University of Southern California Dissertations and Theses

Quantifying the surface chemistry of semiconductor nanocrystals spanning covalent to ionic materials

(USC Thesis Other)

Quantifying the surface chemistry of semiconductor nanocrystals spanning covalent to ionic materials

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content
QUANTIFYING THE SURFACE CHEMISTRY OF SEMICONDUCTOR
NANOCRYSTALS SPANNING COVALENT TO IONIC MATERIALS

by
Sara R. Smock
__________________________________________________________________
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)

December 2021

Copyright 2021 Sara R. Smock

ii

Acknowledgements

The work presented herein would not have been possible without the remarkable
support of my advisors, peers, family, and friends. First, I would like to express my utmost
gratitude to my advisor, Prof. Richard Brutchey, who not only facilitated my research
curiosities, but also supported me with meaningful guidance, advice, and nominations for
additional opportunities. Specifically, under his guidance I was awarded an NSF Graduate
Research Fellowship and the WiSE Cisco Systems Fellowship, which provided me funding
to pursue my research fulltime and attend conferences to share my work all over the
country. I sincerely thank you Richard for all your advice, collaboration, and support over
these past five years.

I would like to thank the National Science Foundation for the Graduate Research
Fellowship and the research grant that funded my work. The financial and academic
support provided me with the opportunity to dedicate time to my research and travel to
present my work at various conferences.

I wish to express my appreciation to my dissertation committee, Prof. Richard
Brutchey, Prof. Travis Williams, Prof. Mark Thompson, Prof. G. K. Surya Prakash, and
Prof. John Platt. You have all been instrumental in my education and growth as a chemist.
Additionally, I want to recognize Prof. Travis Williams for his collaboration and
contribution on several projects and publications. I am truly grateful for the assistance and
challenges that you all have posed to me in classes, the screening, and qualifying exams. I
am also deeply appreciative to my collaborators, Prof. Aaron Rossini and Yunhua Chen at
Iowa State University for your excellent solid-state NMR spectroscopy work. I have
thoroughly enjoyed the projects and noteworthy data that we have been able to observe
together. I also would like to recognize Prof. Susan Kauzlarich and Dr. Katayoon
Tabatabaei from UC Davis for their collaboration and preparation of nanocrystals for us to
study. I would also like to acknowledge my undergraduate advisor at UCSB, Prof. Ram
Seshadri who gave me the opportunity to work in his research lab and inspired me to pursue
chemistry and attend graduate school.

I would next like to thank the Brutchey group–past and present–for making my
graduate school experience a rewarding and fun experience. I want to begin by thanking
my former labmate, mentor, and now dear friend, Emily Roberts, who I met at Open House
weekend and was a major influence on my decision to attend USC. I’ve treasured all our
adventures–Hawaii, Boston, Honduras, and Berkeley–and look forward to many more. I
know I’ve made a lifelong friend. I next want to thank my senior labmates, Carrie
McCarthy, Patrick Cottingham, Lucia Mora, and Haipeng Lu for all your training,
guidance, and advice. Bryce Tappan, thank you for your support and for going through this
process together, I will truly miss listening to podcasts and indie music in lab. Next, I want
to acknowledge my office buddies, Kris Koskela for his riveting news updates and fresh
iii

garden veggies and Emily Williamson for her epic gym lifting sessions. I want to say thanks
to my girl, Lanja Karadaghi, who has become a great friend and person I go to often for
advice. Lastly, I want to say thank you and good luck to the newest group members, Kyle
Crans and Marissa Strumolo. I cannot wait to read your future publications!

I want to thank my UCSB gal pals – Lindsey, Carly, Kylie, Maureen, Jacie, and
Jordan. I miss seeing you all every day, but I want to thank you all for your friendships and
for studying with me over those amazing four years. To Carly and Kylie, thank you for
your constant humor and mischievous nature. To Maureen, Jacie, and Jordan, thank you
for your encouragement, deep conversations, and sincerity. Lastly, to my dear Lindsey
thank you for being my best friend and for your constant love and advice. Love you all!

To James, thank you for your constant support, advice, and patience. Your
encouragement over the past 5 years has helped me more than you could imagine. I will
never forget your patience and support during the challenges of my PhD and for your
celebration during the exciting times. Additionally, I want to thank your parents, Big Mike
and Helen for their support and encouragement. I am very thankful to have had you and
your family by my side throughout this experience.

Finally, I want to thank my family. To my supportive parents, Tom and Joan, thank
you for all your love, sacrifices, and encouragement. I honestly would not have been able
to accomplish half the things I have without you both. I am forever grateful for their
encouragement to pursue my passions. I dedicate this dissertation to you both. Thank you,
Paul, for encouraging your sister with all her nerdy adventures. To Grandpa, thank you for
being supportive and for your reminders of Glenna. I miss her every day. To Sally, thank
you for being a wonderful light and support to me and my family. I want to thank my aunts,
Lisa and Anne for being such inspiring women and for your support of my career choices.

iv

Table of Contents

Acknowledgements ii

List of Tables vii

List of Figures viii

Abstract xix

Chapter 1. Introduction 1
1.1. Quantum Dot Materials 1
1.1.1. Covalent Systems 3
1.1.1.1. Native Surface Chemistry of Covalent
Nanocrystals and Post-Synthetic Treatments 4
1.1.2. Traditional QDs (Mixed Covalent/Ionic Character) 5
1.1.2.1. Native Surface Chemistry of CdSe Nanocrystals
and Post-Synthetic Treatments 6
1.1.3. Ionic QD Materials 11
1.1.3.1. Native Surface Chemistry of CsPbBr3
Nanocrystals 13
1.1.3.2. Post-Synthetic Surface Treatments 15
1.2. Characterization Toolbox for QD Surfaces 20
1.2.1. Nuclear Magnetic Resonance Spectroscopy 20
1.2.2. Other Commonly Utilized Surface Characterization
Techniques 23
1.3. References 24

Chapter 2. Surface Coordination Chemistry of Germanium Nanocrystals
Synthesized by Microwave-Assisted Reduction in Oleylamine
2.1 Abstract 33
2.2 Introduction 33
2.3 Preparation and Characterization of Ge Nanocrystals Prepared via
Microwave-Assisted Reduction of Oleylamine 35
2.4
1
H NMR Spectroscopy to Determine the Surface Ligands of Ge
Nanocrystals 37
2.5 Ligand Exchange with Amines 40
2.6 Ligand Exchange with CTAB 44
v

2.7 Ligand Exchange with Thiol and Carboxylic Acids 46
2.8 Conclusions 52
2.9 References 52

Chapter 3. Probing the Ligand Exchange of N-Heterocyclic Carbene-Capped
Ag2S Nanocrystals with Amines and Carboxylic Acids
3.1 Abstract 56
3.2 Introduction 57
3.3 Preparation and Characterization of NHC-Ag2S Nanocrystals 58
3.4
1
H NMR Spectroscopy to Determine the Surface Ligands of Ag2S
Nanocrystals 60
3.5 Ligand Exchange with Primary Amines 64
3.6 Ligand Exchange with Carboxylic Acids 69
3.7 DFT Analysis of Ligand Exchange 73
3.8 Conclusions 75
3.9 References 76

Chapter 4. Quantifying the Thermodynamics of Ligand Binding to CsPbBr3
Quantum Dots
4.1 Abstract 78
4.2 Introduction 78
4.3 Preparation and Characterization of CsPbBr3 Quantum Dots 80
4.4
1
H NMR Spectroscopy to Determine the Surface Ligands of
CsPbBr3 Quantum Dots 81
4.5 DOSY NMR Spectroscopy 83
4.6 Ligand Exchange with 10-Undecenoic Acid and 10-
Undecenylphosphonic Acid 84
4.7 Ligand Exchange with 10-Undecenamine 89
4.8 Conclusions 93
4.9 References 93

Chapter 5. Surface Termination of CsPbBr3 Perovskite Quantum Dots
Determined by Solid-State NMR Spectroscopy 115
5.1 Abstract 96
5.2 Introduction 97
5.3 Basic Characterization of CsPbBr3 Quantum Dots with and without
10-Undecylphosphonic Acid 100
5.4 Solid-State NMR Characterization of CsPbBr3 Quantum Dots 105
vi

5.4.1
1
H{
14
N} RESPDOR Confirm the Presence of Ligands
Bound on the Surface 105
5.4.2 Double-Quantum Single-Quantum and Cross
Polarization Experiments Confirm the
1
H NMR Signal
Assignments 106
5.4.3
31
P NMR Experiment Confirms Phosphonate Ligand
Surface Binding 108
5.4.4
133
Cs and
207
Pb NMR Experiments Determine the
Surface Termination of CsPbBr3 Quantum Dots 109
5.5 Surface Termination of CsPbBr3 Quantum Dots Determined by
Experimental and Theoretical Surface Modelling 120
5.6 Conclusions 121
5.7 References 122
Bibliography 129

vii

List of Tables

Table 2.1 Summary of diffusion coefficients and calculated solvodynamic
diameters using DOSY NMR, and expected diameters. 39

Table 2.2 % Bound oleylamine from Figure 2.6. 43

Table 2.3 Summary of diffusion coefficients for oleylamine and undecenoic
acid and the solvodynamic diameters determined using DOSY NMR
for Ge nanocrystals synthesized at 250 ˚C. 49

Table 3.1 Table of SEM-EDS results for Ag2S nanocrystals. 59

1. Table
3.2 Diffusion coefficients of Ag2S nanocrystals and their solvodynamic
diameter calculated using diffusion coefficient of the ligand in parentheses and
Stokes-Einstein equation. 63

Table 3.3 Summary of diffusion coefficients of molecular species in CDCl3. 67

Table 4.1 Energy dispersive X-ray spectroscopic data for before and after ligand
exchange for oleic acid and dodecylamine capped QDs (OAc) and
oleylamine and lauric acid capped QDs (OAm) titrated with 10-
undecenoic acid (UDAc), 10-undecenylphosphonic acid (UDPAc),
and undec-10-en-1-amine (UDAm). Average atomic percent is
presented from three different randomly chosen locations on the
sample. 87

Table 5.1 Statistical probability of
207
Pb isotopomers in the Cs termination with
Cs substitution model. 117

Table 5.2 Statistical probability of
207
Pb isotopomers in the Cs termination
without Cs substitution model. 118

Table 5.3 Statistical probability of
207
Pb isotopomers in Pb termination model
(ammonium on Br
-
). 118

viii

List of Figures

Figure 1.1 (A) Examples of several ligand exchange reactions. (B) The
coordination of different types of ligands in Green’s formulation to
metal-chalcogenide nanocrystals. R is an alkyl group; Bu is n-butyl. 2

Figure 1.2 Surface photovoltage spectroscopy data obtained on dodecanethiol
(red) and oleylamine (blue) terminated Ge nanocrystals. The inset
depicts the expanded version of the SPV spectra obtained on
oleylamine-terminated Ge nanocrystals. 5

Figure 1.3 (a) NHC dimer precursor for ligand exchange, where X = Cl, Me, or
H. (b) Plot of ΔR (the change in the apparent excitonic radius of the
QDs) versus the number of equivalents of the XNHC ligands used (red
squares – MeNHC, green triangles – ClNHC, blue circles –HNHC). 9

Figure 1.4 Proposed energy landscape for CdX2-treated CdSe nanocrystals. An
estimation of the electron trap state energies relative to the CdSe
conduction band minimum is taken from the energies of the pre-band
gap features observed in SPV spectra. The energy axis on the figure
is not drawn to scale. 10

Figure 1.5 (a) Photocurrent response for ligand-exchanged stibanate-capped
CdSe nanocrystal films heat treated to 300 °C and as-prepared CdSe
films heat treated to 150 °C. (b) UV–vis spectra of spin-cast
nanocrystal films of as-prepared (green) and ligand exchanged CdSe
(red). 10

Figure 1.6 (a) TEM micrograph of 0-D CsPbBr3 nanocrystals. (b) Colloidal
suspensions of CsPbX3 (X=Cl, Br, I) nanocrystals under UV
excitation. (c) PL spectra of CsPbX3 (X=Cl, Br, I) nanocrystal
suspensions. 12

Figure 1.7 Size-dependent anion/lead ratio of cuboidal CsPbX3 nanocrystals. The
inset shows the anion/lead ratio for experimentally observed
nanocrystal sizes. 14

Figure 1.8 (a) Absorbance and PL spectra of Cs-oleate (red) and DDAB (green)
capped CsPbBr3 nanocrystals. (b) Evolution of PLQY for the two
types of CsPbBr3 nanocrystals. (c) Model of tetramethylammonium
bromide sitting in the A-site of CsPbBr3. 16

Figure 1.9 (a) Graphical depiction of as-prepared CsPbBr3 nanocrystals capped
with traditional long-chain ligands (oleate or bromide and
oleylammonium), and (b) with zwitterions containing both cationic
ix

and anionic groups in one molecule. (c) PLQY of CsPbBr3
nanocrystals terminated with 3-(N,N-
dimethyloctadecylammonio)propanesulfonate (green) and
oleylammonium oleate (black) ligands after a two-step of purification
on day 1 and after day 28. 18

Figure 1.10 NOESY spectrum of the sample purified with acetone in CDCl3. 22

Figure 2.1 Powder XRD patterns of Ge nanocrystals prepared at various
temperatures (210 ˚C, 230 ˚C, 250 ˚C, and 270 ˚C) compared to the
reference pattern (PDF #04-0545) showing the (111), (220, (311),
(400), and (331) reflections of cubic Ge. 36

Figure 2.2 TEM micrographs and size histograms of Ge nanocrystals prepared at
(a) 210 ˚C, (b) 230 ˚C, (c) 250 ˚C, (d) 270 ˚C. 36

Figure 2.3 Representative FT-IR spectra of pure oleylamine (OAm) and
oleylamine-capped Ge nanocrystals prepared at various synthetic
temperatures (210, 230, 250, and 270 ˚C) dispersed in hexanes and
dispensed onto the attenuated total reflection (ATR)-crystal followed
by drying. 37

Figure 2.4 Room-temperature 600 MHz
1
H NMR spectra and fitting of the bound
(d » 5.64 ppm) and physisorbed/free (d » 5.59 ppm) peaks of the
alkenyl region that show the peak broadening and change in chemical
shift associated with ligand binding of oleylamine (OAm), which is
more prominent for nanocrystals synthesized at higher temperatures.
The
1
H NMR spectrum of free oleylamine in toluene-d8 is given for
comparison. 38

Figure 2.5 Room-temperature 600 MHz
1
H NMR spectra of Ge nanocrystal
suspensions in toluene–d8 (7 mg/mL) for samples synthesized at (a)
230 ˚C, (b) 250 ˚C, and (c) 270˚C. The as-synthesized nanocrystals
are capped with native oleylamine (OAm) ligands and the suspensions
are titrated with increasing amounts of undeceneamine (UAm) (0-62
mM). 41

Figure 2.6 Variable-temperature
1
H NMR spectra of 7 mg/mL Ge nanocrystal
suspensions capped with oleylamine, titrated with 1:3 (mol/mol)
undeceneamine to oleylamine in toluene-d8. 43

Figure 2.7
1
H NMR spectra of as-synthesized Ge nanocrystals prepared at 250
˚C (orange), titrated with 1:2 (mol/mol) dodecylamine (DAm) to
oleylamine (cyan), and free dodecylamine (pink) in toluene-d8. 44

x

Figure 2.8 (a) Full
1
H NMR spectra and (b) alkenyl region of as-synthesized Ge
nanocrystals prepared at 250 ˚C (pink) and titrated with 6.3 mM
CTAB in dichloromethane-d2. 45

Figure 2.9 (a) Room-temperature 600 MHz
1
H NMR spectra of Ge nanocrystal
suspension in toluene-d8 (7 mg/mL). The as-synthesized nanocrystals
are capped with native oleylamine ligands and the suspension is
titrated with increasing amounts of free undecenethiol (UTh) (0-30
mM). (b) Superimposed room-temperature
1
H NMR spectra of the
suspension before and after heating at 90 ˚C for ca. 15 min, which
shows both free (F) and a small fraction of bound (B) peaks for
undecenethiol (d » 5.13 and 5.92 ppm). 47

Figure 2.10 Room-temperature 600 MHz
1
H NMR spectra of Ge nanocrystal
suspension in toluene-d8 (7 mg/mL). The as-synthesized nanocrystals
are capped with native oleylamine ligands and the suspension is
titrated with increasing amounts of undecenoic acid (UAc) (0-50
mM). 48

Figure 2.11 600 MHz
1
H NMR spectra of 7 mg/mL Ge nanocrystal suspensions
capped with oleylamine, titrated with 50 mM undecenoic acid in
toluene-d8 at room temperature (black), heated to 90 ˚C and cooled
back to room temperature for ca. 10 min (blue), and heated to 90 ˚C
for 14 h and cooled back to room temperature (purple). The blue and
purple spectrum shows the presence of strongly bound undecenoic
acid after cooling (ca. 5.14 and 5.93 ppm). 49

Figure 2.12 (A) Full
1
H NMR spectrum and (B) alkenyl region of as-synthesized
Ge nanocrystals prepared at 250 ˚C (black) and titrated with 1.2 mM
aqueous HCl heated to 80 ˚C for 4 h (blue) in toluene-d8. The HCl
solution was prepared in D2O and 5 µL of the HCl/D2O solution was
added. 51

Figure 2.13 2D
1
H-
13
C HSQC spectrum of Ge nanocrystal suspension synthesized
at 250 ˚C and titrated with 1.2 mM aqueous HCl in toluene-d8. This
spectrum does not show a cross peak for the N-H proton resonance (d
» 5.18 ppm), demonstrating that these protons are not attached to an
oleyl carbon. 51

Figure 3.1 (a) Schematic of the NHC ligand used. (b) TEM micrograph of NHC-
Ag2S nanocrystals with a measured average size of 7.5 ± 1.2 nm. (c)
Powder X-ray diffraction pattern of NHC-Ag2S nanocrystals, with the
stick pattern of monoclinic Ag2S given below (PDF no. 00-014-0072).
59

xi

Figure 3.2 TGA traces of NHC-Ag2S (blue), oleylamine-Ag2S (pink), and oleic
acid-Ag2S (purple) nanocrystals. 61

Figure 3.3 (a) Solution
1
H NMR spectra of 7.5-nm NHC-Ag2S nanocrystal
suspension (30 µM) and NHC-AgBr in CDCl3 (denoted by *).
Residual octadecene (ODE) reaction solvent is present after
nanocrystal synthesis and purification (denoted by ¡) with a 0.6 mM
ferrocene standard (denoted by +). About 30% of the ligands present
in the nanocrystal suspension are protonated carbene (C2-proton
resonance denoted by D). (b) 2D DOSY NMR spectrum of 7.5-nm
NHC-Ag2S nanocrystal suspension (30 µM) in CDCl3. 61

Figure 3.4
1
H NMR spectra of (a) NHC-Ag2S nanocrystals as reported, and (b)
NHC-Ag2S nanocrystals prepared using carefully dried solvents and
dispersed in dried CDCl3. 62

Figure 3.5 Selective presaturation of
1
H NMR for NHC-Ag2S nanocrystals. (a)
1
H NMR spectrum of NHC-Ag2S as-is without saturation. Selective
presaturation of
1
H NMR spectra of (b) proton resonance at d = 8.07
ppm, (c) the proton resonance at d = 11.65 ppm, (d) proton resonance
at d = 4.63 ppm, (e) the proton resonance at d = 7.68 ppm, and (f) a
duplicate for the saturation of the proton resonance at d = 8.07 ppm.
64
Figure 3.6 (a) Alkenyl region of solution
1
H NMR spectra of 7.5-nm NHC-Ag2S
nanocrystals (30 µM) dispersed in CDCl3 and titrated with 0.3-1.5
µmol oleylamine. (b)
1
H NMR spectra of NHC-Ag2S nanocrystals
(blue) and oleylamine-Ag2S nanocrystals after forced ligand exchange
(pink) dispersed in CDCl3 (denoted by *).
65
Figure 3.7
1
H NMR spectra of 7.5-nm NHC-Ag2S nanocrystal suspension,
titrated with 0.3-1.5 µmol oleylamine (top) and oleylamine (bottom)
in CDCl3. The alkenyl region of the spectra where oleylamine is
titrated into the colloidal. Nanocrystal suspension is shown as an
inset. 66

Figure 3.8 (a)
1
H NMR spectra of oleylamine in toluene-d8 (top) and CDCl3
(bottom). (b) Alkenyl region of
1
H NMR spectra in toluene-d8 (top)
and CDCl3 (bottom) demonstrating that cis-/trans-isomers are
distinguishable in CDCl3, but not toluene-d8. 66

Figure 3.9
1
H NMR spectra of 7.5-nm NHC-Ag2S nanocrystal suspension,
titrated with 1-5 µmol butylamine (top) and butylamine (bottom) in
CDCl3. The -proton region of the spectra where butylamine is
titrated into the colloidal nanocrystal suspension is shown as an inset.
67
xii

Figure 3.10 Powder X-ray diffraction patterns of NHC-Ag2S (blue), oleylamine-
Ag2S (pink), and oleic acid-Ag2S (purple) nanocrystals. 69

Figure 3.11 (a) Alkenyl region of solution
1
H NMR spectra of 7.5-nm NHC-Ag2S
nanocrystals (30 µM) dispersed in CDCl3 and titrated with 0.3-1.6
µmol oleic acid. (b)
1
H NMR spectra of NHC-Ag2S nanocrystals
(blue) and oleic acid-Ag2S nanocrystals after forced ligand exchange
(purple) dispersed in CDCl3 (denoted by *). 70

Figure 3.12
1
H NMR spectra of 7.5-nm NHC-Ag2S nanocrystal suspension,
titrated with 0.3-1.6 µmol oleic acid (top) and oleic acid (bottom) in
CDCl3. The alkenyl region of the spectra where oleic acid is titrated
into the colloidal nanocrystal suspension is shown as an inset. 71

Figure 3.13
1
H NMR spectra of 7.5-nm NHC-Ag2S nanocrystal suspension,
titrated with 1.7-8.7 µmol acetic acid in CDCl3. The methyl resonance
region of the spectra for acetic acid being titrated into the colloidal
nanocrystal suspension is shown as an inset. 72

Figure 3.14
1
H NMR spectra of 7.5-nm oleic acid-Ag2S nanocrystal suspension
(top) and free oleic acid (bottom) in CDCl3 (bottom, black). 73

Figure 3.15 DFT results providing minimum energy structures of ligand-capped
Ag12S6 clusters, formation energies, and cluster-ligand bond distances
for model NHC, amine, and carboxylic acid ligands. Model
visualizations are created using the Envision package.
[1]
74

Figure 3.16 Energy decomposition analysis of ligand-Ag12S6 cluster interactions.
The total interaction energies, obtained as the sum of the frozen,
polarization, and charge transfer terms, are -179.6 kJ/mol, -113.7
kJ/mol, and -81.4 kJ/mol for NHC, amine, and carboxylic acid
ligands, respectively. 75

Figure 4.1 XRD patterns of drop-cast films of CsPbBr3 QDs synthesized with
oleic acid (OAc) and dodecylamine (DAm) [blue] and CsPbBr3 QDs
synthesized with oleylamine (OAm) and lauric acid (LAc) [pink]. 80

Figure 4.2 TEM micrographs of CsPbBr3 QDs (a) OAm and OAc coated QDs
synthesized with ODE as solvent (as reference), (b) OAc and DAm
capped QDs synthesized with DPE as solvent, and (c) OAm and LAc
capped QDs synthesized with DPE as solvent. 81

Figure 4.3
1
H NMR spectrum of 12.4-nm CsPbBr3 QDs (1.6 mM) synthesized
with oleic acid (OAc) and dodecylamine (DAm) and dispersed in
xiii

toluene-d8 (denoted by *) with a 0.3 µM ferrocene standard. Residual
diphenyl ether (DPE) reaction solvent is present after purification. The
individual
1
H NMR spectra of oleic acid and dodecylamine in toluene-
d8 are given for comparison. 82

Figure 4.4 (a) Selective presaturation of
1
H NMR for the solvent and alkenyl
region of oleic acid and dodecylamine capped CsPbBr3 QDs (pink is
1
H NMR spectrum as is, orange is for the saturation of free oleic acid,
green is for the saturation of physisorbed oleic acid, blue is for the
saturation of the bound oleic acid). (b) Zoom in on the alkenyl region
with increased intensities to show the bound and physisorbed peaks
more clearly. 83

Figure 4.5 (a) Room-temperature
1
H NMR spectra of 1.6 mM CsPbBr3 QD
suspension possessing oleic acid (OAc) and dodecylamine native
ligands, titrated with increasing amounts (0–2.9 µM) of 10-
undecenoic acid (UAc) in toluene-d8, showing both free (F) and bound
(B) fractions. (b) Van‘t Hoff plot of 6.1 mM and 3.2 mM CsPbBr3 QD
suspension with 5.1 µM and 4.7 µM 10-undecenoic acid, respectively,
in toluene-d8 at temperatures ranging from 283 K to 325 K. (c)
1
H
NMR spectrum of the alkenyl region of an oleic acid and
dodecylamine capped CsPbBr3 QD suspension with the different
peaks labeled (i.e., bound, physisorbed, and free). (d)
1
H NMR
spectrum of the alkenyl region of an oleylamine and lauric acid capped
CsPbBr3 QD suspension with the different peaks labeled (i.e.,
bound/physisorbed and free). 85

Figure 4.6 Plots of [OAc]F[UDAc]B vs. [OAc]B[UDAc]F. The slope of this plot
is used to determine an average Keq for the ligand exchange between
oleic acid (OAc) and 10-undecenoic acid (UDAc) based on the
equilibrium equation (
!"
=
[$%&]
!
[()%&]
"
[$%&]
"
[()%&]
!
). 86

Figure 4.7 UV-vis absorption (solid line) and photoluminescence (dashed line)
of (a) QD suspensions synthesized with oleic acid and dodecylamine
before and after titration with 10-undecenoic acid. (b) CsPbBr3 QD
suspensions synthesized with oleic acid and dodecylamine before and
after titration with 10-undecenylphosphonoic acid. (c) CsPbBr3 QD
suspensions synthesized with oleylamine and lauric acid before and
after titration with undec-10-en-1-amine. All photoluminescence
spectra were collected at an excitation wavelength of 440 nm. 87
Figure 4.8
1
H NMR spectra of 1.8 mM CsPbBr3 QD suspension possessing oleic
acid (OAc) and dodecylamine native ligands, titrated with increasing
amounts (0–1.9 µM) of 10-undecenylphosphonic acid (UPAc) in
toluene-d8, showing both free (F) and bound (B) fractions. 88

xiv

Figure 4.9
1
H NMR spectrum of 12.9-nm CsPbBr3 QDs (2.5 mM) synthesized
with oleylamine (OAm) and lauric acid (LAc) and dispersed in
toluene-d8 (denoted by *) with a 0.3 µm ferrocene standard. Residual
diphenyl ether (DPE) reaction solvent is present after purification. The
individual
1
H NMR spectra of oleylamine and lauric acid in toluene-
d8 are given for comparison. 89

Figure 4.10 (a) Room-temperature
1
H NMR spectra of 2.5 mM CsPbBr3 QD
suspension possessing oleylamine (OAm) and lauric acid native
ligands, titrated with increasing amounts (0–8.0 µM) of undec-10-en-
1-amine (UAm) in toluene-d8, showing both free (F) and bound (B)
fractions. (b) Van‘t Hoff plot of 4.1 mM and 2.0 mM CsPbBr3 QD
suspension with 7.7 µM and 8.0 µM undec-10-en-1-amine,
respectively, in toluene-d8 at temperatures ranging from 283 K to 325
K. 91

Figure 4.11 UV-vis absorption (solid line) and photoluminescence spectra
(dashed) of CsPbBr3 QDs synthesized with oleic acid and
dodecylamine before and after titration with oleylamine. 92

Figure 5.1 Idealized models of the surface termination of the as-synthesized
orthorhombic CsPbBr3 QDs. Cs and Br atoms are depicted by blue
and yellow spheres, respectively. Pb atoms reside at the center of the
red octahedra formed by Br. The QDs consist of an inorganic core
with CsBr (left) or PbBr2 (right) surface termination. Substitution of a
surface Cs atom with an ammonium ligand gives an ABr terminated
surface (A = Cs or DDA, middle left). Both surfaces are capped by
cationic and anionic organic ligands at the outermost layer. The
anionic X-type oleate and 10-undecenylphosphonate ligands are
assumed to bind to exposed Cs or Pb surface atoms. 99

Figure 5.2 Room-temperature solution
1
H NMR spectra of CsPbBr3 QDs with
10-undecylphosphonic acid (green) and without 10-
undecylphosphonic acid (yellow). The QDs were dispersed in toluene-
d8. The broad resonances centered around d = 5.65 ppm and the sharp
resonances centered at d = 5.54 ppm, correspond to bound and
physisorbed oleate/oleic acid, respectively. The spectrum of CsPbBr3
QDs with undecylphosphonic acid (green) also shows two bound
undecylphosphonate resonances (d = 5.2 and 6.0 ppm) and indicates
the absence of free undecylphosphonic acid. Quantifying the
concentration of bound oleate ligands gives a bound fraction that
amounts to 98% of the total oleic acid present in the system before
addition of undecylphosphonic acid (yellow), which decreases to 87%
upon addition of undecylphosphonic acid (green). The
1
H NMR
xv

spectra of free undecylphosphonic acid (UDPA) and oleic acid in
toluene-d8 are also shown (black). 101

Figure 5.3 (A) MAS
1
H spin echo solid-state NMR spectra of CsPbBr3 QDs with
and without UDPA. The insets show the diagnostic high-frequency
chemical shifts of the vinyl functional groups of UDPA, alkenyl
protons of oleate (OA), and the ammonium group of
dodecylammonium (DDA).
1
H detected 2D dipolar
1
H–
13
C CP-
HETCOR spectra of CsPbBr3 QDs (B) without and (C) with UDPA.
(D) 2D dipolar
1
H →
31
P CP HETCOR of CsPbBr3 QDs with UDPA.
The CP contact time is indicated. All spectra were obtained with a 25
kHz MAS frequency. 102

Figure 5.4 Powder XRD patterns of drop-cast films of CsPbBr3 QDs with 10-
undecylphosphonic acid (green) and without 10-undecylphosphonic
acid (yellow), with the stick pattern for orthorhombic CsPbBr3
provided in black below. 103

Figure 5.5 TEM micrographs of (a) CsPbBr3 QDs without 10-
undecylphosphonic acid and (b) with 10-undecylphosphonic acid. The
average particle size and distribution is indicated on the image.
Particle size distributions were extracted from the images with
ImageJ. 104

Figure 5.6 UV-vis absorption (solid line) and photoluminescence (dashed line)
spectra of CsPbBr3 QD suspensions with (green) and without (yellow)
10-undecylphosphonic acid (UDPA) ligand exchange. The absorption
spectrum of CsPbBr3 QDs with UPDA was normalized to the optical
density of the spectrum of CsPbBr3 QDs without UDPA at 475 nm,
and this normalization ratio was used to qualitatively compare the
photoluminescence intensities. Both photoluminescence spectra were
collected at an excitation wavelength of 400 nm. 104

Figure 5.7
1
H-
14
N RESPDOR spectra and curves for CsPbBr3 QDs. (A, B)
1
H
NMR spectra corresponding to the S and S0 with a 1.0 ms recoupling
time. (C, D) DS/S0 fractions for CsPbBr3 QDs (C) without and (D)
with 10-undecylphosphonic acid. The best-fit distances to the surface
are marked in accordance with the spectra. The CH2 signals are
partially saturated and inverted (see inset) because a selective
saturation pulse was applied prior to the RESPDOR pulse sequence.
This was necessary in order to reduce the CH2
1
H NMR signal and
prevent it from interfering with measurement of the dipolar dephasing
of the ammonium
1
H NMR signal. 105

xvi

Figure 5.8 2D
1
H-
1
H dipolar double-quantum single-quantum (DQ-SQ)
homonuclear correlation spectra of CsPbBr3 QDs (A) without and (B)
with UDPA. The diagonal black dashed lines indicate the
autocorrelation line. The spectra were obtained using the BABA pulse
sequence. 107

Figure 5.9 Direct detected
1
H→
13
C CPMAS solid-state NMR spectra of CsPbBr3
QDs with and without UDPA. 107

Figure 5.10
133
Cs spin echo and 2D
1
H →
133
Cs CP-HETCOR NMR spectra of
CsPbBr3 QDs (A) without and (B) with UDPA.
207
Pb spin echo NMR
spectra and 2D
207
Pb →
1
H CP-HETCOR of CsPbBr3 QDs (C)
without and (D) with UDPA. The CP contact times were 9 and 8 ms
for
133
Cs and
207
Pb CP-HETCOR experiments, respectively.
The
207
Pb NMR signals at 500 and 850 ppm are not real and are
from t1-noise. 110

Figure 5.11
133
Cs spin-echo solid-state NMR spectra and surface-selective 2D
dipolar
1
H→
133
Cs CP-HETCOR spectra of CsPbBr3 QDs (A) without
and (B) with 10-undecylphosphonic acid. Both spectra were obtained
with a 2 ms CP contact time, while those shown in the main text were
obtained with 8 or 9 ms CP contact times. 111

Figure 5.12 Comparison of
133
Cs projections of surface-selective 2D dipolar
1
H→
133
Cs CP-HETCOR spectra of CsPbBr3 QDs with and without
10-undecylphosphonic acid. CP contact times were 2 ms or 9 ms and
are indicated on the spectra. The
133
Cs NMR signals assigned to the
surface Cs atoms show enhanced relative intensities in the spectra
obtained with short contact times, consistent with their closer
proximity to
1
H spins from the surface ligands. 111

Figure 5.13 (A–D) Structural models of the orthorhombic (010) CsPbBr3 surface
used to simulate the
1
H{
133
Cs}/
1
H{
207
Pb} multispin RESPDOR/S-
REDOR curves. Each blue sphere corresponds to a Cs atom while a
blue atom with a purple halo around it indicates a Cs atom in the
subsurface layer. The ammonium H atom is assumed to be directly
above the central atom, in the position indicated by an asterisk. The
experimental ΔS/S0 intensities were plotted as a function of total
recoupling time for CsPbBr3 QDs without and with UDPA surface
ligands. Blue points and orange points correspond to the
1
H{
133
Cs}
RESPDOR and
1
H{
207
Pb} S-REDOR experiments, respectively. Blue
and orange lines are simulated dephasing curves for the
1
H–
133
Cs
and
1
H–
207
Pb spin systems, respectively. The best fit distances are
indicated on the plots and structural models. 113

xvii

Figure 5.14 (A)
1
H{
133
Cs} RESPDOR spectra of CsPbBr3 QDs with UDPA. The
S and S0
1
H NMR spectra are shown for 0.5, 1.0, 2.0, 3.0, and 4.0 ms
total recoupling time, respectively. DANTE pulse trains
9, 10
were used
to selectively excite the high-frequency ammonium
1
H NMR signals
and minimize signals from other
1
H spins. (B)
1
H{
133
Cs} RESPDOR
NMR spectra for CsPbBr3 QDs with UDPA with a 1.0, 2.0, 3.0, and
4.0 ms total recoupling time, respectively. A DANTE pulse train was
used to selectively excite the ammonium
1
H NMR signals at 7 ppm.
However, the other
1
H NMR signals are still visible because the
DANTE pulse train is not perfectly selective. The alkene
1
H NMR
signal from oleate becomes increasingly intense and partially
contributes to
1
H NMR signal from the ammonium groups which was
monitored for dipolar dephasing. This partly causes DS intensities to
appear smaller at longer recoupling times. (C) The model of the
orthorhombic CsPbBr3 surface with Cs termination with Cs
substitution used to simulate the
1
H{
133
Cs}/
1
H{
207
Pb} multi-spin
RESPDOR curves. (D) Experimental
1
H{
133
Cs} and
1
H{
207
Pb}
RESPDOR dephasing curves for CsPbBr3 QDs with 10-
undecenylphosphonate (UDPA) surface ligands. Calculated
1
H{
133
Cs} RESPDOR multi-spin dipolar dephasing curves (solid
lines) are shown for different distances of the
1
H spin above and below
the Cs substitution position. Each blue circle corresponds to a signal
build-up point in
1
H{
133
Cs} RESPDOR experiments while each
orange circle corresponds to a point in
1
H{
207
Pb} RESPDOR
experiments. The calculated
1
H{
207
Pb} RESDOR curve for a distance
of 0 Å is also shown. 116

Figure 5.15 (Left panels) The model of the orthorhombic CsPbBr3 surface with Cs
termination without Cs substitution used to calculate the
1
H{
133
Cs}/
1
H{
207
Pb} multi-spin RESPDOR curves. Experimental
1
H{
133
Cs} and
1
H{
207
Pb} RESPDOR dephasing curves for CsPbBr3
QDs with 10-undecenylphosphonate (UDPA) surface ligands. (Right
panels) Calculated
1
H{
133
Cs} and
1
H{
207
Pb} RESPDOR multi-spin
dipolar dephasing curves (solid lines) are shown for different
distances of the
1
H spin above a surface Br atom. Each blue circle
corresponds to a signal build-up point in
1
H{
133
Cs} RESPDOR
experiments while each orange circle corresponds to a point in
1
H{
207
Pb} RESPDOR experiments. 120

Figure 5.16 Models of the (010) Cs and Br terminated surface of orthorhombic
CsPbBr3 with the ammonium
1
H spin substituted into the central Cs
position indicated by an asterisk.
1
H{
133
Cs} and
1
H{
207
Pb} multi-spin
RESPDOR curves are shown for models with different numbers of
surface cesium atoms. The indicated DDA ligand densities were
calculated by assuming each Cs vacancy is filled by a
xviii

dodecylammonium ligand. This figure illustrates that the
1
H{
133
Cs}
RESPDOR curves are compatible with dodecylammonium densities
between 0.71 and 2.13 dodecylammonium nm
-2
; at higher ligand
densities the
133
Cs dipolar dephasing is significantly reduced. Each
blue sphere corresponds to a Cs atom while the purple halo around it
indicates a Cs atom in the second, sub-surface layer. 121

xix

Abstract
The surface chemistry of nanocrystals is a critical component of the quantum dot
construct, due to its contribution to both the colloidal and chemical stability of quantum
dots and therefore, its underlying optoelectronic properties. In Chapter 1, an introduction
to background and previous work on the surface chemistry of various quantum dots
spanning from covalent to ionic will be presented. Beyond, we will present surface chemistry
investigations performed on various QD materials. In Chapter 2, we will discuss our work on
covalent Ge nanocrystals, in which we tracked ligand exchange reactions via
1
H NMR
spectroscopy and determined that the nanocrystals were coordinated by strongly bound oleylamide
ligands, with covalent X-type Ge-alkyl amide bonds. Our work extended into studying the surface
chemistry of mixed covalent/ionic systems with interesting native L-type carbene ligands on Ag2S
QDs. Finally, we will conclude with our work on the highly ionic CsPbBr3 QDs, which utilized
both solution and solid-state NMR spectroscopy to gain insight into the ligand binding and surface
termination of carboxylate and ammonium-terminated CsPbBr3 quantum dots that have a surface
binding or X2-type. The binding strength of ligands tends to decrease with increasing ionicity of
the semiconductor, which makes room temperature ligand exchange reactions facile to study for
ionic semiconductors with NMR spectroscopy, however, in contrast for highly covalent
semiconductors with covalently bound ligands, are much more difficult to exchange with new
ligands at room temperature. Overall, we have found that the ligand binding of covalent, mixed,
and ionic semiconductors tends to be covalent, mixed, or ionic in nature, respectively.

