Article
pubs.acs.org/JPCC
Surface Charge Affects the Structure of Interfacial Ice
Emmanuel Anim-Danso, Yu Zhang, and Ali Dhinojwala*
Department of Polymer Science, The University of Akron, Akron, Ohio 44325-3909, United States
S Supporting Information
*
ABSTRACT: We have conducted studies on the freezing of
water molecules next to charged surfaces to elucidate the effect
of water orientation on the structure of ice using sum
frequency generation spectroscopy. We observed that when
water is frozen next to a positively charged sapphire surface,
the signal intensity of ice is higher than that of liquid water as
expected from previous theoretical studies. However, when
water is frozen next to a negatively charged sapphire surface
(using NaOH as pH adjuster), the signal intensity decreases.
The same signal attenuation upon freezing is obtained when
cesium hydroxide (CsOH) and tetramethylammonium hydroxide (N(CH3)4OH) are used as pH adjusters. Since Na+, Cs+, and
N(CH3)+4 ions have different hydration properties, the cation specific effect for this attenuation in signal intensity for ice is ruled
out. Experiments on a mica surface (inherently negatively charged) also showed similar attenuation in signal intensity for ice as
negatively charged sapphire surface. We conclude that the orientation of the water molecules next to a surface plays an important
role in the structure of ice. These results have important implications in understanding the strength of ice nucleation and strength
of ice adhesion next to charged surfaces.
■
differences in freezing temperatures.5 We also reported recently
on the direct measurement of the freezing of water next to
charged Al2O3 (sapphire) surface using sum frequency
generation (SFG) spectroscopy.10 SFG, a second-order nonlinear technique, is only active where there is a breakdown in
inversion symmetry, which makes this technique useful to study
the structure of water molecules near surfaces and interfaces.11−21 In these studies, NaOH and HCl were used to adjust
the pH so that we could measure the freezing behavior of water
next to negatively and positively charged sapphire surfaces,
respectively. We observed that, on negatively charged surface,
the SFG intensity of ice was attenuated in comparison with
water signals before freezing. This was opposite of what was
observed next to positively charged sapphire surfaces. We
concluded that protons were disordered in the ice crystals next
to negatively charged surfaces. However, the origin of proton
disorder was not identified. Two hypotheses were suggested.
First, there is a possibility of Na+ ions migrating to the sapphire
interface and influencing the ordering of frozen water
molecules. The second possibility could be due to the
orientation of water and the fact that the freezing of water
when the hydrogen atoms are pointing toward the surface is
inherently different than that where hydrogens are pointing
inward. Experiments using phase-sensitive SFG (PS-SFG),
which is sensitive to the average orientation of water molecules
at interfaces, observed a flip of the OH bonds toward the silica
surface after melting of ice.22 Therefore, in a situation where
INTRODUCTION
Water is the most important liquid on Earth, and its interaction
with charged surfaces influences key processes in biology,
electrochemistry, geology, and technology.1−3 For biomolecular
self-assembly, it has been shown that aggregation of protein
molecules is controlled by the orientation of water molecules,
which is influenced by the net charge on the protein surface and
not by the protein−protein interactions.2 The interactions of
water molecules with positively and negatively charged surfaces
are not the same.4,5 In a dielectric continuum, the strength of
an electric field should not depend on the sign of the charge.
However, it has been pointed out that the orientation of water
does make a difference, and coincidentally, most macromolecules and lipid bilayers in the cells are partially negatively
charged. In addition, the other most important interfaces, oil/
water or hydrophobic/water interfaces, are also suggested to be
negatively charged due to adsorption of OH−.6−8 Scheu et al.4
recently reported an intriguing asymmetry for tetraphenylborate (TPB−) and tetraphenylarsonium (TPA+) ions in water.
Although these two ions are very similar in size and
hydrophobic character, because of the charge differences,
water molecules form stronger hydrogen bonds with negatively
charged TPB− ions compared with TPA+.4,9
Recently, Ehre et al.5 explored the role of charged surfaces in
freezing of water and showed using specular X-ray diffraction
(XRD) that the ice nucleates differently on positively and
negatively charged surfaces. These differences resulted in a
lower freezing temperature of water next to negatively charged
surfaces. This experiment also suggested that the water
molecules pointing with hydrogen facing toward or away
from the surface influence the formation of ice and the
© 2016 American Chemical Society
Received: August 27, 2015
Revised: January 7, 2016
Published: February 2, 2016
3741
DOI: 10.1021/acs.jpcc.5b08371
J. Phys. Chem. C 2016, 120, 3741−3748
Article
The Journal of Physical Chemistry C
experiment, a thin sheet of mica was peeled off from a thicker
mica sheet stuck to the PS film. This prevented contamination
of the mica surface. The study of mica/water interface in a total
internal reflection geometry using index matching layer was
inspired by the recent paper by Abdelmonem et al.28 Here we
have used deuterated polystyrene instead of index matching
gels.
