University of Southern California Dissertations and Theses

An investigation of morphology and transport in amorphous solid water via guest-host interactions

(USC Thesis Other)

An investigation of morphology and transport in amorphous solid water via guest-host interactions

PDF

Download

Open document

Flip pages

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content An Investigation of Morphology and Transport in Amorphous
Solid Water via Guest-Host Interactions
by
Jaimie Elizabeth Stomberg
A Dissertation Presented to the
FACULTY OF THE GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements of the Degree
DOCTOR OF PHILOSOPHY
(Chemistry)
August 2015
Copyright 2015 Jaimie Elizabeth Stomberg

I dedicate this thesis to my husband, Matthew Mills, who is my partner in science and
in life.
“Science is magic that works.” –Kurt Vonnegut
ii
Acknowledgements
First and foremost, I would like to thank my advisors: Curt Wittig and Hanna
Reisler. They have pushed me to think critically about the scientific process and my
data, and I have become a better scientist because of it. Without their guidance, I would
not be who I am today.
A close second in influence and help has been Stephanie McKean, my cohort in
science throughout my entire graduate school career. She has been a much-needed
resource of support, insight, and friendship in a stressful environment, and I will be
forever grateful for her.
I would also like to extend my appreciation to all of the people who have given
me help from the Wittig and Reisler groups throughout the years, as well as the phys-
ical chemistry community in SSC. They have given me valuable information, provided
equipment to fix a broken instrument, or pointed out something I never would have
thought of on my own. In particular, I have benefited from the wisdom of Oscar
Rebolledo-Mayoral and Chris Larson while they worked on this project with me.
I also want to thank the Women in Science and Engineering program (and by ex-
tension the Women in Chemistry group) at USC. Not only did they provide me with
funding, but they gave me the opportunity to be part of a community of women who
are inspiring, supportive, and smart as hell.
Lastly, I would like to thank my family. My father, Keith Stomberg, made sure I
knew how to fix things myself when I was young, which has proven to be a valuable
skill in the lab. My late brother, Wayne Stomberg, inspired me to try harder every
day to improve the world. My fantastic mother, Maribeth Stomberg, has supported me
since birth, and I could never have achieved as much without her presence in my life.
And of course, I never would have made it through graduate school without the steady
iii
presence of my husband, Matthew Mills. He is my rock and I am so thankful to have
him as my partner.
iv
Table of Contents
Dedication ii
Acknowledgements iii
List of Figures vii
List of Tables xiv
Abstract xv
1 Introduction 1
1.1 Water: a Complex Molecule . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.1.1 Amorphous Solid Water . . . . . . . . . . . . . . . . . . . . . . . . . 3
1.1.2 Amorphous Solid Water and Guest Molecule Interactions . . . . . 7
1.2 Amorphous Solid Water in the Interstellar Medium . . . . . . . . . . . . . 10
1.3 NO
2
/N
2
O
4
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
1.4 Thesis Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
Chapter References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
2 Experimental Methods 21
2.1 Time-of-Flight Mass Spectrometry . . . . . . . . . . . . . . . . . . . . . . . 21
2.1.1 Jacobian . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
2.1.2 Resolution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
2.2 Laser-Induced Desorption . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
Chapter References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
3 Experimental Apparatus 33
3.1 Ultra-High Vacuum System . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
3.2 Surface Manipulator . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
3.3 Surface Holder and MgO Surface . . . . . . . . . . . . . . . . . . . . . . . . 36
3.4 Sample Growth . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
3.4.1 Calibration Using Temperature-Programmed Desorption . . . . . 41
3.4.2 Directed Dosing vs. Background Dosing . . . . . . . . . . . . . . . 43
3.5 Time-of-Flight Mass Spectrometer . . . . . . . . . . . . . . . . . . . . . . . 45
3.6 Time Synchronization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
3.7 Fourier-Transform Infrared Spectroscopy . . . . . . . . . . . . . . . . . . . 51
Chapter References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
v
4 Effects of Buried Heat in Amorphous Solid Water Films 55
4.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55
4.2 Experimental . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
4.2.1 Changing to UV: Using NOCl as a Test Species . . . . . . . . . . . 60
4.3 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
4.3.1 Standard Spectra and Data Analysis . . . . . . . . . . . . . . . . . . 63
4.3.2 ASW Spacer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
4.3.3 N
2
O
4
Thickness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
4.3.4 Laser Energy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
4.3.5 Multiphoton Effects . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
4.3.6 ASW Upper Layer Thickness . . . . . . . . . . . . . . . . . . . . . . 76
4.3.7 High Fluence vs. Low Fluence . . . . . . . . . . . . . . . . . . . . . 86
4.3.8 N
2
O
4
Codeposition . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
4.4 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88
4.4.1 Time Scales . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
4.4.2 Fissures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92
4.5 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94
Chapter References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96
5 Introducing a Third Species: Molecular Transport and Isotope Exchange 99
5.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99
5.2 Experimental . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101
5.2.1 CO
2
and C
2
H
2
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102
5.2.2 D
2
O . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102
5.3 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103
5.3.1 Depth Profiling With CO
2
and C
2
H
2
. . . . . . . . . . . . . . . . . . 103
5.3.2 Switching to D
2
O . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107
5.4 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112
5.5 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115
Chapter References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117
6 Future Work 119
6.1 Improvements in Experimental Apparatus . . . . . . . . . . . . . . . . . . 119
6.2 Introducing a Third Species: What Next? . . . . . . . . . . . . . . . . . . . 121
6.3 Gold nanoparticles as Fissure-Creators . . . . . . . . . . . . . . . . . . . . 121
6.4 Experimental Strategy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 124
6.4.1 Surface Preparation and Experimental Adjustments . . . . . . . . . 124
6.5 Initial Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126
6.5.1 Nanoparticle Breakdown . . . . . . . . . . . . . . . . . . . . . . . . 127
6.6 Summary and Future Experiments . . . . . . . . . . . . . . . . . . . . . . . 128
Chapter References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131
A Appendix 132
BIBLIOGRAPHY 134
vi
List of Figures
1.1 A 3D depiction of hexagonal ice (HI), the most common form of crys-
talline ice on earth. Each water molecule participates in 4 hydrogen
bonds with neighboring molecules, which forms a tetrahedral structure
and characteristic long-range order. . . . . . . . . . . . . . . . . . . . . . 2
1.2 (a) Ball-and-stick model of N
2
O
4
. The NN bond is quite weak (around
0.5 eV) [2, 84]. (b) The bcc structure is the dominant structure of crys-
talline N
2
O
4
at low temperatures. Defects, including presence of the
NO
2
and unstable isomer ONONO
2
, start to appear below 20 K [17]. . . 13
2.1 A schematic of a linear two-stage Wiley-McLaren TOFMS system. Ma-
terial enters the region between repeller and extractor plates initially
kept at the same voltage (A). If necessary, the material is ionized (this
can be done by electron impact ionization, photoionization, etc). Sub-
sequently, the voltage is dropped on the extractor plate and ionized
material is extracted and further accelerated by a grounded third plate
into a field-free drift region (C). Ionized material of different masses can
be differentiated by flight times, with more massive ions arriving after
lighter ions (relative mass is indicated by size in the diagram). . . . . . . 22
2.2 Two particles of the same mass, A and B, are ionized in a finite ioniza-
tion region with a width ofDl. The flight path can therefore differ by
a range ofDl/2; in this case, particle A will have a shorter flight path
than particle B. However, particle B will gain more kinetic energy than
particle A. This means that there will be a point in space where particle
B will pass particle A. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
2.3 A particle trajectory, depicted as a blue circle, with initial kinetic energy
U
0
. Only the component in the direction of extraction will affect the
overall kinetic energy of the extracted particle, U
tot
. . . . . . . . . . . . . 27
3.1 An external view of the UHV chamber (not to scale). Key sections are
labeled. The surface manipulator provides xyz translation capabilities.
The window through which UV radiation enters for a typical experi-
ment is designated with a star. The diamond indicates the FTIR pathway. 35
vii
3.2 An exploded view of the substrate holder and associated components.
Two machined copper pieces are electrically isolated from each other
by a custom ceramic spacer and attached to the copper block with long
screws fixed with flat and ceramic hat washers on each side. Two 18-
gauge copper wires are screwed to the machined pieces to allow for
current to flow through the tantalum heater that is bonded to the back
of the substrate holder. A thermocouple is glued to the “hot” side of
the MgO substrate and connected to a wire threaded through the LN
2
reservoir sheath. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
3.3 An AFM image of a cleaved MgO(100) surface. The surface is quite flat
and unblemished, with terraces as the main visible feature along with
a slightly larger abberation. . . . . . . . . . . . . . . . . . . . . . . . . . . 39
3.4 (a) TPD spectra from ASW films (20-100 ML) grown in standard dosing
position (Tier 2). The bulk ASW desorption occurs at 165 K, with a
crystallization peak that appears at 160 K. (b) Integrated areas from the
plots in (a) compared to previously calibrated film thicknesses. . . . . . 42
3.5 FTIR spectra of ASW grown by background and directed dosing, each
at 2 10
7
Torr for 5 minutes (equivalent to 100 ML for background
deposition). Removing the contribution of material that grows on the
back side of the MgO substrate (see text for discussion of this calcula-
tion), directed dosing increases film thickness by a factor of 1.7 when
compared to background dosing. . . . . . . . . . . . . . . . . . . . . . . . 44
3.6 A schematic representation of the TOFMS experiment. The fourth har-
monic from a Continuum Nd:YAG is used. The beam waist is reduced
to 3 mm using a telescope. The radiation enters the chamber and ex-
cites the film at normal incidence. Released material is ionized, ex-
tracted, and then detected using a multi-channel plate (MCP) at a rate
of 100 kHz. The set-up shown is with focused radiation, which allows
for 9 individual “experiments” on one film via translation in the xz
plane, as indicated by the grid. These data can be averaged to improve
S/N. By removing the 50 cm lens, a larger area can be irradiated (shown
in light purple) to allow FTIR experiments of the ablated film. . . . . . . 46
3.7 (a) An example of a full temporal profile. It is the average of data
collected from eight individual spots from a film of 80 L of N
2
O
4
ablated
with focused 1.0 mJ radiation. Each 10 μs segment represents a full
mass spectrum, shown as (b), of a segment of the plume of material that
desorbs from the surface. The data can be analyzed as a full temporal
profile or as individual mass spectra. . . . . . . . . . . . . . . . . . . . . 48
viii
3.8 A schematic of the four SRS DG535 Digital Delay/Pulse Generators that
control the triggering for the TOFMS, chopper wheel, and Nd:YAG laser
flash lamps and Q-switch; only necessary connections and switches
have been depicted for clarity. Refer to Table 3.1 for individual pulse
generator settings. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
3.9 A schematic of the top chamber tier, showing the FTIR pathway. Ra-
diation from the detector is directed and focused through the surface
using flat and parabolic mirrors. It is then focused on a liquid nitrogen-
cooled InSb detector with a second parabolic mirror. All external optics
are encased in plexiglass boxes that are purged with dry air. . . . . . . . 52
4.1 (a) FTIR spectrum of 100 L of N
2
O
4
.The sharp peak at 2960 cm
1
is
a weak combination band [110]. All fundamental bands for N
2
O
4
are
below the detection range of the InSb detector used in these experi-
ments (cutoff at 1850 cm
1
), so this small peak was the only indication
of the presence of N
2
O
4
on the surface. However, it was distinct and
reproducible. (b) FTIR spectrum of 110 L of NOCl (negative peaks are
due to decrease in water vapor signal on the experimental time scale).
The peak centered at 1955 cm
1
is the NO bond stretching mode [62].
We did not detect any other peaks associated with NOCl. While this
particular peak is also representative of NO only, our surface was not
cold enough to condense NO [35]. Furthermore, we were able to desorb
fragments of NOCl successfully with UV radiation. . . . . . . . . . . . . 62
4.2 Mass spectrum of NO
2
, N
2
O
4
, NO, and H
2
O obtained from our exper-
imental apparatus. All ratios given are based on peak areas.(a) 300 K
NO
2
leaked into the UHV chamber gives NO
+
/NO
2
+
= 3.18, O
+
/NO
2
+
= 0.58, and N
+
/NO
2
+
= 0.19. (b) 300 K NO leaked into the chamber
gives O
+
/NO
+
= 0.01 and N
+
/NO
+
= 0.06. The NO
2+
signal at m/q =
15 is a distinctive feature of NO electron impact ionization. (c) 300 K
H
2
O leaked into the chamber gives OH
+
/H
2
O
+
= 0.27, O
+
/H
2
O
+
=
0.02, and H
+
/H
2
O
+
(not shown) = 0.07. (d) The spectrum obtained via
N
2
O
4
sublimation at155 K gives NO
+
/NO
2
+
= 0.9, O
+
/NO
2
+
= 0.1,
and N
+
/NO
2
+
= 0.02. Neither N
2
O
3
+
or N
2
O
4
+
(76 m/q and 92 m/q,
respectively) was detected (not shown). . . . . . . . . . . . . . . . . . . . 66
4.3 (a)The initial design for a TOFMS electrode to minimize the effect of
molecular collisions with the plates. (b) Improved design for a TOFMS
electrode to allow material to pass out of the ionization region. The
stars designate the openings parallel to the substrate through which
desorbing material would enter the ionizing region. . . . . . . . . . . . . 67
ix
4.4 Temporal profiles of NO
+
x
signal with and without an ASW spacer, with
cartoon depictions of deposition films. Both films were irradiated with
a focused 266 nm, 1.0 mJ beam. (a) 80 L N
2
O
4
deposited on a bare MgO
substrate. (b) 80 L N
2
O
4
deposited on 300 ML ASW. . . . . . . . . . . . 69
4.5 Averaged mass spectrum from extractions 2-11 (20-110 μs) of data from
Figure 4.4b. All species detected can be attributed to N
2
O
4
irradiation
and cracking. There is a slight bump at mass 18, but it is insignificant in
comparison to the other peaks and compared to previous experiments
conducted by this lab. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
4.6 Sum of peak areas for NO
+
and NO
2
+
signals from laser ablation of 20-
100 L N
2
O
4
films grown on 300 ML of ASW. The signal increases with
increasing thickness, indicating that absorption within the film hasn’t
been saturated in the range of thicknesses studied. . . . . . . . . . . . . 72
4.7 Sum of peak areas for NO
+
and NO
2
+
signals from laser ablation of
80 L N
2
O
4
dosed on 300 ML H
2
O. The signal generally increases with
increasing laser energy. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
4.8 A representative resolved TOF spectrum of the fast ions from the first
extraction. The spectrum has been shifted to account for the 2 μs delay.
The peaks are labeled with probable species assignments. . . . . . . . . 76
4.9 Nine first-pulse temporal profiles from a single 300 ML D
2
O/80 L N
2
O
4
/2400 ML
H
2
O sandwich (1.0 mJ focused radiation, configuration shown in car-
toon, depicted on common scale). For clarity, water signal is shown in
black and NO
+
x
signal is shown in red. The irregularity of the water
signal, particularly at long times, is indicative of nonthermal processes.
The amount of signal at long times is considerable. It should also be
noted that the NO
+
x
signal does not exhibit this behavior and decays
with time. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
4.10 A water monomer’s kinetic energy compared to its flight time, assum-
ing the molecule left the surface at t = 0. . . . . . . . . . . . . . . . . . . 79
4.11 Temporal profiles comparing first and second incident pulses for upper
H
2
O layer thickness: (1) 600 ML, (2) 1200 ML, and (3) 2400 ML. Each
profile is the average of nine individual profiles (1 mJ, 266 nm). Red
peaks designate NO
+
and NO
2
+
. Black peaks denote H
2
O
+
, OH
+
, and
H
+
. All temporal profiles are on the same ordinate scale. The lower
ASW layer is 300 ML of D
2
O and the N
2
O
4
layer is 80 L (sample com-
position shown at bottom). . . . . . . . . . . . . . . . . . . . . . . . . . . 81
4.12 Mass spectra representing the 100 μs extraction from each temporal
profile in Figure 4.11; refer to the caption of that Figure for composition
details. The peaks in spectrum (a) have been labeled with ion assign-
ments for clarity. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82
x
4.13 Plots of R
N
for each extraction with good S/N from each temporal
profile in Figure 4.11; refer to the caption of that figure for composition
details. After the initial extractions, first pulse R
N
values (filled circles)
level off around unity. Second pulse R
N
values (open triangles) follow
the same qualitative trend but level off at higher values, indicating the
presence of N
2
O
4
in the surface region. . . . . . . . . . . . . . . . . . . . 84
4.14 (a) Overlaid FTIR spectra of a 300 ML D
2
O/80 L N
2
O
4
/2400 ML H
2
O
sandwich (configuration depicted in figure) before and after 70 shots of
6 mJ radiation. A 3 mm beam waist was used to ensure that a reason-
able area was available for FTIR measurements. The red line represents
the spectrum of the undisturbed film; the black line is the spectrum col-
lected after irradiation. The OD stretch peak centered at 2435 cm
1
is virtually unchanged, but the OH peak centered at 3250 cm
1
is
reduced by 14%. (b) 2400 ML of H
2
O was annealed at 165 K for 10 min-
utes to induce crystallization and is shown for comparison. The “after”
spectrum in (a) did not show any observable increase in crystallinity
despite the prolonged exposure to laser radiation. . . . . . . . . . . . . . 85
4.15 Averaged mass spectrum from extractions 2-11 (20-110 μs) of data from
a low fluence experiment (0.2 J/cm
2
) with the same film composition as
Figures 4.4b and 4.5 (300 ML ASW/80 L N
2
O
4
). The difference in the
NO
+
/NO
2
+
ratio compared to Figure 4.5 is striking. . . . . . . . . . . . . 86
4.16 Expanded mass spectrum of 80 L N
2
O
4
codeposited with 240 ML of
ASW and irradiated with focused 1.0 mJ radiation. Five first-shot exper-
iments from the film have been averaged to improve S/N. While small,
clusters up to the protonated water tetramer were visible. No proto-
nated species were detected in any of the experiments where N
2
O
4
was
deposited as a discrete layer. It is also important to note that there is no
signal at m/q = 62, which would indicate the presence of NO
3
+
. . . . . 89
4.17 A cartoon depiction of how material expelled from neighboring fissures
may interact. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
5.1 Cartoon depiction of different sandwich configurations using a third
species as a probe (black represents the MgO substrate). Typical ASW
sandwiches are deposited in layers, with layers of CO
2
or C
2
H
2
de-
posited at different depths within a film. Thicknesses used were 300 ML
for the bottom H
2
O spacer, 80 L for the N
2
O
4
layer, and 100 ML of CO
2
or C
2
H
2
. The top H
2
O layer was 200 ML [consisting of two separate
100 ML sections for configuration (3)]. . . . . . . . . . . . . . . . . . . . . 102
xi
5.2 Mass spectra averaged from extractions 2-21 where CO
2
is underneath
the ASW/N
2
O
4
sandwich [configuration (1) in Figure 5.1]. Mass spec-
trum (a) is after the first laser pulse, while (b) is from the second pulse
incident on the same location (2.0 mJ); CO
2
+
peaks are boxed in red for
clarity. In (a), the very small peak at mass 44 shows how little CO
2
es-
capes the film. This signal increases significantly after the second pulse
on the same spot, indicating a morphological change that allows for
CO
2
to move upward through the film. . . . . . . . . . . . . . . . . . . . 105
5.3 Temporal profiles comparing CO
2
+
signal in configuration (2) versus (3)
(refer to Figure 5.1). Trace (a) is the result of configuration (2); trace
(b) is due to ablation of configuration (3). For both temporal profiles,
data was averaged for the first incident shot on 9 spots on the film
(2.0 mJ). Signal due to H
2
O, N
x
O
y
and CO
2
are designated by black,
red, and blue, respectively. Because CO
2
+
is partially obscured by NO
+
x
signal, a trace of its outline is drawn in blue for clarity. (a) When CO
2
is
deposited directly on top of N
2
O
4
, its temporal profile tracks with the
NO
+
x
signal. (b) Embedding CO
2
in the upper water layer results in a
smaller signal when compared to (a), as well as slower CO
2
+
signal that
tracks more closely with the water. Taken together, the traces indicate
that material removal occurs via two different pathways: mixing with
the hot fluid and scraping of fissure walls. . . . . . . . . . . . . . . . . . 106
5.4 Waterfall plot of averaged mass spectra (8-9 spots, 20 extractions) from
experiments using C
2
H
2
in configuration (3) where the water spacer
between C
2
H
2
and N
2
O
4
ranged from 100 to 0 ML. The main fragments
detected from C
2
H
2
are C
2
H
2
+
, C
2
H
+
, and C
2
+
(indicated by the red
arrow). Change in C
2
H
2
fragment signal is small as the H
2
O spacer
between N
2
O
4
and C
2
H
2
is decreased. However, the complete removal
of a water spacer results in a significant C
2
H
2
signal drop. . . . . . . . . 108
5.5 Overlaid FTIR spectra of a film initially grown at 100 K of 300 ML
of H
2
O deposited on 300 ML of D
2
O (red) that was then heated to
160 K for 1.5 minutes (black). The appearance of a central structure in
both peaks is indicative of HDO exchange occuring; there may be some
contribution from crystallization of the film, but the impact should be
small due to heating the film for <2 minutes. This shape change was
not detected at lower temperatures or when the sample was left to sit
for 20 minutes at 100 K. The negative-peak distortion of the OD peak
is due to background age. . . . . . . . . . . . . . . . . . . . . . . . . . . . 109
xii
5.6 Representative mass spectra showing HDO
+
formed from isotopic scramb-
ing during the ablation process. The spectra are averaged from the first
20 extractions yielding signal from first laser pulse temporal profiles
(1.5 mJ). Mass peaks are labeled in (a) for clarity. The film layer thick-
nesses for both samples (a) and (b) are: 300 ML H
2
O, 80 L N
2
O
4
, and
200 ML each of H
2
O and D
2
O (configurations depicted in cartoons). . . 111
5.7 Plot showing the inverse between R
I
and film thickness. R
I
values were
calculated for sandwich configurations like that depicted in Figure 5.6,
comparing D
2
O on top of a layer of H
2
O or underneath a layer of H
2
O.
The thicknesses reported on the x-axis refer to the thickness of the indi-
vidual D
2
O and H
2
O layers (i.e, the top layer of the sandwich is twice
as thick). There is a clear decrease in R
I
as the thickness of the H
2
O and
D
2
O layers increases. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112
6.1 Cartoon depiction of how gold nanoparticles can be used to form chan-
nels in ASW using 532 nm radiation, which is resonant with 10-80 nm
gold nanoparticles. (1) Nanoparticles are deposited on a bare MgO sur-
face, and particle density is determined using AFM. (2) A layer of N
2
O
4
,
followed by ASW is grown over the nanoparticles. (3) 532 nm radiation
is focused on the surface, heating the nanoparticles via surface plasmon
resonance to remove material. Once SPR heating is no longer able to
remove material, UV radiation (355/266 nm) is used to excite the N
2
O
4
layer. (4) This iterative process can form vertical channels in the ice that
are roughly the diameter of the gold nanoparticle and length of the film.
Once these channels are formed, they can be doped with molecules of
interest and the film can be further probed with spectroscopic techniques.123
6.2 AFM image of a MgO(100) surface wetted with one drop of a solution
containing 20 nm gold nanoparticles with a concentration of approx-
imately 6 10
10
nps/mL. This scan was taken from the center of the
surface, where the density of the particles was 10 nanoparticles per
square micron. In scans taken near the edge of the surface, the density
of nanoparticles was slightly higher and more aggregates of particles
were visible. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125
6.3 A temporal profile of 300 ML of ASW grown over 20 nm gold nanopar-
ticles after irradiation with 6.4 mJ of focused 532 nm light (average of
three spots). The majority of the signal is due to water and water frag-
ments, with some contribution from citrate fragments. . . . . . . . . . . 127
6.4 AFM images of 80 nm gold nanoparticles on the MgO substrate be-
fore (a) and after (b) laser ablation. The particle density of the freshly-
prepared film was approximately 1 particle per micron. The particles in
(b) have experienced several 532 nm pulses at a fluence of up to3 J/cm
2
.129
xiii
List of Tables
3.1 Timing settings for each pulse generator used in the timing set-up. Note:
all of the pulse generators are triggered at 100 kHz except for the pulse
generator triggering the optical chopper, which has a repetition rate of
100 Hz. Refer to labels from Figure 3.8. . . . . . . . . . . . . . . . . . . . . 51
4.1 70 eV ionization cross sections for detected molecules. All values ob-
tained from the NIST database [90], except for N
2
O
4
, which was com-
puted from the equation given in [73]. . . . . . . . . . . . . . . . . . . . . . 65
4.2 R
N
values for different laser energies at different extraction times. The
changing ratio indicates that different species are the result of the des-
orption process, such as collisions; a consistent value would indicate all
species are due to a fragmentation process that occurs in the ionization
region. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
xiv
Abstract
The effects of inserting energy in a buried stratum in amorphous solid water (ASW)
films were investigated using pulsed 266 nm radiation. Material ejected from irradi-
ated films was detected with time-of-flight mass spectrometry (TOFMS). A technique
was developed using N
2
O
4
in conjunction with focused UV radiation (1 J/cm
2
) that
enabled facile introduction of energy in a spatially selective way via an electronic transi-
tion of N
2
O
4
. A variety of experiments were carried out explore the structural changes
induced by the sudden addition of energy to ASW films, and an attempt was made to
characterize the nature of transport within and above the surface of the film.
Layered ASW/N
2
O
4
films, up to 2800 monolayers (ML) thick, were grown on a
MgO(100) substrate. All samples were grown at100 K under ultra-high vacuum con-
ditions to produce porous, high quality films. Once deposited, the films were were irra-
diated with 266 nm radiation that was generated as the fourth harmonic of a Nd:YAG
laser (10 ns pulses, reduced to 1 Hz) focused to a 0.3 mm beam waist. After a sin-
gle laser pulse incident on the film, the N
2
O
4
layer was converted to a hot fluid that
heated the surrounding material. Heating of the film competes with cooling by the
MgO(100) substrate, which acts as an efficient heat sink due to its high thermal con-
ductivity (250 W/mK). Therefore, extreme pressure and heat gradients exist within the
film upon radiation and the film cools quickly upon cessation of the laser pulse.
Despite fast cooling of the film, laser-heated N
2
O
4
fluid, along with water monomer,
was detected at times greater than 1 ms. This was due to a catastrophic structural
change triggered by the temperature and pressure gradients, which resulted in the
formation of fissures. The hot fluid of N
2
O
4
and its photoproducts escaped to ultra-high
vacuum through these fissures, scraping the walls and removing H
2
O molecules. Long
flight times were attributed to collisions occurring above the surface of the film due to
the high density of the escaping material and prevalence of fissures in the irradiated
area. Fissures proved to be robust, with spectra from subsequent pulses on the same
spot resembling spectra of exposed N
2
O
4
.
xv
This model will be explored further by implementing gold nanoparticles as fissure-
creators. Being able to form fissures of a known dimension and density within an ASW
film would allow for more careful analysis of this system and would be the first demon-
stration of induced morphological changes in ASW that are relatively well-defined. An
outline of this technique is presented, along with preliminary results.
xvi
Chapter 1
Introduction
Analyzing amorphous solid water interacting with the guest species N
2
O
4
is the
central focus of this dissertation. Amorphous solid water (ASW) as an experimen-
tal species is useful for modeling liquid water, glassy materials, and highly porous
compounds because of its disordered structure. It is also of great interest to the astro-
chemical community due to its prevalence in the interstellar medium and its unique
characteristics. Therefore, it is necessary to begin with a discussion of ASW properties,
as well as a brief summary of key experiments conducted with this mercurial substance.
1.1 Water: a Complex Molecule
Water is a research interest to scientists because of its omnipresence on Earth and
in the interstellar medium (ISM) [1–6]. It is involved in innumerable mechanisms in
life processes and acts as a near-universal solvent. Because of this, water is one of
the most investigated materials in science. Its simple molecular structure consisting of
two hydrogen atoms bonded to an oxygen atom belies its many unusual properties.
These include its high heat capacity, maximum density at 4

