High-energy X-ray diffraction (XRD) and Fourier transform infrared spectroscopy (FTIR) of amorphous solid water (ASW) were studied during vapor deposition and the heating process. From the diffraction patterns, the oxygen-oxygen pair distribution functions (PDFs) were calculated up to the eighth coordination shell and an r = 23 Å. The PDF of ASW obtained both during vapor deposition at 80 K as well as the subsequent heating are consistent with that of low-density amorphous ice. The formation and temperature-induced collapse of micropores were observed in the XRD data and in the FTIR measurements, more specifically, in the OH stretch and the dangling mode. Above 140 K, ASW crystallizes into a stacking disordered ice, Isd. It is observed that the fourth, fifth, and sixth peaks in the PDF, corresponding to structural arrangements between 8 and 12 Å, are the most sensitive to the onset of crystallization.
High-energy X-ray diffraction (XRD) and Fourier transform infrared spectroscopy (FTIR) of amorphous solid water (ASW) were studied during vapor deposition and the heating process. From the diffraction patterns, the oxygen-oxygen pair distribution functions (PDFs) were calculated up to the eighth coordination shell and an r = 23 Å. The PDF of ASW obtained both during vapor deposition at 80 K as well as the subsequent heating are consistent with that of low-density amorphous ice. The formation and temperature-induced collapse of micropores were observed in the XRD data and in the FTIR measurements, more specifically, in the OH stretch and the dangling mode. Above 140 K, ASW crystallizes into a stacking disordered ice, Isd. It is observed that the fourth, fifth, and sixth peaks in the PDF, corresponding to structural arrangements between 8 and 12 Å, are the most sensitive to the onset of crystallization.
Amorphous solid water (ASW) is the most abundant form of solid
water in astrophysical environments. It is believed to be a major
component of interstellar clouds, comets, and solar-system bodies.[1−4] Due to its porosity and capability of adsorbing gases, ASW is discussed
to participate as a medium in many chemical reactions in outer space
that could play a key role in the earliest stages of planet building.[3] However, its true degree of porosity remains
unclear due to the lack of unambiguous observational data and a variety
of contributing processes during its formation and re-accretion in
outer space. ASW was first produced in the laboratory in 1935 by Burton
and Oliver through water vapor deposition on a cold copper substrate.[5,6] The so-formed ASW can serve as a model system to better understand
part of the contributing processes. Since then, ASW has been intensively
studied.[7−20] The morphologies and physical properties of ASW fabricated in the
laboratory can vary with vapor flux and directionality, substrate
temperature, and water partial pressure.[15,21−23] The effect of irradiation on the structure of amorphous
ice, and in particular on the dangling OH groups, has been studied
using UV,[24] IR,[17,25] and heavy ion bombardment.[26−28] Formation of high-density ASW
has been reported to occur at very low deposition temperatures (<30
K).[29−31] When the deposition temperature is increased from
77 to 200 K, the deposited ice phase changes from porous ASW (p-ASW)[1] to collapsed ASW (c-ASW) and eventually to crystalline
ice. We note that all of these phases exhibit a low density of around
0.94 g·cm–3.[32,33] The heating
process of p-ASW, initially formed at 77 K or below, is accompanied
by a pore collapse, hence transforming from p-ASW to c-ASW at temperatures
above 120 K.[11,21,34,35]ASW is known to be structurally similar
to low-density amorphous
ice (LDA) and hyper-quenched glassy water (HGW). All three types of
solid water have a first diffraction peak at 1.7 Å–1 in neutron scattering[36] and present a
glass transition at around 136 K.[37] However,
it is not yet fully understood how the p-ASW builds up during vapor
deposition nor how the long-range molecular structure changes during
the pore collapse.X-ray diffraction (XRD) and neutron scattering
techniques are powerful
and non-destructive methods to study molecular-level structures of
different materials. They have been previously applied to obtain the
structure factors and the pair distribution functions (PDFs) of low-
and high-density amorphous ices (LDA and HDA).[37−40] As ASW is of high astrophysical
relevance, most studies investigate the vibrational states via Fourier transform infrared spectroscopy (FTIR), while
X-ray and neutron scattering studies are, so far, limited.[11,30,34,36] Taking advantage of the high flux of the high-energy X-rays at beamline
6 ID-D of the Advance Photon Source (APS), we were able to perform in situ X-ray measurements on p-ASW during vapor deposition.The current work presents a combined study of in situ XRD and FTIR on ASW, investigating its formation and the pore collapse
during heating. Oxygen–oxygen PDFs of ASW were determined,
as a function of temperature, up to 23 Å.