The details of the SFG system used in this work have been
described in previous publications.29 In brief, a picosecond
Spectra Physics laser system with tunable ∼3.5 μJ IR beam
−2000−3800 cm−1, 1 ps pulse width, 1 kHz repetition rate, and
a diameter of 100−200 μm is overlapped with a ∼70 μJ visible
beam of 800 nm in wavelength, with a 1 ps pulse width, 1 kHz
repetition rate, and a diameter of 1 mm. A home-built
computer-controlled motorized delay stage is used to maintain
the temporal delay as the IR beam is scanned from 2700 to
3800 cm−1. The SFG signals were acquired using a 5 s
averaging time and at intervals of 10 cm−1. Due to the low IR
beam energy used in the water experiments, laser heating of ice
was negligible. The melting temperature of the ice was close to
0 °C and served as a good way to calibrate the temperature.20
The SFG experiments were performed using equilateral 60°
angle sapphire prisms at or near total internal reflection. The
sapphire prisms are cut with the crystal axis to avoid any
birefringence effect. A photomultiplier tube connected to a
spectrometer 0.5 m in length was used to collect the SFG
signals. The spectra were collected at an IR incidence angle of
16° with respect to the normal of the sapphire prism face. The
incident angle of the visible laser beam was ∼1.5° lower than
the incident angles for the tunable IR laser. The polarization
combinations reported in this work are SSP (s-polarized SFG
output, s-polarized visible input, and p-polarized IR input) and
PPP (p-polarized SFG output, p-polarized visible input, and ppolarized IR input). The SSP and PPP polarizations provide
complementary information about the molecular susceptibility
tensor and are useful in interpreting the orientation of
molecules. A model to interpret SSP polarization results in
internal reflection geometry has been provided in a previous
publication.30
We have used a Lorentzian fitting function to fit our data (eq
1).26
the water molecules are strongly bonded to a negatively
charged surface, such a flip of the water dipoles during freezing
could also cause the protons to disorder. The orientation of
water can create a barrier in forming proton-ordered ice crystals
because it has to overcome the strong electrostatic interactions
to rearrange the water molecules.
To distinguish whether the ions or water orientation is a
dominant factor in attenuation of the ice signals, we present
here experimental data on two other ions which are known to
be weak disrupters of water structure based on the Hofmeister
series,23,24 cesium (Cs+) and tetramethylammonium (N(CH3)+4 ) ions. We use hydroxides of these two ions to adjust
the pH such that the sapphire surface is negatively charged. In
addition, we studied an ammonia/water solution next to
sapphire surface because it is known that ammonia is able to
bond to the aluminol groups and polar order water molecules
with the OH groups of the ice facing toward the bulk.22 This
control of water orientation should exhibit strong protoordered ice structure and a high intensity of the ice peak.
Finally, we will discuss SFG spectra for water and ice next to a
negatively charged mica surface. Since mica is negatively
charged, we can test the effect of surface charge on ice without
any interference from ions, and the structure of ice should be
directly related to the surface charge.
■
EXPERIMENTAL SECTION
Ultrapure water with a resistance of 18.2 MΩ·cm (Millipore
filtration system with deionizing and organic removal columns)
was used in these experiments. CsOH (≥99.9% purity, SigmaAldrich), N(CH4)4OH (25 wt % in H2O, Sigma-Aldrich), and
NH3OH (≥28−30% NH3 basis, Sigma-Aldrich) were used as
obtained for pH adjustments. The pH measurements were
performed with an Oakton pH meter with an epoxy-body pH
electrode.
Sapphire has a high surface energy, and it is easier for
adventitious carbon to adsorb on the surface. Care was taken to
clean the sapphire surfaces by sequentially sonicating in
acetone, ethanol, and deionized water for 1 h each.25−27 The
sapphire prisms were then thoroughly rinsed in deionized water
and blow-dried with dry nitrogen gas. The prisms were then
plasma treated for 4 min. We used a stainless steel cell to hold
the sapphire prisms and water. The volume and surface area of
the cell are ∼300 μL and 0.3 cm2, respectively. The cell was
cleaned using the same procedure we used for sapphire prisms
just before the experiments. Our cooling and heating
experimental setup has been described earlier, and the diagram
of this setup is provided in the Supporting Information (Figure
S1).10,21 The temperature stage was purchased from Instec Inc.
and modified in-house to hold the sapphire prisms. We have
used 0.3 °C/min cooling or heating rate in most cases, and for
some selected exceptions, the cooling or heating rates are
provided in the text or figure captions. The process used to
calibrate the temperature of the sample cell has been provided
in earlier publication.21
Muscovite mica was obtained from Lawrence Company,
New Bedford, MA. To perform total internal reflection
geometry experiment, mica was adhered onto a sapphire
prism by first spin coating a polystyrene film on a clean
sapphire prism. Then a drop of toluene was placed on the top
of the film. A thin sheet of mica, cut to the size of the sapphire,
was carefully placed on the toluene-soaked PS film. This tacky
film was a good adhesive to hold the mica in place. The sample
was then annealed at 120 oC for 12 h. Prior to performing an
ISFG ∝ χeff,NR +
∑
Aq
ωIR − ωq − i Γq
|2
(1)
In eq 1, χeff,NR describes the nonresonant contribution. Aq, Γq,
and ωq are the amplitude, damping constant, and angular
frequency of the qth vibrational resonance, respectively.
The use of total internal reflection geometry and scanning
the IR beam over a broad range of wavelength could result in
Fresnel factors that are a function of wavelength.30,31 The
changes in SFG intensity (SSP polarization) due to the Fresnel
factors are provided in the Supporting Information (Figures S2
and S3). The calculation predicts the SFG intensity for ice to be
a factor of 2 higher than water. The comparison of the SFG
data reported here is not corrected for the changes in Fresnel
factors. However, to understand the influence of the changes in
the SFG intensity due to Fresnel factors, we have shown the
ratios of the amplitude strength and SFG intensities before and
after correcting for the Fresnel factors (Table S1). The incident
angles used in these experiments are provided in Table S2.