C, glass transition, and
many solid phases [3, 7–9]. The exact nature and origin of these properties are still not
well understood and require additional study. Building a cohesive picture is further
complicated by the fact that the intramolecular interactions of water are dominated by
hydrogen bonding, which is notoriously difficult to model [10–12]. Because of this,
1
investigation of water and aqueous systems continues to be a path for scientists to gain
insight into this ubiquitous molecule.
A unique feature of water is its many solid phases; there are at least 15 defined
crystalline states of water [9]. The most common on Earth is hexagonal water, HI,
which is thermodynamically and kinetically favorable at atmospheric pressure. It has
a regular structure that is characterized by a tetrahedral hydrogen-bonding network, as
shown in Figure 1.1. When ice crystallizes below 240 K it can have a different structure,
identified by König in 1944 as cubic ice (CI) [13]. Both hexagonal and cubic ice have a
tetrahedral first coordination shell; the difference arises from the stacking arrangement
of hexagonal bilayers [14].
Figure 1.1: A 3D depiction of hexagonal ice (HI), the most common form of crystalline
ice on earth. Each water molecule participates in 4 hydrogen bonds with neighboring
molecules, which forms a tetrahedral structure and characteristic long-range order.
Condensing water at temperatures of140 K gives rise to solid water with long
range disorder referred to as amorphous solid water (ASW) [15–19]. Due to the low
temperature needed for formation, on Earth it is only found in laboratories. However,
2
it is believed to be the most common form of water in the universe due to its preva-
lence in the ISM. Amorphous solid water is known to be a component of dust grains,
satellites, and planetary rings [4, 6]. Studying its characteristics and interactions with
other molecules is important for understanding the fundamental chemical and physical
reactions occurring in the universe. ASW is also useful in the laboratory environment
because its disordered structure resembles that of the liquid phase, but its molecular
immobility makes it more appealing for study [8]. It has even been used as a substrate
to investigate large molecules and biological compounds, such as RNA, in order to
freeze material in a hydrated solid phase to allow for electron microscopy of embedded
material [20]. However, much about its structure and molecular interactions remain
unknown, so continued study is needed to illuminate properties of this unusual com-
pound. It is necessary to provide an outline of ASW characteristics, which will be given
in the subsequent section.
1.1.1 Amorphous Solid Water
Amorphous solid water is a complex substance, with multiple morphologies. Fur-
thermore, structural changes can occur through temperature changes or other insertions
of energy into the ASW system. This section will consist of a brief summary of ASW
characteristics, as well as concepts and experiments done in this lab and others as this
information is relevant to the results reported later on in this dissertation.
Amorphous ice was first discovered in 1935 by Burton and Oliver in an X-ray diffrac-
tion study of water that had been vapor-deposited on a cryoplate held at 100 K [21]. The
absence of Bragg peaks indicated a long-range internal disorder that is representative
of an amorphous solid. In this way, ASW is analogous in form to liquid water [8]. Since
this initial report, multiple formation methods and unique amorphous phases have
been reported in the literature. The sample preparation method dictates the structure
of the ASW sample, so it is instructive to describe the most common techniques.
3
One of the most popular methods of ASW creation is vapor deposition of water at
temperatures below 140 K [3]. At these temperatures, kinetic factors dominate struc-
tural growth, despite a crystalline structure being more thermodynamically favorable.
Film growth at30 K will produce high-density amorphous solid water (HDA), while
higher temperatures will grow low-density amorphous solid water (LDA) [22, 23]. The
angle of deposition also affects the amorphous structure; the specific effects of this will
be discussed in a later section. Another preparation of ASW samples is the fast quench-
ing of small liquid water drops at cryogenic temperatures [24]. Samples prepared with
this method are sometimes referred to as hyperquenched glassy water (HGW). Finally,
hexagonal ice can undergo a transition to amorphous ice if it is subjected to pressures
in excess of 1 GPa [25, 26]. This process is referred to as pressure-induced amorphiza-
tion (PIA) and produces HDA. There are two mechanisms that have been suggested
for this transition: “melting” of ice to generate a vitreous solid [25], or mechanical
collapse of the HI structure itself [27]. HDA can also undergo another transition to
very high-density amorphous solid water (VHDA) if subjected to increased pressure
and annealing [28, 29]. The work presented in this dissertation deals only with vapor-
deposited ASW (i.e., LDA), but a discussion of the different morphologies is required
for context.
LDA, HDA, and VHDA are the three designations of ASW morphologies that are
well agreed on in the scientific community [3, 7, 9, 30, 31], though there is still some
debate about the distinction between the latter two structures [17]. LDA has an average
density similar to that of hexagonal ice, 0.94 g/cm
3
. In contrast, HDA has a density
of 1.17 g/cm
3
and VHDA has a density of 1.25 g/cm
3
(and even higher densities have
been reported) [3, 31]. Conversely, the hydrogen bond length increases with increasing
density [28, 29]. This seemingly incongruous trend is due to local structure of ASW.
Unique to amorphous water compared to other amorphous materials is that it has lo-
calized order: analysis of high frequency dynamics of ASW has revealed crystal-like
4
phonon excitations that indicate local structure is well-ordered [32]. However, it is im-
portant to emphasize the fact that this is indicative of local order only; amorphous solid
water is designated as such due to its long range disorder. Experiment and simulations
have found that the local structure in ASW adopts a tetrahedral arrangement similar
to that in crystalline ice [30, 33, 34]. Therefore, ASW roughly follows the same local
structure rules for ice published by Bernal and Fowler in 1933 [35]. A water molecule
will have a tetrahedral coordination shell with the surrounding 4 molecules. This net-
work is known as the Walrafen pentamer [34]. The structural differences between the
amorphous phases is due to the presence (or lack thereof) of molecules in the intersti-
tial sites. LDA has an “empty shell” between the first coordination shell and the second
shell. In comparison, HDA has one water molecule in the empty shell and VHDA
has two [23, 29, 34]. The presence of these additional molecules increases the density
while at the same time causes a distortion in the Walrafen pentamer that pushes the
molecules slightly apart, therefore increasing hydrogen bond length. In this way, these
contradictory trends of density and hydrogen bond length are reconciled.
ASW is a metastable form of solid water; simply put, this means that it forms due
to energy restrictions impeding rearrangement to a more thermodynamically favorable
(crystalline) structure. Because ASW is formed due to kinetic factors and not thermo-
dynamic stability, it is highly susceptible to morphological changes. It is possible to
change from one ASW morphology to another through external stress on the system.
For example, amorphous ice can experience relaxation to less strained morphologies.
This occurs in HDA formed from PIA, which will experience a slight change in den-
sity as it structurally relaxes upon annealing (heating) [17]. ASW can also experience
amorphous-amorphous phase transitions via annealing or pressure changes. Defined
as polymorphism, this is a property that is very rare in the chemical world [9, 17, 31].
In addition, temperature-programmed desorption (TPD) studies done by this lab and
others have found unique transformations that occur in the ASW as it is heated. When
ASW is subjected to heating, the structure can abruptly change at specific temperatures
5
[1, 36, 37]. The most significant change occurs around 140 K, when the structure under-
goes a rapid transition via irreversible crystallization. This transition acts as a marker
for the presence of ASW and can be consistently observed in TPD experiments as a
“bump” in the desorption peak [38]. This results from a change in vapor pressure due
to the decrease in free energy when transitioning from ASW to PCI. It is important to
note that the temperature of crystallization is not a set temperature as this is a kinetic
process, not a thermodynamic process. Therefore, the experimental conditions (such as
ramp rate) play a role in the crystallization temperature [36]. Above this temperature,
solid water exists as a mix of cubic ice and hexagonal ice known as polycrystalline ice
(PCI) [14]. Crystallization can be induced at even lower temperatures if the ASW film is
grown on a crystalline water substrate due to the removal of the nucleation activation
barrier [39].
ASW behavior near the crystallization temperature has been the focus of intense
study. Near this temperature it is believed that water undergoes a glass transition to a
highly viscous fluid [18, 40, 41]. A glass transition is usually determined experimen-
tally using differential scanning calorimetry (DSC). When a material goes from its glass
state to a supercooled liquid, there is a significant increase in heat capacity as degrees
of freedom become unfrozen. ASW exhibits such an increase at 136 K; however, the in-
crease is extremely small when compared to the transition behavior of classically glassy
materials [42]. This has resulted in much debate about the temperature of the glass
transition or whether is actually exists at all [7, 43]. Another investigation pathway into
the “glassy water” question is analysis of diffusion rates. Diffusion rates in crystalline
ice are extremely slow: below 10
18
cm
2
/s. TPD experiments by Smith and coworkers
with binary isotopic ices (D
2
O/H
2
O and H
2
16
O/H
2
18
O) found a million-fold increase
in the diffusion rate as evidenced by isotopic mixing products [44]. This is indicative of
a highly viscous liquid-like state that exists prior to crystallization and may point to a
glass transition.
6
1.1.2 Amorphous Solid Water and Guest Molecule Interactions
The interactions of guest molecules with an ASW matrix are of interest to astro-
chemists due to the prevalence of ASW mantles on dust grains in the ISM. Other atoms
and molecules adsorb and/or become trapped in the ASW coating and can undergo
chemical changes in an amorphous ice host. Because of its high porosity, it is facile for
molecules to become trapped in ASW and remain in the host matrix until energy is in-
troduced into the system. The degree of porosity is directly related to growth conditions
such as temperature and deposition angle. Because ASW is formed at low temperatures
that don’t allow for movement of molecules to achieve thermodynamic stability, vapor
deposition growth of ASW follows the simple model of “hit and stick”. When a wa-
ter molecule adsorbs to a surface that is below 155 K, it doesn’t have enough internal
energy to rearrange into a thermodynamically favorable structure, so it has a kinetically-
determined (random) orientation. More molecules will “hit and stick”, forming unique
filamentary structures and a porous film. Background deposition of water vapor on a
surface that is cooled below 140 K forms a LDA film (r = 0.8 g/cm
3
) with a porosity of
20% [45]. The group of Bruce Kay has shown that directed dosing with a molecular
beam can give a range of porosities depending on dosing angle [46, 47]. Samples grown
at normal incidence have a maximum density of 0.94 g/cm
3
and are considered to be
compact, nonporous ice. In contrast, films grown at extreme glancing angles (86

from
normal to the surface) have a maximum porosity of >80% and minimum density of 0.16
g/cm
3
. At incident angles greater than 45

the average pore size begins to increase
along with the increasing porosity. This is due to the growth of “filament” structures
that tilt in the direction of dosing and whose spacing increases with increasing incident
angle, an effect referred to as “shadowing”. As molecules approach the substrate at in-
creasingly oblique angles, filament growth will block a larger segment of the pathway.
This model was corroborated with ballistic deposition simulations [46, 48].
The highly porous nature of ASW significantly increases the surface area of ASW
with respect to crystalline ice, with as much as 10% of water molecules at pore surfaces.
7
[49, 50]. Also, the number of danglingOH groups is much larger for ASW, as evi-
denced by detection of an absorption peak at3700 cm
1
[51, 52]. This is in contrast
with hexagonal ice, where danglingOH groups are limited to the surface of the ma-
terial. The presence ofOH groups encourages uptake of adsorbing materials through
van der Waals interactions, and pores provide room for guest species [53].
Thermal processing of ASW with trapped molecules gives rise to explosive release
of material during crystallization at 155 K, dubbed the “molecular volcano” [38, 54,
55]. Ejection of guest species is possible at the amorphous-to-crystalline transition due
to a relaxation in the ice matrix. Any guest molecules that remain in the film after
crystallization remain trapped in the PCI structure until they can codesorb with water
at 180 K. Some experimental and theoretical studies have posited that the remaining
molecules are trapped in clathrate-hydrate cages within the crystal structure [49, 56].
Nucleation sites and crystal growth has been hotly discussed for the last 20 years,
with debate over whether initiation occurs at the substrate, in the bulk, or at the
ASW/vacuum interface [39, 57, 58]. Building a nucleation model may be further com-
plicated by different mechanisms for different morphologies [59]. Recent compelling
experiments by May and coworkers [60] have made a strong case for crystallization ini-
tializing at the surface of ASW films and then propagating down towards the substrate
when an ASW sample was heated at a constant rate. A monolayer of Ar was dosed on
the substrate surface (graphene) under a 300 ML film of ASW while a monolayer of O
2
was dosed at various levels in the film; TPD experiments were then conducted on the
ternary films. The desorption temperature of Ar remained consistent for each film, but
the desorption temperature for the O
2
layer increased as the layer was moved deeper
in the ASW film. This indicates that cracks initiate at the vacuum interface, propagate
downwards and open a pathway for sequential release of the O
2
layer and then the Ar
layer. As such, the molecular volcano peak is due to the top-down propagation of cracks
during crystallization that initialize at the ASW/vacuum interface. Until this process
occurs, many guest species are effectively contained within the water host matrix. This
8
is significant because molecules can stay in the film above their characteristic desorp-
tion temperatures. This has important implications for chemistry in the ISM that will be
discussed in the next section of this chapter. Guest molecules may also escape through
stress-induced fracturing of the ASW film, in a “bottom-up” mechanism [37, 61]. This
occurs when no direct pathway to the surface is available to species that become mobile
at low temperatures. With enough heat, the pressure of the trapped species can frac-
ture surrounding ASW and be violently ejected from the matrix. Experiments done by
Bar-Nun and associates with cm-thick samples of ASW also saw the ejection of large ice
“needles” along with the guest species [37].
A different method of probing ASW with and without guest molecules is using laser
radiation to induce photodesorption. One method, originally reported by Livingston
and coworkers [62] and put to use in this lab [63], excites the OH stetch in condensed
ASW with l = 2.94 μm. This heats the water, resulting in desorption of material
from a substrate. The process is very selective to the spatial regime being irradiated
due to the limited diffusion of heat during a ns-scale laser pulse and can be used
for depth-profiling of an ASW film. Photodissociation of water itself to form H and
OH can be achieved with wavelengths <160 nm. This reaction is believed to be an
important source of H
2
in the universe [64]. A third mechanism for laser-induced
desorption in ASW films is excitation of a guest molecule. This method was used
effectively by Thrower and coworkers [65–67], who investigated benzene-doped ASW
samples. They were able to instigate an indirect-adsorbate-mediated desorption process
with 250 nm radiation. Even though H
2
O has negligible absorbance in this region,
benzene was able to efficiently transfer energy to the surrounding ASW film and trigger
water codesorption. Such a result implies that a wider range of wavelengths can initiate
physical and chemical change in ASW found in the universe.
9
1.2 Amorphous Solid Water in the Interstellar Medium
The chemistry of the universe is not limited to the surfaces of planets or the interior
of stars: much of it occurs in the space between, which is called the interstellar medium
(ISM). More than 140 unique molecules have been identified to exist in the interstellar
medium by astronomers [1, 2]. This variety of compounds, ranging from diatomics to
small organic molecules, seems in contradiction with the low temperatures (<20 K in
dense molecular clouds) and molecular density ( 10
4
atoms/cm
3
) observed in space
[4]. Furthermore, chemistry in the interstellar medium is more complex than can be
explained solely from gas-phase processes. The diversity of molecules, which includes
organic material and large carbon molecules, is thought to be due to the presence of
silicate and carbonaceous “grains” [2, 4, 68, 69]. These grains provide surfaces on which
material can interact, and may have been the source of prebiotic organic molecules on
Earth [70]. Such grains also play a role in the formation of water in the ISM, which then
collects as a mantle on the surface of the grain [23].
Hydrogen and oxygen are among the most ubiquitous atoms in the interstellar
medium (first and third most abundant, respectively, with a ratio of 1 : 5 10
4
[2]),
providing ample material for water formation. Specifically, water molecules are gener-
ated in dense gas and dust clouds where reactions are facilitated by grains [4, 69]. The
gas-phase mechanism is believed to begin with the creation of molecular hydrogen on
a dust grain surface. This sublimates quickly and can be ionized by cosmic-ray pro-
tons; the H
2
+
ion can then react with another molecular hydrogen to form the triatomic
hydrogen ion. This can then react with oxygen to form either OH
+
or H
2
O
+
, and the
chain continues. In contrast, a solid phase path can occur via diffusive reactions, called
the Langmuir-Hinshelwood (LH) mechanism, on the surfaces of dust grains [2, 4, 69].
The exact mechanism and starting materials are debated, but the end product is water
ice accretion on the surface of the grain. It is believed that ASW makes up to 70% of
surface material on grains in the universe and is mostly high density in structure [3, 23,
71]. While the icy mantle is growing, other molecules can form on the grain through
10
similar surface reactions and get trapped in the ice. Species believed to be a part of these
icy mantles include CO, CO
2
, methanol, and polycyclic aromatic hydrocarbons (PAHs)
[2, 66, 69]. Over the extremely long lifetime of molecular clouds (10
6
-10
8
years), ices
coating grains are subjected to many different external stresses. For example, dust is
bombarded with energetic particles and radiation (e.g. Lyman-a, UV , etc.) from celestial
bodies. Such interactions can cause morphological changes, chemical reactions, desorp-
tion, or other structural transformations and chemical proccessing [1, 6]. Laboratory
experiments have tried to understand these reactions by studying ASW in model sys-
tems under UHV conditions in an effort to mimic the low temperatures and pressures
of space.
ASW films can also facilitate reactions between trapped species, as seen in exper-
iments done by Fesneau and coworkers [72]. When acetone and ammonia were co-
condensed on a gold-plated surface at 20 K, no reaction between the two species was
detected by Fourier transform infrared spectroscopy (FTIR) or mass spectrometry (MS)
detection as both product desorbed at their typical desorption temperatures. Adding
water to the mixture being condensed resulted in the formation of an aminoalcohol
upon heating to 150 K, which was robust even above the water desorption temperature
of 185 K. The ASW host matrix traps the reactants above their desorption tempera-
tures, allowing for the reaction to occur when sufficient thermal energy is available. In
this way, ASW can act as a medium for the formation of more complex molecules and
promote the chemical diversity observed in the ISM.
On the macroscale, ASW covers the surface of certain comets and moons in the
outer solar system and beyond [4, 37]. A case that has attracted scientific interest is
that of Enceladus, a moon of Saturn. Enceladus has a thick crust of ASW covering its
entire surface. During the early phases of the Cassini mission, images were captured
that showed “eruptions” of material from the icy surface, similar to volcanoes. These
spouts of water, ammonia, and other small molecules have been studied to ascertain a
mechanism for these volcanoes and to try and understand what could be happening
11
beneath the ASW crust. The current theory involves gravitational stress due to the orbit
of Enceladus that generates enough heat for a liquid ocean [73–75]. While the planetary
scale is vastly different than that in the laboratory, it prompts scientific interest in how
heating of buried strata affects ASW.
1.3 NO
2
/N
2
O
4
Previous experiments done in this lab have focused on introducing energy to thin
films via the water OH stretch. Subsequent analysis focused on how that energy dis-
rupts the water matrix and was transferred to a dopant species. This method resulted
in an irreversible change in the film triggered by a single IR laser pulse that was able to
liberate water monomer and small clusters from the solid film. Furthermore, there was
preferential release of CO
2
, the dopant, even from the depths of the film because of the
relaxation of the host matrix [63].
In contrast, the following dissertation outlines the study of an ASW/N
2
O
4
system
where the energy is introduced to the film via exciting an electronic state of N
2
O
4
and examining how that energy changes the system. Nitrogen dioxide is a molecule
of interest for a few reasons. First of all, it is part of the NO
x
family of chemicals
that are a key component of pollution from car exhaust and industrial processes. It
can interact with solid water in the troposphere and undergo various changes when
subjected to radiation (e.g. generating O atoms that react with O
2
to form ozone), and
understanding these reactions is extremely important for predicting the impact NO
2
has on the environment [79]. Furthermore, nitrogen compounds have been found in
the ISM [80, 81] and may play a role in chemical processing. Most important for our
work, it acts as a easy-to-use dopant for introducing energy into a film.
When NO
2
is condensed between 20 K and 140 K, it exclusively forms a stable dimer
of the form O
2
NNO
2
without any remaining monomer [76]. N
2
O
4
readily absorbs in
the UV via an electronic transition when adsorbed on a substrate [82, 83], making it an
attractive candidate for condensed-phase introduction of heat via laser radiation. As
12
Figure 1.2: (a) Ball-and-stick model of N
2
O
4
. The NN bond is quite weak (around
0.5 eV) [76, 77]. (b) The bcc structure is the dominant structure of crystalline N
2
O
4
at low
temperatures. Defects, including presence of the NO
2
and unstable isomer ONONO
2
,
start to appear below 20 K [78].
13
the LUMO of N
2
O
4
is the NN antibonding orbital, the weak NN bond between the
NO
2
molecules (0.5 eV [79, 84]) can break apart easily when irradiated (ball-and stick
model shown in Figure 1.2a). This forms two excited NO
2
molecules [82, 83]. The NO
2
can then be used to further excite a film with additional UV photons. In this way, a
multistep film excitation process is easily attained.
The monolayer structure of N
2
O
4
when condensed on a substrate is still under de-
bate as to whether it is ordered [85] or disordered [86]. The bulk structure when con-
densed from the gas phase has been found to be body-centered cubic and belonging to
the space group T
5
h
(Im3). The cubic structure is stable down to temperatures of 20 K;
below this temperature, amorphous structures, unstable isomers, and the monomer
may be present [77, 87]. While the most stable molecular structure of condensed N
2
O
4
is that shown in Figure 1.2b, various isomers have been detected by several groups.
Bolduan and coworkers [78] found evidence of both the O
2
NNO
2
and O
2
NONO iso-
mers of N
2
O
4
when NO
2
was deposited on a copper substrate at 15 K, whereas only the
O
2
NNO
2
dimer was detected in films grown at 80 K. When condensed on ASW films,
N
2
O
4
physisorbs to the H
2
O interface in a disordered monolayer without any chemical
changes. This is apparent in the low binding energy of 0.40 eV between the N
2
O
4
and
H
2
O [87].
Photodissociation experiments conducted by Dixon-Warren and coworkers of ad-
sorbed NO
2
conducted with UV radiation (193-351 nm) have found two distinct des-
orption processes that are substrate-dependent [88]. When N
2
O
4
is photodesorbed from
a metal substrate [in this case Pd(111)], there is a significant “hot” peak that is due to
a charge-transfer mechanism from substrate to desorbing molecule. In contrast, pho-
todesorption of N
2
O
4
multilayers from an insulator (LiF) results in a dominant slower
pathway that is due to direct desorption of the species.
One drawback to spectroscopic investigation of N
2
O
4
is its IR absorption spectrum.
The IR-active modes of solid N
2
O
4
are outside of the range of certain detectors (e.g.,
InSb), with the highest energy fundamental peak at 1257 cm
1
[76]. There are a few
14
combination bands detectable above 1850 cm
1
, but they are extremely weak or can be
obscured by water absorption peaks. This can make N
2
O
4
difficult to examine with
certain FTIR set-ups; this will be discussed more in depth in following chapters.
1.4 Thesis Overview
The research presented in this thesis explores molecular transport using UV radi-
ation in combination with the guest molecule of N
2
O
4
to introduce energy to ASW
films. A time-of-flight (TOF) mass spectrometer and Fourier transform infrared (FTIR)
spectrometer were the main detection tools used. This work presents the following:
N
2
O
4
as an effective guest molecule to introduce energy to ASW in a buried stra-
tum.
A model that qualitatively describes molecular transport and the induction of
fissures in ASW films due to a superheated guest molecule layer.
A direction for future work that focuses on fissures with well-determined shape
and size.
Chapter 2 details the theory behind the experimental methods used, while Chapter
3 describes the specific apparatus used for these experiments. A variety of results
characterizing the N
2
O
4
/ASW system are presented in Chapter 4. Chapter 5 discusses
the introduction of other guest molecules, with a particular focus on D
2
O and isotope
exchange. Finally, the future direction for these experiments along with preliminary
results is covered in Chapter 6.
15
Chapter References
[1] Burke, D.; Brown, W. Phys. Chem. Chem. Phys. 2010, 12, 5947–5969.
[2] Hama, T.; Watanabe, N. Chem. Rev. 2013, 113, 8783–8839.
[3] Angell, C. Annu. Rev. Phys. Chem. 2004, 55, 559–583.
[4] Van Dishoeck, E.; Herbst, E.; Neufeld, D. Chem. Rev. 2013, 113, 9043–9085.
[5] Kang, H. Acc. Chem. Res. 2005, 38, 893–900.
[6] Baragiola, R. Planet. Space Sci. 2003, 51, 953–961.
[7] Debenedetti, P . J. Phys.: Condens. Matter 2003, 15, R1669–R1726.
[8] Angell, C. Science 2008, 319, 582–587.
[9] Malenkov, G. J. Phys: Condens. Matter 2009, 21, 283101.
[10] Rice, S.; Sceats, M. J. Phys. Chem. 1981, 85, 1108–1119.
[11] Vega, C.; Abascal, J.; Conde, M.; Aragones, J. FaradayDiscuss. 2009, 141, 251–276.
[12] Gandyopadhyay, D.; Mohan, S.; Ghosh, S.; Choudhury, N. J. Phys. Chem. B 2013,
117, 8831–8843.
[13] König, H. Z. Kristallogr. 1944, 105, 279–286.
[14] Moore, E.; Molinero, V . Phys. Chem. Chem. Phys. 2011, 13, 20008–20016.
[15] Poole, P .; Essmann, U.; Sciortino, F.; Stanley, H. Phys. Rev. E 1993, 48, 4605–4610.
[16] Johari, G.; Andersson, O. Thermochim. Acta 2007, 461, 14–43.
[17] Loerting, T.; Winkel, K.; Seidl, M.; Bauer, M.; Mitterdorfer, C.; Handle, P .; Salz-
mann, C.; Mayer, E.; Finney, J.; Bowron, D. Phys. Chem. Chem. Phys. 2011, 13,
8737–8794.
[18] Limmer, D.; Chandler, D. PNAS 2014, 111, 9413–9418.
[19] Fanetti, S.; Pagliai, M.; Citroni, M.; Lapini, A.; Scandolo, S.; Righini, R.; Bini, R.
J. Phys. Chem. Lett. 2014, 5, 3804–3809.
[20] Chiu, W. Ann. Rev. Biophys. Biophys. Chem. 1986, 15, 237–257.
[21] Burton, E.; Oliver, W. Nature 1935, 135, 505–506.
[22] Wenzel, J.; Linderstrom-Lang, C.; Rice, S. Science 1975, 187, 428–430.
16
[23] Jenniskens, P .; Blake, D.; Wilson, M.; Pohorille, A.Astrophys.J. 1995,455, 389–401.
[24] Brüggeller, P .; Mayer, E. Nature 1980, 288, 569–571.
[25] Mishima, O.; Calvert, L.; Whalley, E. Nature 1984, 310, 393–395.
[26] Giovambattista, N.; Stanley, H.; Sciortino, F. Phys. Rev. Lett. 2005, 94, 107803.
[27] Tse, J. J. Chem. Phys. 1992, 1992, 5482–5487.
[28] Loerting, T.; Salzmann, C.; Kohl, I.; Mayer, E.; Hallbrucker, A. Phys. Chem. Chem.
Phys. 2001, 3, 5355–5357.
[29] Finney, J.; Bowron, D.; Soper, A.; Loerting, T.; Mayer, E.; Hallbrucker, A. Phys.
Rev. Lett. 2002, 89, 205503(4).
[30] Brovchenko, I.; Oleinikova, A. J. Chem. Phys. 2006, 124, 164505.
[31] Giobambattista, N.; Amann-Winkel, K.; Loerting, T. In Liquid Polymorphism: Ad-
vances in Chemical Physics, Volume 152, Stanely, H., Ed.; John Wiley & Sons, Inc.:
Somerset, NJ, 2013.
[32] Schober, H.; Koza, M.; Tölle, A.; Masciovecchio, C.; Sette, F.; Fujara, F. Phys. Rev.
Lett. 2000, 85, 4100–4103.
[33] Essmann, U.; Geiger, A. J. Chem. Phys. 1995, 103, 4678–4692.
[34] Loerting, T.; Giovambattista J. Phys. Condens. Matter 2006, 18, R919–R977.
[35] Bernal, J.; Fowler, R. J. Chem. Phys. 1933, 1, 515–548.
[36] Smith, R.; Petrik, N.; Kimmel, G.; Kay, B. Acc. Chem. Res. 2012, 45, 33–42.
[37] Laufer, D.; Bar-Nun, A.; Igal, P .; Jacovi, R. Icarus 2013, 222, 73–80.
[38] Smith, R.; Huang, C.; Wong, E.; Kay, B. Phys. Rev. Lett. 1997, 79, 909–912.
[39] Dohnálek, Z.; Ciolli, R.; Kimmel, G.; Stevenson, K.; Smith, R.; Kay, B. J. Chem.
Phys. 1999, 110, 5489–5492.
[40] Mishima, O.; Stanley, H. Nature 1998, 396, 329–335.
[41] Smith, R.; Kay, B. Nature 1999, 398, 788–791.
[42] Ediger, M.; Angell, C.; Nagel, S. J. Phys. Chem. 1996, 100, 13200–13212.
[43] Shephard, J.; Evans, J.; Salzmann, C. J. Phys. Chem. Lett. 2013, 4, 3672–3676.
[44] Smith, R.; Huang, C.; Kay, B. J. Phys. Chem. B. 1997, 101, 6123–6126.
17
[45] Raut, U.; Famá, M.; Teolis, B.; Baragiola, R. J. Chem. Phys. 2007, 127, 204713.
[46] Dohnálek, Z.; Kimmel, G.; Ayotte, P .; Smith, R.; Kay, B. J. Chem. Phys. 2003, 118,
364–372.
[47] Kimmel, G.; Stevenson, K.; Dohnálek, Z.; Smith, R.; Kay, B. J. Chem. Phys. 2001,
114, 5284–5294.
[48] Kimmel, G.; Stevenson, K.; Dohnálek, Z.; Smith, R.; Kay, B. J. Chem. Phys. 2001,
114, 5295–5303.
[49] Devlin, J. J. Geophys. Res. 2001, 106, 33333–33349.
[50] Maté, B.; Rodríguez-Lazcano, Y.; Herrero, V . Phys. Chem. Chem. Phys. 2012, 14,
10595–10602.
[51] Mitlin, S.; Leung, K. J. Phys. Chem. B 2002, 106, 6234–6247.
[52] Cholette, F.; Zubkov, T.; Smith, R.; Dohnálek, Z.; Kay, B.; Ayotte, P . J. Phys. Chem.
B 2009, 113, 4131–4140.
[53] Ayotte, P .; Smith, R.; Stevenson, K.; Dohnálek, Z.; Kimmel, G.; Kay, B. J. Geophys.
Res. 2001, 106, 33387–33392.
[54] Kumi, G.; Malyk, S.; Hawkins, S.; Reisler, H.; Wittig, C. J. Phys. Chem. A 2006,
110, 2097–2105.
[55] May, R.; Smith, R.; Kay, B. J. Phys. Chem. Lett. 2012, 3, 327–331.
[56] Faizullin, M.; Vinogradov, A.; Koverda, V . Chem. Eng. Sci. 2015, 000–000.
[57] Kondo, T.; Kato, H.; Bonn, M.; Kawai, M. J. Chem. Phys. 2007, 127, 094703.
[58] Backus, E.; Grecea, M.; Kleyn, A.; Bonn, M. Phys. Rev. Lett. 2004, 92, 236101.
[59] Safarik, D.; Mullins, C. J. Chem. Phys. 2004, 121, 6003–6010.
[60] May, R.; Smith, R.; Kay, B. J. Chem. Phys. 2013, 138, 104501.
[61] May, R.; Smith, R.; Kay, B. J. Chem. Phys. 2013, 138, 104502.
[62] Livingston, F.; Smith, J.; George, S. Anal. Chem. 2000, 72, 5590–5599.
[63] Rebolledo–Mayoral, O.; Stomberg, J.; McKean, S.; Reisler, H.; Wittig, C. J. Phys.
Chem. C 2012, 116, 563–569.
[64] Yabushita, A.; Kanda, D.; Kawanaka, N.; Kawasaki, M.; Ashfold, M. J. Chem.
Phys. 2006, 125, 133406.
18
[65] Thrower, J.; Burke, D.; Collings, M.; Dawes, A.; Holtom, P .; Jamme, F.; Kendall, P .;
Brown, W.; Clark, I.; Fraser, H.; McCoustra, M.; Mason, N.; Parker, A. Astrophys.
J. 2008, 673, 1233–1239.
[66] Thrower, J.; Collings, M.; McCoustra, M.; Burke, D.; Brown, W.; Dawes, A.;
Holtom, P .; Kendall, P .; Mason, N.; Jamme, F.; Fraser, H.; Clark, I.; Parker, A.
J. Vac. Sci. Tech. 2008, 26, 919–924.
[67] Thrower, J.; Abdulgalil, A.; Collings, M.; McCoustra, M.; Burke, D.; Brown, W.;
Dawes, A.; Holtom, P .; Kendall, P .; Mason, N.; Jamme, F.; Fraser, H.; Rutten, F. J.
Vac. Sci. Technol., A 2010, 28, 799–806.
[68] Draine, B. Annu. Rev. Astron. Astrophys. 2003, 41, 241–289.
[69] Herbst, E. AIP Con. Proceed. 2013, 1543, 15–30.
[70] Chyba, C.; Sagan, C. Nature 1992, 355, 125–132.
[71] Tielens, A.; Hagen, W. Astron. Astrophys. 1982, 114, 245–260.
[72] Fresneau, A.; Danger, G.; Rimola, A.; Theule, P .; Duvernay, F.; Chiavassa, T. MN-
RAS 2014, 443, 2991–3000.
[73] Hansen, C.; Shemansky, D.; Esposito, L.; Stewart, A.; Lewis, B.; Colwell, J.; Hen-
drix, A.; West R.A., W. J.; J.H. Teolis, B.; Magee, B. Geophys. Res. Lett. 2011, 38,
L11202.
[74] Porco, C.; DiNino, D.; Nimmo, F. Astron. J. 2014, 148, 45.
[75] Hurford, A.; Helfenstein, P .; Hoppa, G.; Greenberg, R.; Bills, B. Nature 2007, 447,
292–294.
[76] Andrews, B.; Anderson, A. J. Chem. Phys. 1981, 74, 1534–1537.
[77] Kvick, Å.; McMullan, R.; Newton, M. J. Chem. Phys. 1982, 78, 3754–3761.
[78] Bolduan, F.; Jodl, H.; Loewenschuss, A. J. Res. Nat. Bur. Stand. A 1976, 80A, 143–
166.
[79] Bass A.M.and Ledford Jr., A.; Laufer, A. J. Res. Nat. Bur. Stand. A 1976, 80A, 143–
166.
[80] McGonagle, D.; Ziurys, L.; Irvine, W.; Minh, Y. Astrophys. J. 1990, 359, 121–124.
[81] Halfen, D.; Apponi, A.; Ziurys, L. Astrophys. J. 2001, 561, 244–253.
[82] Bach, W.; Breuer, H. Faraday Discuss. Chem. S. 1974, 58, 237–243.
19
[83] Kawasaki, M.; Kasatani, K.; Hiroyasu, S.; Shinohara, H.; Nobuyuki, N. Chem.
Phys. 1983, 78, 65–74.
[84] Hisatsune, I. J. Phys. Chem. 1961, 65, 2249–2253.
[85] Rieley, H.; Colby, D.; McMurry, D.; Reeman, S. J. Phys. Chem. B 1997, 101, 4982–
4991.
[86] Yabushita, A.; Inoue, Y.; Senta, T.; Kawasaki, M.; Sato, S. J. Chem. Phys. A 2004,
108, 438–446.
[87] Rieley, H.; McMurry, D.; Haq, S. Faraday Trans. 1996, 92, 933–939.
[88] Dixon-Warren, S. J.; Jackson, R.; Polanyi, J.; Rieley, H.; Shapter, J.; Weiss, H. J.
Phys. Chem. 1992, 96, 10983–10994.
20
Chapter 2
Experimental Methods
This chapter gives a brief summary of concepts relevant to the experiments. The
information presented here will give context to results in later chapters and summarize
the theory behind key experimental apparatus detailed in Chapter 3.
2.1 Time-of-Flight Mass Spectrometry
Mass spectrometry (MS) was first implemented using sector mass spectrometers.
These required careful alignment and machining to ensure that the material of inter-
est was directed along the correct path. Size restrictions existed because of the use of
magnetic fields. By exploiting the Lorentz force law, material within a magnetic (or
electric) field is deflected different amounts based on their mass-to-charge ratio, m/q.
Mass spectrometers using electric fields were initially limited by design flaws that lim-
ited mass resolution. In 1955, Wiley and McLaren described an improved two-stage
time-of-flight system that was able to collect mass spectra quickly and efficiently. This
allowed for increased mass resolution and detection of a range of masses in a short
time, which paved the way for its widespread use in laboratories [1]. Because of its ease
of use, mass spectrometry has been utilized in a variety of applications: matrix-assisted
laser desoprtion-ionization (MALDI), chromatography, electrospray ionization systems,
fragment characterization, etc. [2].
21
Time-of-flight (TOF) MS exploits the relationship between mass and energy such
that molecules of different masses can be separated out and detected. An example of a
two-stage system is diagrammed in Figure 2.1.
Figure 2.1: A schematic of a linear two-stage Wiley-McLaren TOFMS system. Material
enters the region between repeller and extractor plates initially kept at the same voltage
(A). If necessary, the material is ionized (this can be done by electron impact ionization,
photoionization, etc). Subsequently, the voltage is dropped on the extractor plate and
ionized material is extracted and further accelerated by a grounded third plate into a
field-free drift region (C). Ionized material of different masses can be differentiated by
flight times, with more massive ions arriving after lighter ions (relative mass is indicated
by size in the diagram).
Molecules of interest enter a region between two plates, referred to as a “repeller”
and an “extractor.” Because of the nature of TOFMS, detected material must be ionized:
either material enters the the extraction region as ions, or it is ionized in that region;
common ionization techniques include photoionization and electron impact bombard-
ment. Then, an electric field accelerates ions to a field-free region. The accelerating
pulse ensures that ions leave the electric-field region with the same kinetic energy. This
results in the ions having velocities that are a function of their charge (q) and mass
(m). Given sufficient drift time, the ions separate according to the relationship m/q
and are collected by a detector. This relationship between mass and drift time can be
22
defined mathematically for the two-stage TOFMS system depicted in Figure 2.1, which
is a schematic of the approach used in this dissertation. The advantage of a two-stage
system will be addressed in Section 2.1.2.
Particles enter the ionization region with an initial kinetic energy, U
0
. There, ma-
terial is ionized via an ionization process, which is electron impact ionization for our
apparatus. The ions gain kinetic energy proportional to the distance it travels in an
electric field. As such, we can state the total kinetic energy of a charged particle in a
two-stage TOF system as
U
tot
= U
0
+qsE
s
+qdE
d
(2.1)
where s and d are the path lengths in the ionization and acceleration regions (A and
B in Figure 2.1, respectively) and E
s
and E
d
are the magnitudes of the electric fields in
those regions. There is no term representing the drift region (C in Figure 2.1) because
it is field free, and thus does not contribute kinetic energy. Time can be introduced by
using
K.E. =
1
2
mv
2
=
1
2
m