The PDFs were then compared with other amorphous ice forms to investigate
potential similarities in the long-range ordering. The OH-stretch
vibration mode from the FTIR spectrum was recorded as a function of
deposition time and temperature, allowing us to follow the changes
in the vibrational spectrum.
Experimental Methods
Sample Preparation
p-ASW samples
for both XRD and FTIR measurements were prepared by vapor deposition
of water molecules onto a cold surface. We used ultrapure Milli-Q
water with 100% H2O composition for the XRD measurements
and an isotope diluted sample for the FTIR measurements, with 5 wt
% HOD in H2O. Deuterium oxide (D2O, 99.9 atom
% D) was purchased from Sigma-Aldrich. For the FTIR measurements,
the sample was additionally degassed by one freeze-pump-thaw cycle.A liquid N2-flow cryostat (JANIS VPF-100, Janis Research
Inc.) was used as a sample environment, which is capable of operating
in vacuum between 77 K and room temperature. Optical windows allowing
for X-ray transmission (Kapton 75 μm and Diamond 50 μm)
or IR transmission (CaF2 1 mm) were used, respectively.
Water was filled and sealed in a glass vial prior to deposition. Water
vapor is guided into the vacuum chamber using a steel tube (1/16″)
controlled by a needle valve. The steel tube is directed toward the
window at around 45°, 10 mm away from the window. Once a vacuum
of at least 10–3 mbar and a temperature of 80 K
were reached, vapor deposition started by opening the valve. Thereby,
p-ASW formed on the cold window surface and grew continuously over
time. Co-deposition of impurities due to the relatively high base
pressure (10–3 mbar), the usage of non-degassed
water, and the use of Kapton windows at the cryostat for the X-ray
measurement cannot be excluded but are not expected to have an impact
on the XRD measurement.The temperature Tm was measured by
a Si diode. For heating, a resistive heater cartridge (50 Ω)
is used. Both are mounted directly at the copper mount of the cold
finger. The temperature is controlled by a PID controller (LakeShore
335). However, the Si diode is placed almost 2 cm away from the window
itself. According to our previous temperature calibration for the
same sample holder, using a powered ice sample, there is an offset
of 10 ± 2 K between the sample temperature (Ts) and the measured temperature (Tm), i.e., Ts ≈ Tm + 10 K (see the Supporting Information of ref (41)) when
using Diamond or Kapton windows. This is also confirmed by FTIR measurements
on p-ASW deposited on a Kapton window (see the Supporting Information
and Figure S1). Additional FTIR measurements
were done using a 1 mm-thick CaF2 window that assured a
better thermal contact between the sample holder and the window. We
assume the Ts at the CaF2 window
to be almost the same as Tm due to a good
thermal contact and a value within the experimental error of ±2
K.
High-Energy X-ray Diffraction (XRD)
To investigate the structural evolution of ASW during the vapor deposition
and heating processes, in situ XRD experiments were
performed at beamline 6 ID-D of the Advance Photon Source (APS). Water
vapor was deposited on Kapton and Diamond windows for approximately
4 and 2 h, respectively. After the vapor deposition, the p-ASW sample
was warmed up to 150 K in steps of 2 K with an overall heating rate
of around 0.33 K/min for XRD measurements. The heating rate is estimated
by taking into account the accumulation time, heating steps, and dark
images taken by the detector in between the measurements.An
unfocused X-ray beam with a diameter of 0.5 mm and a photon energy
of 100 keV was chosen. Diffraction patterns were recorded on a large
2D detector (Perkin-Elmer XRD1621), with 2048 × 2048 pixels and
a pixel size of 200 μm. This allows for a high momentum transfer Qmax (23 Å–1). Cerium dioxide was measured for Q-calibration.