3742
DOI: 10.1021/acs.jpcc.5b08371
J. Phys. Chem. C 2016, 120, 3741−3748
Article
The Journal of Physical Chemistry C
Figure 1. SFG spectra collected in SSP polarization during the cooling cycle of hydrochloric acid (HCl) and sodium hydroxide (NaOH) solutions.
The changes in the SFG spectra were measured for pH 3.3 (a) and 9.8 (b). The SFG data in panels (a) and (b) were collected with temperature
increments of 1 °C (using a 4 °C/min cooling rate) and a 30 min equilibration time. Panel (c) and (d) shows the SFG spectra collected in SSP
polarization during cooling of cesium hydroxide solution (CsOH) and tetramethylammonium hydroxide (N(CH3)4OH) solutions, respectively. The
SFG spectra in panel (c) and (d) were collected at every 1 °C increment and a cooling rate of 0.3 °C/min and a 30 min equilibration. The spectra
for water and ice are shown using symbols of empty squares and solid circles, respectively. The spectra with filled symbols in panel (c) and (d) have
been offset for clarity.
■
the 3450 cm−1 peak is assigned to weak hydrogen-bonding OH
stretching modes of water near the interface.11,32,33 These peaks
are similar to those observed at the water/air interface. The
only difference is the peak at 3720 cm−1. At the water/air
interface, this peak is assigned to “dangling” OH stretching
mode.34 However, at the sapphire/water interface, such a peak
is attributed to free surface OH groups.16,26,29,35 These free
surface OH groups form very strong bonds and hence do not
participate in hydrogen bonding with the interfacial water
molecules.26 These groups are only deprotonated at relatively
high pH.10 This is very unlike the silica silanol groups which,
when immersed in water, even at low pH, are easily
deprotonated (due to silica’s low isoelectric point).11 Recently,
it was reported that, while the assignment for the sapphire
RESULTS AND DISCUSSIONS
Figure 1 shows the spectra of water and ice next to positively
(a) and negatively (b, c, and d) charged sapphire surfaces. The
empty squares and solid circles represent spectra of water and
ice, respectively. The positively charge sapphire was prepared
using HCl solution to adjust the pH to 3.3. The negatively
charged sapphire surfaces were prepared using NaOH, CsOH,
and N(CH3)4OH. The data for HCl and NaOH were discussed
in our previous publication; here we briefly describe the peak
assignments for water and ice next to the sapphire interface.
The water spectrum (Figure 1a) has two peaks in the
hydrogen-bonded region (3000−3600 cm−1). The peak at
3200 cm−1 corresponds to strong hydrogen-bonding OH
stretching modes of water molecules at the solid surface, and
3743
DOI: 10.1021/acs.jpcc.5b08371
J. Phys. Chem. C 2016, 120, 3741−3748
Article
The Journal of Physical Chemistry C
Ishiyama and Morita et al.46,47 on ice/vapor interface also
found that charge transfer between the bilayer stitching water
molecules are important for the SFG signal. We note that there
are currently no MD calculations for the solid/ice interface.
However, due to spectral similarities, we have used the same
assignment for the ice formed next to the sapphire surface.
Based on the above assignment, it is clear that the ice spectra
in Figure 1a are quite different from the spectra collected next
to the negative charged surfaces (Figures 1b, c, and d). While
the signal intensity is expected to increase (Figure 1a), on
negatively charged sapphire surfaces, the ice SFG signal
intensity is about an order of magnitude lower than that of
water. This indicates that the ice is proton disordered at the
interface. Ice formed at atmospheric pressure exists as
hexagonal ice (Ice Ih). Though this ice Ih is crystalline, there
remains some residual entropy that allows the water molecules
to rotate.48 This leads to protons being disordered in the ice
crystal. Pauling48 produced a statistical model for the position
of the hydrogen atoms based on the suggestions by Bernal and
Fowler.49 According to the Bernal−Fowler ice rule, a large
number of configurations exist in ice Ih due the many different
distribution of hydrogens in relation to oxygen atoms. Based on
the SFG results, we conclude that the protons are ordered next
to positively charged surfaces while they could be disordered
next to the negatively charged surfaces. Since these results are
similar for both strong and weak ions, we believe that the
surface disorder of ice is not due to a specific cation effect.
We believe that the decrease in the SFG signal for ice
suggests an intriguing possibility that the orientation of protons
toward the negatively charged sapphire substrate may result in
surface ice layer with disordered protons. Ehre et al.5 reported
on the freezing of water molecules on negatively and positively
charged LiTaO3 surface. Using XRD, the authors observed that
freezing was initiated at the water/vapor and water/LiTaO3
interfaces on negatively and positively charged surfaces,
respectively. They attributed this difference in nucleation sites
to the orientation of water molecules on the LiTaO3 surface.
Although we have not directly measured the orientation of the
water molecules next to the sapphire substrate, Wei et al.22 have
used a phase-sensitive SFG to show that, upon melting, the
hydrogen atoms that were facing the bulk flip to point toward
the silica substrate. This implies that for negative charged
surfaces, the water has to flip its orientation upon freezing. We
believe this would lead to a proton disorder, and this is reflected
in a weaker SFG intensity for ice next to negatively charged
surfaces.