d
t

2
(2.2)
where m is the mass of the particle, v is the velocity of the particle, and t is the flight
time in a specific region. In this way, the flight time for a charged particle in TOFMS is
defined as
T(U
0
,s) = T
s
+T
d
+T
D
(2.3)
Therefore, each ion with a unique m/q will have a characteristic flight time, where each
flight time term can be defined with respect to its mass, travel distance, and kinetic
energy.
Because of the relationship between time and mass inherent in Equation 2.2, each
component in Equation 2.3 will have a
p
m/q term. Therefore, the total flight time of
a particle is proportional to the square root of its mass-to-charge ratio, with constants
dependent on the factors of a specific system. Combining this relationship with param-
eters of a spectific TOF instrument, a mass spectrum can be measured quickly and easily
with high resolution. Furthermore, it is possible to detect a range of masses all at once
23
instead of selecting one mass at a time for detection as with other instruments, such
as quadrupole mass analyzers. A system can be set up to collect entire mass spectra
in very short time windows with fast triggering of the ionization-extraction electronics.
The work detailed in this dissertation employs this method in order to investigate a
plume of material as it evolves in time; the specifics of the arrangement are described
in Chapter 3, Section 3.5.
2.1.1 Jacobian
When converting a time-of-flight spectrum to a mass spectrum, it is necessary to use
a Jacobian transformation. This is because when solving for mass in the aforementioned
relationship, there is a t
2
term. Thus, when converting from time space to mass space, a
Jacobian transformation is required to ensure that the peak areas remain the same. The
equation used to get a good fit with data in the time domain is as follows:
m = At
2
+Bt+C (2.4)
where A, B, C are fitting parameters. The B and C terms are very small, indicating
that deviations from the theoretical relationship between t and
p
m are minor. While
it may appear that a direct conversion from time to mass is possible, it is important
to note that data collected is signal per unit time [3]. If a direct conversion was carried
out, the intervals in the time-of-flight spectrum would not be evenly spaced in the mass
spectrum. Thus, we employ the following relationship
f(t)dt = g(m)dm (2.5)
By using this relationship in conjuction with Equation 2.4, we can determine the term
needed to perform a Jacobian transformation.
dm
dt
= 2At+B (2.6)
g(m) = f(t)
dt
dm
= f(t)
1
2At+B
(2.7)
24
2.1.2 Resolution
The above relationship between flight time and m/q is true for an idealized “point”
ionization and assumes that the material has the same initial kinetic energy while trav-
eling in the same direction. In an ideal TOFMS, there is an ionization “region” where
only material ionized at the center with no other deviations will follow Equation 2.3
exactly. Our system further confounds the ideal scenario as molecules desorb with var-
ious trajectories and kinetic energies. Because of this, there is a range of possible kinetic
energies for each mass, which will result in ions of the same m/q ratio arriving at the
detector at slightly different times. The distribution of flight times results in a broaden-
ing of mass peaks which limits the mass resolution. The three factors affecting TOFMS
resolution are due to time, space, and kinetic energy distributions of the ionized mate-
rial [2]. Each of these come into play in our experimental approach, and as such will be
described briefly herein.
Time distributions occur when ions of the same mass form at different times with the
same kinetic energy. This can happen because material in a Wiley-McLaren TOFMS is
ionized for a specific amount of time, t
i
, and desorbing material can enter the ionization
region at any time during this window. Because of this, ionized particles will travel
through the drift region with a discrete time difference. Mass resolution is defined by
the following equation:
Dm
m
=
2Dt
t
(2.8)
This relationship can be exploited to increase mass resolution by increasing the flight
time, which is done with low accelerating voltage and/or long drift path lengths.
The afore-mentioned ionization region contributes to spatial distribution of ions.
For instance, given a rectangular ionization region as shown in Figure 2.2, ions will be
extracted from the entire width of the rectangle, l. Based on Equation 2.1, material in
the ionization region that is closest to the extractor grid will gain less kinetic energy
than material located at the other extreme of the ionization region. Clearly this spread
of kinetic energies due to the ionization region width, l, will give a range of flight times.
25
Figure 2.2: Two particles of the same mass, A and B, are ionized in a finite ionization
region with a width ofDl. The flight path can therefore differ by a range ofDl/2; in
this case, particle A will have a shorter flight path than particle B. However, particle B
will gain more kinetic energy than particle A. This means that there will be a point in
space where particle B will pass particle A.
However, the difference in flight path distances must be taken into account. Ions leaving
the extraction region with the least kinetic energy will have a flight path of length D
0
;
ions leaving the extraction region with the most kinetic energy will have a flight path
of length D
0
+l. Since ions with more kinetic energy have a longer flight path, their
flight time will be increased accordingly. Furthermore, this results in a point along the
flight path where more energetic ions will pass less energetic ions; this is referred to
as space focusing. In a well-designed system, this point (referred to as the space focus
plane, which is independent of mass) occurs at the detector so as to minimize the effect
of a finite ionization region. However, the location of this plane in a one-stage system
is fixed at a short distance from the extraction region, resulting in a short drift time
and poor mass resolution. A two-stage system addresses this issue by adding a second
extraction region, which allows for the location of the space focus plane to be chosen
by careful selection of extraction and repeller plate voltages [2, 4].
Another way that a range of kinetic energies can be introduced is via the initial
velocity distribution of particles in the ionization region. This is a contributing factor
26
Figure 2.3: A particle trajectory, depicted as a blue circle, with initial kinetic energy U
0
.
Only the component in the direction of extraction will affect the overall kinetic energy
of the extracted particle, U
tot
.
in the system discussed by this dissertation as material is leaving a surface with a
wide range of veolcities. When a particle enters the ionization region with a nonzero
velocity, the kinetic energy component in the direction of the direction of extraction will
contribute to the total kinetic energy of the particle; this is diagrammed in Figure 2.3.
Thus, the total kinetic energy of the particle when it leaves the ionization region will be
a sum:
U
tot
= U
ext
+U
0x
(2.9)
Instead of kinetic energy only being imparted by the extraction process, U
ext
, the initial
kinetic energy component along the extraction trajectory, U
0x
, also contributes. The
variation in U
0x
will result in a range of U
tot
values. This issue is hard to address in
a linear TOFMS as increasing drift length increases peak width. Ions can also have
trajectories in the opposite direction of the extraction path, which means that while
they will be accelerated to the same kinetic energy as other ions of the same mass, they
will have a longer flight time because of their turn-around time in the extraction region.
27
An additional issue that is particulary relevant to this work is that of velocity dis-
tributions affecting velocity resolution. The approach used for experiments described
herein allows for multiple complete TOF spectra to be taken in quick succession; this
process will be discussed in more detail in the following chapter. A rough approx-
imation of material velocity can be made by using the extraction time (which is in
increments of 10 μs) in conjunction with the 2.5 cm flight path from the substrate to the
ionization region. However, there are two obfuscating factors: the finite dimensions of
the ionization region (Figure 2.2) and the temporal length of the ionization process. The
former case will be discussed first. The ionization region in our TOFMS has a length of
1.2 cm along the path from the substrate to the ionization region. Therefore, the flight
path of material that is detected in a single extraction can range from 1.9-3.1 cm. For the
10 μs extraction, the velocity spread is 1200 m/s. The second confounding factor is the
process of ionization. It is not instantaneous: our apparatus is set to ionize material for
7 μs. As such, the material in the 10 μs extraction has a range of flight times (3-10 μs)
that result in a velocity spread of5800 m/s assuming an average flight path of 2.5 cm.
Thus, calculated velocities of material arriving in each extraction of our pulsed TOFMS
system are only a rough approximation.
2.2 Laser-Induced Desorption
Using lasers to remove material from surfaces has a wide variety of applications in
many different fields, such as medicine, material sciences, and astrophysics [5–7]. Re-
ferred to as laser ablation or laser-induced desorption (LID), this process can elucidate
relationships between materials on a molecular level. By carefully chosing a wavelenth
that excites a material of interest, energy can be introduced to that material via laser
radiation. This can result in removal of material, transfer of energy to non-absorbing
species, extreme heating of the substance, dissociation and/or reaction of the material
[8]. Several theories have been put forth with respect to a mechanism for ablation,
and the debate continues; some of the most popular models will be presented in this
28
section. While there isn’t a quantitative match between the material presented in this
section and results reported in Chapters 4 and 5, qualitative descriptions are valuable
for interpreting trends. It should be noted that ablation and desorption are used in-
terchangeably by some groups, but differentiated by others. In the latter case, ablation
refers to a more violent removal of material than desorption; an attempt will be made
to keep language used by the literature discussed consistent.
A model for the effects of laser ablation was first suggested by J. F. Ready in 1965
[9]. It presented a quantitative look at heat transfer in metals and posited that there was
a different mechanism for long and short laser pulses. While these calculations were
based on very high laser intensities ( 10
8
W/cm
2
compared to10 W/cm
2
for our
approach), the findings are instructive to discuss. One of the most impressive results
from the derived equations was the extreme temperature gradients ( 10
8
deg/cm) that
could be induced on short time scales (10
10
deg/s) with laser heating. Also, the shape
of the laser pulse was found to have a significant effect: shorter, higher power pulses
resulted in a higher suface temperature than a longer, lower pulse of the same total
energy. Instead, the latter pulse would result in a smaller temperature gradient and a
larger penetration depth. Another result of interest is that as thermal conductivity of
the material decreases, the peak surface temperature increases with a decreased depth
of penetration.
Laser ablation, with a particular focus on ultraviolet matrix-assisted desorption pro-
cesses (UV-MALDI), has been modelled in the microscale by Zhigilei and Garrison [10–
12]. They utilized a “breathing sphere” model that approximates internal degrees of
freedom to allow for internal energy conversion to translational motion. A differentia-
tion between desorption and ablation is made: desorption is defined as a process where
chiefly monomers desorb from a surface, whereas ablation results in larger clusters of
material being released via photoabsorption. It is possible to move from the desorption
regime to the ablation regime by increasing laser fluence; the crossover point in these
simulations was at 0.7 mJ/cm
2
. However, the fraction of small clusters and monomers
29
will increase in the ablation regime with increasing laser fluence, so they are not only
indicative of a desorption mechanism.
The overall ablation process is as follows. Laser radiation tuned to excite molecules
in a film is absorbed. This energy quickly degrades to heat in the form of translational
motion. However, the temperature rises faster than the system can mechanically relax,
so there is a significant increase in pressure. Once the pressure gradient exceeds the
mechanical strength of the material, individual molecules and/or clusters are ejected.
Rapid heating of the confined material can force the material to temperatures above
the boiling point. This can result in a “phase explosion”, where the excited material is
quickly converted to a hot, gaseous fluid. Such a phase explosion contributes additional
pressure to the system that has already experienced heating at constant volume.
There were a few notable trends in the molecular dynamics (MD) simulations of
laser ablation [11]. The average velocity decreased in relation to the depth of the plume,
indicating that collisions play a significant role in the redistribution of energy as ma-
terial is ablated. This leads to a range of “stream velocities” that require an adjusted
Maxwell-Boltmann distribution to properly account for. Non-absorbing species (in this
case, analyte molecules) have a similar average velocity to the absorbing species (ma-
trix molecules). This follows from the phenomenon of entrainment, which has been
experimentally observed by several groups [13–15].
An opposing view of laser ablation put forth by Perez and coworkers [16] chal-
lenges the existence of a phase explosion. They argued that while phase explosion,
deemed “nontrivial” fragmentation, is possible with ultrashort laser pulses (ps scale),
nanosecond pulses cannot trigger such a transformation. Instead, a slower expansion
of material and efficient thermal conduction inhibit this process and ablation causes a
“trivial” fragmentation.
Their MD simulations used classical Lennard-Jones potential interactions and “car-
rier” atoms that absorb energy (photons) and transfer it to surrounding material. Mod-
elling a 4 μm-thick sample ablated with a 2.5 ns pulse, they found 3 unique regions
30
within the film where effects were dominated by the amount of inertial confinement. At
the surface, a region of weak inertial confinment, material is free to expand into vacuum
when irradiatively heated. The material can pass the melting point into a supercritical
state that generates void nucleation and subsequent clusters (trivial fragmentation) [17,
18]. Farther down in the heated film, homogeneous nucleation occurs during isochoric
heating and triggers a phase explosion (though different in character to that described
by Zhigilei). In the depths of the material, thermal diffusion from effective cooling be-
gins to hamper expansion. In this region, heteregeneous nucleation is the dominant
mechanism that results in the ejection of thick pieces of material (hundreds of nm).
Adsorbate desorption mechanisms fall into two categories: thermal and nonthermal
[19]. Thermal mechanisms are due to irradiative heating and the detected signal pattern
reflects those seen in temperature-programmed desorption. It has a Maxwell-Botzmann
shape, is sensitive to surface temperature, and has a long tail. Nonthermal mechanisms
are due to photoprocesses and are independent of surface temperature. A combination
of these mechanisms results in the removal of material. The process studied in this
dissertation differs from the typical ablation system. Instead of homogeneously heating
a sample with laser radiation, the goal of experiments detailed in this dissertation was
to implant energy in a spatially selective way. Specifically, a layer embedded within a
film absorbs radiation, which decays to heat in picoseconds. Heat can then transfer to
surrounding material. Furthermore, a pressure gradient may be introduced depending
on if a phase change occurs within the absorbing region. As such, the processes in-
volved will not result in simple removal of material from the surface and cannot easily
be described by typical ablation mechanisms. The proposed mechanism for our system
will be described in detail in Chapters 4 and 5.
31
Chapter References
[1] Wiley, W.; McLaren, I. Rev. Sci. Instrum. 1955, 26, 1150–1157.
[2] Time-of-Flight Mass Spectrometry; Cotter, R. J., Ed.; American Chemical Society:
1994.
[3] Mooney, J.; Kambhampati, P . J. Phys. Chem. Lett. 2013, 4, 3316–3318.
[4] Guilhaus, M. J. Mass. Spectrom. 1995, 30, 1519–1532.
[5] Edwards, G. Laser Photonics Rev. 2009, 3, 545–555.
[6] Lippert, T. Plasma Process. Polym. 2005, 2, 525–546.
[7] Livingston, F.; Smith, J.; George, S. Anal. Chem. 2000, 72, 5590–5599.
[8] Ho, W. In Desorption Induced by Electronic Transitions, Betz, G., Varga, P ., Eds.;
Springer-Verlag: Berlin, 1990.
[9] Ready, J. J. Appl. Phys. 1965, 36, 462–468.
[10] Zhigilei, L.; Kodali, P .; Garrison, B. Chem. Phys. Lett. 1997, 276, 269–273.
[11] Zhigilei, L.; Kodali, P .; Garrison, B. J. Phys. Chem. B 1998, 102, 2845–2853.
[12] Zhigilei, L.; Garrison, B. Rapid Commun. Mass Sp. 1998, 12, 1273–1277.
[13] Beavis, R.; Chait, B. Chem. Phys. Lett. 1991, 181, 479–484.
[14] Huth-Fehre, T.; Becker, C. Rapid Commun. Mass Sp. 1991, 5, 378–382.
[15] Pan, Y.; Cotter, R. Org. Mass Spectrom. 1992, 27, 3–8.
[16] Perez, D.; Lewis, L.; Lorazo, P .; Meunier, M. Appl. Phys. Lett. 2006, 89, 141907.
[17] Lorazo, P .; Lewis, L.; Meunier, M. Phys. Rev. Lett. 2003, 91, 225502.
[18] Perez, D.; Lewis, L. Appl. Phys. A 2004, 79, 987–990.
[19] Chuang, T. Surf. Sci. Rep. 1983, 3, 1–105.
32
Chapter 3
Experimental Apparatus
Now that a review of the relevant experimental methods has been dispensed with,
this chapter describes how each concept is put into practice for data collection. The
experimental chamber is discussed first, followed by a description of sample prepara-
tion, and finally details about the instruments and methods used for each experimental
technique are provided.
3.1 Ultra-High Vacuum System
All experiments were carried out in a 3-tier stainless steel ultra-high vacuum (UHV)
chamber that can reach pressures of <310
-10
Torr. This ensures high sample quality,
as monolayer formation from background gases takes several hours under these condi-
tions. Chamber pressure is monitored using a nude ion gauge (Granville-Phillips 330).
The chamber is pumped using a turbomolecular pump (Leybold TurboVac, 600 L/s)
that is backed with a mechanical roughing pump (Welch Model 1397). The 3 tiers
are structured such that different experimental analyses can be carried out in different
levels of the chamber. The top tier (labeled Tier 1 in Figure 3.1), where the surface
manipulator is attached to a gate valve (MDC GV-4000M), is set up to accommodate
Fourier-transform infrared (FTIR) spectroscopy and temperature-programmed desorp-
tion (TPD) experiments. It includes a residual gas analyzer (SRS RGA 300) and has two
ports with calcium fluoride (CaF
2
) windows to accomodate the path of the IR beam
33
from an FTIR spectrometer to an external detector (LN
2
-cooled Nicolet InSb detector).
The main part of the chamber is divided into upper and lower sections (Tier 2 and
Tier 3, respectively), where the upper section is used for laser-induced “eruption” ex-
periments detailed in this dissertation. Tier 2 inclues a window for introducing laser
radiation, two leak valves (MDC ULV-150), as well as a linear time-of-flight mass spec-
trometer. Tier 3 was previously used for molecular beam scattering experiments, but
is no longer in use. It houses a third leak valve that is used for dosing additional test
species during film growth.
To maintain UHV pressure after introduction of water to the chamber, which sticks
to the stainless steel walls and raises the base pressure of the system through outgassing,
the chamber had to be “baked” at regular intervals (bake-out temperatures can be found
in the appendix of reference [1]). This was done with heating tape that was wrapped
around the chamber and controlled with variable transformers (Staco-Variac 3PN1010).
The temperature of the chamber was raised to 110

C for 2-3 days; the temperature
was monitored with K-type thermocouples that were attached at various points on the
chamber. In order to maintain low pressure during the week, a heat lamp (Osram 650
W, Model 64535) installed inside the chamber and plugged into a Variac was left on
during the night to speed up outgassing.
3.2 Surface Manipulator
The top tier of the UHV chamber is equipped with a surface manipulator that is
attached to anxyz translational stage and enables rotation in thexy-plane (VG, modified
by McAllister Technical Services). It has a 600 mm range in the z direction (VG Omniax
Translator, z slide module) and a 25 mm range in both the x and y directions (VG
Omniax Translator, xy stage module). Furthermore, it can rotate 360