XRD patterns were collected by accumulating up to 180 individual 2D
images, each with an exposure time of 1 s. Background scans were acquired
for Kapton and Diamond windows, respectively. After dark image subtraction
and azimuthal integration, the intensity I(Q) vs Q curves for ASW were obtained. The
oxygen–oxygen structure factor SOO(Q) of ASW and its corresponding pair distribution function gOO(r) were derived from the measured intensity I(Q), following the method described in
detail in earlier literature studies.[40,42−44] Briefly, the total structure factor is derived by applying a weighting
function after first subtracting the background and the molecular
form factor.[40] The PDFgetX2 software[45] was used to apply standard corrections for self-absorption,
oblique incidence, and detector efficiency. Finally, the oxygen–oxygen
structure factor SOO(Q) was calculated
by subtracting the oxygen–hydrogen contribution from the total
S(Q), taken from isotope-substituted neutron scattering
data for liquid water.[46] A detailed description
of the data analysis can be found in our previous publication;[40] the current data set was taken during the same
beam time and analyzed in the same way. The LDA data[42] shown in Figure a were collected under identical conditions.
Figure 2
(a) Full-range structure factor, SOO(Q), at 80 K for p-ASW deposited on Kapton at 80 K for 4
h. LDA-I and
LDA-II (Mariedahl et al.)[42] are shown for comparison; both samples were measured under the same
conditions. (b) SOO(Q) curves at a short
Q-range. (c) Corresponding pair distribution function, gOO(r). The inset in (c) shows the short-range correlations.
(d) r2(gOO(r)-1) curves to help visualize the eight coordination shells up to
23 Å.
Fourier Transform Infrared Spectroscopy (FTIR)
Similar
experiments were performed on an FTIR spectrometer (Frontier,
PerkinElmer) using the same sample environment, with both Kapton (see
the Supporting Information) and CaF2 windows. The FTIR transmission spectra of ASW were measured
during vapor deposition for 115 min (Figure ) and 260 min (see the Supporting Information) and the following heating processes.
The p-ASW sample was heated to distinct temperatures between Tm = 80 and 155 K. The sample was kept at each
temperature for 10 min to equilibrate before recording the FTIR spectra.
Additionally, we performed a heating-quenching procedure, hence quenching
to 80 K in between each heating step, to allow for a same-temperature
comparison.
Figure 1
(a) Time evolution of the I(Q) curves of p-ASW on Kapton during vapor deposition at 80 K. The
colored arrow indicates the color code for the deposition time from
0 to 255 min. The inset shows the low Q region (0.42 Å–1 < Q <
1.1 Å–1). (b) Time evolution
of the FTIR spectra of p-ASW on CaF2 during vapor deposition
at 80 K for 115 min. The inset shows the spectra after normalization
based on the 3250 cm–1 peak.
(a) Time evolution of the I(Q) curves of p-ASW on Kapton during vapor deposition at 80 K. The
colored arrow indicates the color code for the deposition time from
0 to 255 min. The inset shows the low Q region (0.42 Å–1 < Q <
1.1 Å–1). (b) Time evolution
of the FTIR spectra of p-ASW on CaF2 during vapor deposition
at 80 K for 115 min. The inset shows the spectra after normalization
based on the 3250 cm–1 peak.Each spectrum was recorded in a spectral range of 6000–1300
cm–1 with a resolution of 2.0 cm–1, using an acquisition of six scans with a total duration of 60 s.
A background spectrum was taken just before the start of the vapor
deposition. Background correction and data acquisition were done using
the software “Spectrum 10.5.1” from PerkinElmer.