To confirm that orientation is playing an important role and
not the cations used to increase pH of the solution, we have
collected SFG spectra for the freezing of water for two
additional systems: an ammonia/water solution and a
negatively charged mica surface. Wei et al. reported that
when an ammonia/water solution was frozen next to a silica
surface, there was a threefold increase in the ice SFG signal
intensity.22 This was explained due to ammonia molecules
hydrogen bonding with the surface silanol groups and orienting
the water molecules with the protons facing the bulk. Su et al.
used similar explanation to describe the strong SFG signal
observed for ice grown on Pt(111) surface.13 Figure 3 shows
the ammonia/water solution and ice spectra next to sapphire
surface. The water spectrum (empty squares) shows highly
ordered water molecules. In aqueous solution, ammonia
deprotonates a small amount of water to form ammonium
hydroxide. These OH− ions deprotonate some of the sapphire
surface OH is correct, some portion of the peak should be
attributed to free water OH oscillators at the sapphire surface.32
We would like to note that there are various interpretations of
the water peaks, especially at the water/air interface.36−39
However, at the solid/water interface, water molecules have
been shown to structure even in the absence of charge.40 In the
presence of charge, the local electric field extends further and
orients water molecules deeper in the bulk.40,41
For the CsOH system, the water spectrum is similar to that
of NaOH solution, and the 3450 cm−1 peak is absent (Figure
1c). There is, however, a small rise at around 3720 cm−1 due to
partial deprotonation of alumnol groups at pH 10. Figures 1d
and 2 show the SFG spectra for a N(CH3)4OH solution with a
Figure 2. SFG spectra collected in PPP polarization during cooling of
tetramethylammonium hydroxide solution (N(CH3)4OH). The water
(empty squares) and ice (solid circles) spectra are collected during the
cooling cycle using a solution of pH 10. The data were collected with
temperature increment of 1 °C (using a 0.3 °C/min cooling rate) and
a 30 min equilibration time, before collecting the SFG spectra. The ice
spectrum has been offset for clarity.
soft cation N(CH3)4+ using SSP and PPP polarizations,
respectively. The liquid spectra in both polarizations are similar
except for the shoulder at 3450 cm−1. In general, SSP
polarization shows stronger symmetric vibration in comparison
to asymmetric vibration. When we freeze the solution, four
peaks are observed in the PPP spectrum at 2933, 3026, 3200,
and 3450 cm−1. The first two peaks, 2933 and 3026 cm−1, are
tentatively assigned to CH3 symmetric and asymmetric stretch,
respectively.42 For the SSP, these two peaks are not clear, and
they are merged with the shoulder of the broad water peaks. It
is possible that the water signal overshadowed any signal from
the C−H stretch of N(CH3)+4 in the water spectrum. It became
apparent only after the significant signal attenuation in ice.
Upon freezing, the ice peak (Figure 1a) which appears at
3150 cm−1 is assigned to hydrogen-bonded stretching modes of
the water molecules. The exact nature of the water molecules
contributing to this intense peak was not known until quite
recently. Through seminal work by Buch et al.43 and Shultz et
al.,44,45 bilayer stitching bonds between water layers on the
basal face (ice/vapor interface) were found to be responsible
for this peak. Molecular dynamics simulation (MD) by
3744
DOI: 10.1021/acs.jpcc.5b08371
J. Phys. Chem. C 2016, 120, 3741−3748
Article
The Journal of Physical Chemistry C
To emphasize the role of negatively charged surface in
controlling the ice formation, we performed experiments on
muscovite mica surface. Mica is an atomically smooth surface
with an intrinsic negative charge density of −0.33 C/m2 due to
the isomorphic substitution of silicon by aluminum atoms.51 In
air, the charge is compensated by potassium and, to a lesser
extent, sodium ions. In aqueous solution, the surface charge is
considerably less than −0.33 C/m2 due to the adsorption of
counterions to the surface charge sites.51 However, in water
with considerably less ions (deionized water), potassium ions
become hydrated and dissociate from the mica surface.52,53 The
surface charge decreases with decreasing pH. The point of zero
charge (PZC) for mica is still under debate, but it is generally
assumed to be 3.54 Based on this, we studied both the water
and ice structures on muscovite mica using deionized water of
pH around 6. The slight acidity is derived from dissolved
carbon dioxide in the solution. Figure 5a shows the spectra of
water on mica in SSP and PPP polarizations. There are two
main peaks. The red-shifted peak at 3150 cm−1 is attributed to
strong hydrogen-bonded water molecules normally referred to
as icelike, and the peak at 3450 cm−1 is assigned to the weak
three- or low-coordinated liquidlike water molecules, similar to
water on sapphire surface. The PPP polarization (empty
squares), which is sensitive to both symmetric and asymmetric
water stretching modes, shows both icelike and liquidlike water
molecules. The SSP polarization (empty circles), which
corresponds to molecules normal to the surface, shows highly
ordered water next to the mica surface. This assessment is in
line with data from scanning polarization force microscopy
(SPFM)55,56 and first principles molecular dynamics simulation
(MD)57 where water molecules adsorbed on mica form stable
icelike monolayers. In the SSP water spectrum (Figure 5a), we
observe an upturn in intensity from 3700−3800 cm−1, and
because this observation is not reproducible (Figure 5b) we do
not assign this upturn to a free OH vibration.