around the z-axis
via a differentially-pumped feedthrough system (VG Rotary Feedthrough, DPRF 25H).
The manipulator consists of a sheathed stainless steel reservoir that is silver-brazed to
a copper block to which the sample holder is attached, shown in Figure 3.2. Liquid
34
Figure 3.1: An external view of the UHV chamber (not to scale). Key sections are
labeled. The surface manipulator provides xyz translation capabilities. The window
through which UV radiation enters for a typical experiment is designated with a star.
The diamond indicates the FTIR pathway.
35
nitrogen is poured into this reservoir, which is open to atmosphere on one end, to
achieve cryogenic temperatures of <110 K. With good thermal contact between the
copper block and surface holder, temperatures as low as 96 K are possible. Even lower
temperatures can be reached by bubbling helium through the liquid nitrogen. Using
this method, surface temperatures of90 K can be reached.
3.3 Surface Holder and MgO Surface
The surface holder consists of two machined pieces of copper and a piece of copper
foil that holds the MgO(100) surface, which are diagrammed in Figure 3.2. The two
pieces of machined copper are attached to the copper block at the end of the stainless
steel liquid nitrogen reservoir by three screws. Ceramic washers (McAllister Technical
Services, size 4-40) and a ceramic spacer (USC machine shop, dimensions given in refer-
ence [1]) ensure that the two sides are electrically isolated from each other. A sapphire
disc (Esco Products, 0.040 in. thick, 0.75 in. diam.) is sandwiched between the machined
pieces and the copper block at the end of the surface manipulator. The disc acts as an
electrical insulator and a “thermal switch”. The thermal conductivity of sapphire in-
creases as its temperature decreases, peaking around 30 K [2]. This allows for efficient
cooling of the surface by the LN
2
reservoir. In contrast, heating is also efficient as the
thermal conductivity of the sapphire drops dramatically while it is heated, which ther-
mally isolates the hot surface from the cold copper block. This is necessary to ensure
that the surface can be heated effectively via a tantalum wire; this heating method will
be described later.
The copper foil (ESPI Metals, 0.012 in. thick) was cut to roughly 14 10 mm, with
an arm extending along the length of the substrate that could be bent around an arm
of one of the machined copper pieces. A 6 6 mm opening was cut in the foil to allow
radiation to pass through the MgO surface once it was in the holder; this was necessary
for performing transmission FTIR experiments. This holder design was adjusted from
a previous design to ameliorate a thermal gradient on the surface [3]. The side of the
36
substrate nearest the copper arm in contact with the LN
2
reservoir was several Kelvin
cooler than the far side, resulting in a “cold” side of the surface and a “hot” side of the
surface. By extending the arm of the holder along an entire side of the MgO surface
instead of having an arm that only extended halfway along the side, the temperature
gradient was reduced significantly. While increasing the amount of material used for
the surface holder could slow heating and cooling times, no significant change in either
were noted after the design change.
A length of tantalum wire (ESPI Metals, 0.015 in. diam.) was glued to the back
of the holder with ceramic glue (Aremco 835M), encircling the window, which acts as
a heater. The wire was encased in segments of ceramic rod (Omega ORX-020132) to
prevent shorting of the circuit. The ends of the wire were inserted in small holes in
the machined copper pieces (Figure 3.2) and secured with small stainless steel screws.
Electrical current can be introduced to the wire through two leads that are attached to
the machined copper pieces. Outside the chamber,8 A are provided from a power
supply (Hewlett-Packard 6259B) for rapid resistive heating. With this arrangement,
temperatures from 90-600 K can be reached.
MgO(100) was chosen as a substrate because it transmits >80% of radiation between
2,000–30,000 cm
1
, which enables transmission FTIR spectroscopy. The MgO(100) sub-
strate was prepared by cleaving it on both sides from a larger piece (10 10 300 mm,
fine-ground polish, MTI Corporation) in dry air to obtain a fresh, smooth surface. This
was later verified by atomic force microscopy (AFM), as shown in Figure 3.3. Polished
surfaces are undesirable for an experimental substrate because small amounts of foreign
material are left behind by the polishing process. The MgO substrate was clamped in
the copper foil, and a K-type thermocouple was affixed using ceramic glue to the “hot”
side opposite the arm that attaches to the copper holder. This ensured that temperature
measurements were an upper bound.
After a fresh surface is made, it is left with oxygen vacancies. These vacancies are
filled once the surface has been installed in the UHV chamber by adding an oxygen
37
Figure 3.2: An exploded view of the substrate holder and associated components. Two
machined copper pieces are electrically isolated from each other by a custom ceramic
spacer and attached to the copper block with long screws fixed with flat and ceramic
hat washers on each side. Two 18-gauge copper wires are screwed to the machined
pieces to allow for current to flow through the tantalum heater that is bonded to the
back of the substrate holder. A thermocouple is glued to the “hot” side of the MgO
substrate and connected to a wire threaded through the LN
2
reservoir sheath.
38
atmosphere of 10
-7
Torr and heating the surface to 600 K for 1 hour. If this is not done,
the vacancies may be filled with hydroxyl groups, which will change the nature of the
MgO(100) surface and its interactions with adsorbates [4, 5]. It is important to fill the
oxygen vacancies after the baking procedure has finished: CO and CO
2
that outgasses
from the walls of the chamber can react with the hot tantalum wire to form tantalum
carbide [6]. This can reduce the heater’s lifetime and increase the amperage needed for
reasonable heating rates.
Figure 3.3: An AFM image of a cleaved MgO(100) surface. The surface is quite flat
and unblemished, with terraces as the main visible feature along with a slightly larger
abberation.
3.4 Sample Growth
All films were grown on a MgO substrate that was cleaned of background gases
which may have condensed between experiments. This was ensured by heating the
39
surface to200 K to desorb water and NO
2
layers. Water samples (Macron Fine Chem-
icals, UHPLC-grade) were degassed with a freeze-pump-thaw cycle that was repeated
until no bubbles were visible upon thawing. Nitrogen dioxide (Aldrich,99.5% purity)
purity was monitored by visual inspection upon freezing of the sample. A pure white
solid indicated high purity; a blue tinge was evidence of N
2
O
3
contaminant. To remove
N
2
O
3
, O
2
was added in excess to the sample vessel and left to react with NO con-
taminant. The sample was then frozen with LN
2
, and the remaining O
2
was pumped
off.
All materials except for NO
2
were introduced to the UHV chamber via background
dosing. Three high-precision leak valves (MDC ULV-150) are affixed to the middle and
bottom tiers of the chamber to allow for dosing of multiple chemicals. The leak valves
were opened until the ionization gauge (Granville-Phillips 330) read a specific pressure,
and then the leak valve was left open for a predetermined amount of time to grow a film
of a known thickness. The pressures and times were determined previously in this lab
[3] and corroborated by the calibration procedure described in the following section.
Based on the calibration for ASW, at 2 10
7
Torr, a monolayer (ML) of material is
deposited every three seconds. Because it is difficult to define what a “monolayer” is
for N
2
O
4
, its thickness is reported in Langmuirs. A Langmuir (L) is defined as exposure
to 10
6
Torr for one second.
Because of the catalytic reaction of NO
2
with metal surfaces that forms NO, a differ-
ent dosing method had to be designed that minimized contact with the stainless steel
interior of the UHV chamber. A 1 cm diameter 25 cm long glass tube was cemented
to a custom stainless steel piece (USC machine shop) that could be affixed to the inte-
rior face of a leak valve with stainless steel screws. By rotating the surface position, we
could achieve directed dosing. Initially, this pathway had to be passivated to ensure
that NO
2
, not NO, was the majority species entering the vacuum chamber. This was
accomplished by opening the NO
2
line leak valve slightly (10
8
Torr) and monitoring
the NO/NO
2
ratio with the RGA until it had stabilized. Despite the position being
40
irrelevant for background dosing, the surface was rotated to face the NO
2
doser at the
beginning of the growth sequence for all films that contained N
2
O
4
to ensure a quick
and consistent dosing process.
3.4.1 Calibration Using Temperature-Programmed Desorption
In order to quantitatively determine film thickness as a function of dosing time,
it is necessary to conduct a TPD experiment of a monolayer of H
2
O to compare to
the bulk. The H
2
O monolayer desorbs at a higher temperature than the bulk (240 K
compared to 165 K [7]), so the spectral feature can be isolated for integration. Once the
value for a monolayer is known, different dosing times can be analyzed and integrated
for comparison, and thickness can be calculated. However, definitively detecting the
monolayer proved to be more difficult that expected, and another calibration method
using TPD had to be devised.
TPD experiments were carried out using an SRS RGA 300 with an attached nose
cone (USC machine shop). A fine mesh was affixed to the8 mm opening of the nose
cone with ceramic glue (Aremco 835M); this ensured that only material that entered
the nose cone could be ionized and detected. Notches were added to the edge of the
nose cone to aid pumping efficiency. This greatly reduced a long tail that was observed
in initial spectra collected with the old nose cone design. The surface was lowered to
the main chamber and H
2
O films were grown at100 K. This was necessary to obtain
results representative of film growth for typical experiments, as films grown when the
surface was in the upper chamber desorbed approximately half as much material with
the same dosing times as when grown in the main chamber, where films are grown for
typical experiments. While a film thickness difference due to dosing location was not
expected, it was most likely due to the water leak valve being located on the main tier.
The surface was then raised to the upper tier and oriented towards the end of the nose
cone, at a distance of 4 mm. A custom LabView program collected signal while the
surface was heated to approximately 300 K at a rate of 1-2 K/s.
41
Figure 3.4: (a) TPD spectra from ASW films (20-100 ML) grown in standard dosing
position (Tier 2). The bulk ASW desorption occurs at 165 K, with a crystallization peak
that appears at 160 K. (b) Integrated areas from the plots in (a) compared to previously
calibrated film thicknesses.
42
A representative set of TPD spectra are shown in Figure 3.4a. The multilayer des-
orption peak reproducibly desorbed around 180 K, which is consistent with TPD data
published by the groups of Kay and Goodman [7–10]. The dosing times compared to
expected film thicknesses showed a linear trend (Figure 3.4b), which was consistent
with previous calibrations done in the lab.
3.4.2 Directed Dosing vs. Background Dosing
The limitations of working with NO
2
due to its reactivity with stainless steel surfaces
led to the inclusion of directed dosing apparatus. Water and other nonreactive species
were still dosed via back-filling the chamber to avoid cross-contamination of dosing
lines. As growth was measured assuming a calibration curve based on background
dosing pressures, it was necessary to determine the growth difference between dosing
methods. Quantitative measurements of films are possible using FTIR; however, direct
measurement of NO
2
is impratical with our equipment as the only absorption peak
within the detection range is a very weak combination band [11]. Water was chosen as
an acceptable analogue for NO
2
as both species have unit sticking coefficients at 100 K
[12, 13]. Films were grown by back-filling the chamber and via the direct pathway at
2 10
7
Torr for 5 minutes, which is equivalent to 60 L (100 ML based on the previously
discussed calibration). The FTIR spectra of the films grown from each pathway were
collected and the area of the broad OH stretch absorption peak centered at 3200 cm
1
for each was calculated. The raw spectra are shown in Figure 3.5.
To account for material dosed on the back side of the MgO substrate, the area of
the background spectrum was divided in half. This value was then subtracted from the
direct dosing spectrum. Using the adjusted peak areas, the direct dosing peak was a
factor of 1.7 larger than the background dosing peak.
43
Figure 3.5: FTIR spectra of ASW grown by background and directed dosing, each at 2
10
7
Torr for 5 minutes (equivalent to 100 ML for background deposition). Removing
the contribution of material that grows on the back side of the MgO substrate (see text
for discussion of this calculation), directed dosing increases film thickness by a factor
of 1.7 when compared to background dosing.
44
3.5 Time-of-Flight Mass Spectrometer
In order to measure a mass spectrum in our UHV chamber, material must be ejected
from the surface of the MgO substrate. This is accomplished by irradiating the surface
with a Nd:YAG laser (Continuum 9010, 1064 nm, 10 ns pulse) triggered at 10 Hz. The
third or fourth harmonic is generated to give 355 or 266 nm radiation that is collimated
to3 mm using a telescope. The radiation is used either at this diameter to irradiate the
whole surface or focused with a 50 cm CaF
2
lens to a diameter of0.3 mm at the MgO
surface. The frequency is also reduced from 10 Hz to 1 Hz using an optical chopper
wheel (ThorLabs MC1000) to enable a higher sampling rate with the analog-to-digital
converter. The radiation then enters the UHV chamber through a CaF
2
window and
strikes the MgO substrate normal to the plane of the substrate and orthogonal to the
flight path of ionized material (entrance port is indicated by a star in Figure 3.1). The
substrate is located approximately 2.5 cm from the center of the ionizing region of the
time-of-flight mass spectrometer.
Our time-of-flight mass spectrometer (TOFMS) was custom-made by Jordan TOF
Products and is a two-stage linear Wiley-McLaren-style mass spectrometer [14]. Upon
irradiation, ablated material enters the extraction region parallel to two plates that are
held at 1800 V . A fraction of the material in this region is ionized by a continuous stream
of 70 eV electrons. After 7 μs, the voltage on the extractor plate is dropped to 1550 V , and
ions are extracted for 3 μs into the acceleration region, where a third accelerator plate
is grounded. The voltage difference results in the ions being accelerated into a 48 cm-
long field-free drift region. Voltages were chosen to optimize mass resolution around
m/q = 18. Ions are detected by a 25-mm chevron multi-channel plate (MCP) supplied
with high voltage divided in fixed ratios (44%, 24%, and 4%). Signals are amplified with
a fast pre-amplifier (SRS D-300 MHz) and then converted using an analog-to-digital
card (Gage CS 8500, 8 bit, 512k samples, triggered at 10 Hz).
The TOFMS can be pulsed at rates up to 200 kHz; for these experiments, it is trig-
gered at 100 kHz, which results in a complete mass spectrum every 10 μs. During one
45
Figure 3.6: A schematic representation of the TOFMS experiment. The fourth harmonic from a Continuum Nd:YAG is used.
The beam waist is reduced to 3 mm using a telescope. The radiation enters the chamber and excites the film at normal
incidence. Released material is ionized, extracted, and then detected using a multi-channel plate (MCP) at a rate of 100 kHz.
The set-up shown is with focused radiation, which allows for 9 individual “experiments” on one film via translation in the xz
plane, as indicated by the grid. These data can be averaged to improve S/N. By removing the 50 cm lens, a larger area can be
irradiated (shown in light purple) to allow FTIR experiments of the ablated film. 46
laser pulse, a sequence of mass spectra is collected, which is referred to herein as a
"temporal profile". By adjusting the sampling rate, mass spectra can be collected over a
300, 600, or 1200 μs window (30, 60, or 120 total mass spectra). However, increasing the
length of the temporal profile window results in a slight decrease in resolution.
Spectra are collected either from nine 0.3-mm-diameter spots on the MgO(100) sur-
face or a single3 mm-diameter spot centered on the surface (Figure 3.6). Multiple
shots on the same surface are possible because of the minimal lateral transfer of energy
during a laser pulse [15]. While the electronic excitation of the film via irradiation de-
cays to heat on a picosecond time scale, the diffusion of that heat through a film is not
as rapid. Thermal diffusion length in a surface is defined by
L
tm
= (2kt
L
)
1/2
(3.1)
where k is the thermal diffusivity of the material and t
L
is the laser pulse length [16].
Thermal diffusivity of amorphous ice can be estimated as
k
a
=
vl
3
(3.2)
with v equal to the velocity of sound and l is the phonon mean free path [17]. Using
a velocity of 3 10
5
cm s
1
and a mean free path of 10 Å gives k
1
10
2
cm
2
/s.
Using this value along with our pulse length of 10 ns results in a thermal diffusion
length of 45 nm. Even assuming a k value based on crystalline ice, whose maximum
reported value is 0.4 cm
2
/s at 100 K and can be considered an upper limit [18], the
thermal diffusion length is 0.9 μm. Thus, translating the surface 1 mm in the xy-plane
allows for irradiation of essentially undisturbed material, and multiple laser passes can
occur on one film. The 3 3 grid pattern shown in Figure 3.6 was the typical radiation
pattern for each film to maximize data collection without getting too close to the copper
holder. In contrast, irradiating the same spot multiple times results in a more complex
set of data as the first laser pulse irreversibly alters the ASW film (this will be discussed
further in Chapter 4). As such, subsequent shots on that spot are interacting with a film
with an ill-defined morphology.
47
Figure 3.7: (a) An example of a full temporal profile. It is the average of data collected
from eight individual spots from a film of 80 L of N
2
O
4
ablated with focused 1.0 mJ
radiation. Each 10 μs segment represents a full mass spectrum, shown as (b), of a seg-
ment of the plume of material that desorbs from the surface. The data can be analyzed
as a full temporal profile or as individual mass spectra.
48
When nine spots are irradiated on a single film, it allows for the temporal profiles
to be averaged together. By averaging the spectra, the signal-to-noise (S/N) can be
greatly improved and film irregularities and fluctuations in laser energy have less of an
impact on data collection. However, the S/N is high enough that individual temporal
profiles can still be used for analysis. An example of a temporal profile is shown in
Figure 3.7a. These profiles can be broken down into the individual mass spectra that
are taken every 10 μs. Each mass spectrum can be thought of as representing the make-
up of a specific section of the plume (Figure 3.7b). In this way, the shape of the temporal
profile gives information about the speed and amount of material leaving the surface,
and individual mass spectra give insight into the make-up of the plume at different
times after the initial laser pulse.
3.6 Time Synchronization
As stated previously, we trigger the TOFMS at 100 kHz. In order to synchronize
this with the firing of the Nd:YAG laser, we have a set of four low-jitter digital pulse
generators (SRS DG535) daisy-chained together as shown in Figure 3.8. This is an
improved set-up from a previous iteration that only used three pulse generators where
the TOFMS trigger had a jitter of160 ns. Because of this, the flight times would
vary from experiment to experiment. By using four pulse generators, we reduced the
jitter in the TOFMS with respect to the laser pulse by a factor of 10 (10 ns) and
improved experiment-to-experiment timing stability. This also increased confidence in
peak assignments due to the decreased variability in flight times for specific masses.
A brief summary of the triggering mechanism follows: the TOFMS pulse generator
runs at 100 kHz, acts as the master, and sets the internal clock for the other three pulse
generators. The output of the TOFMS pulse generator triggers a second pulse generator
that effectively runs at 10 Hz due to the delay settings (Table 3.1) It externally triggers
the flash lamps of the Nd:YAG laser. A third pulse generator is triggered off of the flash
lamps pulse generator. It also effectively runs at 10 Hz, but it controls Q-switching of
49
Figure 3.8: A schematic of the four SRS DG535 Digital Delay/Pulse Generators that
control the triggering for the TOFMS, chopper wheel, and Nd:YAG laser flash lamps
and Q-switch; only necessary connections and switches have been depicted for clarity.
Refer to Table 3.1 for individual pulse generator settings.
50
Table 3.1: Timing settings for each pulse generator used in the timing set-up. Note:
all of the pulse generators are triggered at 100 kHz except for the pulse generator
triggering the optical chopper, which has a repetition rate of 100 Hz. Refer to labels
from Figure 3.8.
A=T+ B=A+ C=A+ D=T+ AB
TOFMS 0 500 ns 0 0 50W, TTL
Flash lamps 62.1 μs 10 μs 0 99.995 ms High Z, TTL
Q-switch 324 μs 10 μs 0 0 High Z, TTL
Chopper wheel 0 10 μs 0 9.995 ms High Z, TTL
the laser which is triggered after a 364 μs delay to optimize the energy output. Finally,
the fourth pulse generator controlling the optical chopper wheel is triggered off of the
TOF pulse generator and runs at 100 Hz.
3.7 Fourier-Transform Infrared Spectroscopy
A Nicolet Protegé 460 spectrometer is located next to the top tier of the UHV cham-
ber and was used in conjunction with an external InSb detector (Nicolet 860) for all ex-
periments detailed in this dissertation. The IR beam was directed and focused through
the surface (which was moved to the top tier for data collection) with a flat mirror and
a parabolic mirror and then refocused on an indium antimonide (InSb) detector with a
second parabolic mirror. The entrance and exit windows on the FTIR tier are CaF
2
to
allow for transmission of the IR beam. A wire-grid polarizer (Molectron, 93-98% purity)
was sometimes placed in the IR beam path to prevent saturation of the detector. The
optics and detector are housed in opaque, lidded plexiglass boxes that are purged with
dry N
2
to decrease absorption due to background gases (e.g. CO
2
, water vapor). Purge
gas for the external boxes and the FTIR bench was provided by a dry air gas generator
(Whatman, FT-IR 75-62) that filtered H
2
O and CO
2
from a house compressed air line.
51
Figure 3.9: A schematic of the top chamber tier, showing the FTIR pathway. Radiation
from the detector is directed and focused through the surface using flat and parabolic
mirrors. It is then focused on a liquid nitrogen-cooled InSb detector with a second
parabolic mirror. All external optics are encased in plexiglass boxes that are purged
with dry air.
52
Spectra were measured using OMNIC software to collect 200 scans per spectrum at
a 1 cm
1
resolution over the range 1850 to 7000 cm
1
. With these settings, the system
was sensitive enough to detect <10 ML of ASW on the MgO substrate. FTIR was used
for checking dosing procedures as well as film changes from heating and irradiation.
For irradiation experiments, a background spectrum was collected in Tier 1, the surface
was moved to Tier 2 for dosing, it was then moved back to Tier 1 for an initial spectrum,
then moved back down to Tier 2 to be irradiated, and then moved up to Tier 1 for the
final spectrum.
53
Chapter References
[1] Suchan, M. M. Molecule-Surface Interactions in HCl/MgO and Water/MgO Sys-
tems., Ph.D. Thesis, University of Southern California, 2001.
[2] Slack, G. A. Phys. Rev. 1962, 126, 427–441.
[3] Malyk, S. Transport and Guest-Host Interactions in Amorphous and Crystalline
Ice., Ph.D. Thesis, University of Southern California, 2009.
[4] Korolik, M.; Suchan, M.; Johnson, M.; Arnold, D.; Reisler, H.; Wittig, C. Chem.
Phys. Lett. 2000, 326, 11–21.
[5] Knözinger, E.; Jacob, K.; Hofmann, P . J. Chem. Soc. Faraday Trans. 1993, 89, 1101–
1107.
[6] Chrysanthou, A.; Grieveson, P . Mat. Sci. Eng. A 1995, 194, L11–L14.
[7] Hawkins, S.; Kumi, G.; Malyk, S.; Reisler, H.; Wittig, C. Chem. Phys. Lett. 2005,
404, 19–24.
[8] Kim, Y. D.; Stultz, J.; Goodman, D. W. J. Phys. Chem. B 2002, 106, 1515–1517.
[9] Günster, J.; Liu, G.; Stultz, J.; Krischok, S.; Goodman, D. W.J.Phys.Chem.B 2000,
104, 5738–5743.
[10] Stirniman, M. J.; Huang, C.; Smith, R. S.; Joyce, S. A.; Kay, B. D. J. Chem. Phys.
1996, 105, 1295–1298.
[11] Mélen, F.; Pokorni, F.; Herman, M. Chem. Phys. Lett. 1992, 194, 181–186.
[12] Dixon-Warren, S. J.; Jackson, R.; Polanyi, J.; Rieley, H.; Shapter, J.; Weiss, H. J.
Phys. Chem. 1992, 96, 10983–10994.
[13] Brown, D.; George, S.; Huang, C.; Wong, E.; Rider, K.; Smith, R.; Kay, B. J. Phys.
Chem. 1996, 100, 4988–4995.
[14] Wiley, W.; McLaren, I. Rev. Sci. Instrum. 1955, 26, 1150–1157.
[15] Burgess, D.; Stair, P .; Weitz, E. J. Vac. Sci. Technol., A 1986, 4, 1362–1366.
[16] Matthias, E.; Reichling, M.; Siegel, J.; Käding, O.; Petzoldt, S.; Skurk, H.; Bizen-
berger, P .; Neske, E. Appl. Phys. A 1994, 58, 129–136.
[17] Smoluchowski, R. Astrophys. J. 1981, 244, L31–L34.
[18] Hobbs, P . V ., Ice physics; Oxford: Clarendon Press: 1974.
54
Chapter 4
Effects of Buried Heat in Amorphous Solid Water
Films
Previous experiments done in this lab had been carried out using IR radiation to
excite amorphous solid water; as such, there was an element of the unknown to how
NO
2
would interact with radiation with our experimental approach. As we would
be exploring a new system and method of energy transfer to ASW films, extensive
testing and characterization of the system was carried out. Important data trends with
various combinations of N
2
O
4
and ASW are summarized in this chapter, and a model
for morphological change and transport in ASW is presented. The process of fine-
tuning signal collection is also documented herein.
4.1 Introduction
Amorphous solid water (ASW) has been the subject of intense study since its orig-
inal discovery in 1935 [1]. It remains fascinating to the scientific community due to its
unusual structure and characteristics, such as its heat capacity behavior and high vis-
cosity at low temperatures, that continue to elude understanding [2]. ASW has come
under further scrutiny as it became apparent that H
2
O is an abundant molecule in the
interstellar medium (ISM), and the corresponding cold temperatures necessitated that
it would commonly exist in the solid phase [3]. ASW is present in diverse astronomical
55
enviroments, from the surface of Enceladus to dust grain mantles, the latter of which it
is believed to be the main component. In the ISM, ASW is subject to a variety of energy
sources that can trigger morphological changes: chemical and thermal processing, gas
phase bombardment, and UV radiation [4]. As such, an understanding of fundamen-
tal properties of this substance is important for basic scientific inquiry as well as for
gaining insight into chemical processes in the universe.
Many studies have addressed the unique changes that occur in ASW when it un-
dergoes morphological changes from thermal heating [5–7]. Previous work done in our
lab has investigated ASW changes using infrared radation as an energy source. We
found that IR absorption in an ASW film initiated efficient ejection of water monomer
and small clusters, as well as the preferential release of guest species [8]. In the experi-
ments documented in this chapter, we wanted to elucidate morphological changes and
molecular transport when stress was exerted by a guest molecule. Specifically, the elec-
tronic excitation of N
2
O
4
was exploited to introduce a very hot layer of material within
an ASW film, which could then disrupt the surrounding matrix. A broad absorption
band due to an electronic transition centered around 250 nm allows for easy excitation
with the fourth harmonic of a Nd:YAG laser (10 ns duration, 266 nm, 10 Hz) that can
then degrade to heat on a ps time scale [9, 10]. This combination gives a facile way to
investigate the effect of heat in a buried stratum, though interpretation proved to not
be trivial.
While these experiments were conducted with microscale films (100-1000 nm),
they may provide some qualitative insight into macroscale events. A specific example
is Enceladus, a moon of Saturn that is believed to have a liquid water ocean located
underneath a thick solid water crust. This theory is based on detection of “plumes”
of molecules (H
2
O is the dominant molecule, along with CO
2
, NH
3
, HCN, CH
2
O, CO,
H
2
, and CH
4
) that erupt from the surface along the “tiger stripes” located on the south
polar terrain [11, 12]. It is believed that tidal stresses provide enough thermal energy
to maintain a liquid ocean. When extreme pressure and heat is generated during the
56
cyclic ebb and flow of this system, the liquid water forces its way through fissures in
the upper ASW layer and escapes.
The material presented in this chapter covers a thorough review of experiments
conducted with this system. Films were grown via vapor deposition between 95 and
110 K, which resulted in a low-density ASW film. Subsequent UV photolysis of a N
2
O
4
layer creates a hot fluid that then contributes to a temperature and pressure gradient
within the film. If covered by a layer of ASW, N
2
O
4
forces its way through fissures to
escape to vacuum, scraping water monomer from the wall during transport.
4.2 Experimental
A schematic of the experimental apparatus used for irradiation of all ASW/N
2
O
4
films discussed in this chapter is shown in Figure 3.6 in the previous chapter. Experi-
ments were conducted in an ultra-high vacuum (UHV) chamber (3 10
10
Torr) using
pulsed 266-nm radiation (10 Hz, 10 ns) in conjunction with a high-repetition time-of-
flight mass spectrometer (TOFMS) from Jordan TOF Products. The UHV chamber has
multiple tiers capable of different types of experiments; these specifications have been
documented thoroughly elsewhere [8, 13]. Briefly, the top tier allows for temperature-
programmed desorption (TPD) and Fourier transform infrared (FTIR) spectroscopy, and
the middle tier houses the equipment for TOF mass spectrometry. The latter two meth-
ods are those used to obtain the data in this chapter.
A MgO(100) substrate was used for all experiments. The MgO surface was obtained
by cleaving a crystal on both sides in dry N
2
. The 0.8 10 10 mm substrate was
then transferred to the UHV chamber. Once vacuum was achieved, oxygen vacancies
in the MgO substrate were filled by backfilling the chamber with O
2
(10
8
Torr) and
annealing the surface to 600 K for an hour. This procedure has been shown to produce
a high quality MgO(100) surface [14].
The substrate was clamped in copper foil (0.012 in. thick) with a 6 6 mm square
cut-out for radiation to pass through. It was then attached to the arm of one of two
57
electrically-isolated copper blocks that were mounted onto a copper-capped stainless
steel cold finger on a LN
2
reservoir (Kurt Lesker, altered by McAllister Technical Ser-
vices). With this configuration, the surface could be cooled to <100 K with LN
2
or 90
K by bubbling He through the LN
2
. The temperature was monitored using a K-type
thermocouple glued (Ceramabond 835-M, Aremco Products) to one side of the MgO
surface. A tantalum wire cemented to the back of the copper housing of the surface
was used for resistive heating. The cryostat was fitted with bellows and was capable of
xyz translation and 360

rotation.
Liquid samples were kept in glass containers with Teflon stopcocks connected to a
glass and Swagelok gas-handling manifold with an attached vacuum line. H
2
O sam-
ples (Macron Fine Chemicals, UHPLC-grade) were degassed with a freeze-pump-thaw
process. NO
2
(Sigma-Aldrich,99.5% purity) purity was checked by freezing the sam-
ple and noting the presence of a pure white solid. The most common contaminant was
N
2
O
3
due to excess NO, which would give a blue cast to the solid [15]. This was ad-
dressed by adding O
2
to the sample to react with NO contaminant. After 4-12 hours,
the sample was frozen (NO
2
freezing point = -11

C) and the remaining O
2
was pumped
off to leave pure NO
2
, evidenced by the observation of a pure white solid when frozen.
All H
2
O/N
2
O
4
films were grown at 100 K. All species were introduced to the surface
using background dosing except for NO
2
. Because of its catalytic reaction with metal
surfaces, a 1 cm ID 23 cm long glass tube was affixed to a precision leak valve for
directed dosing. The surface was rotated from the TOFMS position such that the end
of the tube was approximately 4 cm away and normal to the surface. After heating the
surface to200 K to remove impurities and recooling to 100 K, each layer was dosed
sequentially to avoid mixing of the species within layers. Upon completion of dosing,
the surface was rotated back to TOFMS position as shown in Figure 3.6. Directed dosing
grew films at a faster rate than background dosing; however, background pressure was
used as an indication of dosing rate. A conversion factor of 1.7 to account for this
difference was calculated from TPD experiments used to adjust all N
2
O
4
thicknesses
58
reported herein (refer to Section 3.4.2 in Chapter 3 for a detailed description of this
calculation).
Ultra-violet radiation (266 nm) was generated by pumping second- and fourth-
harmonic crystals with 1064 nm radiation from a Q-switched Nd:YAG laser (Contin-
uum PL9010,500 mJ, 1 ns) with a 10 Hz repetition rate. The beam was reduced
from 8 mm to 3 mm in diameter using a telescope and reduced to 1 Hz with an optical
chopper wheel (ThorLabs MC1000). It was then focused to0.3 mm beam diameter at
the substrate using a 50 cm CaF
2
lens.
The TOFMS, externally triggered at 100 kHz to produce a mass spectrum every
10 μs, consisted of a linear two-stage Wiley-McLaren type extraction with three elec-
trode plates and an electron impact ionizer. Once the film was irradiated, the plume of
material leaving the substrate entered the ionizing region parallel to the repeller and ex-
traction plates. The plates were initially held at 1800 V while the molecular plume was
bombarded with 70 eV electrons from an electron beam positioned25 mm from the
surface. The positive ions formed in this region were extracted every 10 μs by dropping
the potential on the extractor plate to 1550 V for 3 μs. Ions were then accelerated by a
grounded third plate into the 48 cm-long field-free drift tube. Ions were detected at the
end of the drift tube by a multi-channel plate (MCP , -3100 or -3200 V) connected to a
fast pre-amplifier (SRS D-300 MHz). The amplified current from the MCP was recorded
on a computer with an analog-to-digital converter (Gage CS 8500, 8 bit, 512k samples).
Spectra were collected using a custom LabView program (National Instruments) and
further processed using IGOR Pro (WaveMetrics). The set of 30-120 individual mass
spectra collected during a single laser pulse is referred to in this chapter as a “temporal
profile”.
Nine spots in a three by three grid were irradiated per surface to give nine individual
temporal profiles. This gave the option of an averaged spectrum to improve signal-
to-noise and to reduce the effect of laser energy fluctuations and film irregularities.
Because of the limited diffusion of energy in the surface perpendicular to the laser
59
beam, moving the laser position by 1 mm irradiated an “undisturbed” section of the
film [16].
All FTIR spectra were collected using a Nicolet Protegé 460 spectrometer and a LN
2
-
cooled InSb detector. Spectra were measured using OMNIC software over the span of
1850-7000 cm
1
with a 1 cm
1
resolution and averaged over 200 scans. The IR beam
was collimated using an iris to a 2 mm beam diameter to ensure that it only sampled
the part of the surface that was irradiated. To collect an FTIR spectrum, the surface
was raised to the top tier to collect a background. The surface was then lowered to
the middle tier for the dosing procedure and then raised again to collect a baseline
FTIR spectrum. After the surface was lowered again, the 50 cm CaF
2
focusing lens
was removed to give a 3 mm-diameter beam centered on the surface. The surface was
irradiated and raised to take a final spectrum.
4.2.1 Changing to UV: Using NOCl as a Test Species
Initially, the laser was reconfigured to generate 532 nm radiation via a second har-
monic generation (SHG) crystal, as this wavelength is within the wide absorption band
of NO
2
and was assumed to be capable of initiating desorption [17]. However, no sig-
nal was seen, and review of the literature revealed that it was very unlikely that any
NO
2
would be present on the MgO surface; instead, it would all condense as the dimer,
N
2
O
4
[18]. To address this, a third harmonic generation (THG) crystal was installed
to produce 355 nm radiation, which is absorbed by N
2
O
4
[17]. After thorough test-
ing, we were once again unable to get signal with this wavelength. This was not due
to not enough N
2
O
4
condensing on the MgO substrate; desorption was proven to be
facile when hitting the copper holder and FTIR spectra of the surface showed the char-
acteristic combination band at 2960 cm
1
(Figure 4.1a). The ease of signal achieved
from the copper holder is likely the result of substrate-mediated processes that have
been observed by other groups using metal substrates [19]: UV electrons excite valence
60
electrons in the substrate, which can cause a temperature spike and induce thermal
desorption, among other processes.
Since we were only able to detect the presence of N
2
O
4
and not the amount due to
the range of our InSB detector (all N
2
O
4
fundamental peaks fall below 1700 cm
1
[20]), a
switch was made in the absorbing species. NOCl was chosen as an ideal candidate due
to its similar absorption spectrum in the UV [22] and familiarity with its characteristics
due to past use in this lab [23–25].
NOCl was synthesized by first condensing NO and Cl
2
in a 2:1 ratio under cryo-
genic conditions, with NO in slight excess, in a glass vessel using a vacuum line. The
mixture was then left to sit for 24 hours at room temperature to ensure complete reac-
tion. A freeze-pump-thaw purification procedure was carried out using a pentane slush
(131