Results and Discussion
p-ASW Formation and Growth
Figure a shows
the development
of the intensity vs momentum transfer curves, I(Q), of p-ASW deposited on a Kapton window
at 80 K over 255 min. It is observed that two broad diffraction maxima
at Q = 1.71 and 3.05 Å–1 appear after the vapor deposition starts and grow
continuously with deposition time. These two peaks are characteristic
for low-density amorphous ice (LDA) and are correlated to its tetrahedral
structure.[42] The same values were found
for p-ASW and LDA samples in earlier work by X-ray and neutron scattering.[36,40] Thus, we can say that the microstructure of deposited p-ASW is similar
to LDA. We also observe an intensity increase along with deposition
time (see inset) at small scattering angles (Q <
0.7 Å–1). This is related
to p-ASW, as deposited at 80 K, being of heterogeneous nature and
forming a porous matrix. In the literature, the degree of porosity
was estimated using small-angle neutron scattering (SANS)[11,34] at around 0.1 Å–1. However,
we were unable to perform such estimations as in the present study
we have a limited Q-range for we focused on the wide
angle, i.e., at Q values above 0.4 Å–1.Figure b
shows the FTIR spectra of p-ASW deposited on a CaF2 window
at 80 K under similar conditions, recorded in a range of 1300–4000
cm–1. The growth of p-ASW is observed from the intensity
increase of the main peak located at ν3 = 3250 cm–1, assigned to OH-stretching vibrations. Similar results
were also observed for p-ASW deposited on Kapton and Diamond windows
at 80 K (Figures S1 and S2). These vibrations
consist of intermolecularly coupled OH oscillators and are related
to transverse optical vibration modes (TO).[4,47−51] The position and line shape of the here reported spectra are consistent
with literature values,[4,47,48,51] taking into consideration that the exact
position depends on the deposition rate and angle.[4,15] The
data presented in Figure b were recorded for a deposition time of 2 h, instead of 4
h, to avoid saturation of the OH-stretch peak. From the corresponding
absorbance of the OH-stretching vibration at 3250 cm–1 during the 2 h vapor deposition (Figure S3a), it was observed that the absorbance increases almost linearly
after the first 10 min. We determined the thickness of the ASW film
as shown in Figure S3b to increase with
a rate of around 0.016 μm/min, resulting in a total thickness
of 3 μm. To provide a direct comparison to the X-ray data, the
spectra for a 4 h deposition are shown in Figure S4. It is observed that the absorbance of the OH-stretching
vibration starts to saturate after depositing for more than 2 h.The normalized data (inset in Figure b) show that with increasing thickness of
the p-ASW layer, the shoulder at 3400 cm–1 also
grows. This is also visible when normalizing the data to the OD band
(see Figure S3c), as the ratio between
the main peak and the shoulder changes continuously. This band has
been related to different molecular motions. It was identified as
a combination band by Hardin and Harvey,[47] and later related to longitudinal vibrations (LO) by comparison
to the TO–LO splitting in crystalline ices.[48] It has also been suggested, that the line shape of the
OH-stretch region is caused by intermolecular vibrational coupling.[49] In addition, the surface modes at 3549 and 3503
cm–1 are assumed to give a broadening effect to
the blue-shifted side of the OH-stretch signal.[25] Finally, we conclude from our experiments that the observed
increase of the shoulder at 3400 cm–1 is also indicative
of the growing porosity of the p-ASW sample during the deposition
process, as this effect happens prior to the saturation of the main
peak (compare to Figure S4). Besides the
shoulder at 3400 cm–1, during deposition, we also
observe the appearance of the well-studied three-coordinated dangling
bond at 3698 cm–1 (non-hydrogen bonded),[25] which is characteristic for p-ASW.[15] As such, an increase of the high-frequency shoulder
at 3400 cm–1 might be related to the appearance
of weaker hydrogen bonds, i.e., less water molecules are found to
be arranged in a rigid tetrahedral structure. This interpretation
is consistent with the temperature dependence of the IR and Raman
spectra of liquid water, which exhibit a decrease of the high-frequency
spectral component upon cooling.[52]By using a solution of 5 wt % HOD in H2O, another peak
at 2440 cm–1 was observed and assigned to the uncoupled
OD-stretch vibration.[48−51] At around 1600 cm–1, the H–O–H bending
mode appears,[53,54] while the relatively weak peak
at around 2230 cm–1 has been assigned to a bending
+ vibration combination band.[55]
Structure of P-ASW
The structure
factor, SOO(Q), of p-ASW after 255 min
is shown in Figure a,b (blue line). It is compared with two
different LDA samples, namely, the so-called LDA-I (black line) as
derived from unannealed high-density amorphous ice (HDA) by heating
above its phase transition temperature and LDA-II (red line), prepared
through decompression of very high-density amorphous ice at 140 K.[42,56] The hydrogen bond networks of LDA-I and LDA-II were found to deviate
at intermediate length scales.[56,57]Figure a shows SOO(Q) at 80 K for p-ASW, LDA-I, and LDA-II over the full momentum transfer
range, up to Q = 23 Å–1. The similarity of the overall oscillations of the
three SOO(Q) curves suggests that the
structure of the deposited p-ASW sample resembles that of the LDA
structure. However, the intensities for the first three diffraction
maxima (Figure b)
are slightly different for the three differently prepared LDA samples.
In particular, the height of the first two maxima for p-ASW is lower
than those for LDA-I and LDA-II. Note that the minor peaks at around
11 and 14 Å–1 are due to
the low signal-to-noise ratio of the thin ASW layer that is approximately
5 μm thick (Figure S3b). A direct
comparison to eHDA can be found in Figure S5, demonstrating that p-ASW is clearly distinct from high-density
amorphous ice.(a) Full-range structure factor, SOO(Q), at 80 K for p-ASW deposited on Kapton at 80 K for 4
h. LDA-I and
LDA-II (Mariedahl et al.)[42] are shown for comparison; both samples were measured under the same
conditions. (b) SOO(Q) curves at a short
Q-range. (c) Corresponding pair distribution function, gOO(r). The inset in (c) shows the short-range correlations.
(d) r2(gOO(r)-1) curves to help visualize the eight coordination shells up to
23 Å.Figure c shows
the corresponding oxygen–oxygen pair distribution function,
gOO(r), obtained through Fourier transformation
(FT). The difference in SOO(Q) is observed
as a small difference in gOO(r) in the
second coordination shell. The inset in Figure c shows gOO(r) values of the first and second coordination shells at 2.74 and
4.5 Å, respectively. It is well known that
the second coordination shell is connected to the tetrahedrality of
the hydrogen-bond network structure.[40,58] For p-ASW
(blue curve), we observed a slightly higher gOO(r) value at around 3.7 Å, compared
to LDA. The difference in SOO(Q) and the
enhancement in gOO(r) at 3.7 Å are
caused by the microporous nature of the p-ASW sample and relates to
a lower degree of tetrahedrality for p-ASW. The second peak in the
gOO(r) of HDA is also located at this
value and assigned to the appearance of interstitial, non-tetrahedrally
hydrogen-bonded molecules.[40] To emphasize
the coordination shells up to 23 Å, r2(gOO(r)-1) is shown
in Figure d.We were able to resolve up to eight coordination shells due to
the wide Q-range accessed in the measurement. The
small oscillations that are visible in the r2(gOO(r)-1) curve for p-ASW are
caused by the lower signal/noise ratio in the micrometer-thick ice
film. Still, we consider the overall structure of p-ASW to be very
similar compared to LDA-I and LDA-II, as all coordination shells are
found to be located at similar distances. Therefore, we can conclude
that the molecular structure of the vapor-deposited p-ASW sample deposited
at 80 K exists within the same energy landscape megabasin of LDA.
Our results are consistent with previous findings,[40,42,56] and additionally extend the previously explored
coordination shells up to a distance of 23 Å.
Structural Evolution of ASW during the Heating
Process
To investigate the crystallization behavior of p-ASW,
we followed its diffraction patterns during heating in steps of 2
K. Figure a shows
the I(Q) curves of selected temperatures,
in the small Q region (0.42 Å–1 < Q < 1.0 Å–1), of p-ASW when heating it up to Tm = 150 K (assumed that Ts = 160 K). Opposite to the trend observed in the inset of Figure a, the intensity
of the I(Q) curves decreases continuously
with increasing temperature up to Tm =
130 K (assumed that Ts = 140 K, black
curve). This intensity decrease is most obvious at Q < 0.7 Å–1, indicating
that the formed micropores (p-ASW) steadily collapse (c-ASW) when
heating up to Tm = 130 K. Above 130 K,
the intensity at Q < 0.7 Å–1 decreases less with increasing temperature, meaning
that most of the pores collapsed during the previous heating process
(80 K < Tm < 130 K).
Figure 3
(a) I(Q) curves in the small Q region
(SAXS) of selected temperatures of ASW deposited
on Kapton at 80 K for 255 min. The arrow indicates the color code
for Tm from 80 to 150 K. (b) Representative
structure factor, SOO(Q), for ASW (same
as in (a)) during heating. The corresponding gOO(r) and r2(gOO(r)-1) for ASW are plotted in (c) and (d), respectively.
The inset in (c) shows the short-range correlations. *Note: the temperature
was measured at the cold finger (Tm);
we assume the sample temperature (Ts)
to be warmer by 10 ± 2 K (Ts ≈ Tm + 10 K).
(a) I(Q) curves in the small Q region
(SAXS) of selected temperatures of ASW deposited
on Kapton at 80 K for 255 min. The arrow indicates the color code
for Tm from 80 to 150 K. (b) Representative
structure factor, SOO(Q), for ASW (same
as in (a)) during heating. The corresponding gOO(r) and r2(gOO(r)-1) for ASW are plotted in (c) and (d), respectively.
The inset in (c) shows the short-range correlations. *Note: the temperature
was measured at the cold finger (Tm);
we assume the sample temperature (Ts)
to be warmer by 10 ± 2 K (Ts ≈ Tm + 10 K).The structure factor SOO(Q) of ASW
at representative temperatures is plotted in Figure b. The positions of the Bragg peaks for hexagonal
ice (ice Ih)[59] and cubic ice (ice Ic)[60] are marked in Figure b as red and black vertical lines, respectively.
Although the micropore collapse is observed at temperatures below Tm = 130 K (Figure a), no pronounced structural change is seen
on the molecular level, evidenced by the almost identical SOO(Q) curves (Figure b). Crystallization is observed at temperatures above Tm = 130 K (Ts =
140 K) by the appearance of Bragg peaks at Q = 2.79
and 3.28 Å–1, both of which
are related to ice Ic and ice Ih.[42] Some
Bragg peaks strictly related to ice Ih, e.g., Q = 3.04 Å–1,
are observed at higher temperatures. Thus, we conclude that ASW transforms
from LDA to the so-called stacking disordered ice I (ice Isd) at around Ts = 140 K.[60−66] A direct comparison of crystalline ice derived from p-ASW (warmed
to Ts = 160 K), LDA-I (warmed to Ts = 150 K), and LDA-II (warmed to Ts = 160 K) is shown in Figure S6. While the ice Ih peak at Q = 3.04 Å–1 is absent in the crystalline ice derived
from LDA-II, it is most pronounced in the LDA-I sample and appears
as a minor feature in SOO(Q) of ice Isd (p-ASW heated to Ts = 160 K),
featuring a mixture of both the SOO(Q)s
for ice Ih and ice Ic.[42,62] Note that this terminology is
used to discriminate the different topological structures of ice I.[60−62,66,67]The gOO(r) and r2(gOO(r)-1) curves for ASW
at the
selected temperatures are plotted in Figure c,d, respectively. Similar to SOO(Q), almost no changes were observed for the gOO(r) curves at temperatures below Tm = 130 K. Above Tm = 130 K, the height
of the second coordination shell (r = 4.5 Å) increases alongside temperature (inset in Figure c). This observation
can be associated with the ordering of two adjacent tetrahedral structures.
Meanwhile, the broad fifth and sixth coordination shells start to
sharpen. This trend is consistent with the results from super-cooled
water studied by Pathak et al.,[68] where a similar fifth peak of water at r≈ 11 Å is associated with five-membered
pentamer rings in low-density liquid-like structures. Thus, the fifth
and sixth coordination shells are sensitive to the long-range order
structures. At the same time, a feature at r = 5.3 Å appears at Tm >
135
K (Ts > 145 K), indicating the presence
of hexagonal stacking. The previous observation supports the SOO(Q) results in Figure b, which show that ASW changes from LDA to ice Isd.For all X-ray measurements shown in Figures and 3, water vapor
was deposited on Kapton. We also used a Diamond window as an alternative
substrate, where ASW was tracked during vapor deposition and heating
using X-rays. The corresponding SOO(Q),
gOO(r), and r2(gOO(r)-1) are summarized in Figures S7 and S8. It is observed that deposited
p-ASW transformed into ice Isd already during the deposition
process (Figure S7a). This quick crystallization
might be caused by the lower energy barrier to form a crystalline
structure on the diamond surface, which on the other hand, seems unlikely
in a micrometer-thick layer of ice. Another cause could be the weak
adhesion between ice and diamond, leading to a bad thermal contact.
When heating up, the ice Isd restructures and increases
the fraction of ice Ih with increasing temperature evidenced by the
additional Bragg reflections (Figure S8a).In an independent measurement, the FTIR spectra were recorded
to
track the vibrational changes of the OH and OD stretching regions
of ASW during heating. Figure a shows the absorption spectra collected at different temperatures
during the heating process. With the increasing temperature from 80
to 130 K, it is observed that the FTIR peak for the OH stretch mode
at ν3 = 3250 cm–1 narrows, and
the dangling OH group at 3698 cm–1 disappears. Hence,
the microporous structure for p-ASW collapses during heating, in agreement
with the literature studies.[11,15,34,69,70] In the inset of Figure a, the peak positions of OH and OD stretch modes are plotted
as a function of temperature. There, we observe a shift toward lower
frequency at 150 K due to crystallization. Similar peak shift behavior
was reported by Mishima and Suzuki[71] and
Salzmann et al.[12] in their
Raman spectroscopy results. Also, the OH bending mode clearly changes
with crystallization, as seen in the quench-recovered spectra recorded
at 80 K (Figure b).
While the bending mode of ASW shows two separated peaks, they merge
to one broader band at 145 K.[53]Figure c shows the normalized
absorption spectra of OH stretch modes for the ASW sample after quenching
the sample to 80 K from different temperatures. The observed collapse
is even more evident after quenching than in the measurements made
at high temperatures (Figure c).
Figure 4
(a) FTIR spectra of ASW during step-wise heating from 80 to 155
K. Inset in (a) presents the peak positions of the OH- and OD-stretch
modes during heating and quenching. (b, c) FTIR spectra of OH-bend
and OH-stretch modes for ASW, respectively, after quenching** to 80 K from different temperatures. The spectra in (c) is normalized
to the main peak. Note that temperature was measured at the cold finger
(Tm); we assume the sample temperature
(Ts) at the CaF2 window to
be almost the same (Ts = Tm) due to a good thermal contact.
(a) FTIR spectra of ASW during step-wise heating from 80 to 155
K. Inset in (a) presents the peak positions of the OH- and OD-stretch
modes during heating and quenching. (b, c) FTIR spectra of OH-bend
and OH-stretch modes for ASW, respectively, after quenching** to 80 K from different temperatures. The spectra in (c) is normalized
to the main peak. Note that temperature was measured at the cold finger
(Tm); we assume the sample temperature
(Ts) at the CaF2 window to
be almost the same (Ts = Tm) due to a good thermal contact.Above 140 K, the FTIR peaks for both the OD and OH stretch modes
are red-shifted, indicating stronger hydrogen bonds, and can be explained
by the distortion of the hydrogen bonds in ASW compared to crystalline
ice, a phenomenon well discussed in the literature.[47−49] This trend
has been previously observed during crystallization of hyperquenched
glassy water when heating from 40 to 150 K at 5 K/min.[69] In liquid water, a similar behavior has been
reported, where a colder sample exhibits a more red-shifted absorption
frequency.[19] Generally, the positions of
the OD and OH stretch modes on the FTIR spectrum of ice derived from
ASW at 155 K are very similar to the ones for ice Ih,[50] but their widths are much broader. In addition, we observe
that the OH bending mode in the region of 1600 cm–1 is split in two peaks in the ASW spectrum at temperatures below
145 K (Figure ).
Conclusions
In this study, we performed in situ high-energy
XRD experiments for ASW during vapor deposition and heating processes,
in which the ASW’s long-range structure factor and oxygen–oxygen
correlations were determined, up to a distance of r = 23 Å. We have also recorded, in an independent in situ FTIR experiment, the OH and OD stretch vibration
modes, using the same sample chamber as in the XRD measurements. We
have demonstrated that the combination of in situ XRD and FTIR measurements allows for a comprehensive description
of the structural evolution of the vapor-deposited p-ASW.A
microporous structure (p-ASW) was observed to build up during
the vapor deposition at 80 K over 250 min. The growth of the deposited
layer was observed through an increase in the total X-ray intensity
(I(Q)) as well as an increase in
absorbance in the OH-stretch vibrational region at around ν3 = 3250 cm–1. The porous characteristic
of the p-ASW sample was observed through an increased SAXS intensity
(Q < 0.7 Å–1) in the XRD measurements, as well as the appearance of dangling
OH bonds at 3698 cm–1. When increasing the temperature,
the micropores collapse, which is seen as a decrease of the shoulder
at 3400 cm–1, vanishing of the dangling OH bond,
and a decrease in SAXS intensity. This result is consistent with a
previously reported SANS experiment.[11,34] The pair correlation
functions gOO(r) of p-ASW and LDA are
almost identical up to 23 Å (Figure c), which is consistent with the fact that
both ASW and LDA have the same density.[32] The main structural difference is seen in the region around 3.7
Å and indicates the existence of interstitial molecules in the
porous ASW structure, as tetrahedrally coordinated LDA does not have
an enhancement at this distance.[40] The
structure of p-ASW is also clearly distinct from HDA, as can be seen
in a direct comparison in Figure S5.Above a sample temperature of 140 K, we observed that the fourth,
fifth, and sixth peaks in the PDF are the most sensitive to the onset
of crystallization, indicating the growth of the local tetrahedral
network at length scales of 8–16 Å (black curve in Figure d). In particular,
the growth of the fifth shell at 12 Å is related to the formation
of hexagonal rings.[68] At temperatures above
145 K, the second, fourth, and sixth peaks, they each split into two
peaks throughout crystallization (Figure d). By comparing with PDFs of crystalline
ice derived from different LDA ices, we observe that ASW transforms
into ice Isd above 140 K (see the Supporting Information). This study is of relevance for understanding
the microstructure of ASW as a major component involved in chemical
processes in interstellar ices.
Authors: Tamsin L Malkin; Benjamin J Murray; Andrey V Brukhno; Jamshed Anwar; Christoph G Salzmann Journal: Proc Natl Acad Sci U S A Date: 2012-01-09 Impact factor: 11.205
Authors: Anita Zeidler; Philip S Salmon; Henry E Fischer; Jörg C Neuefeind; J Mike Simonson; Thomas E Markland Journal: J Phys Condens Matter Date: 2012-06-27 Impact factor: 2.333
Authors: Christian Mitterdorfer; Marion Bauer; Tristan G A Youngs; Daniel T Bowron; Catherine R Hill; Helen J Fraser; John L Finney; Thomas Loerting Journal: Phys Chem Chem Phys Date: 2014-08-14 Impact factor: 3.676
Authors: Leonardo Del Rosso; Milva Celli; Francesco Grazzi; Michele Catti; Thomas C Hansen; A Dominic Fortes; Lorenzo Ulivi Journal: Nat Mater Date: 2020-02-03 Impact factor: 43.841
Authors: Catherine R Hill; Christian Mitterdorfer; Tristan G A Youngs; Daniel T Bowron; Helen J Fraser; Thomas Loerting Journal: Phys Rev Lett Date: 2016-05-26 Impact factor: 9.161
Authors: Daniel Mariedahl; Fivos Perakis; Alexander Späh; Harshad Pathak; Kyung Hwan Kim; Chris Benmore; Anders Nilsson; Katrin Amann-Winkel Journal: Philos Trans A Math Phys Eng Sci Date: 2019-06-03 Impact factor: 4.226
Authors: Aigerim Karina; Tobias Eklund; Christina M Tonauer; Hailong Li; Thomas Loerting; Katrin Amann-Winkel Journal: J Phys Chem Lett Date: 2022-08-18 Impact factor: 6.888