-en inks of ZnO and CdO give the corresponding phase-pure
cubic sulfides after annealing to 300 °C, the EDT–en inks of these same oxides give a
mixture of cubic and hexagonal phases after annealing to the same temperature. This
result suggests that the phase of these sulfides can be kinetically influenced depending on
the ink formulation.
56
Mixing the oxide inks with stoichiometric amounts of inks of selenium or tellurium
results in one of two scenarios: 1) the mixture remains fully dissolved as an ink (we will
refer to this result as a “compound ink”), or 2) a solid precipitates out of solution. Each of
the compound inks was thermally decomposed at 300–375 °C after solution deposition,
or in the case that a precipitate forms, it was isolated by centrifugation and washed with
isopropyl alcohol. It was found that phase-pure, crystalline Ag
2
Se, Ag
2
Te, and PbTe were
all recoverable by direct precipitation out of solution at room temperature, while phase-
pure Cu
2
Se, ZnSe, CdSe, and Sb
2
Se
3
were recoverable from the corresponding compound
inks via solution deposition and annealing (Figure 2.2 g–m). Sulfur volatilizes at a lower
temperature than selenium, and it is likely that this volatility difference between sulfur
and selenium plays a role in the conversion to selenides upon annealing rather than
sulfides.
6b,10
The prospective use of the compound inks to prepare binary and ternary chalcogenide
semiconductors will now be explored in more detail using Sb
2
Se
3
as a model system. In
recent years, both Sb
2
S
3
and Sb
2
Se
3
have been explored as thin film absorbers for
photovoltaic devices due to their near optimal direct band gaps of 1.7–1.1 eV, high
absorption coefficients in the visible region (α=10
5
cm
−1
), and the fact that they are
comprised of relatively Earth-abundant elements.
11
To demonstrate the utility of the
compound inks, inks of Sb
2
O
3
and selenium in EtSH-en were combined in a 1:3
stoichiometric ratio and then the compound ink was subsequently spin coated onto glass
57
substrates and annealed to 350 °C to give nearly opaque dark gray films of crystalline
Sb
2
Se
3
(Figure 2.3 a).
Figure 2.2. Powder XRD diffraction patterns of crystalline materials recovered from annealing
(a-f) metal oxide inks from three different alkanethiol-en solutions, (g-k) selenium compound
inks from EtSH-en solutions, and (l-m) as-precipitated materials from compound inks with
tellurium.
58
Figure 2.3. (a) Dissolved Sb
2
O
3
and selenium in EtSH–en and a Sb
2
Se
3
film on a glass substrate.
(b) Cross-sectional SEM micrograph of a Sb
2
Se
3
thin film on glass. (c) Linear voltage sweep of p-
type Sb
2
Se
3
with chopped 1-Sun simulated illumination in contact with 0.1m Eu(NO
3
)
3
(aq), and
(d) transient photoresponse of Sb
2
Se
3
under potential control at -700 mV versus Ag wire.
Thermogravimetric analysis (TGA) was used to determine the temperature needed to
decompose the compound ink and convert it to Sb
2
Se
3
. After drying the compound ink to
125 °C, the sample displayed a multistep mass loss of ca. 65 wt % by TGA with an end
point of decomposition observed at 300 °C (Figure 2.4). FT-IR spectroscopy corroborated
the loss of organics after annealing to 300 °C, as evidenced by the complete loss of
strong ν(N-H) and ν(C-H) stretching bands originating from the solvent mixture (Figure
2.5). Powder X-ray diffraction (XRD) confirmed that the films were crystalline and
phase-pure orthorhombic Pbnm Sb
2
Se
3
(Figure 2.2 k), and bands characteristic of the
orthorhombic structure at 84 (translation), 118 (Se-Sb-Se bending), 190 (Se-Sb-Se
59
bending), and 253 cm
−1
(Sb-Se stretching) were observed by Raman spectroscopy (Figure
2.6).
12
Inductively coupled plasma optical emission spectroscopy (ICP-OES) confirmed
that the material possessed a Sb/Se ratio of ca. 0.65, as expected for a near-stoichiometric
material without significant oxide incorporation. This was corroborated by X-ray
photoelectron spectroscopy (XPS) data (Figure 2.7), which shows only one chemical
environment for antimony, and no evidence of an O1s peak, indicating that the material is
Sb
2
Se
3
without significant oxide impurities.
Figure 2.4. TGA traces for five nominal compositions of Sb2Se
3–x
S
x
alloy inks first dried to 125
˚C under flowing nitrogen showing end point of decomposition by ~300 ˚C.
Figure 2.5. FT-IR spectra of compound Sb
2
O
3
/Se inks from EtSH-en (a) dried at room
temperature under nitrogen flow and (b) annealed under nitrogen to 300 ˚C. After annealing to
300 ̊C, there is no evidence for remaining organics from the solvent mixture.
60
Figure 2.6. Raman spectrum of crystalline Sb
2
Se
3
derived from a EtSH-en ink. Bands could be
assigned to Sb
2
Se
3
at 84 cm
-1
(translation), 118 cm
-1
(Se-Sb-Se bending) 190 cm
-1
(Se-Sb-Se
bending), and 153 cm
-1
(Sb-Se Stretching).
Figure 2.7. XPS spectrum of Sb3d/O1s region of Sb
2
Se
3
recovered from an EtSH-en ink showing
a single Sb doublet and the absence of any O1s signal.
Scanning electron microscopy (SEM) images showed that the Sb
2
Se
3
films were ca.
340 nm thick with a polycrystalline morphology (Figure 2.3 b). The optical band gap of
the Sb
2
Se
3
film was estimated by diffuse reflectance UV/vis spectroscopy using an
integrating sphere. Tauc plots suggest that the Sb
2
Se
3
films possess a direct optical band
gap transition of 1.22 eV (i.e., (F(R)hν)
2
versus hν), which is commensurate with values
reported for the bulk material.
11
Photoelectrochemical (PEC) experiments were
conducted on the Sb
2
Se
3
thin films as an indication of how the material might perform in
61
solar energy conversion applications. The transient photocurrent response was measured
with the Sb
2
Se
3
films in contact with 0.1 M Eu(NO
3
)
3
(aq) as the redox mediator, and a
calibrated xenon arc lamp was cycled on and off at regular time intervals to provide 1-
Sun chopped illumination. As shown in Figure 2.3 c, the films clearly give cathodic
photocurrent that increases as the potential is swept to values negative of −200 mV vs Ag
wire, signifying that they demonstrate p-type behavior. DC potential amperometry gave
photocurrents >130 µA cm
−2
at an applied bias of −700 mV vs Ag wire under 1-Sun
illumination (Figure 2.3 d).
The ability to engineer band gap energies of inorganic semiconductors is important
for photovoltaic applications, where the band gap energy of the absorber material can be
optimized to match the peak flux of the solar spectrum.
13
It is possible to engineer the
band gap of a material by creating a solid solution between two isostructural
semiconductors having different band gaps. Solid solutions of Sb
2
Se
3−x
S
x
thin films have
been previously synthesized by selenization of Sb
2
S
3
, chemical bath deposition, atomic
layer deposition, and thermal evaporation;
14
however, all of these methods require energy
intensive (high vacuum) or complex and exacting (precise control over temperature,
solution pH, etc.) conditions for thin film deposition. We prepared five nominal
compositions of Sb
2
Se
3−x
S
x
alloys by mixing the Sb
2
O
3
and selenium inks in alkanethiol–
en solvent mixtures in varying Sb/Se stoichiometric ratios. After annealing to 435 °C,
crystalline and phase-pure Sb
2
Se
3−x
S
x
alloys (x=3, 2.25, 1.50, 0.75, and 0) were
recovered. The actual elemental composition of each alloy matched the nominal
62
composition, within experimental error, as determined by ICP-OES. The powder XRD
patterns for the end members (x=3 and 0) can be indexed to the
orthorhombic Pbnm patterns of crystalline Sb
2
S
3
and Sb
2
Se
3
(PDF# 00-042-1393 and 01-
072-1184, respectively) (Figure 2.8 a). The Sb
2
Se
3−x
S
x
alloys display the same overall
diffraction patterns as the Sb
2
S
3
and Sb
2
Se
3
end-members, indicating they are
isostructural with the Pbnm crystal structure. As expected for the smaller anionic radius
of S
2−
(Shannon–Prewitt anionic radii of S
2−
and Se
2−
are 1.84 and 1.98
Å,
15
respectively), a gradual shift to higher 2θ was observed as the sulfur content in the
Sb
2
Se
3−x
S
x
alloys increased. The measured unit cell volume demonstrates a linear
dependence on the experimentally determined elemental composition (Figure 2.8 b), as it
decreases monotonically with increasing sulfur content. This linear dependence is
consistent with Vegard’s law and demonstrates the compositional homogeneity of these
Sb
2
Se
3−x
S
x
alloys.
Figure 2.8. (a) Powder XRD patterns of five compositions of Sb
2
Se
3−x
S
x
alloys. (b) Lattice
parameters and unit cell volumes of the five solid solutions showing a linear dependence of unit
cell volume on composition.
63
The band gap energies of the alloys also vary as a function of Sb
2
Se
3−x
S
x
alloy
composition, which can be easily controlled with the compound ink formulation. Tauc
plots for the direct optical band gaps derived from UV/vis diffuse reflectance spectra for
each of the five nominal compositions of Sb
2
Se
3−x
S
x
(x = 3, 2.25, 1.50, 0.75, and 0), as
well as bulk Sb
2
S
3
and Sb
2
Se
3
powders, are given in Figure 2.9. There is an obvious
systematic increase in band gap energy as the nominal sulfur content increases (1.22,
1.27, 1.33, 1.52, and 1.63 eV for x = 0, 0.75, 1.5, 2.25, and 3, respectively). As has been
shown previously for Sb
2
Se
3−x
S
x
solid solutions,
16
the change in direct band gap follows a
quadratic dependence on composition of selenium, following the equation E
g,dir.
(x) =
0.03175 x
2
+ 0.04743 x + 1.216 for Sb
2
Se
3−x
S
x
(Figure 2.10). This band gap tunability
from 1.2–1.6 eV demonstrates how easily the compound alkanethiol–en inks can be
utilized to control chalcogenide semiconductor composition and engineer the resulting
band gaps.
Figure 2.9. Tauc plots showing band gap tunablity of five Sb
2
Se
3−x
S
x
alloy compositions as
measured by UV/vis diffuse reflectance spectroscopy. The dotted traces correspond to as-bought
bulk Sb
2
Se
3
(E
g
= 1.2 eV) and Sb
2
S
3
(E
g
= 1.6 eV) powders. The alloys possess direct band baps
of 1.63, 1.52, 1.33, 1.27, 1.22 eV for nominal x = 3, 2.25, 1.5, 0.75, and 0 compositions,
respectively.
64
Figure 2.10. Band gap energy as a function of nominal alloy composition showing a quadratic
dependence E
g,dir.
(x) = 0.03175 x
2
+ 0.04743 x + 1.216 for Sb
2
Se
3–x
S
x
alloys.
2.4 Experimental
2.4.1 General Considerations
All reagents and solvents were used as received. 1,2-Ethylenediamine (en, 99.5%),
ethanethiol (EtSH, 97%), 2-mercaptoethanol (Merc, 99.0%), germanium (IV) oxide
(GeO
2
, 99.99%), and arsenic (III) oxide (As
2
O
3
, 99.995%) were all purchased from
Sigma Aldrich. Cadmium (II) oxide (CdO, 99%) was purchased from Strem Chemicals.
1,2-ethanedithiol (EDT, 98+%), silver (I) oxide (Ag
2
O, 99%), copper (I) oxide (Cu
2
O,
99.9%), zinc (II) oxide (ZnO, 99.99%), tin (II) oxide (SnO, 99%), antimony (III) oxide
(Sb
2
O
3
, 99.999%), lead (II) oxide (PbO, 99.999%), bismuth (III) oxide (Bi
2
O
3
,
99.999%), selenium powder (Se, 99.5%), and tellurium powder (Te, 99.999%) were all
purchased from Alfa Aesar.
65
2.4.2 Preparation of Oxide Inks
Metal oxides were weighed out and added to 1.0 mL of ethylenediamine and 0.25 mL
of alkanethiol. The solutions were agitated for 5 min. If the solid oxide was not totally
dissolved, a magnetic stir bar was added and the solution was stirred for no longer than
24 h. Solubility limits in wt% were determined gravimetrically at room temperature. Se
and Te inks were made by adding bulk Se or Te to 1.0 mL of ethylenediamine and 0.25
mL of ethanethiol. The Se solution was completely dissolved within 2 min of agitation at
room temperature, and the Te required ~15 min of stirring using a magnetic stir bar.
Compound inks were made by combining stoichiometric amounts of dissolved oxide and
Se or Te together with stirring.
2.4.3 Recovery of Products
All oxide solutions were thermalized on a temperature controlled hot plate to 300-
400 ̊C under flowing nitrogen followed by a slow cool down to room temperature. The
compound inks were thermalized to 350-375 ̊C under the same conditions. Sb
2
Se
3–x
S
x
solid solutions were annealed to 435 ̊C. Metal chalcogenide precipitates were isolated by
centrifugation and washing with isopropyl alcohol in triplicate, followed by drying under
a flow of nitrogen for 24 h at room temperature.
66
2.4.4 Sb
2
Se
3
Thin Films and Photoelectrochemistry
1 x 4 cm
2
pieces of FTO/glass (Sigma Aldrich) were cleaned with detergent followed
by rinsing with ethanol, isopropyl alcohol, and acetone. The substrates were cleaned
using a UV-ozone treatment directly before use. The en/EtSH/Sb
2
O
3
/Se stoichiometric
compound ink was spin coated on the FTO/glass using a Laurell Technologies
Corporation WS400Ez-6NPPLITE single-wafer spin processor under flowing nitrogen.
Immediately after the ink was deposited, the substrates were annealed to 350 ̊C under
flowing nitrogen and allowed to cool to room temperature. They were stored under
nitrogen until photoelectrochemistry (PEC) measurements.
An electrolyte solution of 0.1 M aqueous Eu
3
(NO
3
)
3
was prepared using nitrogen-
sparged deionized water. The light source was a xenon arc lamp that was calibrated to 1-
Sun intensity. The PEC experiments were performed with a BASi Epsilon-EC
potentiostat. The Sb
2
Se
3
/FTO/glass electrode was submerged in the electrolyte solution in
a quartz cuvette. A Pt-wire acted as the counter electrode and a Ag-wire acted as a
pseudo-reference electrode and the Sb
2
Se
3
/FTO/glass substrate was the working
electrode.
2.4.5 Spectroscopic Characterization
FT-IR spectra were measured on a Bruker Vertex 80. Samples were prepared by drop
casting the compound Sb
2
O
3
/Se ink onto a ZnSe window and either drying under
67
nitrogen flow or annealing to 300 ̊C under flowing nitrogen. UV-vis diffuse reflectance
spectra were measured with a Perkin Elmer Lamba 950 equipped with a 150 mm
integrating sphere. Raman data were gathered on a Horiba Jobin Yvon spectrometer.
2.4.6 Thermal Gravimetric Analysis (TGA)
TGA was performed on a TA Instruments TGA Q50 instrument with an alumina
crucible under a flowing nitrogen atmosphere with a heating rate of 10 ̊C/min. TGA
samples were pre-dried at 125 ̊C under flowing nitrogen for 5 min, and sample sizes of
~5-10 mg were analyzed.
2.4.7 Powder X-ray diffraction (XRD)
XRD data was collected with a Rigaku Ultima IV diffractometer in parallel beam
geometry (2 mm beam width) using Cu Kα radiation (λ = 1.54 Å).
2.4.8 Elemental Analysis
Inductively coupled plasma optical emission spectroscopy (ICP-OES) was performed
on a Thermo Scientific iCAP 7000 series ICP-OES. Samples were prepared from alloys
recovered from inks of en/EDT/Sb
2
O
3
+ en/EtSH/Se for x = 3, 2.25, 1.5 and 0.75 and
en/EtSH/Sb
2
O
3
+ en/EtSH/Se for x = 3. ~4.5 mg of powder was digested with 2 mL of
68
conc. HNO
3
and filtered through a 0.45 µm filter. This solution was then diluted to 25
mL with 2 M HNO
3
. Each sample was run in triplicate, the presented results are an
average of the three measurements.
2.4.9 X-ray Photoelectron Spectroscopy (XPS)
The high resolution XPS spectrum was acquired using a Kratos Axis Ultra X-ray
photoelectron spectrometer with the analyzer lens in hybrid mode. A monochromatic
aluminum anode with an operating current of 6 mA and voltage of 10 kV was used with a
step size of 0.1 eV, a pass energy of 20 eV, and a pressure range between 1-3 x 10
-8
torr.
The binding energy was referenced to the C1s core level at 284 eV.
2.5 Conclusion
In conclusion, we have demonstrated the ability of the thiol–en solvent system to
readily dissolve ten bulk metal(loid) oxides under ambient conditions. From these ten
inks, six crystalline sulfides (ZnS, Ag
2
S
,
CdS, SnS, Sb
2
S
3
, and PbS) can be recovered
upon mild annealing. The oxide inks may also be stoichiometrically mixed with dissolved
selenium or tellurium to give compound inks that yield crystalline Cu
2
Se, ZnSe, CdSe,
and Sb
2
Se
3
upon annealing, while PbTe, Ag
2
Se, and Ag
2
Te can be isolated as phase-pure
precipitates at room temperature. The compound ink of dissolved Sb
2
O
3
and selenium in
EtSH-en can be used to spin coat Sb
2
Se
3
thin films that give strong photocurrent under 1-
69
Sun illumination, thus highlighting the potential of Sb
2
Se
3
thin films processed from this
solvent system for solar energy-conversion applications. Moreover, homogeneous solid
solutions of Sb
2
Se
3−x
S
x
were readily made from compositionally tailored compound inks
with varying Sb/Se ratios, allowing the direct optical band gap to be tuned with
increasing selenium content from 1.6–1.2 eV. Given these results, we believe this
alkanethiol–en solvent system holds considerable potential for the solution conversion of
cheap oxide starting materials to compositionally controlled chalcogenide
semiconductors.
2.6 References
(1) a) Choi, Y. C.; Lee, D. U.; Noh, J. H.; Kim, E. K.; Seok, S. I. Highly Improved
Sb
2
S
3
Sensitized-Inorganic–Organic Heterojunction Solar Cells and Quantification of
Traps by Deep-Level Transient Spectroscopy. Adv. Funct. Mater. 2014, 24, 3587-3592;
b) Todorov, T.K.; Gunawan, O.; Gokmen, T.; Mitzi, D. B. Solution-processed
Cu(In,Ga)(S,Se)
2
absorber yielding a 15.2% efficient solar cell. Prog. Photovoltaics: Res.
Appl. 2013, 21, 82-87; c) Zhou, H.; Chen, Q.; Li, G.; Luo, S.; Song, T.; Duan, H.; Hong,
Z.; You, J.; Liu, Y.; Yang, Y. Interface engineering of highly efficient perovskite solar
cells. Science 2014, 345, 542-546.
(2) Reuss, R. H.; Chalamala, B. R. in Solution Processing of Inorganic Materials
(Ed.: D. B. Mitzi), Wiley, New York 2009, pp. 1-32.
(3) Wang, R. Y.; Feser, J. P.; Gu, X.; Yu, K. M.; Segalman, R. A.; Majumdar, A.;
Milliron, D. J.; Urban, J. J. Universal and Solution-Processable Precursor to Bismuth
Chalcogenide Thermoelectrics. Chem. Mater. 2010, 22, 1943-1945.
(4) Milliron, D. J.; Raoux, S.; Shelby, R. M.; Jordan-Sweet, J. Solution-phase
deposition and nanopatterning of GeSbSe phase-change materials. Nat. Mater. 2007, 6,
352-356.
(5) Milliron, D. J.; Mitzi, D. B.; Copel, M.; Murray, C. E. Solution-Processed Metal
70
Chalcogenide Films for p-Type Transistors. Chem. Mater. 2006, 18, 587-590.
(6) a) Mitzi, D. B.; Kosbar, L. L.; Murray, C. E.; Copel, M.; Afzali, A. High-mobility
ultrathin semiconducting films prepared by spin coating. Nature 2004, 428, 299-303; b)
Mitzi, D. B.; Solution Processing of Chalcogenide Semiconductors via Dimensional
Reduction. Adv. Mater. 2009, 21, 3141-3158; c) Yuan, M.; Mitzi, D. B.
Solvent properties of hydrazine in the preparation of metal chalcogenide bulk materials
and films. Dalton Trans. 2009, 6078-6088.
(7) a) Webber, D. H.; Buckley, J. J.; Antunez, P. D.; Brutchey, R. L.; Facile
dissolution of selenium and tellurium in a thiol–amine solvent mixture under ambient
conditions. Chem. Sci. 2014, 5, 2498-2502; b) Antunez, P. D.; Torelli, D. A.; Yang, F;
Rabuffetti, F. A.; Lewis, N. S.; Brutchey, R. L.; Low Temperature Solution-Phase
Deposition of SnS Thin Films. Chem. Mater. 2014, 26, 5444-5446; c) Webber, D. H.;
Brutchey, R. L.; Alkahest for V
2
VI
3
Chalcogenides: Dissolution of Nine Bulk
Semiconductors in a Diamine-Dithiol Solvent Mixture. J. Am. Chem. Soc. 2013, 135,
15722-15725; d) Buckley, J. J.; Greaney, M. J.; Brutchey, R. L.; Ligand Exchange of
Colloidal CdSe Nanocrystals with Stibanates Derived from Sb
2
S
3
Dissolved in a Thiol-
Amine Mixture. Chem. Mater. 2014, 26,6311-6317.
(8) Vineyard, B. D. Versatility and the Mechanism of the n-butyl-amine-catalyzed
Reaction of Thiols with Sulfur. J. Org. Chem. 1967, 32, 3833-3836.
(9) Yang, Y.; Wang, G.; Zhao, W.; Tian, Q.; Huang, L.; Pan, D. Solution-Processed
Highly Efficient Cu
2
ZnSnSe
4
Thin Film Solar Cells by Dissolution of Elemental Cu, Zn,
Sn, and Se Powders. ACS Appl. Mater. Interfaces 2015, 7, 460-464.
(10) CRC Handbook of Chemistry and Physics, 95th ed, CRC, Boca Raton, 2015.
(11) a) Choi, Y. C.; Lee, Y. H.; Im, S. H.; Noh, J. H.; Mandal, T. N.; Yang, W. S.;
Seok, S. I. Efficient Inorganic-Organic Heterojunction Solar Cells Employing Sb
2
(S
x
/Se
1-
x
)
3
Graded-Composition Sensitizers. Adv. Energy Mater. 2014, 4, 1301680; b) Zhou, Y.;
Leng, M.; Xia, Z.; Zhong, J.; Song, H.; Liu, X.; Yang, B.; Zhang, J.; Chen, J.; Zhou, K.;
Han, J.; Cheng, Y.; Tang, J. Solution-Processed Antimony Selenide Heterojunction Solar
Cells. Adv. Energy Mater. 2014, 4, 1301846; c) Choi, Y. C.; Lee, D. U.; Noh, J. H.; Kim,
E. K.; Seok, S. I. Highly Improved Sb
2
S
3
Sensitized-Inorganic–Organic Heterojunction
Solar Cells and Quantification of Traps by Deep-Level Transient Spectroscopy. Adv.
Funct. Mater. 2014, 24, 3587–3592.
(12) Efthimiopoulos, I.; Zhang, J.; Kucway, M.; Park, C.; Ewing, R. C.; Wang, Y.
Sb
2
Se
3
Under Pressure. Sci. Rep. 2013, 3, 2665 –2673.
(13) Shockley, W. H.; Queisser, J. Detailed Balance Limit of Efficiency of p-
71
n Junction Solar Cells. J. Appl. Phys. 1961, 32, 510–519.
(14) a) Suárez-Sandoval, D. Y.; Nair, M. T. S.; Nair, P. K.; Photoconductive Antimony
Sulfide-Selenide Thin Films Produced by Heating a Chemically Deposited Se-Sb
2
S
3
Layer. J. Electro-chem. Soc. 2006, 153, C91–C96; b) Calixto-Rodriguez, M.; Garcia, H.
M.; Nair, M. T. S.; Nair, P. K. Antimony Chalcogenide/Lead Selenide Thin Film Solar
Cell with 2.5% Conversion Efficiency Prepared by Chemical Deposition. ECS J. Solid
State Sci. Technol. 2013, 2, Q69–Q73; c) Yang, R. B.; Bachmann, J.; Pippel, E.; Berger,
A.; Woltersdorf, J.; Gosele, U.; Nielsch, K. Pulsed Vapor-Liquid-Solid Growth of
Antimony Selenide and Antimony Sulfide Nanowires. Adv. Mater. 2009, 21, 3170-3174;
d) El-Sayad, E.; Compositional Dependence of the Optical Properties of Amorphous
Sb
2
Se
3−x
S
x
thin films. J. Non-Cryst. Solids 2008, 354, 3806–3811.
(15) Ladd, M. Symmetry of Crystals and Molecules, Vol. 1, Oxford University Press,
New York, 2014, p.45.
(16) Deng, Z.; Mansuripur, M.; Muscat, A. J. Simple Colloidal Synthesis of Single-
Crystal Sb−Se−S Nanotubes with Composition Dependent Band-Gap Energy in the Near-
Infrared. Nano Lett. 2009, 9, 2015–2020.
72
Chapter 3. Earth-Abundant CuSbS
2
Thin Films Solution Processed From
Thiol–Amine Mixtures*
*Published in J. Mater. Chem. C 2016, 4, 6230–6233.
3.1 Abstract
Solution processing is a practical low-cost strategy for depositing semiconductor thin
films. A binary thiol–amine solvent mixture dissolves bulk Cu
2
S and Sb
2
S
3
under ambient
conditions, allowing for solution deposition and low temperature recovery of CuSbS
2
. The
resulting films of Earth-abundant CuSbS
2
possess optoelectronic properties suitable for
photovoltaic applications.
3.2 Introduction
Macroelectronic applications based on semiconductor thin films, such as CdTe solar cells
and flat panel displays containing ITO transparent conducting electrodes, are revolutionizing
the way the world accesses and uses electricity. Currently, such materials are industrially
deposited by relatively expensive and energy intensive techniques, such as vapor transport
deposition and sputtering.
1
As an attractive alternative to these traditional deposition methods,
the solution processing of inks to give semiconductor thin films offers a route that is both low
cost and rapid. There are now many examples, for example, demonstrating how
semiconductors relevant as absorber layers for thin film solar cells can be solution deposited,
such as FeS
2
(iron pyrite), SnS (Herzenbergite), Cu(In,Ga)(S,Se)
2
(CIGS),
CH
3
NH
3
PbI
3
perovskites, and Cu
2
ZnSn(S,Se)
4
(CZTS).
2–6
73
Molecular inks to give semiconductor thin films generally take one of two forms: (i) pre-
prepared molecular complexes that are dissolved as a soluble precursor, or (ii) bulk inorganic
materials with the desired elemental composition that are dissolved to generate soluble
precursors in situ. The direct dissolution of bulk materials is conceptually attractive because it
bypasses the need to pre-synthesize discrete molecular complexes; however, this “dissolve and
recover” process is typically hampered by the insolubility of most metals, metal oxides, and
metal chalcogenides in common solvents. To overcome this challenge, we first reported in
2013 a binary solvent system comprised of a short chain alkanethiol and ethylenediamine
(termed an “alkahest”), which we have since demonstrated to have appreciable solvent power
to rapidly dissolve numerous bulk materials under ambient conditions.
7–12
The resulting inks
can be readily solution processed because of the appropriate solvent volatility, and upon
thermal annealing at low temperatures, give phase-pure semiconductor thin films. For
example, this thiol–amine solvent system has been used by Arnou et al., Zhao et al., and
Zhang et al. to solution process thin films of multinary semiconductors, such as CIS, CIGS,
and CZTS.
13–16
Chalcostibite (CuSbS
2
) is a promising, albeit less studied, ternary semiconductor that is
comprised of relatively Earth-abundant and less toxic elements.
17
This material composition
offers a sustainable alternative to materials that contain environmentally harmful heavy metals
(e.g., Cd or Pb) and/or scarce elements (e.g., Te, In).
17
Theoretically and experimentally, bulk
CuSbS
2
has been shown to have a near ideal direct band gap of 1.4-1.6 eV and a high
absorption coefficient (10
5
cm
−1
) in the visible region, making it a suitable absorber material
74
for single junction photovoltaic devices.
17,18
Power conversion efficiencies of 3.1% have
recently been achieved in Al:ZnO/CdS/CuSbS
2
/Mo/glass solar cell device structures.
19
Thin
films of CuSbS
2
have been previously prepared by solution methods such as sequential
chemical bath deposition of CuS on Sb
2
S
3
films followed by annealing at 400 °C, or by the
deposition of nanocrystal inks; however, these methods are complicated either by the exacting
conditions (i.e., time, pH, concentration, temperature) required for chemical bath deposition,
or depend on the prerequisite hot injection synthesis and purification of CuSbS
2
nanocrystals
to form the ink.
20, 21
More recently, there have been reports of simpler solution deposition
methods for CuSbS
2
thin films using precursor inks from hydrazine, thioglycolic acid and
ethanolamine, or thiourea in alcohol as solvent systems; however, these methods either require
highly toxic solvents, lack thorough investigation of the resulting material, or produce thin
films with impurity phases (such as Cu
2
S), respectively.
6, 22, 23
Deposition of phase-pure
CuSbS
2
thin films is challenging due to (i) other ternary copper antimony sulfide phases (i.e.,
Cu
3
SbS
4
, Cu
12
Sb
4
S
13
) that may be formed depending on the stoichiometry or thermal
conditions used during synthesis, and/or (ii) commonly observed binary impurities, such as
Cu
2
S and Sb
2
S
3
.
23, 24
Herein, we demonstrate the ability of a thiol–amine solvent mixture to
readily dissolve bulk Cu
2
S and Sb
2
S
3
under ambient conditions to generate a stable ink that
can be used to deposit CuSbS
2
thin films via mild annealing. The appropriateness of these
solution-deposited thin films as absorber layers for photovoltaics has been established by
optical and van der Pauw measurements.
75
3.3 Results and Discussion
In a 1 : 4 vol/vol mixture of 2-mercaptoethanol and ethylenediamine, Cu
2
S and Sb
2
S
3
have
room temperature and ambient pressure solubility limits of ca. 124 and 280 mg mL
−1
,
respectively, based on thermogravimetric analysis (TGA, Figure 3.1 a). A clear yellow ink for
CuSbS
2
that is free of visible scattering can be made by co-dissolving bulk Cu
2
S and
Sb
2
S
3
powders simultaneously in this thiol–amine solvent mixture at room temperature and
ambient pressure over the course of 20 min. TGA was used to determine the minimum
temperature required to remove the volatile organic residues and decomposition products at
the endpoint of mass loss. The TGA trace of the CuSbS
2
ink (pre-dried to 160 °C under
nitrogen) indicates that the completion of organic decomposition and volatile loss occurs by
300 °C (Figure 3.1 a). FT-IR spectra of the dried ink before and after annealing corroborates
this, with the observed ν(C–H), ν(O–H), and ν(N–H) stretches in the dried ink disappearing by
300 °C (Figure 3.1 b).
Figure 3.1 (a) TGA traces of inks containing Sb
2
S
3
, Cu
2
S, and Sb
2
S
3
+ Cu
2
S (CuSbS
2
precursor
ink), showing end of mass loss by 300 °C in all cases. (b) FT-IR absorption spectra of
CuSbS
2
precursor ink dried at 100 °C under vacuum and annealed to 300 °C under flowing
nitrogen.
76
In agreement with previous reports, it was necessary to use a slightly nominally Cu-poor
stoichiometry when preparing the ink in order to avoid spurious Cu-rich phases of copper
antimony sulfide.
23
The material recovered from an ink containing a 1 : 1 stoichiometry of
dissolved Cu
2
S to Sb
2
S
3
gives an X-ray diffraction (XRD) pattern that shows diffraction peaks
resulting from a Cu-rich impurity phase and Sb
2
O
3
(Figure 3.2).
Figure 3.2 Powder XRD patterns of alternative ink formulation (Cu:Sb 1:1) and lower annealing
temperature (350 ˚C), revealing impurity phases when either Cu-poor conditions, or a lower annealing
temperature, is used.
A nominal molar ratio of 0.9 : 1 of dissolved Cu
2
S to Sb
2
S
3
, however, gives phase-pure
CuSbS
2
upon annealing to 400 °C, as confirmed by XRD and Raman spectroscopy (Figure
3.3). The XRD pattern of the recovered material matches
orthorhombic Pnma CuSbS
2
(PDF#01-073-3954).
25
A single band at 335 cm
−1
in the Raman
spectrum of the recovered material corresponds to the A
g
mode of CuSbS
2
. Moreover, in
corroboration with the XRD data, no Raman bands are observed for a Cu
12
Sb
4
S
13
(316 and
355 cm
−1
) impurity phase.
26
77
Figure 3.3 (a) Powder XRD data from recovered phase-pure orthorhombic CuSbS
2
and (b)
Raman spectrum of recovered CuSbS
2
, showing a single peak at 335 cm
−1
corresponding to the
A
g
mode.
While TGA and FT-IR suggest that annealing to 300 °C is sufficient to remove all volatile
organic content and decomposition products, XRD studies reveal that a higher annealing
temperature is required to recover phase-pure CuSbS
2
. This is clearly shown by the XRD
pattern of the ink annealed to 350 °C, which shows diffraction peaks for a Cu-rich impurity
phase, while the ink annealed to 400 °C gives a phase-pure diffraction pattern for
orthorhombic CuSbS
2
(Figure 3.2).
Figure 3.4 Rietveld analysis of the powder XRD pattern for recovered CuSbS
2
shown as experimental
data (circles), calculated pattern (red trace), difference data (blue trace), and green tic marks
corresponding to the refined CuSbS
2
phase.
78
Rietveld refinement of the recovered phase-pure CuSbS
2
XRD pattern gives lattice
parameters (a = 6.018(3) Å, b = 3.794(2) Å, c = 14.490(6) Å) that closely match the
previously reported values of PDF# 01-073-3954 (Table 3.1, Figure 3.4, and Figure 3.5).
Table 3.1 Rietveld analysis of X-ray diffraction data of CuSbS
2
.
CuSbS
2
a (Å) 6.018(3) Cu – S(1) × 2 2.330(5)
b (Å) 3.794(2) Cu – S(2) 2.345(13)
c (Å) 14.490(6) Cu – S(2) 2.235(9)
V (Å
3
) 330.9(4) Sb – S(1) 2.427(9)
X
Cu
0.7494(13) Sb – S(1) × 2 3.088(6)
X
Sb
0.2297(5) Sb – S(2) × 2 2.609(6)
X
S(1)
0.6260(13)
X
S(2)
0.1384(17) S(1) – Cu – S(1) 109.0(4)
Z
Cu
0.1720(4) S(1) – Cu – S(2) × 2 110.2(3)
Z
Sb
0.0632(3) S(1) – Cu – S(2) × 2 111.6(3)
Z
S(1)
0.0939(6) S(2) – Cu – S(2) 104.3(4)
Z
S(2)
0.1808(7) Cu – S(1) – Cu 109.0(4)
U
Cu
(Å
2
)
a
1.61 Cu – S(2) – Cu 110.5(5)
U
Sb
(Å
2
)
a
0.37
U
S(1)
(Å
2
)
a
1.72 R
wp
(%) 13.2
U
S(2)
(Å
2
)
a
2.87 χ
2
1.79
a
Atomic
displacement
parameters
are
given
as
100
×
Uiso
79
Figure 3.5 Space-filling polyhedra depiction of refined structure of recovered CuSbS
2
, with CuS
4
tetrahedra, and SbS
5
square pyramids were Cu is pink, Sb is red, and S is aqua.
Based on our previous work exposing the molecular nature of dissolved species in this
solvent system, we expect that upon co-dissolution, bulk Sb
2
S
3
and Cu
2
S are broken down
completely to give molecular solutes that are able to thoroughly mix in the resulting ink,
leaving no Sb
2
S
3
and Cu
2
S in solution.
9,12
Similar to a previous report using hydrazine to
solution process CuSbS
2
, where non-stoichiometric Cu-poor inks are used, excess Sb is driven
off as Sb
2
S
3
as a result of annealing to 400 °C.
23
X-ray photoelectron spectroscopy (XPS) was used to study the elemental composition and
valence states of the recovered material. A high resolution scan for the Sb 3d and O 1s region
shown in Figure 3.6 b reveals a non-Gaussian Sb 3d doublet giving a d
5/2
–d
3/2
peak splitting
of ca. 9.3 eV, indicative of an Sb
3+
environment that can be assigned to the expected
CuSbS
2
phase with a binding energy of 528.8 eV for the Sb 3d
5/2
peak. The slight shoulders at
80
higher binding energies on each of the peaks are likely assignable to surface oxides of Sb,
which is corroborated by the presence of a minor O 1s peak at 532.0 eV.
22
The Cu 2p doublet
has binding energies of 931.5 and 951.2 eV, with a peak splitting of 19.7 eV consistent with
previous reports for Cu
+
in CuSbS
2
(Figure 3.6 c).
22,23
The S 2p doublet peaks at 161.0 and
162.2 (Figure 3.6 d) have a splitting of 1.2 eV, which is consistent with previous reports for
S
2−
in CuSbS
2
.
23
Thus, these XPS results verify the valence state of the recovered material to
be Cu
+
Sb
3+
(S
2−
)
2
. This composition is corroborated with elemental analysis by inductively
coupled plasma atomic emission spectroscopy (ICP-AES), which shows a Cu : Sb ratio of 1:1
for the recovered material.
Figure 3.6 XPS spectra for the recovered CuSbS
2
. (a) Survey scan, and high-resolution scans of
the (b) Sb 3d, (c) Cu 2p, and (d) S 2p regions.
To demonstrate the practicality of this method for producing good quality polycrystalline
thin films, the ink (∼0.16 mM) was sequentially deposited in multiple coats by spin coating
followed by annealing at 350 °C, with a final anneal at 400 °C after the last layer was
deposited. This procedure gives dark gray, specularly reflective thin films that are free of
81
pinholes and microcracks (Figure 3.7 a). A cross-sectional scanning electron microscopy
(SEM) image of a 490 nm thick CuSbS
2
film on silicon that was prepared by depositing two
layers is shown in Figure 3.7 b. Increasing the number of layers deposited results in a thicker
film; for example, a 1.1 µm thick film was made by depositing 9 coats of the ink. Furthermore,
Figure 3.7 (a) Photograph of CuSbS
2
precursor ink and thin film of CuSbS
2
on glass substrate
and (b) cross-sectional SEM micrograph of CuSbS
2
thin film deposited from precursor ink. (c and
d) Optical data derived from transmittance spectrum of CuSbS
2
thin film. (c) Plot of absorption
coefficient (α) as a function of wavelength, and (d) Tauc plot extrapolated to estimate a direct
band gap of 1.6 eV.
a SEM micrograph of the 9-layer sample reveals a uniform film that is denser than the 2-layer
sample, demonstrating an enhancement of film quality as the number of layers is increased
(Figure 3.8).
82
Figure 3.8 SEM micrograph of a CuSbS
2
film made by depositing 9 layers of precursor ink.
The root mean square (rms) roughness and surface topography of the 2- and 9-layer
samples were compared using atomic force microscopy (AFM, Figure 3.9). The rms
roughness decreased from 92 nm for the 2-layer sample to 46 nm for the 9-layer sample,
corroborating the interpretation of increased film quality with additional layers.
Figure 3.9 Plane view AFM micrographs of CuSbS
2
film surfaces made by depositing (a) 2 layers and
(b) 9 layers of CuSbS
2
precursor ink.
The optical properties of the films were studied by transmission UV-vis spectroscopy on
the front of an integrating sphere. The transmission spectrum of a film on glass (Figure 3.10)
was used to calculate the absorption coefficient α as a function of wavelength (Figure 3.7 c),
83
which, consistent with previous reports, reaches 7 × 10
4
cm
−1
by 775 nm (1.6 eV) and
continues to rise to 1.5 × 10
5
cm
−1
at 600 nm (2.1 eV).
23
A Tauc plot derived using the
correlation of α with band bap (E
g
) for a direct band gap semiconductor extrapolated an
optical gap of E
g
= 1.6 eV for the thin film (Figure 3.7 d). This value is consistent with the
theoretical direct band gap of CuSbS
2
that has been previously reported (vide supra).
18
Figure 3.10 Transmittance spectrum of CuSbS
2
thin film on glass substrate.
Electronic properties of the resulting CuSbS
2
thin films were studied by the van der Pauw
method. The free hole concentration and resistivity at 290 K were calculated to be n = 3.18 ×
10
19
cm
−3
and ρ = 3.04 × 10
−3
Ω cm, respectively. These values are consistent with the
material being a heavily doped p-type semiconductor. The source of such a high hole
concentration is likely intrinsic copper vacancy defects, which have been reported to be
dominant in CuSbS
2
due to their low formation energy relative to other defects.
23
We
calculated a hole mobility of µ
h
= 64.6 cm
2
V
−1
s
−1
, which is one of the highest in the range of
values reported in the literature for films of this material prepared by other techniques, further
corroborating that the films are comprised of high quality crystalline CuSbS
2
.
22,23,27,28
Such
84
electronic characteristics, combined with the near-ideal optical band gap, high absorption
coefficient, and the ability to easily tune film thickness through deposition of multiple coats,
suggests that CuSbS
2
thin films deposited using this method would be suitable as absorber
layers in photovoltaic devices.
3.4 Experimental
3.4.1 General Considerations
All reagents and solvents were used as received, except in the case of Sb2S3, which was
ground with a mortar and pestle to facilitate faster dissolution. 1,2-Ethylenediamine (en,
99.5%), 2-mercaptoethanol (Merc, 99.0%), Copper (I) Sulfide (Cu
2
S, 99.99%), and Antimony
(III) Sulfide (Sb
2
S3, 99.995%) were all purchased from Sigma Aldrich.
3.4.2 Thin Film Characterization
Scanning electron microscopy (SEM) was performed on a JEOL JSM-7001F
scanningelectron microscope with an operating voltage of 10 kV. Atomic Force Microscopy
(AFM) was performed on an Agilent AFM 5400 in tapping mode at a scan rate of 0.51 lines
per second. Van der Pauw measurements were performed on a 490 nm thick film of CuSbS
2
deposited on a silicon substrate. The substrate was cut into an approximately square shape and
copper leads were attached at each corner using an indium-gallium eutectic. The sample was
mounted in a Physical Properties Measurement System (Quantum Design), which was used to
control the sample temperature and apply a magnetic field of 10000 Oe. Data were collected
85
using Keithley 2182A nanovoltmeter and a Keithley 6220 current source, controlled by a
Keithley 7065 Hall effect card.
3.4.3 Elemental Analysis and Organic Content of Recovered CuSbS
2
Inductively coupled plasma atomic emission spectroscopy (ICP-AES) was performed by
Galbraith Laboratories Inc. Thermal Gravimetric Analysis (TGA) was performed on a TA
Instruments TGA Q50 instrument with an alumina crucible under a flowing nitrogen
atmosphere with a heating rate of 5 ˚C/min. TGA samples were pre-dried at 160-180 ˚C under
flowing nitrogen prior to TGA analysis. FT-IR spectra were measured on a Bruker Vertex 80.
Samples were prepared by drop casting the CuSbS
2
ink onto a ZnSe window and either drying
under vacuum at 100 ˚C or annealing to 300 ˚C under flowing nitrogen.
3.4.4 Structural and Optical Characterization
Powder X-ray Diffraction (XRD) patterns were collected in the 10−80° 2θ range using a
Rigaku Ultima IV diffractometer operated at 44 mA and 40 kV. Cu Kα radiation (λ = 1.5406
Å) was employed. For structural refinements, the step size and collection time were 0.01° and
6 s per step, respectively. All patterns were recorded under ambient conditions. Rietveld
Structural Refinements
29, 30
were carried out using the General Structure Analysis System
(GSAS) software.
31
The following parameters were refined: (1) scale factor, (2) background,
which was modeled using a shifted Chebyshev polynomial function, (3) sample displacement,
(4) peak shape, which was modeled using a modified Thomson−Cox−Hasting pseudo-Voigt,
32
(5) lattice constants, (6) fractional atomic coordinates, and (7) an isotropic thermal parameter
86
for each chemical species (i.e., UCu, USb, and US). The usual Rwp and χ
2
indicators were
employed to assess the quality of the refined structural models.
33
Raman Spectroscopy was
performed using a Horiba Jobin Yvon spectrometer with powdered CuSbS2. UV-vis
transmittance spectroscopy was performed with a Perkin Elmer Lamba 950 equipped with a
150 mm integrating sphere. The thin film on a glass substrate was placed in front of the
integrating sphere. X-ray Photoelectron Spectroscopy (XPS) was performed using a Kratos
Axis Ultra X-ray photoelectron spectrometer with the analyzer lens in hybrid mode. A
monochromatic aluminum anode with an operating current of 6 mA and voltage of 10 kV was
used with a step size of 0.1 eV, a pass energy of 20 eV, and a pressure range between 1-3 x
10–8 Torr. The binding energy was referenced to the C 1s core level at 284.8 eV.
3.4.5 Preparation of CuSbS
2
Ink
22.7 mg (0.143 mmol) Cu
2
S and 53.7 mg (0.158 mmol) Sb
2
S
3
were dissolved in 0.8 mL
en and 0.2 mL merc using sonication and mixing, resulting in a ca. 0.16 mM clear yellow
solution.
3.4.6 Recovery of CuSbS
2
Powdered CuSbS
2
was recovered by drop-casting the CuSbS
2
precursor ink onto glass,
annealing to 400 ˚C for 10 min under flowing nitrogen, cooling to room temperature under
flowing nitrogen, and scraping off the resulting powder from the slides. Thin Films of CuSbS
2
for PPMS measurements were made by spin coating the precursor ink onto ca. 1 cm x 1 cm
substrates using a Laurell Technologies Corporation WS400Ez-6NPPLITE single-wafer spin
87
processor at 3000 r.p.m. for 1 min in a nitrogen atmosphere. In between coats, the films were
annealed on a hot plate at 350 ˚C and allowed to cool to room temperature before the next coat
was spin coated. After the final coat the films were annealed to 400 ˚C for 10 min and allowed
to cool to room temperature.
3.5 Conclusion
In summary, we have shown that the thiol–amine solvent system can be used to easily
solution process phase-pure CuSbS
2
starting from bulk Cu
2
S and Sb
2
S
3
. The resulting films
have a direct optical band gap of 1.6 eV with an absorption coefficient greater than 1 ×
10
5
cm
−1
in the visible region. Electronic measurements confirm that the material is a heavily
doped p-type semiconductor with free hole concentration and mobility values rivalling
literature values, highlighting the potential of polycrystalline CuSbS
2
films solution processed
from a thiol–amine ink as absorber layers in thin film solar cells.
3.6 References
(1) Mitzi, D. B. Solution Processing of Chalcogenide Semiconductors via Dimensional
Reduction. Adv. Mater. 2009, 21, 3141–3158.
(2) Seefeld, S.; Limpinsel, M., Liu, Y.; Farhi, N.; Weber, A.; Zhang, Y.; Berry, N.; Kwon,
Y. J.; Perkins, C. L.; Hemminger, J. C.; Wu, R.; Law, M. Iron Pyrite Thin Films Synthesized
from an Fe(acac)
3
Ink. J. Am. Chem. Soc. 2013, 135, 4412–4424.
(3) Tanusevski, A. Optical and photoelectric properties of SnS thin films prepared by
chemical bath deposition. Semicond. Sci. Technol. 2003, 18, 501–505.
88
(4) Todorov, T. K.; Gunawan, O.; Gokmen T.; Mitzi, D. B. Solution-processed
Cu(In,Ga)(S,Se)
2
absorber yielding a 15.2% efficient solar cell. Prog. Photovoltaics 2013, 21,
82–87.
(5) Jeon, N. J.; Noh, J. H.; Kim, Y. C.; Yang, W. S.; Ryu, S.; Seok, S. I. Solvent
engineering for high-performance inorganic–organic hybrid perovskite solar cells. Nat. Mater.
2014, 13, 897–903.
(6) Tian, Q.; Wang, G.; Zhao, W.; Chen, Y.; Yang, Y.; Huang, L.; Pan, D. Versatile and
Low-Toxic Solution Approach to Binary, Ternary, and Quaternary Metal Sulfide Thin Films
and Its Application in Cu
2
ZnSn(S,Se)
4
Solar Cells. Chem. Mater. 2014, 26, 3098–3103.
(7) Webber, D. H.; Brutchey, R. L. Alkahest for V
2
VI
3
Chalcogenides: Dissolution of
Nine Bulk Semiconductors in a Diamine-Dithiol Solvent Mixture. J. Am. Chem. Soc.,
2013, 135, 15722–15725.
(8) Webber, D. H.; Buckley, J. J.; Antunez, P. D.; Brutchey, R. L. Facile dissolution of
selenium and tellurium in a thiol–amine solvent mixture under ambient conditions. Chem. Sci.
2014, 5, 2498–2502.
(9) Buckley, J. J.; Greaney M. J.; Brutchey, R. L. Ligand Exchange of Colloidal CdSe
Nanocrystals with Stibanates Derived from Sb
2
S
3
Dissolved in a Thiol-Amine Mixture. Chem.
Mater. 2014, 26, 6311–6317.
(10) Antunez, P. D.; Torelli, D. A.; Yang, F.; Rabuffetti, F. A.; Lewis, N. S.; Brutchey, R.
L. Low Temperature Solution-Phase Deposition of SnS Thin Films. Chem. Mater. 2014, 26,
5444–5446.
(11) McCarthy, C. L.; Webber, D. H.; Schueller E. C.; Brutchey, R. L. Solution-Phase
Conversion of Bulk Metal Oxides to Metal Chalcogenides Using a Simple Thiol–Amine
Solvent Mixture. Angew. Chem., Int. Ed. 2015, 54, 8378–8381.
(12) Buckley, J. J.; McCarthy, C. L.; Del Pilar-Albaladejo, J.; Rasul, G.; Brutchey, R.
L. Dissolution of Sn, SnO, and SnS in a Thiol–Amine Solvent Mixture: Insights into the
Identity of the Molecular Solutes for Solution-Processed SnS. Inorg. Chem. 2016, 55,
3175–3180.
(13) Arnou, P.; Cooper, C. S.; Malkov, A. V.; Bowers, J. W.; Walls, J. M. Solution-
processed CuIn(S,Se)
2
absorber layers for application in thin film solar cells. Thin Solid Films
2015, 582, 31–34.
(14) Arnou, P.; van Hest, M. F. A. M.; Cooper, C. S.; Malkov, A. V.; Walls, J. M.;
Bowers, J. W. Hydrazine-Free Solution-Deposited CuIn(S,Se)
2
Solar Cells by Spray
Deposition of Metal Chalcogenides. ACS Appl. Mater. Interfaces 2016, 8, 11893–11897.
89
(15) Zhao, D.; Tian, Q.; Zhou, Z.; Wang, G.; Meng, Y.; Kou, D.; Zhou, W.; Pan, D.;
Wu, S. Solution-Deposited Pure Selenide CIGSe Solar Cells from Elemental Cu, In, Ga,
and Se. J. Mater. Chem. A 2015, 3, 19263–19267.
(16) Zhang, R.; Szczepaniak, S. M.; Carter, N. J.; Handwerker, C. A.; Agrawal, R. A
Versatile Solution Route to Efficient Cu
2
ZnSn(S,Se)
4
Thin-Film Solar Cells. Chem. Mater.
2015, 27, 2114–2120.
(17) Krishnan, B.; Shaji, S.; Ornelas, R. E. Progress in development of copper antimony
sulfide thin films as an alternative material for solar energy harvesting. J. Mater. Sci.: Mater.
Electron. 2015, 26, 4770–4781.
(18) Yu, L.; Kokenyesi, R. S.; Keszler. D. A.; Zunger, A. Inverse Design of High
Absorption Thin-Film Photovoltaic Materials. Adv. Energy Mater. 2013, 3, 43–48.
(19) Septina, W.; Ikeda, S.; Iga, Y.; Harada, T.; Matsumura, M. Thin film solar cell based
on CuSbS
2
absorber fabricated from an electrochemically deposited metal stack. Thin Solid
Films 2014, 550, 700–704.
(20) Rodríguez-Lazcano, Y.; Nair, M. T. S.; Nair, P. K. CuSbS
2
thin film formed through
annealing chemically deposited Sb
2
S
3
–CuS thin films. J. Cryst. Growth 2001, 223, 399–406.
(21) Suehiro, S.; Horita, K.; Yuasa, M.; Tanaka, T.; Fujita, K.; Ishiwata, Y.; Shimanoe, K.;
Kida, T. Synthesis of Copper–Antimony-Sulfide Nanocrystals for Solution-Processed Solar
Cells. Inorg. Chem. 2015, 54, 7840–7845.
(22) Ramos Aquino, J. A.; Rodriguez Vela, D. L.; Shaji, S.; Avellaneda, D. A.; Krishnan,
B. Spray pyrolysed thin films of copper antimony sulfide as photovoltaic absorber. Phys.
Status Solidi C 2016, 13, 24–29.
(23) Yang, B.; Wang, L.; Han, J.; Zhou, Y.; Song, H.; Chen, S.; Zhong, J.; Lv, L.; Niu, D.;
Tang, J. CuSbS
2
as a Promising Earth-Abundant Photovoltaic Absorber Material: A
Combined Theoretical and Experimental Study. Chem. Mater. 2014, 26, 3135–3143.
(24) Xu, D.; Shen, S.; Zhang, Y.; Gu, H.; Wang, Q. Selective Synthesis of Ternary
Copper–Antimony Sulfide Nanocrystals. Inorg. Chem. 2013, 52, 12958–12962.
(25) Kyono, A.; Kimata, M. Crystal structures of chalcostibite (CuSbS
2
) and emplectite
(CuBiS
2
): Structural relationship of stereochemical activity between chalcostibite and
emplectite. Am. Mineral. 2015, 90, 162–165.
(26) Rath, T.; MacLachlan, A. J.; Brown, M. D.; Haque, S. A. Structural, optical and
charge generation properties of chalcostibite and tetrahedrite copper antimony sulfide thin
films prepared from metal xanthates. J. Mater. Chem. A 2015, 3, 24155–24162.
90
(27) A. W. Welch, A. W.; Zawadzki, P. P.; Lany, S.; Wolden, C. A.; Zakutayev, A. Self-
regulated growth and tunable properties of CuSbS
2
solar absorbers. Sol. Energy Mater. Sol.
Cells 2015, 132, 499–506.
(28) Peccerillo, E.; Major, J.; Phillips, L.; Treharne, R.; Whittles, T. J.; Dhanak, V.;
Halliday, D.; Durose, K. in Characterization of sulfurized CuSbS2 thin films for PV
applications. 2014 IEEE 40th Photovoltaic Specialist Conference (PVSC) IEEE, Denver, CO,
2014, pp. 0266–0269.
(29) Rietveld, H. M. Line profiles of neutron powder-diffraction peaks for structure
refinement. Acta Crystallogr. 1967, 22, 151.
(30) Rietveld, H. M.; A profile refinement method for nuclear and magnetic structures.
J. Appl. Crystallogr. 1969, 2, 65.
(31) Larson, A. C.; Von Dreele, R. B. General Structure Analysis System (GSAS); Los
Alamos National Laboratory Report LAUR 86-748; Los Alamos National Laboratory:
Los Alamos, NM, 2000.
(32) Thompson, P.; Cox, D. E.; Hastings, J. M. Rietveld refinement of Debye–Scherrer
synchrotron X-ray data from Al
2
O
3
. J. Appl. Crystallogr. 1987, 20, 79.
(33) Young, R. A. Ed. The Rietveld Method; IUCr Monographs on Crystallography
(Book 5); Oxford University Press: New York, 1993.
91
Chapter 4 Method for the Solution Deposition of Phase-Pure CoSe
2
as an
Efficient Hydrogen Evolution Reaction Electrocatalyst*
*Published in ACS Energy Lett. 2016, 1, 607–611.
4.1 Abstract
We demonstrate the ability of a thiol−amine solvent mixture to deposit phase-pure
marcasite-type CoSe
2
nanostructured thin films as effective hydrogen evolution reaction
(HER) electrocatalysts. Electrodes are readily prepared by spin coating a precursor ink
onto highly ordered pyrolytic graphite substrates followed by annealing to 350 °C. The
resulting CoSe
2
films have an onset potential for HER of −117 mV vs RHE and Tafel
slopes of ca. 60 mVdec
−1
. Normalization based on electrochemically active surface area
reveals that simple optimization of film thickness, based on the number of layers
deposited, leads to electrodes with better surface utilization. Based on the electrocatalytic
performance of the solution-processed CoSe
2
presented here (η
10mA/cm2
= −272 mV vs
RHE), this approach shows promise as a simple method to deposit a wide range of useful
dichalcogenide electrocatalysts.
4.1 Introduction
The discovery of cost-effective alternatives to noble metals for water splitting
processes (i.e., hydrogen evolution reaction (HER) and oxygen evolution reaction (OER))
has become an important area of research with the growing desire to achieve economical
and sustainable electro- and photocatalytic systems for alternative fuel production.
1, 2
For
92
example, recent investigation into transition-metal dichalcogenides (ME
2
, where M = Fe,
Co, Ni and E = S, Se) has revealed high catalytic activity toward HER, potentially a
result of the reduced coordination number of metal sites at the surfaces of the low-energy
planes that resemble the catalytic active site in hydrogenase metalloenzymes observed in
nature.