C) to remove any excess NO. Finally, the dosing line was passivated using the
same method as described previously for NO
2
, except mass 65 was monitored.
Signal was initially detected after dosing 24 L, but it was very small. Once again,
coverage was confirmed using FTIR spectroscopy (Figure 4.1b), so dosing problems
were not the culprit. Despite trying different combinations of coverage, laser energy,
and radiation spot size, the signal size did not increase. Because of this, we switched to
266 nm radiation by installing a fourth harmonic generation (FHG) crystal in an attempt
to increase the absorption, and thus, signal. This did result in increased signal, and so
after some fine-tuning, NO
2
was reintroduced.
N
2
O
4
signal remained small, and so various dosing combinations were tried: N
2
O
4
on MgO, N
2
O
4
on ASW, N
2
O
4
codeposited with ASW, and ASW-N
2
O
4
“sandwich.”
None of these experiments had a significant improvement in signal, so a sequential
measurement of laser energy was carried out after each optical element. We found that
laser energy was being decreased 10-fold due to a mislabeled mirror and a damaged
CaF
2
lens that was absorbing most of the radiation. After replacing the lens, mirror,
and optimizing the alignment, we were able to get significant signal. Subsequent ex-
periments were carried out with N
2
O
4
and the experimental set-up described in the
61
Figure 4.1: (a) FTIR spectrum of 100 L of N
2
O
4
.The sharp peak at 2960 cm
1
is a weak
combination band [20]. All fundamental bands for N
2
O
4
are below the detection range
of the InSb detector used in these experiments (cutoff at 1850 cm
1
), so this small peak
was the only indication of the presence of N
2
O
4
on the surface. However, it was distinct
and reproducible. (b) FTIR spectrum of 110 L of NOCl (negative peaks are due to
decrease in water vapor signal on the experimental time scale). The peak centered at
1955 cm
1
is the NO bond stretching mode [21]. We did not detect any other peaks
associated with NOCl. While this particular peak is also representative of NO only, our
surface was not cold enough to condense NO [18]. Furthermore, we were able to desorb
fragments of NOCl successfully with UV radiation.
62
the previous section. A return to 355 nm in future experiments with improved optical
elements would likely yield higher signal from NOCl. Furthermore, N
2
O
4
should yield
signal, though higher fluence may be needed to collect data that resembles the results
reported herein.
4.3 Results
266-nm radiation was used to irradiate N
2
O
4
/ASW films of various compositions
and thicknesses, and single-shot and multi-shot experiments were carried out. Time-of-
flight mass spectra were collected during irradiation, and FTIR spectroscopy was used
to monitor the films before and after irradiation. Interpretation of this data required the
collection of individual spectra for all molecules involved: N
2
O
4
, NO
2
, NO, and H
2
O.
These spectra, along with their implications, are presented in Section 4.3.1. Following
a definition of terms, the subsequent sections each present a set of data with a specific
variable analyzed to characterize this system. The individual trends contributed to
development of a model for fissure formation presented in Section 4.4. All TOF mass
spectra and temporal profiles are the result of one pulse on a film; averaging was carried
out for spot-to-spot data, not shot-to-shot (e.g., the first and second pulse data are not
averaged together).
4.3.1 Standard Spectra and Data Analysis
Analysis of TOF spectra involves comparisons between individual spectra as well
as accounting for individual species’ contributions. In order to make such comparisons
with confidence, it was necessary to collect the mass spectra with our apparatus for
all experimentally-introduced molecules instead of relying on spectra published in the
literature. By collecting spectra with our equipment, we avoid errors due to a difference
in experimental method. The spectra obtained are shown in Figure 4.2. It was trivial
to record spectra for H
2
O, NO
2
, and NO by leaking vapor/gas into the UHV chamber
63
while triggering the TOFMS. In contrast, the mass spectrum of N
2
O
4
was not found
in the literature, despite references to its existence [26]. This is due to the gas phase
equilibrium of 2 NO
2
*
)
N
2
O
4
where NO
2
is heavily favored at room temperature,
as well as the weak NN, which makes a gas phase spectrum of pure N
2
O
4
essentially
impossible to achieve through standard means. Therefore, it was necessary to devise a
different method for collecting the mass spectrum of N
2
O
4
.
The method chosen to obtain a “pure” spectrum of N
2
O
4
was to do a thermal des-
orption of a condensed film. As the energy needed for desorption is 0.47 eV [27] and
the bond strength of the NN bond is 0.55 eV [28], it is safe to assume that the material
that desorbs at 155 K has N
2
O
4
as the majority species. This is because in order for
NO
2
to enter the gas phase during sublimation, it would need enough energy to break
the NN bond and then desorb as NO
2
. Such a scenario is energetically unfavorable
at 155 K, the desorption temperature of N
2
O
4
. The mass spectrum for N
2
O
4
is also
presented in Figure 4.2; a thorough search of the literature indicates that this is the first
direct measurement of the cracking pattern of this molecule.
It is apparent in Figure 4.2b that 70 eV electron impact ionization of N
2
O
4
results
in a maximum mass at 46 amu (NO
2
+
). N
2
O
4
+
isn’t expected to be a stable parent ion,
since ionization of an electron from the HOMO results in loss from the NN bonding
orbital. However, N
2
O
3
+
was reported to be a stable parent ion [26, 29], but was notably
absent in all spectra collected. Because the only species present in the N
2
O
4
spectrum
are identical to those produced by NO
2
(NO
2
+
, NO
+
, O
2
+
, O
+
, and N
+
), determining
parentage of molecules in TOF mass spectra presented a challenge. What does differen-
tiate between these two species is the signal ratio between molecules from the cracking
pattern. Therefore, an important number to define is the ratio between the areas of the
NO
+
and NO
2
+
peaks, which we have termed R
N
. NO
2
+
can only be a product from
cracking of NO
2
or N
2
O
4
, whereas NO
+
can result from NO, NO
2
, and N
2
O
4
. As NO is
most likely due to NO
2
photolysis, its contribution to R
N
can be considered negligible
and parentage from NO
2
and N
2
O
4
can be assumed. Furthermore, this ratio is unique
64
for NO
2
and N
2
O
4
, as shown in Figure 4.2a and b, and thus can give insight into the
nature of the material that is desorbing. Two very different ratios are obtained for NO
2
and N
2
O
4
, which are 3.1 and 0.9, respectively. This large difference allows for some
statements to be made about the parent ions.
The NO
2
/N
2
O
4
ratio can be expressed in terms of the respective numbers of ions:
NO
2
+
, NO
+
, O+, and N+, as well as the N
2
O
4
and NO
2
ionization cross sections at 70 eV .
An equation expressing this relationship is given below:
NO
2
N
2
O
4
= 2.08