SFG measurement of adsorbed water molecules was first
performed by Miranda et al.58 The authors observed that at
90% RH (relative humidity), the monolayer formed on the
mica surface was an ordered icelike structure similar to what
was predicted by molecular dynamics simulation.57,58 A recent
study of the water/muscovite mica interface using SFG also
reported the observation of icelike water molecules next to mica
surface, and the trend is consistent with the results reported
here.59 A shoulder near 3400 cm−1 has also been observed for
water next to mica surface, which was not observed in these
results,58,59 and this could be due to differences in experimental
geometry. Figure 5b shows the SFG results for water and the
ice structure next to the mica surface using SFG. Upon freezing,
the SFG signal intensity decreases and the peak is slightly blueshifted. This sudden decrease in signal intensity is consistent
with the hypothesis that proton disordering next to negatively
charged surface reduces the intensity of ice. It is interesting that
liquid water next to negatively charged surfaces is highly
ordered due to the surface electric field. However, molecules in
ice are proton disordered due to the strong interaction of water
with the surface and the mechanism of ice formation. We would
like to comment on the structure of water next to mica just
before freezing. Very recently, Abdelmonem et al. reported,
using second harmonic generation (SHG), that water next to
mica increases its structural order before the freezing
transition.28 We show in Figure 5c the SFG signal intensity
at mica/water interface as a function of temperature during
cooling (blue curve) and heating (red curve) at 0.3 °C/min.
Figure 3. SFG spectra collected in SSP polarization during cooling of
ammonia/water solution. The changes in the SFG spectra were
measured for pH 10. Two freezing rates are presented: 5 °C/min
(empty circles) and 10 °C/min (solid circles). For each rate, the
solution is cooled from 20 to −40 °C. The water spectrum (empty
squares) was collected at 20 °C. This figure shows that the
incorporation of ammonia into the ice structure at the sapphire
surface is rate dependent.
hydroxyl groups to make the surface negatively charged. When
the solution is cooled to form ice, the ice SFG signal intensity
rises, contrary to what is observed on negatively charged
surfaces. This further suggests that for the strong ice peak at
3150 cm−1 to be observed, water molecules have to orient with
their OH bonds pointed toward the bulk. This is consistent
with the hypothesis that some of the NH3 molecules are able to
hydrogen bond to aluminol groups and orient water molecules
at the interface.22 We also find that the incorporation of
ammonia molecules into the ice is rate dependent and the SFG
intensity for ice was higher with a faster freezing rate (Figure
3).22 We have not directly observed the N−H symmetric peak
(at 3312 cm−150) possibly due the high intensity of the ice
signal which may have overshadowed the N−H peak. In Figure
4, we show an illustration of how ammonia is able to order
water molecules in ice at the sapphire surface.
Figure 4. Illustration of NH3-doped ice near sapphire/ice interface.
The bound ammonia molecules are able to order water molecules,
significantly increasing the SFG signal intensity of ice next to sapphire
substrate. This illustration is based on work in ref 22
3745
DOI: 10.1021/acs.jpcc.5b08371
J. Phys. Chem. C 2016, 120, 3741−3748
Article
The Journal of Physical Chemistry C
Figure 5. (a) SFG spectra of water on mica collected in SSP (empty circles) and PPP (empty squares) polarizations at 22 °C. (b) SFG spectra
collected in SSP polarization during cooling of water. The data were collected with temperature increment of 1 °C (using a 0.3 °C/min cooling rate)
and a 30 min equilibration time, before collecting the spectra. The spectra for water (0 °C) (empty circles) and ice (−4 °C) (solid circles) near the
freezing point are shown. (c) Variation in SFG intensity are plotted using SSP polarization and at IR frequency of 3150 cm−1 during the cooling
(blue curve, empty circles) and heating (red curve, empty squares) cycles. These measurements were done at cooling and heating rate of
approximately 0.3 °C/min.
reported median transition temperatures that were very close to
homogeneous nucleation temperature of water (−35 °C),
indicating a very weak influence of the substrate on the
nucleation process.61 Atkinson et al. also reported that Kfeldspar is a better nucleator than mica sheets.60 In comparison,
Steinke et al. observed a higher freezing temperature on mica
(−22 °C).62 Because these studies were conducted using
droplets with smaller volume, smaller surface area (4 × 10−9−4
× 10−6 mL and 3 × 10−6−3 × 10−4 cm2),60−62 and faster
cooling rates, this could be the reason for lower freezing
temperature reported in these studies. The results by
Abdelmonem et al. using similar experimental conditions
reported freezing transition temperature of around −12 °C
for mica and sapphire substrates,28 which are similar to the
freezing temperatures reported here. Experiments are in
progress to understand the influence of surface energy and
volume of the sample cell on freezing transition temperatures.
Our experiment shows no apparent increase in signal intensity
as the temperature is decreased and is different from the SHG
results. Perhaps this subtle change in SHG intensity is not
observed in the SFG experiments.28 Finally, we would like to
comment on the melting temperature of ice. In all cases studied
here, the melting temperature is close to 0 °C. Although
surprising, this implies that the proton order or disorder has no
effect on melting temperature.
Finally, we would like to comment on the freezing
temperature observed here for sapphire and mica surfaces in
comparison to published work in the literature. These studies
used different volume, surface area, cooling rates, and
experimental approach, and this could influence the nucleation
process.60−62 Campbell et al. studied freezing of water droplets
on three substrates (silicon, glass, and mica), and they
measured the number of droplets frozen as the sample was
cooled below −15 °C at a cooling rate of 1 °C/min.61 They
3746
DOI: 10.1021/acs.jpcc.5b08371
J. Phys. Chem. C 2016, 120, 3741−3748
Article
The Journal of Physical Chemistry C
■
(4) Scheu, R.; Rankin, B. M.; Chen, Y.; Jena, K. C.; Ben-Amotz, D.;
Roke, S. Charge asymmetry at aqueous hydrophobic interfaces and
hydration shells. Angew. Chem., Int. Ed. 2014, 53, 9560−9563.