3, 4
Cobalt dichalcogenides in both the pyrite and marcasite phases have shown
particularly high HER activities; for example, Basu et al. have recently reported that the
marcasite phase of CoSe
2
and CoTe
2
both have high catalytic activities toward HER
under acidic conditions.
3, 5, 6
In addition to the discovery of new electrocatalysts, it is also necessary to develop
new methods for preparing these catalysts such that they can be robustly deposited on an
electrode surface with morphology and structure that are optimizable for maximum
utilization with respect to mass and surface area. Current methods for the preparation of
these dichalcogenide catalysts are energy intensive, including thermal
sulfurization/selenization of e-beam evaporated or sputtered transition-metal thin films
and high-pressure solvothermal preparation of bulk or nanostructured dichalcogenides
followed by subsequent deposition onto an electrode.
3-8
An alternate and less expensive
method for the preparation of such electrocatalysts would combine the synthesis and
deposition into a single-step process, ideally under ambient pressure, while permitting the
use of various electrode substrates and allowing for tunablilty with respect to catalyst
loading, composition, and micro- and nanostructuring.
9
93
Solution processing is an excellent candidate for such deposition of dichalcogenide
electrocatalysts; unfortunately, bulk dichalcogenides are insoluble in common solvents,
which complicates a simple “dissolve and recover” process. Recently, we developed a
method that has been extremely effective for solution processing many metal
chalcogenides relevant for optoelectronic applications. This method can be used to
deposit phase-pure chalcogenide thin films from inks made by the dissolution of bulk
metal chalcogenides, elemental sources, or metal oxides in an “alkahest” solvent
composed of ethylenediamine (en) and a short chain thiol.
10-17
To date, there have been
no reports using this method to produce chalcogenides or dichalcogenides that are
functionally competent as catalysts or electrocatalysts. Herein, we demonstrate the use of
the thiol–amine alkahest for the solution deposition of high-quality nanostructured
CoSe
2
thin films with excellent HER activity and stability when compared to other
deposition methods.
4.2 Results and Discussion
The alkahest precursor ink was prepared by the simultaneous dissolution of
Co(OH)
2
and elemental selenium powders in 1:4 vol/vol binary mixture of
mercaptoethanol (merc) and en (25 °C, 1 atm, 20 min). Upon filtering this solution with a
1 µm filter, an optically clear, green ink is produced (Figure 4.1 a). Thermogravimetric
analysis (TGA) was used to determine the ink concentration to be ca. 40 mM and an
appropriate annealing temperature for removal of the organic decomposition products.
The TGA trace of the ink predried to ca. 160 °C indicates that the decomposition end
94
point occurs by 350–400 °C (Figure 4.1 b). The complete loss of organic content is
corroborated by comparing Fourier transfrom infrared (FT-IR) spectra of the dried
precursor ink to that of the ink annealed to 300 °C (Figure 4.1 c). The spectrum of the
dried ink clearly possesses bands for ν(O–H), ν(N–H), and ν(C–H) stretches, while the
spectrum of the annealed ink is featureless from the expulsion and decomposition of
organics from the merc–en solvent system.
Figure 4.1. (a) Picture of the filtered CoSe
2
precursor ink in merc-en solvent system. (b) TGA
trace of the dried CoSe
2
precursor ink (5 ˚C min
-1
under flowing nitrogen). (c) FT-IR spectra of
the dried ink and the ink annealed to 300 ˚C. (d) Rietveld refinement of XRD patterns of CoSe
2
recovered from precursor ink after annealing to 350 ˚C. Experimental (×) and calculated (⎯⎯)
patterns are shown, along with the difference curve (⎯⎯) and tick marks (⏐⏐) corresponding to
the refined phase.
95
Annealing the ink to 350 °C under a nitrogen atmosphere resulted in recovery of
phase-pure marcasite-type CoSe
2
, as verified by the powder X-ray diffraction (XRD)
pattern given in Figure 4.1 d. The XRD pattern can be indexed to the orthorhombic
(Pnnm) marcasite-type phase of CoSe
2
(PDF 01-089-2003). When ethanedithiol (EDT) or
ethanethiol (EtSH) is used as the thiol component of the solvent mixture, phase-pure
marcasite-type CoSe
2
is similarly recovered upon annealing to 350 °C (Figure 4.2).
Figure 4.2. Powder XRD patterns for CoSe
2
recovered by annealing precursor inks to 350 ˚C in
nitrogen atmosphere using (a) merc, (b) EDT, and (c) EtSH.
Rietveld refinement analysis of the recovered CoSe
2
gave lattice parameters of a =
4.881(6) Å, b = 5.883(7) Å, and c = 3.655(4) Å and a Se–Se distance of 2.5116(33) Å,
which closely match values previously reported for marcasite-type CoSe
2
.
5
The refined
pattern compared to experimental data (including difference data) is shown in Figure 4.1
d. The crystallinity of CoSe
2
recovered using this method was further studied using
transmission electron microscopy (TEM) by annealing a dilute precursor ink deposited
onto a silicon nitride grid. TEM analysis reveals small nanoparticulate grains of
nonuniform shape, and a high-resolution TEM (Figure 4.3 a) of a single CoSe
2
particle
96
revealed lattice fringes with a d-spacing of 2.6 Å, corresponding to the (111) lattice
planes for orthorhombic CoSe
2
, and confirming the crystallinity of the resulting
nanostructures.
Figure 4.3. (a) High-resolution TEM image of the CoSe
2
showing lattice fringes corresponding to
the (111) planes of marcasite-type CoSe
2
, and (b) SEM micrograph of CoSe
2
recovered from the
merc-en solvent system.
X-ray photoelectron spectroscopy (XPS) was used to study the elemental
environment and composition of the recovered CoSe
2
(Figure 4.4). The Co 2p
3/2
and
2p
1/2
doublet at 778.0 and 792.9 eV, respectively, possesses a peak splitting of 14.9 eV,
which is indicative of the Co
2+
species in CoSe
2
.
18, 19
The Se 3d
5/2
and 3d
3/2
doublet at
53.6 and 54.2 eV, respectively, corresponds to Se
2
2–
in CoSe
2
.
18
Quantification of the
survey scan was used to measure a surface stoichiometry of CoSe
1.9
for the material
(Figure 4.4 a).
97
Figure 4.4. XPS spectra of recovered CoSe
2
. (a) survey scan; (b) high-resolution Co 2p; and (c)
high-resolution Se 3d region.
The bulk composition of the material was also verified by inductively coupled plasma
optical emission spectroscopy (ICP-OES) and energy dispersive X-ray spectroscopy
(EDX), which both revealed a near-stoichiometric composition of CoSe
2.2
. The
morphology of the recovered CoSe
2
was investigated using scanning electron microscopy
(SEM). A representative micrograph is given in Figure 4.3 b, which shows that the
CoSe
2
is composed of small nanostructured grains (ca. 75–100 nm). This morphology is
98
beneficial for catalysis as the nanosized grains provide a high surface area for catalysis to
occur (vide infra).
High-quality CoSe
2
thin films were deposited by spin coating the precursor ink onto a
substrate followed by annealing at 350 °C in a nitrogen-filled glovebox. In this way,
multiple layers of the material can be deposited without significant dissolution of
previous layers. This is confirmed by the nearly linear increase in film thickness whereby
4, 8, and 12 layer films yield thicknesses of 151, 258, and 391 nm, respectively, as
determined by cross-sectional SEM. Electrocatalytic HER activity studies were
conducted for thin films deposited on highly ordered pyrolytic graphite (HOPG)
electrode substrates using an electrolyte solution of 0.5 M H
2
SO
4
(aq). Figure 4.5 shows
the polarization curve for a film made by depositing four coats of ink.
Figure 4.5. Polarization curve of a 4-layer CoSe
2
film in 0.5 M H
2
SO
4
.
The established benchmarking overpotential required to produce a magnitude current
density of 10 mA cm
–2
(η
10mA/cm2
), corresponding to the current density expected for a
99
10% efficient solar-to-fuel conversion device under 1 sun illumination, is observed at
−327 mV vs RHE.
2
A Tafel plot generated from the polarization curve (Figure 4.6)
indicates that H
2
production has a low onset potential of just −117 mV vs RHE, and the
Tafel slope of 55 mV dec
–1
indicates fast kinetics for HER catalysis.
3
Figure 4.6. Tafel plot for a 4-layer sample where the Tafel slope is 55 mV dec
–1
.
The 4-layer CoSe
2
thin film’s onset potential of −117 mA cm
–2
and low Tafel slope
indicate that the CoSe
2
deposited from the precursor ink is a viable catalyst for HER
under acidic solutions; however, the η
10mA/cm2
value of −327 mV vs RHE is a bit high
when compared to other reports for this and other dichalcogenide materials (e.g.,
η
10mA/cm2
= −200 to −300 mV vs RHE).
4, 6
The source of high η
10mA/cm2
lies in the higher
overpotential regime, where a lower sloping polarization curve is indicative of diffusion-
limited current.
20
To directly study the effects of mass transport on surface and mass
utilization for HER catalysis, we compared the 4-layer thin films’ performance to 8- and
12-layer samples. We will refer to these as 4L, 8L, and 12L, respectively. Atomic force
100
micrographs of these films on HOPG reveal root-mean-square roughnesses of ca. 17–48
nm, and ICP-OES data reveals a linear increase in CoSe
2
loading with increasing number
of layers (Figure 4.7, Figure 4.8 and Table 4.1).
Figure 4.7. Atomic force micrographs of (a) blank HOPG substrate, (b) 4L, (c) 8L, and (d) 12L
samples on HOPG. The rms roughnesses are 101 nm, 48.2 nm, 17.2 nm, and 35.3 nm, for blank
HOPG, 4L, 8L, and 12L samples, respectively.
101
Figure 4.8. Plot of average CoSe
2
loading on HOPG substrates (measured by ICP-OES) and
average film thickness (measured by SEM) as functions of number of coats of precursor ink
revealing a linear increase of catalyst loading and film thickness with increased number of ink
coats.
Table 4.1. Pertinent electrochemical values for three different CoSe
2
thin film thicknesses.
Avg. Film
Thickness
(nm)
Avg.
CoSe
2
Loading
(µg cm
-2
)
η
onset
(mV)
a
η
10 mA/cm2
(mV)
a
Tafel
Slope
(mV
dec
–1
)
C
dl
(mF
cm
–2
)
ECSA
(cm
2
)
b
ECSA-
Corrected
η
10 mA/cm2
(mV)
a
I
o
(mA
cm
-2
)
4L 151 87 -117 -327 55 1.5 105.7 -212
4.03 x
10
-4
8L 258 165 -117 -287 62 4.8 342.9 -252
2.22 x
10
-3
12
L
391 240 -117 -272 61 6.8 482.9 -277
2.05 x
10
-3
a
Potentials are vs. RHE.
b
Using a general specific capacitance of 0.035 mF cm
–2
for all samples.
2
Polarization curves and Tafel plots for the three samples are presented in Figure 4.11
a,b. To study how surface utilization compares between the three samples, we used cyclic
102
voltammetry (CV) to measure the double-layer capacitance (C
dl
), which is directly related
to the electrochemically active surface area (ECSA) (Figure 4.9 and Figure 4.10).
Figure 4.9. Cyclic voltammograms at various scan rates used to derive double layer capacitance
values for (a) 4L, (b) 8L, and (c) 12L films.
Figure 4.10. Scan rate vs. current density difference at 0 mV vs. RHE from CV curves for 4L,
8L, and 12L films. Where C
dl
= slope/2 for y = 2.993x + 0.02366 (4L), y = 9.645x + 0.06266 (8L),
and y = 13.64x + 0.08923 (12L).
Table 4.1 gives a summary of the important electrochemical parameters derived for
each sample with respect to HER. While the onset potentials from the Tafel plots are the
same for all three samples, the values for η
10mA/cm2
vary greatly from −327 mV (4L) to
−287 mV (8L) to −272 mV (12L) as the higher overpotential regions of the polarization
103
curves “fan out” after the initial onset of HER activity. Additionally, ECSA increases as
the number of layers increases. These two trends in the data for η
10mA/cm2
and ECSA may
make sense considering that the 4L sample has the lowest loading of CoSe
2
(Table 4.1)
and therefore would require a higher η
10mA/cm2
and have a lower ECSA. However, when
the polarization curves are normalized with respect to ECSA (Figure 4.11 a), 4L actually
has the lowest η
10mA/cm2
, followed by 8L and 12L.
Figure 4.11. Electrochemical characterization of the catalytic activity toward HER. (a)
polarization curves and (b) Tafel plots for different CoSe
2
loadings. (c) 8 hour stability test
measuring driving potential required to produce a magnitude current density of 20 mA cm
-2
.
Differences in inherent activity of the recovered CoSe
2
between samples can be ruled
out as the cause for this observation as the exchange current densities (I
0
) for all samples
(extrapolated as the x-intercept of the Tafel line) are all nearly the same (Table 4.1).
Thus, a lower ECSA-corrected η
10mA/cm2
for lower CoSe
2
loading can be effectively
rationalized by considering a mass transport-dominated process at high overpotentials.
The 12L thin film has the greatest measured ECSA; however, this measurement is
derived from the low overpotential regime of the measured CV curves where the
observed current density is a measurement of charging current only. When higher
104
overpotentials are applied, resulting in Faradaic current density, the thicker sample
induces a mass transport constraint as solutes no longer have time to diffuse either to or
from the available catalytic surfaces located deeper within the film. This kinetically limits
the portion of the electrode that is able to perform HER, and as a result, the higher
overpotential regime is operating over an ECSA smaller than that measured by CV
experiments. Therefore, thicker samples have poorer surface utilization because the bulk
of the film is not being used for HER.
Figure 4.11 c shows a stability test for the 12L sample in which the overpotential
required to produce a current density magnitude of 20 mA cm
–2
(η
20mA/cm2
) was measured
over the course of 8 h. During stability testing, polarization curves were recorded at 1 h
intervals verifying that curve shape was retained and that no new features were observed,
thus verifying that the catalytic process was not changing over time (Figure 4.12).
Although the η
20mA/cm
2
steadily drops over the course of the first 5 h, this can be attributed
to a rise in pH as protons were used to make H
2
(g). A pH of 0.3 was measured at t = 0
corresponding to the fresh 0.5 M H
2
SO
4
electrolyte solution, while at t = 5 h the
measured pH was 0.5. Therefore, after 5 h, the electrolyte solution was replaced with
fresh 0.5 M H
2
SO
4
, resulting in η
20mA/cm2
returning to the t = 0 value, indicating that the
CoSe
2
material has suitable stability toward HER under acidic solutions.
105
Figure 4.12. Polarization curves taken at 1 h intervals during extended stability tests.
The I
0
values (ca. 10
–3
mA cm
–2
, Table 4.1) observed for the CoSe
2
films presented
here are higher than have been previously reported for thin-film CoSe
2
HER catalysts
(e.g., ca. 10
–5
mA cm
–2
);
4
the Tafel slopes (ca. 60 mV dec
–1
) are relatively low (previous
reports for ME
2
catalysts, including CoSe
2
range between ca. 40–70 mV dec
–1
);
4
and
Faradaic efficiencies for H
2
production were determined to be >99% for all samples. This
demonstrates impressive intrinsic HER activity and kinetics for the solution-deposited
CoSe
2
films.
4
Furthermore, as exemplified by the thinner 4L sample presented here,
better catalyst utilization is observed when CoSe
2
is deposited as thinner films such that
mass transport does not dominate at high current densities. Fortunately, because the
concentration of the ink and solution deposition conditions can be easily tuned using this
method, deposition of CoSe
2
in thin layers on various substrate morphologies should be
easily accessible with simple engineering optimization; this is expected to produce
solution-deposited CoSe
2
with optimal electrocatalytic activity toward HER on a range of
cost-effective electrode substrates.
106
4.3 Experimental
4.3.1 General considerations
All reagents and solvents were used as received. 1,2-Ethylenediamine (en, 99.5%), 2-
mercaptoethanol (Merc, 99.0%), ethanethiol (EtSH, 97%), cobalt (II) chloride (97%), and
cobalt standard solution for ICP-OES (TraceCERT®, 1000 mg L
-1
Co in 2% HNO
3
) were
all purchased from Sigma Aldrich. Sodium hydroxide (98%) was purchased from Macron
Chemicals. 1,2-ethanedithiol (EDT, 98+%) and selenium powder (grey Se, -200 mesh
99.999%) were purchased from Alfa Aesar. The selenium ICP-OES standard solution
(1000 mg L
-1
Se in 2% HNO
3
) was purchased from Perkin Elmer.
4.3.2 Experimental Methods
Co(OH)
2
: 41.9 mg (0.323 mmol) of CoCl
2
and 244.2 mg (6.105 mmol, excess) of
NaOH were dissolved separately in deionized water (30 mL and 10 mL, respectively).
The NaOH solution was added to the CoCl
2
solution and the mixture was allowed to stir
in an ice bath for 30 min as blue-green Co(OH)
2
precipitated out of solution. Solid
Co(OH)
2
was filtered off and washed with cold DI water and allowed to dry in air
overnight before being ground to a fine powder. CoSe
2
precursor ink: 20.3 mg (0.218
mmol) of Co(OH)
2
, and 34.4 mg (0.436 mmol) of Se were added to a vial followed by
1.2 mL of ethylenediamine and 0.3 mL of thiol. The solution was stirred and bath
sonicated until all solids were dissolved over the course of 20 min at room temperature,
yielding a solution that upon filtering with a 1 µm PTFE filter yielded an optically clear
107
ink. Recovery of CoSe
2
: Powder. The filtered precursor ink was drop cast onto
microscope slides, annealed to 350 ˚C, and then allowed to cool to room temperature,
under flowing nitrogen. The resulting CoSe
2
powder was then scraped off the microscope
slides using a razor blade. Thin Films. In a nitrogen glove box, the precursor ink was
filtered directly onto a cleaned HOPG substrate (ca. 1 cm x 4 cm) and spin coated at 2000
r.p.m. for 1 min using a Laurell Technologies Corporation WS400Ez-6NPPLITE single-
wafer spin processor. In between the multiple coats, films were annealed immediately at
350 ˚C on a temperature controlled hot plate for 1 min.
4.3.3 Microscopy:
Scanning electron microscopy (SEM) was performed on a JEOL JSM-7001F
scanning-electron microscope with an operating voltage of 10 kV. SEM was used to
measure film thicknesses by imaging the cross section of 4, 8, and 12 layer films on glass
substrates. Transmission electron microscopy (TEM) analysis was performed on a JEOL
JEM-2100 microscope at an operating voltage of 200 kV, equipped with a Gatan Orius
CCD camera. Samples for TEM analysis were prepared by diluting a standard CoSe
2
precursor ink with en, followed by deposition on 10 nm silicon nitride grids (Ted Pella,
Inc.), followed by annealing at 400 ˚C. Atomic Force Microscopy (AFM) analysis was
performed on an Agilent AFM 5400 in tapping mode at a scan rate of either 0.30 or 0.51
lines per second.
108
4.3.4 Recovered CoSe
2
Composition:
Thermal Gravimetric Analysis (TGA) was performed on a TA Instruments TGA Q50
instrument in an alumina crucible under a flowing nitrogen atmosphere with a heating
rate of 5 ˚C min
–1
. TGA samples were pre-dried at 160 ˚C under flowing nitrogen prior to
TGA analysis. FT-IR spectra were measured on a Bruker Vertex 80. Samples were
prepared by drop casting the filtered CoSe
2
ink onto a ZnSe window and either drying at
room temperature or annealing to 300 ˚C under flowing nitrogen. Scanning electron
microscope-energy dispersive X-ray spectroscopy (SEM-EDX) was used for elemental
analysis on a JEOL JSM-7001F scanning-electron microscope with an operating voltage
of 25 kV and a working distance of 15 mm. Inductively coupled plasma optical emission
spectroscopy ICP-OES) was performed on a Thermo Scientific iCAP 7000 series ICP-
OES. Samples were prepared by digesting ~4.5 mg of the recovered CoSe
2
powder in 2
mL of conc. HNO
3
followed by filtering through a 0.45 µm filter. This solution was then
diluted to 25 mL with 2% by vol. HNO
3
. Each sample was run in triplicate, the presented
results are an average of the three measurements. The loading of CoSe
2
on HOPG
substrates for 4, 8, and 12 layers of ink was also measured by ICP-OES. Samples were
prepared by soaking each substrate in conc. HNO
3
for 45 minutes to digest the deposited
CoSe
2
, followed by dilution to 25 mL. Each sample was run in triplicate, the presented
results are an average of the three measurements. The total mass of cobalt observed by
ICP-OES measurements was converted to mass of CoSe
2
averaged over the surface area
of one side of the HOPG substrate (5 cm
2
) to obtain the reported average CoSe
2
loading
for each sample.
109
4.3.5 Structural and Optical Characterization:
Powder X-ray Diffraction (XRD) was collected with a Rigaku Ultima IV
diffractometer in parallel beam geometry (2 mm or 10 mm beam width) using Cu Kα
radiation (λ = 1.54 Å). Rietveld Structural Refinements
24, 25
were carried out using the
General Structure Analysis System (GSAS) software.
26
The following parameters were
refined: (1) scale factor, (2) background, which was modeled using a shifted Chebyshev
polynomial function, (3) sample displacement, (4) peak shape, which was modeled using
a modified Thomson−Cox−Hasting pseudo-Voigt,
27
(5) lattice constants, (6) fractional
atomic coordinates, and (7) an isotropic thermal parameter for each chemical species (i.e.,
U
Cu
, U
Sb
, and U
S
). The usual R
wp
and χ
2
indicators were employed to assess the quality of
the refined structural models.
28
X-ray Photoelectron Spectroscopy (XPS) was performed
using a Kratos Axis Ultra X-ray photoelectron spectrometer with the analyzer lens in
hybrid mode. A monochromatic aluminum anode with an operating current of 6 mA and
voltage of 10 kV was used with a step size of 0.1 eV, a pass energy of 20 eV, and a
pressure range between 1-3 × 10
–8
Torr. The binding energy was referenced to the C 1s
core level at 284.0 eV.
4.3.6 Electrochemical Measurements
Electrochemical measurements were carried out using a Pine potentiostat in a
nitrogen filled Vacuum Atmospheres glovebox. All measurements were performed in 0.5
110
M H
2
SO
4
(aq) (prepared using 18.2 MΩ·cm resistivity water from a Millipore Synergy
system) using a two-compartment, three-electrode cell. The two compartments were
separated by a fine porosity glass frit. It is important to note that the electrochemical data
presented in this study were not iR corrected. Highly ordered pyrolytic graphite (HOPG)
electrodes (5 cm × 1 cm × 0.3 cm) were used as the working and auxiliary electrodes.
The CoSe
2
catalyst covered only one side of the HOPG working electrode and the
geometric surface area used to derive current densities corresponded to the portion of the
CoSe
2
side that was submerged in the electrolyte solution. The reference electrode was a
Ag/AgCl/saturated KCl(aq) electrode separated from the solution by a Vycor tip.
Polarization data were obtained at a sweep rate of 2 mV s
–1
while rapidly stirring the
solution with a magnetic stir bar. Double layer capacitance (C
dl
) was measured for each
electrode from 55 mV to -145 mV vs. SHE in the two-compartment, three-electrode cell
at scan rates of 5, 10, 25, 50, 100, and 200 mV s
–1
. The preparation of Pt/C Electrodes
was achieved by following a reported literature procedure.
6
20 mg of Pt/C (20% Platinum
on Vulcan XC 72 from Fuel Cell Store) and Nafion solution (10 µL 5 wt%) were
dispersed in 1 mL of 1:1 v/v water/ethanol solvent mixture by 30 minutes of sonication to
form the catalyst ink. The ink (150 µL) was dropcasted on an HOPG electrode covering a
surface area of 1.5 cm
2
. The short-term stability was measured by monitoring the
overpotential to maintain 20 mA cm
–2
of activity over time. After 5 h of controlled
current electrolysis, the pH of the solution in the working compartment of the cell was
measured using a Mettler Toledo pH meter. The pH was recorded as 0.55 as opposed to
111
the initial pH 0.3. Following 5 h, the solutions in the two compartment cell were replaced
with fresh 0.5 M H
2
SO
4
. Quantitative hydrogen yield measurements were performed in
0.5 M H
2
SO
4
(aq) using a two-compartment cell separated by a fine porosity frit. A
current corresponding to 20 mA cm
–2
was passed for ~15-30 min. Using a gas-tight
syringe, 10 mL of gas were withdrawn from the headspace of the working compartment
of the cell and injected into a gas chromatography instrument (Shimadzu GC-2010-Plus)
equipped with a BID detector and a Restek ShinCarbon ST Micropacked column. To
determine the Faradaic efficiency, the theoretical H
2
amount based on total charge flowed
is compared with the GC-detected H
2
produced from controlled-current electrolysis.
4.4 Conclusions
In summary, we have presented a straightforward method to solution deposit a
CoSe
2
HER catalyst under relatively low temperature and ambient pressure conditions
using a thiol–amine solvent system. The resulting films of marcasite-type CoSe
2
are
phase pure and free of organic residues after annealing to 350 °C. The resulting thin films
possess a nanostructured morphology with ECSAs in the range of 105–480
cm
2
depending on the number of layers deposited. We have shown that while more
deposited layers can yield a relatively low η
10mA/cm2
of −272 mV, it ultimately results in
poor surface utilization due to mass transport-dominated HER. Therefore, by using a
thinner film by depositing fewer layers, we demonstrated that surface utilization of the
electrocatalyst is improved, thus demonstrating how this solution processing method
allows for common issues such as mass transport limitations to be easily tuned.
112
Conceivably, even lower driving potentials for this CoSe
2
electrocatalyst may be
achievable by depositing the ink onto porous electrodes, allowing for more efficient
utilization of mass and surface. As has been shown, this thiol–amine solvent system is
incredibly versatile and can be used to solution process many metal chalcogenides.
21-
23
This processing versatility suggests that this method has the capability to deposit other
catalysts or to tune the CoSe
2
catalyst by alloying or doping as a more general approach
for the practical synthesis and deposition of dichalcogenide electrocatalysts.
4.5 References
(1) McKone, J. R.; Marinescu, S. C.; Brunschwig, B. S.; Winkler, J. R.; Gray, H. B.
Earth- Abundant Hydrogen Evolution Electrocatalysts. Chem. Sci. 2014, 5, 865–878.
(2) McCrory, C. C. L.; Jung, S.; Ferrer, I. M.; Chatman, S. M.; Peters, J. C.;
Jaramillo, T. F. Benchmarking Hydrogen Evolving Reaction and Oxygen Evolving
Reaction Electrocatalysts for Solar Water Splitting Devices. J. Am. Chem. Soc. 2015,
137, 4347– 4357.
(3) Faber, M. S.; Lukowski, M. A.; Ding, Q.; Kaiser, N. S.; Jin, S. Earth-Abundant
Metal Pyrites (FeS
2
, CoS
2
, NiS
2
, and Their Alloys) for Highly Efficient Hydrogen
Evolution and Polysulfide Reduction Electrocatalysis. J. Phys. Chem. C 2014, 118,
21347–21356.
(4) Kong, D.; Cha, J. J.; Wang, H.; Lee, H. R.; Cui, Y. First-Row Transition Metal
Dichalcogenide Catalysts for Hydrogen Evolution Reaction. Energy Environ. Sci. 2013,
6, 3553–3558.
(5) Liang, L.; Cheng, H.; Lei, F.; Han, J.; Gao, S.; Wang, C.; Sun, Y.; Qamar, S.;
Wei, S.; Xie, Y. Metallic Single-Unit-Cell Orthorhombic Cobalt Diselenide Atomic
Layers: Robust Water-Electrolysis Catalysts. Angew. Chem. 2015, 127, 12172–12176.
(6) Basu, M.; Zhang, Z.-W.; Chen, C.-J.; Chen, P.-T.; Yang, K.-C.; Ma, C.-G.; Lin,
C. C.; Hu, S.-F.; Liu, R.-S. Heterostructure of Si and CoSe
2
: A Promising Photocathode
113
Based on a Non-Noble Metal Catalyst for Photoelectrochemical Hydrogen Evolution.
Angew. Chem. Int. Ed. 2015, 54, 6211–6216.
(7) Lu, T.-H.; Chen, C.-J.; Basu, M.; Ma, C.-G.; Liu, R.-S. The CoTe
2
Nanostructure: An Efficient and Robust Catalyst for Hydrogen Evolution. Chem.
Commun. 2015, 51, 17012– 17015.
(8) Shi, J.-B.; Wu, P.-F.; Kao, C.-T.; Lee, M.-W.; Chan, C.-C.; Yang, P.-C.; Lin, C.-
L.; Huang, R.-Y.; Huang, Y.-J.; Lin, S.-K.; et al. Influence of Selenization Temperature
on Synthesis, Morphology, and Properties of Nanostructured CoSe
2
Films. Cryst. Res.
Technol. 2015, 50, 155–159.
(9) Tan, S. M.; Pumera, M. Bottom-up Electrosynthesis of Highly Active Tungsten
Sulfide (WS
3–x
) Films for Hydrogen Evolution. ACS Appl. Mater. Interfaces 2016, 8,
3948– 3957.
(10) Webber, D. H.; Brutchey, R. L. Alkahest for V2VI3 Chalcogenides: Dissolution
of Nine Bulk Semiconductors in a Diamine-Dithiol Solvent Mixture. J. Am. Chem. Soc.
2013, 135, 15722–15725.
(11) Webber, D. H.; Buckley, J. J.; Antunez, P. D.; Brutchey, R. L. Facile Dissolution
of Selenium and Tellurium in a Thiol–amine Solvent Mixture under Ambient Conditions.
Chem. Sci. 2014, 5, 2498–2502.
(12) Buckley, J. J.; Greaney, M. J.; Brutchey, R. L. Ligand Exchange of Colloidal
CdSe Nanocrystals with Stibanates Derived from Sb2S3 Dissolved in a Thiol-Amine
Mixture. Chem. Mater. 2014, 26, 6311–6317.
(13) Antunez, P. D.; Torelli, D. A.; Yang, F.; Rabuffetti, F. A.; Lewis, N. S.; Brutchey,
R. L. Low Temperature Solution-Phase Deposition of SnS Thin Films. Chem. Mater.
2014, 26, 5444–5446.
(14) McCarthy, C. L.; Webber, D. H.; Schueller, E. C.; Brutchey, R. L. Solution-Phase
Conversion of Bulk Metal Oxides to Metal Chalcogenides Using a Simple Thiol–Amine
Solvent Mixture. Angew. Chem. Int. Ed. 2015, 54, 8378–8381.
(15) Buckley, J. J.; McCarthy, C. L.; Del Pilar-Albaladejo, J.; Rasul, G.; Brutchey, R.
L. Dissolution of Sn, SnO, and SnS in a Thiol–Amine Solvent Mixture: Insights into the
Identity of the Molecular Solutes for Solution-Processed SnS. Inorg. Chem. 2016, 55,
3175–3180.
114
(16) Zhao, D.; Tian, Q.; Zhou, Z.; Wang, G.; Meng, Y.; Kou, D.; Zhou, W.; Pan, D.;
Wu, S. Solution-Deposited Pure Selenide CIGSe Solar Cells from Elemental Cu, In, Ga,
and Se. J. Mater. Chem. A 2015, 3, 19263–19267.
(17) McCarthy, C. L.; Cottingham, P.; Abuyen, K.; Schueller, E. C.; Culver, S.
P.; Brutchey, R. L. Earth Abundant CuSbS
2
Thin Films Solution Processed from Thiol-
Amine Mixtures J. Mater. Chem. C 2016, 4, 6230– 6233.
(18) Wang, K.; Xi, D.; Zhou, C.; Shi, Z.; Xia, H.; Liu, G.; Qiao, G. CoSe
2
Necklace-
like Nanowires Supported by Carbon Fiber Paper: A 3D Integrated Electrode for the
Hydrogen Evolution Reaction. J. Mater. Chem. A 2015, 3, 9415–9420.
(19) Li, Z.; Xue, H.; Wang, J.; Tang, Y.; Lee, C.-S.; Yang, S. Reduced Graphene
Oxide/Marcasite-Type Cobalt Selenide Nanocrystals as an Anode for Lithium-Ion
Batteries with Excellent Cyclic Performance. ChemElectroChem 2015, 2, 1682–1686.
(20) Zheng, J.; Yan, Y.; Xu, B. Correcting the Hydrogen Diffusion Limitation in
Rotating Disk Electrode Measurements of Hydrogen Evolution Reaction Kinetics. J.
Electrochem. Soc. 2015, 162, F1470–F1481.
(21) Arnou, P.; Cooper, C. S.; Malkov, A. V.; Bowers, J. W.; Walls, J. M. Solution-
Processed CuIn(S,Se)
2
Absorber Layers for Application in Thin Film Solar Cells. Thin
Solid Films 2015, 582, 31–34.
(22) Arnou, P.; van Hest, M. F. A. M.; Cooper, C. S.; Malkov, A. V.; Walls, J. M.;
Bowers, J. W. Hydrazine-Free Solution-Deposited CuIn(S,Se)
2
Solar Cells by Spray
Deposition of Metal Chalcogenides. ACS Appl. Mater. Interfaces 2016, 8, 11893–11897.
(23) Zhang, R.; Szczepaniak, S. M.; Carter, N. J.; Handwerker, C. A.; Agrawal, R. A
Versatile Solution Route to Efficient Cu
2
ZnSn(S,Se)
4
Thin-Film Solar Cells. Chem.
Mater. 2015, 27, 2114–2120.
(24) Rietveld, H. M. Line profiles of neutron powder-diffraction peaks for structure
refinement. Acta Crystallogr, 1967, 22, 151.
(25) Rietveld, H. M.; A profile refinement method for nuclear and magnetic structures.
J. Appl. Crystallogr. 1969, 2, 65.
(26) A. C.Larson, R. B. Von Dreele, General Structure Analysis System (GSAS); Los
Alamos National Laboratory Report LAUR 86-748; Los Alamos National Laboratory:
Los Alamos, NM, 2000.
115
(27) Thompson, P.; Cox, D. E.; Hastings, J. M. Rietveld refinement of Debye–Scherrer
synchrotron X-ray data from Al
2
O
3
. J. Appl. Crystallogr. 1987, 20, 79.
(28) R. A. Young, Ed. The Rietveld Method; IUCr Monographs on Crystallography
(Book 5); Oxford University Press: New York, 1993.
116
Chapter 5. Room Temperature Dissolution of Bulk Elemental Ni and Se for
Solution Deposition of a NiSe
2
HER Electrocatalyst *
*Published in Inorg. Chem. 2017, 56, 10143-10146.
5.1 Abstract
With hydrogen fuel becoming a more viable alternative to fossil fuels comes the need
for inexpensive, low-energy hydrogen production. Here, a low-temperature direct
solution-processing method is presented for the deposition of Earth-abundant pyrite-type
NiSe
2
as an efficient hydrogen evolution reaction (HER) catalyst. Thin films of phase-
pure NiSe
2
are deposited from a precursor ink prepared by room-temperature dissolution
of bulk elemental Ni and Se in a binary thiolamine solvent mixture. The nanostructured
NiSe
2
thin films demonstrate high HER catalytic activity with 100% Faradaic efficiency.
5.2 Introduction
As we work toward a more sustainable global society, innovations in clean energy
and fuel production are paramount. Fuels that produce little or no polluting byproducts
when consumed are of specific importance as we aim to minimize levels of pollution
resulting from the vehicles we drive. Hydrogen fuel cells are the prime example of clean
fuel utilization as they result in electricity generation with water and heat as the sole
byproducts (i.e., 2 H
2
+ O
2
à 2 H
2
O). Several vehicles utilizing fuel cells that are on the
market now (e.g., Toyota Miri, Hyundai Tucson Fuel Cell, and Honda Clarity) are able to
offer ranges (412-589 km per tank) and fueling times (3-5 min) comparable to gasoline
117
powered cars – achievements that affordable battery-powered electric vehicles are still
working toward (several hours to charge, and ranges of 95-539 km).
1–3
With the
technology to utilize this fuel in place, key factors governing the practicality of mass-
adoption of these vehicles will be competitive fuel pricing (e.g. <4 USD/kg H
2
) and
availability/infrastructure.
4
Toward this end, Earth-abundant catalysts with high activity and efficiency toward
HER are being developed as alternatives to the traditional catalysts based on noble
metals, such as platinum, which are highly effective but are limiting for large scale use
due to elemental scarcity and high cost.
5–7
Transition metal dichalcogenides (ME
2
, where
M = Fe, Co, Ni and E = S, Se) have been studied as alternative, Earth-abundant HER
catalysts, and demonstrate high activities and Faradaic efficiencies.
5,8–11
For example,
recent reports on NiSe
2
have shown impressive onset potentials of ca. 70-130 mV vs.
RHE.
12
In addition to potential cost savings achievable by using more Earth-abundant
catalysts, the monetary and energetic cost of producing hydrogen fuel can be further de-
creased by developing inexpensive synthetic methods to produce the catalysts
themselves. For example, savings can be realized by replacing energy and capital
intensive processing methods, like physical vapor deposition, with simple solution
processing methods that are based on cheaper equipment and require less energy to
operate. We developed such a solution processing method based on the dissolution of
bulk inorganic materials using a binary solvent mixture of ethylenediamine (en) and a
118
short chain thiol to produce molecular inks that can be subsequently deposited and
recovered as high-quality thin films.
13–20
In particular for HER catalysis, we
demonstrated the dissolution of Co(OH)
2
and Se to form a precursor ink from which
phase-pure nanostructured marcasite-type CoSe
2
can be recovered.
21
Herein, we
demonstrate how this method can be used to prepare an ink from bulk elemental Ni and
Se, which can be used to prepare highly efficient HER catalyst electrodes.
5.3 Results and Discussion
Elemental Ni and grey Se were dissolved in a 4:1 (vol/vol) ratio of
en/mercaptoethanol (merc) over the course of 48 h at 25 ˚C, resulting in a murky rust-
colored solution (Figure 5.1), which upon filtering yielded a clear rusty orange colored
ink (Figure 5.2 a).
Figure 5.1. Image of unfiltered Ni and Se dissolved in the en/merc solvent mixture.
119
Figure 5.2. (a) Filtered ink prepared from the dissolution of bulk Ni and Se in a mixture of 4:1
(vol/vol) en/merc. (b) TGA trace of the filtered ink. (c) XRD pattern of NiSe
2
recovered from
decomposition of the en/merc ink. (d) FT-IR spectra of the dried ink (blue) and after annealing to
325 ˚C (red) under a N
2
atmosphere.
Thermogravimetric analysis (TGA, Figure 5.2 b) was used to identify that the end of
ink decomposition occurs by 350 ̊C. Annealing the filtered ink to 350 ̊C under a nitrogen
atmosphere resulted in the recovery of phase-pure NiSe
2
, as confirmed by powder X-ray
diffraction (XRD, Figure 5.2 c). Rietveld analysis of the experimental XRD data (Figure
5.3 and Table 5.1) confirms that the structure is adequately de- scribed by the cubic Pa 3
structure, with a lattice parameter of a = 5.9380(4) Å, which is in good agreement with
previous structural reports on cubic pyrite-type NiSe
2
(e.g., a = 5.9576(2) Å and a =
5.965 Å).
22,23
120
Figure 5.3. Rietveld refinement of XRD pattern of NiSe
2
recovered from the precursor ink after
annealing to 350 °C. Experimental (×) and calculated (red) patterns are shown, along with the
difference curve (blue) and tick marks (green) corresponding to the refined phase.
Table 5.1. Rietveld analysis of X-ray diffraction data of NiSe
2
.
CuSbS
2
a (Å)
6.018(3) Cu – S(1) × 2 2.330(5)
b (Å)
3.794(2) Cu – S(2) 2.345(13)
c (Å)
14.490(6) Cu – S(2) 2.235(9)
V (Å
3
)
330.9(4) Sb – S(1) 2.427(9)
X
Cu
0.7494(13) Sb – S(1) × 2 3.088(6)
X
Sb
0.2297(5) Sb – S(2) × 2 2.609(6)
X
S(1)
0.6260(13)
X
S(2)
0.1384(17) S(1) – Cu – S(1) 109.0(4)
Z
Cu
0.1720(4) S(1) – Cu – S(2) × 2 110.2(3)
Z
Sb
0.0632(3) S(1) – Cu – S(2) × 2 111.6(3)
121
Z
S(1)
0.0939(6) S(2) – Cu – S(2) 104.3(4)
Z
S(2)
0.1808(7) Cu – S(1) – Cu 109.0(4)
U
Cu
(Å
2
)
a
1.61 Cu – S(2) – Cu 110.5(5)
U
Sb
(Å
2
)
a
0.37
U
S(1)
(Å
2
)
a
1.72 R
wp
(%) 13.2
U
S(2)
(Å
2
)
a
2.87 χ
2
1.79
a
Atomic
displacement
parameters
are
given
as
100
×
Uiso
Fourier trans- form infrared (FT-IR) spectroscopy was used to verify that this
annealing process results in NiSe
2
with no detectable organic species left behind from the
solvent system. The FT-IR spectrum in blue in Figure 1d shows bands corresponding to
the organic content in an ink dried to just 100 ̊C, while the featureless spectrum in red is
for a sample that was annealed to 325 ̊C and indicates the complete loss of organic
content.
Figure 5.4. SEM-EDS spectrum of NiSe
2
recovered from en/merc ink.
122
The composition of the recovered NiSe
2
was analyzed by scanning electron
microscopy – energy dispersive X-ray spectroscopy (SEM-EDS, Figure 5.4). A
stoichiometry of NiSe
2.06
was found, with ca. 1 at% sulfur remaining from merc in the
solvent system. X-ray photoelectron spectroscopy (XPS) was used to study the surface
composition and elemental environment of the recovered NiSe
2
. Figure 5.6 gives the
high-resolution spectra in the Ni 2p and Se 3d binding energy regions. The binding
energies for Ni 2p
3/2
and 2p
1/2
at 852.3 eV and 869.6 eV (with satellite peaks at 858.6 eV
and 875.3 eV), respectively, have an energy splitting of 17.3 eV. These values are
consistent with what has been previously reported for Ni
2+
in NiSe
2
.
9
Notably, as the Ni
2p peaks have no shoulders, we can rule out the presence of NiO, which would give
multiplet-split Ni 2p signals.
24–26
The position of the Se 3d
5/2
(54.5 eV) and 3d
3/2
(55.2
eV) peaks are also consistent with previous reports on NiSe
2
for Se
2
2–
in the pyrite
structure.
27
The lack of signal around 59-60 eV indicates that there are no oxidized
selenium species (i.e., SeO
x
) present.
Figure 5.5. XPS survey scan of recovered NiSe
2
.
123
Figure 5.6. High-resolution XPS spectra in the (a) Ni 2p and (b) Se 3d regions.
To analyze the catalytic performance of the recovered NiSe
2
, thin films were prepared
in a nitrogen-filled glove box by spin coating multiple layers of the filtered ink on highly
ordered pyrolytic graphite (HOPG) substrates with subsequent annealing to 350 ˚C in
between layers. Linear sweep voltammetry (LSV) in an acidic electrolyte solution of 0.5
M H
2
SO
4
(aq) was used to study the electrocatalytic HER activity of the NiSe
2
. Figure 5.7
a shows the LSV trace of a NiSe
2
film, made by depositing 8 layers of ink, as compared
to a Pt/C reference, with the horizontal dotted line indicating the benchmarking current
density of 10 mA cm
–2
. The NiSe
2
requires an overpotential of 269 mV to reach this
benchmark (i.e., η
10 mA/cm2
= 269 mV), which is comparable to recent reports for NiSe
2
and other transition metal dichalcogenides under similar conditions (e.g., 117 – 272
mV).
5,8,9,12,21,27–29
A Tafel plot was generated from the LSV data (Figure 5.7 b) to reveal
an onset potential of just η
0
= 130 mV, a low Tafel slope of 35 mV dec
–1
, and an
exchange current density of I
0
= 3.8 x 10
–6
mA cm
–2
, which are all comparable to the
values reported for the same materials (e.g., η
0
in the range of ca. 70 – 130 mV, Tafel
124
slopes of ca. 31 – 61 mV dec
–1
, and I
0
on the order of 10
–5
– 10
–3
mA cm
–2
).
5,8,9,12,21,27–30
Additionally, the NiSe
2
operates at an excellent Faradaic efficiency of 100% for HER as
measured by H
2
detection by gas chromatography of the head gas produced. These data
suggest that the inherent catalytic activity of the NiSe
2
recovered directly from the
precursor ink is on par with similar materials prepared by other methods that require
multiple steps between catalyst synthesis and electrode construction. These methods
include electron-beam evaporation, photoinduced cation exchange or etching of
previously synthesized nanoparticles, etching of metal foam, and hydrothermal reactions,
all of which require a subsequent high temperature (450-600 ˚C) sulfurization or
selenization step.
5,8,9,12,21,27–30
In some of these cases, a further step is required to deposit
the synthesized catalyst onto an electrode substrate.
Figure 5.7. (a) LSV traces for NiSe
2
and Pt/C electrodes on HOPG substrates. (b) Tafel plot for
an 8-layered NiSe
2
electrode on HOPG substrate. (c) SEM micrograph of recovered NiSe
2
showing nanostructured morphology.
The recovered NiSe
2
, like several of the previously reported transition metal
dichalcogenide HER electrocatalysts, has a nanostructured morphology. Figure 5.7 c
125
shows an SEM micrograph revealing grain sizes on the order of 200 - 700 nm. In order to
study the implications of this morphology on HER electrocatalysis, the electrocatalytic
properties of samples prepared with different loadings of NiSe
2
(4 and 6 layers of ink)
were investigated and compared to the 8-layered sample discussed above. Polarization
curves were measured for each of the samples and used to generate Tafel plots (Figure
5.10 b), from which relevant electrochemical values were extracted. All three samples (4,
6, and 8 layers) have similar η
0
, I
0
, and Tafel slopes (i.e., ca. 130 mV, ca. 10
-6
mA cm
-2
,
and 35-40 mV dec
-1
, Table 5.2), indicating similar intrinsic HER catalytic activity
comparable to previous reports for transition metal dichalcogenides. The η
10 mA/cm2
values
decrease with increasing NiSe
2
loading, which is expected since increasing the amount of
catalyst provides more surface active sites for performing catalysis, resulting in greater
current generation. To further probe this observation, double layer capacitance (C
dl
)
measurements were used as a proxy to compare the electrochemically active surface area
(ECSA) of the electrodes.
12,31,32
The current response in the potential window 0.1-0.2 V
vs. RHE, which should be due only to double-layer charging and discharging, was
measured at different scan rates (Figure 5.8).
Figure 5.8. Cyclic voltammograms at various scan rates used to derive C
dl
values for 4-, 6-, and
8-layered NiSe
2
electrodes.
126
From the generated plots of non-Faradaic current density vs. scan rate (Figure 5.9),
we extracted C
dl
values of 0.12, 0.23, and 0.30 mF cm
–2
for the 4, 6, and 8 layer samples,
respectively (Table 5.2), revealing a trend of increased ECSA with increased NiSe
2
loading, corroborating the η
10 mA/cm2
trend.
Figure 5.9. Plot of non-Faradaic current density vs. voltage scan rate extracted from the CV plots
in Figure 5.8. C
dl
= slope/2 for y = 0.2402*x + 0.004956 (4 layers), y = 0.4691*x + 0.00514 (6
layers), y = 0.608*x + 0.006348 (8 layers).
Table 5.2. Key values for electrocatalytic activity for various loadings of NiSe
2
on HOPG.
η
10 mA/cm2
(mV)
Tafel
Slope
(mV
dec
-1
)
η
0
(mV)
I
0
(mA/cm
2
)
C
dl
(mF/cm
2
)
ECSA
Corrected
η
10 mA/cm2
(mV)
R
ct
(η =
0.25 V)
(Ω)
R
ct
(η =
0.35 V)
(Ω)
R
ct
(η = 0.45
V) (Ω)
4 L 328 40 127 8.9 x 10
-6
0.12 257 --
3.622
1.741
6 L 277 34 135 1.4 x 10
-6
0.23 258 59.67 3.406 1.315
8 L 269 35 130 3.8 x 10
-6
0.3 -- 30.34 2.633 1.183
127
To further compare surface utilization of the three samples during catalysis, we
corrected the 4 and 6 layer LSV curves based on their relative ECSA compared to the 8
layer sample with the resulting data plotted as the dotted traces in Figure 5.10 a. With the
correction, all three samples have similar η
10 mA/cm2
values, however at higher
overpotentials (η > 270 mV), the samples with lower loadings generate higher ECSA-
corrected current densities. This is characteristic of the onset of a diffusion-limited
regime resulting in decreased surface utilization at higher overpotentials for thicker films
of NiSe
2
.
33
Figure 5.10. HER electrocatalysis characterization of electrodes made with various loadings (4L
= 4 layers, 6L = 6 layers, and 8L = 8 layers) of NiSe
2
on HOPG: (a) polarization curves of as-
measured (solid) samples and ECSA-corrected (dotted) data, and (b) Tafel plots with linear
regions extrapolated to show log(I
0
) at the x-intercept for each of the three loadings.
We used electrochemical impedance spectroscopy (EIS) to further study the kinetics
of the three samples (Nyquist plots, Figure 5.11) from which we observed a general trend
of decreased charge transfer resistance (R
ct
) with increased NiSe
2
loading at each applied
128
potential. Interestingly, as the applied potential is increased, the relative difference in R
ct
values for the different loadings decreases (i.e., 30- 60 Ω at η = 250mV compared to 1.2-
1.7 Ω at η = 450 mV, Tabel S2). This is likely due to diffusion limitations at higher
overpotentials for the thicker 8 layer sample, in agreement with the ECSA-corrected LSV
data. R
ct
decreases with increased potential, as expected when in- creasing the external
driving force for the HER reaction.
Figure 5.11. EIS Nyquist plots (fitted data solid lines) for various loadings of NiSe
2
at (a) η =
0.25 V and (b) η = 0.35 V.
In addition to the solution resistance (R
s
) and R
ct
components, the equivalent circuit
model used to fit the EIS data (Figure 5.12) includes an overpotential-independent
component (R
1
-CPE
1
) that has previously been associated with the porosity of the
electrode, or the contact between the electrode surface and the catalyst layer.
34–37
This
component contributes minimal resistance to the systems (0.2-0.5 Ω) and does not appear
to trend with catalyst loading.
129
Figure 5.12. Equivalent circuit model used to fit the EIS data. The two-time constant serial model
(2TS) was employed where R
s
is the solution resistance and R
ct
-CPE
2
is related to the charge
transfer reaction. R
1
-CPE
1
is unrelated to HER kinetics and remains relatively unchanged with
overpotential. R
1
-CPE
1
has been associated with the porosity of the electrode or the contact
between the electrode and the catalyst layer.
38–41
5.4 Experimental
5.4.1 General Considerations
All reagents and solvents were used as received. 1,2- ethylenediamine (en, 99.5%), 2-
mercaptoethanol (merc, 99.0%), and nickel powder (-100 mesh 99%) were all purchased
from Sigma Aldrich. Selenium powder (grey Se, -200 mesh 99.999%) was purchased
from Alfa Aesar. Ethylenediamine and mercaptoethanol are classified as flammable,
toxic and/or corrosive; therefore care needs to be taken when working with the solvent
mixtures.
5.4.2 Ink Preparation and NiSe
2
Recovery
To prepare the precursor ink, 0.8 mL en and 0.2 mL merc were added to elemental Ni
(20 mg) and Se (26.9 mg). After 48 h of stirring at 25 ˚C, the solution was a murky rust
color, and some of the Ni was remained on the stir bar. The solution was filtered with a 1
µm syringe filter before deposition. Powder samples were prepared by drop casting the
130
filtered ink onto a glass slide followed by annealing to ca. 350 ˚C and cooling to room
temperature under flowing nitrogen. The resulting NiSe
2
powder was scraped off the
slide using a razor blade for further characterization. Thin films were prepared in a
nitrogen-filled glove box by filtering the ink directly onto a HOPG substrate (ca. 1 cm x 4
cm) followed by spin coating using a Laurell Technologies Corporation
WS400Ez6NPPLITE single-wafer spin processor. Eight coats of ink were deposited with
immediate annealing at 350 ˚C on a hot plate between coats.
5.4.3 Material Characterization
Scanning electron microscopy (SEM) was performed on a JEOL JSM-7001F
scanning electron microscope with an operating voltage of 7 kV. Thermal gravimetric
analysis (TGA) was performed on a TA Instruments TGA Q50 instrument in an alumina
crucible under a flowing nitrogen atmosphere with a heating rate of 5 ˚C min
–1
. Samples
were pre-dried to 155 ˚C under flowing nitrogen prior to TGA analysis. Fourier transform
infrared (FT-IR) spectra were measured on a Bruker Vertex 80. Samples were prepared
by drop casting the filtered NiSe
2
ink onto a ZnSe window and either drying to 100 ˚C or
annealing to 325 ˚C under flowing nitrogen. SEM-energy dispersive X-ray spectroscopy
(SEM-EDS) was used for elemental analysis on a JEOL JSM-7001F scanning-electron
microscope with an operating voltage of 10 kV and a working distance of 15 mm.
Powder X-ray Diffraction (XRD) was collected with a Rigaku Ultima IV diffractometer
in S2 parallel beam geometry (10 mm beam width) using Cu Kα radiation (λ = 1.54 Å).
Rietveld structural refinements
42, 43
were carried out using the General Structure Analysis
131
System (GSAS) software.
44
The following parameters were refined: (1) scale factor, (2)
background, which was modeled using a shifted Chebyshev polynomial function, (3)
polarization fraction, (4) peak shape, which was modeled using a modified Finger-
CoxJephcoat Pseudo-Voigt function,
45
(5) lattice constants, (6) fractional atomic
coordinates, and (7) an isotropic thermal parameter for each chemical species (i.e., UNi,
and USe). The usual Rwp and χ
2
indicators were employed to assess the quality of the
refined structural models.
46
X-ray photoelectron spectroscopy (XPS) was performed
using a Kratos Axis Ultra X-ray photoelectron spectrometer with the analyzer lens in
hybrid mode. A monochromatic aluminum anode with an operating current of 6 mA and
voltage of 10 kV was used with a step size of 0.1 eV, a pass energy of 40 eV, and a
pressure range between 1-3 × 10
–8
Torr. The binding energy was referenced to the C 1s
core level at 284.8 eV.
5.4.4 Electrochemical Measurements
Electrochemistry experiments were carried out using a VersaSTAT 3 potentiostat. All
measurements were performed in 0.5 M H
2
SO
4
(aq) (prepared using 18.2 MΩ·cm
resistivity water from a Millipore Synergy system) using a two-compartment,
threeelectrode cell. Highly ordered pyrolytic graphite (HOPG) electrodes (5 cm × 1 cm ×
0.3 cm) were used as the working and auxiliary electrodes. The NiSe
2
catalyst covered
only one side of the HOPG working electrode and the geometric surface area used to
derive current densities corresponded to the portion of the NiSe
2
side that was submerged
in the electrolyte solution. The two compartments were separated by a fine porosity glass
132
frit. The reference electrode, placed in a separate compartment and connected by a Vycor
tip, was based on an aqueous Ag/AgCl/saturated KCl electrode. The reference electrode
in aqueous media was calibrated externally relative to ferrocenecarboxylic acid (Fc-
COOH) at pH 7.0, with the Fe
3+/2+
couple at 0.28 V vs Ag/AgCl. All potentials reported
in this paper were converted to the reversible hydrogen electrode (RHE) by adding a
value of (0.205 + 0.059 × pH) V. Polarization data were obtained at a sweep rate of 2 mV
s
–1
while rapidly stirring the solution with a magnetic stir bar. The preparation of Pt/C
electrodes was achieved by following a reported literature procedure.
47
20 mg of Pt/C
(20% Pt on Vulcan XC 72 from Fuel Cell Store) and Nafion solution (10 µL 5 wt%) were
dispersed in 1 mL of 1:1 vol/vol water/ethanol solvent mixture by 30 min of bath
sonication to form the catalyst ink. The ink (150 µL) was dropcasted on an HOPG
electrode covering a surface area of 1.5 cm
2
.
Cyclic voltammograms for double layer capacitance measurements were taken in a
potential window between 0.1 and 0.2 V vs RHE in 0.5 M H
2
SO
4
at scan rates of 20, 40,
60, 80, 100, and 150 mV/s. The total current density obtained from the current density
difference (Δj = ja – jc) at 0.15 V vs RHE was plotted against the scan rate. The slope is
twice the value of the double layer capacitance. For Tafel analysis, polarization curves
were measured in 0.5 M H2SO4 solution with a scan rate of 2 mV s
–1
.
Electrochemical impedance spectroscopy (EIS) measurements were carried out at
different overpotentials (η = 0.25-0.45 V) in the frequency range of 200 kHz – 0.1 Hz
with 10 mV sinusoidal perturbations in 0.5 M H
2
SO
4
solutions. Experimental EIS data
were analyzed and fitted with the ZSimpWin software.
133
The obtained polarization curves were corrected by the iR loss according to the
following equation:
E
corr
= E
mea
– iR
s
Where E
corr
is iR-corrected potential, E
mea
is the experimentally measured potential, and
R
s
is the solution resistance extracted from the fitted EIS data.
Quantitative hydrogen yield measurements were performed in 0.5 M H
2
SO
4
(aq)
using a two-compartment cell separated by a fine porosity frit. A current corresponding to
20 mA cm
–2
was passed for ~15-30 min. Using a gas-tight syringe, 10 mL of gas was
withdrawn from the headspace of the working compartment of the cell and injected into a
gas chromatography instrument (Shimadzu GC-2010-Plus) equipped with a BID detector
and a Restek ShinCarbon ST Micropacked column. To determine the Faradaic efficiency,
the theoretical H
2
amount based on total charge flowed is compared with the GC-detected
H
2
produced from controlled-current electrolysis.
5.5 Conclusions
Bulk elemental Ni and Se were readily dissolved in an en/merc solvent mixture that
was solution processed to yield phase-pure NiSe
2
films that facilitated HER catalysis with
100% Faradaic efficiency and low η
10 mA/cm2
values. The onset potential, exchange current
density, Tafel slope, and specific HER electrocatalytic activity of the films proved to be
comparable to reports for similar transition metal dichalcogenide electrocatalysts, which
re- quire multiple synthetic steps utilizing more energy intensive conditions. This work
demonstrates how the thiol-amine solvent system is a valuable tool in the develop- ment
134
of inexpensive solution processing methods for a variety of alternative Earth-abundant
catalysts. Furthermore, such innovations in materials processing bring powerful
contributions toward achieving low-cost hydrogen fuel for fuel-cell vehicles and other
applications.
5.6 References
(1) U.S. Department of Energy; Office of Energy Efficiency and Renewable Energy;
U.S. Environmental Prodection Agency. Model Year 2017 Fuel Economy Guide; 2017.
(2) Nonobe, Y. Development of the Fuel Cell Vehicle Mirai. IEEJ Trans. Electr.
Electron. Eng. 2017, 12, 5–9.
(3) Thomas, C. E. Sustainable Transportation Options for the 21st Century and
beyond: A Comprehensive Comparison of Alternatives to the Internal Combustion
Engine; Springer International Publishing: Switzerland, 2015.
(4) U.S. Department of Energy. Fuel Cell Technologies Office: Hydrogen Production
https://energy.gov/eere/fuelcells/hydrogenproduction (accessed Apr 26, 2017).
(5) Faber, M. S.; Lukowski, M. A.; Ding, Q.; Kaiser, N. S.; Jin, S. Earth-Abundant
Metal Pyrites (FeS
2
, CoS
2
, NiS
2
, and Their Alloys) for Highly Efficient Hydrogen
Evolution and Polysulfide Reduction Electrocatalysis. J. Phys. Chem. C 2014, 118,
21347–21356.
(6) Zeng, M.; Li, Y. Recent Advances in Heterogeneous Electrocatalysts for the
Hydrogen Evolution Reaction. J. Mater. Chem. A 2015, 3, 14942–14962.
(7) Faber, M. S.; Jin, S. Earth-Abundant Inorganic Electrocatalysts and Their
Nanostructures for Energy Conversion Applications. Energy Environ. Sci. 2014, 7, 3519–
3542.
(8) Kong, D.; Cha, J. J.; Wang, H.; Lee, H. R.; Cui, Y. First-Row Transition Metal
Dichalcogenide Catalysts for Hydrogen Evolution Reaction. Energy Environ. Sci. 2013,
6, 3553–3558.
(9) Kwak, I. H.; Im, H. S.; Jang, D. M.; Kim, Y. W.; Park, K.; Lim, Y. R.; Cha, E. H.;
Park, J. CoSe
2
and NiSe
2
Nanocrystals as Superior Bifunctional Catalysts for
135
Electrochemical and Photoelectrochemical Water Splitting. ACS Appl. Mater. Interfaces
2016, 8, 5327–5334.
(10) Zhuo, J.; Cabán-Acevedo, M.; Liang, H.; Samad, L.; Ding, Q.; Fu, Y.; Li, M.; Jin,
S. High-Performance Electrocatalysis for Hydrogen Evolution Reaction Using Se-Doped
Pyrite-Phase Nickel Diphosphide Nanostructures. ACS Catal. 2015, 5, 6355–6361.
(11) Zhang, H.; Ding, Q.; He, D.; Liu, H.; Liu, W.; Li, Z.; Yang, B.; Zhang, X.; Lei,
L.; Jin, S. A P-Si/NiCoSe
x
Core/Shell Nanopillar Array Photocathode for Enhanced
Photoelectrochemical Hydrogen Production. Energy Environ. Sci. 2016, 9, 3113–3119.
(12) Wang, F.; Li, Y.; Shifa, T. A.; Liu, K.; Wang, F.; Wang, Z.; Xu, P.; Wang, Q.;
He, J. Selenium-Enriched Nickel Selenide Nanosheets as a Robust Electrocatalyst for
Hydrogen Generation. Angew. Chem. Int. Ed. 2016, 55, 6919–6924.
(13) Webber, D. H.; Brutchey, R. L. Alkahest for V
2
VI
3
Chalcogenides: Dissolution
of Nine Bulk Semiconductors in a Diamine-Dithiol Solvent Mixture. J. Am. Chem. Soc.
2013, 135, 15722–15725.
(14) Webber, D. H.; Buckley, J. J.; Antunez, P. D.; Brutchey, R. L. Facile Dissolution
of Selenium and Tellurium in a Thiol–amine Solvent Mixture under Ambient Conditions.
Chem. Sci. 2014, 5, 2498– 2502.
(15) Antunez, P. D.; Torelli, D. A.; Yang, F.; Rabuffetti, F. A.; Lew- is, N. S.;
Brutchey, R. L. Low Temperature Solution-Phase Deposition of SnS Thin Films. Chem.
Mater. 2014, 26, 5444–5446.
(16) Buckley, J. J.; Greaney, M. J.; Brutchey, R. L. Ligand Exchange of Colloidal
CdSe Nanocrystals with Stibanates Derived from Sb
2
S
3
Dissolved in a Thiol-Amine
Mixture. Chem. Mater. 2014, 26, 6311– 6317.
(17) McCarthy, C. L.; Webber, D. H.; Schueller, E. C.; Brutchey, R. L. Solution-Phase
Conversion of Bulk Metal Oxides to Metal Chalcogenides Using a Simple Thiol–Amine
Solvent Mixture. Angew. Chem. Int. Ed. 2015, 54, 8378–8381.
(18) McCarthy, C. L.; Cottingham, P.; Abuyen, K.; Schueller, E. C.; Culver, S. P.;
Brutchey, R. L. Earth Abundant CuSbS
2
Thin Films Solution Processed from Thiol–
amine Mixtures. J. Mater. Chem. C 2016, 4, 6230–6233.
(19) Buckley, J. J.; McCarthy, C. L.; Del Pilar-Albaladejo, J.; Rasul, G.; Brutchey, R.
L. Dissolution of Sn, SnO, and SnS in a Thiol–Amine Solvent Mixture: Insights into the
Identity of the Molecular Solutes for Solution-Processed SnS. Inorg. Chem. 2016, 55,
3175–3180.
136
(20) McCarthy, C. L.; Brutchey, R. L. Solution Processing of Chalcogenide Materials
Using Thiol–amine “alkahest” Solvent Systems. Chem. Commun. 2017, 53, 4888–4902.
(21) McCarthy, C. L.; Downes, C. A.; Schueller, E. C.; Abuyen, K.; Brutchey, R. L.
Method for the Solution Deposition of Phase-Pure CoSe
2
as an Efficient Hydrogen
Evolution Reaction Electrocatalyst. ACS Energy Lett. 2016, 1, 607–611.
(22) Yuan, B.; Luan, W.; Tu, S. One-Step Solvothermal Synthesis of Nickel Selenide
Series: Composition and Morphology Control. CrystEngComm 2012, 14, 2145–2151.
(23) Zheng, L.J.; Zhang, B.-P.; Han, C.G.; Pei, J.; Chen, N. Mechanical Alloying-
Spark Plasma Sintering Synthesis and Thermoelectric Properties of N-Type NiSe
2+x
Semiconductors: Analysis of Intrinsic Defects and Phase Structures. J. Mater. Sci. Mater.
Electron. 2016, 27, 8363–8369.
(24) Moulder, J. F.; Stickle, W. F.; Sobol, P. E.; Bomben, K. D. Handbook of X-Ray
Photoelectron Spectroscopy; Chastain, J., Ed.; Physical Electronics Division, Perkin-
Elmer Corp.: Eden Prairie, Minnesota, 1979.
(25) Grosvenor, A. P.; Biesinger, M. C.; Smart, R. S. C.; McIntyre, N. S. New
Interpretations of XPS Spectra of Nickel Metal and Oxides. Surf. Sci. 2006, 600, 1771–
1779.
(26) Bonomo, M.; Marrani, A. G.; Novelli, V.; Awais, M.; Dowling, D. P.; Vos, J. G.;
Dini, D. Surface Properties of Nanostructured NiO Undergoing Electrochemical
Oxidation in 3-Methoxy-Propionitrile. Appl. Surf. Sci. 2017, 403, 441–447.
(27) Liang, J.; Yang, Y.; Zhang, J.; Wu, J.; Dong, P.; Yuan, J.; Zhang, G.; Lou, J.
Metal Diselenide Nanoparticles as Highly Active and Stable Electrocatalysts for the
Hydrogen Evolution Reaction. Nanoscale 2015, 7, 14813–14816.
(28) Kong, D.; Wang, H.; Lu, Z.; Cui, Y. CoSe
2
Nanoparticles Grown on Carbon Fiber
Paper: An Efficient and Stable Electrocatalyst for Hydrogen Evolution Reaction. J. Am.
Chem. Soc. 2014, 136, 4897– 4900.
(29) Liang, H.; Li, L.; Meng, F.; Dang, L.; Zhuo, J.; Forticaux, A.; Wang, Z.; Jin, S.
Porous Two-Dimensional Nanosheets Converted from Layered Double Hydroxides and
Their Applications in Electrocatalytic Water Splitting. Chem. Mater. 2015, 27, 5702–
5711.
(30) Zhou, H.; Yu, F.; Liu, Y.; Sun, J.; Zhu, Z.; He, R.; Bao, J.; Goddard, W. A.; Chen,
S.; Ren, Z. Outstanding Hydrogen Evolution Reaction Catalyzed by Porous Nickel
Diselenide Electrocatalysts. Energy Environ. Sci. 2017, 10, 1487–1497.
137
(31) Faber, M. S.; Dziedzic, R.; Lukowski, M. A.; Kaiser, N. S.; Ding, Q.; Jin, S.
High-Performance Electrocatalysis Using Metallic Cobalt Pyrite (CoS2) Micro- and
Nanostructures. J. Am. Chem. Soc. 2014, 136, 10053–10061.
(32) McCrory, C. C. L.; Jung, S.; Ferrer, I. M.; Chatman, S. M.; Pe- ters, J. C.;
Jaramillo, T. F. Benchmarking Hydrogen Evolving Reaction and Oxygen Evolving
Reaction Electrocatalysts for Solar Water Splitting Devices. J. Am. Chem. Soc. 2015,
137, 4347–4357.
(33) Zheng, J.; Yan, Y.; Xu, B. Correcting the Hydrogen Diffusion Limitation in
Rotating Disk Electrode Measurements of Hydrogen Evolution Reaction Kinetics. J.
Electrochem. Soc. 2015, 162, F1470– F1481.
(34) Kucernak, A. R. J.; Sundaram, V. N. N. Nickel Phosphide: The Effect of
Phosphorus Content on Hydrogen Evolution Activity and Corrosion Resistance in Acidic
Medium. J. Mater. Chem. A 2014, 2, 17435–17445.
(35) Navarro-Flores, E.; Chong, Z.; Omanovic, S. Characterization of Ni, NiMo, NiW
and NiFe Electroactive Coatings as Electrocatalysts for Hydrogen Evolution in an Acidic
Medium. J. Mol. Catal. Chem. 2005, 226, 179–197.
(36) Wang, X.; Kolen’ko, Y. V.; Bao, X. Q.; Kovnir, K.; Liu, L. One-Step Synthesis of
Self-Supported Nickel Phosphide Nanosheet Array Cathodes for Efficient
Electrocatalytic Hydrogen Generation. Angew. Chem. Int. Ed. 2015, 54, 8188–8192.
(37) Vrubel, H.; Moehl, T.; Grätzel, M.; Hu, X. Revealing and Accelerating Slow
Electron Transport in Amorphous Molybdenum Sulphide Particles for Hydrogen
Evolution Reaction. Chem. Commun. 2013, 49, 8985–8987.
(38) Kucernak, A. R. J.; Sundaram, V. N. N. Nickel Phosphide: The Effect of
Phosphorus Content on Hydrogen Evolution Activity and Corrosion Resistance in Acidic
Medium. J. Mater. Chem. A 2014, 2, 17435–17445.
(39) Navarro-Flores, E.; Chong, Z.; Omanovic, S. Characterization of Ni, NiMo, NiW
and NiFe Electroactive Coatings as Electrocatalysts for Hydrogen Evolution in an Acidic
Medium. J. Mol. Catal. Chem. 2005, 226, 179–197.
(40) Wang, X.; Kolen’ko, Y. V.; Bao, X.-Q.; Kovnir, K.; Liu, L. One-Step Synthesis
of Self-Supported Nickel Phosphide Nanosheet Array Cathodes for Efficient
Electrocatalytic Hydrogen Generation. Angew. Chem. Int. Ed. 2015, 54, 8188–8192.
138
(41) Vrubel, H.; Moehl, T.; Grätzel, M.; Hu, X. Revealing and Accelerating Slow
Electron Transport in Amorphous Molybdenum Sulphide Particles for Hydrogen
Evolution Reaction. Chem. Commun. 2013, 49, 8985–8987.
(42) Rietveld, H. M. Line Profiles of Neutron Powder-Diffraction Peaks for Structure
Refinement. Acta Crystallogr. 1967, 22, 151–152.
(43) Rietveld, H. M. A Profile Refinement Method for Nuclear and Magnetic
Structures. J. Appl. Crystallogr. 1969, 2, 65–71.
(44) Larson, A. C.; Von Dreele, R. B. General Structure Analysis System (GSAS); Los
Alamos National Laboratory Report LAUR 86-784; Los Alamos National Laboratory:
Los Alamos, NM, 1994.
(45) Thompson, P.; Cox, D. E.; Hastings, J. B. Rietveld Refinement of Debye–
Scherrer Synchrotron X-Ray Data from Al
2
O
3
. J. Appl. Crystallogr. 1987, 20, 79–83.
(46) The Rietveld Method; Young, R. A., Ed.; IUCr Monographs on Crystallography
(Book 5); Oxford University Press: New York, 1993.
(47) Tang, C.; Cheng, N.; Pu, Z.; Xing, W.; Sun, X. NiSe Nanowire Film Supported on
Nickel Foam: An Efficient and Stable 3D Bifunctional Electrode for Full Water Splitting.
Angew. Chem. Int. Ed. 2015, 54, 9351–9355.
139
Chapter 6. Solution Deposited Cu
2
BaSnS
4-x
Se
x
from a Thiol-Amine Solvent
Mixture*
*Published in Chem. Mater., 2018, 30, pp 304–308.
6.1 Abstract
The discovery that cation antisite defects in Cu
2
ZnSn(S,Se)
4
(CZTS) thin film
photovoltaic absorbers are a major contributor to lower than expected device efficiencies
has spurred recent interest in material design to develop similar materials that maintain
the desirable optoelectronic properties of CZTS, while mitigating this source of loss.
Recent reports have shown that this can be accomplished by replacing zinc with a larger
cation of the same valency, such as barium, to give Cu
2
BaSn(S,Se)
4
. Until this point,
Cu
2
BaSn(S,Se)
4
has only been synthesized by solid-state and physical deposition
methods. Herein, we use a thiol-amine solvent mixture to dissolve bulk Cu
2
S, BaS, and
SnO to prepare a precursor ink that upon low temperature annealing yields phase-pure
trigonal Cu
2
BaSnS
4-x
Se
x
(x = 0, 1, 2, 3) alloys with tunable direct band gaps from 1.56 –
1.86 eV.
6.2 Introduction
The design of functional Earth-abundant materials has become an important
component in efforts to develop more sustainable solutions to global issues, especially in
the energy sector where scalable technologies such as photovoltaic (PV) cells, batteries,
and fuel cell catalysts are so important for tackling the challenges associated with global
climate change. In this vein, chalcogenide materials have been studied for a wide gamut
140
of applications in optoelectronics, theromoelectics, and catalysis.
1–6
Specifically, the
multinary semiconductor chalcogenide Cu
2
ZnSn(S,Se)
4
(CZTS) and its derivatives have
shown strong promise as thin film PV absorber materials with suitable direct band gaps
(1.0-1.8 eV) and high absorption coefficients in the visible range (~10
5
cm
–1
).
1
Unfortunately, devices employing CZTS absorber materials suffer from depressed open
circuit potentials (V
OC
) stemming from cation anitsite defects (i.e., Cu
Zn
and Zn
Cu
), which
dominate as a result of their low formation energies, a consequence of the similarities in
size and valence of Zn
2+
and Cu
+
.
7
Recently, several groups have worked on designing new materials that maintain the
beneficial optoelectronic properties of CZTS, while improving upon the antisite defect
challenges.
8–18
As a result of this effort, Cu
2
BaSnS
4-x
Se
x
has been studied on the premise
that replacing Zn
2+
with a much larger cation, such as Ba
2+
, would render the anitsite
defects energetically unfavorable. Several groups have performed theoretical calculations
on the material yielding results that support this hypothesis, and there have been multiple
experimental reports on the preparation of this material using solid-state and physical
deposition methods.
9–12,14–16
Rather than the kesterite structure that persists through all
S/Se stoichiometreis of CZTS, Cu
2
BaSnS
4-x
Se
x
takes on a trigonal phase with space
group P3
1
for x = 0 – 3 and an orthorhombic phase with Ama2 symmetry for values of x >
3.
10,14
Band gap values for the pure sulfide material have been reported to range from
1.95 – 2.05 eV , while solid solutions up to x = 3 give rise to a range of band gaps from
1.55 – 2.05 eV .
10,12,14–16,19
PV devices using this material as the absorber layer have
141
shown a drastic increase in efficiency from 1.6% to over 5% in less than two years, this
material possesses promise for thin film devices with continued device engineering and
optimization.
10,18
While the development of Earth-abundant materials addresses the environmental and
financial costs of acquiring raw materials, the costs associated with processing crude
materials into useful form factors still needs to be addressed. Dissolution of bulk
materials to form molecular inks that can be solution deposited offers the excellent
advantages of low capital and energetic investment. Additionally, if the desired crystalline
material is recoverable upon low-temperature annealing, this continues to reduce the
energetic and financial cost of production, in addition to opening the range of useable
substrates to include plastics. Unfortunately, dissolution of bulk inorganic materials is
very difficult or impossible using common solvents, and thus direct solution processing
of bulk materials has not been widely used in industrial applications. Recently, however,
our group developed the “alkahest” method, which utilizes a binary mixture of a short
chain thiol and ethylenediamine (en) as a solvent system that is capable of dissolving
numerous bulk materials.
20–28
More importantly, phase-pure chalcogenide materials can
be recovered from these inks upon mild annealing, making solution processing of thin
films for applications such as photovoltaics possible. Since its initial report, the alkahest
method has been shown to dissolve over 65 bulk materials including bulk chalcogens,
transition metal chalcogenides, and transition metal oxides.
29
This method has been used
to solution process the two most promising multinary chalcogenide absorbers for thin
142
film PV , CZTS and Cu(In,Ga)(S,Se)
2
, yielding devices with efficiencies up to 8.0% and
12.2%, respectively.
30,31
Herein, we demonstrate that Cu
2
BaSnS
4
may be solution
deposited using the alkahest method, thus demonstrating the feasibility of its solution
deposition.
6.3 Results and Discussion
Bulk Cu
2
S, BaS, and SnO (1:1:1 mol/mol/mol) were used as the precursors to
formulate the Cu
2
BaSnS
4
precursor ink in a 1:4 (vol/vol) ethanedithiol (EDT)/en solvent
mixture. All components were added to a nitrogen-purged vial and were stirred at 50 ˚C
for 11 d to yield a viscous yellow solution; full dissolution was confirmed by gently
heating the vial with a heat gun to give a free-flowing, clear yellow solution. BaS appears
to be the slowest precursor to dissolve as inks prepared in less time yielded Cu
2
BaSnS
4
with a BaS impurity phase by powder X-ray diffraction (XRD, Figure 6.1).
Figure 6.1. Powder XRD pattern of material recovered from a precursor ink that was only
allowed to stir for 3 d at 50 ˚C; the BaS impurity is likely due to incomplete dissolution of the
bulk BaS precursor.
143
To facilitate the decomposition and volatilization of the ink, and subsequent recovery
of the target crystalline phase, the ink was annealed under a nitrogen atmosphere to 350
˚C. The thermogravimetric analysis (TGA) trace in Figure 6.2 a indicates the end of
organic decomposition by ca. 315 ˚C, while the concomitant loss of FT-IR bands, when
increasing the temperature from 100 ˚C to 350 ˚C, indicates that little organic content is
left behind in the material (Figure 6.2 b). The strongest FTIR peaks in the 100 ˚C
spectrum correspond to signals expected from EDT (e.g., 2970 cm
-1
ν
C-H
stretch, 2930
cm
-1
ν
C-H
stretch, 1430 cm
-1
ν
CH2
bend), and there are no signals above 3000 cm
-1
where
ν
NH2
is expected, indicating that upon mild heating the EDT remains intact while the en
either begins decomposing or volatilizes.
32–34
The lower intensity peaks are thus perhaps
coming from a small amount of en (e.g., ν
C-N
stretching 1100-1200 cm
-1
) and organic
decomposition products.
34
In agreement with previous reports that have shown evidence
of deprotonation of EDT in similar alkahest solvent mixtures, there is no peak near 2570
cm
-1
where the ν
S-H
stretch is expected.
21,32,33
However, the ν
C-S
stretch can be seen at 675
cm
-1
shifted to lower wavenumbers compared to what is expected for protonated EDT
(i.e., ca. 699 cm
-1
).
Figure 6.2. Ink decomposition as observed by (a) TGA trace of Cu
2
BaSnS
4
precursor ink and (b)
FT-IR spectra for Cu
2
BaSnS
4
inks dried to 100 ˚C and annealed to 350 ˚C.
144
After cooling the annealed ink to room temperature, a dark maroon colored powder
was collected. The recovered material was confirmed to be phase-pure Cu
2
BaSnS
4
by
XRD (Cu Kα radiation, λ = 1.5406 Å), as shown in Figure 6.3. Lattice parameters of a =
6.40 Å and c = 15.95 Å, and a unit cell volume V = 566.03 Å
3
, were determined from the
d-spacing of the (100) and (103) reflections at 15.97˚ and 23.15˚, respectively. These unit
cell parameters are similar to previously reported experimental values for Cu
2
BaSnS
4
(e.g., a = 6.366 Å and c = 15.828 Å).
10
Figure 6.3. Powder XRD pattern of phase-pure Cu
2
BaSnS
4
recovered from an EDT/en ink by
annealing at 350 ˚C.
The optical band gap of the recovered Cu
2
BaSnS
4
material was measured by diffuse
reflectance UV-vis spectroscopy. A Tauc plot derived from this data is presented in
Figure 6.4 with the linear portion extrapolated to reveal a direct band gap of E
g, dir.
= 1.86
eV , which is on the lower end of the range of theoretically calculated (ca. 1.7 eV) and
experimentally determined band gaps reported for this material (vida supra).
9,10
145
Figure 6.4. Tauc plot generated from the diffuse reflectance UV-vis spectrum for Cu
2
BaSnS
4
recovered from EDT/en ink at 350 ˚C.
X-ray photoelectron spectroscopy (XPS) was used to gain insight into the valence
states of the material. Figure 6.5 shows the high-resolution spectra for the Cu 2p, Ba 3d,
Sn 3d, and S 2p regions, including peak fittings.
Figure 6.5. High-resolution XPS data for the recovered Cu
2
BaSnS
4
. (a) Cu 2p region, (b) Ba 3d
region, (c) Sn 3d region, and (d) S 2p region. The open circles correspond to the collected data,
the green traces correspond to peaks fitted and assigned to the bulk Cu
2
BaSnS
4
, the red traces are
assigned to surface oxide impurities, and the black traces are the envelopes resulting from the
fitted peaks.
146
Table 6.1. Fitted peak positions and splitting values corresponding to the high-resolution XPS
spectra in Figure 6.5.
Element
Peak
Splitting (eV)
Peak ID
Binding
Energy
(eV)
Cu 19.7
2p
1/2
951.17
2p
3/2
931.38
Ba 15.3
3d
3/2
794.92
3d
5/2
779.62
Sn 8.4
3d
3/2
494.12
3d
5/2
485.72
Sn (SnO
x
) 8.4
3d
3/2
494.62
3d
5/2
486.22
S 1.2
2p
1/2
162.59
2p
3/2
161.41
S (SO
x
2–
) 1.2
2p
1/2
164.90
2p
3/2
163.72
Table 6.1 gives the fitted peak positions and peak splitting values for each high-
resolution spectrum. The Cu 2p and Ba 3d regions can each be fit by a single doublet set
of peaks (951.17 eV and 931.38 eV for Cu 2p, and 794.92 eV and 779.62 eV for Ba 3d),
indicating a single environment for both Cu and Ba in the material. The splitting and peak
positions observed are similar to Ge and Yan’s report for Cu
2
BaSnS
4
and are corroborated
by the expected values for Cu
+
and Ba
2+
.
16,35,36
Additionally, the lack of strong satellite
peaks in the Cu 2p region rules out the likelihood of significant Cu
2+
species being
present. The Sn 3d region can be fit with two sets of doublets each with a splitting of 8.4
eV . This observation of two different Sn environments has been previously observed by
XPS in Cu
2
BaSnS
4
that was prepared by co-sputtering vacuum deposition, and was
attributed to SnO
x
species on the surface of the material. In the previous report, the SnO
x
layer could be removed by cleaning the surface with argon ions to yield a single doublet
147
signal indicating a single Sn environment in the bulk of the material.
16
Since the
recovered Cu
2
BaSnS
4
samples in our case were not kept in an inert environment, the
observation of surface SnO
x
species, not observable by XRD, is also reasonable. This is
corroborated by the presence of oxygen signals in the XPS survey scan of the material
(Figure 6.6). In the S 2p region, we observe a complex set of peaks that can be
deconvoluted into two sets of doublets, each with a splitting of 1.2 eV , in addition to a
broad signal at higher binding energies, which could not be readily deconvoluted into its
corresponding doublet. A similar spectrum was observed in the Cu
2
BaSnS
4
prepared by
the co-sputtering method with the higher binding energy doublet set being assigned to
surface oxidation of the bulk material, and the broad high-binding energy signal assigned
to surface (SO
3,4
)
2–
species. In that case, both were removed by surface argon cleaning to
yield a single doublet indicating one sulfur species in the bulk material.
16
Figure 6.6. XPS survey scan of Cu
2
BaSnS
4
recovered from an EDT/en ink.
148
With regards to applications, a valuable feature of chalcogenide semiconductor
materials is the ability to tune the optoelectronic properties of the material by tuning
mixed chalcogen compositions to achieve solid solutions. We have shown this to be
especially straightforward using the alkahest method for preparing Sb
2
S
3-x
Se
x
solid
solutions, where bulk elemental selenium precursor is fully incorporated into the material
and sulfur from the thiol in the solvent system fills the remainder of chalcogen
stoichiometry.
24
As discussed above, reports on Cu
2
BaSnS
4
have shown the ability to
incorporate Se into the trigonal structure giving stoichiometries of Cu
2
BaSnS
4-x
Se
x
where
x = 0 – 3, and corresponding band gaps ranging from 1.55 – 2.05 eV .
10,12,14–16,19
To demonstrate the feasibility of the alkahest method for preparing more complex
multinary solid solutions, we set out to make a series of Cu
2
BaSnS
4-x
Se
x
alloys with
nominal x = 0, 1, 2, and 3. The precursor inks were formulated by adding the appropriate
amount of elemental Se powder to otherwise identical formulations of Cu
2
S, BaS, and
SnO precursors in 1:4 (vol/vol) EDT/en. The addition of Se appeared to increase the
solubility of the other precursors, an observation that has been made for other bulk
solutes in the presence of elemental chalcogen in previous reports.
21,37
The inks were
stirred at 50-70 ˚C for 8-11 d followed by annealing to 350 ˚C under a nitrogen
atmosphere. Elemental analysis by SEM-EDX showed that the value of x in the recovered
Cu
2
BaSnS
4-x
Se
x
materials was less than the nominal ratio assumed based on the amount
of selenium used in the ink formulation (i.e., nominal values of x = 0, 1, 2, and 3 yielded
experimental values of x = 0, 0.9, 1.4, and 2.0, respectively).
Even when an excess of Se
149
was used in the ink formulation (i.e., aiming for x = 4.75), the Se content of the recovered
material did not surpass the x = 3 boundary and the orthorhombic phase was not
achieved.
Figure 6.7. (a) Powder XRD patterns collected for each nominal composition of Cu
2
BaSnS
4-x
Se
x
,
the zoomed region to the right shows more clearly the shifting of peaks to lower 2θ values with
increasing selenium content. (b) Plot of calculated a and c lattice parameters and unit cell volume
vs experimental composition showing a near linear increase in all unit cell parameters with
increased Se content.
Figure 6.7 a shows the XRD patterns for the solid solution samples corresponding to
nominal values of x = 0, 1, 2, and 3, verifying that each is phase pure with the same
trigonal structure. The zoomed plot shows the general trend of peaks shifting to lower 2θ
values as x increases, as expected for increasing the proportion of larger Se atoms
150
uniformly throughout the materials. Figure 6.7 b shows how the lattice parameters (a and
c) and unit cell volume change with increasing experimental x, showing a nearly linear
correlation. Tauc plots derived from diffuse reflectance measurements were used to
determine the optical band gaps of each material (Figure 6.8 a), which range from 1.86 –
1.56 eV for the four compositions. A plot of experimental x vs. E
g,dir.
also indicates a
nearly linear trend of decreasing band gap with increasing x (Figure 6.8 b). Thus, by
simply adjusting the initial Se precursor content in the ink, corresponding semiconductors
with tunable band gaps can be easily prepared.
Figure 6.8. (a) Tauc plots of each solid solution composition with extrapolated band gaps denoted
and (b) plot of direct band gap vs experimental composition in Cu
2
BaSnS
4-x
Se
x
showing a nearly
linear decrease in E
g,dir.
with increasing x.
6.4 Experimental
6.4.1 General considerations
All reagents were used as received, solvents were sparged with nitrogen for 20 min
prior to use. 1,2-Ethylenediamine (en, 99.5%), copper (I) sulfide (Cu
2
S, 99.99%), and
151
barium sulfide (BaS, 99.9%) were all purchased from Sigma Aldrich. 1,2-Ethanedithiol
(EDT, 98+%), tin (II) oxide (SnO, 99%), and selenium powder (Se, -200 mesh, 99.999%)
were all purchased from Alfa Aesar.
6.4.2 Content of Recovered Cu
2
BaSnS
4-x
Se
x
Thermal Gravimetric Analysis (TGA) was performed on a TA Instruments TGA Q50
instrument in an alumina crucible under a flowing nitrogen atmosphere with a heating
rate of 5 ˚C min
–1
. TGA samples were pre-dried at 150-160 ˚C under flowing nitrogen
prior to TGA analysis. FT-IR spectra were measured on a Bruker Vertex 80. Samples
were prepared by drop casting the Cu
2
BaSnS
4-x
Se
x
ink onto a ZnSe window and either
drying to 100 ˚C or annealing to 350 ˚C under flowing nitrogen. Scanning electron
microscope-energy dispersive X-ray spectroscopy (SEM-EDX) was used for elemental
analysis on a JEOL JSM-7001F scanning-electron microscope with an operating voltage
of 23 kV and a working distance of 15 mm.
6.4.3 Structural and Optical Characterization
Powder X-ray Diffraction (XRD) patterns were collected in the 10−80° 2θ range using
a Rigaku Miniflex 600 diffractometer operated at 15 mA and 40 kV. Cu Kα radiation (λ
= 1.5406 Å) was employed. The step size and collection time were 0.05
°
and 1.5 s step
–1
,
respectively. All patterns were recorded under ambient conditions. UV-vis diffuse
reflectance spectroscopy was performed with a Perkin Elmer Lamba 950 equipped with a
150 mm integrating sphere. Cu
2
BaSnS
4-x
Se
x
powders were diluted with MgO powder for
analysis. X-ray Photoelectron Spectroscopy (XPS) was performed using a Kratos Axis
152
Ultra X-ray photoelectron spectrometer with the analyzer lens in hybrid mode. A
monochromatic aluminum anode with an operating current of 6 mA and voltage of 10 kV
was used with a step size of 0.1 eV, a pass energy of 20 eV, and a pressure range between
1-3 x 10
–8
Torr. The binding energy was referenced to the C 1s core level at 284.8 eV.
6.4.4 Preparation of Cu
2
BaSnS
4-x
Se
x
Inks
Cu
2
BaSnS
4
: 31.8 mg (0.2 mmol) Cu
2
S, 33.9 mg (0.2 mmol) BaS, and 26.9 mg (0.2
mmol) SnO were stirred at 50 ˚C in a mixture of 1.6 mL en and 0.4 mL EDT for 11 d,
resulting in a viscous, murky yellow solution that upon heating with an electric heat gun
became clear and free flowing. Cu
2
BaSnS
4-x
Se
x
inks were made using the same
formulation as the pure sulfide ink, with the addition of each corresponding
stoichiometric amount of Se powder, i.e., 15.8 mg (0.2 mmol) Se for x = 1, 31.6 mg (0.4
mmol) Se for x = 2, and 47.4 mg (0.6 mmol) Se for x = 3. These inks appeared to have a
faster dissolution time and were stirred at 50-70 ˚C for 8-11 d to yield viscous murky
solutions with a gradient of colors becoming more reddish with increased Se content.
These inks also became clear and free flowing upon heating with a heat gun.
6.4.5 Recovery of Cu
2
BaSnS
4-x
Se
x
Powdered Cu
2
BaSnS
4-x
Se
x
was recovered by drop-casting the Cu
2
BaSnS
4-x
Se
x
precursor inks (heated with heat gun to reduce viscosity) onto glass microscope slides,
annealing to 350 ˚C under flowing nitrogen, cooling to room temperature under flowing
nitrogen, and scraping off the resulting powder from the slides. Thin Films of Cu
2
BaSnS
4
were prepared using a precursor solution with similar formulation as the x = 0 inks for
153
powder samples, except in order to achieve a less viscous consistency, the en:EDT ratio
was increased to 9:1, and after the dissolution was complete, the ink was diluted with 2-
methoxyethanol in a 1:1 vol/vol ratio and sonicated to yield a homogeneous consistency.
The films were spin coated in two coats on a ca. 1 cm x 1 cm glass substrate using a
Laurell Technologies Corporation WS400Ez-6NPPLITE single-wafer spin processor at
1600 r.p.m. for 1 min in a warm (ca. 45 ˚C) nitrogen atmosphere. In between coats, the
films were annealed on a hot plate at 350 ˚C and allowed to cool to room temperature
before the second coat was spin coated. SEM images (Figure 6.9) of the films were
collected using a JEOL JSM-7001F scanning-electron microscope with an operating
voltage of 10 kV.
Figure 6.9. SEM micrographs of solution deposited Cu
2
BaSnS
4
thin film.
6.5 Conclusions
We have demonstrated for the first time a solution deposition method for the
preparation of Cu
2
BaSnS
4
. The straightforward alkahest method was used to dissolve
154
bulk Cu
2
S, BaS, and SnO in an EDT/en solvent mixture forming inks that upon mild
thermal annealing produces phase-pure Cu
2
BaSnS
4
. By incorporating elemental Se into
the ink formulation, Cu
2
BaSnS
4-x
Se
x
solid solutions were prepared with direct band gaps
ranging from 1.86 – 1.56 eV for compositions up to x = 2.02. This work demonstrates the
value of the alkahest method for the solution deposition of complex, compositionally
well-controlled multinary semiconductors. Moving forward, this method has the capacity
to be used as an inexpensive solution processing alternative for the development and
preparation of functional materials.
6.6 References
(1) Coughlan, C.; Ibáñez, M.; Dobrozhan, O.; Singh, A.; Cabot, A.; Ryan, K. M.
Compound Copper Chalcogenide Nanocrystals. Chem. Rev. 2017, 117, 5865–6109.
(2) Regulacio, M. D.; Han, M.-Y . Multinary I-III-VI
2
and I
2
-II-IV-VI
4
Semiconductor
Nanostructures for Photocatalytic Applications. Acc. Chem. Res. 2016, 49, 511–519.
(3) Tan, G.; Zhao, L.-D.; Kanatzidis, M. G. Rationally Designing High-Performance
Bulk Thermoelectric Materials. Chem. Rev. 2016, 116, 12123–12149.
(4) Gao, M. R.; Zheng, Y .-R.; Jiang, J.; Yu, S.-H. Pyrite-Type Nanomaterials for
Advanced Electrocatalysis. Acc. Chem. Res. 2017, 50, 2194–2204.
(5) Shao, M.; Chang, Q.; Dodelet, J.-P.; Chenitz, R. Recent Advances in
Electrocatalysts for Oxygen Reduction Reaction. Chem. Rev. 2016, 116, 3594–3657.
(6) Dasgupta, N. P.; Meng, X.; Elam, J. W.; Martinson, A. B. F. Atomic Layer
Deposition of Metal Sulfide Materials. Acc. Chem. Res. 2015, 48, 341–348.
(7) Chen, S.; Walsh, A.; Gong, X.; Wei, S. Classification of Lattice Defects in the
Kesterite Cu
2
ZnSnS
4
and Cu
2
ZnSnSe
4
Earth-Abundant Solar Cell Absorbers. Adv. Mater.
2013, 25, 1522–1539.
155
(8) Gershon, T.; Lee, Y . S.; Antunez, P.; Mankad, R.; Singh, S.; Bishop, D.; Gunawan,
O.; Hopstaken, M.; Haight, R. Photovoltaic Materials and Devices Based on the Alloyed
Kesterite Absorber (Ag
x
Cu
1–x
)
2
ZnSnSe
4
. Adv. Energy Mater. 2016, 1502468.
(9) Hong, F.; Lin, W.; Meng, W.; Yan, Y . Trigonal Cu
2
-II-Sn-VI
4
(II = Ba, Sr and VI =
S, Se) Quaternary Compounds for Earth-Abundant Photovoltaics. Phys. Chem. Chem.
Phys. 2016, 18, 4828–4834.
(10) Shin, D.; Saparov, B.; Zhu, T.; Huhn, W. P.; Blum, V .; Mitzi, D. B.
BaCu
2
Sn(S,Se)
4
: Earth-Abundant Chalcogenides for Thin-Film Photovoltaics. Chem.
Mater. 2016, 28, 4771–4780.
(11) Ge, J.; Yu, Y .; Yan, Y . Earth-Abundant Orthorhombic BaCu
2
Sn(Se
x
S
1–x
)
4
(X ≈
0.83) Thin Film for Solar Energy Conversion. ACS Energy Lett. 2016, 1, 583–588.
(12) Shin, D.; Ngaboyamahina, E.; Zhou, Y .; Glass, J. T.; Mitzi, D. B. Synthesis and
Characterization of an Earth-Abundant Cu
2
BaSn(S,Se)
4
Chalcogenide for
Photoelectrochemical Cell Application. J. Phys. Chem. Lett. 2016, 7, 4554–4561.
(13) Xiao, Z.; Meng, W.; Li, J. V .; Yan, Y . Distant-Atom Mutation for Better Earth-
Abundant Light Absorbers: A Case Study of Cu
2
BaSnSe
4
. ACS Energy Lett. 2017, 2, 29–
35.
(14) Ge, J.; Yu, Y .; Yan, Y . Earth-Abundant Trigonal BaCu
2
Sn(Se
x
S
1−x
)
4
(X = 0–0.55)
Thin Films with Tunable Band Gaps for Solar Water Splitting. J. Mater. Chem. A 2016, 4,
18885–18891.
(15) Ge, J.; Roland, P. J.; Koirala, P.; Meng, W.; Young, J. L.; Petersen, R.; Deutsch, T.
G.; Teeter, G.; Ellingson, R. J.; Collins, R. W.; et al. Employing Overlayers To Improve
the Performance of Cu
2
BaSnS
4
Thin Film Based Photoelectrochemical Water Reduction
Devices. Chem. Mater. 2017, 29, 916–920.
(16) Ge, J.; Yan, Y . Synthesis and Characterization of Photoelectrochemical and
Photovoltaic Cu
2
BaSnS
4
Thin Films and Solar Cells. J. Mater. Chem. C 2017, 5, 6406–
6419.
(17) Zhu, T.; Huhn, W. P.; Wessler, G. C.; Shin, D.; Saparov, B.; Mitzi, D. B.; Blum, V .
I
2
–II–IV–VI
4
(I = Cu, Ag; II = Sr, Ba; IV = Ge, Sn; VI = S, Se): Chalcogenides for Thin-
Film Photovoltaics. Chem. Mater. 2017, DOI: 10.1021/acs.chemmater.7b02638.
(18) Shin, D.; Zhu, T.; Huang, X.; Gunawan, O.; Blum, V .; Mitzi, D. B. Earth-
Abundant Chalcogenide Photovoltaic Devices with over 5% Efficiency Based on a
Cu
2
BaSn(S,Se)
4
Absorber. Adv. Mater. 2017, 29, 1606945.
156
(19) Ge, J.; Grice, C. R.; Yan, Y . Cu-Based Quaternary Chalcogenide Cu
2
BaSnS
4
Thin
Films Acting as Hole Transport Layers in Inverted Perovskite CH
3
NH
3
PbI
3
Solar Cells. J.
Mater. Chem. A 2017, 5, 2920–2928.
(20) Webber, D. H.; Brutchey, R. L. Alkahest for V
2
VI
3
Chalcogenides: Dissolution of
Nine Bulk Semiconductors in a Diamine-Dithiol Solvent Mixture. J. Am. Chem. Soc.
2013, 135, 15722–15725.
(21) Webber, D. H.; Buckley, J. J.; Antunez, P. D.; Brutchey, R. L. Facile Dissolution
of Selenium and Tellurium in a Thiol–amine Solvent Mixture under Ambient Conditions.
Chem. Sci. 2014, 5, 2498–2502.
(22) Antunez, P. D.; Torelli, D. A.; Yang, F.; Rabuffetti, F. A.; Lewis, N. S.; Brutchey,
R. L. Low Temperature Solution-Phase Deposition of SnS Thin Films. Chem. Mater.
2014, 26, 5444–5446.
(23) Buckley, J. J.; Greaney, M. J.; Brutchey, R. L. Ligand Exchange of Colloidal
CdSe Nanocrystals with Stibanates Derived from Sb
2
S
3
Dissolved in a Thiol-Amine
Mixture. Chem. Mater. 2014, 26, 6311–6317.
(24) McCarthy, C. L.; Webber, D. H.; Schueller, E. C.; Brutchey, R. L. Solution-Phase
Conversion of Bulk Metal Oxides to Metal Chalcogenides Using a Simple Thiol–Amine
Solvent Mixture. Angew. Chem. Int. Ed. 2015, 54, 8378–8381.
(25) McCarthy, C. L.; Cottingham, P.; Abuyen, K.; Schueller, E. C.; Culver, S. P.;
Brutchey, R. L. Earth Abundant CuSbS
2
Thin Films Solution Processed from Thiol–
amine Mixtures. J. Mater. Chem. C 2016, 4, 6230–6233.
(26) McCarthy, C. L.; Downes, C. A.; Schueller, E. C.; Abuyen, K.; Brutchey, R. L.
Method for the Solution Deposition of Phase-Pure CoSe
2
as an Efficient Hydrogen
Evolution Reaction Electrocatalyst. ACS Energy Lett. 2016, 1, 607–611.
(27) Buckley, J. J.; McCarthy, C. L.; Del Pilar-Albaladejo, J.; Rasul, G.; Brutchey, R.
L. Dissolution of Sn, SnO, and SnS in a Thiol–Amine Solvent Mixture: Insights into the
Identity of the Molecular Solutes for Solution-Processed SnS. Inorg. Chem. 2016, 55,
3175–3180.
(28) McCarthy, C. L.; Downes, C. A.; Brutchey, R. L. Room Temperature Dissolution
of Bulk Elemental Ni and Se for Solution Deposition of a NiSe
2
HER Electrocatalyst.
Inorg. Chem. 2017, 56, 10143–10146.
(29) McCarthy, C. L.; Brutchey, R. L. Solution Processing of Chalcogenide Materials
Using Thiol–amine “alkahest” Solvent Systems. Chem. Commun. 2017, 53, 4888–4902.
157
(30) Yang, Y .; Wang, G.; Zhao, W.; Tian, Q.; Huang, L.; Pan, D. Solution-Processed
Highly Efficient Cu
2
ZnSnSe
4
Thin Film Solar Cells by Dissolution of Elemental Cu, Zn,
Sn, and Se Powders. ACS Appl. Mater. Interfaces 2015, 7 (1), 460–464.
(31) Zhao, X.; Lu, M.; Koeper, M. J.; Agrawal, R. Solution-Processed Sulfur Depleted
Cu(In, Ga)Se
2
Solar Cells Synthesized from a Monoamine–dithiol Solvent Mixture. J.
Mater. Chem. A 2016, 4, 7390–7397.
(32) Hayashi, M.; Shiro, Y .; Oshima, T.; Murata, H. The Vibrational Assignment,
Rotational Isomerism and Force Constants of 1,2-Ethanedithiol. Bull. Chem. Soc. Jpn.
1965, 38, 1734–1740.
(33) Li, Y .-S.; Li, S. FTIR Spectra of HSCH
2
CH
2
SH, CH
3
SCH
2
SH, and CH
3
SSCH
3
in
Argon and Nitrogen Matrices. Spectrochim. Acta Part Mol. Spectrosc. 1994, 50, 509–
519.
(34) Lin-Vien, D.; Colthup, N. B.; Fateley, W. G.; Grasselli, J. G. The Handbook of
Infrared and Raman Characteristic Frequencies of Organic Molecules; Academic Press:
Boston, 1991.
(35) Moulder, J. F.; Stickle, W. F.; Sobol, P. E.; Bomben, K. D. Handbook of X-Ray
Photoelectron Spectroscopy; Chastain, J., Ed.; Physical Electronics Division, Perkin-
Elmer Corp.: Eden Prairie, Minnesota, 1979.
(36) Jin, X.; Li, J.; Chen, G.; Xue, C.; Liu, W.; Zhu, C. Preparation of Cu
2
ZnSnS
4
-
Based Thin Film Solar Cells by a Combustion Method. Sol. Energy Mater. Sol. Cells
2016, 146, 16–24.
(37) Zhao, D.; Tian, Q.; Zhou, Z.; Wang, G.; Meng, Y .; Kou, D.; Zhou, W.; Pan, D.;
Wu, S. Solution-Deposited Pure Selenide CIGSe Solar Cells from Elemental Cu, In, Ga,
and Se. J. Mater. Chem. A 2015, 3, 19263–19267.
158
Appendix A. Dissolution of Nickel Precursors in Alkahest Solvent Mixtures
For Solution Deposition of Nickel Sulfides
A.1. Introduction
There has been a rich discussion in the literature of the many observed phases and
stoichiometries of nickel sulfide spanning decades.
1-4
Researchers have carefully
determined the phase boundaries that make up the complex Ni-S phase diagram shown in
Figure A.1.
4
Due to the narrow phase spaces, specifically in the 40 - 70 at% sulfur range
at low temperatures, targeting a specific phase using a low-temperature synthesis can be
difficult.
Figure A.1. Ni-S phase diagram.
4
159
Herein are described the preliminary findings pertaining to experiments aimed at
elucidating the relationship between the solute species generated upon dissolution of
various nickel precursors (i.e., Ni, Ni(acac)
2
, Ni(COD)
2
, Ni(OH)
2
, and NiCl
2
) in thiol-
amine mixtures and the phase and stoichiometry of nickel sulfide recovered upon
annealing the precursor inks. We were interested in controlling the phase/stoichiometry
of nickel sulfides through the alkahest method because (1) we wanted to better
understand the mechanism at play for the dissolution of bulk materials in these systems,
and (2) because many nickel sulfides have been used as noble-metal-free electrocatalysts
and battery electrodes.
A.2. Results and Discussion
A.2.1. General Considerations
The Ni(acac)
2
used in the studies was likely in the hydrated form (i.e.,
Ni(acac)
2
•2H
2
O) as it was stored under ambient conditions. As these are preliminary
results, not all experiments were done in replicate to test for reproducibility; these results
are presented here to give initial insights into how these systems behave. All dissolutions
were carried out in 1:4 thiol/en mixtures, at room temperature, and under ambient
atmosphere unless otherwise specified. All annealing was done under a flowing nitrogen
atmosphere.
160
A.2.2. Dissolution and Recovery Experiments
Initial studies in this work began by experimenting with different nickel precursors in
merc/en and EDT/en solvent mixtures. After dissolution and removal of unreacted
precursor, the inks were annealed to ca. 350 ˚C. In early studies, the exact ramping, end
temperature, and duration of annealing were not thought to be important, but later these
parameters surfaced as possible drivers for the recovery of different nickel sulfide phases.
A summary of the dissolution and recovery experiments for the five can be found in
Table A.1.
Bulk “nickel sulfide” purchased from Alfa Aesar was identified as Ni
x
S
6
by XRD, as
shown in Figure A.2. This bulk material was stirred at room temperature in a mixture of
merc/en and over the course of 10 d only a small portion of the starting material was
dissolved. The mixture was filtered and annealed to 350 ˚C to yield phase pure α-NiS as
observed by XRD (Figure A). This change of stoichiometry and phase suggests
dissolution of the bulk nickel sulfide down to molecular solutes in the alkahest mixture.
161
Table A.1. Examples of the appearance of inks prepared from five nickel precursors in merc/en
or EDT/en inks. The approximate time allowed for dissolution at room temperature and all phases
that have been recovered from each ink are also listed.
Merc/en EDT/en
Ni + S
Dissolution Time Days Days
Recovered Phases
α-NiS
α/β-NiS mixture
α-NiS
Ni(acac)
2
Deep blue/green
(no photo available)
Dissolution Time Minutes Minutes
Recovered Phases
Ni
3
S
2
+ Ni
9
S
8
impurity
α-NiS
α-NiS + β-NiS impurity
Ni(COD)
2
Dissolution Time 4 h 1 d Minutes
Recovered Phases α-NiS α-NiS
Ni(OH)
2
Dissolution Time Minutes Minutes
Recovered Phases α-NiS + Cubic NiS
2
impurity α-NiS
NiCl
2
Dissolution Time Minutes Minutes
Recovered Phases Ni
9
S
8
Ni
9
S
8
162
Figure A.2. XRD of as-purchased “Nickel Sulfide” from Alfa Aesar (black trace) and of the
recovered material from a merc/en ink made using the bulk nickel sulfide as the precursor
annealed to 350 ˚C.
Dissolution of elemental Ni + S in merc/en solvent mixtures occurs at a slow rate
(over days) even at elevated temperatures (e.g, 50-95 ˚C), and in most cases the Ni metal
does not fully dissolve. The inks are thus filtered prior to deposition and recovery. Initial
dissolution and recovery experiments indicated that Ni + S dissolved in merc/en gave
varying ratios of α-NiS + β-NiS depending on the conditions of dissolution (i.e., time and
temperature) with the recovery temperature being kept at ca. 350 ˚C. Although initially it
appeared that higher dissolution temperatures and longer dissolution time lead to a higher
portion of the β-NiS, the results were not reproducible (Figure A.3). Most significantly,
phase pure α-NiS was recovered at higher temperatures of dissolution than the mainly β-
NiS material (Figure A.3, top pattern).
163
Figure A.3. XRD patterns for various dissolution conditions for Ni + S in en/merc inks. The
dissolution reaction temperature and time are given above each plot.
Ni(acac)
2
dissolved completely in en to give a clear magenta solution owing its color
to [Ni(en)
3
]
2+
ions (discussed further below). Even if the reaction was not stirred, the
Ni(acac)
2
and en reacted in a heterogeneous fashion to yield a purple solid, the XRD of
which did not appear to have any major Ni(acac)
2
diffraction peaks Figure A.4.
Figure A.4. XRD patterns of the bulk Ni(acac)
2
(bottom) and the [Ni(en)
3
]
2+
salt heterogeneous
reaction product when Ni(acac)
2
is added to en without stirring (top).
164
When Ni(acac)
2
was dissolved in a mixture of merc/en, a similar magenta solution
was observed. Several experiments were conducted in hopes of deriving a relationship
between concentration or en/merc ratio and the recovered phase upon annealing,
however, there was no reproducible trend observed in the powder XRD data (Figure A.5
a, Figure A.6). In EDT/en mixtures, Ni(acac)
2
dissolves to give a deep greenish blue
solution. Upon filtering and annealing this ink to 350 ˚C, phase pure α-NiS was recovered
irrespective of ink concentration or annealing temperature (Figure A.5 b).
Figure A.5. XRD patterns for recovered materials from Ni(acac)
2
inks with varying
concentrations and annealing temperatures as denoted on each of the patterns. (a) merc/en inks
and (b) EDT/en inks.
165
Figure A.6. XRD patterns of the recovered materials from 100 mg mL
-1
Ni(acac)
2
inks with
varying en/merc ratios as denoted above each pattern.
When merc was added to Ni(acac)
2
in the absence of en, a deep red viscous solution
was obtained. With dissolution in EDT, a black solid precipitated out of solution. Both of
these observations indicate reaction of the thiol with Ni(acac)
2
in the absence of en.
When the merc solution was annealed to ca. 350 ˚C, phase pure α-NiS was recovered
(Figure A.7).
Figure A.7. XRD pattern of recovered material from 80 mg mL
-1
Ni(acac)
2
in merc after
centrifugation and annealing to 352 ˚C.
166
When Ni(COD)
2
was added to en, and deep red solution was produced. Adding EDT
to this solution resulted in a dark green solution, while adding merc initially resulted in an
orange solution. After filtering and annealing to 350 ˚C, both inks yielded phase pure α-
NiS (Figure A.8). Interestingly, as shown in Figure A.8, when left stirring at room
temperature the merc/en ink slowly changed from clear orange to murky pink over the
course of 24 h. Single crystal diffraction of crystals grown from this pink solution
verified the presence of the same [Ni(en)
3
]
2+
solute present in the other nickel precursor
merc/en inks, as discussed below in the single crystal X-ray diffraction section.
Figure A.8. XRD patterns from materials recovered by annealing Ni(COD)
2
in merc/en and
EDT/en inks to 353 ˚C after centrifugation.
Figure A.9. Images of a Ni(COD)
2
ink at various stages of dissolution. (a) 10 min, (b) 4 h, and
(c) 24 h.
167
Ni(OH)
2
was prepared by reacting NiCl
2
with NaOH in water to precipitate Ni(OH)
2
.
In merc/en, Ni(OH)
2
gave a murky pink solution, while in EDT/en it gave a dark green
solution. Filtering and annealing to 350 ˚C resulted in a mixture of α-NiS and seemingly
cubic NiS
2
in the case of the merc/en ink, while the EDT/en ink gave phase pure α-NiS
(Figure A.10).
Figure A.10. XRD patterns from materials recovered by annealing Ni(OH)
2
in merc/en and
EDT/en inks to 350 ˚C.
NiCl
2
was dissolved at room temperature within 15 min in merc/en and EDT/en to
yield clear bright magenta and deep greenish blue inks, respectively. Upon annealing,
Ni
9
S
8
was recovered from both inks Figure A.11.
Figure A.11. XRD patterns from materials recovered by annealing NiCl
2
in merc/en and EDT/en
inks to 353 ˚C.
168
A.2.2. UV-vis Experiments
As a control experiment to show displacement of acac ligands from Ni(acac)
2
by en,
we obtained UV-vis spectra of an aqueous solution of Ni(acac)
2
with incremental
equivalents of en added (Figure A.12 a). Figure A.12 b shows the peak maxima shifts as
the solute species progresses from zero en ligands to three en ligands, after which point
(ca. 3-4 equivalents of en) the peaks no longer shift, corresponding to the solution
changing color from blue to magenta.
Figure A.12. (a) UV-vis spectra of increasing equivalents of en in an aqueous Ni(acac)
2
solution
and (b) plot of peak maxima shifts for the spectra, showing a plateau after ca. 3-4 eq of en.
Figure A.13 a gives the spectra of five merc/en inks containing different nickel
precursors. All precursors give similar spectra with the same two peaks (ca. 545 nm and
885 nm), as seen in the final spectra for the control experiment (vida supra), varying only
in their intensities and onset of near-UV absorption, aside from Ni(COD)
2
which has an
additional peak around 400 nm (hence the orange color of the Ni(COD)
2
ink). This
indicates that upon dissolution in a merc/en ink, all of the different Ni precursors are
converted into [Ni(en)
3
]
2+
species in solution. As such, this suggests that the solute
169
species is not the cause for different phases upon annealing, since each of these
precursors was shown to give different phases/stoichiometries of nickel sulfide upon
annealing (Table A.1). Similarly, when the same precursors are dissolved in EDT/en
mixtures, they all give similar UV-vis spectra, again varying only in intensity, and
suggesting that once dissolved, all precursors yield the same solute (Figure A.13 b).
Interestingly, these spectra appear to be just slightly red shifted from the two absorbance
peaks observed for a solution of Ni(acac)
2
in EDT only, with no amine present. This may
mean that the EDT/en dissolution species are similar to the EDT-only species, which is
likely to have thiol coordination, as opposed to full en coordination as observed for the
merc/en inks.
Figure A.13. UV-vis spectra of nickel precursors in mixtures of (a) merc/en and (b)
EDT/en.
A.2.3. Single Crystal X-ray Diffraction
X-ray quality single crystals were grown from some of the amine-thiol inks either by
slow evaporation of the solvent system, or by layering an antisolvent (i.e., toluene or
170
hexane). Figures A.14 a-c show the crystal structures as observed by single crystal X-ray
diffraction. As predicted by the UV-vis data, all merc/en solutes contain a [Ni(en)
3
]
2+
ion,
which is charge balanced by two monothiolate molecules. In the case of Ni + S, there are
two water molecules associated with the structure, whereas the Ni(COD)
2
and Ni(acac)
2
structures have a single en molecule associated. The structural data obtained for the ink
corresponding Ni(acac)
2
dissolved in EDT/en (Figure A.14 c) was not absolutely
conclusive, and only the basic backbone of the molecules could be discerned; charge
assignments were unclear. Still, the solutes observed in Figure A.14 c for a Ni(acac)
2
dissolved in EDT/en ink indicate a less straightforward dissolution mechanism for the
EDT/en ink compared to the merc/en inks.
Figure A.14. Structures discerned in the single crystal X-ray diffraction data for crystals
grown out of (a) merc/en Ni(acac)
2
and merc/en Ni(COD)
2
, (b) merc/en Ni + S, and
(c) EDT/en Ni(acac)
2
inks. For (c) only the basic structures could be obtained from the
data.
Ni
N
H
2
H
2
N
NH
2
NH
2
H
2
N
H
2
N
2+
OH S
2
N
N
Ni
S
S
S
S
Ni
N
H
2
H
2
N
NH
2
NH
2
H
2
N
H
2
N
2+
OH S
2
H
2
N NH
2
2 H
2
O
(a)
(b)
(c)
171
A.3. Conclusions
From the UV-vis and single crystal X-ray diffraction studies, it can be concluded that
merc/en and EDT/en give the same respective dissolved solutes, regardless of nickel
precursor (except in the case of Ni(COD)
2
, where an intermediate is formed before the
final nickel species, as observed by an orange-colored mixture upon initial dissolution,
which slowly changes to the purple color observed for all other nickel precursors). Thus,
within a set of merc/en or EDT/en inks , the identity of the nickel solute does not appear
to be the main determining factor in governing which phase is recovered upon annealing.
The concentration of the inks may play a role, however, in the cases where different
phases were obtained at different concentrations, but we were unable to identify any
reproducible trends for the precursor/solvent combinations that were studied.
A.4. Future Work
In the case that this work is continued, it is important to be aware of the issues
encountered while working on the project. For example, the complex Ni–S phase diagram
with the many stoichiometries and phases of nickel sulfides, e.g., Ni
3
S
2
, Ni
4
S
3
, Ni
6
S
5
,
Ni
7
S
6
, Ni
9
S
8
, Ni
3
S
4
, NiS
2
, NiS, make phase-pure materials difficult to recover
reproducibly, even under seemingly identical processing conditions. Reproducing
experiments with the utmost precision is imperative. Key parameters to be mindful of are
ink concentration, thiol/amine ratio, reaction temperature (even room temperature may
fluctuate from day to day), reaction time, purity and dryness of solvent, method of
172
removing undissolved precursor form the ink before annealing (e.g., syringe filtering or
centrifugation), ramping rate of the annealing system, final annealing temperature, and
duration of annealing. Additionally, several of the inks tend to solidify (a known
characteristic of alkahest mixtures), but can be liquefied by gentle heating using a heat
gun. Further studies on the effect of ink concentration on recovered phase should be
conducted, as those presented here do not cover all of the precursors; perhaps broader
ranges of concentration could also be tested.
As part of this project, a Li-ion battery trial was performed using alkahest-processed
α-NiS as the cathode, however, after ca. 75% through the first cycle, the cell short-
circuited, reaching only ca. 40% of its theoretical capacitance. More careful battery
engineering may be a solution to this problem in future work. Additionally,
electrocatalytic studies should be conducted using similar methods to those described for
CoSe
2
and NiSe
2
in Chapters 4 and 5, respectively, in this thesis.
A. 5. References
(1) Kullerud, G.; Yund, R. A. The Ni-S System and Related Minerals. J. Petrol. 1962, 3,
126–175.
(2) Lin, R. Y.; Hu, D. C.; Chang, Y. A. Thermodynamics and Phase Relationships of
Transition Metal-Sulfur Systems: II. The Nickel-Sulfur System. Metall. Trans. B 1978, 9,
531–538.
(3) Waldner, P.; Pelton, A. D. Thermodynamic Modeling of the Ni–S System. Z. Für
Met. 2004, 95, 672–681.
(4) Okamoto, H. J. Ni-S (Nickel-Sulfur). Phase Equilib. Diffus. 2009, 30, 123.
173
Bibliography
“Best Research-Cell Efficiencies” Chart; National Renewable Energy Laboratory -
National Center for Photovoltaics, 2016.
Antunez, P. D.; Torelli, D. A.; Yang, F; Rabuffetti, F. A.; Lewis, N. S.; Brutchey, R. L.;
Low Temperature Solution-Phase Deposition of SnS Thin Films. Chem. Mater. 2014, 26,
5444-5446.
Arnou, P.; Cooper, C. S.; Malkov, A. V.; Bowers, J. W.; Walls, J. M. Solution-Processed
CuIn(S,Se)
2
Absorber Layers for Application in Thin Film Solar Cells. Thin Solid Films
2015, 582, 31–34.
Arnou, P.; Cooper, C. S.; Uličná, S.; Abbas, A.; Eeles, A.; Wright, L. D.; Malkov, A. V.;
Walls, J. M.; Bowers, J. W. Solution Processing of CuIn(S,Se)
2
and Cu(In,Ga)(S,Se)
2
Thin Film Solar Cells Using Metal Chalcogenide Precursors. Thin Solid Films 2016,
DOI: 10.1016/j.tsf.2016.10.011.
Arnou, P.; van Hest, M. F. A. M.; Cooper, C. S.; Malkov, A. V.; Walls, J. M.; Bowers, J.
W. Hydrazine-Free Solution-Deposited CuIn(S,Se)
2
Solar Cells by Spray Deposition of
Metal Chalcogenides. ACS Appl. Mater. Interfaces 2016, 8, 11893–11897.
Basu, M.; Zhang, Z.-W.; Chen, C.-J.; Chen, P.-T.; Yang, K.-C.; Ma, C.-G.; Lin, C. C.;
Hu, S.-F.; Liu, R.-S. Heterostructure of Si and CoSe
2
: A Promising Photocathode Based
on a Non-Noble Metal Catalyst for Photoelectrochemical Hydrogen Evolution. Angew.
Chem. Int. Ed. 2015, 54, 6211–6216.
Bonomo, M.; Marrani, A. G.; Novelli, V.; Awais, M.; Dowling, D. P.; Vos, J. G.; Dini,
D. Surface Properties of Nanostructured NiO Undergoing Electrochemical Oxidation in
3-Methoxy-Propionitrile. Appl. Surf. Sci. 2017, 403, 441–447.
Buckley, J. J.; Greaney M. J.; Brutchey, R. L. Ligand Exchange of Colloidal CdSe
Nanocrystals with Stibanates Derived from Sb
2
S
3
Dissolved in a Thiol-Amine Mixture.
Chem. Mater., 2014, 26, 6311–6317.
Buckley, J. J.; McCarthy, C. L.; Del Pilar-Albaladejo, J.; Rasul, G.; Brutchey, R. L.
Dissolution of Sn, SnO, and SnS in a Thiol–Amine Solvent Mixture: Insights into the
Identity of the Molecular Solutes for Solution-Processed SnS. Inorg. Chem. 2016, 55,
3175–3180.
174
Calixto-Rodriguez, M.; Garcia, H. M.; Nair, M. T. S.; Nair, P. K. Antimony
Chalcogenide/Lead Selenide Thin Film Solar Cell with 2.5% Conversion Efficiency
Prepared by Chemical Deposition. ECS J. Solid State Sci. Technol. 2013, 2, Q69–Q73.
Chen, G.; Wang, W.; Wang, C.; Ding, T.; Yang, Q. Controlled Synthesis of Ultrathin
Sb
2
Se
3
Nanowires and Application for Flexible Photodetectors. Adv. Sci. 2015, 2,
1500109.
Chen, S.; Walsh, A.; Gong, X.; Wei, S. Classification of Lattice Defects in the Kesterite
Cu
2
ZnSnS
4
and Cu
2
ZnSnSe
4
Earth-Abundant Solar Cell Absorbers. Adv. Mater. 2013,
25, 1522–1539.
Chern, G. C.; Lauks, I. Spin-coated Amorphous Chalcogenide Films. J. Appl. Phys.
1982, 53, 6979–6982.
Choi, Y. C.; Lee, D. U.; Noh, J. H.; Kim, E. K.; Seok, S. I. Highly Improved
Sb
2
S
3
Sensitized-Inorganic–Organic Heterojunction Solar Cells and Quantification of
Traps by Deep-Level Transient Spectroscopy. Adv. Funct. Mater. 2014, 24, 3587-3592.
Choi, Y. C.; Lee, Y. H.; Im, S. H.; Noh, J. H.; Mandal, T. N.; Yang, W. S.; Seok, S. I.
Efficient Inorganic-Organic Heterojunction Solar Cells Employing Sb
2
(S
x
/Se
1-x
)
3
Graded-
Composition Sensitizers. Adv. Energy Mater. 2014, 4, 1301680.
Cooper, C. S.; Arnou, P.; Wright, L. D.; Uličná, S.; Walls, J. M.; Malkov, A. V.; Bowers,
J. W. An Innovative Approach for Fabrication of Cu
2
ZnSnSe
4
Absorber Layers Using
Solutions of Elemental Metal Powders. Thin Solid Films 2016, DOI:
10.1016/j.tsf.2016.12.026.
Coughlan, C.; Ibáñez, M.; Dobrozhan, O.; Singh, A.; Cabot, A.; Ryan, K. M. Compound
Copper Chalcogenide Nanocrystals. Chem. Rev. 2017, 117, 5865–6109.
CRC Handbook of Chemistry and Physics, 95th ed, CRC, Boca Raton, 2015.
Dasgupta, N. P.; Meng, X.; Elam, J. W.; Martinson, A. B. F. Atomic Layer Deposition of
Metal Sulfide Materials. Acc. Chem. Res. 2015, 48, 341–348.
Deng, Z.; Mansuripur, M.; Muscat, A. J. Simple Colloidal Synthesis of Single-Crystal
Sb−Se−S Nanotubes with Composition Dependent Band-Gap Energy in the Near-
Infrared. Nano Lett. 2009, 9, 2015–2020.
Efthimiopoulos, I.; Zhang, J.; Kucway, M.; Park, C.; Ewing, R. C.; Wang, Y.
175
Sb
2
Se
3
Under Pressure. Sci. Rep. 2013, 3, 2665 –2673.
Faber, M. S.; Dziedzic, R.; Lukowski, M. A.; Kaiser, N. S.; Ding, Q.; Jin, S. High-
Performance Electrocatalysis Using Metallic Cobalt Pyrite (CoS2) Micro- and
Nanostructures. J. Am. Chem. Soc. 2014, 136, 10053–10061.
Faber, M. S.; Jin, S. Earth-Abundant Inorganic Electrocatalysts and Their Nanostructures
for Energy Conversion Applications. Energy Environ. Sci. 2014, 7, 3519–3542.
Faber, M. S.; Lukowski, M. A.; Ding, Q.; Kaiser, N. S.; Jin, S. Earth-Abundant Metal
Pyrites (FeS
2
, CoS
2
, NiS
2
, and Their Alloys) for Highly Efficient Hydrogen Evolution
and Polysulfide Reduction Electrocatalysis. J. Phys. Chem. C 2014, 118, 21347–21356.
Feser, J. P.; Chan, E. M.; Majumdar, A.; Segalman, R. A.; Urban, J. J. Ultralow Thermal
Conductivity in Polycrystalline CdSe Thin Films with Controlled Grain Size. Nano Lett.
2013, 13, 2122–2127.
Gao, M.-R.; Zheng, Y.-R.; Jiang, J.; Yu, S.-H. Pyrite-Type Nanomaterials for Advanced
Electrocatalysis. Acc. Chem. Res. 2017, 50, 2194–2204.
Ge, J.; Grice, C. R.; Yan, Y. Cu-Based Quaternary Chalcogenide Cu
2
BaSnS
4
Thin Films
Acting as Hole Transport Layers in Inverted Perovskite CH
3
NH
3
PbI
3
Solar Cells. J.
Mater. Chem. A 2017, 5, 2920–2928.
Ge, J.; Roland, P. J.; Koirala, P.; Meng, W.; Young, J. L.; Petersen, R.; Deutsch, T. G.;
Teeter, G.; Ellingson, R. J.; Collins, R. W.; et al. Employing Overlayers To Improve the
Performance of Cu
2
BaSnS
4
Thin Film Based Photoelectrochemical Water Reduction
Devices. Chem. Mater. 2017, 29, 916–920.
Ge, J.; Yan, Y. Synthesis and Characterization of Photoelectrochemical and Photovoltaic
Cu
2
BaSnS
4
Thin Films and Solar Cells. J. Mater. Chem. C 2017, 5, 6406–6419.
Ge, J.; Yu, Y.; Yan, Y. Earth-Abundant Orthorhombic BaCu
2
Sn(Se
x
S
1–x
)
4
(X ≈ 0.83)
Thin Film for Solar Energy Conversion. ACS Energy Lett. 2016, 1, 583–588.
Ge, J.; Yu, Y.; Yan, Y. Earth-Abundant Trigonal BaCu
2
Sn(Se
x
S
1−x
)
4
(X = 0–0.55) Thin
Films with Tunable Band Gaps for Solar Water Splitting. J. Mater. Chem. A 2016, 4,
18885–18891.
Ge, Z.; Zhao, L.; Wu, D.; Liu, X.; Zhang, B.; Li, J.; He, J. Low-Cost, Abundant Binary
Sulfides as Promising Thermoelectric Materials. Mater. Today 2016, 19, 227–239.
176
Gershon, T.; Lee, Y. S.; Antunez, P.; Mankad, R.; Singh, S.; Bishop, D.; Gunawan, O.;
Hopstaken, M.; Haight, R. Photovoltaic Materials and Devices Based on the Alloyed
Kesterite Absorber (Ag
x
Cu
1–x
)
2
ZnSnSe
4
. Adv. Energy Mater. 2016, 1502468.
Global Flexible Electronics Market 2016-2020; Technavio, 2016.
Global Thin-Film Semiconductor Deposition Market 2016-2020; Technavio, 2016.
Green, M. A.; Emery, K.; Hishikawa, Y.; Warta, W.; Dunlop, E. D.; Levi, D. H.; Ho-
Baillie, A. W. Y. Solar Cell Efficiency Tables (Version 49). Prog. Photovolt. Res. Appl.
2017, 25, 3–13.
Grosvenor, A. P.; Biesinger, M. C.; Smart, R. S. C.; McIntyre, N. S. New Interpretations
of XPS Spectra of Nickel Metal and Oxides. Surf. Sci. 2006, 600, 1771–1779.
Hasan, M. R.; Arinze, E. S.; Singh, A. K.; Oleshko, V. P.; Guo, S.; Rani, A.; Cheng, Y.;
Kalish, I.; Zaghloul, M. E.; Rao, M. V.; et al. An Antimony Selenide Molecular Ink for
Flexible Broadband Photodetectors. Adv. Electron. Mater. 2016, 2, 1600182.
Hayashi, M.; Shiro, Y.; Oshima, T.; Murata, H. The Vibrational Assignment, Rotational
Isomerism and Force Constants of 1,2-Ethanedithiol. Bull. Chem. Soc. Jpn. 1965, 38,
1734–1740.
Hong, F.; Lin, W.; Meng, W.; Yan, Y. Trigonal Cu
2
-II-Sn-VI
4
(II = Ba, Sr and VI = S,
Se) Quaternary Compounds for Earth-Abundant Photovoltaics. Phys. Chem. Chem. Phys.
2016, 18, 4828–4834.
Jackson, P.; Wuerz, R.; Hariskos, D.; Lotter, E.; Witte, W.; Powalla, M. Effects of Heavy
Alkali Elements in Cu(In,Ga)Se
2
Solar Cells with Efficiencies up to 22.6%. Phys. Status
Solidi RRL 2016, 10, 583–586.
Jeon, N. J.; Noh, J. H.; Kim, Y. C.; Yang, W. S.; Ryu, S.; Seok, S. I. Solvent engineering
for high-performance inorganic–organic hybrid perovskite solar cells. Nat. Mater.,
2014, 13, 897–903.
Jin, X.; Li, J.; Chen, G.; Xue, C.; Liu, W.; Zhu, C. Preparation of Cu
2
ZnSnS
4
-Based Thin
Film Solar Cells by a Combustion Method. Sol. Energy Mater. Sol. Cells 2016, 146, 16–
24.
Kong, D.; Cha, J. J.; Wang, H.; Lee, H. R.; Cui, Y. First-Row Transition Metal
177
Dichalcogenide Catalysts for Hydrogen Evolution Reaction. Energy Environ. Sci. 2013,
6, 3553–3558.
Kong, D.; Wang, H.; Lu, Z.; Cui, Y. CoSe
2
Nanoparticles Grown on Carbon Fiber Paper:
An Efficient and Stable Electrocatalyst for Hydrogen Evolution Reaction. J. Am. Chem.
Soc. 2014, 136, 4897– 4900.
Kovalenko, M. V.; Scheele, M.; Talapin, D. V. Colloidal Nanocrystals with Molecular
Metal Chalcogenide Surface Ligands. Science 2009, 324, 1417–1420.
Kovalenko, M. V.; Spokoyny, B.; Lee, J.; Scheele, M.; Weber, A.; Perera, S.; Landry, D.;
Talapin, D. V. Semiconductor Nanocrystals Functionalized with Antimony Telluride
Zintl Ions for Nanostructured Thermoelectrics. J. Am. Chem. Soc. 2010, 132, 6686–6695.
Krishnan, B.; Shaji, S.; Ornelas, R. E. Progress in development of copper antimony
sulfide thin films as an alternative material for solar energy harvesting. J. Mater. Sci.:
Mater. Electron., 2015, 26, 4770–4781.
Kucernak, A. R. J.; Sundaram, V. N. N. Nickel Phosphide: The Effect of Phosphorus
Content on Hydrogen Evolution Activity and Corrosion Resistance in Acidic Medium. J.
Mater. Chem. A 2014, 2, 17435–17445.
Kullerud, G.; Yund, R. A. The Ni-S System and Related Minerals. J. Petrol. 1962, 3,
126–175.
Kwak, I. H.; Im, H. S.; Jang, D. M.; Kim, Y. W.; Park, K.; Lim, Y. R.; Cha, E. H.; Park,
J. CoSe
2
and NiSe
2
Nanocrystals as Superior Bifunctional Catalysts for Electrochemical
and Photoelectrochemical Water Splitting. ACS Appl. Mater. Interfaces 2016, 8, 5327–
5334.
Kyono, A.; Kimata, M. Crystal structures of chalcostibite (CuSbS
2
) and emplectite
(CuBiS
2
): Structural relationship of stereochemical activity between chalcostibite and
emplectite. Am. Mineral., 2015, 90, 162–165.
Ladd, M. Symmetry of Crystals and Molecules, Vol. 1, Oxford University Press, New
York, 2014, p.45.
Larson, A. C.; Von Dreele, R. B. General Structure Analysis System (GSAS); Los Alamos
National Laboratory Report LAUR 86-784; Los Alamos National Laboratory: Los
Alamos, NM, 1994.
178
Li, Y.-S.; Li, S. FTIR Spectra of HSCH
2
CH
2
SH, CH
3
SCH
2
SH, and CH
3
SSCH
3
in Argon
and Nitrogen Matrices. Spectrochim. Acta Part Mol. Spectrosc. 1994, 50, 509–519.
Li, Z.; Xue, H.; Wang, J.; Tang, Y.; Lee, C.-S.; Yang, S. Reduced Graphene
Oxide/Marcasite-Type Cobalt Selenide Nanocrystals as an Anode for Lithium-Ion
Batteries with Excellent Cyclic Performance. ChemElectroChem 2015, 2, 1682–1686.
Liang, H.; Li, L.; Meng, F.; Dang, L.; Zhuo, J.; Forticaux, A.; Wang, Z.; Jin, S. Porous
Two-Dimensional Nanosheets Converted from Layered Double Hydroxides and Their
Applications in Electrocatalytic Water Splitting. Chem. Mater. 2015, 27, 5702–5711.
Liang, J.; Yang, Y.; Zhang, J.; Wu, J.; Dong, P.; Yuan, J.; Zhang, G.; Lou, J. Metal
Diselenide Nanoparticles as Highly Active and Stable Electrocatalysts for the Hydrogen
Evolution Reaction. Nanoscale 2015, 7, 14813–14816.
Liang, L.; Cheng, H.; Lei, F.; Han, J.; Gao, S.; Wang, C.; Sun, Y.; Qamar, S.; Wei, S.;
Xie, Y. Metallic Single-Unit-Cell Orthorhombic Cobalt Diselenide Atomic Layers:
Robust Water-Electrolysis Catalysts. Angew. Chem. 2015, 127, 12172–12176.
Lin, R. Y.; Hu, D. C.; Chang, Y. A. Thermodynamics and Phase Relationships of
Transition Metal-Sulfur Systems: II. The Nickel-Sulfur System. Metall. Trans. B 1978, 9,
531–538.
Lin-Vien, D.; Colthup, N. B.; Fateley, W. G.; Grasselli, J. G. The Handbook of Infrared
and Raman Characteristic Frequencies of Organic Molecules; Academic Press: Boston,
1991.
Liu, F.; Zhu, J.; Hu, L.; Zhang, B.; Yao, J.; Nazeeruddin, M. K.; Grätzel, M.; Dai, S.
Low-Temperature, Solution-Deposited Metal Chalcogenide Films as Highly Efficient
Counter Electrodes for Sensitized Solar Cells. J. Mater. Chem. A 2015, 3, 6315–6323.
Liu, Y.; Yao, D.; Shen, L.; Zhang, H.; Zhang, X.; Yang, B. Alkylthiol-Enabled Se
Powder Dissolution in Oleylamine at Room Temperature for the Phosphine-Free
Synthesis of Copper-Based Quaternary Selenide Nanocrystals. J. Am. Chem. Soc. 2012,
134, 7207–7210.
Lu, T.-H.; Chen, C.-J.; Basu, M.; Ma, C.-G.; Liu, R.-S. The CoTe
2
Nanostructure: An
Efficient and Robust Catalyst for Hydrogen Evolution. Chem. Commun. 2015, 51,
17012– 17015.
Ma, Y.; Vartak, P. B.; Nagaraj, P.; Wang, R. Y. Thermoelectric Properties of Copper
Chalcogenide Alloys Deposited via the Solution-Phase Using a Thiol–amine Solvent
179
Mixture. RSC Adv. 2016, 6, 99905–99913.
McCarthy, C. L.; Brutchey, R. L. Solution Processing of Chalcogenide Materials Using
Thiol–amine “alkahest” Solvent Systems. Chem. Commun. 2017, 53, 4888–4902.
McCarthy, C. L.; Cottingham, P.; Abuyen, K.; Schueller, E. C.; Culver, S.
P.; Brutchey, R. L. Earth Abundant CuSbS
2
Thin Films Solution Processed from Thiol-
Amine Mixtures J. Mater. Chem. C 2016, 4, 6230 – 6233.
McCarthy, C. L.; Downes, C. A.; Brutchey, R. L. Room Temperature Dissolution of Bulk
Elemental Ni and Se for Solution Deposition of a NiSe
2
HER Electrocatalyst. Inorg.
Chem. 2017, 56, 10143–10146.
McCarthy, C. L.; Downes, C. A.; Schueller, E. C.; Abuyen, K.; Brutchey, R. L. Method
for the Solution Deposition of Phase-Pure CoSe
2
as an Efficient Hydrogen Evolution
Reaction Electrocatalyst. ACS Energy Lett. 2016, 1, 607–611.
McCarthy, C. L.; Webber, D. H.; Schueller E. C.; Brutchey, R. L. Solution-Phase
Conversion of Bulk Metal Oxides to Metal Chalcogenides Using a Simple Thiol–Amine
Solvent Mixture. Angew. Chem., Int. Ed., 2015, 54, 8378–8381.
McCrory, C. C. L.; Jung, S.; Ferrer, I. M.; Chatman, S. M.; Pe- ters, J. C.; Jaramillo, T. F.
Benchmarking Hydrogen Evolving Reaction and Oxygen Evolving Reaction
Electrocatalysts for Solar Water Splitting Devices. J. Am. Chem. Soc. 2015, 137, 4347–
4357.
McKone, J. R.; Marinescu, S. C.; Brunschwig, B. S.; Winkler, J. R.; Gray, H. B. Earth-
Abundant Hydrogen Evolution Electrocatalysts. Chem. Sci. 2014, 5, 865–878.
Milliron, D. J.; Mitzi, D. B.; Copel, M.; Murray, C. E. Solution-Processed Metal
Chalcogenide Films for p-Type Transistors. Chem. Mater. 2006, 18, 587-590.
Milliron, D. J.; Raoux, S.; Shelby, R. M.; Jordan-Sweet, J. Solution-Phase Deposition
and Nanopatterning of GeSbSe Phase-Change Materials. Nat. Mater. 2007, 6, 352–356.
Miskin, C. K.; DuboisCamacho, A.; Reese, M. O.; Agrawal, R. A Direct Solution
Deposition Approach to CdTe Thin Films. J. Mater. Chem. C 2016, 4, 9167–9171.
Miskin, C. K.; Yang, W.; Hages, C. J.; Carter, N. J.; Joglekar, C. S.; Stach, E. A.;
Agrawal, R. 9.0% Efficient Cu
2
ZnSn(S,Se)
4
Solar Cells from Selenized Nanoparticle
Inks. Prog. Photovolt. Res. Appl. 2015, 23, 654–659.
180
Mitzi, D. B. N
4
H
9
Cu
7
S
4
: A Hydrazinium-Based Salt with a Layered Cu
7
S
4
- Framework.
Inorg. Chem. 2007, 46, 926–931.
Mitzi, D. B. Polymorphic One-Dimensional (N
2
H
4
)
2
ZnTe: Soluble Precursors for the
Formation of Hexagonal or Cubic Zinc Telluride. Inorg. Chem. 2005, 44, 7078–7086.
Mitzi, D. B. Solution Processing of Chalcogenide Semiconductors via Dimensional
Reduction. Adv. Mater. 2009, 21, 3141–3158.
Mitzi, D. B. Solution Processing of Inorganic Materials; John Wiley & Sons: Hoboken,
New Jersey, 2008.
Mitzi, D. B. Synthesis, Structure, and Thermal Properties of Soluble Hydrazinium
Germanium(IV) and Tin(IV) Selenide Salts. Inorg. Chem. 2005, 44, 3755–3761.
Mitzi, D. B.; Copel, M.; Chey, S. J. Low-Voltage Transistor Employing a High-Mobility
Spin-Coated Chalcogenide Semiconductor. Adv. Mater. 2005, 17, 1285–1289.
Mitzi, D. B.; Copel, M.; Murray, C. E. High-Mobility P-Type Transistor Based on a
Spin-Coated Metal Telluride Semiconductor. Adv. Mater. 2006, 18, 2448–2452.
Mitzi, D. B.; Kosbar, L. L.; Murray, C. E.; Copel, M.; Afzali, A. High-Mobility Ultrathin
Semiconducting Films Prepared by Spin Coating. Nature 2004, 428, 299–303.
Mitzi, D. B.; Raoux, S.; Schrott, A. G.; Copel, M.; Kellock, A.; Jordan-Sweet, J.
Solution-Based Processing of the Phase-Change Material KSb5S8. Chem. Mater. 2006,
18, 6278–6282.
Mitzi, D. B.; Solution Processing of Chalcogenide Semiconductors via Dimensional
Reduction. Adv. Mater. 2009, 21, 3141-3158.
Mitzi, D. B.; Yuan, M.; Liu, W.; Kellock, A. J.; Chey, S. J.; Deline, V.; Schrott, A. G. A
High-Efficiency Solution-Deposited Thin-Film Photovoltaic Device. Adv. Mater. 2008,
20, 3657–3662.
Moulder, J. F.; Stickle, W. F.; Sobol, P. E.; Bomben, K. D. Handbook of X-Ray
Photoelectron Spectroscopy; Chastain, J., Ed.; Physical Electronics Division, Perkin-
Elmer Corp.: Eden Prairie, Minnesota, 1979.
Navarro-Flores, E.; Chong, Z.; Omanovic, S. Characterization of Ni, NiMo, NiW and
181
NiFe Electroactive Coatings as Electrocatalysts for Hydrogen Evolution in an Acidic
Medium. J. Mol. Catal. Chem. 2005, 226, 179–197.
Nonobe, Y. Development of the Fuel Cell Vehicle Mirai. IEEJ Trans. Electr. Electron.
Eng. 2017, 12, 5–9.
Okamoto, H. J. Ni-S (Nickel-Sulfur). Phase Equilib. Diffus. 2009, 30, 123.
Peccerillo, E.; Major, J.; Phillips, L.; Treharne, R.; Whittles, T. J.; Dhanak, V.; Halliday,
D.; Durose, K. in Characterization of sulfurized CuSbS2 thin films for PV applications.
2014 IEEE 40th Photovoltaic Specialist Conference (PVSC), IEEE, Denver, CO, 2014,
pp. 0266–0269.
Powell, A. V. and Vaqueiro, P. in Thermoelectric Materials and Devices, eds. I.
Nandhakumar, N. M. White and S. Beeby, The Royal Society of Chemistry, Cambridge,
UK, 2016, ch. 2, pp. 27–59.
Qu, Y.; Zoppi, G.; Beattie, N. S. The Role of Nanoparticle Inks in Determining the
Performance of Solution Processed Cu
2
ZnSn(S,Se)
4
Thin Film Solar Cells. Prog.
Photovolt. 2016, 24, 836–845.
Ramos Aquino, J. A.; Rodriguez Vela, D. L.; Shaji, S.; Avellaneda, D. A.; Krishnan, B.
Spray pyrolysed thin films of copper antimony sulfide as photovoltaic absorber. Phys.
Status Solidi C, 2016, 13, 24–29.
Rath, T.; MacLachlan, A. J.; Brown, M. D.; Haque, S. A. Structural, optical and charge
generation properties of chalcostibite and tetrahedrite copper antimony sulfide thin films
prepared from metal xanthates. J. Mater. Chem. A, 2015, 3, 24155–24162.
Regulacio, M. D.; Han, M.-Y. Multinary I-III-VI
2
and I
2
-II-IV-VI
4
Semiconductor
Nanostructures for Photocatalytic Applications. Acc. Chem. Res. 2016, 49, 511–519.
Reuss, R. H.; Chalamala, B. R. in Solution Processing of Inorganic Materials (Ed.: D. B.
Mitzi), Wiley, New York 2009, pp. 1-32.
Rietveld, H. M. A Profile Refinement Method for Nuclear and Magnetic Structures. J.
Appl. Crystallogr. 1969, 2, 65–71.
Rietveld, H. M. Line profiles of neutron powder-diffraction peaks for structure
refinement. Acta Crystallogr, 1967, 22, 151.
Rietveld, H. M. Line Profiles of Neutron Powder-Diffraction Peaks for Structure
182
Refinement. Acta Crystallogr. 1967, 22, 151–152.
Rietveld, H. M.; A profile refinement method for nuclear and magnetic structures. J.
Appl. Crystallogr. 1969, 2, 65.
Rodríguez-Lazcano, Y.; Nair, M. T. S.; Nair, P. K. CuSbS
2
thin film formed through
annealing chemically deposited Sb
2
S
3
–CuS thin films. J. Cryst. Growth, 2001, 223, 399–
406.
Rose, D. H.; Hasoon, F. S.; Dhere, R. G.; Albin, D. S.; Ribelin, R. M.; Li, X. S.;
Mahathongdy, Y.; Gessert, T. A.; Sheldon, P. Fabrication Procedures and Process
Sensitivities for CdS/CdTe Solar Cells. Prog. Photovolt. Res. Appl. 1999, 7, 331–340.
Scheer, R.; Schock, H. Chalcogenide Photovoltaics: Physics, Technologies, and Thin
Film Devices; Wiley-VHC: Weinheim, Germany, 2011.
Seefeld, S.; Limpinsel, M., Liu, Y.; Farhi, N.; Weber, A.; Zhang, Y.; Berry, N.; Kwon, Y.
J.; Perkins, C. L.; Hemminger, J. C.; Wu, R.; Law, M. Iron Pyrite Thin Films Synthesized
from an Fe(acac)
3
Ink. J. Am. Chem. Soc., 2013, 135, 4412–4424.
Septina, W.; Ikeda, S.; Iga, Y.; Harada, T.; Matsumura, M. Thin film solar cell based on
CuSbS
2
absorber fabricated from an electrochemically deposited metal stack. Thin Solid
Films, 2014, 550, 700–704.
Shao, M.; Chang, Q.; Dodelet, J.-P.; Chenitz, R. Recent Advances in Electrocatalysts for
Oxygen Reduction Reaction. Chem. Rev. 2016, 116, 3594–3657.
Shi, J.-B.; Wu, P.-F.; Kao, C.-T.; Lee, M.-W.; Chan, C.-C.; Yang, P.-C.; Lin, C.-L.;
Huang, R.-Y.; Huang, Y.-J.; Lin, S.-K.; et al. Influence of Selenization Temperature on
Synthesis, Morphology, and Properties of Nanostructured CoSe
2
Films. Cryst. Res.
Technol. 2015, 50, 155–159.
Shin, D.; Ngaboyamahina, E.; Zhou, Y.; Glass, J. T.; Mitzi, D. B. Synthesis and
Characterization of an Earth-Abundant Cu
2
BaSn(S,Se)
4
Chalcogenide for
Photoelectrochemical Cell Application. J. Phys. Chem. Lett. 2016, 7, 4554–4561.
Shin, D.; Saparov, B.; Zhu, T.; Huhn, W. P.; Blum, V.; Mitzi, D. B. BaCu
2
Sn(S,Se)
4
:
Earth-Abundant Chalcogenides for Thin-Film Photovoltaics. Chem. Mater. 2016, 28,
4771–4780.
Shin, D.; Zhu, T.; Huang, X.; Gunawan, O.; Blum, V.; Mitzi, D. B. Earth-Abundant
Chalcogenide Photovoltaic Devices with over 5% Efficiency Based on a Cu
2
BaSn(S,Se)
4
183
Absorber. Adv. Mater. 2017, 29, 1606945.
Shockley, W. H.; Queisser, J. Detailed Balance Limit of Efficiency of p-n Junction Solar
Cells. J. Appl. Phys. 1961, 32, 510–519.
Suárez-Sandoval, D. Y.; Nair, M. T. S.; Nair, P. K.; Photoconductive Antimony Sulfide-
Selenide Thin Films Produced by Heating a Chemically Deposited Se-Sb
2
S
3
Layer. J.
Electro-chem. Soc. 2006, 153, C91–C96.
Suehiro, S.; Horita, K.; Yuasa, M.; Tanaka, T.; Fujita, K.; Ishiwata, Y.; Shimanoe, K.;
Kida, T. Synthesis of Copper–Antimony-Sulfide Nanocrystals for Solution-Processed
Solar Cells. Inorg. Chem., 2015, 54, 7840–7845.
Sun, Y.; Rogers, J. A. Inorganic Semiconductors for Flexible Electronics. Adv. Mater.
2007, 19, 1897–1916.
Tan, G.; Zhao, L.-D.; Kanatzidis, M. G. Rationally Designing High-Performance Bulk
Thermoelectric Materials. Chem. Rev. 2016, 116, 12123–12149.
Tan, S. M.; Pumera, M. Bottom-up Electrosynthesis of Highly Active Tungsten Sulfide
(WS
3–x
) Films for Hydrogen Evolution. ACS Appl. Mater. Interfaces 2016, 8, 3948–
3957.
Tang, C.; Cheng, N.; Pu, Z.; Xing, W.; Sun, X. NiSe Nanowire Film Supported on Nickel
Foam: An Efficient and Stable 3D Bifunctional Electrode for Full Water Splitting.
Angew. Chem. Int. Ed. 2015, 54, 9351–9355.
Tanusevski, A. Optical and photoelectric properties of SnS thin films prepared by
chemical bath deposition. Semicond. Sci. Technol., 2003, 18, 501–505.
The Rietveld Method; Young, R. A., Ed.; IUCr Monographs on Crystallography (Book
5); Oxford University Press: New York, 1993.
Thomas, C. E. Sustainable Transportation Options for the 21st Century and beyond: A
Comprehensive Comparison of Alternatives to the Internal Combustion Engine; Springer
International Publishing: Switzerland, 2015.
Thompson, P.; Cox, D. E.; Hastings, J. B. Rietveld Refinement of Debye–Scherrer
Synchrotron X-Ray Data from Al
2
O
3
. J. Appl. Crystallogr. 1987, 20, 79–83.
Thompson, P.; Cox, D. E.; Hastings, J. M. Rietveld refinement of Debye–Scherrer
184
synchrotron X-ray data from Al
2
O
3
. J. Appl. Crystallogr. 1987, 20, 79.
Tian, Q.; Cui, Y.; Wang, G.; Pan, D. A Robust and Low-Cost Strategy to Prepare
Cu
2
ZnSnS
4
Precursor Solution and Its Application in Cu2ZnSn(S,Se)
4
Solar Cells. RSC
Adv. 2015, 5, 4184–4190.
Tian, Q.; Huang, L.; Zhao, W.; Yang, Y.; Wang, G.; Pan, D. Metal Sulfide Precursor
Aqueous Solutions for Fabrication of Cu
2
ZnSn(S,Se)
4
Thin Film Solar Cells. Green
Chem. 2015, 17, 1269–1275.
Tian, Q.; Wang, G.; Zhao, W.; Chen, Y.; Yang, Y.; Huang, L.; Pan, D. Versatile and
Low-Toxic Solution Approach to Binary, Ternary, and Quaternary Metal Sulfide Thin
Films and Its Application in Cu
2
ZnSn(S,Se)
4
Solar Cells. Chem. Mater. 2014, 26, 3098–
3103.
Todorov, T. K.; Gunawan, O.; Gokmen T.; Mitzi, D. B. Solution-processed
Cu(In,Ga)(S,Se)
2
absorber yielding a 15.2% efficient solar cell. Prog. Photovoltaics,
2013, 21, 82–87.
Trends 2016 In Photovoltaic Applications; International Energy Agency, 2016.
U.S. Department of Energy; Office of Energy Efficiency and Renewable Energy; U.S.
Environmental Prodection Agency. Model Year 2017 Fuel Economy Guide; 2017.
U.S. Department of Energy. Fuel Cell Technologies Office: Hydrogen Production
https://energy.gov/eere/fuelcells/hydrogenproduction (accessed Apr 26, 2017).
Vineyard, B. D. Versatility and the Mechanism of the N-Butyl-Amine-Catalyzed
Reaction of Thiols with Sulfur. J. Org. Chem. 1967, 32, 3833–3836.
Vrubel, H.; Moehl, T.; Grätzel, M.; Hu, X. Revealing and Accelerating Slow Electron
Transport in Amorphous Molybdenum Sulphide Particles for Hydrogen Evolution
Reaction. Chem. Commun. 2013, 49, 8985–8987.
Waldner, P.; Pelton, A. D. Thermodynamic Modeling of the Ni–S System. Z. Für Met.
2004, 95, 672–681.
Walker, B. C.; Agrawal, R. Contamination-Free Solutions of Selenium in Amines for
Nanoparticle Synthesis. Chem. Commun. 2014, 50, 8331–8334.
Wang, F.; Li, Y.; Shifa, T. A.; Liu, K.; Wang, F.; Wang, Z.; Xu, P.; Wang, Q.; He, J.
185
Selenium-Enriched Nickel Selenide Nanosheets as a Robust Electrocatalyst for Hydrogen
Generation. Angew. Chem. Int. Ed. 2016, 55, 6919–6924.
Wang, K.; Xi, D.; Zhou, C.; Shi, Z.; Xia, H.; Liu, G.; Qiao, G. CoSe
2
Necklace-like
Nanowires Supported by Carbon Fiber Paper: A 3D Integrated Electrode for the
Hydrogen Evolution Reaction. J. Mater. Chem. A 2015, 3, 9415–9420.
Wang, R. Y.; Feser, J. P.; Gu, X.; Yu, K. M.; Segalman, R. A.; Majumdar, A.; Milliron,
D. J.; Urban, J. J. Universal and Solution-Processable Precursor to Bismuth Chalcogenide
Thermoelectrics. Chem. Mater. 2010, 22, 1943-1945.
Wang, W.; Winkler, M. T.; Gunawan, O.; Gokmen, T.; Todorov, T. K.; Zhu, Y.; Mitzi,
D. B. Device Characteristics of CZTSSe Thin-Film Solar Cells with 12.6% Efficiency.
Adv. Energy Mater. 2014, 4, 1301465.
Wang, X.; Kolen’ko, Y. V.; Bao, X. Q.; Kovnir, K.; Liu, L. One-Step Synthesis of Self-
Supported Nickel Phosphide Nanosheet Array Cathodes for Efficient Electrocatalytic
Hydrogen Generation. Angew. Chem. Int. Ed. 2015, 54, 8188–8192.
Wang, X.; Kolen’ko, Y. V.; Bao, X.-Q.; Kovnir, K.; Liu, L. One-Step Synthesis of Self-
Supported Nickel Phosphide Nanosheet Array Cathodes for Efficient Electrocatalytic
Hydrogen Generation. Angew. Chem. Int. Ed. 2015, 54, 8188–8192.
Webber, D. H.; Brutchey, R. L. Alkahest for V
2
VI
3
Chalcogenides: Dissolution of Nine
Bulk Semiconductors in a Diamine-Dithiol Solvent Mixture. J. Am. Chem. Soc. 2013,
135, 15722–15725.
Webber, D. H.; Buckley, J. J.; Antunez, P. D.; Brutchey, R. L. Facile Dissolution of
Selenium and Tellurium in a Thiol–amine Solvent Mixture under Ambient Conditions.
Chem. Sci. 2014, 5, 2498– 2502.
Welch, A. W.; Zawadzki, P. P.; Lany, S.; Wolden, C. A.; Zakutayev, A. Self-regulated
growth and tunable properties of CuSbS
2
solar absorbers. Sol. Energy Mater. Sol. Cells,
2015, 132, 499–506.
Xiao, Z.; Meng, W.; Li, J. V.; Yan, Y. Distant-Atom Mutation for Better Earth-Abundant
Light Absorbers: A Case Study of Cu
2
BaSnSe
4
. ACS Energy Lett. 2017, 2, 29–35.
Xu, D.; Shen, S.; Zhang, Y.; Gu, H.; Wang, Q. Selective Synthesis of Ternary Copper–
Antimony Sulfide Nanocrystals. Inorg. Chem., 2013, 52, 12958–12962.
186
Yang, B.; Wang, L.; Han, J.; Zhou, Y.; Song, H.; Chen, S.; Zhong, J.; Lv, L.; Niu, D.;
Tang, J. CuSbS
2
as a Promising Earth-Abundant Photovoltaic Absorber Material: A
Combined Theoretical and Experimental Study. Chem. Mater., 2014, 26, 3135–3143.
Yang, R. B.; Bachmann, J.; Pippel, E.; Berger, A.; Woltersdorf, J.; Gosele, U.; Nielsch,
K. Pulsed Vapor-Liquid-Solid Growth of Antimony Selenide and Antimony Sulfide
Nanowires. Adv. Mater. 2009, 21, 3170-3174; d) El-Sayad, E.; Compositional
Dependence of the Optical Properties of Amorphous Sb
2
Se
3−x
S
x
thin films. J. Non-Cryst.
Solids 2008, 354, 3806–3811.
Yang, Y.; Kang, X.; Huang, L.; Pan, D. Tuning the Band Gap of Cu
2
ZnSn(S,Se)
4
Thin
Films via Lithium Alloying. ACS Appl. Mater. Interfaces 2016, 8, 5308–5313.
Yang, Y.; Kang, X.; Huang, L.; Wei, S.; Pan, D. A General Water-Based Precursor
Solution Approach to Deposit Earth Abundant Cu
2
ZnSn(S,Se)
4
Thin Film Solar Cells. J.
Power Sources 2016, 313, 15–20.
Yang, Y.; Wang, G.; Zhao, W.; Tian, Q.; Huang, L.; Pan, D. Solution-Processed Highly
Efficient Cu
2
ZnSnSe
4
Thin Film Solar Cells by Dissolution of Elemental Cu, Zn, Sn, and
Se Powders. ACS Appl. Mater. Interfaces 2015, 7, 460–464.
Young, R. A. Ed. The Rietveld Method; IUCr Monographs on Crystallography (Book 5);
Oxford University Press: New York, 1993.
Yu, L.; Kokenyesi, R. S.; Keszler. D. A.; Zunger, A. Inverse Design of High Absorption
Thin-Film Photovoltaic Materials. Adv. Energy Mater., 2013, 3, 43–48.
Yuan, B.; Luan, W.; Tu, S. One-Step Solvothermal Synthesis of Nickel Selenide Series:
Composition and Morphology Control. CrystEngComm 2012, 14, 2145–2151.
Yuan, M.; Mitzi, D. B. Solvent properties of hydrazine in the preparation of
metal chalcogenide bulk materials and films. Dalton Trans. 2009, 6078-6088.
Zeng, M.; Li, Y. Recent Advances in Heterogeneous Electrocatalysts for the Hydrogen
Evolution Reaction. J. Mater. Chem. A 2015, 3, 14942–14962.
Zhai, T.; Ye, M.; Li, L.; Fang, X.; Liao, M.; Li, Y.; Koide, Y.; Bando, Y.; Golberg, D.
Single-Crystalline Sb
2
Se
3
Nanowires for High-Performance Field Emitters and
Photodetectors. Adv. Mater. 2010, 22, 4530–4533.
187
Zhang, H.; Ding, Q.; He, D.; Liu, H.; Liu, W.; Li, Z.; Yang, B.; Zhang, X.; Lei, L.; Jin, S.
A P-Si/NiCoSe
x
Core/Shell Nanopillar Array Photocathode for Enhanced
Photoelectrochemical Hydrogen Production. Energy Environ. Sci. 2016, 9, 3113–3119.
Zhang, R.; Cho, S.; Lim, D. G.; Hu, X.; Stach, E. A.; Handwerker, C. A.; Agrawal, R.
Metal–metal Chalcogenide Molecular Precursors to Binary, Ternary, and Quaternary
Metal Chalcogenide Thin Films for Electronic Devices. Chem. Commun. 2016, 52, 5007–
5010.
Zhang, R.; Szczepaniak, S. M.; Carter, N. J.; Handwerker, C. A.; Agrawal, R. A Versatile
Solution Route to Efficient Cu
2
ZnSn(S,Se)
4
Thin Film Solar Cells. Chem. Mater. 2015,
27, 2114–2120.
Zhang, T.; Yang, Y.; Liu, D.; Tse, S. C.; Cao, W.; Feng, Z.; Chen, S.; Qian, L. High
Efficiency Solution-Processed Thin-Film Cu(In,Ga)(Se,S)
2
Solar Cells. Energy Environ.
Sci. 2016, 9, 3674–3681.
Zhao, D.; Fan, Q.; Tian, Q.; Zhou, Z.; Meng, Y.; Kou, D.; Zhou, W.; Wu, S. Eliminating
Fine-Grained Layers in Cu(In,Ga)(S,Se)
2
Thin Films for Solution-Processed High
Efficiency Solar Cells. J. Mater. Chem. A 2016, 4, 13476–13481.
Zhao, D.; Tian, Q.; Zhou, Z.; Wang, G.; Meng, Y.; Kou, D.; Zhou, W.; Pan, D.; Wu, S.
Solution-Deposited Pure Selenide CIGSe Solar Cells from Elemental Cu, In, Ga, and Se.
J. Mater. Chem. A 2015, 3, 19263–19267.
Zhao, W.; Pan, D.; Liu, S. Kesterite Cu
2
Zn(Sn,Ge)(S,Se)
4
Thin Film with Controlled Ge-
Doping for Photovoltaic Application. Nanoscale 2016, 8, 10160–10165.
Zhao, X.; Lu, M.; Koeper, M. J.; Agrawal, R. Solution-Processed Sulfur Depleted Cu(In,
Ga)Se
2
Solar Cells Synthesized from a Monoamine–dithiol Solvent Mixture. J. Mater.
Chem. A 2016, 4, 7390–7397.
Zheng, J.; Yan, Y.; Xu, B. Correcting the Hydrogen Diffusion Limitation in Rotating
Disk Electrode Measurements of Hydrogen Evolution Reaction Kinetics. J. Electrochem.
Soc. 2015, 162, F1470– F1481.
Zheng, L.J.; Zhang, B.-P.; Han, C.G.; Pei, J.; Chen, N. Mechanical Alloying-Spark
Plasma Sintering Synthesis and Thermoelectric Properties of N-Type NiSe
2+x
Semiconductors: Analysis of Intrinsic Defects and Phase Structures. J. Mater. Sci. Mater.
Electron. 2016, 27, 8363–8369.
Zhou, H.; Chen, Q.; Li, G.; Luo, S.; Song, T.; Duan, H.; Hong, Z.; You, J.; Liu, Y.; Yang,
188
Y. Interface engineering of highly efficient perovskite solar cells. Science 2014, 345,
542-546.
Zhou, H.; Yu, F.; Liu, Y.; Sun, J.; Zhu, Z.; He, R.; Bao, J.; Goddard, W. A.; Chen, S.;
Ren, Z. Outstanding Hydrogen Evolution Reaction Catalyzed by Porous Nickel
Diselenide Electrocatalysts. Energy Environ. Sci. 2017, 10, 1487–1497.
Zhou, Y.; Leng, M.; Xia, Z.; Zhong, J.; Song, H.; Liu, X.; Yang, B.; Zhang, J.; Chen, J.;
Zhou, K.; Han, J.; Cheng, Y.; Tang, J. Solution-Processed Antimony Selenide
Heterojunction Solar Cells. Adv. Energy Mater. 2014, 4, 1301846.
Zhu, T.; Huhn, W. P.; Wessler, G. C.; Shin, D.; Saparov, B.; Mitzi, D. B.; Blum, V. I
2
–
II–IV–VI
4
(I = Cu, Ag; II = Sr, Ba; IV = Ge, Sn; VI = S, Se): Chalcogenides for Thin-
Film Photovoltaics. Chem. Mater. 2017, DOI: 10.1021/acs.chemmater.7b02638.
Zhuo, J.; Cabán-Acevedo, M.; Liang, H.; Samad, L.; Ding, Q.; Fu, Y.; Li, M.; Jin, S.
High-Performance Electrocatalysis for Hydrogen Evolution Reaction Using Se-Doped
Pyrite-Phase Nickel Diphosphide Nanostructures. ACS Catal. 2015, 5, 6355–6361.
Abstract (if available)
Abstract
In the face of anthropogenic climate change we must begin to think and act based on a systems approach rather than addressing problems as discrete events. Regarding the design of products, we should not only design for sustainability in the direct use of the product, but also for low‐impact fabrication as well as possible reuse, upcycling, or recycling. Specifically, for the design of functional inorganic materials, such as catalysts and semiconductors, we must consider the various impacts, beyond functionality, of choosing specific elements and sources to incorporate into materials. A second aspect of a systems approach to functional materials is the design of versatile, low‐impact, high‐throughput fabrication processes. Currently, the industry standards for fabricating functional inorganic materials are energetically and financially burdensome, requiring capital‐intensive hardware and energy‐intensive deposition processes. As such, there is room for improvements toward more cost‐ and energy‐effective manufacturing methods. ❧ The deposition of molecular precursor inks prepared from bulk materials is an ideal approach to material preparation, as it requires only a single dissolution step followed by direct deposition onto a suitable substrate. Additionally, the composition of molecular inks can be easily tuned by simply changing the formulation of the ink, and the very nature of such precursor inks facilitates thorough mixing on the molecular level to yield more compositionally homogeneous films. With these advantages in mind, our group developed a novel binary solvent system comprised of ethylenediamine (en) and a short chain thiol that is capable of dissolving bulk inorganic precursors, including chalcogenides, oxides, and elemental materials to give solution processable molecular inks. Since its initial discovery, this “alkahest” solution processing method has been employed for the deposition of a wide range of Earth‐abundant functional inorganic thin films including semiconductors for photovoltaic (PV) devices, electrocatalysts, thermoelectrics, and photodetectors. This method is an example for which a systems approach has been used to couple Earth‐abundant compositions with sustainable manufacturing and the possibility for reuse or recycling. ❧Since the seminal report on the solvent system, wherein the dissolution of nine bulk V₂VI₃ chalcogenide semiconductors were shown to dissolve at room temperature, under ambient pressure, and over a short time, we and others have worked toward broadening the scope of soluble precursors. We studied the solubility of ten bulk oxides, showing that upon annealing their respective alkahest inks phase pure sulfides could be recovered. In some cases, when elemental selenium or tellurium was added to the ink, selenides or tellurides could be made. Using Sb₂O₃ and selenium as an example system, we showed how varying the nominal content of selenium in the ink formulation could lead to a series of Sb₂Se₃₋ₓSₓ alloys with tunable band gaps from 1.2 – 1.6 eV. Using Sb₂S₃ and Cu₂S as precursors, we prepared high quality thin films of CuSbS₂, an Earth‐abundant alternative to CuInGaS₂ as a PV absorber. We demonstrated that the alkahest‐processed CuSbS₂ films had optoelectronic properties promising for application in PV devices. ❧ In addition to absorber materials for PV, we applied the alkahest method to the preparation of Earth‐abundant alternatives to platinum group metal electrocatalysts for the hydrogen evolution reaction (HER). We first prepared nano‐structured marcasite‐type CoSe₂ by annealing an ink formulated using Co(OH)₂ and elemental selenium as precursors. Thin films of CoSe₂ that were spin coated onto Highly Ordered Pyrolytic Graphite (HOPG) facilitated HER with 100% Faradaic efficiency and key electrocatalysis parameters on par with previous reports for similar materials prepared by other methods. By comparing samples with various CoSe₂ loadings, we were able to demonstrate further value of the alkahest method in the ability to easily vary the film thickness by simply adjusting the number of ink coats, which lead to optimization of catalyst utilization. Next we prepared pyrite‐type NiSe₂ using elemental nickel and selenium as precursors. Comparison of the HER performance for various NiSe₂ loadings showed good catalyst utilization for all loadings, indicating the ability to use thicker films to achieve higher hydrogen production per unit of geometric surface area. ❧ In our most recent work, we reported on the first solution synthesis of Cu₂BaSn(S,Se)₄ (CBTS). This material has gained recent interest as an alternative to Cu₂ZnSn(S,Se)₄ (CZTS) because of its higher theoretical PV performance stemming from the rendering of antisite defects energetically unfavorable upon replacement of zinc with a larger atom, barium. We used Cu₂S, BaS, and SnO as precursors in an ethanedithiol/en‐based formulation, which after several days of mixing gave a thick solution that could be annealed to yield CBTS. Thorough material characterization verified the phase purity, elemental composition, elemental oxidation states, and band gap of the CBTS. Although the as‐made ink was viscous and not able to be spin coated directly, we were able to prepare uniform CBTS thin films by diluting the ink with 2-methoxyethanol. We were also able to show that by incorporating elemental selenium into the ink formulation a set of Cu₂BaSnS₄₋ₓSeₓ alloys with experimentally determined values from x = 0 – 2, could be prepared with tunable band gaps from 1.56-1.86 eV, thus demonstrating a straightforward handle for engineering optimized photovoltaic absorbers.
Linked assets
University of Southern California Dissertations and Theses
Conceptually similar
PDF
Solution processed functional chalcogenide thin films and their molecular solutes from thiol-amine inks
PDF
Solution‐phase synthesis and deposition of earth‐abundant metal chalcogenide semiconductors
PDF
Semiconductor inks for solution processed electronic thin‐films
PDF
The chemistry of polymorphism in semiconductor nanocrystals
PDF
Synthesis and characterization of metal chalcogenide semiconductor nanocrystals using dialkyl dichalcogenide precursors
PDF
Solution phase synthesis routes to functional nanomaterials for energy storage
PDF
Perovskite chalcogenides: emerging semiconductors for visible to infrared optoelectronics
PDF
Expanding the library of surface ligands for semiconductor nanocrystal synthesis and photovoltaic applications
PDF
Organic photovoltaics: from materials development to device application
PDF
Fabrication and study of organic and inorganic optoelectronics using a vapor phase deposition (VPD)
PDF
Accessible cytotoxic and antiviral drug analogues: improved synthetic approaches to isoindolinones and bioisosteric difluoromethylated nucleotides, and the search for therapeutic organotelluranes
Asset Metadata
Creator
McCarthy, Carrie L.
(author)
Core Title
Solution processing of chalcogenide functional materials using thiol–amine “alkahest” solvent systems
School
College of Letters, Arts and Sciences
Degree
Doctor of Philosophy
Degree Program
Chemistry
Publication Date
04/06/2018
Defense Date
02/27/2018
Publisher
University of Southern California
(original),
University of Southern California. Libraries
(digital)
Tag
alkahest,amine,chalcogenide,functional materials,materials,OAI-PMH Harvest,semiconductor,solution processing,thin film,thiol
Language
English
Contributor
Electronically uploaded by the author
(provenance)
Advisor
Brutchey, Richard (
committee chair
), Thompson, Mark (
committee member
), West, Joshua (
committee member
)
Creator Email
carriemccarthy8@gmail.com
Permanent Link (DOI)
https://doi.org/10.25549/usctheses-c89-368
Unique identifier
UC11670317
Identifier
etd-McCarthyCa-6157.pdf (filename),usctheses-c89-368 (legacy record id)
Legacy Identifier
etd-McCarthyCa-6157.pdf
Dmrecord
368
Document Type
Dissertation
Rights
McCarthy, Carrie L.
Type
texts
Source
University of Southern California
(contributing entity),
University of Southern California Dissertations and Theses
(collection)
Access Conditions
The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the a...
Repository Name
University of Southern California Digital Library
Repository Location
USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA
Tags
alkahest
amine
chalcogenide
functional materials
materials
semiconductor
solution processing
thin film
thiol