NO
+
/NO
2
+

F
NO
2
+
N
2
O
4
+
F
NO
+
N
2
O
4
+
F
NO
+
NO
2
+

NO
+
/NO
2
+

F
NO
2
+
NO
2
+
(4.1)
where F
X
+
Y
+
is the percentage of ion X
+
that derives from parent ion Y
+
. The number
2.08 is the ratio of the ionization cross sections for N
2
O
4
and NO
2
(see Table 4.1). The
full derivation of this expression is given in Appendix A.
Table 4.1: 70 eV ionization cross sections for detected molecules. All values obtained
from the NIST database [30], except for N
2
O
4
, which was computed from the equation
given in [31].
Molecule ionization cross section (10
16
cm
2
)
N
2
O
4
7.34
NO
2
3.53
NO 2.81
N
2
2.51
O
2
2.41
O 1.36
H
2
O 2.28
There were concerns about molecular collisions with the electrodes in the ionization
region due to the wide angle of desorption. Such collisions could result in dissocia-
tion of the dimer to form NO
2
and skew the results. To address this, an alternative
65
Figure 4.2: Mass spectrum of NO
2
, N
2
O
4
, NO, and H
2
O obtained from our experimental apparatus. All ratios given are based
on peak areas.(a) 300 K NO
2
leaked into the UHV chamber gives NO
+
/NO
2
+
= 3.18, O
+
/NO
2
+
= 0.58, and N
+
/NO
2
+
= 0.19.
(b) 300 K NO leaked into the chamber gives O
+
/NO
+
= 0.01 and N
+
/NO
+
= 0.06. The NO
2+
signal at m/q = 15 is a distinctive
feature of NO electron impact ionization. (c) 300 K H
2
O leaked into the chamber gives OH
+
/H
2
O
+
= 0.27, O
+
/H
2
O
+
= 0.02, and
H
+
/H
2
O
+
(not shown) = 0.07. (d) The spectrum obtained via N
2
O
4
sublimation at155 K gives NO
+
/NO
2
+
= 0.9, O
+
/NO
2
+
= 0.1, and N
+
/NO
2
+
= 0.02. Neither N
2
O
3
+
or N
2
O
4
+
(76 m/q and 92 m/q, respectively) was detected (not shown).
66
repelling electrode was designed with a narrow slit parallel to the substrate to allow
through only molecules that were desorbing close to a normal trajectory from the sur-
face. The first design was made symmetrical to avoid a large distortion of the electrical
field during extraction (Figure 4.3a). However, this had the unintended consequence of
increasing the number of molecular collisions (observed as an increase in NO
+
signal),
as material would hit the flap on the opposite side of the ionization region from the
surface and subsequently be extracted. This was made apparent by an increase in R
N
from 0.9 to 1.0 in all experiments involving the first electrode design. The electrode was
modified such that material could more easily escape the ionization region, shown in
Figure 4.3b. The resulting spectrum resembled those that had been taken with the stan-
dard electrode, indicating that spectra collected under typical conditions were reliable.
Therefore, the original electrode from Jordan TOF Products was reinstalled and used
for all data presented in this dissertation.
Figure 4.3: (a)The initial design for a TOFMS electrode to minimize the effect of molec-
ular collisions with the plates. (b) Improved design for a TOFMS electrode to allow
material to pass out of the ionization region. The stars designate the openings parallel
to the substrate through which desorbing material would enter the ionizing region.
To better understand the interactions between N
2
O
4
and ASW, sets of experiments
were designed and carried out to see how different variables affected the TOFMS signal.
These sets of data will be discussed individually in the following sections and will
contribute to a cohesive model described in Section 4.4.
67
4.3.2 ASW Spacer
The starting point for irradiation experiments after getting sufficient signal was dos-
ing a thin film of N
2
O
4
on a bare MgO surface and then introducing an ASW "spacer"
underneath. Representative data are shown in Figure 4.4.
There was an overall increase in NO
+
x
species signal due to the addition of an ASW
film underneath the N
2
O
4
layer. Integrating the peak areas of all NO
+
and NO
2
+
peaks
in the temporal profiles gives a 75% increase in detected material with an ASW spacer
compared to without. Such an increase in signal of the absorbing species points to
an improvement in energy transfer efficiency. To explain this phenomenon, a brief
discussion of thermal properties of the experimental materials is needed.
Magnesium oxide acts as an excellent thermal conductor in our exerimental appa-
ratus. At 100 K, MgO has a thermal conductivity of250 W/mK [32]. When energy
is introduced to the N
2
O
4
film via laser radiation, it is quickly “drained away” due to
the MgO substrate being an efficient thermal condutor. In contrast, ASW is a very poor
thermal conductor, whose thermal conductivity is1 W/mK [33–35]. When ASW is
between the N
2
O
4
film and the MgO substrate, it acts as a thermal barrier and a slower
decay of heat in the absorbing layer results. Because of this, more N
x
O
y
species can
successfully leave the substrate.
Investigation of the individual mass spectra reveals that only derivatives of N
2
O
4
are
present in significant quantities, as shown in Figure 4.5; almost no mass 18 (indicative
of H
2
O
+
) is detected for films dosed with an ASW spacer unless very high fluences are
used (>1.0 J/cm
2
).
This contrasts with results from Yang and Gudipati[36], who developed a method
for removing species underneath an absorbing film. They started with a film of D
2
O
mixed with polyaromatic hydrocarbons (PAHs) and then a layer of H
2
O on top. Using
IR radiation to excite the OH stretch of the top layer of water, they were able to remove
the underlayer of D
2
O and PAHs. In our case, little to no material is being removed
from the ASW layer. Thermally, we are unlikely to remove material because any heat
68
Figure 4.4: Temporal profiles of NO
+
x
signal with and without an ASW spacer, with car-
toon depictions of deposition films. Both films were irradiated with a focused 266 nm,
1.0 mJ beam. (a) 80 L N
2
O
4
deposited on a bare MgO substrate. (b) 80 L N
2
O
4
deposited
on 300 ML ASW.
69
Figure 4.5: Averaged mass spectrum from extractions 2-11 (20-110 μs) of data from
Figure 4.4b. All species detected can be attributed to N
2
O
4
irradiation and cracking.
There is a slight bump at mass 18, but it is insignificant in comparison to the other
peaks and compared to previous experiments conducted by this lab.
entering the ASW layer is dissipated too quickly for any water molecules to desorb. The
other possible mechanism would be a “shockwave”, as described by Yang and Gudipati.
As we are dosing very thin layers of the absorbing species (<200 nm) compared to Yang
et al ( 3 μm), this mechanism may be too small to contribute detectable signal with
our current experimental method.
Another feature of these data is the overall shape of the temporal profile. The shape
of the NO
+
x
temporal profiles in Figure 4.4 is characteristic of an unobstructed pathway
to vacuum. While the modulation of the NO
+
/NO
2
+
ratio across a temporal profile
differs for various N
2
O
4
and ASW spacer thicknesses, the qualitative shape remains the
same. This will become significant when discussing trends with obstructed pathways,
i.e. a layer of material over a N
2
O
4
layer.
The first extraction of the temporal profile shown in Figure 4.4 that yields a resolved
mass spectrum is the 20 μs extraction. Assuming a molecule leaves the surface at
70
t = 0 and using 20 μs as an approximation of travel time to the ionization region gives
a molecular speed of 1.2 10
5
cm/s. More careful calculations cannot be made
because the large ionization region obfuscates exact travel time. In contrast, material
detected in the final extraction (travel time = 300 μs assuming t = 0) has a speed of
8 10
3
cm/s. Given these speeds, a NO
2
molecule in the first extraction with signal
has a kinetic energy of 2800 cm
1
, whereas in the final extraction it has a kinetic energy
of 12.5 cm
1
. Low kinetic energies for the later extractions are believed to be from the
tail of a warm distribution: the material in the final extraction represents 1% of the total
material detected.
It is important to determine the identity of material that leaves the substrate. The
ratio R
N
has already been established in Section 4.3.1 and can be used to determine the
dominant species in the ionization region. It is obvious from qualitative inspection of
the NO
+
x
peaks that R
N
starts off high and then decreases in later extractions and levels
off. The average value of R
N
at later times is 0.95, which is very close to the value of
0.9 found for N
2
O
4
. Clearly, the source of material detected in the late time extractions
is predominantly N
2
O
4
with a small contribution from NO
2
.
4.3.3 N
2
O
4
Thickness
The next trend examined was that of N
2
O
4
thickness: we wanted to determine
the point at which the amount of N
2
O
4
irradiated gave enough signal to be detected
without flooding our UHV system with NO
2
from long dosing times. This set of data
is shown in Figure 4.6. The relatively linear trend shows that with increasing thickness,
more material is being removed instead of a constant amount. It is assumed that there
must be some point of saturation where the amount of material leaving the surface
would level off, but that point is >100 L. This trend is indicative of the high fluence
being used, but may also be a result of the heat gradient induced by the MgO heat sink.
By increasing the overall thickness of the irradiated film, cooling rates are slower and
may enable more N
2
O
4
to leave the substrate.
71
Figure 4.6: Sum of peak areas for NO
+
and NO
2
+
signals from laser ablation of 20-100 L
N
2
O
4
films grown on 300 ML of ASW. The signal increases with increasing thickness,
indicating that absorption within the film hasn’t been saturated in the range of thick-
nesses studied.
72
4.3.4 Laser Energy
Laser energy may seem like a trivial trend to examine as logic would dictate that
with higher fluence, the amount of desorbing species would increase until the absorp-
tion is saturated. This trend was indeed verified, as shown in Figure 4.7.
Figure 4.7: Sum of peak areas for NO
+
and NO
2
+
signals from laser ablation of 80 L
N
2
O
4
dosed on 300 ML H
2
O. The signal generally increases with increasing laser en-
ergy.
Besides the most obvious result, there are some other trends that result from in-
creased laser energy. One is the appearance of a very small peak at mass 18 that corre-
sponds to H
2
O
+
. This is significant as it shows that as more energy is introduced to the
film, more mixing occurs at the ASW/N
2
O
4
interface. This allows for a small amount
of the ASW underlayer to escape with the hot N
2
O
4
fluid.
Another trend of note is the change in the R
N
value with increasing energy. As
discussed previously, R
N
is a way of characterizing the fraction of material leaving the
surface as N
2
O
4
or NO
2
. Representative R
N
values for the laser energies studied are
summarized in Table 4.1. For all laser energies used with bare N
2
O
4
films, R
N
is at
73
a maximum in the first extraction where resolved signal is detected (20 μs) and then
decays and levels out at later extractions. In contrast, the maximum R
N
value increases
with increasing energy, with R
N
= 4.0 for a laser energy of 2.5 mJ. This is above the
R
N
value of 3.1 calculated for the NO
2
gas phase spectrum. This reveals that there is a
higher proportion of NO
+
being detected than if a majority of the material leaving the
surface was NO
2
. The most likely source for the excess NO
+
is the presence of NO as a
photoproduct after the initial irradiation. NO production would logically increase with
increasing laser fluence, thus increasing the value of R
N
.
The average R
N
reached for the tail of the temporal profile also increases with in-
creasing laser energy, topping out at 1.8. This is likely due to increased production
of NO
2
as a photoproduct. At the lowest energy, the R
N
is almost the same as that for
N
2
O
4
thermal desorption, which shows that that make-up of the plume is mostly N
2
O
4
.
While it seems extremely unlikely that excited material leaves the surface after ablation
as N
2
O
4
, this may be indicative of N
x
O
y
that has undergone collisions in the dense fluid
that leaves the surface to reform the N
2
O
4
dimer, resulting in the long tail. Furthermore,
N
2
O
4
molecules that absorb radiation can heat surrounding molecules that get carried
out as the dimer (entrainment). With increasing laser energy, more NO
2
is formed.
Also, higher energy collisions may be less likely to result in recombination by the time
the plume enters the ionization region. This would result in a higher proportion of NO
+
and a higher R
N
value.
Lastly, increasing the laser energy results in the growth of a lumpy peak in the first
extraction. This peak was observed in several experiments and was initially dismissed
as sodium ions from light reflections. However, closer inspection of the signal revealed
that it had structure; analysis of this peak will be discussed in the following section.
The presence of this peak wasn’t always consistent, but followed the general trend of
increasing in size with increasing laser energy.
74
Table 4.2: R
N
values for different laser energies at different extraction times. The chang-
ing ratio indicates that different species are the result of the desorption process, such
as collisions; a consistent value would indicate all species are due to a fragmentation
process that occurs in the ionization region.
Laser Energy R
N
(mJ) 20 μs 50 μs 100 μs 150 μs 200 μs
0.5 1.7 1.4 1.1 1.1 1.1
1.0 2.0 1.5 1.3 1.2 1.2
1.5 2.8 2.0 1.6 1.6 1.5
2.0 3.6 2.4 1.8 1.8 1.8
2.5 4.0 2.4 1.9 1.8 1.8
4.3.5 Multiphoton Effects
As just discussed, a broad peak with structure was often detected in the first ex-
traction of experiments with N
2
O
4
films without any material on top. Because of the
timing set-up of our experiment, the first extraction contains material that has not been
subjected to electron bombardment. As such, all material detected in the first extraction
consist of molecules that are ionized during the laser ablation process. This was con-
firmed by conducting TOFMS experiments without the electron gun and still detecting
a fast, broad peak. Furthermore, the material must be leaving the surface at extremely
high energy to be detected in the first 10 μs after the laser pulse. Assuming a distance
between the surface and the nearest edge of the ionization region is 2.5 cm, the slowest
speed for this material would be 2500 m/s.
Our typical TOFMS timing scheme was not sufficient to resolve the fast ion peak
and instead was continuously extracting the material over the 3 μs extraction pulse.
This was addressed by shortening the extraction time to 1 μs and introducing a 2 μs
delay to select the “middle” of the fast ion plume. The ions present are NO
+
, NO
2
+
,
and the recombination product O
2
+
. NO
+
is the dominant species by far, illustrating
75
Figure 4.8: A representative resolved TOF spectrum of the fast ions from the first ex-
traction. The spectrum has been shifted to account for the 2 μs delay. The peaks are
labeled with probable species assignments.
that this material is the result of multi-photon absorption. As stated previously, when
N
2
O
4
absorbs a UV photon, it will split into two excited monomers. In order to pro-
duce ionized NO, the NO
2
monomers would need to absorb another photon in a 1+1
resonance-enchanced multiphoton ionization (REMPI)-type process. The presence of
this material speaks to the high fluence of the laser and how much energy is being
introduced to the film. Detection of O
2
+
in this initial region is also illuminating. As
a recombination product, it needs to form from collisions of O atoms released from
cracked N
2
O
4
or NO
2
. In order for such collisions to occur, the material leaving the
surface must be part of a dense fluid.
4.3.6 ASW Upper Layer Thickness
We were interested in how efficient the energy transfer would be between N
2
O
4
and ASW. As the ASW underlayer absorbed the energy with little change, the next
step was to make N
2
O
4
/ASW “sandwiches.” These consisted of an ASW film with a
76
layer of N
2
O
4
dosed on top, and then another layer of ASW. This would also give us
the ability to spatially select regions within the ASW to deposit energy, which was a
unique approach to investigating energy transfer in ASW. While there is the possiblility
of diffusion of N
2
O
4
into the porous ASW, it is most likely small enough to consider
N
2
O
4
a segregated layer. Furthermore, while reactions between ASW and N
2
O
4
can
occur, no product species (e.g. HNO
2
and HNO
3
[37, 38]) were detected by either
TOFMS or FTIR spectroscopy. This agrees with N
2
O
4
photodesorption experiments
conducted by other groups [26, 39].
These data provided a wealth of insight into transport, as well as morphological
changes in the ASW film. It was expected that increasing the thickness of the top ASW
layer would result in decreased overall signal until no more N
2
O
4
could escape because
its transport through the ASW film would be quenched. Instead, we were able to dose
2400 ML of ASW on top of a 80 L N
2
O
4
layer without seeing a noticeable decrease in
signal. Also, there was a large water signal with a long tail that was present with every
thickness tested. The thickest sandwiches will be discussed first.
For experiments carried out with a 2400 ML ASW top layer, very irregular behavior
was recorded in the temporal profile (Figure 4.9). Instead of the characteristic peak and
decay seen with exposed N
2
O
4
and thinner sandwiches, there are “bursts” of water
monomer out to >1 ms. These molecules cannot be leaving the surface at t = 0 and
traveling in a straight path to the detector because the kinetic energies would be too low;
this simple fact is illustrated by Figure 4.10. For instance, a water molecule detected
at 200 μs would have a kinetic energy <10 cm
1
if it left the surface at t = 0. As
temporal profiles in Figure 4.9 show significant signal out to 1200 μs, the signal cannot
be considered the tail of a warm distribution; another explanation is needed.
The only viable options to explain this result are: 1) the surface of the film loses
thermal contact with the MgO and material is able to evaporate after t = 0; 2) very
large water clusters are released and are detected as monomers due to evaporation
in the ionization region; or 3) the fluid entering vacuum is so dense that collisions
77
Figure 4.9: Nine first-pulse temporal profiles from a single 300 ML D
2
O/80 L N
2
O
4
/2400 ML H
2
O sandwich (1.0 mJ focused
radiation, configuration shown in cartoon, depicted on common scale). For clarity, water signal is shown in black and NO
+
x
signal is shown in red. The irregularity of the water signal, particularly at long times, is indicative of nonthermal processes.
The amount of signal at long times is considerable. It should also be noted that the NO
+
x
signal does not exhibit this behavior
and decays with time.
78
Figure 4.10: A water monomer’s kinetic energy compared to its flight time, assuming
the molecule left the surface at t = 0.
are frequent and the path of the material from the substrate to the ionization region is
increased due to following a circuitous route. Due to the very fast transfer of heat to the
substrate, the first option is considered implausible. The presence of large water clusters
was appealing as it has been observed by other groups working on similar systems [40,
41], but we saw no evidence of smaller clusters (including the protonated dimer) in
any of the experiments except for codeposition of N
2
O
4
with ASW. As we know we
are capable of detecting water clusters containing up to nine molecules [8], it seems
unlikely that we wouldn’t detect smaller clusters if we were releasing large clusters.
Large clusters also do not fit with the isotope exchange detected when introducing a
layer of D
2
O (see Chapter 5). Instead, we believe the third option is the best fit with
these data and will be presented as a potential mechanism.
When N
2
O
4
is subjected to UV radiation, it heats up quickly and will break into two
NO
2
molecules [9, 10]. In the case of a sandwich, this results in a trapped, very hot layer
that is under extreme pressure. The hot N
2
O
4
fluid mixes with the ASW immediately
above it and creates fractures and crevices in the porous ASW through which it can
79
now escape. As the N
2
O
4
shoots up the fractures, it scrapes the walls and mixes with
the surrounding water molecules. Once the fluid escapes, it can collide with the plumes
of material escaping from neighboring fissures.
The fissures created through this mechanism are robust, as evident when compar-
ing the first and second shot incident on the same spot. Example temporal profiles are
shown in Figure 4.11. Rows 1-3 are data collected from sandwiches with upper ASW
layer thicknesses of 600, 1200, and 2400 ML. The structure of the water monomer tem-
poral profile after radiation incident on a sandwich (Figure 4.11) has a lumpy shape.
The NO
+
x
peaks that arise from N
2
O
4
, NO
2
, and NO (henceforth referred to as N
x
O
y
)
have a different temporal profile form. The shapes of the temporal profile are revealing
about transport of species through the ASW film.
The first pulse on a fresh film has a large water signal with a delayed and small
NO
+
x
signal (black and red, respectively). The second pulse on the same spot results in
a much larger NO
+
x
signal and a small, delayed water signal. The first pulse temporal
profile reveals that the N
2
O
4
is impeded by the upper ASW layer, resulting in a slower
arrival time and fewer N
x
O
y
species escaping. As the fissures have formed during the
first pulse, N
x
O
y
can easily escape during the second laser pulse without impediment
as indicated by the fast arrival time and increased signal. Furthermore, less water
monomer is liberated from the film with subsequent shots. Comparing the second pulse
temporal profile with that of exposed N
2
O
4
, shown in Figure 4.5, reveals similarities in
shape. This indicates that N
x
O
y
species from the second shot (and those following) are
able to leave the film via an unobstructed path.
A quantitative comparison between first and second shot data can be made by ex-
amining R
N
values for both profiles, shown in Figure 4.13; these values also reflect
similarities. After the first few extractions, R
N
averages out close to unity for first pulse
data. This is much lower than the value for NO
2
, 3.1. We can conclude with some confi-
dence that N
2
O
4
is the dominant species leaving the film. The second shot R
N
averages
are slightly higher, which means that NO
2
has increased its contribution to the plume
80
Figure 4.11: Temporal profiles comparing first and second incident pulses for upper
H
2
O layer thickness: (1) 600 ML, (2) 1200 ML, and (3) 2400 ML. Each profile is the
average of nine individual profiles (1 mJ, 266 nm). Red peaks designate NO
+
and
NO
2
+
. Black peaks denote H
2
O
+
, OH
+
, and H
+
. All temporal profiles are on the same
ordinate scale. The lower ASW layer is 300 ML of D
2
O and the N
2
O
4
layer is 80 L
(sample composition shown at bottom).
81
Figure 4.12: Mass spectra representing the 100 μs extraction from each temporal profile
in Figure 4.11; refer to the caption of that Figure for composition details. The peaks in
spectrum (a) have been labeled with ion assignments for clarity.
82
leaving the substrate. This is especially apparent in the first 10 extractions, where R
N
values are very high. If we make the asssumption that NO
2
is produced from photodis-
sociation of gaseous N
2
O
4
during the length of the laser pulse (10 ns), this would point
to N
2
O
4
being located near the surface on subsequent pulses.
FTIR was also utilized to get a rough estimate of how much water was being re-
moved after ablation. Detecting material removal after a single pulse was not possible
with the current InSb detector and fluence conditions. To ensure that we were only
measuring depletion of the top ASW layer, D
2
O was used for the bottom spacer. The
laser beam waist was increased to 3 mm to ablate a sufficient area for FTIR measure-
ments. As such, laser fluence was approximately an order of magnitude smaller than
typical experimental conditions.
After 70 pulses on the center of the surface, there was a 14% decrease in the OH
peak (Figure 4.14a). It is also noteworthy that the OD peak is practically identical in
intensity before and after irradiation; this is consistent with our previously discussed
results investigating the role of the ASW underlayer. Examination of the individual
temporal profiles from this series of data show no evidence of a D
2
O
+
peak (m/q = 20).
There was also no observable increase in crystallinity of the ASW film, so morphological
changes appear to be limited to the growth of fissures and any changes in crystallinity
must be very small.
Another interesting feature of this data set was the presence of NO
+
x
signal in mass
spectra even in the 70
th
pulse. Instead of removing all absorbing species after a few
laser pulses, thin layers are removed. This is further evidence of the efficient cooling of
the film via the MgO surface. While a hot fluid is created throughout the depth of the
absorbing species when a film is irradiated, a gradient is quickly established that results
in the recondensation of material at the bottom of the absorbing layer. In this way, the
top of the N
2
O
4
layer is removed with each pulse, but material gets more difficult to
remove as the depth of the absorbing layer is probed.
83
Figure 4.13: Plots of R
N
for each extraction with good S/N from each temporal profile
in Figure 4.11; refer to the caption of that figure for composition details. After the initial
extractions, first pulse R
N
values (filled circles) level off around unity. Second pulse R
N
values (open triangles) follow the same qualitative trend but level off at higher values,
indicating the presence of N
2
O
4
in the surface region.
84
Figure 4.14: (a) Overlaid FTIR spectra of a 300 ML D
2
O/80 L N
2
O
4
/2400 ML H
2
O sand-
wich (configuration depicted in figure) before and after 70 shots of 6 mJ radiation. A
3 mm beam waist was used to ensure that a reasonable area was available for FTIR mea-
surements. The red line represents the spectrum of the undisturbed film; the black line
is the spectrum collected after irradiation. The OD stretch peak centered at 2435 cm
1
is virtually unchanged, but the OH peak centered at 3250 cm
1
is reduced by 14%.
(b) 2400 ML of H
2
O was annealed at 165 K for 10 minutes to induce crystallization and
is shown for comparison. The “after” spectrum in (a) did not show any observable
increase in crystallinity despite the prolonged exposure to laser radiation.
85
4.3.7 High Fluence vs. Low Fluence
Almost all experiments conducted in this lab involving irradiation of ASW or species
in ASW have used high-fluence beam arrangements. A majority of the experiments
discussed in this dissertation were carried out using a 50 cm CaF
2
lens to focus the
laser beam waist to <0.5 mm in conjunction with laser pulse energies of 1.0-3.0 mJ,
giving a laser fluence at the surface of1-3 J/cm
2
. In comparison, the few published
papers that have looked at N
2
O
4
photodesorption have used laser fluences almost a
factor of 10 smaller [18, 29, 39]. Because we were seeing a “long tail” feature that other
labs had not seen, the laser set-up was changed to accomodate low fluence. The CaF
2
lens was removed from the beam path, and the beam was trimmed to a 1 mm beam
waist using an iris. Due to the larger beam size, a grid of four spots was used instead
of nine to ensure that each laser pulse was incident on an undisturbed section of the
deposited film.
Figure 4.15: Averaged mass spectrum from extractions 2-11 (20-110 μs) of data from
a low fluence experiment (0.2 J/cm
2
) with the same film composition as Figures 4.4b
and 4.5 (300 ML ASW/80 L N
2
O
4
). The difference in the NO
+
/NO
2
+
ratio compared to
Figure 4.5 is striking.
86
The first comparison must be made between the R
N
ratios for high fluence vs. low
fluence. Low fluence experiments have a much higher population of NO
2
+
, as shown
in Figure 4.15. Another noticeable feature in the low fluence data is impact of an ASW
spacer. With low fluence, the signal increase is even more significant at around 600%.
This strengthens the argument that ASW is acting as an effective insulator to the thermal
properties of MgO.
Finally, an ASW/N
2
O
4
sandwich was also subjected to low fluence. We were able to
detect small signal, but it was very slow and was never the result of the first laser pulse
incident on a film. This was revealing about the nature of N
x
O
y
species’ escape from
under the top ASW layer. As discussed in the previous section and further fleshed out
in Section 4.4, the proposed mechanism for escape is travel of hot N
x
O
y
species through
ASW channels and out into vacuum. In the low fluence situation, this is clearly not hap-
pening after one pulse. Instead, the N
2
O
4
is only able to travel a certain distance up
through the ASW layer before recondensing on walls and/or forming plugs in the chan-
nels, preventing escape. Then, on subsequent shots, N
2
O
4
is able to escape because it is
closer to the surface of the ASW. Tellingly, the shape of the temporal profile resembled
that of exposed N
2
O
4
such as in Figure 4.4, which follows from the proposed mecha-
nism. Furthermore, this indicates that there is a fluence cut-off for N
2
O
4
escape from
sandwiches; determining this boundary requires further experimentation.
4.3.8 N
2
O
4
Codeposition
Past experiments in this lab involved a codeposition of ASW and another species
[8]. However, ASW was the absorbing species, resulting in a relatively uniform initial
distribution of energy throughout the irradiated area. In those experiments, we typ-
ically saw protonated clusters along with protonated monomer. In constrast, clusters
had not been detected in experiments using N
2
O
4
as the absorbing species. We were
also interested in seeing if N
2
O
4
signal would decrease due to isolated NO
2
molecules
depositing at lower concentrations.
87
A representative mass spectrum is shown in Figure 4.16. With codeposition of a1:2
ratio of N
2
O
4
to ASW (based on Langmuirs), the first evidence of protonated clusters
was detected. The ablation mechanisms laid out by Perez and Lewis [42] provide a
likely explanation for this result. When the N
2
O
4
is deposited in a single layer, it results
in a high-energy environment that should only give monomers. The small presence
of clusters in the codeposition experiments indicates that a slightly “softer” desorption
process is occurring. Also, decreasing the ratio to 1:9 still gave significant N
x
O
y
signal
after irradiation. This indicates that H
2
O does not significantly inhibit dimer formation;
indeed, dimerization may be encouraged within the pores of ASW.
It is worthwhile to note that products due to chemical reaction, such as HNO
3
, are
still not detected despite codeposition allowing for maximum interaction between N
2
O
4
and H
2
O. This is apparent from the absence of any peak at mass 62. We tentatively
conclude that if chemical reactions occur within the film, the concentration of product
species is small.
4.4 Discussion
Little previous work has been done on photodesorption of ASW via a dopant
molecule. One system that was looked at in depth by the group of M. R. S. McCoustra
was that of UV-irradiated benzene and water [43–45]. Their set-up included a sapphire
surface and thinner layers of material (several nm thick compared to hundreds in the
present work). Based on their results, they defined three processes for photo-induced
desorption: direct absorbate-mediated, indirect absorbate-mediated, and substrate me-
diated (this process was due to their sapphire substrate absorbing slightly in the UV).
As MgO does not absorb in the UV , we can ignore the last process and focus on the first
two. In the case of direct absorbate-mediated desorption, it is simply defined: material
that is excited by the incident radiation desorbs. When applied to our system, N
2
O
4
is
desorbed via this mechanism as the 266-nm radiation excited the afore-mentioned elec-
tronic transition. Indirect absorbate-mediated desorption, on the other hand, is when
88
Figure 4.16: Expanded mass spectrum of 80 L N
2
O
4
codeposited with 240 ML of ASW
and irradiated with focused 1.0 mJ radiation. Five first-shot experiments from the film
have been averaged to improve S/N. While small, clusters up to the protonated water
tetramer were visible. No protonated species were detected in any of the experiments
where N
2
O
4
was deposited as a discrete layer. It is also important to note that there is
no signal at m/q = 62, which would indicate the presence of NO
3
+
.
89
energy is transferred from the excited molecules to molecules nearby in the film. This
process describes the H
2
O desorption that we see (though the interpretation of this is
much more complicated than this definition implies).
Putting together a model for the data presented in this chapter that reconciles the
various observables into a cohesive whole was a daunting task. It is important to begin
with a summary of the key findings from the previous sections. Heating N
2
O
4
multilay-
ers grown on a 100 K MgO(100) substrate with 266-nm radiation results in desorption
of material. While the 266-nm irradiation of gaseous N
2
O
4
results in photolysis and
the production of NO
2
with translational and internal energy, the dense environments
of our systems favor recombination accompanied by the evolution of heat on a short
time scale. Therefore, N
2
O
4
and its surrroundings are heated by the laser pulse. The
efficient removal of material is increased when the N
2
O
4
layer is grown atop an ASW
layer because the ASW acts as a thermal insulator between the absorbing material and
the substrate. Exposed N
2
O
4
has a qualitatively different temporal profile than buried
N
2
O
4
, which can be examined quantitatively by looking at the value of R
N
as defined
in Section 4.3.1.
At high fluence ( 1 J/cm
2
), the presence of an ASW film on top of the N
2
O
4
doesn’t
significantly impede the desorption of N
x
O
y
species. Despite increasing the upper ASW
layer to thickesses of 2400 ML, desorption of N
2
O
4
was robust. It was determined that
the NO
+
x
species detected orginated from N
2
O
4
due to the small R
N
values that resem-
bled that of the pure N
2
O
4
mass spectrum. Accompanying N
2
O
4
desorption was water
monomer, which differed from previous experiments conducted in our lab that detected
protonated water clusters. The total signal of water monomer increased with increas-
ing ASW thickness, which was counter to expectations. The lack of protonated clusters
points to a different mechanism for material release into vacuum. Our system also
differs from the widely studied “molecular volcano” peak that is also associated with
ejection of molecules from ASW [6, 46]. The latter occurs after homogenous thermal
heating of ASW and embedded species until the crystallization temperature is reached.
90
The experiments presented in this dissertation involve extremely rapid, pulsed heating
of a layer of material on or embedded in ASW, which triggers transport of hot fluid to
vacuum.
Finally, low fluence (<0.2 J/cm
2
) radiation instigates desorption of N
x
O
y
species,
but only after 2 or 3 laser pulses incident on the same area. The set of data taken as
a whole reveals that excited material escapes to vacuum via cracks in the ASW film.
A cohesive model is described below, starting with a discussion of the unusually long
time scales measured.
4.4.1 Time Scales
It was surprising when we first measured significant signal at 300 μs; it was even
more surprising when the time scale was expanded to 1.2 ms and signal was still con-
siderable. The “lumpy” character of the water monomer temporal profile for thick
sandwiches was also intriguing in its variablility in structure, as well as its reproducibly
lumpy character. It was clear that H
2
O molecules couldn’t have enough energy to leave
the surface at t = 0 and arrive in extractions after even 50 μs. For a water molecule
arriving in the 1.2 ms extraction assuming no collisions, the kinetic energy would be
0.33 cm
1
. Another possibility would be that water monomer leave the film at t > 0.
This desorption mechanism would require a loss of thermal contact with the 100 K
MgO substrate, which was determined to be unlikely after heat calculations were con-
ducted [47]. Finally, the release of large clusters could produce bursts of signal at longer
times because they will have long transit times and may evaporate in the beam. How-
ever, the complete absence of even the protonated dimer in any mass spectra (excepting
the codeposition experiments) refutes this mechanism. Instead, we propose that wa-
ter molecules undergo collisions before entering the ionization region. Many collisions
would result in a complicated flight path from the surface to the ionization region,
which is detected as a long flight time. This is consistent with a high fluid density
regime and a high density of fissures, which is expanded on in the next section.
91
4.4.2 Fissures
An intriguing result is the temporal profile of N
x
O
y
signal in a sandwich film com-
paring the first laser pulse to subsequent pulses on the same location. The profile for
the first pulse incident on sandwiches with an upper layer thickness >600 ML shows
a delay in the release of material, with a signal maximum from N
x
O
y
at 100-150 μs.
Such a distribution is logical, as the impeding layer of ASW would sap energy from the
escaping N
x
O
y
species as they create pathways to escape to vacuum. The formation of
fissures is favored because ASW must maintain its macroscopic column density. The
profile of N
x
O
y
from the second and subsequent laser pulses resembles that of exposed
N
2
O
4
films upon irradiation. This demonstrates that once fissures are formed, they do
not collapse or become plugged. As such, they are accessible pathways for transport
of the remaining N
2
O
4
to vacuum when excited by subsequent pulses incident on the
same spot. Fissures must also be relatively direct pathways from the N
2
O
4
layer to
vacuum, as too many twists and turns would result in N
2
O
4
freezing out on the walls.
The hot fluid released from fissures can collide with material released from neigh-
boring fissures (Figure 4.17). The density of fissure openings per unit area in combina-
tion with the density of the fluid released necessitates the occurance of collisions. We
can assume that the hot fluid traveling through fissures is dense as low density fluid
would easily freeze to the walls and be unable to escape to vacuum in any significant
amount. Collisions above the film will change the trajectories of species, giving a spa-
tial inhomogeneity in the ionization region. Such an evironment would explain the
characteristic lumps seen at long times in temporal profiles.
Increasing the thickness of the upper ASW layer does not significantly impede the
escape of N
x
O
y
species. However, it does increase the amount of water that enters
vacuum upon irradiation. This follows from the fact that the heated fluid must travel
over a longer fissure pathway. As the fluid travels through fissures, it scrapes the walls
and frees water monomer.
92
Figure 4.17: A cartoon depiction of how material expelled from neighboring fissures
may interact.
Low fluence experiments still exhibit material release, but require two or more laser
pulses to instigate release. While this clearly points to a fluence threshold that needs to
be achieved in order for material to escape the film in one pulse, the fact that material
can still be desorbed points to arrested transport within the ASW after termination of
the laser pulse. Otherwise stated, hot N
2
O
4
moves upward in the film, but does not
have enough energy to complete the journey to vacuum and recondenses at a higher
level in the film. A second (or third) laser pulse allows for continued transport and
escape of material that has already moved nearer the surface of the ASW layer.
Hot fluid is able to reach vacuum with a single laser pulse at high fluence. Due to the
high thermal conductivity of MgO, cooling efficiency is high enough that any material
left in the fissures after the cessation of a laser pulse would quickly recondense in the
film. Based on the assumption that material that escapes the film must do so before the
end of the laser pulse, a rough calculation can be made of minimum fluid speed. Given
a 1000-nm thick ASW layer, hot fluid travelling through fissures must have a speed
of 100 m/s or more in order to escape the film. With thinner samples, lower speeds
are necessary for transport from the buried layer to the surface and so more N
x
O
y
can
93
escape to vacuum. This is consistent with the observed decrease in N
x
O
y
signal for
sandwiches with increasing upper ASW layer thickness.
4.5 Conclusions
An experimental technique has been presented for studying energy transfer in lay-
ered films of ASW and N
2
O
4
that gives insight into molecular transport. 266-nm fo-
cused radiation excited an electronic transition in N
2
O
4
which quickly degraded to
heat. Uncovered N
2
O
4
was desorbed with ease; covered N
2
O
4
in a “sandwich” config-
uration created structural changes in the upper ASW layer and induced the ejection of
water monomer along with release of N
x
O
y
species.
The nature of the data prevents the development of a definitive, quantitative model.
However, a qualitative understanding of the results is presented that reconciles the ex-
perimental trends uncovered in this study. Building a robust quantitative model would
require further investigations in the laboratory, coupled with computational simula-
tions, which is left as future work.
To summarize, the model presented in this chapter is as follows: Laser heating of
a buried N
2
O
4
layer creates fissures in the upper ASW layer to enable transport to
vacuum. Dense fluid escapes through these fissures without freezing on the walls, re-
sulting in jets of material spewing out of the film. The creation of fissures is irreversible,
as evidenced by continued removal of N
2
O
4
after the first laser pulse. Increasing the
thickness of the upper ASW layer does not significantly impede the escape of N
x
O
y
species; it does increase the amount of water that enters vacuum upon irradiation due
to greater fissure area. Long tails of material detected at unphysical times after the
initial laser pulse are the result of molecular collisions above the film surface. Such
collisions occur due to the prevalence of fissures combined with the high density of the
escaping fluid.
94
To probe these morphological changes more effectively, some equipment changes
will need to be implemented for increased dectection abilities. Furthermore, generat-
ing fissures of known dimensions would be valuable. Preliminary experiments using
gold nanoparticles have been attempted and may help progress understanding of this
system.
95
Chapter References
[1] Burton, E.; Oliver, W. Nature 1935, 135, 505–506.
[2] Angell, C. Annu. Rev. Phys. Chem. 2004, 55, 559–583.
[3] Hama, T.; Watanabe, N. Chem. Rev. 2013, 113, 8783–8839.
[4] Burke, D.; Brown, W. Phys. Chem. Chem. Phys. 2010, 12, 5947–5969.
[5] May, R.; Smith, R.; Kay, B. J. Phys. Chem. Lett. 2012, 3, 327–331.
[6] May, R.; Smith, R.; Kay, B. J. Chem. Phys. 2013, 138, 104501.
[7] May, R.; Smith, R.; Kay, B. J. Chem. Phys. 2013, 138, 104502.
[8] Rebolledo–Mayoral, O.; Stomberg, J.; McKean, S.; Reisler, H.; Wittig, C. J. Phys.
Chem. C 2012, 116, 563–569.
[9] Bach, W.; Breuer, H. Faraday Discuss. Chem. S. 1974, 58, 237–243.
[10] Kawasaki, M.; Kasatani, K.; Hiroyasu, S.; Shinohara, H.; Nobuyuki, N. Chem.
Phys. 1983, 78, 65–74.
[11] Hurford, A.; Helfenstein, P .; Hoppa, G.; Greenberg, R.; Bills, B. Nature 2007, 447,
292–294.
[12] Porco, C.; DiNino, D.; Nimmo, F. Astron. J. 2014, 148, 45.
[13] Hawkins, S.; Kumi, G.; Malyk, S.; Reisler, H.; Wittig, C. Chem. Phys. Lett. 2005,
404, 19–24.
[14] Korolik, M.; Suchan, M.; Johnson, M.; Arnold, D.; Reisler, H.; Wittig, C. Chem.
Phys. Lett. 2000, 326, 11–21.
[15] Andrews, B.; Anderson, A. J. Chem. Phys. 1981, 74, 1534–1537.
[16] Loparo, J.; Roberts, S.; Tokmakoff, A. J. Chem. Phys. 2006, 125, 194521.
[17] Vandaele, A.; Hermans, C.; Simon, P .; Carleer, M.; Colin, R.; Fally, S.; Mérienne,
M.; Jenourvrier, A.; Coquart, B. J. Quant. Spectrosc. Radiat. Transfer 1998, 59, 171–
184.
[18] Dixon-Warren, S. J.; Jackson, R.; Polanyi, J.; Rieley, H.; Shapter, J.; Weiss, H. J.
Phys. Chem. 1992, 96, 10983–10994.
[19] Hasselbrink, E.; Jakubith, S.; Nettesheim, S.; Wolf, M.; Cassuto, A.; Ertl, G. J.
Chem. Phys. 1990, 92, 3154–3169.
96
[20] Mélen, F.; Pokorni, F.; Herman, M. Chem. Phys. Lett. 1992, 194, 181–186.
[21] Hisatsune, I.; Miller Jr., P . J. Chem. Phys. 1963, 38, 49–54.
[22] Tyndall, G.; Stedman, K.; Schneider, W.; Burrows, J.; Moortgat, G. J. Photochem.
1987, 36, 133–139.
[23] Ogai, A.; Qian, C.; Iwata, L.; Reisler, H. Chem. Phys. Lett. 1988, 146, 367–374.
[24] Bai, Y.; Ogai, A.; Qian, C.; Iwata, L.; Segal, G.; Reisler, H. J. Chem. Phys. 1989, 90,
3903–3914.
[25] Qian, C.; Ogai, A.; Iwata, L.; Reisler, H. J. Chem. Phys. 1990, 92, 4296–4307.
[26] Rieley, H.; McMurry, D.; Haq, S. Faraday Trans. 1996, 92, 933–939.
[27] Russ, F. Z. Phys. Chem. 1914, 82, 217–222.
[28] Hisatsune, I. J. Phys. Chem. 1961, 65, 2249–2253.
[29] Rieley, H.; Colby, D.; McMurry, D.; Reeman, S. J. Phys. Chem. B 1997, 101, 4982–
4991.
[30] Linstrom, P ., Mallard, W., Eds. NIST Chemistry WebBook, NIST Standard Refer-
ence Database Number 69., http://webbook.nist.gov (accessed 05/04/2015).
[31] Y.K., K.; Rudd, M. Phys. Rev. A 1994, 50, 3954–3966.
[32] Slack, G. A. Phys. Rev. 1962, 126, 427–441.
[33] Andersson, O.; Inaba, A. Phys. Chem. Chem. Phys. 2005, 7, 1441–1449.
[34] English, N.; Tse, J.; Gallagher, R. Phys. Rev. B 2010, 83, 092201.
[35] English, N.; Tse, J. Phys. Rev. B 2011, 83, 184114.
[36] Yang., R.; Gudipati, M. J. Chem. Phys. 2014, 140, 104202.
[37] Kim, S.; Kang, H. J. Phys. Chem. Lett. 2010, 1, 3085–3089.
[38] Lignell, H.; Varner, M.; Finlayson-Pitts, B.; Benny Gerber, R. J. Chem. Phys. 2012,
405, 52–59.
[39] Susa, A.; Koda, S. J. Atmos. Chem. 2003, 43, 197–217.
[40] Bar-Nun, A.; Dror, J.; Kochavi, E.; Laufer, D. Phys. Rev. B 1987, 35, 2427–2435.
97
[41] Thrower, J.; Burke, D.; Collings, M.; Dawes, A.; Holtom, P .; Jamme, F.; Kendall,
P .; Brown, W. A.; Clark, I.; Fraser, H.; McCoustra, M.; Mason, N.; Parker, A.
Astrophys. J. 2008, 673, 1233–1239.
[42] Perez, D.; Lewis, L.; Lorazo, P .; Meunier, M. Appl. Phys. Lett. 2006, 89, 141907.
[43] Thrower, J.; Burke, D.; Collings, M.; Dawes, A.; Holtom, P .; Jamme, F.; Kendall, P .;
Brown, W.; Clark, I.; Fraser, H.; McCoustra, M.; Mason, N.; Parker, A. Astrophys.
J. 2008, 673, 1233–1239.
[44] Thrower, J.; Collings, M.; McCoustra, M.; Burke, D.; Brown, W. A.; Dawes, A.;
Holtom, P .; Kendall, P .; Mason, N.; Jamme, F.; Fraser, H.; Clark, I.; Parker, A. J.
Vac. Sci. Technol., A 2008, 26, 919–924.
[45] Thrower, J.; Abdulgalil, A.; Collings, M.; McCoustra, M.; Burke, D.; Brown, W.;
Dawes, A.; Holtom, P .; Kendall, P .; Mason, N.; Jamme, F.; Fraser, H.; Rutten, F. J.
Vac. Sci. Technol., A 2010, 28, 799–806.
[46] Smith, R.; Huang, C.; Wong, E.; Kay, B. Phys. Rev. Lett. 1997, 79, 909–912.
[47] McKean, S. M. Energy Transfer in Amorphous Solid Water: Light-Mediated Ex-
pulsion of N
2
O
4
Guest Molecules., Ph.D. Thesis, University of Southern Califor-
nia, 2015.
98
Chapter 5
Introducing a Third Species: Molecular Transport
and Isotope Exchange
There are inherent limitations to the experimental approach described in this dis-
sertation. We were unable to measure the excitation state of desorbed species, or to
investigate structural changes on the surface except with FTIR spectroscopy. FTIR spec-
troscopy itself is limited in the types of changes we can detect: the amount of material
removed, bulk changes, and species with strong peaks in the 1850-7000 cm
1
detection
range. This last limitation meant that we had to focus on changes in the OH stretch
signal, as N
2
O
4
only has one weak combination band that fell in the detection region
and was mainly used as an indicator of the presence or absence of the species itself.
To address the subtleties of transport within the ASW sandwiches, a set of molecules