(5) Ehre, D.; Lavert, E.; Lahav, M.; Lubomirsky, I. Water freezes
differently on positively and negatively charged surfaces of pyroelectric
materials. Science 2010, 327, 672−675.
(6) Tian, C. S.; Shen, Y. R. Structure and charging of hydrophobic
material/water interfaces studied by phase-sensitive sum-frequency
vibrational spectroscopy. Proc. Natl. Acad. Sci. U.S.A. 2009, 106,
15148−15153.
(7) Tian, C. S.; Shen, Y. R. Recent progress on sum-frequency
spectroscopy. Surf. Sci. Rep. 2014, 69, 105−131.
(8) Samson, J.-S.; Scheu, R.; Smolentsev, N.; Rick, S. W.; Roke, S.
Sum frequency spectroscopy of the hydrophobic nanodroplet/water
interface: Absence of hydroxyl ion and dangling OH bond signatures.
Chem. Phys. Lett. 2014, 615, 124−131.
(9) Thormählen, I.; Straub, J.; Grigull, U. Refractive index of water and
its dependence on wavelength, temperature, and density; American
Chemical Society and the American Institute of Physics for the
National Bureau of Standards: Gaithersburg, MD, 1985.
(10) Anim-Danso, E.; Zhang, Y.; Alizadeh, A.; Dhinojwala, A.
Freezing of water next to solid surfaces probed by infrared-visible sum
frequency generation spectroscopy. J. Am. Chem. Soc. 2013, 135,
2734−2740.
(11) Shen, Y.; Ostroverkhov, V. Sum-frequency vibrational spectroscopy on water interface: polar orientation of water molecules at
interfaces. Chem. Rev. 2006, 94, 1140.
(12) Richmond, G. L. Molecular bonding and interactions at aqueous
surfaces as probed by vibrational sum frequency spectroscopy. Chem.
Rev. 2002, 102, 2693−2724.
(13) Su, X.; Lianos, L.; Shen, Y. R.; Somorjai, G. A. Surface-induced
ferroelectric ice on Pt(111). Phys. Rev. Lett. 1998, 80, 1533−1536.
(14) Hsu, P. Y.; Dhinojwala, A. Contact of oil with solid surfaces in
aqueous media probed using sum frequency generation spectroscopy.
Langmuir 2012, 28, 2567−2573.
(15) Rangwalla, H.; Schwab, A. D.; Yurdumakan, B.; Yablon, D. G.;
Yeganeh, M. S.; Dhinojwala, A. Molecular structure of an alkyl-sidechain polymer−water interface: Origins of contact angle hysteresis.
Langmuir 2004, 20, 8625−8633.
(16) Yeganeh, M. S.; Dougal, S. M.; Pink, H. S. Vibrational
spectroscopy of water at liquid/solid interfaces: Crossing the
isoelectric point of a solid surface. Phys. Rev. Lett. 1999, 83, 1179−
1182.
(17) Du, Q.; Freysz, E.; Shen, Y. R. Surface vibrational spectroscopic
studies of hydrogen bonding and hydrophobicity. Science 1994, 264,
826−828.
(18) Jena, K. C.; Covert, P. A.; Hore, D. K. The effect of salt on the
water structure at a charged solid surface: Differentiating second- and
third-order nonlinear contributions. J. Phys. Chem. Lett. 2011, 2,
1056−1061.
(19) Ye, S.; Ma, S.; Wei, F.; Li, H. Intramolecular vibrational coupling
in water molecules revealed by compatible multiple nonlinear
vibrational spectroscopic measurements. Analyst 2012, 137, 4981−
4987.
(20) Anim-Danso, E.; Zhang, Y.; Dhinojwala, A. Freezing and
melting of salt hydrates next to solid surfaces probed by infrared-visible
sum frequency generation spectroscopy. J. Am. Chem. Soc. 2013, 135,
8496−8499.
(21) Zhang, Y.; Anim-Danso, E.; Dhinojwala, A. The effect of a solid
surface on the segregation and melting of salt hydrates. J. Am. Chem.
Soc. 2014, 136, 14811−14820.
(22) Wei, X.; Miranda, P. B.; Zhang, C.; Shen, Y. R. Sum-frequency
spectroscopic studies of ice interfaces. Phys. Rev. B: Condens. Matter
Mater. Phys. 2002, 66, 085401.
(23) Marcus, Y. Effect of ions on the structure of water: Structure
making and breaking. Chem. Rev. 2009, 109, 1346−1370.
(24) Marcus, Y. The hydration entropies of ions and their effects on
the structure of water. J. Chem. Soc., Faraday Trans. 1 1986, 82, 233−
242.
CONCLUSIONS
Water and ice in contact with charged surfaces are ubiquitous
on Earth, and here we show how that the orientation of water
molecules next to negatively charged surfaces has a profound
effect on the surface structure of ice. By using surface-sensitive
infrared−visible sum frequency generation spectroscopy
(SFG), we show that at high pH the ice signals are attenuated
compared to the water signals. This attenuation is not due to
the cations added to increase the pH. However, it is due to the
orientation of water molecules in contact with negatively
charged surfaces, which leads to a proton disordered ice phase.
We have shown using ammonia that if this orientation of water
was reversed and in the same direction as neutral and positively
charged surfaces, the ice signals are actually enhanced several
orders of magnitude compared to water. On mica, which is
inherently negatively charged, the ice signals were again
attenuated compared to liquid water, indicating that it is
negative charge which is responsible for the attenuation rather
than the cations added to increase the pH. In comparison, the
ice signals are much larger than water signals next to positively
charged surfaces. These results have important implications in
designing surfaces to reduce ice adhesion,5 understanding
processes involved in cloud seeding, and movement of glaciers.
■
ASSOCIATED CONTENT
S Supporting Information
*
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.jpcc.5b08371.
Additional descriptions, tables, and figures as described in
the text (PDF)
■
AUTHOR INFORMATION
Corresponding Author
*E-mail: [email protected].
Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS
We thank Edward Laughlin, Anish Kurian, and Liehui Ge for
help in designing the temperature stage. We appreciate help
from Adrian Defante, Saranshu Singla, and Sukhmanjot Kaur
for help with the mica experiments. We thank Mena Klittich,
Dona Foster, Nishad Dhopatkar, Saranshu Singla, and
Sukhmanjot Kaur for helpful discussions. This work was
supported by NSF-DMR GOALI Grant. We also thank GE
Global Research for providing support to fund the design of the
temperature cell for SFG.
■
REFERENCES
(1) Water. A comprehensive treatise; Franks, F., Ed.; Plenum: New
York, 1973; Vol. 1, The physics and physical chemistry of water. xx,
596 pp., illus. Vol. 2, Water in crystalline hydrates; Aqueous solutions
of simple nonelectrolytes. xx, 684 pp., illus. Vol. 3, Aqueous solutions
of simple electrolytes. xviii, 472 pp. Allen, L. C. Physical chemistry of
water. Science 1974, 184, 152.
(2) Chong, S.-H.; Ham, S. Interaction with the surrounding water
plays a key role in determining the aggregation propensity of proteins.
Angew. Chem. 2014, 126, 4042−4045.
(3) Gaigeot, M.-P.; Sulpizi, M. Liquid−solid interfaces: Structure and
dynamics from spectroscopy and simulations. J. Phys.: Condens. Matter
2012, 24, 120301.
3747
DOI: 10.1021/acs.jpcc.5b08371
J. Phys. Chem. C 2016, 120, 3741−3748
Article
The Journal of Physical Chemistry C
(25) Rabung, T.; Schild, D.; Geckeis, H.; Klenze, R.; Fanghänel, T.
Cm(III) Sorption onto Sapphire (alpha-Al2O3) Single Crystals. J. Phys.
Chem. B 2004, 108, 17160−17165.
(26) Zhang, L.; Tian, C.; Waychunas, G. A.; Shen, Y. R. Structures
and charging of alpha-alumina (0001)/water interfaces studied by
sum-frequency vibrational spectroscopy. J. Am. Chem. Soc. 2008, 130,
7686−7694.
(27) Braunschweig, B.; Eissner, S.; Daum, W. Molecular structure of
a mineral/water interface: Effects of surface nanoroughness of Al2O3
(0001). J. Phys. Chem. C 2008, 112, 1751−1754.
(28) Abdelmonem, A.; Lützenkirchen, J.; Leisner, T. Probing icenucleation processes on the molecular level using second harmonic
generation spectroscopy. Atmos. Meas. Technol. Discuss. 2015, 8, 5265−
5282.
(29) Nanjundiah, K.; Hsu, P. Y.; Dhinojwala, A. Understanding
rubber friction in the presence of water using sum-frequency
generation spectroscopy. J. Chem. Phys. 2009, 130, 024702.
(30) Li, G.; Dhinojwala, A.; Yeganeh, M. S. Interference effect from
buried interfaces investigated by angular-dependent infrared-visible
sum frequency generation technique. J. Phys. Chem. C 2011, 115,
7554−7561.
(31) Liljeblad, J. F.; Tyrode, E. Vibrational sum frequency
spectroscopy studies at solid/liquid interfaces: Influence of the
experimental geometry in the spectral shape and enhancement. J.
Phys. Chem. C 2012, 116, 22893−22903.
(32) Boulesbaa, A.; Borguet, E. Vibrational dynamics of interfacial
water by free induction decay sum frequency generation (FID-SFG) at
the Al2O3(1120)/H2O interface. J. Phys. Chem. Lett. 2014, 5, 528−533.
(33) Tyrode, E.; Liljeblad, J. F. D. Water structure next to ordered
and disordered hydrophobic silane monolayers: a vibrational sum
frequency spectroscopy study. J. Phys. Chem. C 2013, 117, 1780−1790.
(34) Allen, H. C.; Casillas-Ituarte, N. N.; Sierra-Hernandez, M. R.;
Chen, X.; Tang, C. Y. Shedding light on water structure at air-aqueous
interfaces: Ions, lipids, and hydration. Phys. Chem. Chem. Phys. 2009,
11, 5538−5549.
(35) Prasad, S.; Zhu, H.; Kurian, A.; Badge, I.; Dhinojwala, A.
Interfacial segregation in polymer blends driven by acid-base
interactions. Langmuir 2013, 29, 15727−15731.
(36) Sovago, M.; Campen, R. K.; Wurpel, G. W. H.; Müller, M.;
Bakker, H. J.; Bonn, M. Vibrational response of hydrogen-bonded
interfacial water is dominated by intramolecular coupling. Phys. Rev.
Lett. 2008, 100, 173901.
(37) Gan, W.; Wu, D.; Zhang, Z.; Feng, R.-R.; Wang, H.-F.
Polarization and experimental configuration analyses of sum frequency
generation vibrational spectra, structure, and orientational motion of
the air/water interface. J. Chem. Phys. 2006, 124, 114705.