Chapter 1. Introduction

1.1. Quantum Dot Materials
Bulk semiconductors have proven useful in various optoelectronic applications, including
photovoltaics,
[2–4]
photodetectors,
[5]
and light emitters.
[6]
While bulk semiconductors are heavily
researched, their nanoscale counterparts have gained large attention due to the interesting physics
that result with their small size. Properties such as size-tunable or composition-tunable band gaps
make zero-dimensional quantum dots (QDs), or semiconductor nanocrystals, of great interest for
many of the same optoelectronic applications.
[7–10]
Quantum dots are a comprised of an inorganic
core and an organic ligand shell, the latter of which contributes to both the colloidal and chemical
stability of the QD and its underlying optoelectronic properties, making it a critically important
component of the QD construct.
Despite the advantageous properties of QDs, there are still significant challenges with
respect to their utilization, including the presence of surface trap states, which can cause
photoluminescence quantum yield (PLQY) losses and blinking, and the insulating ligand shell that
impedes charge transport in QD thin films.
[11,12]
Fundamental studies into the surface coordination
chemistry of colloidal QDs have enabled our ability to rationally choose ligands or perform ligand
exchange reactions that passivate surface trap states and/or enable charge transport.
[13]

The covalent bond classification is a conventional concept used in organometallic
chemistry to categorize the bonding between metals and ligands.
[14,15]
This nomenclature was first
translated from Malcom L. H. Green’s classification of organometallic bonding to describe
nanocrystal-ligand interactions by Owen in 2015 (Figure 1.1).
[16]
Ligands are labelled as L-, X-,
or Z-type ligands and are dependent on the number of electrons (2, 1, or 0, respectively) that the
1
2

neutral ligand donates to the nanocrystal-ligand bond. Note that the word “nanocrystal” was
carefully selected because nanocrystal surfaces are more complex and can often feature multiple
adsorption sites, including metal and non-metal atoms. L-type ligands are Lewis bases and tend to
coordinate to surface metal atoms through a dative bond, whereas Z-type ligands are Lewis acids
and typically coordinate to surface non-metal atoms. X-type ligands can interact with either metal
or nonmetal atoms and are more dependent on the binding affinity of the ligand to the coordinated
atom. The L-, X-, Z- nomenclature will be used throughout this dissertation.

Figure 1.1. (A) Examples of several ligand exchange reactions. (B) The coordination of different types of
ligands in Green’s formulation to metal-chalcogenide nanocrystals. R is an alkyl group; Bu is n-butyl.
Adapted with permission from ref [16]. Copyright (2015) Science.

There are several ways to classify material systems of colloidal QDs, but herein, QDs will
be divided into one of three categories based upon their relative ionicity: covalent, intermediate,
and ionic systems. The crystal structure of covalent semiconductors is composed of covalent bonds
that are symmetrical and identical, with silicon and germanium being two purely covalent
semiconductors. Due to the covalent nature of the crystal structure, there is also a covalent nature
in the surface and ligand interaction, yielding tightly bound ligands. Intermediate, or traditional,
semiconductors (e.g., II-VI and IV-VI binaries) have mixed covalent and ionic character. Some
3

examples of traditional semiconductors include CdSe QDs that have a rich history and will be
explored in more detail below. Lastly, there are ionic QDs, such as the newer lead halide
perovskites, that have highly ionic bonding within the structure and therefore, have highly ionic
binding with ligands as well.

1.1.1. Covalent Systems
First discovered in 1824 by Swedish chemist, Jöns Jacob Berzelius, silicon has become the
most common commercially used semiconductor material used in solar cells and electronic
chips.
[17]
Silicon has an indirect bandgap of 1.1 eV and earth abundance ca. 28%.
[17]
Silicon QDs
have received attention because their size-dependent optoelectronic properties give rise to
applications for optoelectronic devices and light emitting devices.
[18–21]
More specifically, Si QDs
are tunable to emit from the ultraviolet (UV) to the infrared (IR). Due to their biocompatibility,
photoluminescence, and resistance to photobleaching, Si QDs have the potential to be utilized as
a biological imaging agents; therefore, there have been many studies on surface functionalization
to ensure safety and excellent properties.
[22,23]
Due to silicon’s success, many researchers have
looked for Si-alternatives to improve upon existing technologies and performance. One alternative
is another group IV semiconductor, germanium, due to its superior absorption coefficient, charge
transport capabilities, large Bohr exciton radius (ca. 24 nm), and therefore, its size-tunable band
gap.
[24,25]
Germanium has a wider spectral tunability than silicon making it potentially more
versatile.
[26]
Based on these excellent properties, Ge QDs have been studied in many optoelectronic
applications, including bioimaging and solar energy conversion.
[27–30]

4

1.1.1.1. Native Surface Chemistry of Covalent Nanocrystals and Post-Synthetic Treatments
Like most QDs, silicon’s properties are highly sensitive to its surface chemistry. One of
the biggest challenges for Si QDs is the relatively poor size and surface definition. Additionally,
there is limited research on the surface chemistry of Si QDs besides hydrogen-terminated Si, which
is the most commonly utilized synthetic preparation.
[26]
More traditional alkyl organic ligands have
been utilized for Si QD preparation by Kauzlarich and coworkers.
[20,31]
The preparation involves
a metathesis reaction of magnesium silicide with silicon tetrachloride and subsequent termination
with alkyl groups, resulting in a Si–C covalent bond.
[31]
Kauzlarich also showed that Cl-terminated
Si QDs could be ligand exchanged to generate alkyl or aromatic termination (via Grignard reagent)
or terminated with aminopropyltrimethoxysilane (via annealing) to form more crystalline Si
QDs.
[32]

For Ge QDs, the most frequently utilized preparation is the high-temperature chemical
reduction of Ge(II) and Ge(IV) precursors in primary alkylamines.
[33–35]
While this ligand choice
yields air-stable Ge QDs, the long aliphatic carbon chain inhibits charge injection and transport.
As a result, ligand exchange reactions for Ge QDs are necessary, yet there remain few examples
in the literature. Kauzlarich and coworkers sought to solve the insulation-problem by ligand
exchanging these QDs with thiol ligands, which resulted in better charge separation and, therefore,
better photovoltaic response (Figure 1.2).
[24]
Similar to silicon, germanium has highly covalent
surface ligand bond interactions, which makes any sort of equilibrium induced ligand exchange
difficult at room temperature.
[36]
However, Wheeler et al. presented a slightly different binding
motif for ligands on the surface for nanocrystals prepared with a gas-plasma reactor.
[37]
This
difference in ligand binding, compared to previous reports by Kauzlarich,
[24]
is a result of the
synthetic conditions and are functionalized with cationic alkylammonium ligands, rather than
5

conventionally covalently bound ligands, thus demonstrating the importance and effect of
synthetic preparation and the “exceptions to the rule.” In order to rationally design and execute
ligand exchange reactions for Ge and other main group nanocrystals moving forward, a better
understanding of the fundamental coordination chemistry of the nanocrystal surface must be
achieved, which was the goal for Chapter 2.

Figure 1.2. Surface photovoltage spectroscopy data obtained on dodecanethiol (red) and oleylamine (blue)
terminated Ge nanocrystals. The inset depicts the expanded version of the SPV spectra obtained on
oleylamine-terminated Ge nanocrystals. Reproduced with permission from ref [24]. Copyright (2014)
American Chemical Society.

1.1.2 Traditional QDs (Mixed Covalent/Ionic Character)
Any semiconductor composed of two or more elements will have some mixed covalent and
ionic character due to the lack of symmetry in the bonding from the presence of two or more
elements with varying electronegativity. There are many QDs with the traditional or mixed
covalent/ionic character, including II-VI and IV-VI semiconductors; however, to narrow down the
large content of this classification, the surface chemistry of CdSe will be the only material explored
in this section. The surface chemistry of CdSe also has an enormous amount of literature, however,
6

this system is often used to make universal predictions for other traditional systems and is
therefore, a great example to review.
CdSe QDs have been explored in a wide variety of applications, including solar energy
conversion
[38,39]
and light-emitting diodes.
[40]
The excellent optoelectronic properties that allow
for such applications is due to its tunable band gap achieved through size control, with a Bohr
exciton radius (ca. 5.6 nm).
[41]
In general, optical properties of QDs are controlled by quantum
confinement, which refers to the delocalization of an exciton. However, the efficiency of the
transport or recombination of the electron hole pair can be negatively impacted by defects and
surface trap states. The optoelectronic properties of traditional QDs, including CdSe, are often
plagued by these trap states, making it crucial to understand and manipulate the surface chemistry
when developing high quality devices.

1.1.2.1. Native Surface Chemistry of CdSe Nanocrystals and Post-Synthetic Treatments
The two common crystal structures of CdSe can exist in either metastable zinc blende or
the thermodynamically stable hexagonal wurtzite. The two crystal structures can be obtained by
varying the reaction conditions, such as using carboxylic acid ligands to obtain zinc blende or
phosphines to obtain the wurtzite structure.
[42–44]
The surfaces of II-VI nanocrystals, including
CdSe, tend to be metal-rich and therefore require an excess of anionic ligands to maintain charge
balance.
[45–47]
Ligand exchange reactions have been shown to work for X-, L-, and Z-type ligands,
which are all discussed with examples below.
CdSe QDs are traditionally prepared with long chain X-type ligands, such as a carboxylic
acid (oleic acid). Although long chain ligands yield stable and monodisperse colloids, they are not
ideal for many optoelectronic applications. Ligand exchange reactions are performed for
7

coordinating ligands with different functional groups or chain length, which can alter the solvent
dispersability of the QDs. To understand mechanisms and principles that direct ligand exchange,
Knauf et al. monitored the ligand exchange of oleate-capped CdSe QDs with incoming oleate,
phosphonic acid, and thiol ligands with solution NMR spectroscopy.
[48]
Both 10-undecenoic acid
and 10-undecenyl phosphonic acid underwent a 1:1 exchange, however 10-undecenoic acid
underwent an equilibrium exchange (Keq = 0.83), whereas 10-undecenyl phosphonic acid
underwent irreversible exchange with the native oleate. Unlike the acids, 10-undecenethiol
irreversibly displaced oleate with a ligand exchange stoichiometry of 1:1.6 free OA to bound thiol.
Specifically, thiol showed an initially higher exchange ratio (1:2.3), suggesting thiols initially bind
to open sites on the QD surface before displacing native oleate. These results provided new insight
to X-type ligand exchange reactions for traditional QDs and demonstrated a robust method for
investigating organic ligands via solution
1
H NMR spectroscopy. Hens et al. used
1
H and
31
P NMR
to show the superior binding strength of monodentate octadecylphosphonate ligands to the surface
of wurtzite CdSe nanocrystals, as compared to oleate-capped CdSe.
[45]
Similar to Knauf and
coworkers, they discovered that phosphonic acid will readily exchange with the carboxylic acid
ligand, but the reverse does not occur at room temperature. For zinc-blende CdSe, Jiang and
coworkers showed that the equilibrium constant of the exothermic ligand exchange reaction with
the desorption of native Z-type cadmium oleate with L-type alkylamines of various lengths (i.e.,
n-butylamine, n-octylamine, and n-dodecylamine) decreases with increasing alkyl chain length.
[49]

Additionally, it is important to note that despite the desorption of cadmium oleate on the surface,
the UV-vis absorption and fluorescence spectra were almost unchanged, indicating minimal
etching from this process. Because n-butylamine had the shortest carbon chain, van der Waals
interactions between neighboring ligands were the weakest among the three alkylamine ligands
8

and resulted in a positive enthalpy, whereas the longer amines studied has a negative enthalpy.
This demonstrates that surface chemistry of ligands is dependent on both the interaction of ligands
with the surface of the nanocrystal and ligand-ligand interactions.
N-Heterocyclic carbene (NHC) ligands are another interesting class of L-type ligands that
bind to the surface through the electron pair on the carbene carbon.
[50]
NHCs are sterically and
electronically tunable. The R-groups on the 1,3-positions of the imidazole ring can be tuned with
different length alkyl chains or functional groups and the imidazole can be modified as well, such
as with benzimidazole. NHCs are a part of a group of ligands, known as exciton-delocalizing
ligands (EDLs), that couple quantum-confined electrons within semiconductor nanocrystals to
their immediate environments.
[51]
Exciton delocalization is desirable because this permits the
extraction of an excitonic hole from a QD, which allows for quick photoreduction of a QD,
migration of excitons through a QD film, and multihole photo-oxidation with QDs.
[52–54]
Weiss
and coworkers performed a ligand exchange with NHCs on CdSe QDs by introducing an NHC-
dimer to a colloidal suspension of oleate-capped CdSe at room temperature for ca. 16 h.
[53]
Upon
exchange, there is a redshift in the absorbance spectrum up to 0.111 eV, which increased with the
p-acidity of the X-group on the backbone of the imidazole (Figure 1.3). The trend in the absorbance
is consistent with the proposed mechanism for exciton delocalization due to the mixing of π-
orbitals on the carbene carbon with the QD valence band. The stable colloidal suspensions of
MeNHC and HNHC maintained 87% of their initial PL line widths, demonstrated potential
usefulness as colorometric sensors, and the lack of hole trapping by the NHC ligands also suggests
that they may greatly enhance the rates of direct charge carrier extraction to NHC-linked molecular
acceptors in photocatalytic applications.
9

Figure 1.3. (a) NHC dimer precursor for ligand exchange, where X = Cl, Me, or H. (b) Plot of ΔR (the
change in the apparent excitonic radius of the QDs) versus the number of equivalents of the XNHC ligands
used (red squares – MeNHC, green triangles – ClNHC, blue circles –HNHC). Reproduced with permission
from ref [53]. Copyright (2020) American Chemical Society.

The use of shorter ligands, such as short alkyl organic ligands (e.g., tert-butylthiol) or small
inorganic anions (e.g., halides) is a highly utilized method for producing nanocrystals for solution
processed electronics.
[55]
One challenge with ligand exchange with small anionic ligands is that
hazardous chemicals, such as pyrophoric phosphines and hydrazine, and inert atmospheres are
often needed in addition to perform successful ligand exchange. However, Brutchey et al.
demonstrated that Z-type CdX2 (X = Cl, Br, I) ligands were able to displace native ligands using
methanol and simple work up to form colloidal suspensions for producing films with shallower
trap states, particularly for CdI2 (Figure 1.4).
[56]
Another example of ligand exchange with
inorganic anions performed by Brutchey et al. demonstrated the usefulness of stibanates for ligand
capping of CdSe nanocrystals.
[57]
In this publication, they opted to use their thiol-amine solvent
system as an alternative to using hydrazine to digest Sb2S3, which produced colloidally stable
stibanate-capped CdSe nanocrystals. As a result, they found an improved interparticle coupling
10

compared to the as-prepared nanocrystals, which lead to markedly higher electrochemical
photocurrent generation (Figure 1.5).

Figure 1.4. Proposed energy landscape for CdX 2-treated CdSe nanocrystals. An estimation of the electron
trap state energies relative to the CdSe conduction band minimum is taken from the energies of the pre-
band gap features observed in SPV spectra. The energy axis on the figure is not drawn to scale. Reproduced
with permission from ref [56]. Copyright (2015) American Chemical Society.

Figure 1.5. (a) Photocurrent response for ligand-exchanged stibanate-capped CdSe nanocrystal films heat
treated to 300 °C and as-prepared CdSe films heat treated to 150 °C. (b) UV–vis spectra of spin-cast
nanocrystal films of as-prepared (green) and ligand exchanged CdSe (red). Reproduced with permission
from ref [57]. Copyright (2014) American Chemical Society.

11

1.1.3. Ionic QD Materials
[58]

The most popular highly ionic semiconductor are halide perovskites. Perovskites are any
material in the same structure type family as the mineral CaTiO3, which was discovered in 1839
and named after the Russian mineralogist Lev Perovski.
[59]
Over the past 100 years, the ABO3
oxide perovskites (A=12-coordinate cation, B=6-coordinate cation), where the [BO6] octahedra
are linked by corner sharing in three dimensions, have been intensively studied because of their
multifunctionality.
[60]
The perovskite structure class can be more broadly defined as ABX3,
[61]

where the X anion can also be a chalcogenide (e.g., BaZrS3),
[62]
a pnictide (e.g., LaWN3),
[63]
or a
halide (e.g., CsPbX3, X=Cl, Br, I).
[64]
The ABX3 halide perovskites, while known since the late
nineteenth century,
[65]
have undergone a renaissance because of the remarkable photovoltaic
performance of MAPbI3-based thin films (MA=methylammonium). Lead halide perovskite solar
absorbers have garnered massive attention because of their high power conversion efficiencies
(PCEs), currently >25%.
[66]
These performance achievements have partly been enabled by post-
deposition treatments for chemical passivation. Treatment with Lewis bases (e.g., pyridine,
[67]

phosphine oxide
[68]
) and alkylammonium halides (e.g., methylammonium iodide,
[69]

ethylammonium iodide
[70]
) help minimize non-radiative recombination and enhance device
performance; however, an atomistic understanding of how these passivants coordinate to the
surface is often lacking. The small surface area-to-volume ratio of large grained polycrystalline
halide perovskites obscure these efforts.
Colloidal lead halide perovskite nanocrystals, on the other hand, have large surface area-
to-volume ratios because of their small size, which increases the importance of their surfaces on
ensuing optoelectronic properties, making them an ideal platform for studying surface chemistry.
Approximately 20% of the atoms in a 10 nm CsPbBr3 nanocrystal are within the first surface layer
12

of the particle. The surface chemistry of semiconductor nanocrystals is affected by their exposed
crystal facets,
[71,72]
stoichiometry,
[73,74]
organic ligands,
[51,75,76]
and ionicity.
[36,37,77]
Unlike more
mixed/traditional semiconductor nanocrystals (e.g., CdSe, InP), lead halide perovskites are
markedly ionic, which affects their environmental stability,
[78]
electronic structure and defect
tolerance,
[79]
and ligand binding and fluxionality.
[80]
These unique characteristics make lead halide
perovskite nanocrystals a rich area of study, with the all-inorganic CsPbX3 nanocrystals being the
most studied because of their excellent optoelectronic properties, including size- and composition-
dependent band gaps, high PLQYs, narrow emission line widths that cover a wide color gamut,
and suppressed blinking.
[81,82]
These properties make CsPbX3 nanocrystals appealing for use in
optical devices; for example, solar cells, low threshold lasers, light emitting diodes (LEDs), single
photon emitters, and X-ray scintillators.
[83]

Kovalenko synthesized CsPbX3 nanocrystals by the hot-injection of a Cs(oleate) solution
in 1-octadecene into a solution of PbX2, oleic acid, and oleylamine at 140-200 ˚C.
[84]
This synthesis
results in morphologically well-defined cuboidal nanocrystals ranging in edge length from 4-15
nm and has become the prototypical preparation method for 0-D CsPbX3 nanocrystals. The PL
emission wavelength can be tuned by varying nanocrystal size and composition (Figure 1.6).

Figure 1.6. (a) TEM micrograph of 0-D CsPbBr 3 nanocrystals. (b) Colloidal suspensions of CsPbX 3 (X=Cl,
Br, I) nanocrystals under UV excitation. (c) PL spectra of CsPbX 3 (X=Cl, Br, I) nanocrystal suspensions.
Reproduced with permission from ref [84]. Copyright (2015) American Chemical Society.

13

1.1.3.1. Native Surface Chemistry of CsPbBr3 Nanocrystals
Understanding the surface chemistry of halide perovskite nanocrystals is crucial to
rationally improving and tuning their photophysical properties. Prior to discussing structural
features of the CsPbBr3 nanocrystal surface, it is necessary to understand the bulk structure. Using
both Rietveld refinement and pair distribution function analyses of X-ray total scattering data,
Brutchey and co-workers determined that the crystal structure of colloidal 9 nm CsPbBr3
nanocrystals is orthorhombic (Pnma)
[85,86]
up until a temperature of 50 ˚C < Tg–b < 59 ˚C, where a
transition to a tetragonal phase (P4/mbm) occurs, followed by a higher-temperature transition to
the cubic (Pm3
%
m) phase at 108 ˚C < Tb–a < 117 ˚C.
[87]
The 0-D CsPbBr3 nanocrystals obtained by
the Protesescu hot-injection method typically adopt a cuboidal morphology, whereby the (001)
and (110) facets have been predicted to be terminate the nanocrystal surface.
[79]
Given this crystal
structure and particle morphology, two types of surfaces can occur – a CsBr terminated surface
and a PbBr2 terminated surface. In recent experimental reports, high-resolution TEM (HR-TEM)
images revealed that cuboidal CsPbBr3 nanocrystals exhibit (001) and (110) crystal lattice planes
that compose the surface facets of these particles.
[88–90]
Given the high fraction of atoms that reside at the surface of nanocrystals, measuring the
particle elemental stoichiometry gives insight into the surface composition.
[73,74]
Ravi et al. applied
variable energy X-ray photoelectron spectroscopy (XPS) to probe the composition of 10 nm
CsPbBr3 nanocrystals in a layer-by-layer fashion,
[90]
and found that the Br:Cs elemental ratio is
∼5 at the surface of the CsPbBr3 nanocrystals and decreases to the expected ratio of ~3 in the bulk
which clearly suggests a surface excess of Br. However, if some of the surface Cs-sites were
occupied by alkylammonium cations a Br:Pb ratio of >3 and a Cs:Pb ratio of ~1 could arise for a
CsBr termination. Kovalenko and Infante computationally predicted that the Br:Pb ratio should be
14

ca. 3.2 for CsBr terminated CsPbBr3 nanocrystals in the size range of 7-11 nm (Figure 1.7). They
confirmed this prediction experimentally using inductively coupled plasma-optical emission
spectroscopy (ICP-OES), demonstrating that 9 nm CsPbBr3 nanocrystals give Br:Pb ratios of 3.2-
3.5.
[91]
Kovalenko and Infante also suggested that PbBr2 termination of the cuboidal CsPbBr3
nanocrystals is unlikely because it requires a much denser ligand packing, which would encounter
significant steric hindrance, and disruption of the Pb
2+
octahedral coordination. CsBr termination
of the CsPbBr3 nanocrystals was confirmed and will be discussed in Chapter 5 using solid-state
NMR spectroscopy.
[92]

Figure 1.7. Size-dependent anion/lead ratio of cuboidal CsPbX 3 nanocrystals. The inset shows the
anion/lead ratio for experimentally observed nanocrystal sizes. Reproduced with permission from ref [91].
Copyright (2019) American Chemical Society.

The native ligands on CsPbBr3 nanocrystal surfaces consist of long-chain primary
alkylammonium ligands (oleylammonium) and long-chain carboxylate ligands (oleate).
Considering a pseudo-cubic lattice parameter (0.587 nm), a theoretical monolayer value of 5.8
ligands nm
–2
was previously calculated for CsPbBr3 assuming that the surface is solely passivated
by inorganic CsBr atoms.
[80]
However, for long-chain organic ligands with 0.3−0.5 nm
2
ligand
−1

15

footprints,
[93,94]
the maximum allowable ligand density is on the order of 2−3 ligands nm
−2
.
[92]

Indeed, most measured values of native ligand densities of CsPbBr3 nanocrystals fall within this
range.
[80,92,93]

The native ligands affect particle morphology,
[88]
stability,
[95]
and solution dispersibility,
and also passivate electronic trap states. However, the native surface ligands are labile, even in
nonpolar solvents, which contributes to the poor stability of CsPbX3 nanocrystals.
[80]
The dynamic
equilibrium of ligand binding to CsPbBr3 nanocrystals was quantified using solution
1
H NMR
spectroscopy, which is further discussed in Chapter 4.
[93]
The native oleylammonium and oleate
ligands dynamically exchange with free alkylamines and carboxylic acids in solution. Giansante
and co-workers explored a library of amine ligands with variable pKb, steric hindrance, and chain
length with the goal of achieving thermodynamically stable surface coordination and effective
passivation by exchange of native amine ligands on CsPbBr3 nanocrystals.
[96]
For
oleylammonium-capped CsPbBr3 nanocrystals, the stoichiometric addition of short (C4-C8),
strongly basic primary alkylamines results in binding to the nanocrystal surface after proton
transfer from free oleylammonium in solution to improve colloidal stability and give near-unity
PLQYs.

1.1.3.2. Post-Synthetic Surface Treatments
Didodecyldimethylammonium bromide (DDAB), a quaternary ammonium salt, has been
shown to be a very effective passivating ligand for CsPbBr3 nanocrystals. Further, the charge of
DDAB is pH independent. Manna et al. demonstrated that the post-synthetic addition of DDAB to
CsPbBr3 nanocrystals with native Cs-oleate ligands results in superior colloidal stability and near-
unity PLQYs (Figure 1.8).
[97]
They hypothesize that these beneficial effects result from weaker
16

ligand-solvent interactions for DDAB over Cs-oleate, which drives the quaternary ammonium salt
to the nanocrystal surface. Manna and Krahne followed this initial work with a temperature-
dependent PL study comparing CsPbBr3 nanocrystals treated with DDAB versus nanocrystals with
Cs-oleate or oleylammonium oleate on their surface.
[98]
The photophysics of the DDAB-treated
sample was found to behave differently; the time-resolved PL spectra maintained monoexponential
decay down to low temperatures and the PL spectrum at 4 K possessed significantly less low-
energy tailing. The authors hypothesized that DDAB-treated CsPbBr3 nanocrystals have a lower
trap state density resulting from more efficient surface passivation.

Figure 1.8. (a) Absorbance and PL spectra of Cs-oleate (red) and DDAB (green) capped CsPbBr 3
nanocrystals. (b) Evolution of PLQY for the two types of CsPbBr 3 nanocrystals. (c) Model of
tetramethylammonium bromide sitting in the A-site of CsPbBr 3. Reproduced with permission from ref [97].
Copyright (2019) American Chemical Society.

Kovalenko and Infante showed that post-synthetic ligand treatment of CsPbBr3
nanocrystals with DDAB and PbBr2 significantly increases the PLQY from 60-70% (for the as-
prepared nanocrystals with oleic acid and oleylamine) to 90-100% post-treatment.
[91]
The authors
performed both DDAB and PbBr2 and DDAB-only ligand treatments and found that the
combination of both DDAB and PbBr2 resulted in a 10-20% higher PLQY and prolonged colloidal
stability. The authors rationalized the improved PLQY and colloidal stability to result from the
combination of PbBr2 for “healing” of the (PbBr2) nanocrystal surface layer, presumably through
passivation of bromine vacancies (VBr), and superior steric stabilization by DDAB. Despite the
17

positive effects of DDAB treatments on PLQY and colloidal stability, there are still outstanding
issues. Manna and Krahne noted that the PL intensity of DDAB-treated CsPbBr3 nanocrystals
decreases with increasing temperature from 250 K to room temperature, suggesting that more
thermally robust ligand passivation is needed for device applications working above room
temperature.
[98]
Additionally, Giansante et al. reported spectrophotometric data that suggest
nanocrystal restructuring upon the addition of substoichiometric amounts of DDAB.
[96]
An
atomistic understanding of this restructuring based on empirical evidence is currently missing.
Zwitterionic ligands have become an interesting and effective strategy for stabilization and
surface passivation. Zwitterions have two advantages over traditional oleylammonium oleate
ligands: (i) zwitterions contain both cationic and anionic groups and there is no chance of
neutralization by Brønsted acid-base equilibria, and (ii) ligand binding is stabilized by the chelate
effect.
[99,100]
Kovalenko et al. explored commercially available zwitterions (i.e., sulfobetaines,
phosphocholines, γ-amino acids) for passivation of CsPbBr3 nanocrystals (Figure 1.9). They found
that sulfobetaine-capped CsPbBr3 nanocrystals formed more concentrated suspensions (50-100 mg
mL
–1
), remained more colloidally stable upon several washing cycles with polar nonsolvents, and
retained higher PLQYs for extended periods, as compared to standard oleate- and oleylammonium-
capped CsPbBr3 (Figure 1.9 c). DFT calculations were used to study sulfobetaine binding to a
CsPbBr3 nanocrystal surface and compare against oleylammonium bromide and oleylammonium
oleate binding. The anionic sulfonate group of sulfobetaine binds to sub-surface Pb (similar to
bromide and oleate), while the cationic quaternary ammonium group of sulfobetaine occupies
surface A-site vacancies (similar to oleylammonium). The quaternary ammonium group of
sulfobetaine is still well accommodated into the A-site, despite its steric bulk. The binding energies
18

for the traditional ligand ion pairs and the sulfobetaine zwitterion were comparable (40-45 kcal
mol
-1
), suggesting that the improved properties are a result of the chelate effect.

Figure 1.9. (a) Graphical depiction of as-prepared CsPbBr 3 nanocrystals capped with traditional long-chain
ligands (oleate or bromide and oleylammonium), and (b) with zwitterions containing both cationic and
anionic groups in one molecule. (c) PLQY of CsPbBr 3 nanocrystals terminated with 3-(N,N-
dimethyloctadecylammonio)propanesulfonate (green) and oleylammonium oleate (black) ligands after a
two-step of purification on day 1 and after day 28. Reproduced with permission from ref [99]. Copyright
(2018) American Chemical Society.

Kovalenko et al. further examined the utility of soy lecithin, a natural phospholipid, which
contains a physical mixture of zwitterions with various combinations of both saturated and
unsaturated hydrocarbon chains. Soy lecithin binds to the nanocrystal surface through both
19

quaternary ammonium and dialkylphosphate functionalities resulting in tighter ligand binding and
superior colloidal stability as compared to standard oleylammonium oleate ligands, allowing for a
wide range of colloid concentrations ranging from ultradilute (ng/mL) to ultraconcentrated
conditions (>400 mg/mL).
[101]
The authors rationalized the high colloidal stability of the lecithin-
stabilized CsPbBr3 nanocrystals using the Alexander-De Gennes model of polymeric interactions,
which attributes the increased colloidal stability to increased particle-particle repulsion resulting
from the tight ligand binding, high grafting density (1.8 nm
–2
), and ligand chain polydispersity.
[102–
104]
The resulting lecithin-stabilized CsPbBr3 nanocrystals demonstrated high PLQYs, equal to or
better than the aforementioned first-generation zwitterions.
Oleic acid has been replaced with other long-chain Brønsted acids, such as alkylphosphonic
acids, to stabilize CsPbX3 nanocrystals. Alkylphosphonates bind tightly to CsPbBr3 nanocrystals,
allowing for excellent surface passivation, improved PLQYs, and superior colloidal
stability.
[105,106]
Undercoordinated surface lead atoms are a cause of surface trap states for CsPbX3
nanocrystals and due to the relatively soft Lewis acid character of lead, a complementary soft
Lewis base, such as a phosphonate, is highly effective for surface passivation.
[107]
Alivisatos and
co-workers found addition of hexylphosphonic acid improved the PLQY from 76% (with a native
ligand shell of oleylammonium oleate) to 98% (with a ligand shell of oleylammonium
hexylphosphonate) for CsPbBr3 nanocrystals.
[107]
In Chapter 4, we used solution
1
H NMR
spectroscopy to show that an alkenylphosphonic acid (10-undecylphosphonic acid) irreversibly
binds and displaces oleic acid from the CsPbBr3 nanocrystal surface, quantitatively confirming the
tighter binding of softer phosphonate ligands over oleate ligands.
[93]
In Chapter 5, the binding of
this same alkenylphosphonate ligand on CsPbBr3 nanocrystals using solid-state NMR
spectroscopy was further explored.
[92]
As predicted by Alivisatos and co-workers, soft Lewis base
20

sulfonate ligands also stabilize CsPbBr3 nanocrystals and passivate surface traps.
[107]
Zeng et al.
reported that benzenesulfonate ligands result in CsPbBr3 nanocrystals with high PLQYs (>90%)
that maintain structural and optoelectronic stability through eight purification cycles and storage
>5 months.
[108]

1.2. Characterization Toolbox for QD Surfaces
Due to the impact surface ligands have on the properties and performances of quantum
dots, experimental characterization and theoretical modelling of surfaces is very important to
understand how the ligands are affecting the QDs. There are many instrumental techniques used
to directly and indirectly characterize and quantify ligand exchange reactions, including NMR,
thermogravimetric analysis (TGA), UV-vis and photoluminescence spectroscopies, FT-IR
spectroscopy, and surface modelling.

1.2.1. Nuclear Magnetic Resonance Spectroscopy
One method that is most often utilized for monitoring the ligand exchange reactions of
organic ligands in situ is nuclear magnetic resonance (NMR) spectroscopy. Not only does NMR
allow for qualitative observation of incoming and outgoing ligands in a colloidal suspension, but
it also allows for quantitative measurements if an internal standard is used. There are many NMR
spectroscopic techniques that are utilized in a traditional ligand exchange study: (1) one-
dimensional
1
H NMR, (2) 2D diffusion ordered spectroscopy (DOSY), and (3) nuclear Overhauser
effect spectroscopy (NOESY).
The one-dimensional
1
H NMR spectrum is incredibly useful for the study of nanocrystal
suspensions. Qualitatively, the proton resonances will typically shift either upfield or downfield
21

(depending on the solvent) and broaden if the organic ligands are interacting with the QD surfaces.
This line broadening is due to different relaxation behavior between free and bound molecules. In
particular, ligands that are bound to the surface act like larger molecules and have a slower T2
relaxation, also known as transverse relaxation or spin-spin relaxation. This results in broadened
peaks, whereas free ligands can tumble more quickly in solution and have faster T2 relaxation,
resulting in sharper peaks. The
1
H NMR spectrum can also provide quantitative information. In
particular, if an internal standard is used, the concentration of bound and free ligands present (if
the peaks are distinguishable) can be calculated with peak integration. These concentrations can
then be used to calculate the ligand exchange equilibrium (eq. 1), where IL represents the incoming
ligand, NL represents the native ligand, F is free, and B is bound:

!"
=
[*+]
"
[,+]
!
[*+]
!
[,+]
"
(1)
As previously discussed, ligands that are bound to the surface diffuse more slowly in
solution compared to those that are free in solution. This behavior can also be measured with
DOSY. From a 2D DOSY spectrum, diffusion coefficients are extracted for all molecules in the
suspension. This is particularly useful for observing ligands attached to QD surfaces because
tightly bound ligands will have a smaller diffusion coefficient than free ligands. Additionally, these
diffusion coefficients (D) are inversely related to the solvodynamic radius (rs), which includes the
QD and ligand shell. The Stokes-Einstein equation (eq. 2) used for spherical particles is dependent
on the Boltzmann constant (kB), temperature (T), and solvent viscosity (h).
=
-
"
.
/012
#
(2)
While
1
H NMR and DOSY are useful for gathering information about the binding of
ligands, there are two additional NMR spectroscopic techniques often used in literature to analyze
22

ligand binding. 2D Nuclear Overhauser effect spectroscopy (NOESY) is a useful addition to ligand
exchange experiments if the exchange reaction occurs quickly (on the NMR time scale). On the
2D NOESY spectrum, there is the phase of the peaks on the diagonal and there are cross peaks. If
the cross peaks are positive (marked by the same color of the diagonal), the ligands are bound. If
the cross peaks are negative (marked by a different color of the diagonal), the ligands are free and
are possibly exchanging quickly. De Roo et al. demonstrated the utility of 2D NOESY for CsPbBr3
QDs that were prepared with oleylamine and oleic acid and showed that oleic acid and octadecene
(the non-coordinating solvent) were not interacting with the QD surface, whereas oleylamine does
and it binds as an oleylammonium bromide pair of X-type ligands (Figure 1.10).

Figure 1.10. NOESY spectrum of the sample purified with acetone in CDCl 3. Reproduced with permission
from ref [80]. Copyright (2016) American Chemical Society.

Solid-state NMR is another valuable technique to prove the surface chemistry and ligand
binding of QDs, which will be discussed in more detail in Chapter 5 for CsPbBr3 QDs. Due to the
inherent disorder and highly dynamic surfaces of most QDs, solid-state NMR is very useful for
determining the which surface atoms the ligands bind to, if the nuclei present in the QD are NMR
23

active. With solid-state NMR, the scalar and dipolar couplings of NMR active nuclei provide
valuable proximity and connectivity information that can often be used to determine the surface
composition and atom-ligand connectivity. The information gathered from solid-state NMR versus
solution NMR is unique because internuclear distances between ligands and the surface atoms can
be probed and certain resonances that may overlap with NMR solvents can be avoided because
there is no NMR solvent present in solid-state NMR.

1.2.2. Other Commonly Utilized Surface Characterization Techniques
There are many characterization methods used to monitor the surface chemistry of QDs
that focus on the ligand composition, the optical properties of the QDs, or the interaction between
the QDs and the ligands.
TGA is a quantitative technique that measures the weight percent of organic content present
in a dried sample of QDs. Used in conjunction with other methods, such as NMR or UV-vis, the
data provided by TGA can be used to determine the ratio of organic ligands to nanoparticles and
can be used to calculate the isolated yield of the reaction. To study the colloidal stability and
optoelectronic properties of QDs, UV-vis and PL spectroscopies are measured to monitor any
changes upon ligand exchange. For example, if a ligand exchange reaction caused etching of the
QD surface a blue shift would be observed in both the absorption and PL spectra. Fourier-transform
infrared (FT-IR) spectroscopy is a technique that is used to observe the absorption of organic
compounds in the infrared region of the electromagnetic spectrum and can be used to monitor and
identify organic surface ligands. FT-IR can identify the ligands present, because ligands have
specific functional groups that are attached to the QD surface and typically have a unique
“fingerprint.” Additionally, if the goal of the ligand exchange reaction was to completely remove
24

a ligand, FT-IR is useful in identifying if any of the native ligand remains on the QD surface. FT-
IR is frequently used in conjunction with
1
H and
13
C NMR confirm the presence or absence of
organic ligands.
Lastly, surface modelling through density functional theory (DFT) is exploited in many
literature investigations of the surface chemistry of QDs because the calculated binding affinities
can yield valuable insight as to why a surface ligand passivates the QD surface well (or poorly).
Additionally, surface modelling of QDs can be calculated to determine which cation or anion site
ligands are binding to. This is typically corroborated with experimental data, such as Resonance-
Echo Saturation-Pulse DOuble-Resonance (RESPDOR) curves from solid-state NMR
spectroscopy.

1.3. References
[1] F. Manby, T. Miller, P. Bygrave, F. Ding, T. Dresselhaus, F. Batista-Romero, A.
Buccheri, C. Bungey, S. Lee, R. Meli, K. Miyamoto, C. Steinmann, T. Tsuchiya, M.
Welborn, T. Wiles, Z. Williams, Entos: A Quantum Molecular Simulation Package,
2019.
[2] S. Irvine, in Springer Handbook of Electronic and Photonic Materials (Eds.: S. Kasap, P.
Capper), Springer International Publishing, Cham, 2017, pp. 1–1.
[3] A. Kojima, K. Teshima, Y. Shirai, T. Miyasaka, J. Am. Chem. Soc. 2009, 131, 6050–
6051.
[4] A. K. Jena, A. Kulkarni, T. Miyasaka, Chem. Rev. 2019, 119, 3036–3103.
[5] T. Yu, F. Wang, Y. Xu, L. Ma, X. Pi, D. Yang, Adv. Mater. 2016, 28, 4912–4919.
[6] X. Zhou, J. Qiao, Z. Xia, Chem. Mater. 2021, 33, 1083–1098.
25

[7] C.-H. M. Chuang, P. R. Brown, V. Bulović, M. G. Bawendi, Nat. Mater. 2014, 13, 796–
801.
[8] P. Guyot-Sionnest, M. M. Ackerman, X. Tang, J. Chem. Phys. 2019, 151, 060901.
[9] J. Owen, L. Brus, J. Am. Chem. Soc. 2017, 139, 10939–10943.
[10] P. Wangyang, C. Gong, G. Rao, K. Hu, X. Wang, C. Yan, L. Dai, C. Wu, J. Xiong, Adv.
Opt. Mater 2018, 6, 1701302.
[11] H. Lu, G. M. Carroll, N. R. Neale, M. C. Beard, ACS Nano 2019, acsnano.8b09815.
[12] M. V. Kovalenko, Nat. Nanotech 2015, 10, 994–997.
[13] J. Owen, Science 2015, 347, 615–616.
[14] M. L. H. Green, G. Parkin, J. Chem. Educ. 2014, 91, 807–816.
[15] M. L. H. Green, J. Organomet. Chem. 1995, 500, 127–148.
[16] J. Owen, Science 2015, 347, 615–616.
[17] M. Woodhouse, R. Jones-Albertus, D. Feldman, R. Fu, K. Horowitz, D. Chung, D.
Jordan, S. Kurtz, On the Path to SunShot: The Role of Advancements in Solar
Photovoltaic Efficiency, Reliability, and Cost, National Renewable Energy Laboratory,
2016.
[18] R. K. Baldwin, K. A. Pettigrew, E. Ratai, M. P. Augustine, S. M. Kauzlarich, Chem.
Commun. 2002, 1822–1823.
[19] P. M. Fauchet, Mater. Today 2005, 8, 26–33.
[20] K. A. Pettigrew, Q. Liu, P. P. Power, S. M. Kauzlarich, Chem. Mater. 2003, 15, 4005–
4011.
[21] J. M. Buriak, Phil. Trans. R. Soc. A. 2006, 364, 217–225.
[22] Z. F. Li, E. Ruckenstein, Nano Lett. 2004, 4, 1463–1467.
26

[23] L. Wang, V. Reipa, J. Blasic, Bioconjugate Chem. 2004, 15, 409–412.
[24] E. Muthuswamy, J. Zhao, K. Tabatabaei, M. M. Amador, M. A. Holmes, F. E. Osterloh,
S. M. Kauzlarich, Chem. Mater. 2014, 26, 2138–2146.
[25] E. Fok, M. Shih, A. Meldrum, J. G. C. Veinot, Chem. Commun. 2004, 386.
[26] K. Dohnalová, T. Gregorkiewicz, K. Kůsová, J. Phys.: Condens. Matter 2014, 26,
173201.
[27] F. Li, J. Wang, S. Sun, H. Wang, Z. Tang, G. Nie, Small 2015, 11, 1954–1961.
[28] J.-W. Luo, A. Franceschetti, A. Zunger, Nano Lett. 2008, 8, 3174–3181.
[29] U. Kortshagen, R. Anthony, R. Gresback, Z. Holman, R. Ligman, C.-Y. Liu, L.
Mangolini, S. A. Campbell, Pure Appl. Chem. 2008, 80, 1901–1908.
[30] S. Bhattacharjee, I. M. C. M. Rietjens, M. P. Singh, T. M. Atkins, T. K. Purkait, Z. Xu, S.
Regli, A. Shukaliak, R. J. Clark, B. S. Mitchell, G. M. Alink, A. T. M. Marcelis, M. J.
Fink, J. G. C. Veinot, S. M. Kauzlarich, H. Zuilhof, Nanoscale 2013, 5, 4870.
[31] C.-S. Yang, R. A. Bley, S. M. Kauzlarich, H. W. H. Lee, G. R. Delgado, J. Am. Chem.
Soc. 1999, 121, 5191–5195.
[32] T. M. Atkins, M. C. Cassidy, M. Lee, S. Ganguly, C. M. Marcus, S. M. Kauzlarich, ACS
Nano 2013, 7, 1609–1617.
[33] H. Gerung, S. D. Bunge, T. J. Boyle, C. J. Brinker, S. M. Han, Chem. Commun. 2005,
1914–1916.
[34] T. N. Lambert, N. L. Andrews, H. Gerung, T. J. Boyle, J. M. Oliver, B. S. Wilson, S. M.
Han, Small 2007, 3, 691–699.
[35] D. D. Vaughn, J. F. Bondi, R. E. Schaak, Chem. Mater. 2010, 22, 6103–6108.
27

[36] S. R. Smock, K. Tabatabaei, T. J. Williams, S. M. Kauzlarich, R. L. Brutchey, Nanoscale
2020, 12, 2764–2772.
[37] L. M. Wheeler, A. W. Nichols, B. D. Chernomordik, N. C. Anderson, M. C. Beard, N. R.
Neale, Nano Lett. 2016, 16, 1949–1954.
[38] I. Robel, V. Subramanian, M. Kuno, P. V. Kamat, J. Am. Chem. Soc. 2006, 128, 2385–
2393.
[39] K. S. Leschkies, R. Divakar, J. Basu, E. Enache-Pommer, J. E. Boercker, C. B. Carter, U.
R. Kortshagen, D. J. Norris, E. S. Aydil, Nano Lett. 2007, 7, 1793–1798.
[40] S. Sadeghi, S. Khabbaz Abkenar, C. W. Ow-Yang, S. Nizamoglu, Sci Rep 2019, 9,
10061.
[41] M. Nirmal, L. Brus, Acc. Chem. Res. 1999, 32, 407–414.
[42] X. Peng, M. C. Schlamp, A. V. Kadavanich, A. P. Alivisatos, J. Am. Chem. Soc. 1997,
119, 7019–7029.
[43] D. V. Talapin, A. L. Rogach, A. Kornowski, M. Haase, H. Weller, Nano Lett. 2001, 1,
207–211.
[44] Y. C. Cao, J. Wang, J. Am. Chem. Soc. 2004, 126, 14336–14337.
[45] R. Gomes, A. Hassinen, A. Szczygiel, Q. Zhao, A. Vantomme, J. C. Martins, Z. Hens, J.
Phys. Chem. Lett. 2011, 2, 145–152.
[46] J. M. Luther, J. M. Pietryga, ACS Nano 2013, 7, 1845–1849.
[47] J. Jasieniak, P. Mulvaney, J. Am. Chem. Soc. 2007, 129, 2841–2848.
[48] R. R. Knauf, J. C. Lennox, J. L. Dempsey, Chem. Mater. 2016, 28, 4762–4770.
[49] M. Liu, Y.-Y. Wang, Y. Liu, F.-L. Jiang, J. Phys. Chem. C 2020, 124, 4613–4625.
28

[50] M. K. Samantaray, V. Katiyar, D. Roy, K. Pang, H. Nanavati, R. Stephen, R. B. Sunoj, P.
Ghosh, Eur. J. Inorg. Chem. 2006, 2006, 2975–2984.
[51] M. T. Frederick, E. A. Weiss, ACS Nano 2010, 4, 3195–3200.
[52] M. S. Azzaro, A. Dodin, D. Y. Zhang, A. P. Willard, S. T. Roberts, Nano Lett. 2018, 18,
3259–3270.
[53] D. E. Westmoreland, R. López-Arteaga, E. A. Weiss, J. Am. Chem. Soc. 2020, 142,
2690–2696.
[54] S. Lian, J. A. Christensen, M. S. Kodaimati, C. R. Rogers, M. R. Wasielewski, E. A.
Weiss, J. Phys. Chem. C 2019, 123, 5923–5930.
[55] M. V. Kovalenko, L. Manna, A. Cabot, Z. Hens, D. V. Talapin, C. R. Kagan, V. I.
Klimov, A. L. Rogach, P. Reiss, D. J. Milliron, P. Guyot-Sionnnest, G. Konstantatos, W.
J. Parak, T. Hyeon, B. A. Korgel, C. B. Murray, W. Heiss, ACS Nano 2015, 9, 1012–
1057.
[56] M. J. Greaney, E. Couderc, J. Zhao, B. A. Nail, M. Mecklenburg, W. Thornbury, F. E.
Osterloh, S. E. Bradforth, R. L. Brutchey, Chem. Mater. 2015, 27, 744–756.
[57] J. J. Buckley, M. J. Greaney, R. L. Brutchey, Chem. Mater. 2014, 26, 6311–6317.
[58] S. R. Smock, Y. Chen, A. J. Rossini, R. L. Brutchey, Acc. Chem. Res. 2021, 54, 707–718.
[59] G. Rose, J. Prakt. Chem. 1840, 19, 459–468.
[60] A. S. Bhalla, R. Guo, R. Roy, Mater. Res. Innov. 2000, 4, 3–26.
[61] J. Breternitz, S. Schorr, Adv. Energy Mater. 2018, 8, 1802366.
[62] S. Niu, H. Huyan, Y. Liu, M. Yeung, K. Ye, L. Blankemeier, T. Orvis, D. Sarkar, D. J.
Singh, R. Kapadia, J. Ravichandran, Adv. Mater. 2017, 29, 1604733.
29

[63] K. R. Talley, C. L. Perkins, D. R. Diercks, G. L. Brennecka, A. Zakutayev,
arXiv:2001.00633 [cond-mat] 2020.
[64] C. C. Stoumpos, C. D. Malliakas, J. A. Peters, Z. Liu, M. Sebastian, J. Im, T. C.
Chasapis, A. C. Wibowo, D. Y. Chung, A. J. Freeman, B. W. Wessels, M. G. Kanatzidis,
Cryst. Growth Des. 2013, 13, 2722–2727.
[65] H. L. Wells, Z. Anorg. Chem. 1893, 3, 195–210.
[66] “Best Research-Cell Efficiencies,” can be found under
https://www.nrel.gov/pv/assets/pdfs/best-research-cell-efficiencies.20200708.pdf, n.d.
[67] N. K. Noel, A. Abate, S. D. Stranks, E. S. Parrott, V. M. Burlakov, A. Goriely, H. J.
Snaith, ACS Nano 2014, 8, 9815–9821.
[68] D. W. deQuilettes, S. Koch, S. Burke, R. K. Paranji, A. J. Shropshire, M. E. Ziffer, D. S.
Ginger, ACS Energy Lett. 2016, 1, 438–444.
[69] Z. Hawash, S. R. Raga, D.-Y. Son, L. K. Ono, N.-G. Park, Y. Qi, J. Phys. Chem. Lett.
2017, 8, 3947–3953.
[70] E. A. Alharbi, A. Y. Alyamani, D. J. Kubicki, A. R. Uhl, B. J. Walder, A. Q. Alanazi, J.
Luo, A. Burgos-Caminal, A. Albadri, H. Albrithen, M. H. Alotaibi, J.-E. Moser, S. M.
Zakeeruddin, F. Giordano, L. Emsley, M. Grätzel, Nat. Commun 2019, 10, 3008.
[71] I. K. Hewavitharana, S. L. Brock, ACS Nano 2017, 11, 11217–11224.
[72] H. Choi, S. Kim, J. M. Luther, S.-W. Kim, D. Shin, M. C. Beard, S. Jeong, Chem. Eur. J.
2017, 23, 17707–17713.
[73] A. J. Morris-Cohen, M. T. Frederick, G. D. Lilly, E. A. McArthur, E. A. Weiss, J. Phys.
Chem. Lett. 2010, 1, 1078–1081.
30

[74] I. Moreels, K. Lambert, D. De Muynck, F. Vanhaecke, D. Poelman, J. C. Martins, G.
Allan, Z. Hens, Chem. Mater. 2007, 19, 6101–6106.
[75] C. Bullen, P. Mulvaney, Langmuir 2006, 22, 3007–3013.
[76] J. J. Buckley, E. Couderc, M. J. Greaney, J. Munteanu, C. T. Riche, S. E. Bradforth, R. L.
Brutchey, ACS Nano 2014, 8, 2512–2521.
[77] V. Grigel, L. K. Sagar, K. De Nolf, Q. Zhao, A. Vantomme, J. De Roo, I. Infante, Z.
Hens, Chem. Mater. 2018, 30, 7637–7647.
[78] K. Sandeep, K. Y. Gopika, M. R. Revathi, Phys. Status Solidi RRL 2019, 13, 1900387.
[79] S. ten Brinck, I. Infante, ACS Energy Lett. 2016, 1, 1266–1272.
[80] J. De Roo, M. Ibáñez, P. Geiregat, G. Nedelcu, W. Walravens, J. Maes, J. C. Martins, I.
Van Driessche, M. V. Kovalenko, Z. Hens, ACS Nano 2016, 10, 2071–2081.
[81] M. V. Kovalenko, L. Protesescu, M. I. Bodnarchuk, Science 2017, 358, 745–750.
[82] H. Huang, M. I. Bodnarchuk, S. V. Kershaw, M. V. Kovalenko, A. L. Rogach, ACS
Energy Lett. 2017, 2, 2071–2083.
[83] F. Yan, S. T. Tan, X. Li, H. V. Demir, Small 2019, 15, 1902079.
[84] L. Protesescu, S. Yakunin, M. I. Bodnarchuk, F. Krieg, R. Caputo, C. H. Hendon, R. X.
Yang, A. Walsh, M. V. Kovalenko, Nano Lett. 2015, 15, 3692–3696.
[85] P. Cottingham, R. L. Brutchey, Chem. Commun. 2016, 52, 5246–5249.
[86] P. Cottingham, R. L. Brutchey, Chem. Commun. 2016, 28, 7574–7577.
[87] P. Cottingham, R. L. Brutchey, Chem. Mater. 2018, 30, 6711–6716.
[88] A. Pan, B. He, X. Fan, Z. Liu, J. J. Urban, A. P. Alivisatos, L. He, Y. Liu, ACS Nano
2016, 10, 7943–7954.
[89] M. C. Brennan, M. Kuno, S. Rouvimov, Inorg. Chem. 2019, 58, 1555–1560.
31

[90] V. K. Ravi, P. K. Santra, N. Joshi, J. Chugh, S. K. Singh, H. Rensmo, P. Ghosh, A. Nag,
J. Phys. Chem. Lett. 2017, 8, 4988–4994.
[91] M. I. Bodnarchuk, S. C. Boehme, S. ten Brinck, C. Bernasconi, Y. Shynkarenko, F.
Krieg, R. Widmer, B. Aeschlimann, D. Günther, M. V. Kovalenko, I. Infante, ACS
Energy Lett. 2019, 4, 63–74.
[92] Y. Chen, S. R. Smock, A. H. Flintgruber, F. A. Perras, R. L. Brutchey, A. J. Rossini, J.
Am. Chem. Soc. 2020, 142, 6117–6127.
[93] S. R. Smock, T. J. Williams, R. L. Brutchey, Angew. Chem. Int. Ed. 2018, 57, 11711–
11715.
[94] A. Antanovich, A. Prudnikau, A. Matsukovich, A. Achtstein, M. Artemyev, J. Phys.
Chem. C 2016, 120, 5764–5775.
[95] S. Wang, L. Du, Z. Jin, Y. Xin, H. Mattoussi, J. Am. Chem. Soc. 2020, 142, 12669–
12680.
[96] D. Quarta, M. Imran, A.-L. Capodilupo, U. Petralanda, B. van Beek, F. De Angelis, L.
Manna, I. Infante, L. De Trizio, C. Giansante, J. Phys. Chem. Lett. 2019, 10, 3715–3726.
[97] M. Imran, P. Ijaz, L. Goldoni, D. Maggioni, U. Petralanda, M. Prato, G. Almeida, I.
Infante, L. Manna, ACS Energy Lett. 2019, 4, 819–824.
[98] P. Ijaz, M. Imran, M. M. Soares, H. C. N. Tolentino, B. Martín-García, C. Giannini, I.
Moreels, L. Manna, R. Krahne, J. Phys. Chem. Lett. 2020, 11, 2079–2085.
[99] F. Krieg, S. T. Ochsenbein, S. Yakunin, S. ten Brinck, P. Aellen, A. Süess, B. Clerc, D.
Guggisberg, O. Nazarenko, Y. Shynkarenko, S. Kumar, C.-J. Shih, I. Infante, M. V.
Kovalenko, ACS Energy Lett. 2018, 3, 641–646.
[100] G. Schwarzenbach, Helv. Chim. Acta 1952, 35, 2344–2359.
32

[101] F. Krieg, Q. K. Ong, M. Burian, G. Rainò, D. Naumenko, H. Amenitsch, A. Süess, M. J.
Grotevent, F. Krumeich, M. I. Bodnarchuk, I. Shorubalko, F. Stellacci, M. V. Kovalenko,
J. Am. Chem. Soc. 2019, 141, 19839–19849.
[102] P. G. De Gennes, Macromolecules 1981, 14, 1637–1644.
[103] P. G. De Gennes, Macromolecules 1982, 15, 492–500.
[104] P. G. De Gennes, Adv. Colloid Interface Sci. 1987, 27, 189–209.
[105] Y. Tan, Y. Zou, L. Wu, Q. Huang, D. Yang, M. Chen, M. Ban, C. Wu, T. Wu, S. Bai, T.
Song, Q. Zhang, B. Sun, ACS Appl. Mater. Interfaces 2018, 10, 3784–3792.
[106] W. Koh, S. Park, Y. Ham, ChemistrySelect 2016, 1, 3479–3482.
[107] D. P. Nenon, K. Pressler, J. Kang, B. A. Koscher, J. H. Olshansky, W. T. Osowiecki, M.
A. Koc, L.-W. Wang, A. P. Alivisatos, J. Am. Chem. Soc. 2018, 140, 17760–17772.
[108] D. Yang, X. Li, W. Zhou, S. Zhang, C. Meng, Y. Wu, Y. Wang, H. Zeng, Adv. Mater.
2019, 1900767.

33

Chapter 2. Surface Coordination Chemistry of Germanium Nanocrystals Synthesized
by Microwave-Assisted Reduction in Oleylamine

*Published in Nanoscale 2020, 12, 2764-2772.

2.1. Abstract
As surface ligands play a critical role in the colloidal stability and optoelectronic properties
of semiconductor nanocrystals, we used solution NMR experiments to investigate the surface
coordination chemistry of Ge nanocrystals synthesized by a microwave-assisted reduction of GeI2
in oleylamine. The as-synthesized Ge nanocrystals are coordinated to a fraction of strongly bound
oleylamide ligands (with covalent X-type Ge–NHR bonds) and a fraction of more weakly bound
(or physisorbed) oleylamine, which readily exchanges with free oleylamine in solution. The
fraction of strongly bound oleylamide ligands increases with increasing synthesis temperature,
which also correlates with better colloidal stability. Thiol and carboxylic acid ligands bind to the
Ge nanocrystal surface only upon heating, suggesting a high kinetic barrier to surface binding.
These incoming ligands do not displace native oleylamide ligands but instead appear to coordinate
to open surface sites, confirming that the as-prepared nanocrystals are not fully passivated. These
findings will allow for a better understanding of the surface chemistry of main group nanocrystals
and the conditions necessary for ligand exchange to ultimately maximize their functionality.

2.2. Introduction
Nanocrystals of Ge, a group IV semiconductor, have been investigated as an alternative to
Si nanocrystals for many years as a result of their superior absorption coefficient, charge transport
capabilities, narrow bulk band gap (0.67 eV at 300 K), and fairly large Bohr exciton radius (ca. 24
34

nm) that allows for a size-tunable band gap.
[1,2]
These collective properties make Ge nanocrystals
appealing for a range of optoelectronic applications, including bioimaging and solar energy
conversion.
[3–5]
It is well established that the properties of semiconductor nanocrystals are highly
dependent on the surface ligands that fill out the coordination shell of their surface atoms.
[6–10]
The
long-chain aliphatic native ligands that are typically present from the nanocrystal synthesis (e.g.,
C18 oleylamine) impart solution dispersibility in nonpolar organic solvents; however, many
potential applications of Ge nanocrystals require the exchange of such native ligands with new
ligands that will, for example, improve their charge mobility and conductivity in nanocrystal
solids,
[11]
or allow them to be dispersed in polar, biologically relevant media.
[12]

Beginning with the work of Boyle et al. in 2005,
[13]
synthesizing colloidal Ge nanocrystals
by the high-temperature chemical reduction of Ge(II) and/or Ge(IV) precursors in long-chain
aliphatic primary amines has become common practice,
[12,14–16]
resulting in nanocrystals
passivated by surface-coordinated amine ligands. While only a few ligand exchange reactions on
these Ge nanocrystals have been reported thus far, it has been demonstrated that the native primary
amine ligands can be at least partially exchanged with thiols,
[12,17]
polyethyleneimine,
[11]
and
polysulfides
[18]
via room temperature exchange reactions. When colloidal Ge nanocrystals are
synthesized by the chemical reduction of GeI2 in oleylamine by microwave heating, instead of
conventional heating, the resulting nanocrystals are similarly supported by surface oleylamine
ligands;
[19,20]
however, it was reported that these native oleylamine ligands can only be efficiently
exchanged with thiols after surface reduction with hydrazine followed by heating the nanocrystals
in neat thiol.
[1]
In order to rationally design and execute ligand exchange reactions for Ge and other
main group nanocrystals moving forward, a better understanding of the fundamental coordination
chemistry of the nanocrystal surface must be achieved.
35

Herein, we use solution NMR spectroscopy to determine the binding mode of the native
oleylamine ligands to Ge nanocrystals synthesized via microwave-assisted reduction. Based on
exchange reactions with various amine/ammonium, thiol, and carboxylic acid ligands, we posit
that the as-synthesized Ge nanocrystals are coordinated to a fraction of strongly bound oleylamide
ligands (with covalent, X-type Ge-alkyl amide (Ge–NHR) bonds) and a fraction of more weakly
bound (or physisorbed) oleylamine. Metal–amide surface ligands have been previously speculated
to support other main group (i.e., In, Ga, Bi, Sb and Sn) nanocrystals prepared by the reduction of
metal amide precursors formed in situ,
[21,22]
and have been experimentally confirmed in the case
of Si nanocrystals;
[23]
however, this provides the first evidence for their existence in Ge
nanocrystals.

2.3. Preparation and Characterization of Ge Nanocrystals Prepared via Microwave-Assisted
Reduction of Oleylamine
The Ge nanocrystals used here were synthesized by the microwave-assisted reduction of
GeI2 in oleylamine, without additional ligands or solvents, at a series of reaction temperatures (i.e.,
210, 230, 250, and 270 ˚C) for 1 h.
[1,19,24,25]
The nanocrystal products all adopted the expected
cubic diamondoid structure of Ge, as assessed by powder X-ray diffraction, with nanocrystal sizes
that increase with increasing reaction temperature (Figure 2.1-2.2). The resulting nanocrystals
were purified by four dispersion/precipitation cycles with one cycle of toluene and methanol,
followed by three cycles of hexanes and methanol, except for the Ge nanocrystals synthesized at
210 ˚C, which were only washed twice because of their relatively poor colloidal stability. Indeed,
we qualitatively observed that suspensions of Ge nanocrystals synthesized at higher temperatures
36

generally possessed better colloidal stability. FT-IR spectra of the resulting nanocrystals confirm
that oleylamine acts as the passivating ligand (Figure 2.3).

Figure 2.1. Powder XRD patterns of Ge nanocrystals prepared at various temperatures (210 ˚C, 230 ˚C,
250 ˚C, and 270 ˚C) compared to the reference pattern (PDF #04-0545) showing the (111), (220, (311),
(400), and (331) reflections of cubic Ge.

37

Figure 2.2. TEM micrographs and size histograms of Ge nanocrystals prepared at (a) 210 ˚C, (b) 230 ˚C,
(c) 250 ˚C, (d) 270 ˚C.

Figure 2.3. Representative FT-IR spectra of pure oleylamine (OAm) and oleylamine-capped Ge
nanocrystals prepared at various synthetic temperatures (210, 230, 250, and 270 ˚C) dispersed in hexanes
and dispensed onto the attenuated total reflection (ATR)-crystal followed by drying.

2.4.
1
H NMR Spectroscopy to Determine the Surface Ligands of Ge Nanocrystals
To better understand the effect that reaction temperature has on colloidal stability, and
ligand binding to the Ge nanocrystals, solution
1
H NMR spectra were collected from Ge
nanocrystal samples synthesized at 210, 230, 250, and 270 ˚C. It is observed that the peak shapes
of the resonances associated with the alkenyl protons of the native oleylamine ligands (d = 5.4-5.7
ppm) asymmetrically broaden with increasing reaction temperature, accompanied by a downfield
shift (Figure 2.4). This broadening and downfield shift generally imply a ligand interaction with
the nanocrystal surface.
[26,27]
Deconvolution of the peaks associated with the alkenyl protons of
oleylamine into two populations of ligand species (i.e., strongly bound oleylamine at d » 5.64 ppm
38

and less strongly bound, or physisorbed, oleylamine at d » 5.59 ppm) is given in Figure 2.4. This
analysis returns a 18%, 51%, and 70% strongly bound fraction for the Ge nanocrystals synthesized

Figure 2.4. Room-temperature 600 MHz
1
H NMR spectra and fitting of the bound (d » 5.64 ppm) and
physisorbed/free (d » 5.59 ppm) peaks of the alkenyl region that show the peak broadening and change in
chemical shift associated with ligand binding of oleylamine (OAm), which is more prominent for
nanocrystals synthesized at higher temperatures. The
1
H NMR spectrum of free oleylamine in toluene-d 8 is
given for comparison.

at 230, 250, and 270 ˚C, respectively, thereby demonstrating that the number fraction of strongly
bound oleylamine increases with higher synthetic temperatures. Due to the colloidal instability of
the nanocrystals synthesized at 210 ˚C, the
1
H NMR peak intensity is too small to reliably integrate
a strongly bound ligand fraction. Nonetheless, based on a qualitative visual inspection of the
chemical shift and peak shape, there is a larger fraction of free or loosely bound oleylamine present
39

in the suspension of Ge nanocrystals synthesized at 210 ˚C. The overall ligand density of native
oleylamine for the most colloidally stable Ge nanocrystals synthesized at 270 ˚C were calculated
using thermogravimetric analysis to be between 1.3-1.8 oleylamine nm
-2
when considering either
just the strongly bound oleylamine or the total oleylamine population, respectively. This range of
oleylamine surface densities is well below that of a theoretical monolayer of 2.8 oleylamine nm
-2

calculated for this particle size using an oleylamine footprint of 0.36 nm
2
.
[28]
The low surface
coverage is consistent with previously published conclusions based on property measurements that
these Ge nanocrystals are incompletely passivated.
[1]

In order to corroborate the increasing fraction of strongly bound oleylamine for
nanocrystals synthesized at higher temperatures, diffusion ordered NMR spectroscopy (DOSY)
was performed to calculate diffusion coefficients and solvodynamic diameters using the Stokes-
Einstein equation.
[26]
The experimentally determined diffusion coefficients and calculated
solvodynamic diameters are presented in Table 2.1, which indicate that all the Ge nanocrystal
suspensions have significantly smaller diffusion coefficients for oleylamine (< 300 µm s
–1
) than
free oleylamine by itself (887 ± 2 µm s
–1
). The smaller oleylamine diffusion coefficients in the
nanocrystal suspensions imply the ligands are interacting with the Ge nanocrystal surfaces,
returning a weighted mean diffusion coefficient between the bound and unbound states. As such,
the calculated solvodynamic diameters are smaller than the expected diameters because the
Table 2.1. Summary of diffusion coefficients and calculated solvodynamic diameters using DOSY NMR,
and expected diameters
40

Synthetic temperature Oleylamine diffusion
coefficient (µm s
–1
)
Calculated solvodynamic
diameter (nm)
Expected diameter
(nm)
a
free oleylamine 887 ± 2
210 ˚C 277 ± 50 2.8 4.9 – 5.4
230˚C 187 ± 4 4.2 5.4 – 5.9
250 ˚C 165 ± 3 4.8 6.0 – 6.5
270 ˚C 119 ± 5 6.6 7.0 – 7.5
a
Expected diameter = core Ge nanocrystal size from TEM + 1.5-2 nm estimated from oleylamine ligand
shell, which is predicted to be about half of the ligand length.
[29]

solvodynamic diameters are calculated using a diffusion coefficient that includes both bound and
physisorbed/free oleylamine, giving an overall larger diffusion coefficient for oleylamine than
expected for a strictly bound state and, therefore, a smaller solvodynamic diameter. More
importantly, the diffusion coefficients decrease with increasing synthetic temperature, which
reflect and confirm that there is a higher fraction of strongly bound oleylamine in Ge nanocrystal
samples synthesized at higher temperatures, directly correlating with the better colloidal stability
of those suspensions.

2.5. Ligand Exchange with Amines
We then shifted our attention to ligand exchange reactions on the Ge nanocrystals,
beginning with a primary alkylamine. If the native oleylamine ligands are bound to the Ge
nanocrystal surface through a dative L-type ligand interaction, it is expected that an incoming
primary amine could displace the native ligand. Undeceneamine was selected for ligand exchange,
because of its spectroscopically distinct vinylic proton resonances that do not overlap with the
internal alkenyl proton resonances of oleylamine.
[30]
This permits the binding of oleylamine and
undeceneamine to be followed concurrently and thereby allows all changes in ligand binding to be
monitored. Upon titrating increasing amounts of the new undeceneamine ligand into a Ge
nanocrystal suspension at room temperature, the peaks corresponding to the vinylic protons of
41

undeceneamine shift upfield (Figure 2.5). This behavior has previously been attributed to
physisorbed, or weakly bound “interdigitated” ligands,
[31]
and as more undeceneamine is

Figure 2.5. Room-temperature 600 MHz
1
H NMR spectra of Ge nanocrystal suspensions in toluene–d 8 (7
mg/mL) for samples synthesized at (a) 230 ˚C, (b) 250 ˚C, and (c) 270˚C. The as-synthesized nanocrystals
are capped with native oleylamine (OAm) ligands and the suspensions are titrated with increasing amounts
of undeceneamine (UAm) (0-62 mM).

introduced, the ratio of free to interdigitated ligand increases as more amine competes for those
physisorbed “sites”. If strongly bound, undeceneamine is expected to have broad peaks downfield
from the free peaks; this is not observed. Upon titration with undeceneamine, the alkenyl
resonances for the less strongly bound oleylamine also undergo a chemical shift upfield, increasing
the separation between the peaks for the strongly bound and progressively more free oleylamine
ligands. For all the Ge nanocrystal samples synthesized at various temperatures, the broad peak
associated with strongly bound oleylamine does not significantly change shape or intensity upon
undeceneamine titration, implying that the incoming undeceneamine ligands are not displacing the
42

strongly bound native oleylamine ligands, but rather are only competing for binding with weakly
physisorbed oleylamine. Therefore, undeceneamine does not bind tightly to the nanocrystal surface
and does not ligand exchange with strongly bound oleylamine.
DOSY was performed and diffusion coefficients were calculated for the undeceneamine
titration experiments shown in Figure 2.5. The diffusion coefficients for undeceneamine were
calculated to be 768 ± 20, 908 ± 10, and 963 ± 30 µm s
–1
for the suspensions of Ge nanocrystals
synthesized at 230, 250, and 270 ˚C, respectively, compared to the diffusion coefficient of free
undeceneamine, which is 1335 ± 8 µm s
–1
. The comparatively high diffusion coefficients for
undeceneamine in the nanocrystal suspensions reflect that undeceneamine is not binding strongly
as a ligand to the Ge nanocrystal surface, which agrees with the
1
H NMR results. The diffusion
coefficients for oleylamine after the undeceneamine titration were 301 ± 10, 232 ± 5, and 137 ± 4
µm s
–1
, for the nanocrystals synthesized at 230, 250, and 270 ˚C, respectively. The diffusion
coefficient and solvodynamic diameter based on oleylamine changes more with undeceneamine
titration for nanocrystal samples synthesized at lower temperatures, because the fraction of
strongly bound oleylamine is lower (vide supra). Variable-temperature
1
H NMR was then
performed on the suspension of Ge nanocrystals synthesized at 270 ˚C that had been titrated with
undeceneamine. The variable-temperature
1
H NMR showed that at higher temperatures (e.g., 90
˚C), the peaks corresponding to weakly bound oleylamine shift upfield suggesting less ligand
entanglement with the surface upon heating (Figure 2.6) The bound fraction of oleylamine
becomes harder to integrate as the temperature cools; nonetheless, the strongly bound fraction of
oleylamine remains fairly consistent at the various temperatures probed (Table 2.2).
43

Figure 2.6. Variable-temperature
1
H NMR spectra of 7 mg/mL Ge nanocrystal suspensions capped with
oleylamine, titrated with 1:3 (mol/mol) undeceneamine to oleylamine in toluene-d 8.

Table 2.2. %Bound oleylamine from Figure 2.6.
Temperature %Bound OAm
90 ˚C 89
50 ˚C 84
26 ˚C 85
-10 ˚C 79
-54 ˚C 76

To validate that the amine ligand exchange behavior is not unique to undeceneamine and
is not influenced by the terminal vinylic group, dodecylamine was titrated into a suspension of Ge
nanocrystals synthesized at 250 ˚C (Figure 2.7). Dodecylamine does not possess spectroscopically
unique NMR peaks that allow for easy integration; however, upon the titration of dodecylamine,
a peak separation and upfield chemical shift of the alkenyl protons corresponding to weakly bound
oleylamine is observed as before. This is consistent with dodecylamine competing for binding with
physisorbed oleylamine. Moreover, the fraction of strongly bound oleylamine remains statistically
unchanged before and after titration with dodecylamine. Since saturated primary alkylamines also
do not exchange the strongly bound native oleylamine, the vinylic group of undeceneamine is not
involved in the ligand exchange behavior, as expected.
44

Figure 2.7.
1
H NMR spectra of as-synthesized Ge nanocrystals prepared at 250 ˚C (orange), titrated with
1:2 (mol/mol) dodecylamine (DAm) to oleylamine (cyan), and free dodecylamine (pink) in toluene-d 8.

2.6. Ligand Exchange with CTAB
To further probe the nature of the native oleylamine ligand binding,
cetyltrimethylammonium bromide (CTAB) was next introduced in a
1
H NMR titration experiment.
Wheeler et al. previously showed that plasma-synthesized, hydride-terminated Ge nanocrystals
that are functionalized with oleylamine result in ionic ligand binding of oleylammonium cations
to a negatively charged Ge surface.
[32]
While undeceneamine and docecylamine should both be
able to displace ionically bound oleylammonium ligands via proton exchange, we further verified
that this binding mode is not operative here through titration with CTAB, which would also enable
ligand exchange of cationic oleylammonium species if they are present on the nanocrystal surface.
Upon titration of CTAB into a dichloromethane-d2 suspension of Ge nanocrystals synthesized at
250 ˚C, there appears to be no major changes to the alkenyl resonances of oleylamine; however,
this peak (ca. 5.33 ppm) overlaps with the residual solvent peak (ca. 5.32 ppm) (Figure 2.8). Due
to this peak overlap, DOSY was utilized to probe changes in the diffusion coefficient for bound
45

oleylamine and CTAB. The diffusion coefficient of oleylamine in dichloromethane-d2 did not
change significantly upon CTAB titration (i.e., 184 ± 5 µm s
–1
vs. 184 ± 8 µm s
–1
before and after
CTAB titration). Similarly, the diffusion coefficient of CTAB in the titration was determined to
be 842 ± 3 µm s
–1
, compared to 841 ± 3 µm s
–1
for free CTAB in dichloromethane-d2. Due to the
identical diffusion coefficients of CTAB in the presence of Ge nanocrystals and that of free CTAB,
we can conclude that CTAB is not binding to the nanocrystal surface.

Figure 2.8. (a) Full
1
H NMR spectra and (b) alkenyl region of as-synthesized Ge nanocrystals prepared at
250 ˚C (pink) and titrated with 6.3 mM CTAB in dichloromethane-d 2.

The complete lack of ligand exchange observed for the strongly bound native oleylamine
ligand fraction on the Ge nanocrystal surface with either neutral primary amines or ammonium
cations suggests that neither a dative L-type nor ionic bonding motif are operative. Instead, this
implies that the strongly bound native ligands are binding as oleylamide, with a covalent X-type
Ge–NHR bond, which is calculated to be ca. 60 kcal/mol.
[33]
Indeed, N–H bond dissociation has
been theoretically predicted to occur on Ge(100)-2´1 surfaces to give covalent Ge–N bonds,
[34]

which may be occurring during the microwave-assisted synthesis of these Ge nanocrystals. The
46

generation of H2 has been experimentally observed during the microwave-assisted reduction of
GeI2 in oleylamine,
[24]
which could result from N–H bond dissociation.

2.7. Ligand Exchange with Thiol and Carboxylic Acids
To explore exchange reactions with more acidic ligands, thiols and carboxylic acids were
next titrated into a suspension of Ge nanocrystals synthesized at 250 ˚C. Thiols have previously
been used as a supporting ligands for Ge nanocrystals to provide colloidal stability and electronic
passivation.
[1]
In that work, dodecanethiol was installed on the Ge nanocrystal synthesis by first
reducing off the native ligands with hydrazine and subsequently heating the nanocrystals at 150
˚C in an excess of thiol. Here, undecenethiol was titrated into a suspension of Ge nanocrystals at
room temperature and it was observed that the alkenyl peaks from oleylamine in the
1
H NMR
spectrum did not change. Likewise, the resonances corresponding to the vinylic protons on
undecenethiol did not shift upfield with increasing concentration, as was observed with
undeceneamine, and there is no evidence of undecenethiol binding as an X-type thiolate via proton
exchange with oleylamide (Figure 2.9). This implies that at room temperature there is no ligand
exchange or interaction with the Ge nanocrystals. Variable-temperature
1
H NMR of the Ge
nanocrystal suspension titrated with undecenethiol revealed a reduction in free undecenethiol with
a concomitant appearance of bound undecenethiol upon heating to 90 ˚C, as evidenced by small
resonances downfield of the free species (at ca. 5.13 and 5.92 ppm, Figure 2.9b). The amount of
undecenethiol that binds is small and therefore integration is approximate; nonetheless, total
binding of undecenethiol is ca. 14% relative to free undecenethiol. Consistent with this result, after
heating the Ge nanocrystal suspension titrated with undecenethiol at 90 ˚C for 2 h, two diffusion
47

Figure 2.9. (a) Room-temperature 600 MHz
1
H NMR spectra of Ge nanocrystal suspension in toluene-d 8
(7 mg/mL). The as-synthesized nanocrystals are capped with native oleylamine ligands and the suspension
is titrated with increasing amounts of free undecenethiol (UTh) (0-30 mM). (b) Superimposed room-
temperature
1
H NMR spectra of the suspension before and after heating at 90 ˚C for ca. 15 min, which
shows both free (F) and a small fraction of bound (B) peaks for undecenethiol (d » 5.13 and 5.92 ppm).

coefficients were detected for undecenethiol – 72 ± 10 µm s
-1
for bound undecenethiol(ate) and
1051 ± 11 µm s
-1
for free undecenethiol. The diffusion coefficient for oleylamine remains virtually
unchanged after heating the Ge nanocrystal suspension titrated with undecenethiol, with diffusion
coefficients of 146 ± 5 and 144 ± 5 µm s
-1
before and after heating, respectively. In combination
with
1
H NMR, this suggests that there is a high kinetic barrier for ligand exchange with
undecenethiol even with an exergonic thermodynamic driving force,
[1]
and that the incoming thiol
ligands may be occupying uncoordinated surface sites as X-type thiolates (via deprotonation by
physisorbed oleylamine) rather than exchanging oleylamide.
In order to further investigate the ligand exchange processes for Ge nanocrystals, undecenoic acid
was investigated as an even more acidic ligand (Figure 2.10). The
1
H NMR resonances
corresponding to the vinylic protons of undecenoic acid show a slight upfield chemical shift with
increasing concentration at room temperature, however, the change in chemical shift is not as
48

dramatic as that observed with the undeceneamine titration. With increasing concentrations of
undecenoic acid, the alkenyl oleylamine peak shifts slightly upfield and a set of small peaks appear
downfield of the free acid (at ca. 5.14 and 5.93 ppm) that correspond to strongly bound X-type
undecenoate. The suspension of titrated Ge nanocrystals was then heated in situ, and with
increasing temperature, the two peaks corresponding to strongly bound undecenoate more clearly
emerge downfield of their free species (Figure 2.11). After 14 h of heating at 90 ˚C, ca. 43% of
the undecenoic acid present is bound relative to the free fraction. After heating with titrated
undecenoic acid, the oleylamine alkenyl peak shape does not change greatly, suggesting that
undecenoic acid also may be primarily binding as X-type undecenoate (via deprotonation by
physisorbed oleylamine) to undercoordinated or available surface sites rather than exchanging the
strongly bound oleylamide species. Diffusion coefficients for undecenoic acid in the Ge
nanocrystal suspension were collected before and after heating to compare ligand binding (Table
2.3). The bound undecenoate peaks correspond to a diffusion coefficient of 109 ± 6 µm s
–

1
, which

Figure 2.10. Room-temperature 600 MHz
1
H NMR spectra of Ge nanocrystal suspension in toluene-d 8 (7
mg/mL). The as-synthesized nanocrystals are capped with native oleylamine ligands and the suspension is
titrated with increasing amounts of undecenoic acid (UAc) (0-50 mM).
49

Figure 2.11. 600 MHz
1
H NMR spectra of 7 mg/mL Ge nanocrystal suspensions capped with oleylamine,
titrated with 50 mM undecenoic acid in toluene-d 8 at room temperature (black), heated to 90 ˚C and cooled
back to room temperature for ca. 10 min (blue), and heated to 90 ˚C for 14 h and cooled back to room
temperature (purple). The blue and purple spectrum shows the presence of strongly bound undecenoic acid
after cooling (ca. 5.14 and 5.93 ppm).

Table 2.3. Summary of diffusion coefficients for oleylamine and undecenoic acid and the solvodynamic
diameters determined using DOSY NMR for Ge nanocrystals synthesized at 250 ˚C
Ligand
exchange
temperature
As-synthesized
oleylamine
diffusion coefficient
(µm s
-1
)
Oleylamine
diffusion
coefficient after
titration (µm s
-1
)
Undecenoic acid
diffusion coefficient
after titration (µm s
-1
)
Solvodynamic
diameter based
on oleylamine
(nm)
Room
temperature
187 ± 4 168 ± 3 551 ± 10 4.6
90 ˚C for 14 h 187 ± 4 161 ± 4 109 ± 6 (bound),
676 ± 10 (free)
4.8

is much lower than that of free undecenoic acid (945 ± 10 µm s
–1
), revealing that undecenoic acid
is indeed binding to the Ge nanocrystals upon heating. The diffusion coefficient of oleylamine
actually decreases upon heating the Ge nanocrystal suspension, implying a slightly higher fraction
of oleylamine binding, (i.e., 187 µm s
–1
vs. 161 µm s
–1
before and after heating). Again, this
demonstrates that the kinetic barrier to ligand exchange of the native oleylamide ligands is large,
even with highly acidic ligands.
50

Interestingly, when the titration was repeated with 10-undecylphosphonic acid, the
1
H
NMR spectrum only show bound phosphonic acid vinylic proton peaks. Due to limited solubility
of 10-undecylphosphonic acid in toluene-d8 we did not add excess. We show that the acidity of
the ligand does allow for the ligands to bind despite some sort of kinetic barrier that the thiol and
carboxylic acid are unable to overcome without the use of heating. Additionally, the amount of
integrated bound and free oleylamine after the titration and heating does not change. So, although
the phosphonic acid ligand can bind more readily than thiol and carboxylic acid, it is still not acidic
enough at this concentration to displace bound oleylamine.
In an effort to explore the effect of stronger acids on oleylamine-capped Ge nanocrystals,
we attempted to partially strip the bound oleylamine ligands via HCl titration. Upon the addition
of HCl (1.2 mM) and 4 h heating at 80 ˚C, 0.97 mM of bound oleylamine becomes
physisorbed/free. Additionally, we see the growth of a sharper peak upfield from the alkenyl
protons (Figure 2.12). This peak we attributed to the oleylammonium R-NH3
+
protons. To
corroborate this assumption and rule out the possibility of it being the alkenyl protons, we ran a
HSQC to confirm that these protons were not connected to any carbons (Figure 2.13). This
titration experiment demonstrates that the amine ligands can be removed from the nanocrystal
surface in the presence of a strong acid.

51

Figure 2.12. (A) Full
1
H NMR spectrum and (B) alkenyl region of as-synthesized Ge nanocrystals prepared
at 250 ˚C (black) and titrated with 1.2 mM aqueous HCl heated to 80 ˚C for 4 h (blue) in toluene-d 8. The
HCl solution was prepared in D 2O and 5 µL of the HCl/D 2O solution was added.

Figure 2.13. 2D
1
H-
13
C HSQC spectrum of Ge nanocrystal suspension synthesized at 250 ˚C and titrated
with 1.2 mM aqueous HCl in toluene-d 8. This spectrum does not show a cross peak for the N-H proton
resonance (d » 5.18 ppm), demonstrating that these protons are not attached to an oleyl carbon.

2.8. Conclusions
In conclusion, we investigated the surface chemistry of Ge nanocrystals synthesized via a
microwave reduction of GeI2 in oleylamine at various temperatures. We conclude that the strongly
52

bound native ligands were oleylamides, possessing a covalent X-type Ge–NHR bond, and that the
fraction of these strongly bound ligands on the surface increased with increasing synthesis
temperatures and translated into more colloidally stable nanocrystal suspensions. A high kinetic
barrier to ligand exchange with thiol and carboxylic acid ligands exists, and ligand binding is only
observed upon heating without the displacement, or exchange, of a significant amount of the native
oleylamide ligands. This confirms that the Ge nanocrystals, as synthesized, do not possess fully
passivated surfaces. These findings lead to a better understanding of the surfaces of these Ge
nanocrystals, and by extension other main group nanocrystals postulated to possess M–NHR
surface ligands, which will allow for rational surface modification procedures to maximize
nanocrystal functionality for a given application.

2.9. References
[1] E. Muthuswamy, J. Zhao, K. Tabatabaei, M. M. Amador, M. A. Holmes, F. E. Osterloh, S.
M. Kauzlarich, Chem. Mater. 2014, 26, 2138–2146.
[2] E. Fok, M. Shih, A. Meldrum, J. G. C. Veinot, Chem. Commun. 2004, 386.
[3] F. Li, J. Wang, S. Sun, H. Wang, Z. Tang, G. Nie, Small 2015, 11, 1954–1961.
[4] J.-W. Luo, A. Franceschetti, A. Zunger, Nano Lett. 2008, 8, 3174–3181.
[5] U. Kortshagen, R. Anthony, R. Gresback, Z. Holman, R. Ligman, C.-Y. Liu, L. Mangolini,
S. A. Campbell, Pure Appl. Chem. 2008, 80, 1901–1908.
[6] S. R. Smock, T. J. Williams, R. L. Brutchey, Angew. Chem. Int. Ed. 2018, 57, 11711–
11715.
[7] C. Bullen, P. Mulvaney, Langmuir 2006, 22, 3007–3013.
53

[8] M. J. Greaney, E. Couderc, J. Zhao, B. A. Nail, M. Mecklenburg, W. Thornbury, F. E.
Osterloh, S. E. Bradforth, R. L. Brutchey, Chem. Mater. 2015, 27, 744–756.
[9] J. J. Buckley, E. Couderc, M. J. Greaney, J. Munteanu, C. T. Riche, S. E. Bradforth, R. L.
Brutchey, ACS Nano 2014, 8, 2512–2521.
[10] M. J. Greaney, J. Joy, B. A. Combs, S. Das, J. J. Buckley, S. E. Bradforth, R. L. Brutchey,
J. Chem. Phys. 2019, 151, 074704.
[11] D. A. Ruddy, P. T. Erslev, S. E. Habas, J. A. Seabold, N. R. Neale, J. Phys. Chem. Lett.
2013, 4, 416–421.
[12] T. N. Lambert, N. L. Andrews, H. Gerung, T. J. Boyle, J. M. Oliver, B. S. Wilson, S. M.
Han, Small 2007, 3, 691–699.
[13] H. Gerung, S. D. Bunge, T. J. Boyle, C. J. Brinker, S. M. Han, Chem. Commun. 2005,
1914–1916.
[14] D. D. Vaughn, J. F. Bondi, R. E. Schaak, Chem. Mater. 2010, 22, 6103–6108.
[15] D. A. Ruddy, J. C. Johnson, E. R. Smith, N. R. Neale, ACS Nano 2010, 4, 7459–7466.
[16] R. J. Alan Esteves, S. Hafiz, D. O. Demchenko, Ü. Özgür, I. U. Arachchige, Chem.
Commun. 2016, 52, 11665–11668.
[17] A. L. Holmes, J. Hütges, A. Reckmann, E. Muthuswamy, K. Meerholz, S. M. Kauzlarich,
J. Phys. Chem. C 2015, 119, 5671–5678.
[18] A. T. Kerr, D. Placencia, M. E. Gay, J. E. Boercker, D. Soto, M. H. Davis, N. A. Banek,
E. E. Foos, J. Phys. Chem. C 2017, 121, 22597–22606.
[19] E. Muthuswamy, A. S. Iskandar, M. M. Amador, S. M. Kauzlarich, Chem. Mater. 2012,
25, 1416–1422.
54

[20] A. Bernard, K. Zhang, D. Larson, K. Tabatabaei, S. M. Kauzlarich, Inorg. Chem. 2018, 57,
5299–5306.
[21] M. He, L. Protesescu, R. Caputo, F. Krumeich, M. V. Kovalenko, Chem. Mater. 2015, 27,
635–647.
[22] K. Kravchyk, L. Protesescu, M. I. Bodnarchuk, F. Krumeich, M. Yarema, M. Walter, C.
Guntlin, M. V. Kovalenko, J. Am. Chem. Soc. 2013, 135, 4199–4202.
[23] G. M. Carroll, R. Limpens, N. R. Neale, Nano Lett. 2018, 18, 3118–3124.
[24] K. Tabatabaei, A. L. Holmes, K. A. Newton, E. Muthuswamy, R. Sfadia, S. A. Carter, S.
M. Kauzlarich, Chem. Mater. 2019, 31, 7510–7521.
[25] K. Tabatabaei, H. Lu, B. M. Nolan, X. Cen, C. E. McCold, X. Zhang, R. L. Brutchey, K.
van Benthem, J. Hihath, S. M. Kauzlarich, Chem. Mater. 2017, 29, 7353–7363.
[26] Z. Hens, J. C. Martins, Chem. Mater. 2013, 25, 1211–1221.
[27] L. E. Marbella, J. E. Millstone, Chem. Mater. 2015, 27, 2721–2739.
[28] S. Mourdikoudis, L. M. Liz-Marzán, Chem. Mater. 2013, 25, 1465–1476.
[29] N. Geva, J. J. Shepherd, L. Nienhaus, M. G. Bawendi, T. Van Voorhis, J. Phys. Chem. C
2018, 122, 26267–26274.
[30] R. R. Knauf, J. C. Lennox, J. L. Dempsey, Chem. Mater. 2016, 28, 4762–4770.
[31] B. Fritzinger, R. K. Capek, K. Lambert, J. C. Martins, Z. Hens, J. Am. Chem. Soc. 2010,
132, 10195–10201.
[32] L. M. Wheeler, A. W. Nichols, B. D. Chernomordik, N. C. Anderson, M. C. Beard, N. R.
Neale, Nano Lett. 2016, 16, 1949–1954.
[33] J. S. Kachian, K. T. Wong, S. F. Bent, Acc. Chem. Res. 2010, 43, 346–355.
[34] P. Prayongpan, C. Michael Greenlief, Surf. Sci. 2009, 603, 1055–1069.
55

Chapter 3. Probing the Ligand Exchange of N-Heterocyclic Carbene-Capped Ag2S
Nanocrystals with Amines and Carboxylic Acids

*Submitted manuscript.

3.1. Abstract
N-Heterocyclic carbenes (NHCs) are versatile L-type ligands that have been shown to
stabilize coinage metal chalcogenide nanocrystals, such as Ag2S, remarkably well. However, very
little research has been done on the interaction between NHC ligands and coinage metal
chalcogenide nanocrystal surfaces and subsequent ligand exchange reactions. Herein, solution
1
H
NMR methods were used to monitor ligand exchange reactions on stoichiometric Ag2S nanocrystal
platforms with various primary amine and carboxylic acid ligands. Despite the introduction of new
ligands, the native NHC ligands remain tightly bound to the Ag2S nanocrystal surface and are not
displaced at room temperature. Primary amine and carboxylic acid ligands only demonstrated
quantitative ligand exchange after heating the samples with excess incoming ligand, which implies
a strong NHC-Ag binding energy. Density functional theory affirms that a model NHC ligand
binds the strongest to a Ag12S6 cluster surface, followed by amine and carboxylic acid binding;
computational analysis is therefore in line with the absence of NHC displacement observed in
experiments. Both the bulky sterics of the C14-alkyl chains on the NHC and the strong binding
energies of NHC to the Ag2S surface contribute to the superior colloidal stability over conventional
long-chain amine or carboxylic acid ligands (many months vs. hours).

56

3.2. Introduction
N-Heterocyclic carbenes (NHCs) are noteworthy L-type ligands because they have easily
tunable steric and electronic properties, excellent chemical stability, and s-donation that results in
strong coordinating ability to a wide variety of metals.
[1]
NHCs display a singlet ground-state
electronic configuration due to the adjacent σ-electron-withdrawing and π-electron-donating
nitrogen atoms within the ligand framework that stabilize the structure by lowering the energy of
the σ-orbital in the highest occupied molecular orbital (HOMO), while simultaneously donating
electron density into the empty p-orbital in the lowest unoccupied molecular orbital (LUMO).
[1]

Resulting NHC-metal bonds typically possess high kinetic stability and large thermodynamic bond
strengths, as compared to more traditional L-type ligands, due to their superior electron donating
properties.
[2]

As a result of these desirable characteristics, NHC ligands are of interest for
functionalization and stabilization of colloidal nanocrystal surfaces. NHCs have been employed to
sterically stabilize a range of metal nanocrystals, including Au, Ag, Pd, Pt, Ru, Ir, and Cu
nanocrystals.
[3–9]
Nanocrystal size,
[4]
polydispersity,
[10]
and catalytic activity
[11]
can be tuned using
a range of different NHC ligands. For example, there have been several reports demonstrating the
tunability of NHC ligands to have an effect on various catalytic reactions, such as the activity and
selectivity of CO2 electroreduction with NHC-Au or Pd nanocrystals,
[12]
and the selectivity and
activity of hydrogenation of olefins with NHC-Pd or Cu nanocrystals.
[13–15]

Comparatively little work has been done on non-metallic NHC-stabilized nanocrystals
(e.g., binary metal chalcogenide or phosphides). We were the first to report the synthesis of NHC-
stabilized coinage metal chalcogenide (i.e., Ag2S, Ag2Se, Cu2–xS, CuS, Cu2–xSe and CuSe) and
phosphide (i.e., Cu3–xP) nanocrystals starting from NHC-MBr (M = Ag, Cu) precursors in the
57

absence of any additional supporting ligands.
[6,16,17]
We observed large differences in the colloidal
stability of NHC-Ag nanocrystals (~ 1 week) vs. NHC-Ag2S nanocrystals (> 6 months).
[16]
Density
functional theory (DFT) calculations revealed significantly larger binding affinities between the
NHC and the Ag2S nanocrystal surface over the Ag nanocrystal surface, with NHC binding
energies of Eb = 0.57-0.70 eV and Eb < 0.30 eV, respectively. The only feasible calculated C–Ag
bond length for the NHC-Ag surfaces was approximately 2.59 Å, whereas the C–Ag bond for
NHC-Ag2S nanocrystals ranged from 2.26-2.47 Å, thus indicating tighter ligand binding. Lastly,
we found that the relatively strong ligand binding of the NHC-Ag2S nanocrystals was additionally
reflected in the density of states, in which there was a strong modification of the carbene C orbitals
observed in the valence band.
[16]

While NHC ligands appear to provide excellent steric stabilization, the fundamental ligand
exchange chemistry of NHC-stabilized metal chalcogenide nanocrystals currently remains
unstudied. Herein, we explore the exchange of native NHC ligands with common amine and
carboxylic acid ligands on a colloidal Ag2S nanocrystal platform. Solution NMR spectroscopy was
used to explore binding of the native NHC ligands to the Ag2S nanocrystal surface and monitor
the system upon introduction of extraneous ligands. DFT calculations were carried out to support
the empirical observations. These results help explain the exceptional colloidal stability that has
been previously reported for these nanocrystals and reveal conditions needed to enact ligand
exchange.

3.3. Preparation and Characterization of NHC-Ag2S Nanocrystals
It was previously reported that a metathesis reaction between the bromo[1,3-
(ditetradecyl)benzimidazol-2-ylidene]silver(I) (NHC-AgBr) complex and
58

bis(trimethylsilyl)sulfide ((Me3Si)2S) under ambient conditions yields colloidal Ag2S
nanocrystals. In the absence of any extraneous ligands, the NHC ligands end up coordinated to the
nanocrystal surface and provide excellent colloidal stability.
[16]
Herein, we utilized the same
synthetic preparation for NHC-Ag2S nanocrystals. The resulting nanocrystals were confirmed by
powder XRD to be phase-pure monoclinic Ag2S (PDF no. 00-014-0072) (Figure 3.1), with an
approximate grain size of 5 nm, as determined by Scherrer analysis. An elemental ratio of Ag/S =
2.14 was obtained from SEM-EDS (Table 3.1), revealing the anticipated stoichiometric chemical
composition of Ag2S nanocrystals. The resulting as-prepared nanocrystals possess an average
diameter of 7.5 ± 1.2 nm (N = 300 counts) and a spherical morphology, as observed by TEM
analysis (Figure 3.1).

Figure 3.1. (a) Schematic of the NHC ligand used. (b) TEM micrograph of NHC-Ag 2S nanocrystals with
a measured average size of 7.5 ± 1.2 nm. (c) Powder X-ray diffraction pattern of NHC-Ag 2S nanocrystals,
with the stick pattern of monoclinic Ag 2S given below (PDF no. 00-014-0072).

Table 3.1. Table of SEM-EDS results for Ag 2S nanocrystals.
Sample Atomic% Ag Atomic% S Elemental Ratio Ag/S
NHC-Ag2S 68.2 31.8 2.14
oleylamine-Ag2S 67.7 32.3 2.09
oleic acid-Ag2S 66.7 33.3 2.00

59

3.4.
1
H NMR Spectroscopy to Determine the Surface Ligands of Ag2S Nanocrystals
Suspensions of NHC-Ag2S nanocrystals are colloidally stable under inert atmosphere for
several months, which can be attributed to steric stabilization from surface-bound NHC ligands.
The ligand density was determined to be ca. 1.5 NHC nm
–2
by TGA (Figure 3.2), which is similar
to the previously reported value for NHC-Ag2S nanocrystals.
[16]
This ligand density is appropriate
for the bulky sterics of the long tetradecyl C14-alkyl chains in the 1,3-positions of the NHC, which
have an end corner-to-end corner distance of ca. 1 nm.
[4]
Ligand binding of the native NHC ligands
to the Ag2S nanocrystals was further probed using solution
1
H NMR spectroscopy. The
1
H NMR
spectrum of an NHC-Ag2S nanocrystal suspension indicates binding of the carbene to the surface,
as shown by the significant broadening of peaks due to the restricted rotational freedom of bound
ligands on the surface (Figure 3.3). The asymmetry of the peaks most likely results from the
presence of NHC both tightly bound to the nanocrystal surface (i.e., the downfield and broader
peaks) and less tightly bound or physisorbed (i.e., those closer to the chemical shift of NHC-AgBr
with less broad peaks). We hypothesize that the physisorbed ligands result from protonated
carbene that would not be expected to strongly bind to the nanocrystal surface. There is ca. 30%
protonated carbene present relative to the entire ligand population by integration of the distinctive
C2-proton resonance at d = 11.7 ppm (denoted by D in Figure 3.3), which correlates closely to
integration of the physisorbed NHC peaks in the
1
H NMR spectrum, which are ca. 28% relative to
the broader downfield resonances. After using carefully dried solvents for synthesis, work-up, and
colloidal dispersion for NMR analysis, the fraction of protonated carbene can be decreased and the
asymmetry of the broadened peaks, is reduced (Figure 3.4).
60

Figure 3.2. TGA traces of NHC-Ag 2S (blue), oleylamine-Ag 2S (pink), and oleic acid-Ag 2S (purple)
nanocrystals.

Figure 3.3. (a) Solution
1
H NMR spectra of 7.5-nm NHC-Ag 2S nanocrystal suspension (30 µM) and
NHC-AgBr in CDCl 3 (denoted by *). Residual octadecene (ODE) reaction solvent is present after
nanocrystal synthesis and purification (denoted by ¡) with a 0.6 mM ferrocene standard (denoted by +).
About 30% of the ligands present in the nanocrystal suspension are protonated carbene (C2-proton
resonance denoted by D). (b) 2D DOSY NMR spectrum of 7.5-nm NHC-Ag 2S nanocrystal suspension (30
µM) in CDCl 3.
61

Figure 3.4.
1
H NMR spectra of (a) NHC-Ag 2S nanocrystals as reported, and (b) NHC-Ag 2S nanocrystals
prepared using carefully dried solvents and dispersed in dried CDCl 3.

Diffusion ordered NMR spectroscopy (DOSY) was performed on a suspension of NHC-
Ag2S nanocrystals to gather binding information for the NHC ligands present (Figure 3.3). Similar
to previous reports,
[6]
the average diffusion coefficient of the NHC ligand in CDCl3 is 279 µm
2
s
-1
,
which correlates to a solvodynamic diameter of 2.9 nm (diameter of both the nanocrystal and the
organic ligand shell) using the Stokes-Einstein equation. The diffusion coefficient of NHC-Ag2S
is significantly smaller than the NHC-AgBr precursor (621 µm
2
s
–1
) (Table 3.2). The decrease in
the experimental diffusion coefficient confirms the interaction of NHC with the Ag2S nanocrystal
surface; however, the experimental diffusion coefficient as determined by the Stokes-Einstein
equation is higher than would be expected for a 7.5 nm spherical nanocrystal with a 2 nm organic
ligand shell (theoretical diffusion coefficient = 86 µm
2
s
–1
). This could mean that either (1) there
are two separate and distinct populations of strongly bound NHC and completely free, protonated
NHC present, or (2) the NHC ligands are fluxional on the nanocrystal surface. Using the
62

Table 3.2. Diffusion coefficients of Ag 2S nanocrystals and their solvodynamic diameter calculated using
diffusion coefficient of the ligand in parentheses and Stokes-Einstein equation.
Sample NHC Diffusion
Coefficient (µm
2

s
–1
) (before
titration)
NHC Diffusion
Coefficient (µm
2

s
–1
) (after
titration)
Solvodynamic
Diameter (nm)
(Ligand)
Incoming Ligand
Diffusion
Coefficient (µm
2

s
–1
)
NHC-Ag 2S 279 N/A 2.9 (NHC) N/A
oleylamine
titration
285 285 2.9 (NHC) 625
oleylamine-Ag 2S N/A N/A 1.0 (oleylamine) 780
butylamine
titration
298 304 2.7 (NHC) 763
oleic acid titration 249 250 3.3 (NHC) 406
oleic acid-Ag 2S N/A N/A 1.5 (oleic acid) 557
acetic acid
titration
267 273 3.0 (NHC) 834

experimentally determined concentrations of ~70% strongly bound NHC and ~30% protonated
carbene, this yields a calculated average diffusion coefficient of 243 µm
2
s
–1
, which is in good
agreement with the experimental value and suggests that the former scenario is a possibility.
While the DOSY spectra may imply two distinct ligand populations, we also performed
presaturation (PRESAT) NMR experiments, in which various proton resonances were selectively
saturated, to probe potential exchange dynamics between the broader bound and sharper
physisorbed resonances (Figure 3.5).
[18]
It was found that when the broader bound proton
resonances were saturated, the intensity of the sharper physisorbed resonances were quenched.
Conversely, when the sharper physisorbed resonances were saturated, the intensity of the broader
bound resonances decreased. This suggests that some degree of exchange between the bound and
physisorbed states does occur on the PRESAT timescale (i.e., 2 s). Therefore, there is some degree
of exchange occurring between the bound and physisorbed NHC ligand states that is slower than
the DOSY timescale.
63

Figure 3.5. Selective presaturation of
1
H NMR for NHC-Ag 2S nanocrystals. (a)
1
H NMR spectrum of
NHC-Ag 2S as-is without saturation. Selective presaturation of
1
H NMR spectra of (b) proton resonance at
d = 8.07 ppm, (c) the proton resonance at d = 11.65 ppm, (d) proton resonance at d = 4.63 ppm, (e) the
proton resonance at d = 7.68 ppm, and (f) a duplicate for the saturation of the proton resonance at d = 8.07
ppm.

3.5. Ligand Exchange with Primary Amines
After characterizing the native NHC ligand binding on the Ag2S nanocrystals, we shifted
our attention to ligand exchange reactions, beginning with a common L-type ligand in colloidal
nanocrystal chemistry — primary alkylamines. Specifically, we chose oleylamine because it
possesses spectroscopically distinct alkenyl resonances that do not overlap with the native NHC
ligand in the
1
H NMR spectrum. Additionally, oleylamine is commonly found in many colloidal
nanocrystal synthesis or ligand exchange reactions, including with Ag2S nanocrystals, therefore
making it an interesting ligand to explore.
[19,20]
Upon titrating increasing amounts of the new
oleylamine ligand into a NHC-Ag2S colloidal nanocrystal suspension at room temperature, there
are no significant differences observed for the native NHC ligands or oleylamine by
1
H NMR
(Figures 3.6-3.7). We note that there are two peaks located in the alkenyl region for oleylamine at
5.35 and 5.38 ppm, which are attributed to the cis- and trans-isomers of oleylamine in CDCl3,
[21]

64

Figure 3.6. (a) Alkenyl region of solution
1
H NMR spectra of 7.5-nm NHC-Ag 2S nanocrystals (30 µM)
dispersed in CDCl 3 and titrated with 0.3-1.5 µmol oleylamine. (b)
1
H NMR spectra of NHC-Ag 2S
nanocrystals (blue) and oleylamine-Ag 2S nanocrystals after forced ligand exchange (pink) dispersed in
CDCl 3 (denoted by *).

respectively, that are not resolved in toluene-d8 (Figure 3.8). To clarify if oleylamine is interacting
at all with the NHC-Ag2S nanocrystal surface, we performed DOSY NMR and calculated a
diffusion coefficient of 625 µm
2
s
–1
for oleylamine titrated into the nanocrystal suspension (Table
3.2), which is smaller than the diffusion coefficient for free oleylamine in CDCl3 (925 µm
2
s
–1
).
This implies that there may be some degree of interaction between oleylamine and the Ag2S
nanocrystal surface or NHC ligand shell, most likely arising from interdigitation of oleylamine in
the ligand shell. Importantly, the NHC ligand diffusion coefficient remains unchanged (285 µm
2

s
–1
) when compared to the as-prepared NHC-Ag2S nanocrystals (279 µm
2
s
–1
), suggesting no
change to NHC ligand binding upon introduction of oleylamine (Table 3.2). From this, it is unclear
whether oleylamine does not exchange the native NHC ligands off the Ag2S nanocrystal surface
based on the binding strength of the NHC ligands, or because it simply cannot access the surface
because of the combined steric bulk of the NHC and oleylamine.

65

Figure 3.7.
1
H NMR spectra of 7.5-nm NHC-Ag 2S nanocrystal suspension, titrated with 0.3-1.5 µmol
oleylamine (top) and oleylamine (bottom) in CDCl 3. The alkenyl region of the spectra where oleylamine is
titrated into the colloidal. Nanocrystal suspension is shown as an inset.

Figure 3.8. (a)
1
H NMR spectra of oleylamine in toluene-d 8 (top) and CDCl 3 (bottom). (b) Alkenyl region
of
1
H NMR spectra in toluene-d 8 (top) and CDCl 3 (bottom) demonstrating that cis-/trans-isomers are
distinguishable in CDCl 3, but not toluene-d 8.

To evaluate both possibilities, we titrated n-butylamine, a short C4 primary amine, into the
colloidal suspension of Ag2S nanocrystals at room temperature. Butylamine should be small
enough to access the nanocrystal surface if the combined steric bulk of the NHC and the long-
chain oleylamine ligand prohibited the larger primary amine from accessing the surface. The
66

resulting
1
H NMR spectrum does reveal a downfield shift in the a-proton resonance of butylamine,
implying ligand binding to unpassivated sites on the nanocrystal surface or entanglement with the
native NHC ligand shell (Figure 3.9). Correspondingly, the diffusion coefficient of butylamine
decreased significantly from 1933 µm
2
s
–1
for free butylamine to 763 µm
2
s
–1
(Tables 3.2-3.3);
however, the diffusion coefficient for the NHC ligands remain unchanged (298 µm
2
s
–1
vs. 304
µm
2
s
–1
). The lack of clear displacement of NHC by amines suggests that although the NHC ligands
are sterically bulky, the main reason for lack of ligand exchange is likely due to the binding
strength of the NHC ligands, which we explored further using DFT (vide infra).

Figure 3.9.
1
H NMR spectra of 7.5-nm NHC-Ag 2S nanocrystal suspension, titrated with 1-5 µmol
butylamine (top) and butylamine (bottom) in CDCl 3. The a-proton region of the spectra where butylamine
is titrated into the colloidal nanocrystal suspension is shown as an inset.

Table 3.3. Summary of diffusion coefficients of molecular species in CDCl 3.
Sample Diffusion Coefficient (µm
2
s
–1
)
NHC-Br 608
NHC-AgBr 621
oleylamine 925
butylamine 1933
oleic acid 647
acetic acid 2134
67

Since neither oleylamine nor butylamine were successful in displacing the native NHC
ligands at room temperature, we tried a forced ligand exchange reaction by heating NHC-Ag2S
nanocrystals at elevated temperatures in the presence of excess oleylamine (85 ˚C, 2 h). After work
up, the resulting Ag2S nanocrystals maintained the monoclinic crystal structure and displayed a
distinct TGA mass loss trace compared to the starting NHC-Ag2S nanocrystals (Figures 3.2, 3.10).
The size of the resulting ligand exchanged nanocrystals changed slightly (6.2 ± 1.4 nm) but is
within a standard deviation of the starting NHC-Ag2S nanocrystals (7.5± 1.2 nm). Additionally,
the ligand exchanged oleylamine-Ag2S nanocrystal elemental ratio was unchanged (Ag/S = 2.09),
as determined by SEM-EDS.
1
H NMR suggests a quantitative ligand exchange due to the lack of
any NHC resonances, particularly in the diagnostic aromatic region (ca. 7.5-8.6 ppm) (Figure 3.6).
The
1
H NMR spectrum is consistent with an oleylamine-passivated surface, as seen by the
signature alkenyl resonances at 5.3 ppm. Qualitatively, the resonances for the ligand exchanged
oleylamine-Ag2S nanocrystals are significantly less broad than for the NHC-Ag2S nanocrystals;
therefore, the oleylamine-Ag2S nanocrystals are less strongly bound and have more free ligand
present. As a result, the oleylamine-Ag2S nanocrystals could only be washed once because they
were not very colloidally stable, thus demonstrating the added benefit of NHC surface termination.
The diffusion coefficient of the oleylamine-Ag2S nanocrystals was 708 µm
2
s
–1
, which is smaller
than free oleylamine (925 µm
2
s
–1
) but larger than what would be expected for strongly bound
oleylamine (ca. 71 µm
2
s
–1
) (Tables 3.2-3.3), which is again reflective of the fluxional binding
behavior of oleylamine. The same forced ligand exchange (85 ˚C, 2 h) procedure was repeated for
butylamine, which resulted in nanocrystals that completely crashed out of hexanes after work-up.
This suggests that ligand exchange with butylamine was successful, however, butylamine is not
large enough to provide steric stabilization for a stable colloidal suspension, as expected.
68

Figure 3.10. Powder X-ray diffraction patterns of NHC-Ag 2S (blue), oleylamine-Ag 2S (pink), and oleic
acid-Ag 2S (purple) nanocrystals.

3.6. Ligand Exchange with Carboxylic Acids
Due to the lack of facile ligand exchange with amines, we turned to carboxylic acids with
the idea that a lower pKa may protonate and displace the native NHC ligand. Oleic acid was
selected because it is the analogous C18 carboxylic acid to oleylamine and also possesses the
spectroscopically distinct alkenyl resonances that do not overlap with the native NHC ligand. Upon
titration of oleic acid into a suspension of NHC-Ag2S nanocrystals in CDCl3 at room temperature,
the incoming oleic acid resonances are sharp and appear unbound; for example, the alkenyl
resonances at 5.3 ppm do not shift and remain sharp (Figures 3.11-3.12). Additionally, the NHC
resonances remain unchanged, particularly in the aromatic region (ca. 7.5-8.6 ppm). To clarify if
oleic acid is interacting at all with the nanocrystal surface, we performed DOSY NMR, which
69

Figure 3.11. (a) Alkenyl region of solution
1
H NMR spectra of 7.5-nm NHC-Ag 2S nanocrystals (30 µM)
dispersed in CDCl 3 and titrated with 0.3-1.6 µmol oleic acid. (b)
1
H NMR spectra of NHC-Ag 2S
nanocrystals (blue) and oleic acid-Ag 2S nanocrystals after forced ligand exchange (purple) dispersed in
CDCl 3 (denoted by *).

returned a diffusion coefficient of 406 µm
2
s
–1
, which is smaller than free oleic acid (647 µm
2
s
–1
)
(Tables 3.2-3.3). Similar to oleylamine, this implies that there may be some degree of interaction
between oleic acid and the NHC ligand shell, most likely through interdigitation as the NHC ligand
diffusion coefficient remains unchanged (ca. 250 µm
2
s
–1
). Again, it is unclear whether oleic acid
does not bind to nanocrystal surface because it cannot displace the NHC based on binding strength,
or because it simply cannot access the surface due to steric bulk. Therefore, we performed an
analogous titration with acetic acid in order to explore if smaller carboxylic acids could displace
the native NHC ligands.
We titrated in acetic acid, a short C2-carboxylic acid, into a suspension of the NHC-Ag2S
nanocrystals at room temperature. The
1
H NMR spectrum does not show any clear indication of
ligand exchange, as the resonances corresponding to the native NHC ligands, such as the aromatic
protons at ca. 7.5-8.6 ppm, remain unchanged (Figure 3.13). However, the diffusion coefficient
of the acetic acid greatly decreased from 2134 µm
2
s
–1
for free acetic acid to 834 µm
2
s
–1
(Tables
70

Figure 3.12.
1
H NMR spectra of 7.5-nm NHC-Ag 2S nanocrystal suspension, titrated with 0.3-1.6 µmol
oleic acid (top) and oleic acid (bottom) in CDCl 3. The alkenyl region of the spectra where oleic acid is
titrated into the colloidal nanocrystal suspension is shown as an inset.

3.2-3.3). Similar to butylamine, this implies that there may be some degree of interaction between
acetic acid and the nanocrystal surface or NHC ligand shell; however, the NHC ligand diffusion
coefficient remains unchanged (from 267 before titration to 273 µm
2
s
–1
after titration). Thus, while
it appears that both oleic acid and acetic acid have some association with the surface, they do not
displace the native NHC ligands and bind as X-type carboxylates to liberate protonated carbene at
room temperature.
Because neither carboxylic acid titrations resulted in displacement of the native NHC
ligands, we tried a forced ligand exchange reaction by heating a suspension of NHC-Ag2S
nanocrystals at elevated temperatures in the presence of excess oleic acid (85 ˚C, 2 h). After work-
up, the resulting oleic acid-Ag2S nanocrystals had a distinct TGA mass loss trace compared to
NHC-Ag2S nanocrystals (Figure 3.2). By TEM analysis, the size of the oleic acid-Ag2S
nanocrystals was 9.2 ± 2.6 nm, which is slightly larger than the native NHC-Ag2S nanocrystals
71

Figure 3.13.
1
H NMR spectra of 7.5-nm NHC-Ag 2S nanocrystal suspension, titrated with 1.7-8.7 µmol
acetic acid in CDCl 3. The methyl resonance region of the spectra for acetic acid being titrated into the
colloidal nanocrystal suspension is shown as an inset.

with a larger polydispersity, suggesting some degree of particle ripening. However, the
nanocrystals remain stoichiometric, with Ag/S = 2.00 by SEM-EDS.
1
H NMR shows quantitative
ligand exchange due to the lack of any NHC resonances, particularly in the diagnostic aromatic
region for the NHC ligand (ca. 7.5-8.6 ppm) (Figure 3.11). The
1
H NMR spectrum is consistent
with an oleic acid-passivated surface, as evidenced by the signature alkenyl resonance at 5.3 ppm.
The resonances for the oleic acid-Ag2S nanocrystal suspension are all quite sharp in the
1
H NMR
spectrum, except for the carboxylic acid proton, which is observed at 10.9 ppm and is very broad
(Figure 3.14). The diffusion coefficient for oleic acid in the resulting Ag2S nanocrystal suspension
was 557 µm
2
s
–1
, which is slightly smaller than that for free oleic acid in CDCl3 (647 µm
2
s
–1
)
(Tables 3.2-3.3). Similar to the oleylamine-Ag2S nanocrystals, the oleic acid-Ag2S nanocrystal
suspensions were not colloidally stable and completely crashed out of solution after about 10 h.
The relative colloidal instability of oleic acid-Ag2S and oleylamine-Ag2S nanocrystal suspensions
72

is important to compare to the colloidal stability of NHC-Ag2S nanocrystal suspensions, which
can remain colloidally stable for many months.

Figure 3.14.
1
H NMR spectra of 7.5-nm oleic acid-Ag 2S nanocrystal suspension (top) and free oleic acid
(bottom) in CDCl 3 (bottom, black).

3.7. DFT Analysis of Ligand Exchange
The ligand-Ag12S6 cluster formation energies, structures, and distance between cluster Ag
and the coordinating ligand atom are shown in Figure 3.15. The model NHC ligand binds the
strongest to the Ag12S6 cluster with a formation energy of -160.3 kJ/mol and the model carboxylic
acid binds the weakest (formation energy = -82.6 kJ/mol). Computational analysis is therefore in
line with the absence of NHC displacement observed in experiments.
To understand the driving forces behind the stronger binding of NHC to the nanocrystal
surface, we employ energy decomposition analysis (EDA), which breaks down ligand-cluster
interactions into three key constituents. The ‘frozen’ energies correspond to the cumulative effect
of Pauli repulsions, electrostatics, and dispersion interactions between the two fragments.

73

Ligand Geometry Formation energy (kJ/mol) Distance (Å)
NHC -160.3 Ag-C = 2.190
Amine

-118.6 Ag-N = 2.381
Carboxylic acid

-82.6 Ag-O = 2.329
Figure 3.15. DFT results providing minimum energy structures of ligand-capped Ag 12S 6 clusters, formation
energies, and cluster-ligand bond distances for model NHC, amine, and carboxylic acid ligands. Model
visualizations are created using the Envision package.
[22]

Polarization energies represent intrafragment relaxation that occurs when one fragment polarizes
the other. Finally, charge transfer energies correspond to the stabilization associated with complete
interfragment relaxation and charge transfer. As shown in Figure 3.16, we find that favorable
NHC binding is driven by both interfragment charge transfer as well as polarization interactions.
While the amine and carboxylic acid ligands lead to similar polarization and charge transfer terms,
the frozen term in the amine (dominated by dispersion) leads to more favorable binding compared
to the neutral carboxylic acid.

74

Figure 3.16. Energy decomposition analysis of ligand-Ag 12S 6 cluster interactions. The total interaction
energies, obtained as the sum of the frozen, polarization, and charge transfer terms, are -179.6 kJ/mol, -
113.7 kJ/mol, and -81.4 kJ/mol for NHC, amine, and carboxylic acid ligands, respectively.

3.8. Conclusions
In conclusion, we investigated the surface chemistry of colloidal Ag2S nanocrystals
synthesized with N-heterocyclic carbene ligands. The native NHC ligands are both strongly bound
and sterically bulky, which provides excellent colloidal stability and prevents room temperature
ligand exchange reactions with traditional long chain oleylamine or oleic acid ligands. Although
analogous smaller ligands, such as butylamine and acetic acid, may interact with the nanocrystal
surface or interdigitate the native ligand shell more so than the long-chain analogs, neither displace
the native NHC at room temperature either. Only upon heating and in the presence of excess
incoming ligand can ligand exchange readily occur to afford complete displacement of the NHC
ligands. The parent NHC-Ag2S nanocrystals are much more colloidally stable (i.e., for many
months) than either the resulting oleylamine- or oleic acid-Ag2S nanocrystals from forced ligand
75

exchanges (i.e., hours), thus demonstrating the utility of NHC ligands. These trends in colloidal
stability are well supported by reasonable computational models of this system.

3.9. References
[1] M. N. Hopkinson, C. Richter, M. Schedler, F. Glorius, Nature 2014, 510, 485–496.
[2] R. H. Crabtree, J. Organomet. Chem. 2005, 690, 5451–5457.
[3] J. Vignolle, T. D. Tilley, Chem. Commun. 2009, 7230–7232.
[4] X. Ling, S. Roland, M.-P. Pileni, Chem. Mater. 2015, 27, 414–423.
[5] S. G. Song, C. Satheeshkumar, J. Park, J. Ahn, T. Premkumar, Y. Lee, C. Song,
Macromolecules 2014, 47, 6566–6571.
[6] H. Lu, Z. Zhou, O. V. Prezhdo, R. L. Brutchey, J. Am. Chem. Soc. 2016, 138, 14844–
14847.
[7] E. A. Baquero, S. Tricard, J. C. Flores, E. de Jesús, B. Chaudret, Angew. Chem. Int. Ed.
2014, 53, 13220–13224.
[8] P. Lara, O. Rivada-Wheelaghan, S. Conejero, R. Poteau, K. Philippot, B. Chaudret,
Angew. Chem. Int. Ed. 2011, 50, 12080–12084.
[9] M. Valero, D. Bouzouita, A. Palazzolo, J. Atzrodt, C. Dugave, S. Tricard, S. Feuillastre,
G. Pieters, B. Chaudret, V. Derdau, Angew. Chem. Int. Ed. 2020, 59, 3517–3522.
[10] M. J. MacLeod, J. A. Johnson, J. Am. Chem. Soc. 2015, 137, 7974–7977.
[11] Z. Cao, D. Kim, D. Hong, Y. Yu, J. Xu, S. Lin, X. Wen, E. M. Nichols, K. Jeong, J. A.
Reimer, P. Yang, C. J. Chang, J. Am. Chem. Soc. 2016, 138, 8120–8125.
[12] L. Zhang, Z. Wei, S. Thanneeru, M. Meng, M. Kruzyk, G. Ung, B. Liu, J. He, Angew.
Chem. Int. Ed. 2019, 58, 15834–15840.
76

[13] C. Richter, K. Schaepe, F. Glorius, B. J. Ravoo, Chem. Commun. 2014, 50, 3204.
[14] A. Ferry, K. Schaepe, P. Tegeder, C. Richter, K. M. Chepiga, B. J. Ravoo, F. Glorius,
ACS Catal. 2015, 5, 5414–5420.
[15] N. Kaeffer, H.-J. Liu, H.-K. Lo, A. Fedorov, C. Copéret, Chem. Sci. 2018, 9, 5366–5371.
[16] H. Lu, R. L. Brutchey, Chem. Mater. 2017, 29, 1396–1403.
[17] B. A. Tappan, K. Chen, H. Lu, S. M. Sharada, R. L. Brutchey, ACS Appl. Mater.
Interfaces 2020, 12, 16394–16401.
[18] S. R. Smock, T. J. Williams, R. L. Brutchey, Angew. Chem. Int. Ed. 2018, 57, 11711–
11715.
[19] X. Hou, X. Zhang, W. Yang, Y. Liu, X. Zhai, Mater. Res. Bull 2012, 47, 2579–2583.
[20] D. Ruiz, M. Mizrahi, H. D. A. Santos, D. Jaque, C. M. S. Jones, J. Marqués-Hueso, C.
Jacinto, F. G. Requejo, A. Torres-Pardo, J. M. González-Calbet, B. H. Juárez, Nanoscale
2019, 11, 9194–9200.
[21] D. Baranov, M. J. Lynch, A. C. Curtis, A. R. Carollo, C. R. Douglass, A. M. Mateo-
Tejada, D. M. Jonas, Chem. Mater. 2019, 31, 1223–1230.
[22] F. Manby, T. Miller, P. Bygrave, F. Ding, T. Dresselhaus, F. Batista-Romero, A.
Buccheri, C. Bungey, S. Lee, R. Meli, K. Miyamoto, C. Steinmann, T. Tsuchiya, M.
Welborn, T. Wiles, Z. Williams, Entos: A Quantum Molecular Simulation Package,
2019.

77

Chapter 4. Quantifying the Thermodynamics of Ligand Binding to CsPbBr3
Quantum Dots

*Published in Angew. Chem. Int. Ed. 2018, 57, 11711-11715.

4.1. Abstract
Cesium lead halide perovskites are an emerging class of quantum dots (QDs) that have
shown promise in a variety of applications; however, their properties are highly dependent on their
surface chemistry. To this point, the thermodynamics of ligand binding remain unstudied.
1
H NMR
methods are used to quantify the thermodynamics of ligand exchange on CsPbBr3 QDs. Both oleic
acid and oleylamine native ligands dynamically interact with the CsPbBr3 QD surface, having
individual surface densities of 1.2–1.7 nm
-2
. 10-Undecenoic acid undergoes an exergonic exchange
equilibrium with bound oleate (Keq = 1.97) at 25 ˚C while 10-undecenylphosphonic acid undergoes
an irreversible ligand exchange. Undec-10-en-1-amine exergonically exchanges with oleylamine
(Keq = 2.52) at 25 ˚C. Exchange occurs with carboxylic acids, phosphonic acids, and amines on
CsPbBr3 QDs without etching the nanocrystal surface; increases in steady-state PL intensities are
correlated with more strongly bound conjugate base ligands.

4.2. Introduction
Quantum dot (QD) properties are heavily influenced by the coordination environment of
their surface.
[1]
The ligands coordinated to their surfaces affect both the chemical and colloidal
stability of the QD, in addition to its ensuing electronic structure. An up-and-coming class of QDs
is the colloidal lead halide perovskites (APbX3, where A = CH3NH3
+
, CH(NH2)2
+
, Cs
+
; X = Cl
–
,
Br
–
, I
–
) because of their excellent optoelectronic properties, such as bright photoluminescence (PL)
78

with narrow spectral line widths that cover a wide color gamut.
[2–4]
These properties have made
lead halide perovskite QDs intriguing for use in a multitude of optical devices, such as LEDs
[5–7]

and solar cells.
[3,8]
Unlike the more traditional II-VI QDs, lead halide perovskite QDs are much
more ionic in their bonding, including bonding with ligands. Accordingly, polar solvents promote
ligand desorption, resulting in loss of colloidal stability and PL quantum yield,
[9]
and in some
cases, QD dissolution may occur. Because of these challenges, and the relative newness of these
QD materials, more in depth and quantitative studies on the surface chemistry of these lead halide
perovskite QDs are needed.
De Roo et al.
[10]
analyzed the surface ligands of as-prepared CsPbBr3 QDs synthesized by
the prototypical hot-injection method developed by Protesescu et al.,
[11]
whereby equal volumes
of oleylamine and oleic acid were added after QD purification to promote colloidal stability. They
determined, using solution
1
H NMR spectroscopy, that ligand binding is highly dynamic, and that
oleylamine selectively binds to the surface as oleylammonium bromide in an NC(X)2 binding
motif. Only in the presence of excess oleylamine added after purification does oleic acid bind to
the surface, in the form of oleylammonium oleate. However, as a result of the nearly identical
chemical environments of the diagnostic alkenyl protons for both oleylamine and oleic acid,
quantitative analysis of the separate native ligands was rendered impossible. Herein, for the first
time, we quantify the thermodynamics of ligand binding to CsPbBr3 QDs using a method
developed by Knauf et al. through the use of carboxylic acid, phosphonic acid, and amine-based
ligands possessing a terminal vinyl group that is spectroscopically distinct from the internal alkenyl
protons of the native oleylamine and oleic acid ligands.
[12]
This allows the free and bound fractions
of both ligands to be simultaneously tracked, thereby providing valuable thermodynamic data on
ligand binding for this important class of QD materials.
79

4.3. Preparation and Characterization of CsPbBr3 Quantum Dots
The CsPbBr3 QDs used here were synthesized using a modified hot-injection method as
first reported by Protesescu et al.,
[11]
with diphenyl ether being substituted for 1-octadecene as the
primary reaction solvent as to avoid spectral overlap in the vinylic region of the
1
H NMR spectra
(vide infra). To study carboxylate surface binding, CsPbBr3 QDs were synthesized with oleic acid
and dodecylamine, thereby eliminating spectral overlap in the internal alkenyl region of the
1
H
NMR spectrum between the native ligands. In contrast to the previous work by De Roo et al.,
[10]

no additional ligands were added during or after QD purification. The resulting CsPbBr3 QDs were
characterized by powder X-ray diffraction (XRD) and transmission electron microscopy (TEM) to
confirm the orthorhombic phase
[13]
and cuboidal morphology of the 12.4-nm QDs obtained
through our modified synthesis (Figures 4.1, 4.2a).

Figure 4.1. XRD patterns of drop-cast films of CsPbBr 3 QDs synthesized with oleic acid (OAc) and
dodecylamine (DAm) [blue] and CsPbBr 3 QDs synthesized with oleylamine (OAm) and lauric acid (LAc)
[pink].

80

Figure 4.2. TEM micrographs of CsPbBr 3 QDs (a) OAm and OAc coated QDs synthesized with ODE as
solvent (as reference), (b) OAc and DAm capped QDs synthesized with DPE as solvent, and (c) OAm and
LAc capped QDs synthesized with DPE as solvent.

4.4.
1
H NMR Spectroscopy to Determine the Surface Ligands of CsPbBr3 Quantum Dots
A representative solution
1
H NMR spectrum of a suspension of purified CsPbBr3 QDs is
given in Figure 4.3; as is typical in colloidal QD systems, the resonances corresponding to the
coordinated ligands are shifted and broadened relative to that of their free ligand counterparts. We
focus specifically on the diagnostic alkenyl region (d = 5.4–5.9 ppm) of the
1
H NMR spectrum,
which allows for ready analysis of oleic acid binding without interference from the saturated amine
ligand. Line broadening and a downfield shift are observed in the alkenyl region of oleic acid (d =
5.73 ppm vs 5.54 ppm for free oleic acid), which implies ligand interaction with the QD surface.
Since excess dodecylamine is not present in this system, this result reveals that an excess of amine
is not required for oleic acid binding to the CsPbBr3 QD surface, as has been previously reported
81

Figure 4.3.
1
H NMR spectrum of 12.4-nm CsPbBr 3 QDs (1.6 mM) synthesized with oleic acid (OAc) and
dodecylamine (DAm) and dispersed in toluene-d 8 (denoted by *) with a 0.3 µM ferrocene standard.
Residual diphenyl ether (DPE) reaction solvent is present after purification. The individual
1
H NMR spectra
of oleic acid and dodecylamine in toluene-d 8 are given for comparison.

(albeit for a slightly different synthesis and purification method).
[10]
Additionally, there is a third
alkenyl peak (d = 5.65 ppm) that we assign to physisorbed oleic acid, possessing both a chemical
shift and line width intermediate between the chemisorbed and the free ligand peaks. This
physisorbed species can be described as being entangled (or interdigitated) in the primary, more
tightly bound ligand shell, as has been previously observed with oleic acid bound to CdSe QDs.
[14]

To investigate the exchange amongst the three different alkenyl resonances of oleic acid, we
utilized selective presaturation, and saturated each
1
H NMR peak (bound, physisorbed, and free)
to understand the exchange dynamics between the three states (Figure 4.4). It was found that when
the free peak was saturated, neither the bound nor free peak intensities are altered, implying that
the exchange between free and physisorbed or bound oleic acid is slow (i.e., > 2 s). Additionally,
it was demonstrated that when the bound proton peak was saturated, the physisorbed proton peak
intensity decreased. Conversely, when the physisorbed proton peak was saturated, the bound
82

Figure 4.4. (a) Selective presaturation of
1
H NMR for the solvent and alkenyl region of oleic acid and
dodecylamine capped CsPbBr 3 QDs (pink is
1
H NMR spectrum as is, orange is for the saturation of free
oleic acid, green is for the saturation of physisorbed oleic acid, blue is for the saturation of the bound oleic
acid). (b) Zoom in on the alkenyl region with increased intensities to show the bound and physisorbed peaks
more clearly.

proton peak intensity decreased. This suggests that exchange between the bound and physisorbed
state is happening within a 2 s timescale. Based on the concentration of CsPbBr3 QDs determined
by UV-vis spectroscopy, the bound oleate surface density was calculated to be 1.2–1.5 oleate nm
-2

by integrating the bound alkenyl resonance of oleic acid against an internal ferrocene standard.
Because the ligands bind in a NC(X)2 motif, we can assume that for every oleate, there is a
dodecylammonium ion pair, therefore leading to an overall ligand density of 2.4–3.0 ligands nm
-2
,
which is similar to the previously calculated theoretical monolayer value of 2.9 ligands nm
-2
.
[10]

4.5. DOSY NMR Spectroscopy
Diffusion ordered NMR spectroscopy (DOSY) was performed on a suspension of the
CsPbBr3 QDs to gather binding information for oleic acid. The average diffusion coefficient for
the peaks in the alkenyl region was 327 µm
2
s
-1
for oleic acid, which is considerably smaller than
83

the value for free oleic acid in toluene-d8 (610 µm
2
s
-1
). The decrease in the diffusion coefficient
confirms interaction with the QD surface; however, this diffusion coefficient, as determined by the
Stokes-Einstein equation, is higher than what would be expected for a 12.4 nm cuboidal QD. This
means that the ligands are fluxional on the CsPbBr3 QD surface, manifesting as an average
diffusion coefficient between various bound and free states. Quantifying the experimentally
determined concentrations of the bound oleate ligands gives a bound fraction that typically ranges
between 20-30% of the total oleic acid present in the system (for QD concentrations ranging from
1.6-6.1 mM), corroborating our interpretation of the DOSY data.

4.6. Ligand Exchange with 10-Undecenoic Acid and 10-Undecenylphosphonic Acid
To obtain more quantitative information about ligand binding to the CsPbBr3 QD surface,
exchange reactions were performed with long-chain carboxylic acid and phosphonic acid ligands
that possess a terminal vinyl group. In these studies, the free and bound fractions of both the native
and incoming ligands can be quantified, allowing the surface equilibrium and associated
thermodynamic parameters to be measured. Upon titration of 10-undecenoic acid into a suspension
of CsPbBr3 QDs at room temperature, the bound and free states of oleic acid (d = 5.73 ppm bound,
5.54 ppm free) and 10-undecenoic acid (d = 5.32 and 6.12 ppm bound, 5.13 and 5.90 ppm free)
are well resolved in the solution
1
H NMR spectrum. As increasing amounts of 10-undecenoic acid
are titrated into the QD suspension, the amount of bound oleate decreases as the amount of bound
10-undecenoate increases (Figure 4.5a). In addition, the alkenyl peak that is attributed to
physisorbed oleic acid sharpens and shifts upfield toward the free oleic acid peak with increasing
concentrations of 10-undecenoic acid, as expected (Figure 4.5c). Quantification of the bound and
84

Figure 4.5. (a) Room-temperature
1
H NMR spectra of 1.6 mM CsPbBr 3 QD suspension possessing oleic
acid (OAc) and dodecylamine native ligands, titrated with increasing amounts (0–2.9 µM) of 10-undecenoic
acid (UAc) in toluene-d 8, showing both free (F) and bound (B) fractions. (b) Van‘t Hoff plot of 6.1 mM
and 3.2 mM CsPbBr 3 QD suspension with 5.1 µM and 4.7 µM 10-undecenoic acid, respectively, in toluene-
d 8 at temperatures ranging from 283 K to 325 K. (c)
1
H NMR spectrum of the alkenyl region of an oleic
acid and dodecylamine capped CsPbBr 3 QD suspension with the different peaks labeled (i.e., bound,
physisorbed, and free). (d)
1
H NMR spectrum of the alkenyl region of an oleylamine and lauric acid capped
CsPbBr 3 QD suspension with the different peaks labeled (i.e., bound/physisorbed and free).

free fractions of oleic acid and the incoming ligand over the titration series gives an average
equilibrium constant Keq of 1.97 ± 0.10, which is in close agreement with the average Keq of 1.93
± 0.08 obtained by plotting [oleic acidF][10-undecenoateB] vs [oleateB][10-undecenoic acidF]
85

Figure 4.6. Plots of [OAc] F[UDAc] B vs. [OAc] B[UDAc] F. The slope of this plot is used to determine an
average K eq for the ligand exchange between oleic acid (OAc) and 10-undecenoic acid (UDAc) based on
the equilibrium equation (
!"
=
[$%&]
!
[()%&]
"
[$%&]
"
[()%&]
!
).

(Figure 4.6). Addition of free oleic acid to the QD suspension after the titration series with 10-
undecenoic acid leads to an increase in bound oleate and a concomitant decrease in bound 10-
undecenoate (Keq ca. 0.75), establishing that the exchange is reversible, but not as favorable in the
reverse direction. We also observe that the ratio of [oleic acidF] to [10-undecenoateB] is 0.94:1.00,
which suggests a small amount of 10- undecenoic acid binds to free sites prior to the displacement
of oleic acid. The equilibrium constant for exchange with 10-undecenoic acid equates to a DG of
-1.7 ± 0.1 kJ mol
-1
, signifying the reaction is exergonic at room temperature. Variable temperature
1
H NMR spectra were acquired between 283 and 325 K on a concentrated suspension (6.1 mM
QDs) and a less concentrated suspension (3.2 mM QDs) of CsPbBr3 QDs with ca. 5 µM of added
10-undecenoic acid. A van’t Hoff plot of these data revealed an endothermic reaction (DH = 11.0
kJ mol
-1
) with a positive DS (Figure 4.5b), indicating that ligand exchange with 10-undecenoic
acid is spontaneous at elevated temperatures, including room temperature.
The average ratio of Cs to Pb for the CsPbBr3 QDs (as determined by energy dispersive X-
ray spectroscopy) before and after titration with 10-undecenoic acid (1.00:0.95 and 1.00:0.94,
86

respectively) remains unchanged. In the same way, the energy of band-edge PL emission at 2.42
eV does not change, suggesting there is no etching occurring during exchange (Table 4.1). The
relative steady-state PL intensity also does not change after exchange with 10-undecenoic acid
(Figure 4.7a), as expected given that the amount of free oleate to bound 10-undecenoate is ca. 1:1.

Table 4.1. Energy dispersive X-ray spectroscopic data for before and after ligand exchange for oleic acid
and dodecylamine capped QDs (OAc) and oleylamine and lauric acid capped QDs (OAm) titrated with 10-
undecenoic acid (UDAc), 10-undecenylphosphonic acid (UDPAc), and undec-10-en-1-amine (UDAm).
Average atomic percent is presented from three different randomly chosen locations on the sample.

Trial Cs Pb Br Cs:Pb
OAc Initial 18.21 17.33 64.46 1.00:0.95
OAc-UDAc 18.51 17.43 64.05 1.00:0.94
OAc-UDPAc 17.56 18.8 63.63 1.00:0.93
OAm Initial 15.79 19.50 64.72 1.0:1.2
OAm-UDAm 16.25 20.02 63.65 1.0:1.2

Figure 4.7. UV-vis absorption (solid line) and photoluminescence (dashed line) of (a) QD suspensions
synthesized with oleic acid and dodecylamine before and after titration with 10-undecenoic acid. (b)
CsPbBr 3 QD suspensions synthesized with oleic acid and dodecylamine before and after titration with 10-
undecenylphosphonoic acid. (c) CsPbBr 3 QD suspensions synthesized with oleylamine and lauric acid
before and after titration with undec-10-en-1-amine. All photoluminescence spectra were collected at an
excitation wavelength of 440 nm.

Exchange of the native oleic acid with 10-undecenylphosphonic acid was performed under
similar conditions. The bound and free states of 10-undecenylphosphonic acid (d = 5.21 and 5.98
87

ppm bound, 5.10 and 5.89 ppm free) are well resolved in the solution
1
H NMR spectrum (Figure
4.8). As increasing amounts of 10-undecenylphosphonic acid are titrated into the QD suspension,
the amount of bound oleate decreases as the amount of bound 10-undecenylphosphonoate
increases. Unlike exchange with carboxylic acid, however, this exchange is not in equilibrium
because the phosphonate displays only a bound peak until a critical concentration of 1 µM
phosphonic acid is reached. The relative amount of oleic acid that becomes liberated from the QD
surface upon phosphonate binding is 1.35:1, indicating that, on average, more than one proton per
incoming 10-undecenylphosphonic acid is being deprotonated as oleic acid is displaced to
maintain charge neutrality. The average Cs to Pb composition for the CsPbBr3 QDs before and
after titration with 10-undecenylphosphonic acid was 1.00:0.95 and 1.00:0.93, respectively, which
implies no change in the composition of the surface (Table 4.1). The energy of band-edge PL
emission at 2.42 eV does not change, again suggesting that no etching occurs during exchange
(Figure 4.7b). Interestingly, the relative steady-state PL intensity increases with 10-
undecenylphosphonic acid addition. The increase of PL may result from irreversible and tighter
ligand binding by the 10-undecenylphosphonic acid as compared to the dynamic carboxylic acid
binding.

Figure 4.8.
1
H NMR spectra of 1.8 mM CsPbBr 3 QD suspension possessing oleic acid (OAc) and
dodecylamine native ligands, titrated with increasing amounts (0–1.9 µM) of 10-undecenylphosphonic acid
(UPAc) in toluene-d 8, showing both free (F) and bound (B) fractions.
88

4.7. Ligand Exchange with 10-Undecenamine
We now turn our attention to amine binding to the CsPbBr3 QD surface. To investigate the
amine binding, oleylamine and lauric acid (a saturated carboxylic acid that is absent of any alkenyl
protons) were used to synthesize the QDs. This synthetic preparation yielded orthorhombic
CsPbBr3, as before; however, the morphology of the resulting QDs varies slightly from the oleic
acid and dodecylamine preparation, forming nanoplatelets rather than cuboids (Figure S2c). This
morphology has been previously reported for CsPbBr3 QDs when using a long chain amine with
a shorter carboxylic acid.
[15]
Focusing on the alkenyl region (d = 5.4-5.9 ppm) of
1
H NMR
spectrum, there are two distinct peaks for oleylamine (Figure 4.9, 4.5d). The upfield peak
corresponds to free oleylamine (d = 5.55 ppm), while there is a broad downfield peak (d = 5.68
ppm) corresponding to oleylamine interacting with the QD surface. Based on the concentration of
CsPbBr3 QDs determined by UV-vis spectroscopy, the surface density of oleylamine was then
calculated to be 1.4–1.7 oleylammonium nm
-2
by integrating the bound alkenyl resonance of

Figure 4.9.
1
H NMR spectrum of 12.9-nm CsPbBr 3 QDs (2.5 mM) synthesized with oleylamine (OAm)
and lauric acid (LAc) and dispersed in toluene-d 8 (denoted by *) with a 0.3 µm ferrocene standard. Residual
diphenyl ether (DPE) reaction solvent is present after purification. The individual
1
H NMR spectra of
oleylamine and lauric acid in toluene-d 8 are given for comparison.

89

oleylamine against an internal ferrocene standard, suggesting an overall oleylammonium laurate
surface density of 2.8–3.4 nm
-2
. Oleylamine binds differently than oleic acid as evidenced by there
only being two peaks in the alkenyl region, whereas oleic acid had three (i.e., chemisorbed,
physisorbed, and free). To further investigate this, DOSY NMR was collected, and the average
diffusion coefficient was determined to be 309 µm
2
s
-1
, which is significantly smaller than the
value for free oleylamine (894 µm
2
s
-1
), implying dynamic interaction with the surface.
To gain insight into exchange processes occurring with oleylamine, an analogous
experiment to the aforementioned ligand exchange procedure was performed with undec-10-en-1-
amine. Upon titration of undec-10-en-1-amine into the QD suspension, the bound and free states
of oleylamine (d = 5.69 ppm bound, 5.55 ppm free) and undec-10-en-1-amine (d = 5.17 and 5.79
ppm bound, 5.07 and 5.85 ppm free) are all resolved in the
1
H NMR spectrum. With increasing
amounts of undec-10-en-1-amine, the broad alkenyl peak for oleylamine sharpens and gradually
shifts upfield into the free oleylamine peak (Figure 4.10). This behavior mimics that of the
physisorbed oleic acid, leading to the conclusion that the broad peak represents a relatively weakly
bound oleylammonium without an accompanying strongly bound fraction. To investigate this
further, additional lauric acid was titrated into the QD suspension to see if the excess carboxylic
acid would shift the acid/base equilibrium and cause oleylamine to associate more strongly with
the surface in an ion pair NC(X)2 motif. These experiments indeed showed a downfield shift and
broadening of the bound oleylamine peak, meaning that the ratio of physisorbed-to-free oleylamine
increases with added carboxylic acid.
90

Figure 4.10. (a) Room-temperature
1
H NMR spectra of 2.5 mM CsPbBr 3 QD suspension possessing
oleylamine (OAm) and lauric acid native ligands, titrated with increasing amounts (0–8.0 µM) of undec-
10-en-1-amine (UAm) in toluene-d 8, showing both free (F) and bound (B) fractions. (b) Van‘t Hoff plot of
4.1 mM and 2.0 mM CsPbBr 3 QD suspension with 7.7 µM and 8.0 µM undec-10-en-1-amine, respectively,
in toluene-d 8 at temperatures ranging from 283 K to 325 K.

An average equilibrium constant Keq of 2.52 ± 0.06 was calculated for amine exchange.
Addition of free oleylamine to the QD suspension after the titration series with undec-10-en-1-
amine leads to an increase in bound oleylammonium and a concomitant decrease in bound undec-
10-en-1-ammonium (Keq ca. 0.5), suggesting that the exchange is reversible, but not as favorable
in the reverse direction. The equilibrium constant for amine exchange gives a DG of -2.1 ± 0.1 kJ
mol
-1
at room temperature, signifying the reaction is favorable, as with the carboxylic acid
exchange reaction. Variable temperature
1
H NMR spectra were acquired between 283 and 325 K
on a concentrated (4.1 mM QDs) and less concentrated (2.0 mM QDs) suspension of CsPbBr3 QDs
with ~8 µM added undec-10-en-1-amine. A van’t Hoff plot of these data yielded an average DH =
-15.9 kJ mol
-1
and a negative DS, which implies that this reaction will become less spontaneous
with increasing temperature. While the enthalpy and entropy terms are opposite in sign to the
exchange reaction with carboxylic acid, both ligand exchange processes are overall favorable at
room temperature.
91

The average ratio of Cs to Pb for the CsPbBr3 QDs synthesized with oleylamine before and
after titration with undec-10-en-1-amine remained 1.2:1.0 (Table 4.1), and the energy of the band-
edge steady-state PL emission at 2.41 eV does not change, suggesting that no etching occurs during
ligand exchange. However, there is an increase in the PL intensity after titration with undec-10-
en-1-amine (Figure 4.7c), which may be a result of tighter carboxylate binding at higher amine
concentrations due to a shift in the acid-base equilibrium, which has been previously observed.
[10]

To further investigate, CsPbBr3 synthesized with oleic acid was titrated with dodecylamine to see
if the excess amine would have any effect on the alkenyl region of oleic acid, and indeed, there is
an increase in the bound fraction of oleate ligands correlating with an increase in PL intensity upon
dodecylamine addition (Figure 4.11).

Figure 4.11. UV-vis absorption (solid line) and photoluminescence spectra (dashed) of CsPbBr 3 QDs
synthesized with oleic acid and dodecylamine before and after titration with oleylamine.

92

4.8. Conclusions
In summary, we studied the surface chemistry of CsPbBr3 QDs capped with oleic acid and
oleylamine native ligands. Ligand exchange reactions for carboxylic acids, amines, and
phosphonic acids were performed on CsPbBr3 QDs. Using
1
H NMR spectroscopy, we determined
thermodynamic parameters for carboxylic acid and amine exchange with the native ligands, and it
was determined that both reactions are in dynamic equilibrium. In contrast, phosphonic acid
irreversibly exchanges the native oleate ligands, resulting in tightly bound phosphonate. The
quantitative investigation of ligand exchange parameters via NMR methods gives needed insight
into the surface chemistry of CsPbBr3 QDs, allowing for better understanding of the effects of
carboxylic acids, amines, and phosphonic acids on the optoelectronic properties of these QDs.
Indeed, consistent with prior empirical observations,
[16,17]
we have demonstrated that more
strongly bound conjugate base ligands correlate with increases in steady-state PL intensity. This
may be a result of the more strongly bound ligands in this ionic system (with a higher on:off ratio)
better passivating surface defects and reducing surface-trap-assisted quenching.
[18]
It is expected
that this method can be utilized to study ligand binding for other lead halide perovskite QDs, with
varying A-site cations and X-site anions, in future studies.

4.9. References

[1] R. L. Brutchey, Z. Hens, M. V. Kovalenko, Blackwell Science, Oxford, 2015, pp. 233–
271.
[2] F. Zhu, L. Men, Y. Guo, Q. Zhu, U. Bhattacharjee, P. M. Goodwin, J. W. Petrich, E. A.
Smith, J. Vela, ACS Nano 2015, 9, 2948–2959.
93

[3] A. Swarnkar, A. R. Marshall, E. M. Sanehira, B. D. Chernomordik, D. T. Moore, J. A.
Christians, T. Chakrabarti, J. M. Luther, Science 2016, 354, 92–95.
[4] L. Protesescu, S. Yakunin, M. I. Bodnarchuk, F. Bertolotti, N. Masciocchi, A. Guagliardi,
M. V. Kovalenko, J. Am. Chem. Soc. 2016, 138, 14202–14205.
[5] A. Swarnkar, R. Chulliyil, V. K. Ravi, M. Irfanullah, A. Chowdhury, A. Nag, Angew.
Chem. Int. Ed. 2015, 54, 15424–15428.
[6] T. Xuan, X. Yang, S. Lou, J. Huang, Y. Liu, J. Yu, H. Li, K.-L. Wong, C. Wang, J.
Wang, Nanoscale 2017, 9, 15286–15290.
[7] L. Zhang, X. Yang, Q. Jiang, P. Wang, Z. Yin, X. Zhang, H. Tan, Y. (Michael) Yang, M.
Wei, B. R. Sutherland, et al., Nat. Commun. 2017, 8, 15640.
[8] E. M. Sanehira, A. R. Marshall, J. A. Christians, S. P. Harvey, P. N. Ciesielski, L. M.
Wheeler, P. Schulz, L. Y. Lin, M. C. Beard, J. M. Luther, Sci. Adv. 2017, 3, eaao4204.
[9] F. Krieg, S. T. Ochsenbein, S. Yakunin, S. ten Brinck, P. Aellen, A. Süess, B. Clerc, D.
Guggisberg, O. Nazarenko, Y. Shynkarenko, et al., ACS Energy Lett. 2018, 3, 641–646.
[10] J. De Roo, M. Ibáñez, P. Geiregat, G. Nedelcu, W. Walravens, J. Maes, J. C. Martins, I.
Van Driessche, M. V. Kovalenko, Z. Hens, ACS Nano 2016, 10, 2071–2081.
[11] L. Protesescu, S. Yakunin, M. I. Bodnarchuk, F. Krieg, R. Caputo, C. H. Hendon, R. X.
Yang, A. Walsh, M. V. Kovalenko, Nano Lett. 2015, 15, 3692–3696.
[12] R. R. Knauf, J. C. Lennox, J. L. Dempsey, Chem. Mater. 2016, 28, 4762–4770.
[13] P. Cottingham, R. L. Brutchey, Chem. Commun. 2016, 52, 5246–5249.
[14] B. Fritzinger, R. K. Capek, K. Lambert, J. C. Martins, Z. Hens, J. Am. Chem. Soc. 2010,
132, 10195–10201.
94

[15] A. Pan, B. He, X. Fan, Z. Liu, J. J. Urban, A. P. Alivisatos, L. He, Y. Liu, ACS Nano
2016, 10, 7943–7954.
[16] H. Huang, B. Chen, Z. Wang, T. F. Hung, A. S. Susha, H. Zhong, A. L. Rogach, Chem.
Sci 2016, 7, 5699–5703.
[17] X. Yuan, X. Hou, J. Li, C. Qu, W. Zhang, J. Zhao, H. Li, Phys. Chem. Chem. Phys. 2017,
19, 8934–8940.
[18] D. Yang, X. Li, H. Zeng, Adv. Mater. Interfaces 2018, 5, 1701662.

95

Chapter 5. Surface Termination of CsPbBr3 Perovskite Quantum Dots Determined
by Solid-State NMR Spectroscopy

*Published in J. Am. Chem. Soc. 2020, 142, 6117-6127.

5.1. Abstract
Cesium lead halide perovskite quantum dots (QDs) have gained significant attention as
next-generation optoelectronic materials; however, their properties are highly dependent on their
surface chemistry. The surfaces of cuboidal CsPbBr3 QDs have been intensively studied by both
theoretical and experimental techniques, but fundamental questions still remain about the atomic
termination of the QDs. The binding sites and modes of ligands at the surface remain unproven.
Herein, we demonstrate that solid-state NMR spectroscopy allows the unambiguous assignment
of organic surface ligands via
1
H,
13
C, and
31
P NMR. Surface-selective
133
Cs solid-state NMR
spectra show the presence of an additional
133
Cs NMR signal with a unique chemical shift that is
attributed to Cs atoms terminating the surface of the particle and that are likely coordinated by
carboxylate ligands. Dipolar dephasing curves that report on the distance between the surface
ammonium ligands and Cs and Pb were recorded using double resonance
1
H{
133
Cs} RESPDOR
and
1
H{
207
Pb} S-REDOR experiments. Model QD surface slabs with different possible surface
terminations were generated from the CsPbBr3 crystal structure, and theoretical dipolar dephasing
curves considering all possible
1
H–
133
Cs/
207
Pb spin pairs were then calculated. Comparison of the
calculated and experimental dephasing curves indicates the particles are CsBr terminated (not
PbBr2 terminated) with alkylammonium ligands substituting into some surface Cs sites, consistent
with the surface-selective
133
Cs NMR experiments. These results highlight the utility of high-
resolution solid-state NMR spectroscopy for studying ligand binding and the surface structure of
96

nanomaterials.

5.2. Introduction
Colloidal all-inorganic cesium lead halide perovskite (CsPbX3, X = Cl, Br, or I) quantum
dots (QDs) are intensively studied as a result of their excellent optoelectronic properties, including
tunable band gaps, high photoluminescence quantum yields (PLQYs), narrow emission line
widths, suppressed blinking, and high defect tolerance.
[1–3]
These properties have led to a large
variety of promising applications, such as light-emitting diodes (LEDs),
[4,5]
displays,
[6]
solar
cells,
[7]
and lasers.
[8]
To date, cesium lead halide QD-based solar cells have demonstrated a record
certified power conversion efficiency of 16.6%.
[9]

The optoelectronic properties of QDs are highly dependent on their surface coordination
chemistry.
[10–14]
For instance, surface ligands play a critical role in passivating midgap electronic
trap states and enabling high PLQY.
[10,11,13,15]
Judicious selection of surface ligands can improve
stability
[16]
and improve the quality of lead halide perovskite thin films formed from deposition of
QD precursors.
[17]
Surface reconstruction, as well as ligand absorption–desorption, happen
frequently at the surface of colloidal CsPbBr3 QDs due to their intrinsically ionic and highly
dynamic nature.
[18,19]
Therefore, developing an atomistic description of the surface structure of
CsPbX3 QDs is of paramount importance. Previously, the thermodynamics of ligand binding to
CsPbBr3 QDs were extensively studied by solution
1
H NMR spectroscopy.
[20]
It was shown that
oleic acid and dodecylamine native ligands dynamically interact with the surface with average
ligand densities of 2.4–3.0 ligands nm
-2
. Titration of 10-undecenylphosphonic acid (UDPA) into a
colloidal CsPbBr3 QDs suspension results in irreversible tight binding of phosphonate by
replacement of the native oleate. The stronger phosphonate ligand binding results in improved
97

surface passivation and higher PLQY.
[20]
Fundamental questions still remain, however, regarding
the ligand coordination chemistry at the QD surface and how the surface is terminated.
[21–23]

Bodnarchuk et al. previously used density functional theory calculations to calculate
different possible surface terminations of CsPbBr3 QDs and predict the effects of different surface
terminations on the electronic structures and resulting optoelectronic properties.
[22]
They proposed
that freshly prepared, highly luminescent cuboidal CsPbBr3 QDs have a structure described as
[CsPbBr3](PbBr2){AX′}, which corresponds to a CsPbBr3 core terminated by a PbBr2 inner shell
and an outer shell composed of monovalent cations (A = Cs
+
or ammonium ligands) and anions
(X′ = Br
–
and carboxylic acid ligands).
[22]
The PbBr2 terminated surface [CsPbBr3](CsBr){PbBr2}
was thought to be highly unlikely because of the energetically unfavorable distortion of the
Pb
2+
octahedral coordination and the high density of X-type ligands, such as oleic acid, required
to passivate those surface Pb atoms.
[22]
This termination is also predicted to promote the formation
of charge-trap sites after only 25% ligand loss.
[22]
Physical evidence regarding the termination of
these QDs is still lacking, and experimental methods are needed to prove whether as-synthesized
CsPbX3 QDs are terminated with PbBr2 layers or AX layers (where A is Cs or alkylammonium).
The positions of the organic ligands on the QD surfaces are also unknown.
Normally, electron microscopy, UV/vis spectroscopy, X-ray diffraction, or total scattering
methods are used to characterize CsPbX3 QDs. However, these techniques provide limited
structural information about the surface of the QDs because the surfaces are inherently disordered
and highly dynamic. Because all elements of cesium lead halide perovskites have accessible NMR
active nuclei, solid-state NMR is potentially an ideal technique for this task. Chemical shifts are
very sensitive to the local chemical environment of a given nucleus, and the measurements of
scalar and dipolar couplings between nuclear spins provide valuable connectivity and/or proximity
98

Figure 5.1. Idealized models of the surface termination of the as-synthesized orthorhombic CsPbBr 3 QDs.
Cs and Br atoms are depicted by blue and yellow spheres, respectively. Pb atoms reside at the center of the
red octahedra formed by Br. The QDs consist of an inorganic core with CsBr (left) or PbBr 2 (right) surface
termination. Substitution of a surface Cs atom with an ammonium ligand gives an ABr terminated surface
(A = Cs or DDA, middle left). Both surfaces are capped by cationic and anionic organic ligands at the
outermost layer. The anionic X-type oleate and 10-undecenylphosphonate ligands are assumed to bind to
exposed Cs or Pb surface atoms.

information.
[24–29]
Indeed, solid-state NMR spectroscopy has proven to be a valuable tool to study
cation dynamics
[30–35]
and the structure of bulk lead halide perovskites.
[31,36–42]
To the best of our
knowledge, solid-state NMR spectroscopy has only been applied to study lead halide perovskite
QDs in a handful of cases. Piveteau et al.
[43]
recorded DNP enhanced spectra of
133
Cs from
perovskite CsPbBr3 QDs. In addition, Brown et al.
[44]
have recently applied 2D solid-state
1
H–
31
P
solid-state NMR experiments to probe the binding of octylphosphonate ligands to CsPbBr3 QDs.
However, these studies have not provided direct information on the surface termination of
CsPbBr3 QDs. Here, solid-state NMR spectroscopy is applied to characterize the surface of
precipitated cuboidal CsPbBr3 QDs passivated with dodecylammonium, oleate, and 10-
undecenylphosphonate ligands. Notably, surface-selective
133
Cs and
207
Pb solid-state NMR
experiments suggest that the surface of the cuboidal CsPbBr3 particles are terminated by a CsBr
shell (Figure 5.1).
1
H{
133
Cs} RESPDOR and
1
H{
207
Pb} S-REDOR measurements confirm this
99

proposition and further suggest that alkylammonium ligands substitute into Cs sites at the surface
of the particles.

5.3. Basic Characterization of CsPbBr3 Quantum Dots with and without 10-
Undecylphosphonic Acid
The CsPbBr3 QDs used in this study were synthesized using a modified version of the hot-
injection synthesis initially reported by Protesescu et al.
[1]
Here, dodecylamine was substituted
for oleylamine to avoid alkenyl proton overlap with oleic acid, and diphenyl ether is substituted
for 1-octadecene to avoid vinylic proton overlap with 10-undecenylphosphonic acid in the
1
H
NMR spectra (vide infra). The resulting QDs possess a native ligand shell comprised of
dodecylammonium and oleate that dynamically bind to the QD surface.
[19,20]
The solution
1
H
NMR spectrum of the as-synthesized QDs clearly shows the diagnostic alkenyl protons of oleic
acid in the chemical shift window of δ = 5.4–5.8 ppm, with the resonance for bound oleate being
broadened and shifted downfield relative to that of free oleic acid (Figure 5.2). Based on the
concentration of CsPbBr3 QDs determined by UV–vis spectroscopy and integration of the bound
alkenyl resonance of oleate against an internal standard in the solution
1
H NMR spectrum, the
bound oleate ligand density was calculated to be 0.62 oleate nm
-2
.
[19]
However, from the
solution
1
H NMR spectra, it is difficult to estimate the relative concentration dodecylammonium
ligands due to overlap of residual solvent and ligand signals. Integration of the ammonium and
oleic acid vinyl signals in the 50 kHz MAS
1
H spin echo spectrum indicates a dodecylammonium
to oleate ratio of roughly 1.7:1 (see Figure 5.3 and discussion of
1
H solid-state NMR spectra
below). Using this ratio and the oleic acid ligand density from solution
1
H NMR, we obtain an
100

Figure 5.2. Room-temperature solution
1
H NMR spectra of CsPbBr 3 QDs with 10-undecylphosphonic acid
(green) and without 10-undecylphosphonic acid (yellow). The QDs were dispersed in toluene-d 8. The broad
resonances centered around d = 5.65 ppm and the sharp resonances centered at d = 5.54 ppm, correspond
to bound and physisorbed oleate/oleic acid, respectively. The spectrum of CsPbBr 3 QDs with
undecylphosphonic acid (green) also shows two bound undecylphosphonate resonances (d = 5.2 and 6.0
ppm) and indicates the absence of free undecylphosphonic acid. Quantifying the concentration of bound
oleate ligands gives a bound fraction that amounts to 98% of the total oleic acid present in the system before
addition of undecylphosphonic acid (yellow), which decreases to 87% upon addition of undecylphosphonic
acid (green). The
1
H NMR spectra of free undecylphosphonic acid (UDPA) and oleic acid in toluene-d 8 are
also shown (black).

ligands nm
-2
. This ligand density is less than the theoretical monolayer coverage of 5.8 ligands nm
-
2
previously estimated for CsPbBr3 QDs.
[19]
However, this prior theoretical ligand density was
calculated assuming that the surface is solely passivated by Cs and Br anions. Organic ligands such
as oleic acid and alkyl ammoniums have much larger footprints on the order of 0.3–0.5
nm
2
ligand
-1
;
[20,45]
hence, the maximum total ligand density should be on the order of 2–3 ligands
nm
-2
, in reasonable agreement with the measured values. A ligand density of 1.7 ligands nm
-2
is
fully consistent with the surface models proposed below.
101

Figure 5.3. (A) MAS
1
H spin echo solid-state NMR spectra of CsPbBr 3 QDs with and without UDPA. The
insets show the diagnostic high-frequency chemical shifts of the vinyl functional groups of UDPA, alkenyl
protons of oleate (OA), and the ammonium group of dodecylammonium (DDA).
1
H detected 2D
dipolar
1
H–
13
C CP-HETCOR spectra of CsPbBr 3 QDs (B) without and (C) with UDPA. (D) 2D dipolar
1
H
→
31
P CP HETCOR of CsPbBr 3 QDs with UDPA. The CP contact time is indicated. All spectra were
obtained with a 25 kHz MAS frequency.

A fraction of the as-synthesized CsPbBr3 QDs was treated with 10-undecenylphosphonic
acid during QD purification to perform a partial exchange of the native oleate ligands on the
surface for the phosphonate ligand.
[20]
After treatment with 10-undecenylphosphonic acid, the
solution
1
H NMR spectrum of the CsPbBr3 QDs shows clear resonances for the distinct vinylic
protons of bound phosphonate at δ = 5.2 and 6.0 ppm, with no evidence of free phosphonic acid
in solution (Figure 5.2), confirming that the phosphonic acid binds irreversibly to the
102

surface.
[20]
The relative amount of oleic acid that becomes liberated from the QD surface upon
phosphonate binding is 0.7:1 (mol/mol). The
1
H solid-state NMR spectrum (Figure 5.3a) indicates
the dodecylammoium to acid ligand ratio drops from 1.7:1 to 1.3:1 after 10-undecenylphosphonic
acid exchange, suggesting that on average, more phosphonates bind than oleic acid is exchanged
off.
Powder X-ray diffraction (PXRD) confirms that CsPbBr3 QDs crystallize in the
expected Pnma orthorhombic structure (Figure 5.4).
[46]
Transmission electron microscopy
(TEM) shows that the QDs possess a cuboidal morphology with average edge lengths of 9.6 ± 1.3
and 8.9 ± 1.4 nm for QDs that have and have not undergone ligand exchange with 10-
undecenylphosphonic acid, respectively (Figure 5.5). This is in qualitative agreement with QD

Figure 5.4. Powder XRD patterns of drop-cast films of CsPbBr 3 QDs with 10-undecylphosphonic acid
(green) and without 10-undecylphosphonic acid (yellow), with the stick pattern for orthorhombic CsPbBr 3
provided in black below.

103

Figure 5.5. TEM micrographs of (a) CsPbBr 3 QDs without 10-undecylphosphonic acid and (b) with 10-
undecylphosphonic acid. The average particle size and distribution is indicated on the image. Particle size
distributions were extracted from the images with ImageJ.

Figure 5.6. UV-vis absorption (solid line) and photoluminescence (dashed line) spectra of CsPbBr 3 QD
suspensions with (green) and without (yellow) 10-undecylphosphonic acid (UDPA) ligand exchange. The
absorption spectrum of CsPbBr 3 QDs with UPDA was normalized to the optical density of the spectrum of
CsPbBr 3 QDs without UDPA at 475 nm, and this normalization ratio was used to qualitatively compare the
photoluminescence intensities. Both photoluminescence spectra were collected at an excitation wavelength
of 400 nm.

104

sizes calculated using the photoluminescence emission spectra for CsPbBr3 with phosphonate
(10.5 nm, λmax = 512 nm) and without phosphonate (9.3 nm, λmax = 508 nm, Figure 5.6).
[47]
The
QD solutions were dried overnight under vacuum at room temperature, and the precipitated QDs
were used for solid-state NMR experiments.

5.4. Solid-State NMR Characterization of CsPbBr3 Quantum Dots
5.4.1.
1
H{
14
N} RESPDOR Confirm the Presence of Ligands Bound on the Surface
We began our investigation of the surface structures of CsPbBr3 QDs with
1
H solid-state
NMR experiments because these experiments provide the highest sensitivity and can differentiate
diagnostic functional groups of the surface ligands. Prior to ligand exchange with 10-
undecenylphosphonic acid, the main observable NMR signal covers a chemical shift range of ca.
0–2.5 ppm and corresponds to protons within the long aliphatic alkyl chain of the native
dodecylammonium and oleate ligands (Figure 5.3a). The signal that resonates at ca. 5.3 ppm is
attributed to the alkenyl protons from oleate. The broad signal centered at ca. 7.0 ppm is assigned
to ammonium protons from dodecylammonium. This assignment was confirmed with
1
H{
14
N}
RESPDOR
[48]
experiments that showed
14
N dipolar dephasing for these
1
H signals (Figure 5.7).
Note that the ammonium
1
H NMR signals are hard to observe in our
1
H solution NMR
experiments
[20]
because of overlap with the signals from residual toluene. There are two
additional
1
H NMR signals at ca. 4.9 and 5.7 ppm (Figure 5.3a) observed in the spectrum of the
CsPbBr3 QDs that were ligand exchanged with 10-undecenylphosphonic acid. These signals are
assigned to distinct vinylic protons from the 10-undecenylphosphonic acid ligands. These
chemical shifts are consistent with those observed in the solution
1
H NMR spectrum of a
105

CsPbBr3 QD suspension. The
1
H spin echo solid-state NMR spectra confirm binding of
dodecylammonium, oleate, and 10-undecenylphosphonate to the CsPbBr3 QD surface.

Figure 5.7.
1
H-
14
N RESPDOR spectra and curves for CsPbBr 3 QDs. (A, B)
1
H NMR spectra corresponding
to the S and S 0 with a 1.0 ms recoupling time. (C, D) DS/S 0 fractions for CsPbBr 3 QDs (C) without and (D)
with 10-undecylphosphonic acid. The best-fit distances to the surface are marked in accordance with the
spectra. The CH 2 signals are partially saturated and inverted (see inset) because a selective saturation pulse
was applied prior to the RESPDOR pulse sequence. This was necessary in order to reduce the CH 2
1
H NMR
signal and prevent it from interfering with measurement of the dipolar dephasing of the ammonium
1
H
NMR signal.

5.4.2. Double-Quantum Single-Quantum and Cross-Polarization Experiments Confirm the
1
H NMR Signal Assignments
2D double-quantum single-quantum (DQ-SQ) dipolar
1
H–
1
H homonuclear correlation
experiments
[49–51]
show the expected correlations for the different surface ligand groups, helping
to confirm the
1
H NMR signal assignments (Figure 5.8). The assignments of
1
H NMR signals of
106

the surface ligands were also confirmed with
1
H →
13
C cross-polarization magic angle spinning
(CPMAS) and proton detected 2D
1
H{
13
C} CP-HETCOR NMR spectra
[52]
(Figure 5.3b-c and
Figure 5.9). The 1D and 2D
13
C solid-state NMR spectra show all of the resonances expected for
dodecylammonium, oleate, and/or 10-undecenylphosphonate. The −NH3
+
dodecylammonium
1
H
NMR signals are likely absent from the 2D
1
H{
13
C} CP HETCOR spectra because the ammonium
protons are distant from most carbon atoms.

Figure 5.8. 2D
1
H-
1
H dipolar double-quantum single-quantum (DQ-SQ) homonuclear correlation spectra
of CsPbBr 3 QDs (A) without and (B) with UDPA. The diagonal black dashed lines indicate the
autocorrelation line. The spectra were obtained using the BABA pulse sequence.

Figure 5.9. Direct detected
1
H→
13
C CPMAS solid-state NMR spectra of CsPbBr 3 QDs with and without
UDPA.

107

5.4.3.
31
P NMR Experiment Confirms Phosphonate Ligand Surface Binding
A
1
H →
31
P CP-HETCOR experiment was performed to identify potential binding sites for
10-undecenylphosphonate on the CsPbBr3 QD surface and understand why ligand exchange with
10-undecenylphosphonate can enhance optical properties. Recently, Matthews and coworkers
reported the
31
P solid-state NMR spectra of octylphosphonic acid-capped CsPbBr3 QDs and
observed a primary
31
P peak at 25 ppm.
[44]
On the basis of the
31
P chemical shifts,
[53,54]
and the
HETCOR experiments, this peak was assigned to singly deprotonated, monoanionic surface
phosphonate groups. Holland et al. observed that SnO2 particles capped with 2-
carboxyethanephosphonic acid gave rise to broad
31
P NMR signals with a similar chemical shift.
They ascribed the broadening of the
31
P NMR signals to the presence of bi- and tridentate dianionic
surface phosphonate groups.
[55]
However, dianionic phosphonates should resonate at ca. 15
ppm.
[44,53,54]

The 2D
1
H–
31
P HETCOR spectrum of CsPbBr3 QDs with 10-undecenylphosphonate
shows an intense, relatively narrow
31
P NMR signal at 26 ppm, and there is also a broader signal
that extends to 20 ppm (Figure 5.3d). The peak at 26 ppm is assigned to singly deprotonated,
monoanionic phosphonate because it correlates to high-frequency
1
H NMR peaks at ca. 9.5 and
11 ppm that should correspond to acid protons.
[44]
The narrowing of the
31
P peak at 26 ppm could
arise because a singly deprotonated phosphonic acid may coordinate to the surface in a
monodentate fashion, permitting rotation that could help to average inhomogeneous broadening.
The presence of two acid
1
H NMR signals suggest that there are distinct binding sites on the
surface for the monoanionic 10-undecenylphosphonate. A part of the broader
31
P resonance
correlates to
1
H NMR signals at ca. 7.5 ppm, which should correspond to the −NH3
+
protons. This
108

correlation would suggest that at least some of the surface phosphonates are ion paired with the
positively charged ammonium ligands on the QD surface, as has been proposed previously.
[44]

5.4.4.
133
Cs and
207
Pb NMR Experiments Determine the Surface Termination of CsPbBr3
Quantum Dots
133
Cs direct excitation (spin echo) and surface-selective 2D dipolar
1
H →
133
Cs CP-
HETCOR NMR experiments were performed to obtain more definitive information about possible
surface terminations of the QDs (Figure 5.10 a, b). It has previously been shown that
133
Cs
isotropic chemical shifts are sensitive probes of structure for bulk cesium lead halide perovskites
and other related inorganic cesium phases.
[36,42,56]
The
133
Cs spin echo NMR spectra show signals
from all Cs species present in the sample, including Cs in the bulk of the QD, surface Cs sites, and
Cs in inorganic impurity phases. The intense
133
Cs NMR signal at 100 ppm is attributed to Cs ions
in the bulk of CsPbBr3 because this peak has the highest intensity and a similar chemical shift was
reported in prior studies of microcrystalline CsPbBr3.
[42,43,56]
The chemical shift of the very low
intensity
133
Cs NMR signal at ca. 235 ppm is assigned to a small amount of Cs4PbBr6 not observed
by powder XRD.
[42]
The broad, low-intensity
133
Cs NMR signal centered at ca. 170 ppm is
attributed to Cs that resides on the surface of the QDs. Consistent with this hypothesis, the signal
at 170 ppm attributed to surface Cs shows higher relative intensity in a 2D
1
H →
133
Cs CP-
HETCOR spectrum (Figure 5.10 and Figure 5.11). This unique
133
Cs chemical shift could occur
because these sites may be coordinated by carboxylate or phosphonate ligands. As expected,
the
133
Cs signals assigned to surface Cs atoms show higher relative intensity in short contact
time
1
H →
133
Cs CP-HETCOR spectra (Figure 5.12). Previously, Kovalenko and coworkers
obtained a DNP-enhanced
1
H →
133
Cs CPMAS spectrum of CsPbBr3 QDs.
[43]
Consistent with
109

the
133
Cs NMR spectra shown here, they observed
133
Cs NMR signals at ca. 170 and 100 ppm,
with the signal at 170 ppm showing higher relative intensity in the CPMAS spectrum. However, a
broad and intense signal centered at 0 ppm was also observed.
[43]
Cs cations in aqueous solution
resonate at 0 ppm, hence, we speculate their signal may have arisen from Cs
+
dissolved in water
in the DNP solvent or in the mesoporous silica used to disperse the particles.

Figure 5.10.

133
Cs spin echo and 2D
1
H →
133
Cs CP-HETCOR NMR spectra of CsPbBr 3 QDs (A) without
and (B) with UDPA.
207
Pb spin echo NMR spectra and 2D
207
Pb →
1
H CP-HETCOR of CsPbBr 3 QDs (C)
without and (D) with UDPA. The CP contact times were 9 and 8 ms for
133
Cs and
207
Pb CP-HETCOR
experiments, respectively. The
207
Pb NMR signals at 500 and 850 ppm are not real and are from t 1-noise.

110

Figure 5.11.
133
Cs spin-echo solid-state NMR spectra and surface-selective 2D dipolar
1
H→
133
Cs CP-
HETCOR spectra of CsPbBr 3 QDs (A) without and (B) with 10-undecylphosphonic acid. Both spectra were
obtained with a 2 ms CP contact time, while those shown in the main text were obtained with 8 or 9 ms CP
contact times.

Figure 5.12. Comparison of
133
Cs projections of surface-selective 2D dipolar
1
H→
133
Cs CP-HETCOR
spectra of CsPbBr 3 QDs with and without 10-undecylphosphonic acid. CP contact times were 2 ms or 9 ms
and are indicated on the spectra. The
133
Cs NMR signals assigned to the surface Cs atoms show enhanced
relative intensities in the spectra obtained with short contact times, consistent with their closer proximity to
1
H spins from the surface ligands.

111

The
1
H dimension of the 2D
1
H →
133
Cs CP-HETCOR spectrum shows signals from the
ligand, including −NH3
+
(δ = 6.5 ppm), CH2 groups adjacent to the carboxylate group of the oleate
anions (α-CH2, δ = 3.0 ppm), and the CH2 groups further along the alkyl chain (δ = 1.2 ppm). The
enhanced relative intensity of the −NH3
+
and α-CH2
1
H NMR signals in the projection of the
HETCOR spectrum as compared to the
1
H spin echo spectrum is expected. These groups should
be directly adjacent to the surface of the CsPbBr3 QDs and hence more strongly dipole coupled to
surface and bulk-like, subsurface Cs atoms. If surface Cs atoms are coordinated by oleate or 10-
undecenylphosphonate then they will be >5.0 Å away from ammonium groups (this is confirmed
by
1
H{
133
Cs} RESPDOR experiments, see Figure 5.13 and discussion below). Therefore, the
surface Cs atoms should show the most intense correlations to CH2 signals of the oleate or 10-
undecenylphosphonate alkyl chains.
207
Pb solid-state NMR experiments were also performed. The direct excitation
207
Pb spin
echo spectrum shows a single broad peak centered at ca. 200 ppm. The 200 ppm chemical shift of
CsPbBr3 observed here is similar to the shift of 400 ppm reported for bulk MAPbBr3,
[37]
and the
shift of ca. 250 ppm reported for bulk CsPbBr3.
[42]
Surface-selective
207
Pb NMR spectra were
recorded by using proton detected
207
Pb →
1
H CP-HETCOR experiments.
[40,52]
Proton detection
provides a significant sensitivity gain, allowing the observation of the fraction of
207
Pb that is
proximate to the surface ligands.
[40,52]
Even with
1
H detection these experiments were challenging
and each required between 6 and 9 h of spectrometer time. Notably, the
207
Pb projection from the
surface-selective 2D
207
Pb →
1
H CP-HETCOR spectrum is similar to the
207
Pb spin echo
spectrum, suggesting that the majority of lead atoms are present in a bulk-like environment (i.e.,
within PbBr6 octahedra). Like the
1
H →
133
Cs correlation experiment, the
207
Pb →
1
H 2D NMR

112

Figure 5.13. (A–D) Structural models of the orthorhombic (010) CsPbBr 3 surface used to simulate
the
1
H{
133
Cs}/
1
H{
207
Pb} multispin RESPDOR/S-REDOR curves. Each blue sphere corresponds to a Cs
atom while a blue atom with a purple halo around it indicates a Cs atom in the subsurface layer. The
ammonium H atom is assumed to be directly above the central atom, in the position indicated by an asterisk.
The experimental ΔS/S 0 intensities were plotted as a function of total recoupling time for CsPbBr 3 QDs
without and with UDPA surface ligands. Blue points and orange points correspond to the
1
H{
133
Cs}
RESPDOR and
1
H{
207
Pb} S-REDOR experiments, respectively. Blue and orange lines are simulated
dephasing curves for the
1
H–
133
Cs and
1
H–
207
Pb spin systems, respectively. The best fit distances are
indicated on the plots and structural models.

113

spectrum shows correlations to the −NH3
+
and α-CH2 groups of the oleate and alkylammonium
surface ligands.
Given that unique surface signals were only observed for
133
Cs, we suspect that the
cuboidal CsPbBr3 QDs are primarily terminated by Cs rather than Pb. It is important to keep in
mind, however, that surface
207
Pb NMR signals may have a large chemical shift anisotropy (CSA)
due to asymmetry at the lead coordination environment if PbBr5O, PbBr4O2, PbBr3O3, etc. sites
were present because of coordination by carboxylate or phosphonate groups. If Pb surface sites
with large CSA existed, they would likely have inefficient
207
Pb →
1
H cross-polarization transfers
and would not be detected. One possible solution is to acquire 2D
1
H–
207
Pb correlation spectra
with the dipolar heteronuclear multiple quantum coherence (D-HMQC) pulse sequence. D-HMQC
experiments work well even when the indirectly detected spin has a large CSA.
[57]
We were unable
to observe any signal using a
1
H{
207
Pb} D-HMQC experiment, likely because of the small
1
H–
207
Pb dipolar couplings. To definitively probe the composition and structure of the surface, we
used
1
H{
207
Pb} S-REDOR and
1
H{
133
Cs} RESPDOR experiments
[48,58]
to measure dipolar
couplings between the surface −NH3
+
protons and proximate surface and subsurface
207
Pb
and
133
Cs spins (Figure 5.13). The heteronuclear dipolar couplings measured by RESPDOR/S-
REDOR can then be related to internuclear distances. We note that dipolar dephasing experiments
such as RESPDOR or REDOR have previously been used to identify and model the structure of
surfaces, including determination of ligand binding sites.
[59–62]
RESPDOR generally shows
substantial dipolar dephasing, even when anisotropic interactions broaden the spectrum of the
recoupled nucleus beyond detection.
[63]
In the RESPDOR/S-REDOR experiments, only the
1
H
NMR signal from the ammonium groups of dodecylammonium was monitored because the
ammonium groups should sit directly on the surface and have substantial dipole couplings to
114

proximate
133
Cs and
207
Pb spins on and below the surface. The majority of other
1
H spins of the
surface ligands will be distant from the surface and have negligible dipole couplings to
133
Cs
and
207
Pb spins. DANTE pulse trains
[64,65]
were used to selectively excite the high frequency
ammonium
1
H NMR signals and suppress other
1
H NMR signals. However, alkene
1
H NMR
signals from oleate and 10-undecenylphosphonate ligands also become increasingly intense as the
recoupling time increases and these alkene signals partially overlap with the ammonium group
signals (Figure 5.14 a-b). This signal overlap causes ΔS/S0 intensities to become smaller than
expected at longer recoupling times. Additionally, fluctuations in the MAS frequency will also
cause the latter points in the dipolar dephasing curve to exhibit reduced intensity. Hence, fitting of
the dipolar dephasing curves focuses on the initial points with total recoupling times of less than
1.5 ms.
Both
1
H{
133
Cs} RESPDOR and
1
H{
207
Pb} S-REDOR experiments show significant
dipolar dephasing for the ammonium
1
H NMR signal (Figure 5.13). The dephasing from
133
Cs is
more significant than that from
207
Pb because
133
Cs has a natural isotropic abundance of 100%,
while the natural isotopic abundance of
207
Pb is only 22.1%. However, even taking these
differences in isotopic abundance into consideration, the dipolar dephasing experiments clearly
suggest that the surface is Cs terminated, for the reasons described below. To fit the dipolar
dephasing curves, structural models were constructed by cleaving the known orthorhombic crystal
structure of CsPbBr3 parallel to the (010) plane to obtain Cs or Pb terminated slabs (Figure 5.13
a-d). In the Pb terminated slabs the surface Pb atoms would be capped by the X′ ligands (oleate
or10-undecenylphosphonate). A model featuring Cs termination, but with replacement of central
and corner Cs atoms by an ammonium, was also considered (Figure 5.13 a). The depth and width
of the slabs was restricted to ca. 10–12 Å because at these distances the
1
H–X dipolar couplings
115

Figure 5.14 (A)
1
H{
133
Cs} RESPDOR spectra of CsPbBr 3 QDs with UDPA. The S and S 0
1
H NMR spectra
are shown for 0.5, 1.0, 2.0, 3.0, and 4.0 ms total recoupling time, respectively. DANTE pulse trains
9, 10
were
used to selectively excite the high-frequency ammonium
1
H NMR signals and minimize signals from other
1
H spins. (B)
1
H{
133
Cs} RESPDOR NMR spectra for CsPbBr 3 QDs with UDPA with a 1.0, 2.0, 3.0, and 4.0
ms total recoupling time, respectively. A DANTE pulse train was used to selectively excite the ammonium
1
H NMR signals at 7 ppm. However, the other
1
H NMR signals are still visible because the DANTE pulse
train is not perfectly selective. The alkene
1
H NMR signal from oleate becomes increasingly intense and
partially contributes to
1
H NMR signal from the ammonium groups which was monitored for dipolar
dephasing. This partly causes DS intensities to appear smaller at longer recoupling times. (C) The model of
the orthorhombic CsPbBr 3 surface with Cs termination with Cs substitution used to simulate the
1
H{
133
Cs}/
1
H{
207
Pb} multi-spin RESPDOR curves. (D) Experimental
1
H{
133
Cs} and
1
H{
207
Pb} RESPDOR
dephasing curves for CsPbBr 3 QDs with 10-undecenylphosphonate (UDPA) surface ligands. Calculated
1
H{
133
Cs} RESPDOR multi-spin dipolar dephasing curves (solid lines) are shown for different distances of
the
1
H spin above and below the Cs substitution position. Each blue circle corresponds to a signal build-up
point in
1
H{
133
Cs} RESPDOR experiments while each orange circle corresponds to a point in
1
H{
207
Pb}
RESPDOR experiments. The calculated
1
H{
207
Pb} RESDOR curve for a distance of 0 Å is also shown.

become very weak and make insignificant contributions to the predicted dephasing. The
1
H NMR
signal of the positively charged ammonium group was monitored in the dipolar dephasing
116

experiments; therefore, the
1
H spin in the model slabs is assumed to lie on top of or in between
negatively charged Br atoms because such a configuration should maximize favorable electrostatic
interactions and minimize energy.
For a given orientation of each slab,
133
Cs and
207
Pb analytical dipolar dephasing curves
were calculated for each
1
H–X spin pair. The total dephasing curve was then obtained by taking
the product of all dephasing curves for a given orientation of the slab with respect to the external
magnetic field.
[61,66,67]
This procedure was then repeated for 200 discrete values of the α, β, and γ
Euler angles, representing all possible orientations of the slab and the initial rotor phase relative to
the magnetic field, to obtain the total powder-averaged dipolar dephasing curves. To account for
the 22.1% natural isotopic abundance of
207
Pb, multiple
1
H{
207
Pb} S-REDOR curves were
calculated with the numbers of
207
Pb spins varying from zero up to the total number of lead atoms
(i.e., four Pb atoms, five Pb atoms, six Pb atoms, and four Pb atoms for structural model in Figure
5.13 a-d, respectively). The total dephasing curve was then obtained by summing together all
calculated curves, with each curve weighted by the calculated statistical probability of each
isotopomer occurring (Table 5.1-5.3).
[61]

Table 5.1. Statistical probability of
207
Pb isotopomers in the Cs termination with Cs substitution model.
Number of
207
Pb
Probability weighing
factor
Number of unique
RESPDOR curves
Total probability
0 (22.1%)
0
×(77.9%)
4
1 (22.1%)
0
× (77.9%)
4
×1 ≈ 0.37
1 (22.1%)
1
×(77.9%)
3
4 (22.1%)
1
× (77.9%)
3
×4 ≈ 0.42
2 (22.1%)
2
×(77.9%)
2
6 (22.1%)
2
× (77.9%)
2
×6 ≈ 0.18
3 (22.1%)
3
×(77.9%)
1
4 (22.1%)
3
× (77.9%)
1
×4 ≈ 0.03
4 (22.1%)
4
×(77.9%)
0
1 (22.1%)
4
× (77.9%)
0
×1 ≈ 0.002
Sum - 16 1

117

Table 5.2. Statistical probability of
207
Pb isotopomers in the Cs termination without Cs substitution
model.
Number of
207
Pb
Probability weighing
factor
Number of
unique
RESPDOR
curves
Total probability
0 (22.1%)
0
×(77.9%)
5
1 (22.1%)
0
× (77.9%)
5
×1 ≈ 0.29
1 (22.1%)
1
×(77.9%)
4
5 (22.1%)
1
× (77.9%)
4
×5 ≈ 0.41
2 (22.1%)
2
×(77.9%)
3
10 (22.1%)
2
× (77.9%)
3
×10 ≈ 0.23
3 (22.1%)
3
×(77.9%)
2
10 (22.1%)
3
× (77.9%)
2
×10 ≈ 0.06
4 (22.1%)
4
×(77.9%)
1
5 (22.1%)
4
× (77.9%)
1
×5 ≈ 0.009
5 (22.1%)
5
×(77.9%)
0
1 (22.1%)
5
× (77.9%)
0
×1 ≈ 0.0005
Sum - 32 1

Table 5.3. Statistical probability of
207
Pb isotopomers in Pb termination model (ammonium on Br
-
).
Number of
207
Pb
Probability weighing
factor
Number of
unique
RESPDOR
curves
Total probability
0 (22.1%)
0
×(77.9%)
6
1 (22.1%)
0
× (77.9%)
6
×1 ≈ 0.23
1 (22.1%)
1
×(77.9%)
5
6 (22.1%)
1
× (77.9%)
5
×6 ≈ 0.38
2 (22.1%)
2
×(77.9%)
4
15 (22.1%)
2
× (77.9%)
4
×15 ≈ 0.27
3 (22.1%)
3
×(77.9%)
3
20 (22.1%)
3
× (77.9%)
3
×20 ≈ 0.10
4 (22.1%)
4
×(77.9%)
2
15 (22.1%)
4
× (77.9%)
2
×15 ≈ 0.02
5 (22.1%)
5
×(77.9%)
1
6 (22.1%)
5
× (77.9%)
1
×6 ≈ 0.002
6 (22.1%)
6
×(77.9%)
0
1 (22.1%)
6
× (77.9%)
0
×1 ≈ 0.0001
Sum - 64 1

Among the four models considered, the one that gives the best simultaneous fit of the
1
H{
133
Cs}
RESPDOR and
1
H{
207
Pb} S-REDOR curves is obtained with Cs surface termination where the
ammonium
1
H spin has been substituted at a Cs A-site position (Figure 5.13 a and Figure 5. 13
c-d). The substitution of an ammonium group within a pocket of bromide ions created by a Cs
vacancy is reasonable considering similar binding sites for alkylammonium ligands are observed
in the single crystal X-ray diffraction structures of 2D Ruddlesden–Popper lead halide
phases.
[68]
Kovalenko and coworkers have recently used wide-angle total X-ray scattering and
TEM to determine that colloidal CsPbBr3 nanoplatelets, different from the cuboidal morphology
studied here, are likely CsBr terminated with Cs vacancies.
[69]
Also, models with alkylammonium
118

ligands substituting into surface Cs sites have previously been considered as one of the most likely
surface terminations for CsPbBr3 QDs.
[23]
In comparison, the Cs-terminated surface without
substitution of ammonium into the Cs position gives an adequate fit of the
133
Cs dipolar dephasing
curve, however, the subsurface Pb sites are very distant and essentially no
207
Pb dipolar dephasing
is predicted, in contradiction with the experiment (Figure 5.13 b and Figure 5.15). Alternatively,
if the
1
H spin is moved closer to the surface to reproduce the experimental
207
Pb dipolar dephasing,
then the predicted
133
Cs dephasing is overestimated (Figure 5.13 b); only with both Cs termination
and substitution of Cs for ammonium groups can we simultaneously fit both
1
H{
133
Cs}
and
1
H{
207
Pb} dipolar dephasing curves. Finally, the Pb-terminated surface predicts faster
207
Pb
dephasing than is observed experimentally and can, therefore, be ruled out as a plausible structural
model (Figure 5.13 c-d). In summary, the RESPDOR/S-REDOR data provides the first
experimental evidence for a [CsPbBr3](PbBr2){ABr} termination, where the cation A in the outer
shell is Cs
+
or dodecylammonium. In this model, the anionic oleate or phosphonate ligands are
primarily associated with Cs
+
cations in the outer shell. The similarity of the
133
Cs and
207
Pb NMR
spectra and dipolar dephasing curves for CsPbBr3 QDs with and without 10-undecenylphosphonic
acid would suggest that there is no major reorganization of the surface caused by ligand exchange
of oleate for 10-undecenylphosphonate, which is consistent with prior results that show neither the
Cs:Pb ratio nor the band edge PL emission energy change upon ligand exchange.
[20]

119

Figure 5.15. (Left panels) The model of the orthorhombic CsPbBr 3 surface with Cs termination without Cs
substitution used to calculate the
1
H{
133
Cs}/
1
H{
207
Pb} multi-spin RESPDOR curves. Experimental
1
H{
133
Cs} and
1
H{
207
Pb} RESPDOR dephasing curves for CsPbBr 3 QDs with 10-undecenylphosphonate
(UDPA) surface ligands. (Right panels) Calculated
1
H{
133
Cs} and
1
H{
207
Pb} RESPDOR multi-spin dipolar
dephasing curves (solid lines) are shown for different distances of the
1
H spin above a surface Br atom.
Each blue circle corresponds to a signal build-up point in
1
H{
133
Cs} RESPDOR experiments while each
orange circle corresponds to a point in
1
H{
207
Pb} RESPDOR experiments.

5.5. Surface Termination of CsPbBr3 Quantum Dots Determined by Experimental and
Theoretical Surface Modelling
The [CsPbBr3](PbBr2){ABr} surface model is consistent with a number of other
experimental observations. For a CsPbBr3 QD with an average edge length of ∼10 nm, a Cs-
surface termination necessitates Br/Pb and Cs/Pb ratios of ∼3.2 and ∼1.2, respectively, in
120

agreement with previous experimentally determined values.
[21,22]
We note that Cs-surface
termination also results in a slight excess of positive charge.
[21,22]
Charge compensation should
occur through loss of surface Cs, rather than addition of excess Br, because it has been
computationally predicted that addition of excess bromide results in low PLQY because of
formation of localized trap states above the valence band edge.
[21]
Finally, solution
1
H NMR
indicated a dodecylammonium ligand density of 1.0 dodecylammonium nm
-2
. The model
in Figure 5.13 a corresponds to a density of 1.06 dodecylammonium nm
-2
. The calculated

Figure 5.16. Models of the (010) Cs and Br terminated surface of orthorhombic CsPbBr 3 with the
ammonium
1
H spin substituted into the central Cs position indicated by an asterisk.
1
H{
133
Cs} and
1
H{
207
Pb} multi-spin RESPDOR curves are shown for models with different numbers of surface cesium
atoms. The indicated DDA ligand densities were calculated by assuming each Cs vacancy is filled by a
dodecylammonium ligand. This figure illustrates that the
1
H{
133
Cs} RESPDOR curves are compatible with
dodecylammonium densities between 0.71 and 2.13 dodecylammonium nm
-2
; at higher ligand densities the
133
Cs dipolar dephasing is significantly reduced. Each blue sphere corresponds to a Cs atom while the purple
halo around it indicates a Cs atom in the second, sub-surface layer.

121

1
H{
133
Cs} RESPDOR curves for [CsPbBr3](PbBr2){ABr} models are compatible with ligand
densities between 0.7 and 2.1 dodecylammonium nm
-2
(Figure 5.16), where ligand densities
higher than 0.7 dodecylammonium nm
-2
are achieved by replacing the corner Cs atoms by
dodecylammonium ligands as in Figure 5.13a.

5.6. Conclusions
In summary, we have shown that the surface chemistry of CsPbBr3 QDs can be interrogated
by state-of-the-art solid-state NMR experiments. Specifically, we show that dodecylammonium,
oleate, and/or 10-undecenylphosphonate are all present as surface ligands. Furthermore, we
determine the internuclear distances between dodecylammonium −NH3
+
protons to surface and
subsurface Cs and Pb sites using dipolar dephasing experiments. Simulation of multi-spin dipolar
dephasing curves suggests that the CsPbBr3 QDs are Cs terminated with dodecylammonium
ligands substituting into surface Cs A-sites, thereby providing an atomistic picture of QD
termination that is [CsPbBr3](PbBr2){ABr}, where A is Cs or dodecylammonium. These findings
are in agreement with the surface-selective
207
Pb and
133
Cs HETCOR solid-state NMR
experiments as well as previous computational models of QD surfaces.
[22]
This study demonstrates
that solid-state NMR spectroscopy should be useful to better understand how aging of the QDs,
ligand exchange, and chemical treatments alter surface structure and affect optical properties.
More generally, our results also highlight the utility and future prospects of surface
characterization of nanomaterials via solid-state NMR spectroscopy.

122

5.7. References
[1] L. Protesescu, S. Yakunin, M. I. Bodnarchuk, F. Krieg, R. Caputo, C. H. Hendon, R. X.
Yang, A. Walsh, M. V. Kovalenko, Nano Lett. 2015, 15, 3692–3696.
[2] M. V. Kovalenko, L. Protesescu, M. I. Bodnarchuk, Science 2017, 358, 745–750.
[3] H. Huang, M. I. Bodnarchuk, S. V. Kershaw, M. V. Kovalenko, A. L. Rogach, ACS
Energy Lett. 2017, 2, 2071–2083.
[4] K. Lin, J. Xing, L. N. Quan, F. P. G. de Arquer, X. Gong, J. Lu, L. Xie, W. Zhao, D.
Zhang, C. Yan, W. Li, X. Liu, Y. Lu, J. Kirman, E. H. Sargent, Q. Xiong, Z. Wei, Nature
2018, 562, 245–248.
[5] Z. Wei, J. Xing, J. Phys. Chem. Lett. 2019, 10, 3035–3042.
[6] M. Lu, Y. Zhang, S. Wang, J. Guo, W. W. Yu, A. L. Rogach, Adv. Funct. Mater. 2019,
29, 1902008.
[7] Q. Tai, K.-C. Tang, F. Yan, Energy Environ. Sci. 2019, 12, 2375–2405.
[8] K. Wang, S. Wang, S. Xiao, Q. Song, Adv. Opt. Mater. 2018, 6, 1800278.
[9] M. Hao, Y. Bai, S. Zeiske, L. Ren, J. Liu, Y. Yuan, N. Zarrabi, N. Cheng, M. Ghasemi,
P. Chen, M. Lyu, D. He, J.-H. Yun, Y. Du, Y. Wang, S. Ding, A. Armin, P. Meredith, G.
Liu, H.-M. Cheng, L. Wang, Nat. Energy 2020, 5, 79–88.
[10] R. Brutchey, Z. Hens, M. V. Kovalenko, in Chemistry of Organo-Hybrids (Eds.: B.
Charleux, C. Copéret, E. Lacôte), John Wiley & Sons, Inc., Hoboken, NJ, USA, 2015,
pp. 233–271.
[11] J. Owen, Science 2015, 347, 615–616.
123

[12] J. Pan, Y. Shang, J. Yin, M. De Bastiani, W. Peng, I. Dursun, L. Sinatra, A. M. El-Zohry,
M. N. Hedhili, A.-H. Emwas, O. F. Mohammed, Z. Ning, O. M. Bakr, J. Am. Chem. Soc.
2018, 140, 562–565.
[13] J. Pan, L. N. Quan, Y. Zhao, W. Peng, B. Murali, S. P. Sarmah, M. Yuan, L. Sinatra, N.
M. Alyami, J. Liu, E. Yassitepe, Z. Yang, O. Voznyy, R. Comin, M. N. Hedhili, O. F.
Mohammed, Z. H. Lu, D. H. Kim, E. H. Sargent, O. M. Bakr, Adv. Mater. 2016, 28,
8718–8725.
[14] Q. A. Akkerman, G. Rainò, M. V. Kovalenko, L. Manna, Nat. Mater. 2018, 17, 394–405.
[15] B. Chen, P. N. Rudd, S. Yang, Y. Yuan, J. Huang, Chem. Soc. Rev. 2019, 48, 3842–3867.
[16] J. Pan, S. P. Sarmah, B. Murali, I. Dursun, W. Peng, M. R. Parida, J. Liu, L. Sinatra, N.
Alyami, C. Zhao, E. Alarousu, T. K. Ng, B. S. Ooi, O. M. Bakr, O. F. Mohammed, J.
Phys. Chem. Lett. 2015, 6, 5027–5033.
[17] L. M. Wheeler, E. M. Sanehira, A. R. Marshall, P. Schulz, M. Suri, N. C. Anderson, J. A.
Christians, D. Nordlund, D. Sokaras, T. Kroll, S. P. Harvey, J. J. Berry, L. Y. Lin, J. M.
Luther, J. Am. Chem. Soc. 2018, 140, 10504–10513.
[18] V. K. Ravi, P. K. Santra, N. Joshi, J. Chugh, S. K. Singh, H. Rensmo, P. Ghosh, A. Nag,
J. Phys. Chem. Lett. 2017, 8, 4988–4994.
[19] J. De Roo, M. Ibáñez, P. Geiregat, G. Nedelcu, W. Walravens, J. Maes, J. C. Martins, I.
Van Driessche, M. V. Kovalenko, Z. Hens, ACS Nano 2016, 10, 2071–2081.
[20] S. R. Smock, T. J. Williams, R. L. Brutchey, Angew. Chem. Int. Ed. 2018, 57, 11711–
11715.
[21] S. ten Brinck, I. Infante, ACS Energy Lett. 2016, 1, 1266–1272.
124

[22] M. I. Bodnarchuk, S. C. Boehme, S. ten Brinck, C. Bernasconi, Y. Shynkarenko, F.
Krieg, R. Widmer, B. Aeschlimann, D. Günther, M. V. Kovalenko, I. Infante, ACS
Energy Lett. 2019, 4, 63–74.
[23] G. Almeida, I. Infante, L. Manna, Science 2019, 364, 833–834.
[24] D. D. Lovingood, R. Achey, A. K. Paravastu, G. F. Strouse, J. Am. Chem. Soc. 2010,
132, 3344–3354.
[25] H. Virieux, M. Le Troedec, A. Cros-Gagneux, W.-S. Ojo, F. Delpech, C. Nayral, H.
Martinez, B. Chaudret, J. Am. Chem. Soc. 2012, 134, 19701–19708.
[26] M. P. Hanrahan, E. L. Fought, T. L. Windus, L. M. Wheeler, N. C. Anderson, N. R.
Neale, A. J. Rossini, Chem. Mater. 2017, 29, 10339–10351.
[27] S. Cadars, B. J. Smith, J. D. Epping, S. Acharya, N. Belman, Y. Golan, B. F. Chmelka,
Phys. Rev. Lett. 2009, 103, 136802.
[28] R. Wang, C. I. Ratcliffe, X. Wu, O. Voznyy, Y. Tao, K. Yu, J. Phys. Chem. C 2009, 113,
17979–17982.
[29] J. P. Yesinowski, in Solid State NMR (Ed.: J.C.C. Chan), Springer Berlin Heidelberg,
Berlin, Heidelberg, 2011, pp. 229–312.
[30] G. M. Bernard, R. E. Wasylishen, C. I. Ratcliffe, V. Terskikh, Q. Wu, J. M. Buriak, T.
Hauger, J. Phys. Chem. A 2018, 122, 1560–1573.
[31] A. Karmakar, A. M. Askar, G. M. Bernard, V. V. Terskikh, M. Ha, S. Patel, K. Shankar,
V. K. Michaelis, Chem. Mater. 2018, 30, 2309–2321.
[32] W. M. J. Franssen, S. G. D. van Es, R. Dervişoğlu, G. A. de Wijs, A. P. M. Kentgens, J.
Phys. Chem. Lett. 2017, 8, 61–66.
125

[33] O. Knop, R. E. Wasylishen, M. A. White, T. S. Cameron, M. J. M. V. Oort, Can. J.
Chem. 1990, 68, 412–422.
[34] Q. Xu, T. Eguchi, H. Nakayama, N. Nakamura, M. Kishita, Z. Naturforsch. A 1991, 46,
240–246.
[35] D. H. Fabini, T. A. Siaw, C. C. Stoumpos, G. Laurita, D. Olds, K. Page, J. G. Hu, M. G.
Kanatzidis, S. Han, R. Seshadri, J. Am. Chem. Soc. 2017, 139, 16875–16884.
[36] D. J. Kubicki, D. Prochowicz, A. Hofstetter, S. M. Zakeeruddin, M. Grätzel, L. Emsley,
J. Am. Chem. Soc. 2017, 139, 14173–14180.
[37] B. A. Rosales, L. Men, S. D. Cady, M. P. Hanrahan, A. J. Rossini, J. Vela, Chem. Mater.
2016, 28, 6848–6859.
[38] D. J. Kubicki, D. Prochowicz, A. Hofstetter, P. Péchy, S. M. Zakeeruddin, M. Grätzel, L.
Emsley, J. Am. Chem. Soc. 2017, 139, 10055–10061.
[39] B. A. Rosales, M. P. Hanrahan, B. W. Boote, A. J. Rossini, E. A. Smith, J. Vela, ACS
Energy Lett. 2017, 2, 906–914.
[40] M. P. Hanrahan, L. Men, B. A. Rosales, J. Vela, A. J. Rossini, Chem. Mater. 2018, 30,
7005–7015.
[41] A. M. Askar, G. M. Bernard, B. Wiltshire, K. Shankar, V. K. Michaelis, J. Phys. Chem. C
2017, 121, 1013–1024.
[42] A. Karmakar, M. S. Dodd, X. Zhang, M. S. Oakley, M. Klobukowski, V. K. Michaelis,
Chem. Commun. 2019, 55, 5079–5082.
[43] L. Piveteau, T.-C. Ong, A. J. Rossini, L. Emsley, C. Copéret, M. V. Kovalenko, J. Am.
Chem. Soc. 2015, 137, 13964–13971.
126

[44] A. A. M. Brown, T. J. N. Hooper, S. A. Veldhuis, X. Y. Chin, A. Bruno, P. Vashishtha, J.
N. Tey, L. Jiang, B. Damodaran, S. H. Pu, S. G. Mhaisalkar, N. Mathews, Nanoscale
2019, 11, 12370–12380.
[45] A. Antanovich, A. Prudnikau, A. Matsukovich, A. Achtstein, M. Artemyev, J. Phys.
Chem. C 2016, 120, 5764–5775.
[46] P. Cottingham, R. L. Brutchey, Chem. Commun. 2016, 52, 5246–5249.
[47] M. C. Brennan, J. E. Herr, T. S. Nguyen-Beck, J. Zinna, S. Draguta, S. Rouvimov, J.
Parkhill, M. Kuno, J. Am. Chem. Soc. 2017, 139, 12201–12208.
[48] Z. Gan, Chem. Commun. 2006, 4712–4714.
[49] M. Feike, D. E. Demco, R. Graf, J. Gottwald, S. Hafner, H. W. Spiess, J. Magn. Reson. A
1996, 122, 214–221.
[50] I. Schnell, A. Lupulescu, S. Hafner, D. E. Demco, H. W. Spiess, J. Magn. Reson. 1998,
133, 61–69.
[51] I. Schnell, Prog. Nucl. Magn. Reson. Spectrosc. 2004, 45, 145–207.
[52] J. W. Wiench, C. E. Bronnimann, V. S.-Y. Lin, M. Pruski, J. Am. Chem. Soc. 2007, 129,
12076–12077.
[53] L. C. Thomas, The Identification of Functional Groups in Organophosphorus
Compounds, Academic Press, 1974.
[54] M. M. Crutchfield, E. J. Griffith, M. Grayson, Topics in Phosphorus Chemistry,
Interscience, 1967.
[55] G. P. Holland, R. Sharma, J. O. Agola, S. Amin, V. C. Solomon, P. Singh, D. A. Buttry,
J. L. Yarger, Chem. Mater. 2007, 19, 2519–2526.
127

[56] D. J. Kubicki, D. Prochowicz, A. Pinon, G. Stevanato, A. Hofstetter, S. M. Zakeeruddin,
M. Grätzel, L. Emsley, J. Mater. Chem. A 2019, 7, 2326–2333.
[57] A. J. Rossini, M. P. Hanrahan, M. Thuo, Phys. Chem. Chem. Phys. 2016, 18, 25284–
25295.
[58] L. Chen, Q. Wang, B. Hu, O. Lafon, J. Trébosc, F. Deng, J.-P. Amoureux, Phys. Chem.
Chem. Phys. 2010, 12, 9395.
[59] F. A. Perras, J. D. Padmos, R. L. Johnson, L.-L. Wang, T. J. Schwartz, T. Kobayashi, J.
H. Horton, J. A. Dumesic, B. H. Shanks, D. D. Johnson, M. Pruski, J. Am. Chem. Soc.
2017, 139, 2702–2709.
[60] Y.-Y. Hu, A. Rawal, K. Schmidt-Rohr, Proc. Natl. Acad. Sci. U. S. A. 2010, 107, 22425–
22429.
[61] F. A. Perras, Z. Wang, T. Kobayashi, A. Baiker, J. Huang, M. Pruski, Phys. Chem. Chem.
Phys. 2019, 21, 19529–19537.
[62] P. Berruyer, M. Lelli, M. P. Conley, D. L. Silverio, C. M. Widdifield, G. Siddiqi, D.
Gajan, A. Lesage, C. Copéret, L. Emsley, J. Am. Chem. Soc. 2017, 139, 849–855.
[63] S. S. Nagarkar, H. Kurasho, N. T. Duong, Y. Nishiyama, S. Kitagawa, S. Horike, Chem.
Commun. 2019, 55, 5455–5458.
[64] G. Bodenhausen, R. Freeman, G. A. Morris, J. Magn. Reson. 1976, 23, 171–175.
[65] G. A. Morris, R. Freeman, J. Magn. Reson. 1978, 29, 433–462.
[66] J. M. Goetz, J. Schaefer, J. Magn. Reson. 1997, 127, 147–154.
[67] X. Lu, O. Lafon, J. Trébosc, J.-P. Amoureux, J. Magn. Reson. 2012, 215, 34–49.
[68] C. C. Stoumpos, D. H. Cao, D. J. Clark, J. Young, J. M. Rondinelli, J. I. Jang, J. T. Hupp,
M. G. Kanatzidis, Chem. Mater. 2016, 28, 2852–2867.
128

[69] F. Bertolotti, G. Nedelcu, A. Vivani, A. Cervellino, N. Masciocchi, A. Guagliardi, M. V.
Kovalenko, ACS Nano 2019, 13, 14294–14307.

129

Bibliography

[1] F. Manby, T. Miller, P. Bygrave, F. Ding, T. Dresselhaus, F. Batista-Romero, A. Buccheri,
C. Bungey, S. Lee, R. Meli, K. Miyamoto, C. Steinmann, T. Tsuchiya, M. Welborn, T.
Wiles, Z. Williams, Entos: A Quantum Molecular Simulation Package, 2019.
[2] S. Irvine, in Springer Handbook of Electronic and Photonic Materials (Eds.: S. Kasap, P.
Capper), Springer International Publishing, Cham, 2017, pp. 1–1.
[3] A. Kojima, K. Teshima, Y. Shirai, T. Miyasaka, J. Am. Chem. Soc. 2009, 131, 6050–6051.
[4] A. K. Jena, A. Kulkarni, T. Miyasaka, Chem. Rev. 2019, 119, 3036–3103.
[5] T. Yu, F. Wang, Y. Xu, L. Ma, X. Pi, D. Yang, Adv. Mater. 2016, 28, 4912–4919.
[6] X. Zhou, J. Qiao, Z. Xia, Chem. Mater. 2021, 33, 1083–1098.
[7] C.-H. M. Chuang, P. R. Brown, V. Bulović, M. G. Bawendi, Nat. Mater. 2014, 13, 796–
801.
[8] P. Guyot-Sionnest, M. M. Ackerman, X. Tang, J. Chem. Phys. 2019, 151, 060901.
[9] J. Owen, L. Brus, J. Am. Chem. Soc. 2017, 139, 10939–10943.
[10] P. Wangyang, C. Gong, G. Rao, K. Hu, X. Wang, C. Yan, L. Dai, C. Wu, J. Xiong, Adv.
Opt. Mater 2018, 6, 1701302.
[11] H. Lu, G. M. Carroll, N. R. Neale, M. C. Beard, ACS Nano 2019, acsnano.8b09815.
[12] M. V. Kovalenko, Nat. Nanotech 2015, 10, 994–997.
[13] J. Owen, Science 2015, 347, 615–616.
[14] M. L. H. Green, G. Parkin, J. Chem. Educ. 2014, 91, 807–816.
[15] M. L. H. Green, J. Organomet. Chem. 1995, 500, 127–148.
[16] J. Owen, Science 2015, 347, 615–616.
130

[17] M. Woodhouse, R. Jones-Albertus, D. Feldman, R. Fu, K. Horowitz, D. Chung, D. Jordan,
S. Kurtz, On the Path to SunShot: The Role of Advancements in Solar Photovoltaic
Efficiency, Reliability, and Cost, National Renewable Energy Laboratory, 2016.
[18] R. K. Baldwin, K. A. Pettigrew, E. Ratai, M. P. Augustine, S. M. Kauzlarich, Chem.
Commun. 2002, 1822–1823.
[19] P. M. Fauchet, Mater. Today 2005, 8, 26–33.
[20] K. A. Pettigrew, Q. Liu, P. P. Power, S. M. Kauzlarich, Chem. Mater. 2003, 15, 4005–
4011.
[21] J. M. Buriak, Phil. Trans. R. Soc. A. 2006, 364, 217–225.
[22] Z. F. Li, E. Ruckenstein, Nano Lett. 2004, 4, 1463–1467.
[23] L. Wang, V. Reipa, J. Blasic, Bioconjugate Chem. 2004, 15, 409–412.
[24] E. Muthuswamy, J. Zhao, K. Tabatabaei, M. M. Amador, M. A. Holmes, F. E. Osterloh, S.
M. Kauzlarich, Chem. Mater. 2014, 26, 2138–2146.
[25] E. Fok, M. Shih, A. Meldrum, J. G. C. Veinot, Chem. Commun. 2004, 386.
[26] K. Dohnalová, T. Gregorkiewicz, K. Kůsová, J. Phys.: Condens. Matter 2014, 26, 173201.
[27] F. Li, J. Wang, S. Sun, H. Wang, Z. Tang, G. Nie, Small 2015, 11, 1954–1961.
[28] J.-W. Luo, A. Franceschetti, A. Zunger, Nano Lett. 2008, 8, 3174–3181.
[29] U. Kortshagen, R. Anthony, R. Gresback, Z. Holman, R. Ligman, C.-Y. Liu, L. Mangolini,
S. A. Campbell, Pure Appl. Chem. 2008, 80, 1901–1908.
[30] S. Bhattacharjee, I. M. C. M. Rietjens, M. P. Singh, T. M. Atkins, T. K. Purkait, Z. Xu, S.
Regli, A. Shukaliak, R. J. Clark, B. S. Mitchell, G. M. Alink, A. T. M. Marcelis, M. J. Fink,
J. G. C. Veinot, S. M. Kauzlarich, H. Zuilhof, Nanoscale 2013, 5, 4870.
131

[31] C.-S. Yang, R. A. Bley, S. M. Kauzlarich, H. W. H. Lee, G. R. Delgado, J. Am. Chem. Soc.
1999, 121, 5191–5195.
[32] T. M. Atkins, M. C. Cassidy, M. Lee, S. Ganguly, C. M. Marcus, S. M. Kauzlarich, ACS
Nano 2013, 7, 1609–1617.
[33] H. Gerung, S. D. Bunge, T. J. Boyle, C. J. Brinker, S. M. Han, Chem. Commun. 2005,
1914–1916.
[34] T. N. Lambert, N. L. Andrews, H. Gerung, T. J. Boyle, J. M. Oliver, B. S. Wilson, S. M.
Han, Small 2007, 3, 691–699.
[35] D. D. Vaughn, J. F. Bondi, R. E. Schaak, Chem. Mater. 2010, 22, 6103–6108.
[36] S. R. Smock, K. Tabatabaei, T. J. Williams, S. M. Kauzlarich, R. L. Brutchey, Nanoscale
2020, 12, 2764–2772.
[37] L. M. Wheeler, A. W. Nichols, B. D. Chernomordik, N. C. Anderson, M. C. Beard, N. R.
Neale, Nano Lett. 2016, 16, 1949–1954.
[38] I. Robel, V. Subramanian, M. Kuno, P. V. Kamat, J. Am. Chem. Soc. 2006, 128, 2385–
2393.
[39] K. S. Leschkies, R. Divakar, J. Basu, E. Enache-Pommer, J. E. Boercker, C. B. Carter, U.
R. Kortshagen, D. J. Norris, E. S. Aydil, Nano Lett. 2007, 7, 1793–1798.
[40] S. Sadeghi, S. Khabbaz Abkenar, C. W. Ow-Yang, S. Nizamoglu, Sci Rep 2019, 9, 10061.
[41] M. Nirmal, L. Brus, Acc. Chem. Res. 1999, 32, 407–414.
[42] X. Peng, M. C. Schlamp, A. V. Kadavanich, A. P. Alivisatos, J. Am. Chem. Soc. 1997, 119,
7019–7029.
[43] D. V. Talapin, A. L. Rogach, A. Kornowski, M. Haase, H. Weller, Nano Lett. 2001, 1,
207–211.
132

[44] Y. C. Cao, J. Wang, J. Am. Chem. Soc. 2004, 126, 14336–14337.
[45] R. Gomes, A. Hassinen, A. Szczygiel, Q. Zhao, A. Vantomme, J. C. Martins, Z. Hens, J.
Phys. Chem. Lett. 2011, 2, 145–152.
[46] J. M. Luther, J. M. Pietryga, ACS Nano 2013, 7, 1845–1849.
[47] J. Jasieniak, P. Mulvaney, J. Am. Chem. Soc. 2007, 129, 2841–2848.
[48] R. R. Knauf, J. C. Lennox, J. L. Dempsey, Chem. Mater. 2016, 28, 4762–4770.
[49] M. Liu, Y.-Y. Wang, Y. Liu, F.-L. Jiang, J. Phys. Chem. C 2020, 124, 4613–4625.
[50] M. K. Samantaray, V. Katiyar, D. Roy, K. Pang, H. Nanavati, R. Stephen, R. B. Sunoj, P.
Ghosh, Eur. J. Inorg. Chem. 2006, 2006, 2975–2984.
[51] M. T. Frederick, E. A. Weiss, ACS Nano 2010, 4, 3195–3200.
[52] M. S. Azzaro, A. Dodin, D. Y. Zhang, A. P. Willard, S. T. Roberts, Nano Lett. 2018, 18,
3259–3270.
[53] D. E. Westmoreland, R. López-Arteaga, E. A. Weiss, J. Am. Chem. Soc. 2020, 142, 2690–
2696.
[54] S. Lian, J. A. Christensen, M. S. Kodaimati, C. R. Rogers, M. R. Wasielewski, E. A. Weiss,
J. Phys. Chem. C 2019, 123, 5923–5930.
[55] M. V. Kovalenko, L. Manna, A. Cabot, Z. Hens, D. V. Talapin, C. R. Kagan, V. I. Klimov,
A. L. Rogach, P. Reiss, D. J. Milliron, P. Guyot-Sionnnest, G. Konstantatos, W. J. Parak,
T. Hyeon, B. A. Korgel, C. B. Murray, W. Heiss, ACS Nano 2015, 9, 1012–1057.
[56] M. J. Greaney, E. Couderc, J. Zhao, B. A. Nail, M. Mecklenburg, W. Thornbury, F. E.
Osterloh, S. E. Bradforth, R. L. Brutchey, Chem. Mater. 2015, 27, 744–756.
[57] J. J. Buckley, M. J. Greaney, R. L. Brutchey, Chem. Mater. 2014, 26, 6311–6317.
[58] S. R. Smock, Y. Chen, A. J. Rossini, R. L. Brutchey, Acc. Chem. Res. 2021, 54, 707–718.
133

[59] G. Rose, J. Prakt. Chem. 1840, 19, 459–468.
[60] A. S. Bhalla, R. Guo, R. Roy, Mater. Res. Innov. 2000, 4, 3–26.
[61] J. Breternitz, S. Schorr, Adv. Energy Mater. 2018, 8, 1802366.
[62] S. Niu, H. Huyan, Y. Liu, M. Yeung, K. Ye, L. Blankemeier, T. Orvis, D. Sarkar, D. J.
Singh, R. Kapadia, J. Ravichandran, Adv. Mater. 2017, 29, 1604733.
[63] K. R. Talley, C. L. Perkins, D. R. Diercks, G. L. Brennecka, A. Zakutayev,
arXiv:2001.00633 [cond-mat] 2020.
[64] C. C. Stoumpos, C. D. Malliakas, J. A. Peters, Z. Liu, M. Sebastian, J. Im, T. C. Chasapis,
A. C. Wibowo, D. Y. Chung, A. J. Freeman, B. W. Wessels, M. G. Kanatzidis, Cryst.
Growth Des. 2013, 13, 2722–2727.
[65] H. L. Wells, Z. Anorg. Chem. 1893, 3, 195–210.
[66] “Best Research-Cell Efficiencies,” can be found under
https://www.nrel.gov/pv/assets/pdfs/best-research-cell-efficiencies.20200708.pdf, n.d.
[67] N. K. Noel, A. Abate, S. D. Stranks, E. S. Parrott, V. M. Burlakov, A. Goriely, H. J. Snaith,
ACS Nano 2014, 8, 9815–9821.
[68] D. W. deQuilettes, S. Koch, S. Burke, R. K. Paranji, A. J. Shropshire, M. E. Ziffer, D. S.
Ginger, ACS Energy Lett. 2016, 1, 438–444.
[69] Z. Hawash, S. R. Raga, D.-Y. Son, L. K. Ono, N.-G. Park, Y. Qi, J. Phys. Chem. Lett. 2017,
8, 3947–3953.
[70] E. A. Alharbi, A. Y. Alyamani, D. J. Kubicki, A. R. Uhl, B. J. Walder, A. Q. Alanazi, J.
Luo, A. Burgos-Caminal, A. Albadri, H. Albrithen, M. H. Alotaibi, J.-E. Moser, S. M.
Zakeeruddin, F. Giordano, L. Emsley, M. Grätzel, Nat. Commun 2019, 10, 3008.
[71] I. K. Hewavitharana, S. L. Brock, ACS Nano 2017, 11, 11217–11224.
134

[72] H. Choi, S. Kim, J. M. Luther, S.-W. Kim, D. Shin, M. C. Beard, S. Jeong, Chem. Eur. J.
2017, 23, 17707–17713.
[73] A. J. Morris-Cohen, M. T. Frederick, G. D. Lilly, E. A. McArthur, E. A. Weiss, J. Phys.
Chem. Lett. 2010, 1, 1078–1081.
[74] I. Moreels, K. Lambert, D. De Muynck, F. Vanhaecke, D. Poelman, J. C. Martins, G. Allan,
Z. Hens, Chem. Mater. 2007, 19, 6101–6106.
[75] C. Bullen, P. Mulvaney, Langmuir 2006, 22, 3007–3013.
[76] J. J. Buckley, E. Couderc, M. J. Greaney, J. Munteanu, C. T. Riche, S. E. Bradforth, R. L.
Brutchey, ACS Nano 2014, 8, 2512–2521.
[77] V. Grigel, L. K. Sagar, K. De Nolf, Q. Zhao, A. Vantomme, J. De Roo, I. Infante, Z. Hens,
Chem. Mater. 2018, 30, 7637–7647.
[78] K. Sandeep, K. Y. Gopika, M. R. Revathi, Phys. Status Solidi RRL 2019, 13, 1900387.
[79] S. ten Brinck, I. Infante, ACS Energy Lett. 2016, 1, 1266–1272.
[80] J. De Roo, M. Ibáñez, P. Geiregat, G. Nedelcu, W. Walravens, J. Maes, J. C. Martins, I.
Van Driessche, M. V. Kovalenko, Z. Hens, ACS Nano 2016, 10, 2071–2081.
[81] M. V. Kovalenko, L. Protesescu, M. I. Bodnarchuk, Science 2017, 358, 745–750.
[82] H. Huang, M. I. Bodnarchuk, S. V. Kershaw, M. V. Kovalenko, A. L. Rogach, ACS Energy
Lett. 2017, 2, 2071–2083.
[83] F. Yan, S. T. Tan, X. Li, H. V. Demir, Small 2019, 15, 1902079.
[84] L. Protesescu, S. Yakunin, M. I. Bodnarchuk, F. Krieg, R. Caputo, C. H. Hendon, R. X.
Yang, A. Walsh, M. V. Kovalenko, Nano Lett. 2015, 15, 3692–3696.
[85] P. Cottingham, R. L. Brutchey, Chem. Commun. 2016, 52, 5246–5249.
[86] P. Cottingham, R. L. Brutchey, Chem. Commun. 2016, 28, 7574–7577.
135

[87] P. Cottingham, R. L. Brutchey, Chem. Mater. 2018, 30, 6711–6716.
[88] A. Pan, B. He, X. Fan, Z. Liu, J. J. Urban, A. P. Alivisatos, L. He, Y. Liu, ACS Nano 2016,
10, 7943–7954.
[89] M. C. Brennan, M. Kuno, S. Rouvimov, Inorg. Chem. 2019, 58, 1555–1560.
[90] V. K. Ravi, P. K. Santra, N. Joshi, J. Chugh, S. K. Singh, H. Rensmo, P. Ghosh, A. Nag,
J. Phys. Chem. Lett. 2017, 8, 4988–4994.
[91] M. I. Bodnarchuk, S. C. Boehme, S. ten Brinck, C. Bernasconi, Y. Shynkarenko, F. Krieg,
R. Widmer, B. Aeschlimann, D. Günther, M. V. Kovalenko, I. Infante, ACS Energy Lett.
2019, 4, 63–74.
[92] Y. Chen, S. R. Smock, A. H. Flintgruber, F. A. Perras, R. L. Brutchey, A. J. Rossini, J.
Am. Chem. Soc. 2020, 142, 6117–6127.
[93] S. R. Smock, T. J. Williams, R. L. Brutchey, Angew. Chem. Int. Ed. 2018, 57, 11711–
11715.
[94] A. Antanovich, A. Prudnikau, A. Matsukovich, A. Achtstein, M. Artemyev, J. Phys. Chem.
C 2016, 120, 5764–5775.
[95] S. Wang, L. Du, Z. Jin, Y. Xin, H. Mattoussi, J. Am. Chem. Soc. 2020, 142, 12669–12680.
[96] D. Quarta, M. Imran, A.-L. Capodilupo, U. Petralanda, B. van Beek, F. De Angelis, L.
Manna, I. Infante, L. De Trizio, C. Giansante, J. Phys. Chem. Lett. 2019, 10, 3715–3726.
[97] M. Imran, P. Ijaz, L. Goldoni, D. Maggioni, U. Petralanda, M. Prato, G. Almeida, I. Infante,
L. Manna, ACS Energy Lett. 2019, 4, 819–824.
[98] P. Ijaz, M. Imran, M. M. Soares, H. C. N. Tolentino, B. Martín-García, C. Giannini, I.
Moreels, L. Manna, R. Krahne, J. Phys. Chem. Lett. 2020, 11, 2079–2085.
136

[99] F. Krieg, S. T. Ochsenbein, S. Yakunin, S. ten Brinck, P. Aellen, A. Süess, B. Clerc, D.
Guggisberg, O. Nazarenko, Y. Shynkarenko, S. Kumar, C.-J. Shih, I. Infante, M. V.
Kovalenko, ACS Energy Lett. 2018, 3, 641–646.
[100] G. Schwarzenbach, Helv. Chim. Acta 1952, 35, 2344–2359.
[101] F. Krieg, Q. K. Ong, M. Burian, G. Rainò, D. Naumenko, H. Amenitsch, A. Süess, M. J.
Grotevent, F. Krumeich, M. I. Bodnarchuk, I. Shorubalko, F. Stellacci, M. V. Kovalenko,
J. Am. Chem. Soc. 2019, 141, 19839–19849.
[102] P. G. De Gennes, Macromolecules 1981, 14, 1637–1644.
[103] P. G. De Gennes, Macromolecules 1982, 15, 492–500.
[104] P. G. De Gennes, Adv. Colloid Interface Sci. 1987, 27, 189–209.
[105] Y. Tan, Y. Zou, L. Wu, Q. Huang, D. Yang, M. Chen, M. Ban, C. Wu, T. Wu, S. Bai, T.
Song, Q. Zhang, B. Sun, ACS Appl. Mater. Interfaces 2018, 10, 3784–3792.
[106] W. Koh, S. Park, Y. Ham, ChemistrySelect 2016, 1, 3479–3482.
[107] D. P. Nenon, K. Pressler, J. Kang, B. A. Koscher, J. H. Olshansky, W. T. Osowiecki, M.
A. Koc, L.-W. Wang, A. P. Alivisatos, J. Am. Chem. Soc. 2018, 140, 17760–17772.
[108] D. Yang, X. Li, W. Zhou, S. Zhang, C. Meng, Y. Wu, Y. Wang, H. Zeng, Adv. Mater.
2019, 1900767.
[109] M. J. Greaney, E. Couderc, J. Zhao, B. A. Nail, M. Mecklenburg, W. Thornbury, F. E.
Osterloh, S. E. Bradforth, R. L. Brutchey, Chem. Mater. 2015, 27, 744–756.
[110] M. J. Greaney, J. Joy, B. A. Combs, S. Das, J. J. Buckley, S. E. Bradforth, R. L. Brutchey,
J. Chem. Phys. 2019, 151, 074704.
[111] D. A. Ruddy, P. T. Erslev, S. E. Habas, J. A. Seabold, N. R. Neale, J. Phys. Chem. Lett.
2013, 4, 416–421.
137

[112] D. A. Ruddy, J. C. Johnson, E. R. Smith, N. R. Neale, ACS Nano 2010, 4, 7459–7466.
[113] R. J. Alan Esteves, S. Hafiz, D. O. Demchenko, Ü. Özgür, I. U. Arachchige, Chem.
Commun. 2016, 52, 11665–11668.
[114] A. L. Holmes, J. Hütges, A. Reckmann, E. Muthuswamy, K. Meerholz, S. M. Kauzlarich,
J. Phys. Chem. C 2015, 119, 5671–5678.
[115] A. T. Kerr, D. Placencia, M. E. Gay, J. E. Boercker, D. Soto, M. H. Davis, N. A. Banek,
E. E. Foos, J. Phys. Chem. C 2017, 121, 22597–22606.
[116] E. Muthuswamy, A. S. Iskandar, M. M. Amador, S. M. Kauzlarich, Chem. Mater. 2012,
25, 1416–1422.
[117] A. Bernard, K. Zhang, D. Larson, K. Tabatabaei, S. M. Kauzlarich, Inorg. Chem. 2018, 57,
5299–5306.
[118] M. He, L. Protesescu, R. Caputo, F. Krumeich, M. V. Kovalenko, Chem. Mater. 2015, 27,
635–647.
[119] K. Kravchyk, L. Protesescu, M. I. Bodnarchuk, F. Krumeich, M. Yarema, M. Walter, C.
Guntlin, M. V. Kovalenko, J. Am. Chem. Soc. 2013, 135, 4199–4202.
[120] G. M. Carroll, R. Limpens, N. R. Neale, Nano Lett. 2018, 18, 3118–3124.
[121] K. Tabatabaei, A. L. Holmes, K. A. Newton, E. Muthuswamy, R. Sfadia, S. A. Carter, S.
M. Kauzlarich, Chem. Mater. 2019, 31, 7510–7521.
[122] K. Tabatabaei, H. Lu, B. M. Nolan, X. Cen, C. E. McCold, X. Zhang, R. L. Brutchey, K.
van Benthem, J. Hihath, S. M. Kauzlarich, Chem. Mater. 2017, 29, 7353–7363.
[123] Z. Hens, J. C. Martins, Chem. Mater. 2013, 25, 1211–1221.
[124] L. E. Marbella, J. E. Millstone, Chem. Mater. 2015, 27, 2721–2739.
[125] S. Mourdikoudis, L. M. Liz-Marzán, Chem. Mater. 2013, 25, 1465–1476.
138

[126] N. Geva, J. J. Shepherd, L. Nienhaus, M. G. Bawendi, T. Van Voorhis, J. Phys. Chem. C
2018, 122, 26267–26274.
[127] B. Fritzinger, R. K. Capek, K. Lambert, J. C. Martins, Z. Hens, J. Am. Chem. Soc. 2010,
132, 10195–10201.
[128] J. S. Kachian, K. T. Wong, S. F. Bent, Acc. Chem. Res. 2010, 43, 346–355.
[129] P. Prayongpan, C. Michael Greenlief, Surf. Sci. 2009, 603, 1055–1069.
[130] M. N. Hopkinson, C. Richter, M. Schedler, F. Glorius, Nature 2014, 510, 485–496.
[131] R. H. Crabtree, J. Organomet. Chem. 2005, 690, 5451–5457.
[132] J. Vignolle, T. D. Tilley, Chem. Commun. 2009, 7230–7232.
[133] X. Ling, S. Roland, M.-P. Pileni, Chem. Mater. 2015, 27, 414–423.
[134] S. G. Song, C. Satheeshkumar, J. Park, J. Ahn, T. Premkumar, Y. Lee, C. Song,
Macromolecules 2014, 47, 6566–6571.
[135] H. Lu, Z. Zhou, O. V. Prezhdo, R. L. Brutchey, J. Am. Chem. Soc. 2016, 138, 14844–
14847.
[136] E. A. Baquero, S. Tricard, J. C. Flores, E. de Jesús, B. Chaudret, Angew. Chem. Int. Ed.
2014, 53, 13220–13224.
[137] P. Lara, O. Rivada-Wheelaghan, S. Conejero, R. Poteau, K. Philippot, B. Chaudret, Angew.
Chem. Int. Ed. 2011, 50, 12080–12084.
[138] M. Valero, D. Bouzouita, A. Palazzolo, J. Atzrodt, C. Dugave, S. Tricard, S. Feuillastre,
G. Pieters, B. Chaudret, V. Derdau, Angew. Chem. Int. Ed. 2020, 59, 3517–3522.
[139] M. J. MacLeod, J. A. Johnson, J. Am. Chem. Soc. 2015, 137, 7974–7977.
[140] Z. Cao, D. Kim, D. Hong, Y. Yu, J. Xu, S. Lin, X. Wen, E. M. Nichols, K. Jeong, J. A.
Reimer, P. Yang, C. J. Chang, J. Am. Chem. Soc. 2016, 138, 8120–8125.
139

[141] L. Zhang, Z. Wei, S. Thanneeru, M. Meng, M. Kruzyk, G. Ung, B. Liu, J. He, Angew.
Chem. Int. Ed. 2019, 58, 15834–15840.
[142] C. Richter, K. Schaepe, F. Glorius, B. J. Ravoo, Chem. Commun. 2014, 50, 3204.
[143] A. Ferry, K. Schaepe, P. Tegeder, C. Richter, K. M. Chepiga, B. J. Ravoo, F. Glorius, ACS
Catal. 2015, 5, 5414–5420.
[144] N. Kaeffer, H.-J. Liu, H.-K. Lo, A. Fedorov, C. Copéret, Chem. Sci. 2018, 9, 5366–5371.
[145] H. Lu, R. L. Brutchey, Chem. Mater. 2017, 29, 1396–1403.
[146] B. A. Tappan, K. Chen, H. Lu, S. M. Sharada, R. L. Brutchey, ACS Appl. Mater. Interfaces
2020, 12, 16394–16401.
[147] X. Hou, X. Zhang, W. Yang, Y. Liu, X. Zhai, Mater. Res. Bull 2012, 47, 2579–2583.
[148] D. Ruiz, M. Mizrahi, H. D. A. Santos, D. Jaque, C. M. S. Jones, J. Marqués-Hueso, C.
Jacinto, F. G. Requejo, A. Torres-Pardo, J. M. González-Calbet, B. H. Juárez, Nanoscale
2019, 11, 9194–9200.
[149] D. Baranov, M. J. Lynch, A. C. Curtis, A. R. Carollo, C. R. Douglass, A. M. Mateo-Tejada,
D. M. Jonas, Chem. Mater. 2019, 31, 1223–1230.
[150] R. L. Brutchey, Z. Hens, M. V. Kovalenko, Blackwell Science, Oxford, 2015, pp. 233–
271.
[151] F. Zhu, L. Men, Y. Guo, Q. Zhu, U. Bhattacharjee, P. M. Goodwin, J. W. Petrich, E. A.
Smith, J. Vela, ACS Nano 2015, 9, 2948–2959.
[152] A. Swarnkar, A. R. Marshall, E. M. Sanehira, B. D. Chernomordik, D. T. Moore, J. A.
Christians, T. Chakrabarti, J. M. Luther, Science 2016, 354, 92–95.
[153] L. Protesescu, S. Yakunin, M. I. Bodnarchuk, F. Bertolotti, N. Masciocchi, A. Guagliardi,
M. V. Kovalenko, J. Am. Chem. Soc. 2016, 138, 14202–14205.
140

[154] A. Swarnkar, R. Chulliyil, V. K. Ravi, M. Irfanullah, A. Chowdhury, A. Nag, Angew.
Chem. Int. Ed. 2015, 54, 15424–15428.
[155] T. Xuan, X. Yang, S. Lou, J. Huang, Y. Liu, J. Yu, H. Li, K.-L. Wong, C. Wang, J. Wang,
Nanoscale 2017, 9, 15286–15290.
[156] L. Zhang, X. Yang, Q. Jiang, P. Wang, Z. Yin, X. Zhang, H. Tan, Y. (Michael) Yang, M.
Wei, B. R. Sutherland, E. H. Sargent, J. You, Nat. Commun. 2017, 8, 15640.
[157] E. M. Sanehira, A. R. Marshall, J. A. Christians, S. P. Harvey, P. N. Ciesielski, L. M.
Wheeler, P. Schulz, L. Y. Lin, M. C. Beard, J. M. Luther, Sci. Adv. 2017, 3, eaao4204.
[158] H. Huang, B. Chen, Z. Wang, T. F. Hung, A. S. Susha, H. Zhong, A. L. Rogach, Chem. Sci
2016, 7, 5699–5703.
[159] X. Yuan, X. Hou, J. Li, C. Qu, W. Zhang, J. Zhao, H. Li, Phys. Chem. Chem. Phys. 2017,
19, 8934–8940.
[160] D. Yang, X. Li, H. Zeng, Adv. Mater. Interfaces 2018, 5, 1701662.
[161] K. Lin, J. Xing, L. N. Quan, F. P. G. de Arquer, X. Gong, J. Lu, L. Xie, W. Zhao, D. Zhang,
C. Yan, W. Li, X. Liu, Y. Lu, J. Kirman, E. H. Sargent, Q. Xiong, Z. Wei, Nature 2018,
562, 245–248.
[162] Z. Wei, J. Xing, J. Phys. Chem. Lett. 2019, 10, 3035–3042.
[163] M. Lu, Y. Zhang, S. Wang, J. Guo, W. W. Yu, A. L. Rogach, Adv. Funct. Mater. 2019, 29,
1902008.
[164] Q. Tai, K.-C. Tang, F. Yan, Energy Environ. Sci. 2019, 12, 2375–2405.
[165] K. Wang, S. Wang, S. Xiao, Q. Song, Adv. Opt. Mater. 2018, 6, 1800278.
141

[166] M. Hao, Y. Bai, S. Zeiske, L. Ren, J. Liu, Y. Yuan, N. Zarrabi, N. Cheng, M. Ghasemi, P.
Chen, M. Lyu, D. He, J.-H. Yun, Y. Du, Y. Wang, S. Ding, A. Armin, P. Meredith, G. Liu,
H.-M. Cheng, L. Wang, Nat. Energy 2020, 5, 79–88.
[167] R. Brutchey, Z. Hens, M. V. Kovalenko, in Chemistry of Organo-Hybrids (Eds.: B.
Charleux, C. Copéret, E. Lacôte), John Wiley & Sons, Inc., Hoboken, NJ, USA, 2015, pp.
233–271.
[168] J. Pan, Y. Shang, J. Yin, M. De Bastiani, W. Peng, I. Dursun, L. Sinatra, A. M. El-Zohry,
M. N. Hedhili, A.-H. Emwas, O. F. Mohammed, Z. Ning, O. M. Bakr, J. Am. Chem. Soc.
2018, 140, 562–565.
[169] J. Pan, L. N. Quan, Y. Zhao, W. Peng, B. Murali, S. P. Sarmah, M. Yuan, L. Sinatra, N.
M. Alyami, J. Liu, E. Yassitepe, Z. Yang, O. Voznyy, R. Comin, M. N. Hedhili, O. F.
Mohammed, Z. H. Lu, D. H. Kim, E. H. Sargent, O. M. Bakr, Adv. Mater. 2016, 28, 8718–
8725.
[170] Q. A. Akkerman, G. Rainò, M. V. Kovalenko, L. Manna, Nat. Mater. 2018, 17, 394–405.
[171] B. Chen, P. N. Rudd, S. Yang, Y. Yuan, J. Huang, Chem. Soc. Rev. 2019, 48, 3842–3867.
[172] J. Pan, S. P. Sarmah, B. Murali, I. Dursun, W. Peng, M. R. Parida, J. Liu, L. Sinatra, N.
Alyami, C. Zhao, E. Alarousu, T. K. Ng, B. S. Ooi, O. M. Bakr, O. F. Mohammed, J. Phys.
Chem. Lett. 2015, 6, 5027–5033.
[173] L. M. Wheeler, E. M. Sanehira, A. R. Marshall, P. Schulz, M. Suri, N. C. Anderson, J. A.
Christians, D. Nordlund, D. Sokaras, T. Kroll, S. P. Harvey, J. J. Berry, L. Y. Lin, J. M.
Luther, J. Am. Chem. Soc. 2018, 140, 10504–10513.
[174] G. Almeida, I. Infante, L. Manna, Science 2019, 364, 833–834.
142

[175] D. D. Lovingood, R. Achey, A. K. Paravastu, G. F. Strouse, J. Am. Chem. Soc. 2010, 132,
3344–3354.
[176] H. Virieux, M. Le Troedec, A. Cros-Gagneux, W.-S. Ojo, F. Delpech, C. Nayral, H.
Martinez, B. Chaudret, J. Am. Chem. Soc. 2012, 134, 19701–19708.
[177] M. P. Hanrahan, E. L. Fought, T. L. Windus, L. M. Wheeler, N. C. Anderson, N. R. Neale,
A. J. Rossini, Chem. Mater. 2017, 29, 10339–10351.
[178] S. Cadars, B. J. Smith, J. D. Epping, S. Acharya, N. Belman, Y. Golan, B. F. Chmelka,
Phys. Rev. Lett. 2009, 103, 136802.
[179] R. Wang, C. I. Ratcliffe, X. Wu, O. Voznyy, Y. Tao, K. Yu, J. Phys. Chem. C 2009, 113,
17979–17982.
[180] J. P. Yesinowski, in Solid State NMR (Ed.: J.C.C. Chan), Springer Berlin Heidelberg,
Berlin, Heidelberg, 2011, pp. 229–312.
[181] G. M. Bernard, R. E. Wasylishen, C. I. Ratcliffe, V. Terskikh, Q. Wu, J. M. Buriak, T.
Hauger, J. Phys. Chem. A 2018, 122, 1560–1573.
[182] A. Karmakar, A. M. Askar, G. M. Bernard, V. V. Terskikh, M. Ha, S. Patel, K. Shankar,
V. K. Michaelis, Chem. Mater. 2018, 30, 2309–2321.
[183] W. M. J. Franssen, S. G. D. van Es, R. Dervişoğlu, G. A. de Wijs, A. P. M. Kentgens, J.
Phys. Chem. Lett. 2017, 8, 61–66.
[184] O. Knop, R. E. Wasylishen, M. A. White, T. S. Cameron, M. J. M. V. Oort, Can. J. Chem.
1990, 68, 412–422.
[185] Q. Xu, T. Eguchi, H. Nakayama, N. Nakamura, M. Kishita, Z. Naturforsch. A 1991, 46,
240–246.
143

[186] D. H. Fabini, T. A. Siaw, C. C. Stoumpos, G. Laurita, D. Olds, K. Page, J. G. Hu, M. G.
Kanatzidis, S. Han, R. Seshadri, J. Am. Chem. Soc. 2017, 139, 16875–16884.
[187] D. J. Kubicki, D. Prochowicz, A. Hofstetter, S. M. Zakeeruddin, M. Grätzel, L. Emsley, J.
Am. Chem. Soc. 2017, 139, 14173–14180.
[188] B. A. Rosales, L. Men, S. D. Cady, M. P. Hanrahan, A. J. Rossini, J. Vela, Chem. Mater.
2016, 28, 6848–6859.
[189] D. J. Kubicki, D. Prochowicz, A. Hofstetter, P. Péchy, S. M. Zakeeruddin, M. Grätzel, L.
Emsley, J. Am. Chem. Soc. 2017, 139, 10055–10061.
[190] B. A. Rosales, M. P. Hanrahan, B. W. Boote, A. J. Rossini, E. A. Smith, J. Vela, ACS
Energy Lett. 2017, 2, 906–914.
[191] M. P. Hanrahan, L. Men, B. A. Rosales, J. Vela, A. J. Rossini, Chem. Mater. 2018, 30,
7005–7015.
[192] A. M. Askar, G. M. Bernard, B. Wiltshire, K. Shankar, V. K. Michaelis, J. Phys. Chem. C
2017, 121, 1013–1024.
[193] A. Karmakar, M. S. Dodd, X. Zhang, M. S. Oakley, M. Klobukowski, V. K. Michaelis,
Chem. Commun. 2019, 55, 5079–5082.
[194] L. Piveteau, T.-C. Ong, A. J. Rossini, L. Emsley, C. Copéret, M. V. Kovalenko, J. Am.
Chem. Soc. 2015, 137, 13964–13971.
[195] A. A. M. Brown, T. J. N. Hooper, S. A. Veldhuis, X. Y. Chin, A. Bruno, P. Vashishtha, J.
N. Tey, L. Jiang, B. Damodaran, S. H. Pu, S. G. Mhaisalkar, N. Mathews, Nanoscale 2019,
11, 12370–12380.
[196] M. C. Brennan, J. E. Herr, T. S. Nguyen-Beck, J. Zinna, S. Draguta, S. Rouvimov, J.
Parkhill, M. Kuno, J. Am. Chem. Soc. 2017, 139, 12201–12208.
144

[197] Z. Gan, Chem. Commun. 2006, 4712–4714.
[198] M. Feike, D. E. Demco, R. Graf, J. Gottwald, S. Hafner, H. W. Spiess, J. Magn. Reson. A
1996, 122, 214–221.
[199] I. Schnell, A. Lupulescu, S. Hafner, D. E. Demco, H. W. Spiess, J. Magn. Reson. 1998,
133, 61–69.
[200] I. Schnell, Prog. Nucl. Magn. Reson. Spectrosc. 2004, 45, 145–207.
[201] J. W. Wiench, C. E. Bronnimann, V. S.-Y. Lin, M. Pruski, J. Am. Chem. Soc. 2007, 129,
12076–12077.
[202] L. C. Thomas, The Identification of Functional Groups in Organophosphorus Compounds,
Academic Press, 1974.
[203] M. M. Crutchfield, E. J. Griffith, M. Grayson, Topics in Phosphorus Chemistry,
Interscience, 1967.
[204] G. P. Holland, R. Sharma, J. O. Agola, S. Amin, V. C. Solomon, P. Singh, D. A. Buttry, J.
L. Yarger, Chem. Mater. 2007, 19, 2519–2526.
[205] D. J. Kubicki, D. Prochowicz, A. Pinon, G. Stevanato, A. Hofstetter, S. M. Zakeeruddin,
M. Grätzel, L. Emsley, J. Mater. Chem. A 2019, 7, 2326–2333.
[206] A. J. Rossini, M. P. Hanrahan, M. Thuo, Phys. Chem. Chem. Phys. 2016, 18, 25284–25295.
[207] L. Chen, Q. Wang, B. Hu, O. Lafon, J. Trébosc, F. Deng, J.-P. Amoureux, Phys. Chem.
Chem. Phys. 2010, 12, 9395.
[208] F. A. Perras, J. D. Padmos, R. L. Johnson, L.-L. Wang, T. J. Schwartz, T. Kobayashi, J. H.
Horton, J. A. Dumesic, B. H. Shanks, D. D. Johnson, M. Pruski, J. Am. Chem. Soc. 2017,
139, 2702–2709.
145

[209] Y.-Y. Hu, A. Rawal, K. Schmidt-Rohr, Proc. Natl. Acad. Sci. U. S. A. 2010, 107, 22425–
22429.
[210] F. A. Perras, Z. Wang, T. Kobayashi, A. Baiker, J. Huang, M. Pruski, Phys. Chem. Chem.
Phys. 2019, 21, 19529–19537.
[211] P. Berruyer, M. Lelli, M. P. Conley, D. L. Silverio, C. M. Widdifield, G. Siddiqi, D. Gajan,
A. Lesage, C. Copéret, L. Emsley, J. Am. Chem. Soc. 2017, 139, 849–855.
[212] S. S. Nagarkar, H. Kurasho, N. T. Duong, Y. Nishiyama, S. Kitagawa, S. Horike, Chem.
Commun. 2019, 55, 5455–5458.
[213] G. Bodenhausen, R. Freeman, G. A. Morris, J. Magn. Reson. 1976, 23, 171–175.
[214] G. A. Morris, R. Freeman, J. Magn. Reson. 1978, 29, 433–462.
[215] J. M. Goetz, J. Schaefer, J. Magn. Reson. 1997, 127, 147–154.
[216] X. Lu, O. Lafon, J. Trébosc, J.-P. Amoureux, J. Magn. Reson. 2012, 215, 34–49.
[217] C. C. Stoumpos, D. H. Cao, D. J. Clark, J. Young, J. M. Rondinelli, J. I. Jang, J. T. Hupp,
M. G. Kanatzidis, Chem. Mater. 2016, 28, 2852–2867.
[218] F. Bertolotti, G. Nedelcu, A. Vivani, A. Cervellino, N. Masciocchi, A. Guagliardi, M. V.
Kovalenko, ACS Nano 2019, 13, 14294–14307.

Abstract (if available)

Abstract The surface chemistry of nanocrystals is a critical component of the quantum dot construct, due to its contribution to both the colloidal and chemical stability of quantum dots and therefore, its underlying optoelectronic properties. In Chapter 1, an introduction to background and previous work on the surface chemistry of various quantum dots spanning from covalent to ionic will be presented. Beyond, we will present surface chemistry investigations performed on various QD materials. In Chapter 2, we will discuss our work on covalent Ge nanocrystals, in which we tracked ligand exchange reactions via ¹H NMR spectroscopy and determined that the nanocrystals were coordinated by strongly bound oleylamide ligands, with covalent X-type Ge-alkyl amide bonds. Our work extended into studying the surface chemistry of mixed covalent/ionic systems with interesting native L-type carbene ligands on Ag₂S QDs. Finally, we will conclude with our work on the highly ionic CsPbBr₃ QDs, which utilized both solution and solid-state NMR spectroscopy to gain insight into the ligand binding and surface termination of carboxylate and ammonium-terminated CsPbBr₃ quantum dots that have a surface binding or X₂-type. The binding strength of ligands tends to decrease with increasing ionicity of the semiconductor, which makes room temperature ligand exchange reactions facile to study for ionic semiconductors with NMR spectroscopy, however, in contrast for highly covalent semiconductors with covalently bound ligands, are much more difficult to exchange with new ligands at room temperature. Overall, we have found that the ligand binding of covalent, mixed, and ionic semiconductors tends to be covalent, mixed, or ionic in nature, respectively.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Nanocrystal surface engineering as a route towards improved photovoltaics

PDF

Expanding the library of surface ligands for semiconductor nanocrystal synthesis and photovoltaic applications

PDF

The chemistry of polymorphism in semiconductor nanocrystals

PDF

Semiconductor inks for solution processed electronic thin‐films

PDF

Vibrational spectroscopy and molecular orientation at interfaces: correlating the structure of a material's surface to its properties

PDF

Exploring temporal changes in surface species on weathered feldspar mineral surfaces using solid-state NMR spectroscopy

PDF

Non-radiative energy transfer for photovoltaic solar energy conversion: lead sulfide quantum dots on silicon nanopillars

PDF

Understanding organic and inorganic semiconductor surfaces by coherent nonlinear spectroscopy

PDF

The implementation of data-driven techniques for the synthesis and optimization of colloidal inorganic nanocrystals

PDF

SQUID-based CW NMR system for measuring the magnetization of ³He films

PDF

Exciton dynamics in photovoltaic materials

PDF

Synthesis and surface modification of perovskite-based nanocrystals for use in high energy density nanocomposites

PDF

Interfacial polarization and photobasicity: spectroscopic studies of the electrode environment and photochemical effects

PDF

Development of selective covalent probes to identify and modulate protein targets

PDF

Functionalization of nanodiamond surface for magnetic sensing application

PDF

Cooperative effect of ionic liquid solvents in microfluidic devices for fabricating monodisperse metal nanoparticles

PDF

Solution processed functional chalcogenide thin films and their molecular solutes from thiol-amine inks

PDF

Solution phase synthesis routes to functional nanomaterials for energy storage

PDF

The selective targeting of Ras: a tale of kinetics, mRNA display, and vaccines

PDF

Vibrational sum frequency spectroscopy of molecules on metal, semiconductor, and aqueous surfaces

Asset Metadata

Creator Smock, Sara Rose (author)

Core Title Quantifying the surface chemistry of semiconductor nanocrystals spanning covalent to ionic materials

School College of Letters, Arts and Sciences

Degree Doctor of Philosophy

Degree Program Chemistry

Degree Conferral Date 2021-12

Publication Date 10/08/2021

Defense Date 08/06/2021

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag CsPbBr₃,ligands,nanocrystals,NMR,NMR spectroscopy,nuclear magnetic resonance,OAI-PMH Harvest,quantum dots,surface chemistry

Format application/pdf (imt)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Brutchey, Richard L. (committee chair), Platt, John (committee member), Williams, Travis J. (committee member)

Creator Email sararsmock@gmail.com,smock@usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-oUC16023060

Unique identifier UC16023060

Legacy Identifier etd-SmockSaraR-10150

Document Type Dissertation

Format application/pdf (imt)

Rights Smock, Sara Rose

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 author, as the original true and official version of the work, but does not grant the reader permission to use the work if the desired use is covered by copyright. It is the author, as rights holder, who must provide use permission if such use is covered by copyright. The original signature page accompanying the original submission of the work to the USC Libraries is retained by the USC Libraries and a copy of it may be obtained by authorized requesters contacting the repository e-mail address given.

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

Repository Email cisadmin@lib.usc.edu