were introduced to profile the effect of energy transfer at different depths during the
irradiation process.
5.1 Introduction
Building a model for molecular transport in amorphous solid water has proven to
be a difficult task due to the variety of morphologies of ASW and its complex interac-
tions with guest molecules [1–3]. Work done by Kay and coworkers has shown that a
thermally-induced “molecular volcano,” where guest species are energetically ejected
99
from an ASW matrix upon heating, is due to top-down growth of fissures during crys-
tallization [4]. These cracks provide a pathway from the pores containing the trapped
species to vacuum; material that is not along the path of these fissures remains trapped
in the crystalline ice until it can co-desorb with the water bulk at a higher temperature.
Another method of transport that still requires study is the “bottom-up” pathway.
The groups of Kay [5] and Bar-Nun [6] have also investigated this phenomenon, which
is the mechanism attributed to bursts of material that are released below crystallization
temperature upon heating. The tentative explanation put forth is the structural collapse
of ASW due to extreme pressure of the escaping material during heating. The work
described in the previous chapter presented a model where laser-heating of a buried
stratum induced catastrophic changes in the upper ASW layer with resulting fissures al-
lowing the release of a hot fluid into vacuum. However, working with only two species
gave somewhat limited insight into how energy is transferred through the film during
this process. This chapter focuses on studies where a third species was introduced to
better probe energy transfer in ASW sandwiches. By placing a thin film of another
material in the sandwich, depth profiling can be carried out.
Carbon dioxide and acetylene are attractive probe molecules because they are read-
ily available. Furthermore, both molecules are known to exist in the ISM and may
interact with ASW on the surface of dust grains in molecular clouds [7]. Carbon diox-
ide as a guest molecule has been utilized previously by this lab in multiple studies [8, 9]
and most recently was found to be preferentially released when the surrounding ASW
matrix was disturbed with laser radiation [10]. This mobility can be exploited in order
to learn more about structural changes within the system studied herein.
Alternatively, deuterated water (D
2
O) is useful as an isotopically marked species
that can be differentiated from H
2
O but does not affect the properties of a film sig-
nificantly. However, these two types of water can undergo proton exchange to form
HDO:
100
2 H
2
O+ 2 D
2
O! H
2
O+ 2 HDO+ D
2
O. (5.1)
The Grotthuss mechanism explains this process through proton-hopping; excess pro-
tons from the autoionization of water travel along a cluster of water molecules by “hop-
ping” from molecule to molecule [11]. In the solid phase, isotope exchange is not as
facile but follows similar rules. It is mediated by defects that migrate through the solid
that originate as excess protons or Bjerrum defects [12]. Bjerrum defects are a distortion
of a normal hydrogen bond and exist as two varities: D defects have two protons, L
defects have no protons. Also, the rate of H/D exchange can be increased through the
introduction of H
+
or OH
–
[13–15]. If exchange occured under standard experimental
conditions during sample formation, it would be useless as a probe molecule as D
2
O
and H
2
O could no longer be considered discrete layers. However, exchange is not ob-
served on experimental time scales below 120 K [16]. As experiments are carried out
at 90-100 K, isotopic exchange can act as an indicator of film mechanisms and changes
during and after ablation that encourage diffusion, collisions, etc.
D
2
O and H
2
O interactions are also of interest to the astrochemical community due
to prevalence of water in the ISM. Furthermore, isotopic composition observed in inter-
stellar water compared to terrestrial water has been investigated in order to understand
the origin of water on Earth [17, 18]. In this work, D
2
O is utilized as a probe species
in ASW under astronomically relevant conditions and the implications of isotope ex-
change following irradiation with UV radiation are discussed.
5.2 Experimental
A brief summary of experimental methods will be given, followed by details of
changes made for this set of experiments. Refer to the experimental section in the
previous chapter (Section 4.2) for a detailed description of the procedure used for data
collection. 266-nm radiation was used to irradiate layered ASW/N
2
O
4
films containing
101
a third molecular species (CO
2
, C
2
H
2
, or D
2
O) grown on a MgO(100) substrate kept at
90 or 100 K. The previously established convention of using Langmuirs (L) for N
2
O
4
thickness and monolayers (ML) for all other species will be continued in this chapter.
Material leaving the film upon irradiation was collected using a pulsed TOFMS, which
collected a full mass spectrum every 10 μs. A grid of nine spots were irradiated on
each film, and the temporal profiles obtained from each could be averaged together to
increase S/N.
5.2.1 CO
2
and C
2
H
2
Films with CO
2
and C
2
H
2
were grown at90 K; all other films were grown at 100 K.
Background dosing was used to deposit thin layers of CO
2
or C
2
H
2
(100 ML) at different
depths in ASW/N
2
O
4
sandwiches, depicted in Figure 5.1.
Figure 5.1: Cartoon depiction of different sandwich configurations using a third species
as a probe (black represents the MgO substrate). Typical ASW sandwiches are de-
posited in layers, with layers of CO
2
or C
2
H
2
deposited at different depths within a
film. Thicknesses used were 300 ML for the bottom H
2
O spacer, 80 L for the N
2
O
4
layer, and 100 ML of CO
2
or C
2
H
2
. The top H
2
O layer was 200 ML [consisting of two
separate 100 ML sections for configuration (3)].
5.2.2 D
2
O
Split films of H
2
O and D
2
O were grown via background deposition as the top layer
of ASW/N
2
O
4
sandwiches at100 K. The order of the layers was swapped per ex-
perimental needs. FTIR spectroscopy was also used to inspect H/D exchange. After
102
an initial background spectrum was collected, layers of H
2
O and D
2
O were deposited
by background deposition while the substrate was in Tier 1 (refer to Section 3.1 for a
diagram). Spectra were then collected after heating the film to a maximum of 160 K for
1.5 minutes.
5.3 Results
The first set of data presented are preliminary results that need more exploration
in future experiments. It is important to note that CO
2
and C
2
H
2
experiments were
conducted before a re-calibration of the ion gauge. Also, laser energy was adjusted
using an iris before installation of a telescope to collimate the beam, which may have
resulted in a difference in beam diameter incident on the surface. As such, thicknesses
of material are not comparable with other experiments discussed in the rest of this
dissertation, and area of the surface that was irradiated was not as well regulated.
However, the tight focusing of the beam ensures that area changes were likely minimal.
It can be assumed that experiments conducted at the same laser energy are comparable
to each other. Despite this, the data are worth discussion due to some intriguing trends.
Next, isotopic data will be discussed, which is of good quality and displays interesting
characteristics.
5.3.1 Depth Profiling With CO
2
and C
2
H
2
Due to our lab’s experience with CO
2
, it was chosen as the first candidate for a probe
species. We had seen previously that it was highly mobile at low temperatures and
would be preferentially released from ASW if there was any significant morphological
change [10]. Therefore, it was an attractive choice because any disturbance in the ASW
morphology that would produce pathways to vacuum could be detected as CO
2
signal.
Films were dosed with a probe layer of CO
2
in each of the three configurations
shown in Figure 5.1 and then irradiated with focused 266-nm radiation. As expected,
103
almost no CO
2
was liberated from the ASW underlayer when irradiating films in config-
uration (1). However, the second shot did show an increase in CO
2
signal (Figure 5.2).
This is an intriguing result that points to structural change occuring in the bottom ASW
spacer. First, it is possible that there are slight morphological changes in the ASW
spacer that connect existing pores and provide a pathway for the CO
2
to escape during
the second laser pulse. Second, enough energy may enter the ASW/CO
2
spacer that
the CO
2
can diffuse up through the impeding ASW (and remaining N
2
O
4
) on the first
pulse but freezes before reaching vacuum. Upon the reintroduction of energy via a
second pulse, CO
2
becomes mobile again and a fraction is able to escape the film due
to its proximity to vacuum.
Comparing configuration (2) and configuration (3) provided further insight into
material removal processes. More CO
2
was detected when it was dosed on top of the
N
2
O
4
layer than when it was sandwiched in the upper ASW layer. Furthermore, there
is a noticeable difference in the shape of the CO
2
+
temporal profile when comparing
configuration (2) and configuration (3) (Figure 5.3). When directly above the N
2
O
4
,
the CO
2
+
signal tracks the NO
+
x
signal closely, with an apparent peak in signal in the
earliest extractions. In contrast, configuration (3) results in CO
2
+
signal that peaks
around 100 μs, and tracks more closely with the H
2
O
+
signal. These differences point
to two different pathways to material removal: mixing with the hot N
2
O
4
layer and
“scraping” of the walls as the heated fluid escapes. When CO
2
is dosed next to the
N
2
O
4
it mixes with the hot fluid and comes spewing out along with it via entrainment,
whereas CO
2
buried in the upper H
2
O layer is only removed as hot fluid passes by
the stratum and scrapes away material. The signal size difference indicates that mixing
with hot N
x
O
y
species is a more effective means of escape from the film compared to
scraping the walls of the fissures and subsequent entrainment.
This series of experiments was repeated with acetylene as the guest species. The
initial goal was to test whether or not the results were CO
2
-specific. However, acetylene
ended up providing another unique set of data. When acetylene was buried under the
104
Figure 5.2: Mass spectra averaged from extractions 2-21 where CO
2
is underneath
the ASW/N
2
O
4
sandwich [configuration (1) in Figure 5.1]. Mass spectrum (a) is af-
ter the first laser pulse, while (b) is from the second pulse incident on the same location
(2.0 mJ); CO
2
+
peaks are boxed in red for clarity. In (a), the very small peak at mass
44 shows how little CO
2
escapes the film. This signal increases significantly after the
second pulse on the same spot, indicating a morphological change that allows for CO
2
to move upward through the film.
105
Figure 5.3: Temporal profiles comparing CO
2
+
signal in configuration (2) versus (3)
(refer to Figure 5.1). Trace (a) is the result of configuration (2); trace (b) is due to
ablation of configuration (3). For both temporal profiles, data was averaged for the first
incident shot on 9 spots on the film (2.0 mJ). Signal due to H
2
O, N
x
O
y
and CO
2
are
designated by black, red, and blue, respectively. Because CO
2
+
is partially obscured by
NO
+
x
signal, a trace of its outline is drawn in blue for clarity. (a) When CO
2
is deposited
directly on top of N
2
O
4
, its temporal profile tracks with the NO
+
x
signal. (b) Embedding
CO
2
in the upper water layer results in a smaller signal when compared to (a), as well
as slower CO
2
+
signal that tracks more closely with the water. Taken together, the traces
indicate that material removal occurs via two different pathways: mixing with the hot
fluid and scraping of fissure walls.
106
ASW spacer, we obtained data similar to that of CO
2
; i.e., no C
2
H
2
escapes after the first
laser pulse (though signal was not detected in subsequent pulses, either). However, the
data differed significantly when the acetylene spacer was dosed on top of the N
2
O
4
layer. Whereas this configuration resulted in the largest signal of CO
2
+
, there was very
little C
2
H
2
+
present in the mass spectrum. This was true even with multiple laser
pulses incident on the same spot. Finally, the largest C
2
H
2
+
signal was detected when
it was sandwiched in the top ASW layer. It is worth nothing that the shape of the
C
2
H
2
temporal profile resembles that of CO
2
when it is in the same configuration,
strengthening the supposition of material removal from fissure walls.
We tried to address this strange, yet reproducible, result of C
2
H
2
“disappearing”
in configuration (2) by reducing the ASW spacer between the acetylene and the N
2
O
4
layer to see if there was a signal trend. It was found that C
2
H
2
+
signal is present even
when there is only 20 ML of ASW between it and the N
2
O
4
layer, but is significantly
reduced or disappears altogether with no spacer (Figure 5.4). There are a few possible
explanations for this observation. The obvious explanation is that the acetylene doesn’t
stick to the N
2
O
4
layer; however, this seems unlikely under experimental conditions.
Another possibility is that acetylene is reacting with the sea of hot N
2
O
4
ions and frag-
ments that form when the film is irradiated. This explanation is made more attractive
by the observation that C
2
H
2
signal is not detected after two or more laser pulses on
one spot when in configuration (1), which could point to C
2
H
2
being consumed in the
N
2
O
4
layer. Despite this, no possible product fragments (e.g. C
2
H
2
O
+
) were detected
in any experiments with C
2
H
2
. As such, this observation requires further study for
elucidation.
5.3.2 Switching to D
2
O
Because of the conflicting results from CO
2
and C
2
H
2
, D
2
O was chosen as the new
test species. Deuterated water is a good probe since it behaves similarly to H
2
O, and
isotope exchange is extremely slow at cryogenic temperatures [15]. As our films were
107
Figure 5.4: Waterfall plot of averaged mass spectra (8-9 spots, 20 extractions) from
experiments using C
2
H
2
in configuration (3) where the water spacer between C
2
H
2
and
N
2
O
4
ranged from 100 to 0 ML. The main fragments detected from C
2
H
2
are C
2
H
2
+
,
C
2
H
+
, and C
2
+
(indicated by the red arrow). Change in C
2
H
2
fragment signal is small as
the H
2
O spacer between N
2
O
4
and C
2
H
2
is decreased. However, the complete removal
of a water spacer results in a significant C
2
H
2
signal drop.
108
grown and kept at 100 K, H/D exchange does not occur without an external trigger.
This was verified under our experimental conditions via FTIR monitoring of a layered
binary ice of H
2
O and D
2
O. Isotope exchange results in the sharpening of central
features in both the OD and OH stretch peaks at 2400 and 3300 cm
1
, respectively
[15, 19]. Within the time frame of a typical irradiation experiment (10-20 minutes), no
exchange was detected. Even heating the film above the crystallization temperature
of 140 K didn’t result in measureable isotope exchange unless heating was maintained
for over a minute (Figure 5.5). This showed that if any HDO
+
was detected in TOFMS
experiments, it must occur during the irradiation process.
Figure 5.5: Overlaid FTIR spectra of a film initially grown at 100 K of 300 ML of H
2
O
deposited on 300 ML of D
2
O (red) that was then heated to 160 K for 1.5 minutes (black).
The appearance of a central structure in both peaks is indicative of HDO exchange
occuring; there may be some contribution from crystallization of the film, but the impact
should be small due to heating the film for <2 minutes. This shape change was not
detected at lower temperatures or when the sample was left to sit for 20 minutes at
100 K. The negative-peak distortion of the OD peak is due to background age.
Deuterated water was grown along with H
2
O in layered configurations as depicted
in Figure 5.6. When irradiated, HDO
+
was detected as a significant product, indicating
109
isotope exchange was occuring. The size of HDO
+
signal in Figure 5.6 implies that
extensive mixing is occuring as hot N
x
O
y
species are traveling through fissures. It is
important to note that assuming that material is removed as hot fluid scrapes the fissure
walls, H/D exchange can only happen in the upper half of the sandwich. Allowing for
such a short area for exchange to occur (approximately 140 nm for the spectra shown
in Figure 5.6) emphasizes the energetic nature of material removal and efficiency of
mixing during irradiation. It also discourages a model involving the release of large
clusters, as that would prohibit efficient isoptope exchange. Another interesting feature
of Figure 5.6 is how similar the HDO
+
signal is for both spectra despite the different
film configurations. This points to a process where material is removed along the entire
length of the fissure instead of a fraction of the fissure. Furthermore, it is evidence
against significant evaporation from the surface of the film as a mechanism, as there is
no appreciable change in signal when D
2
O is on top or underneath H
2
O.
Every mass spectrum detected from this set of experiments, with upper layer thick-
ness ranging from 200-1200 ML, showed some amount of isotope exchange. Also, the
placement of the D
2
O probe layer did not have a significant impact on the amount
of isotopic mixing observed, i.e., having D
2
O above or below the H
2
O layer did not
change the amount of HDO present. However, the amount of isotope exchange was
dependent on the thickness of the upper H
2
O/D
2
O layer. Because quantitative analysis
of isotopically mixed species is complicated by the impossibility of collecting a pure
HDO spectrum, a ratio of detected ions was used. The ratio chosen was HDO
+
/D
2
O
+
(values based on peak areas), as mass 19 and 20 can only originate from these species.
It should be noted that mass 19 can also be assigned to the protonated water monomer,
H
3
O
+
, but this species was not detected in any experiments with segretated N
2
O
4
lay-
ers. Therefore, it was assumed to have a negligible contribution to mass 19 signal.
This ratio was designated R
I
and used as an indicator of the percentage of mixing that
occured during irradiation of the film. Figure 5.7 is a plot of R
I
vs. the thickness of
D
2
O/H
2
O layer thickness. There is a clear decrease in R
I
as the thickness of D
2
O and
110
Figure 5.6: Representative mass spectra showing HDO
+
formed from isotopic scramb-
ing during the ablation process. The spectra are averaged from the first 20 extractions
yielding signal from first laser pulse temporal profiles (1.5 mJ). Mass peaks are labeled
in (a) for clarity. The film layer thicknesses for both samples (a) and (b) are: 300 ML
H
2
O, 80 L N
2
O
4
, and 200 ML each of H
2
O and D
2
O (configurations depicted in car-
toons).
111
H
2
O as increased. This trend points to less efficient mixing of species as the upper layer
increases in thickness.
Figure 5.7: Plot showing the inverse between R
I
and film thickness. R
I
values were
calculated for sandwich configurations like that depicted in Figure 5.6, comparing D
2
O
on top of a layer of H
2
O or underneath a layer of H
2
O. The thicknesses reported on the
x-axis refer to the thickness of the individual D
2
O and H
2
O layers (i.e, the top layer of
the sandwich is twice as thick). There is a clear decrease in R
I
as the thickness of the
H
2
O and D
2
O layers increases.
5.4 Discussion
Three different molecules were used as probes to better understand the effect of
extreme heating of a buried stratum in amorphous solid water. The significance of
the results will be discussed in the order in which they were presented, and then a
summary of implications for the proposed model will be given.
Carbon dioxide and acetylene proved to be important for gaining insight into heat
transfer in ASW, despite some differences in behavior. As previously established, when
112
buried N
2
O
4
is heated with UV radiation, it becomes an extremely hot fluid that frac-
tures the ASW layer to form fissures through which material can escape. Material
immediately surrounding the N
2
O
4
layer will become heated and mixed with the hot
fluid produced, but the extent of this mixing is inherently limited due to poor thermal
conductivity within the film. This is deduced from the different behavior of CO
2
when
placed at different depths within the film. Carbon dioxide that is in direct thermal con-
tact with the hot N
2
O
4
layer is able to escape along with the hot fluid through fissures
with relative ease compared to CO
2
embedded in the upper ASW layer, as is appar-
ent from the change in overall signal and temporal profile shape. This contrasts with
the “top-down” process described by May and coworkers [4], where material closer
to the surface is preferentially released as cracks propagate downward from the vac-
uum interface. It may be expected that even though the “bottom-up” mechanism of
N
x
O
y
expulsion from fissures is robust, the mobility of CO
2
would mitigate the effect
on signal size (i.e. depth would play a more dominant role). However, CO
2
escapes
more efficiently with the hot N
x
O
y
despite the increased distance of travel to vacuum.
This reinforces how energetic a process is occuring during irradiation with 266-nm laser
pulses.
Introducing D
2
O as a probe species gave additional insight to the molecular trans-
port model. However, data interpretation was complicated by the fact that it is difficult
to deconvolute the populations of each isotopic parent species as the cracking pattern
of pure HDO is impossible to obtain due to the statistical populations of 1:1:2 that exist
when equal parts H
2
O and D
2
O are mixed. Despite this, some qualitative statements
can be made about these data.
It was found by analysis of FTIR spectra taken over experimental time scales that
isotope exchange does not occur within our samples at 100 K. Therefore, it can only
occur either in the film during the laser heating and subsequent release of N
2
O
4
or
directly above the film surface. Laser heating of the film itself is insufficient to induce
exchange on a significant scale. Experiments carried out by by Park and coworkers [20]
113
investigated the rate of isotope exchange in thin binary films exposed to HCl at 95 and
140 K. Even with the benefit of the HCl catalyst and a low activation energy (9.8 kJ/mol),
exchange was only observed at 140 K and the process continued for at least 10 minutes
without reaching complete exchange. Based on calculations conducted in our lab [21],
thin upper ASW films (300 ML) can reach a surface temperature of 195 K and internal
temperatures above that, but cool back to 100 K within 150 ns. Thicker films (>600 ML)
heat and cool more slowly, but reach a lower maximum temperature of <120 K. Neither
of these cases engenders H/D exchange from laser heating, so other explanations are
needed for this observation.
It is plausible that some exchange occurs due to gas-phase collisions as the hot mix
of chemicals leaves the film. Past and current data show that collisions are a feature of
our system due to the density of fissures at the ASW/vacuum interface and the fluid
that escapes from the fissures. Experimentally, it is detected as an unphysically long
“tail” in temporal profile data that can extend for over 1 ms. However, the orientation
and energy for these collisions to result in isotope exchange would likely make this
path a small contributor to the overall population of HDO.
Isotope exchange can also occur as hot N
2
O
4
carves out channels in the ASW film,
scraping the walls and encouraging mixing of H
2
O and D
2
O. This is a likely scenario
that is further supported by the similarity of exchange signal independent of D
2
O ar-
rangement. It also indicates that water molecules are removed from the fissure walls
along the entire path length.
The large fraction of HDO
+
detected in relation to the D
2
O peak (see Figures 5.6
and 5.7) reveals that mixing is extremely high and indicates a very dense plume of
material. Furthermore, it reinforces the proposal that large chunks of material are not
being released in this mechanism, as that would prohibit extensive isotope exchange.
Interestingly, the mixing ratio decreases with increasing film thickness. Such a trend
may point to wider fissures that provide fewer opportunities for isotope exchange, but
114
to be more definitive on this subject would be speculative. Further study is needed to
understand this trend as it relates to the heating of a buried stratum model.
5.5 Conclusions
This chapter has presented data that builds on the “fissure” model detailed in Chap-
ter 4. The extreme heating of a buried stratum was analyzed by introducing a probe
species (CO
2
, C
2
H
2
, or D
2
O) at different levels within a typical “sandwich” configura-
tion. Focused 266-nm radiation was used to irradiate the film, and the behavior of the
probe species was tracked relative to N
x
O
y
and H
2
O fragments. Carbon dioxide and
acetylene gave valuable insight into the effective quenching of the film by MgO acting
as a heat sink. Furthermore, CO
2
behaved differently when placed above the N
2
O
4
layer
compared to being embedded in the upper ASW layer. This implied that there were two
driving forces behind material removal. Molecules in direct contact with the hot N
2
O
4
fluid become entrained and escape from the ASW film with similar speeds. Material in
upper layers is removed as the hot fluid scrapes the walls. Introducing D
2
O provided
further insight into transport and molecular interactions within the film. The presence
of H/D exchange products supported scraping of fissure walls along the entire path to
vacuum.
Continued study of these systems using FTIR spectroscopy will be beneficial for
quantitatively measuring shot-by-shot material removal. In particular, the absorption
bands of CO
2
centered at 2350 and 3800 cm
1
are easily detected and their decay can be
measured against that of ASW at 3400 cm
1
. Also, further experimentation is required
to explore the result of acetylene’s possible reaction during irradiation. By varying
placement and thickness of C
2
H
2
layers within ASW films, we may be able to determine
the nature of its interaction with its environment and determine details of any reactions
that may take place. Understanding what chemistry can occur during the expulsion of
material from the film may have implications for Enceladus, with its jets of molecularly-
rich water plumes. Knowing what species can survive a trip from a buried ocean
115
through a thick crust of ASW can give insight into the material that is detected in these
“volcanoes.”
116
Chapter References
[1] Loerting, T.; Winkel, K.; Seidl, M.; Bauer, M.; Mitterdorfer, C.; Handle, P .; Salz-
mann, C.; Mayer, E.; Finney, J.; Bowron, D. Phys. Chem. Chem. Phys. 2011, 13,
8737–8794.
[2] Hama, T.; Watanabe, N. Chem. Rev. 2013, 113, 8783–8839.
[3] Manca, C.; Martin, C.; Roubin, P . J. Phys. Chem. B 2003, 107, 8929–8934.
[4] May, R.; Smith, R.; Kay, B. J. Chem. Phys. 2013, 138, 104501.
[5] May, R.; Smith, R.; Kay, B. J. Chem. Phys. 2013, 138, 104502.
[6] Laufer, D.; Bar-Nun, A.; Igal, P .; Jacovi, R. Icarus 2013, 222, 73–80.
[7] Herbst, E. AIP Con. Proceed. 2013, 1543, 15–30.
[8] Kumi, G.; Malyk, S.; Hawkins, S.; Reisler, H.; Wittig, C. J. Phys. Chem. A 2006,
110, 2097–2105.
[9] Malyk, S.; Kumi, G.; Reisler, H.; Wittig, C. J. Phys. Chem. A 2007, 111, 13365–
13370.
[10] Rebolledo–Mayoral, O.; Stomberg, J.; McKean, S.; Reisler, H.; Wittig, C. J. Phys.
Chem. C 2012, 116, 563–569.
[11] De Grotthuss, C. Ann. Chim. 1806, 58, 54–73.
[12] Collier, W.; Ritzhaupt, G.; Devlin, J. J. Phys. Chem. 1984, 88, 363–368.
[13] Lee, C.; Lee, P .; Kim, Y.; Kang, H. J. Chem. Phys. 2007, 127, 084701.
[14] Kim, J.; Kim, Y.; Kang, H. J. Chem. Phys. 2009, 131, 0044705.
[15] Blomquist, J.; Andersson, M.; Uvdal, P . Phys. Rev. Lett. 2011, 107, 216101.
[16] Gálvez, Ó.; Maté, B.; Herrero, V .; Escribano, R. Astrophys. J. 2011, 738, 133–138.
[17] Robert, F. Science 2001, 293, 1056–1058.
[18] Sarafian, A.; Nielsen, S.; Marschall, H.; McCubbin, F.; Monteleone, B. Science
2014, 31, 623–626.
[19] Cwiklik, L.; Devlin, J.; Buch, V . J. Phys. Chem. A 2009, 113, 7482–7490.
[20] Park, S.; Jung, K.; Kang, H. J. Chem. Phys. 2004, 121, 2765–2774.
117
[21] McKean, S. M. Energy Transfer in Amorphous Solid Water: Light-Mediated Ex-
pulsion of N
2
O
4
Guest Molecules., Ph.D. Thesis, University of Southern Califor-
nia, 2015.
118
Chapter 6
Future Work
This dissertation has dealt mainly with the development of a qualitative model that
explains how a hot fluid (N
2
O
4
and derivative species) forces its way through an ASW
film. Due to the many variables and unknowns in the experimental approach and pro-
posed mechanism, currently it is not possible to be quantitative about the changes that
occur in the ASW, not to mention its morphology after irradiation. Some of these issues
can be addressed by making adjustments to equipment in the laboratory. A change in
experimental strategy is required as well to address other questions that remain. Pro-
posed experiments would utilize gold nanoparticles to create fissures of known length
and diameter, which can then be doped with molecules of interest. This chapter dis-
cusses changes in experimental apparatus first, then discusses increased exploration of
systems already investigated, and finally covers the implementation of gold nanoparti-
cles as fissure-creators along with preliminary results.
6.1 Improvements in Experimental Apparatus
There are some practical changes that can address experimental shortcomings. The
easiest to implement would be the purchase and installation of a mercury-cadmium-
telluride (MCT) detector for the FTIR spectrometer. MCT detectors are sensitive over
the range 600-8000 cm
1
, which includes the primary absorption bands of N
2
O
4
[1],
and as such would improve our detection capabilities. The successful implementation
119
of a MCT detector may require a change in substrate material and UHV chamber win-
dows along the FTIR beam path, as MgO and CaF
2
drop to 50% transmission around
1250 cm
1
and 1050 cm
1
, respectively. A material that would allow access to lower en-
ergy absorption bands is BaF
2
, though implementation may not be easy. BaF
2
windows
are available from several vendors, but there may be some trial and error required for
substrate preparation. The cleaving method used to produce clean MgO substrates may
not be so facile with a softer material like BaF
2
.
Another change that would expand experimental possibilities would be the instal-
lation of a helium cryostat. Two closed-cycle refrigerators are already available to the
lab, but may need refurbishing. This would extend our temperature range to20 K
and would result in improved control over substrate temperature. Such a change would
allow for a much larger range of molecules to condense on the surface, including non-
reactive gases such as N
2
, Ar, Kr, and CH
4
. This opens up an experimental route to
investigating fissure capacity by inert gas uptake, a method used by other groups re-
searching ASW morphology [2, 3]. After a film with fissures is generated, a probe gas
can be deposited to fill the fissures. A suitable choice would be N
2
, which does not
form multilayers on an ASW surface at 27 K and has been used by Bruce Kay with
success [4]. Then, TPD experiments can be conducted to measure the overall signal of
the probe gas (in this case, N
2
) and the capacity of the fissures may be determined in a
quantitative way. It may be possible to use different probe gases to determine informa-
tion about the size and nature of the fissures (e.g. to look at how much danglingOH
is present, desorption temperatures of the probe gas could be examined).
A final addition to the experimental system would be the introduction of a cryo-
genic quartz crystal microbalance (CQCM). This would allow for exact measurements
of film thickness after vapor deposition, which has proven to be difficult with our cur-
rent equipment. While the low temperature limit of a CQCM is typically around 75 K
(though some have been reported in the literature to be used at 30-40 K [5]), it would
120
still be extremely useful for calibrating films grown under standard experimental con-
ditions.
6.2 Introducing a Third Species: What Next?
D
2
O, CO
2
, and C
2
H
2
have already been utilized as probe molecules for understand-
ing energy transfer in ASW films (refer to Chapter 5 for a discussion of these exper-
iments). However, the scope of these experiments was relatively limited and did not
fully explore what could be learned using these materials.
12
CO
2
and
13
CO
2
are attractive species for probing material removal quantitatively
because of their strong absorption peaks at 2350 cm
1
and 2280 cm
1
, respectively. If
12
CO
2
is codeposited with N
2
O
4
and the film is subsequently ablated, removal on a shot-
by-shot basis can be tracked by FTIR for both species. In addition,
13
CO
2
can be dosed
as an embedded stratum in the top ASW layer. This way, a comparison can be made
between material that is removed via entrainment in the hot fluid versus material that
is removed from the fissure walls. Another direction to pursue would be deposition of
13
CO
2
and
12
CO
2
at different levels in the ASW in order to examine fissure formation.
Exploring various combinations would invariably invite more ideas for experimental
investigation.
6.3 Gold nanoparticles as Fissure-Creators
Gold nanoparticles are an increasingly useful scientific tool in chemistry, medicine,
and materials science. Their shape and size can dictate very specific absortion spectra
via inducement of surface plasmonic resonance (SPR). In conjunction with short laser
pulses, this can result in very fast heating of the particles as energy absorbed gets
converted to heat. Being able to tailor gold nanoparticles with respect to their size,
shape, and absorption range has enabled a variety of medical applications involving
selective damage, such as destroying cancer cells[6].
121
The ability to selectively remove material is also appealing for introducing structural
features to a film, which is proposed for the future direction of this project. In other
words, gold nanoparticles can be used to create fissures of a known size in an ASW film
in a step-wise fashion. First, a MgO surface would be doped with gold nanoparticles.
The coverage would be kept low enough that the average interparticle distance is at least
an order of magnitude larger than the nanoparticle diameter to minimize side effects
due to particle interaction. After installation of the surface in the vacuum chamber, a
layer of N
2
O
4
followed by a ASW layer would be grown over the nanoparticles. Then,
the surface would be subjected to 532 nm radiation which excites the plasmon resonance
of the nanoparticles [7]. The excitation quickly degrades to heat on a ps timescale [8],
which rapidly heats the gold nanoparticles and material above it. Molecules above
the gold nanoparticles desorb due to this heat, forming vertical channels in the ASW
film (Figure 6.1). Because of the known diameters of the nanoparticles and the known
thickness of the ASW film, it is possible to approximate the dimensions of these fissures.
The nature of this method makes the density and diameter of fissures customizable
by adjusting nanoparticle doping concentration and size. Different fissure shapes can
be achieved by utilizing nanorods in place of spheres, which are easily excited with
1064 nm radiation. Thicker films can be achieved iteratively, by adding layers of ASW
until nanoparticle heating no longer removes material. Then, UV radiation (355 or
266 nm) can be used to excite the N
2
O
4
layer and make longer fissures.
Once channels of known dimension are formed, they can be doped with molecules
of interest and further probed using LID, FTIR, and TPD techniques. This “bottom-up”
method of creating fissures differs greatly from the standard procedures for introduc-
ing morphological changes in ASW, which include extreme pressure (GPa) and thermal
and radiative heating [9]. Instead of triggering morphological changes throughout the
entire film that are difficult to quantify, a size and shape limit is put in place. Further-
more, the distribution density of fissures can be easily adjusted by changing the density
122
Figure 6.1: Cartoon depiction of how gold nanoparticles can be used to form channels
in ASW using 532 nm radiation, which is resonant with 10-80 nm gold nanoparticles. (1)
Nanoparticles are deposited on a bare MgO surface, and particle density is determined
using AFM. (2) A layer of N
2
O
4
, followed by ASW is grown over the nanoparticles.
(3) 532 nm radiation is focused on the surface, heating the nanoparticles via surface
plasmon resonance to remove material. Once SPR heating is no longer able to remove
material, UV radiation (355/266 nm) is used to excite the N
2
O
4
layer. (4) This itera-
tive process can form vertical channels in the ice that are roughly the diameter of the
gold nanoparticle and length of the film. Once these channels are formed, they can be
doped with molecules of interest and the film can be further probed with spectroscopic
techniques.
123
of nanoparticle distribution on the substrate. This would be the first example of exper-
imentally grown ASW fissures with known dimensions and distribution, which could
be characterized quantitatively.
6.4 Experimental Strategy
In order to test the validity of this novel technique, some preliminary experiments
were carried out. One goal was to test the ease of sample preparation with our existing
apparatus. Another goal was assessing the feasibility of the proposed method, i.e., can
gold nanoparticles remove water via 532 nm laser radiation? In order to do this, it
was necessary to prepare MgO substrates with deposited gold nanoparticles and install
them in our UHV chamber. Then, ASW films could be grown and irradiated with minor
changes to our typical experimental set-up.
6.4.1 Surface Preparation and Experimental Adjustments
A new surface had to be prepared, as water was found to not wet aged pieces of
MgO(100). Since no FTIR would be carried out in these initial tests, we only cleaved one
side to make a new substrate from a block of MgO(100). The smoothness of the MgO
substrate was then analysed using atomic force microscopy (AFM), and it was found
to be relatively smooth with terraces as the main defect feature (refer to Figure 3.3 in
Chapter 3).
Gold nanoparticles (10, 20, 40, and 80 nm; citrate-capped spheres in DI water;
Nanopartz) were deposited by wetting the surface of the MgO(100) substrate with
one drop of a solution whose concentration was diluted with distilled water to 4–
610
10
nps/mL. Evaporation was accelerated by directing a flow of dry N
2
at the
freshly-wetted surface. This deposition method resulted in a surface density of ap-
proxmately 1-10 nanoparticles per square micron (Figures 6.2 and 6.4a). The surface
was reinstalled in the chamber in the typical configuration (refer to Chapter 3 for a
124
Figure 6.2: AFM image of a MgO(100) surface wetted with one drop of a solu-
tion containing 20 nm gold nanoparticles with a concentration of approximately
6 10
10
nps/mL. This scan was taken from the center of the surface, where the density
of the particles was 10 nanoparticles per square micron. In scans taken near the edge
of the surface, the density of nanoparticles was slightly higher and more aggregates of
particles were visible.
125
more detailed description), but the oxygen annealing procedure was not carried out to
avoid possible melting, migration, and sintering of nanoparticles on the substrate. We
were able to successfully make substrates with 20 and 80 nm nanoparticles, but had
deposition issues with 10 and 40 nm nanoparticles. The 10 nm nanoparticles appeared
to be too mobile for the wetting deposition method, and as such aggregated in the
area that evaporated last instead of sticking to the surface where initially deposited.
While trying to make a surface with 40 nm nanoparticles, we ran into repeated issues
of contamination and were forced to abandon the attempt.
6.5 Initial Results
We were able to test the efficacy of the proposed method by using 532 nm radiation
to irradiate the surface, which was easily generated as the second harmonic of our ex-
isting Nd:YAG laser (refer to Figure 3.6 in Chapter 3 for a schematic of the laser set-up).
The simplest system, consisting of ASW grown over gold nanoparticles, showed modest
signal with one laser pulse (Figure 6.3). This verified that SPR heating of gold nanopar-
ticles is capable of selectively removing ASW from a film under typical experimental
conditions.
Mixed in with the water fragments were CO
+
and CO
2
+
, which were evidence of the
citrate ligands breaking apart and flying off of the gold nanoparticles. The appearance
of the citrate fragments was decreased by increasing the thickness of the ASW film. This
may be due to citrate fragments becoming “caught” in longer channels or the dispersion
of heat through a thicker film limiting their fracture and movement.
We attempted to find the film thickness cut-off for ASW removal using 20 nm gold
nanoparticles by testing films of thicknesses ranging from 300-1200 ML. We were able to
consistently get signal with a 300 ML-thick film, which is roughly equivalent to 100 nm
of material on the nanoparticles. Six hundred monolayers of ASW also resulted in
signal, but no signal was detected with 1200 ML (480 nm). These results couldn’t be
126
Figure 6.3: A temporal profile of 300 ML of ASW grown over 20 nm gold nanoparticles
after irradiation with 6.4 mJ of focused 532 nm light (average of three spots). The
majority of the signal is due to water and water fragments, with some contribution
from citrate fragments.
repeated rigorously as we had an issue with nanoparticle degradation, discussed in the
following section.
Another interesting finding came from introducing a N
2
O
4
layer. Irradiating a film
with ASW underneath a layer of N
2
O
4
resulted in less desorbing material than when
N
2
O
4
was underneath an ASW layer. Understanding this preliminary finding requires
further study and speculation will not be engaged in at this time.
6.5.1 Nanoparticle Breakdown
While we were able to get measureable desorption signal, which increased with
nanoparticle size and film thickness (to a cutoff point), the nanoparticles seemed to
have an irradiation lifetime. The alignment of our laser, combined with the precision
movement of the surface in the x and z direction, resulted in repeated irradiation of
specific spots on the surface. After conducting an experiment on a designated location
127
on the surface, returning to that location with a new film would have either smaller
signal or no signal at all. Adjusting the position slightly to a spot that had not bee
irradiated in previous experiments gave strong signal. This indicated that distortion of
the particles was occuring such that they were no longer absorbing at 532 nm. This was
confirmed by several AFM images that appear to show conglomerated (20 nm) or oblit-
erated particles (80 nm) in irradiated areas. Before and after images of 80 nm particles
are shown in Figure 6.4. In the case of the 80 nm particles, the dramatic “exploded”
appearance coincides with the findings of other groups working with nanoparticles in
the liquid phase [6, 10, 11].
6.6 Summary and Future Experiments
Based on preliminary results, it is possible to form channels of known diameter in
ASW using gold nanoparticles with 532 nm radiation. The method we used had issues,
and would benefit from the following refinements:
fine-tuning of the nanoparticle deposition process to accomodate more sizes and
even distribution
careful study of laser fluence to determine a balance between signal and nanopar-
ticle degradation
use of gold nanoparticles capped with a more robust material, such as silica
characterization of film thicknesses to obtain the optimum nanoparticle size/ASW
film thickness combination
testing of gold nanorods to create wider fissures via irradiation with 1064 nm light
Once surfaces can be easily installed and used repeatedly to create films, more
complex experiments can be carried out. For example, after fissures have been created in
an ASW film, they can be doped with a radiation-absorbing molecule. This film can then
128
Figure 6.4: AFM images of 80 nm gold nanoparticles on the MgO substrate before (a)
and after (b) laser ablation. The particle density of the freshly-prepared film was ap-
proximately 1 particle per micron. The particles in (b) have experienced several 532 nm
pulses at a fluence of up to3 J/cm
2
.
129
be irradiated and energy transfer to ASW could be analyzed. Also, the fissure model
proposed in Chapter 4 could be explored further by varying the distribution density
of nanoparticles on the surface. By lowering the density of the nanoparticles, and thus
fissures, the structure of temporal profiles from resulting films could be examined and
contrasted with the data presented in this dissertation (which is posited to be the result
of a high fissure density system). The complicated nature of ASW as a system invites
further study, and the ideas presented in this chapter can improve characterization and
understanding.
130
Chapter References
[1] Mélen, F.; Pokorni, F.; Herman, M. Chem. Phys. Lett. 1992, 194, 181–186.
[2] Bar-Nun, A.; Dror, J.; Kochavi, E.; Laufer, D. Phys. Rev. B 1987, 35, 2427–2435.
[3] Bar-Nun, A.; Kleinfeld, I.; Kochavi, E. Phys. Rev. B 1988, 38, 7749–7754.
[4] Smith, R.; Zubkov, T.; Dohnálek, Z.; Kay, B.J.Phys.Chem.B 2009,113, 4000–4007.
[5] Raut, U.; Famá, M.; Teolis, B.; Baragiola, R. J. Chem. Phys. 2007, 127, 204713.
[6] Letfullin, R.; Joenathan, C.; George, T.; Zharov, V . Nanomedicine 2006, 1, 473–480.
[7] Fox, M., Optical Properties of Solids; Oxford University Press: 2010.
[8] Nedyalkov, N.; Imamova, S.; Atanasov, P .; Toshkova, R.; Gardeva, E.; Yossifova,
L.; Alexandrov, M.; Obara, M. Appl. Surf. Sci. 2011, 257, 5456–5459.
[9] Angell, C. Annu. Rev. Phys. Chem. 2004, 55, 559–583.
[10] Cavicchi, R. E.; Meier, D.; Presser, C.; Prabhu, V .; Guha, S. J. Phys. Chem. C 2013,
117, 10866–10875.
[11] Werner, D.; Hashimoto, S. J. Phys. Chem. C 2011, 115, 5063–5072.
131
Appendix A
Appendix
An expression is derived for the ratio of the NO
2
and N
2
O
4
concentrations that
reach the ionization region of the mass spectrometer. The expression is in terms of
ionization cross sections, parent ion fragmentation patterns, and relative ion signals.
For example, use is made of measured quantities such as NO
+
/NO
2
+
, where NO
+
and
NO
2
+
denote the areas of the respective ion signals. It is assumed that the microchannel
plate detector has equal sensitivity for each ion. To begin, expressions for the amounts
(concentrations) of NO
2
+
, NO
+
, and O
+
are written
NO
2
+
= s
ion
N
2
O
4
F
NO
2
+
N
2
O
4
+
N
2
O
4
+s
ion
NO
2
F
NO
2
+
NO
2
+
NO
2
(A.1)
NO
+
= s
ion
N
2
O
4
F
NO
+
N
2
O
4
+
N
2
O
4
+s
ion
NO
2
F
NO
+
NO
2
+
NO
2
+s
ion
NO
F
NO
+
NO
+
NO (A.2)
O
+
= s
ion
N
2
O
4
F
O
+
N
2
O
4
+
N
2
O
4
+s
ion
NO
2
F
O
+
NO
2
+
NO
2
+s
ion
H
2
O
F
O
+
H
2
O
+
+s
ion
NO
F
O
+
NO
+
NO+s
ion
O
O (A.3)
where s
ion
X
is the 70 eV ionization cross section for the neutral species X, and F
Y
+
X
+
is the
fraction of daughter ion Y
+
that arises from fragmentation of the parent ion X
+
.
The use of equals signs rather than proportionality constants in Equations A.1 –
A.3 is in anticipation of the fact that these constants would cancel were they carried
along. The relatively small amount of O
2
+
that appears in spectra shall be ignored.
Likewise, it is assumed that the NO and O concentrations are negligible. This is a
safe assumption for long arrival times at the ionization region, though less so for short
arrival times. To proceed, the last term in Equation A.2 and the last two terms in
132
Equation A.3 are dropped, in which case Equation A.3 is decoupled from Equations
A.1 and A.2. Considering Equation A.2, the relationship NO
+
= (NO
+
/NO
2
+
)NO
2
+
is introduced. Combining Equations A.1 and A.2 then yields