(38) Nihonyanagi, S.; Yamaguchi, S.; Tahara, T. Water hydrogen
bond structure near highly charged interfaces is not like ice. J. Am.
Chem. Soc. 2010, 132, 6867−6869.
(39) Skinner, J. L.; Pieniazek, P. A.; Gruenbaum, S. M. Vibrational
spectroscopy of water at interfaces. Acc. Chem. Res. 2012, 45, 93−100.
(40) Dewan, S.; Carnevale, V.; Bankura, A.; Eftekhari-Bafrooei, A.;
Fiorin, G.; Klein, M. L.; Borguet, E. Structure of water at charged
interfaces: A molecular dynamics study. Langmuir 2014, 30, 8056−
8065.
(41) Gaigeot, M.-P.; Sprik, M.; Sulpizi, M. Oxide/water interfaces:
how the surface chemistry modifies interfacial water properties. J.
Phys.: Condens. Matter 2012, 24, 124106.
(42) Bottger, G. L.; Geddes, A. L. The infrared spectra of the
crystalline tetramethylammonium halides. Spectrochim. Acta 1965, 21,
1701−1708.
(43) Buch, V.; Tarbuck, T.; Richmond, G. L.; Groenzin, H.; Li, I.;
Shultz, M. J. Sum frequency generation surface spectra of ice, water,
and acid solution investigated by an exciton model. J. Chem. Phys.
2007, 127, 204710.
(44) Groenzin, H.; Li, I.; Buch, V.; Shultz, M. J. The single-crystal,
basal face of ice I(h) investigated with sum frequency generation. J.
Chem. Phys. 2007, 127, 214502.
(45) Barnett, I. L.; Groenzin, H.; Shultz, M. J. Hydrogen bonding in
the hexagonal ice surface. J. Phys. Chem. A 2011, 115, 6039−6045.
(46) Ishiyama, T.; Takahashi, H.; Morita, A. Origin of vibrational
spectroscopic response at ice surface. J. Phys. Chem. Lett. 2012, 3,
3001−3006.
(47) Ishiyama, T.; Morita, A. A direct evidence of vibrationally
delocalized response at ice surface. J. Chem. Phys. 2014, 141, 18C503.
(48) Pauling, L. The structure and entropy of ice and of other crystals
with some randomness of atomic arrangement. J. Am. Chem. Soc. 1935,
57, 2680−2684.
(49) Bernal, J. D.; Fowler, R. H. A theory of water and ionic solution,
with particular reference to hydrogen and hydroxyl Ions. J. Chem. Phys.
1933, 1, 515−548.
(50) Simonelli, D.; Shultz, M. J. Sum frequency generation
orientation analysis of molecular ammonia on the surface of
concentrated solutions. J. Chem. Phys. 2000, 112, 6804−6816.
(51) Rojas, O. J. Encyclopedia of Surface and Colloid Science; Marcel
Dekker, Inc, 2002; pp 517−535.
(52) Claesson, P.; Herder, P.; Stenius, P.; Eriksson, J.; Pashley, R. An
ESCA and AES study of ion-exchange on the basal plane of mica. J.
Colloid Interface Sci. 1986, 109, 31−39.
(53) Herder, P. C.; Claesson, P. M.; Herder, C. E. Adsorption of
cationic surfactants on muscovite mica as quantified by means of
ESCA. J. Colloid Interface Sci. 1987, 119, 155−167.
(54) Scales, P. J.; Grieser, F.; Healy, T. W. Electrokinetics of the
muscovite mica−aqueous solution interface. Langmuir 1990, 6, 582−
589.
(55) Hu, J.; Xiao, X. D.; Ogletree, D. F.; Salmeron, M. Imaging the
condensation and evaporation of molecularly thin films of water with
nanometer resolution. Science 1995, 268, 267−269.
(56) Hu, J.; Xiao, X.-d.; Ogletree, D.; Salmeron, M. The structure of
molecularly thin films of water on mica in humid environments. Surf.
Sci. 1995, 344, 221−236.
(57) Odelius, M.; Bernasconi, M.; Parrinello, M. Two dimensional ice
adsorbed on mica surface. Phys. Rev. Lett. 1997, 78, 2855−2858.
(58) Miranda, P. B.; Xu, L.; Shen, Y. R.; Salmeron, M. Icelike water
monolayer adsorbed on mica at room temperature. Phys. Rev. Lett.
1998, 81, 5876−5879.
(59) Yang, Z.; Bertram, A. K.; Chou, K. C. Why do sulfuric acid
coatings influence the ice nucleation properties of mineral dust
particles in the atmosphere? J. Phys. Chem. Lett. 2011, 2, 1232−1236.
(60) Atkinson, J. D.; Murray, B. J.; Woodhouse, M. T.; Whale, T. F.;
Baustian, K. J.; Carslaw, K. S.; Dobbie, S.; O’Sullivan, D.; Malkin, T. L.
The importance of feldspar for ice nucleation by mineral dust in
mixed-phase clouds. Nature 2013, 498, 355−358.
(61) Campbell, J. M.; Meldrum, F. C.; Christenson, H. K. Is ice
nucleation from supercooled water insensitive to surface roughness? J.
Phys. Chem. C 2015, 119, 1164−1169.
(62) Steinke, I. Ice nucleation properties of mineral dusts. Ph.D. thesis,
Ruperto-Carola University of Heidelberg, Germany, p 213.
3748
DOI: 10.1021/acs.jpcc.5b08371
J. Phys. Chem. C 2016, 120, 3741−3748