NO
+
/NO
2
+

s
ion
N
2
O
4
F
NO
2
+
N
2
O
4
+
N
2
O
4
+s
ion
NO
2
F
NO
2
+
NO
2
+
NO
2

= s
ion
N
2
O
4
F
NO
+
N
2
O
4
+
N
2
O
4
+s
ion
NO
2
F
NO
+
NO
2
+
NO
2
(A.4)
Separating this equation into separate terms for NO
2
and N
2
O
4
gives
NO
2
s
ion
NO
2

NO
+
/NO
2
+

F
NO
2
+
NO
2
+
F
NO
+
NO
2
+

= N
2
O
4
s
ion
N
2
O
4

F
NO
+
N
2
O
4
+

NO
+
/NO
2
+

F
NO
2
+
N
2
O
4
+

(A.5)
Further rearrangement yields the desired ratio
NO
2
N
2
O
4
= 2.08

NO
+
/NO
2
+

F
NO
2
+
N
2
O
4
+
F
NO
+
N
2
O
4
+
F
NO
+
NO
2
+

NO
+
/NO
2
+

F
NO
2
+
NO
2
+
(A.6)
where s
ion
N
2
O
4
/s
ion
NO
2
= 2.08 has been used (Table 4.1). Equation A.6 can be used to obtain
relative NO
2
and N
2
O
4
concentrations when NO and O are in low enough concentration
to justify ignoring them.
133
BIBLIOGRAPHY
Andersson, O.; Inaba, A. Phys. Chem. Chem. Phys. 2005, 7, 1441–1449.
Andrews, B.; Anderson, A. J. Chem. Phys. 1981, 74, 1534–1537.
Angell, C. Annu. Rev. Phys. Chem. 2004, 55, 559–583.
Angell, C. Science 2008, 319, 582–587.
Ayotte, P .; Smith, R.; Stevenson, K.; Dohnálek, Z.; Kimmel, G.; Kay, B. J. Geophys. Res.
2001, 106, 33387–33392.
Bach, W.; Breuer, H. Faraday Discuss. Chem. S. 1974, 58, 237–243.
Backus, E.; Grecea, M.; Kleyn, A.; Bonn, M. Phys. Rev. Lett. 2004, 92, 236101.
Bai, Y.; Ogai, A.; Qian, C.; Iwata, L.; Segal, G.; Reisler, H. J. Chem. Phys. 1989, 90, 3903–
3914.
Gandyopadhyay, D.; Mohan, S.; Ghosh, S.; Choudhury, N. J. Phys. Chem. B 2013, 117,
8831–8843.
Baragiola, R. Planet. Space Sci. 2003, 51, 953–961.
Bar-Nun, A.; Dror, J.; Kochavi, E.; Laufer, D. Phys. Rev. B 1987, 35, 2427–2435.
Bar-Nun, A.; Kleinfeld, I.; Kochavi, E. Phys. Rev. B 1988, 38, 7749–7754.
Bass A.M.and Ledford Jr., A.; Laufer, A. J. Res. Nat. Bur. Stand. A 1976, 80A, 143–166.
Beavis, R.; Chait, B. Chem. Phys. Lett. 1991, 181, 479–484.
Bernal, J.; Fowler, R. J. Chem. Phys. 1933, 1, 515–548.
Blomquist, J.; Andersson, M.; Uvdal, P . Phys. Rev. Lett. 2011, 107, 216101.
Bolduan, F.; Jodl, H.; Loewenschuss, A. J. Res. Nat. Bur. Stand. A 1976, 80A, 143–166.
Brown, D.; George, S.; Huang, C.; Wong, E.; Rider, K.; Smith, R.; Kay, B. J. Phys. Chem.
1996, 100, 4988–4995.
134
Brovchenko, I.; Oleinikova, A. J. Chem. Phys. 2006, 124, 164505.
Brüggeller, P .; Mayer, E. Nature 1980, 288, 569–571.
Burgess, D.; Stair, P .; Weitz, E. J. Vac. Sci. Technol., A 1986, 4, 1362–1366.
Burke, D.; Brown, W. Phys. Chem. Chem. Phys. 2010, 12, 5947–5969.
Burton, E.; Oliver, W. Nature 1935, 135, 505–506.
Cavicchi, R. E.; Meier, D.; Presser, C.; Prabhu, V .; Guha, S. J. Phys. Chem. C 2013, 117,
10866–10875.
Chiu, W. Ann. Rev. Biophys. Biophys. Chem. 1986, 15, 237–257.
Cholette, F.; Zubkov, T.; Smith, R.; Dohnálek, Z.; Kay, B.; Ayotte, P . J.Phys.Chem.B 2009,
113, 4131–4140.
Chrysanthou, A.; Grieveson, P . Mat. Sci. Eng. A 1995, 194, L11–L14.
Chuang, T. Surf. Sci. Rep. 1983, 3, 1–105.
Chyba, C.; Sagan, C. Nature 1992, 355, 125–132.
Collier, W.; Ritzhaupt, G.; Devlin, J. J. Phys. Chem. 1984, 88, 363–368.
Time-of-Flight Mass Spectrometry; Cotter, R. J., Ed.; American Chemical Society: 1994.
Cwiklik, L.; Devlin, J.; Buch, V . J. Phys. Chem. A 2009, 113, 7482–7490.
Debenedetti, P . J. Phys.: Condens. Matter 2003, 15, R1669–R1726.
Devlin, J. J. Geophys. Res. 2001, 106, 33333–33349.
Dixon-Warren, S. J.; Jackson, R.; Polanyi, J.; Rieley, H.; Shapter, J.; Weiss, H. J. Phys.
Chem. 1992, 96, 10983–10994.
Dohnálek, Z.; Ciolli, R.; Kimmel, G.; Stevenson, K.; Smith, R.; Kay, B.J.Chem.Phys. 1999,
110, 5489–5492.
Dohnálek, Z.; Kimmel, G.; Ayotte, P .; Smith, R.; Kay, B.J.Chem.Phys. 2003,118, 364–372.
Draine, B. Annu. Rev. Astron. Astrophys. 2003, 41, 241–289.
Ediger, M.; Angell, C.; Nagel, S. J. Phys. Chem. 1996, 100, 13200–13212.
Edwards, G. Laser Photonics Rev. 2009, 3, 545–555.
English, N.; Tse, J.; Gallagher, R. Phys. Rev. B 2010, 83, 092201.
135
English, N.; Tse, J. Phys. Rev. B 2011, 83, 184114.
Essmann, U.; Geiger, A. J. Chem. Phys. 1995, 103, 4678–4692.
Faizullin, M.; Vinogradov, A.; Koverda, V . Chem. Eng. Sci. 2015, 000–000.
Fanetti, S.; Pagliai, M.; Citroni, M.; Lapini, A.; Scandolo, S.; Righini, R.; Bini, R. J. Phys.
Chem. Lett. 2014, 5, 3804–3809.
Finney, J.; Bowron, D.; Soper, A.; Loerting, T.; Mayer, E.; Hallbrucker, A. Phys. Rev. Lett.
2002, 89, 205503(4).
Fox, M., Optical Properties of Solids; Oxford University Press: 2010.
Fresneau, A.; Danger, G.; Rimola, A.; Theule, P .; Duvernay, F.; Chiavassa, T. MNRAS
2014, 443, 2991–3000.
Gálvez, Ó.; Maté, B.; Herrero, V .; Escribano, R. Astrophys. J. 2011, 738, 133–138.
Guilhaus, M. J. Mass. Spectrom. 1995, 30, 1519–1532.
Giovambattista, N.; Stanley, H.; Sciortino, F. Phys. Rev. Lett. 2005, 94, 107803.
Giobambattista, N.; Amann-Winkel, K.; Loerting, T. In Liquid Polymorphism: Advances in
ChemicalPhysics,Volume152, Stanely, H., Ed.; John Wiley & Sons, Inc.: Somerset, NJ,
2013.
De Grotthuss, C. Ann. Chim. 1806, 58, 54–73.
Günster, J.; Liu, G.; Stultz, J.; Krischok, S.; Goodman, D. W. J. Phys. Chem. B 2000, 104,
5738–5743.
Halfen, D.; Apponi, A.; Ziurys, L. Astrophys. J. 2001, 561, 244–253.
Hama, T.; Watanabe, N. Chem. Rev. 2013, 113, 8783–8839.
Hansen, C.; Shemansky, D.; Esposito, L.; Stewart, A.; Lewis, B.; Colwell, J.; Hendrix, A.;
West R.A., W. J.; J.H. Teolis, B.; Magee, B. Geophys. Res. Lett. 2011, 38, L11202.
Hasselbrink, E.; Jakubith, S.; Nettesheim, S.; Wolf, M.; Cassuto, A.; Ertl, G.J.Chem.Phys.
1990, 92, 3154–3169.
Hawkins, S.; Kumi, G.; Malyk, S.; Reisler, H.; Wittig, C. Chem. Phys. Lett. 2005, 404, 19–
24.
Herbst, E. AIP Con. Proceed. 2013, 1543, 15–30.
Hisatsune, I. J. Phys. Chem. 1961, 65, 2249–2253.
136
Hisatsune, I.; Miller Jr., P . J. Chem. Phys. 1963, 38, 49–54.
Ho, W. In Desorption Induced by Electronic Transitions, Betz, G., Varga, P ., Eds.; Springer-
Verlag: Berlin, 1990.
Hobbs, P . V ., Ice physics; Oxford: Clarendon Press: 1974.
Hurford, A.; Helfenstein, P .; Hoppa, G.; Greenberg, R.; Bills, B. Nature 2007, 447, 292–
294.
Huth-Fehre, T.; Becker, C. Rapid Commun. Mass Sp. 1991, 5, 378–382.
Iess, L.; Stevenson, D. J.; Parisi, M.; Hemingway, D.; Jacobson, R. A.; Lunine, J. I.;
Nimmo, F.; Armstrong, J. W.; Asmar, S. W.; Ducci, M.; Tortora, P . Science 2014, 344,
78–80.
Jenniskens, P .; Blake, D.; Wilson, M.; Pohorille, A. Astrophys. J. 1995, 455, 389–401.
Johari, G.; Andersson, O. Thermochim. Acta 2007, 461, 14–43.
Jones, L.; Swanson, B.; Agnew, S. J. Chem. Phys. 1985, 82, 4389–4390.
Kang, H. Acc. Chem. Res. 2005, 38, 893–900.
Kawasaki, M.; Kasatani, K.; Hiroyasu, S.; Shinohara, H.; Nobuyuki, N.Chem.Phys. 1983,
78, 65–74.
Y.K., K.; Rudd, M. Phys. Rev. A 1994, 50, 3954–3966.
Kim, Y. D.; Stultz, J.; Goodman, D. W. J. Phys. Chem. B 2002, 106, 1515–1517.
Kim, J.; Kim, Y.; Kang, H. J. Chem. Phys. 2009, 131, 0044705.
Kim, S.; Kang, H. J. Phys. Chem. Lett. 2010, 1, 3085–3089.
Kimmel, G.; Stevenson, K.; Dohnálek, Z.; Smith, R.; Kay, B. J. Chem. Phys. 2001, 114,
5284–5294.
Kimmel, G.; Stevenson, K.; Dohnálek, Z.; Smith, R.; Kay, B. J. Chem. Phys. 2001, 114,
5295–5303.
Knözinger, E.; Jacob, K.; Hofmann, P . J. Chem. Soc. Faraday Trans. 1993, 89, 1101–1107.
Kondo, T.; Kato, H.; Bonn, M.; Kawai, M. J. Chem. Phys. 2007, 127, 094703.
König, H. Z. Kristallogr. 1944, 105, 279–286.
Korolik, M.; Suchan, M.; Johnson, M.; Arnold, D.; Reisler, H.; Wittig, C.Chem.Phys.Lett.
2000, 326, 11–21.
137
Kumi, G.; Malyk, S.; Hawkins, S.; Reisler, H.; Wittig, C. J.Phys.Chem.A 2006, 110, 2097–
2105.
Kvick, Å.; McMullan, R.; Newton, M. J. Chem. Phys. 1982, 78, 3754–3761.
Laufer, D.; Bar-Nun, A.; Igal, P .; Jacovi, R. Icarus 2013, 222, 73–80.
Lee, C.; Lee, P .; Kim, Y.; Kang, H. J. Chem. Phys. 2007, 127, 084701.
Letfullin, R.; Joenathan, C.; George, T.; Zharov, V . Nanomedicine 2006, 1, 473–480.
Lignell, H.; Varner, M.; Finlayson-Pitts, B.; Benny Gerber, R. J. Chem. Phys. 2012, 405,
52–59.
Limmer, D.; Chandler, D. PNAS 2014, 111, 9413–9418.
Linstrom, P ., Mallard, W., Eds. NIST Chemistry WebBook, NIST Standard Reference
Database Number 69., http://webbook.nist.gov (accessed 05/04/2015).
Lippert, T. Plasma Process. Polym. 2005, 2, 525–546.
Livingston, F.; Smith, J.; George, S. Anal. Chem. 2000, 72, 5590–5599.
Loerting, T.; Giovambattista J. Phys. Condens. Matter 2006, 18, R919–R977.
Loerting, T.; Salzmann, C.; Kohl, I.; Mayer, E.; Hallbrucker, A. Phys. Chem. Chem. Phys.
2001, 3, 5355–5357.
Loerting, T.; Winkel, K.; Seidl, M.; Bauer, M.; Mitterdorfer, C.; Handle, P .; Salzmann, C.;
Mayer, E.; Finney, J.; Bowron, D. Phys. Chem. Chem. Phys. 2011, 13, 8737–8794.
Loparo, J.; Roberts, S.; Tokmakoff, A. J. Chem. Phys. 2006, 125, 194521.
Lorazo, P .; Lewis, L.; Meunier, M. Phys. Rev. Lett. 2003, 91, 225502.
Malenkov, G. J. Phys: Condens. Matter 2009, 21, 283101.
Malyk, S.; Kumi, G.; Reisler, H.; Wittig, C. J. Phys. Chem. A 2007, 111, 13365–13370.
Malyk, S. Transport and Guest-Host Interactions in Amorphous and Crystalline Ice.,
Ph.D. Thesis, University of Southern California, 2009.
Manca, C.; Martin, C.; Roubin, P . J. Phys. Chem. B 2003, 107, 8929–8934.
Manca, C.; Martin, C.; Roubin, P . Chem. Phys. 2004, 300, 53–62.
Maté, B.; Rodríguez-Lazcano, Y.; Herrero, V . Phys. Chem. Chem. Phys. 2012, 14, 10595–
10602.
138
Matthias, E.; Reichling, M.; Siegel, J.; Käding, O.; Petzoldt, S.; Skurk, H.; Bizenberger, P .;
Neske, E. Appl. Phys. A 1994, 58, 129–136.
May, R.; Smith, R.; Kay, B. J. Phys. Chem. Lett. 2012, 3, 327–331.
May, R.; Smith, R.; Kay, B. J. Chem. Phys. 2013, 138, 104501.
May, R.; Smith, R.; Kay, B. J. Chem. Phys. 2013, 138, 104502.
McGonagle, D.; Ziurys, L.; Irvine, W.; Minh, Y. Astrophys. J. 1990, 359, 121–124.
McKean, S. M. Energy Transfer in Amorphous Solid Water: Light-Mediated Expulsion
of N
2
O
4
Guest Molecules., Ph.D. Thesis, University of Southern California, 2015.
Mélen, F.; Pokorni, F.; Herman, M. Chem. Phys. Lett. 1992, 194, 181–186.
Mitlin, S.; Leung, K. J. Phys. Chem. B 2002, 106, 6234–6247.
Mishima, O.; Calvert, L.; Whalley, E. Nature 1984, 310, 393–395.
Mishima, O.; Stanley, H. Nature 1998, 396, 329–335.
Mooney, J.; Kambhampati, P . J. Phys. Chem. Lett. 2013, 4, 3316–3318.
Moore, E.; Molinero, V . Phys. Chem. Chem. Phys. 2011, 13, 20008–20016.
Nedyalkov, N.; Imamova, S.; Atanasov, P .; Toshkova, R.; Gardeva, E.; Yossifova, L.;
Alexandrov, M.; Obara, M. Appl. Surf. Sci. 2011, 257, 5456–5459.
Ogai, A.; Qian, C.; Iwata, L.; Reisler, H. Chem. Phys. Lett. 1988, 146, 367–374.
Pan, Y.; Cotter, R. Org. Mass Spectrom. 1992, 27, 3–8.
Park, S.; Jung, K.; Kang, H. J. Chem. Phys. 2004, 121, 2765–2774.
Perez, D.; Lewis, L. Appl. Phys. A 2004, 79, 987–990.
Perez, D.; Lewis, L.; Lorazo, P .; Meunier, M. Appl. Phys. Lett. 2006, 89, 141907.
Poole, P .; Essmann, U.; Sciortino, F.; Stanley, H. Phys. Rev. E 1993, 48, 4605–4610.
Porco, C.; DiNino, D.; Nimmo, F. Astron. J. 2014, 148, 45.
Qian, C.; Ogai, A.; Iwata, L.; Reisler, H. J. Chem. Phys. 1990, 92, 4296–4307.
Raut, U.; Famá, M.; Teolis, B.; Baragiola, R. J. Chem. Phys. 2007, 127, 204713.
Ready, J. J. Appl. Phys. 1965, 36, 462–468.
139
Rebolledo–Mayoral, O.; Stomberg, J.; McKean, S.; Reisler, H.; Wittig, C. J. Phys. Chem. C
2012, 116, 563–569.
Rice, S.; Sceats, M. J. Phys. Chem. 1981, 85, 1108–1119.
Rieley, H.; McMurry, D.; Haq, S. Faraday Trans. 1996, 92, 933–939.
Rieley, H.; Colby, D.; McMurry, D.; Reeman, S. J. Phys. Chem. B 1997, 101, 4982–4991.
Robert, F. Science 2001, 293, 1056–1058.
Russ, F. Z. Phys. Chem. 1914, 82, 217–222.
Safarik, D.; Mullins, C. J. Chem. Phys. 2004, 121, 6003–6010.
Sarafian, A.; Nielsen, S.; Marschall, H.; McCubbin, F.; Monteleone, B. Science 2014, 31,
623–626.
Schober, H.; Koza, M.; Tölle, A.; Masciovecchio, C.; Sette, F.; Fujara, F. Phys. Rev. Lett.
2000, 85, 4100–4103.
Shalit, A.; Perakis, F.; Hamm, P . J. Phys. Chem. B 2013, 117, 15512–15518.
Shephard, J.; Evans, J.; Salzmann, C. J. Phys. Chem. Lett. 2013, 4, 3672–3676.
Slack, G. A. Phys. Rev. 1962, 126, 427–441.
Smith, R.; Huang, C.; Wong, E.; Kay, B. Phys. Rev. Lett. 1997, 79, 909–912.
Smith, R.; Huang, C.; Kay, B. J. Phys. Chem. B. 1997, 101, 6123–6126.
Smith, R.; Kay, B. Nature 1999, 398, 788–791.
Smith, R.; Zubkov, T.; Dohnálek, Z.; Kay, B. J. Phys. Chem. B 2009, 113, 4000–4007.
Smith, R.; Petrik, N.; Kimmel, G.; Kay, B. Acc. Chem. Res. 2012, 45, 33–42.
Smoluchowski, R. Astrophys. J. 1981, 244, L31–L34.
Starr, D.; Weis, C.; Yamamoto, S.; Nilsson, A.; Bluhm, H. J. Phys. Chem. C 2009, 113,
7355–7363.
Stirniman, M. J.; Huang, C.; Smith, R. S.; Joyce, S. A.; Kay, B. D. J.Chem.Phys. 1996, 105,
1295–1298.
Suchan, M. M. Molecule-Surface Interactions in HCl/MgO and Water/MgO Systems.,
Ph.D. Thesis, University of Southern California, 2001.
Susa, A.; Koda, S. J. Atmos. Chem. 2003, 43, 197–217.
140
Thrower, J.; Burke, D.; Collings, M.; Dawes, A.; Holtom, P .; Jamme, F.; Kendall, P .; Brown,
W.; Clark, I.; Fraser, H.; McCoustra, M.; Mason, N.; Parker, A.Astrophys.J. 2008,673,
1233–1239.
Thrower, J.; Collings, M.; McCoustra, M.; Burke, D.; Brown, W.; Dawes, A.; Holtom, P .;
Kendall, P .; Mason, N.; Jamme, F.; Fraser, H.; Clark, I.; Parker, A. J. Vac. Sci. Tech.
2008, 26, 919–924.
Tielens, A.; Hagen, W. Astron. Astrophys. 1982, 114, 245–260.
Thrower, J.; Collings, M.; McCoustra, M.; Burke, D.; Brown, W. A.; Dawes, A.; Holtom,
P .; Kendall, P .; Mason, N.; Jamme, F.; Fraser, H.; Clark, I.; Parker, A. J. Vac. Sci.
Technol., A 2008, 26, 919–924.
Thrower, J.; Burke, D.; Collings, M.; Dawes, A.; Holtom, P .; Jamme, F.; Kendall, P .; Brown,
W. A.; Clark, I.; Fraser, H.; McCoustra, M.; Mason, N.; Parker, A. Astrophys. J. 2008,
673, 1233–1239.
Thrower, J.; Abdulgalil, A.; Collings, M.; McCoustra, M.; Burke, D.; Brown, W.; Dawes,
A.; Holtom, P .; Kendall, P .; Mason, N.; Jamme, F.; Fraser, H.; Rutten, F. J. Vac. Sci.
Technol., A 2010, 28, 799–806.
Tse, J. J. Chem. Phys. 1992, 1992, 5482–5487.
Tyndall, G.; Stedman, K.; Schneider, W.; Burrows, J.; Moortgat, G. J.Photochem. 1987, 36,
133–139.
Vandaele, A.; Hermans, C.; Simon, P .; Carleer, M.; Colin, R.; Fally, S.; Mérienne, M.;
Jenourvrier, A.; Coquart, B. J. Quant. Spectrosc. Radiat. Transfer 1998, 59, 171–184.
Vega, C.; Abascal, J.; Conde, M.; Aragones, J. Faraday Discuss. 2009, 141, 251–276.
Van Dishoeck, E.; Herbst, E.; Neufeld, D. Chem. Rev. 2013, 113, 9043–9085.
Wenzel, J.; Linderstrom-Lang, C.; Rice, S. Science 1975, 187, 428–430.
Werner, D.; Hashimoto, S. J. Phys. Chem. C 2011, 115, 5063–5072.
Wiley, W.; McLaren, I. Rev. Sci. Instrum. 1955, 26, 1150–1157.
Yang., R.; Gudipati, M. J. Chem. Phys. 2014, 140, 104202.
Yabushita, A.; Inoue, Y.; Senta, T.; Kawasaki, M.; Sato, S. J. Chem. Phys. A 2004, 108,
438–446.
Yabushita, A.; Kanda, D.; Kawanaka, N.; Kawasaki, M.; Ashfold, M. J.Chem.Phys. 2006,
125, 133406.
141
Zhigilei, L.; Kodali, P .; Garrison, B. Chem. Phys. Lett. 1997, 276, 269–273.
Zhigilei, L.; Kodali, P .; Garrison, B. J. Phys. Chem. B 1998, 102, 2845–2853.
Zhigilei, L.; Garrison, B. Rapid Commun. Mass Sp. 1998, 12, 1273–1277.
142

Abstract (if available)

Abstract The effects of inserting energy in a buried stratum in amorphous solid water (ASW) films were investigated using pulsed 266 nm radiation. Material ejected from irradiated films was detected with time‐of‐flight mass spectrometry (TOFMS). A technique was developed using N₂O₄ in conjunction with focused UV radiation (∼1 J/cm²) that enabled facile introduction of energy in a spatially selective way via an electronic transition of N₂O₄. A variety of experiments were carried out explore the structural changes induced by the sudden addition of energy to ASW films, and an attempt was made to characterize the nature of transport within and above the surface of the film. ❧ Layered ASW/N₂O₄ films, up to 2800 monolayers (ML) thick, were grown on a MgO(100) substrate. All samples were grown at ∼100 K under ultra‐high vacuum conditions to produce porous, high quality films. Once deposited, the films were were irradiated with 266 nm radiation that was generated as the fourth harmonic of a Nd:YAG laser (10 ns pulses, reduced to 1 Hz) focused to a 0.3 mm beam waist. After a single laser pulse incident on the film, the N₂O₄ layer was converted to a hot fluid that heated the surrounding material. Heating of the film competes with cooling by the MgO(100) substrate, which acts as an efficient heat sink due to its high thermal conductivity (250 W/mK). Therefore, extreme pressure and heat gradients exist within the film upon radiation and the film cools quickly upon cessation of the laser pulse. ❧ Despite fast cooling of the film, laser‐heated N₂O₄ fluid, along with water monomer, was detected at times greater than 1 ms. This was due to a catastrophic structural change triggered by the temperature and pressure gradients, which resulted in the formation of fissures. The hot fluid of N₂O₄ and its photoproducts escaped to ultra‐high vacuum through these fissures, scraping the walls and removing H₂O molecules. Long flight times were attributed to collisions occurring above the surface of the film due to the high density of the escaping material and prevalence of fissures in the irradiated area. Fissures proved to be robust, with spectra from subsequent pulses on the same spot resembling spectra of exposed N₂O₄. ❧ This model will be explored further by implementing gold nanoparticles as fissure‐creators. Being able to form fissures of a known dimension and density within an ASW film would allow for more careful analysis of this system and would be the first demonstration of induced morphological changes in ASW that are relatively well‐defined. An outline of this technique is presented, along with preliminary results.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Energy transfer in amorphous solid water: light-mediated expulsion of N₂O₄ guest molecules

PDF

Transport and guest-host interactions in amorphous and crystalline ice

PDF

Fourier transform infrared studies of guest-host interactions in ice

PDF

Dissociation energy and dynamics of HCl and aromatic-water clusters

PDF

A multifaceted investigation of tris(2-phenylpyridine)iridium

PDF

Development and implementation of methods for sliced velocity map imaging. Studies of overtone-induced dissociation and isomerization dynamics of hydroxymethyl radical (CH₂OH and CD₂OH)

PDF

Photodissociation dynamics of atmospherically relevant small molecules in molecular beams

PDF

Dissociation energy and dynamics of water clusters

PDF

Photodissociation dynamics of group V hydrides

PDF

Ultrafast electronic deactivation of DNA bases in aqueous solution

PDF

Fabrication, deposition, and characterization of size-selected metal nanoclusters with a magnetron sputtering gas aggregation source

$Study of rotation and phase separation in ³He, ⁴He, and mixed ³He/⁴He droplets by X-ray diffraction$

PDF

Study of rotation and phase separation in ³He, ⁴He, and mixed ³He/⁴He droplets by X-ray diffraction

PDF

Effect of surfactants on the growth and departure of bubbles from solid surfaces

PDF

A technique for assessing water cluster sizes by pickup momentum transfer and a new source with loading system for lithium clusters

PDF

Microclimate and building energy performance

PDF

Adiabatic and non-adiabatic molecular dynamics in nanoscale systems: theory and applications

PDF

I. Microwave-assisted synthesis of phosphonic acids; II. Design and synthesis of polymerase β lyase domain inhibitors

PDF

The brain's virtuous cycle: an investigation of gratitude and good human conduct

PDF

Non-linear surface spectroscopy of photoswitches annd photovoltaics

PDF

Nucleotide analogs and molecular probes for LFER, time-resolved crystallography, FRET, and EPR studies of human DNA polymerase: probing mechanism, conformation and structure

Asset Metadata

Creator Stomberg, Jaimie Elizabeth (author)

Core Title An investigation of morphology and transport in amorphous solid water via guest-host interactions

School College of Letters, Arts and Sciences

Degree Doctor of Philosophy

Degree Program Chemistry

Publication Date 07/21/2015

Defense Date 06/22/2015

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

Tag amorphous solid water,astrochemistry,ASW,mass spectrometry,nitrogen tetroxide,OAI-PMH Harvest

Format application/pdf (imt)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Reisler, Hanna (committee chair), Wittig, Curt F. (committee chair), Kresin, Vitaly V. (committee member)

Creator Email j.e.stomberg@gmail.com,stomberg@usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c3-602112

Unique identifier UC11301268

Identifier etd-StombergJa-3660.pdf (filename),usctheses-c3-602112 (legacy record id)

Legacy Identifier etd-StombergJa-3660.pdf

Dmrecord 602112

Document Type Dissertation

Format application/pdf (imt)

Rights Stomberg, Jaimie Elizabeth

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the a...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA