Crystallization of thin water films on graphite: Effects of n-hexane, formaldehyde, acetone, and methanol additives

Applied Surface Science 357 (2015) 1809–1815

Contents lists available at ScienceDirect

Applied Surface Science journal homepage: www.elsevier.com/locate/apsusc

Crystallization of thin water films on graphite: Effects of n-hexane, formaldehyde, acetone, and methanol additives Ryutaro Souda International Center for Materials Nanoarchitectonics, National Institute for Materials Science, 1-1 Namiki, Tsukuba, Ibaraki 305-0044, Japan

a r t i c l e

i n f o

Article history: Received 22 July 2015 Received in revised form 1 October 2015 Accepted 5 October 2015 Available online 26 October 2015 Keywords: Secondary ion mass spectrometry Temperature programmed desorption Amorphous solid water Crystallization

a b s t r a c t Interactions of molecular additives with amorphous solid water have been investigated using time-offlight secondary ion mass spectrometry and temperature programmed desorption. The crystallization temperature of water on a clean graphite substrate decreases from the bulk value of 160 K to 150 K when water deposition temperature increases from 20 K to 100 K. This phenomenon is induced by the formation of a specifically oriented water layer at the interface, as evidenced by that a submonolayer of n-hexane adspecies on graphite quenches this behavior. Thermal desorption spectra of additives reflect their hydration forms. The n-hexane molecules are trapped in the interior of a porous water film via hydrophobic hydration and released explosively during crystallization. The thermal desorption spectra of methanol resemble those of water from multilayer films because methanol can enter the hydrogenbond network of water via hydrophilic hydration. The hydration of formaldehyde is hydrophobic in nature despite the presence of the polar carbonyl group. Features of both hydrophilic and hydrophobic hydrations are identifiable in acetone–water interactions; the branching ratio depends on the water preparation method and substrate. © 2015 Elsevier B.V. All rights reserved.

1. Introduction Water in nanoconfined geometry is ubiquitous in nature, and understanding of its properties is important in many research fields. Unique physical and chemical properties of water in the bulk originate from a highly polar hydrogen bond, and they are modified significantly upon nanoconfinement because of the interplay between the intermolecular hydrogen bond and the water-substrate bond [1–10]. On the other hand, additives in water induce structural transformation in nanoscale regions via the formation of solvation shells [11–15]. The solvation force competes with forces of surface segregation and adhesion to the substrate in the nanoconfined geometry. Thin films of amorphous solid water (ASW) can be formed by deposition of water molecules onto cold substrates. The structure of ASW deposited at temperatures well below the glass-transition temperature, Tg , is characterized by a microporous structure, as evidenced by the ability to incorporate a large amount of adspecies in thin film interior [16,17]. The occurrence of liquidlike mobility in nanoconfined ASW films has been investigated using time-of-flight secondary ion mass spectrometry (TOF-SIMS) as a function of temperature [18–21]. The molecules become mobile at around bulk Tg

E-mail address: [email protected] http://dx.doi.org/10.1016/j.apsusc.2015.10.024 0169-4332/© 2015 Elsevier B.V. All rights reserved.

of water (136 K) when thin films are deposited onto a Ni(1 1 1) substrate [18]. However, a water monolayer tends to form droplets in the sub-Tg region (∼110–120 K) on hydrophobic substrates [19,20]. The dewetting temperature of ASW depends on film thickness on highly oriented pyrolytic graphite (HOPG) although it is fixed at around 160 K on Ni(1 1 1) [21]. The crystallization kinetics of thin ASW films has been discussed based on isothermal desorption [22–25], temperatureprogrammed desorption (TPD) [25–31], and reflection absorption infrared spectroscopy (RAIRS) [32–35]. The crystallization process is also substrate dependent at specific film thickness [22,23], although the contrary is also suggested that crystallization is initiated from the free surface [33]. The crystallization temperature, Tc , of ASW is around 160 K on hydrophilic metal substrates [26]. However, the nucleation kinetics can be controlled by increasing the deposition temperature of water molecules on CCl4 multilayer films because mobility of the adsorbing water molecules is high on hydrophobic surfaces [31], suggesting that the morphology of initial ASW films is related to the crystallization kinetics. The desorption kinetics of water also depends on hydrophilicity and hydrophobicity of substrates [30,36,37]. The water TPD spectra from a hydrophobic graphite substrate are complicated in coverage dependence [28,37]. It is proposed that water on monolayer graphite form a new ice polymorph that consists of two flat hexagonal sheets of water molecules [10]. It is also suggested that

R. Souda / Applied Surface Science 357 (2015) 1809–1815

orientation of the water molecules in the first monolayer on HOPG can be changed with increasing temperature [37]. Consequently, nanoconfinement effects appear more evidently on hydrophobic substrates because of weak substrate–water interactions. In this paper, interactions of simple molecular additives with water and their effects on the phase transition of thin ASW films on the HOPG substrate are investigated. The crystallization kinetics is discussed based on TPD spectra of water and molecular additives from ASW films [31], as well as the occurrence of film dewetting using TOF-SIMS [35]. Regarding additives, nonpolar molecules are believed to have no effects on water crystallization kinetics although methanol changes the properties of ASW films significantly [38,39]. However, no systematic studies have been performed concerning the effects of other molecular additives on nanoconfined water. To this end, interactions of water with nonpolar aprotic (n-hexane), polar aprotic (formaldehyde, acetone), and polar protic (methanol) molecules are explored to gain insights not only into the phase transition of nanoconfined water but also into hydration of these additives.

25000 20000

(a)

15000 18 amu (TPD)

10000

Intensity (arb. units)

1810

-

OH C2

5000

-

0 25000 20000

(b)

15000 10000

2. Experiment Experiments were performed in an ultrahigh vacuum (UHV) chamber with a base pressure of <1 × 10−10 Torr. The HOPG substrate was purchased from Panasonic Inc. (PGC grade). It was attached to a Ta holder using Ta strips after cleavage by the Scotch Tape method. The sample holder was mounted on the end of a Cu cold finger extended from a closed-cycle helium refrigerator. Temperature was controlled using a digital temperature controller by monitoring temperature of the cold finger close to the sample position using Au(Fe)-chromel thermocouples. For cleaning, the sample was heated several times in UHV to approx. 1300 K by electron bombardment from behind. Sample cleanliness was confirmed by the absence of any impurity peaks in TOF-SIMS. Water, n-hexane (96%), acetone (99.5%), and methanol (99.8%) were degassed by several freeze-pump-thaw cycles. Gaseous formaldehyde was obtained by heating a solid paraformaldehyde sample that was outgassed thoroughly before use. Thin films were deposited onto a cold HOPG substrate by backfilling the chamber with the gases admitted through high precision leak valves. TPD spectra were recorded using a quadrupole mass spectrometer (HIDEN, IDP 300S) placed in a differentially pumped housing; a retractable orifice at the end of the housing was placed ∼3 mm apart from the sample surface. For TOF-SIMS measurements, a primary beam of 2 keV He+ ions was generated in an electron-impacttype ion gun (Specs, IQE 12/38) and was chopped into pulses using electrostatic deflectors. Negatively charged secondary ions ejected perpendicularly to the sample surface were detected using a microchannel plate after passing through a field-free TOF tube. To extract low-energy secondary ions efficiently, a bias voltage (−500 V) was applied to the sample. The temperature was ramped at a rate of 5 K min−1 for both TPD and TOF-SIMS measurements. The fluence of He+ in TOF-SIMS measurements was restricted below 1 × 1012 ions cm−2 to minimize surface decomposition. 3. Experimental results 3.1. n-Hexane The morphology of initially flat ASW films changes during the glass-liquid transition and crystallization. TOF-SIMS reveals the droplet formation straightforwardly by monitoring ion yields from the substrate as a function of temperature [18]. In the present study, TOF-SIMS spectra of negative ions are focused because no

5000 0 100

120

140

160

180

200

T (K) Fig. 1. Temperature-programmed TOF-SIMS intensities of anions (OH− and C2 − ) and TPD spectra of water (18 amu) from ASW films formed by exposure of 20 L H2 O to a clean HOPG substrate at (a) 20 K and (b) 100 K.

positive ions are sputtered from the clean HOPG surface [21]. Temperature-programmed TOF-SIMS intensities of OH− and C2 − ions are plotted in Fig. 1, together with TPD spectra of water (18 amu). The ASW films were formed by deposition of 20 L (Langmuir; 1 L = 1 × 10−6 Torr s) of H2 O onto the clean HOPG substrate at (a) 20 K and (b) 100 K. It has been determined that 1 monolayer of water is attained at exposure of ca. 2.5 L H2 O for HOPG based on the saturation point of ion evolution curves [21]. In general, the monolayer thus determined by TOF-SIMS differs from that by TPD because the former (latter) corresponds to a crowded (scattered) monolayer interacting mainly with adspecies (substrate). In this study, effects of adspecies at typical exposure of 1 L are investigated. The coverage of adspecies is estimated based on the TPD spectra. In Fig. 1, the film morphology change is identified as an increase of the C2 − intensity. This occurs at ca. 160 and 150 K, respectively, for the ASW films deposited at 20 and 100 K. A characteristic peak is also observed in the OH− intensity during film dewetting. The TPD spectrum of water in Fig. 1(a) exhibits a bump at 158 K. The decrease in the water TPD signal at this temperature is due to crystallization of the film: Crystalline ice is more stable and thus has a lower vapor pressure than ASW, leading to the decreased signal. The crystallization and dewetting are coupled because nucleation occurs in a liquidlike phase [18]. Thus, when the film crystallizes it also dewets. However, the bump in Fig. 1(b) is not clearly recognizable. Probably, this is because the desorption yield of water is small during crystallization at around 150 K. Consequently, the results in Fig. 1 indicate that Tc is controlled by the deposition temperature of the water molecules on HOPG. No such a behavior is observed using metal substrates like Ni(1 1 1) [21]. Fig. 2 displays logarithmic plots of TPD spectra of n-hexane adsorbed on HOPG and two differently tailored ASW films. Here, a porous ASW film (20 L) that is deposited directly onto HOPG at

R. Souda / Applied Surface Science 357 (2015) 1809–1815

6

10

5

(50L@20K) (200L@20K)/(50L@100K) (100L@20K)/(50L@100K) (50L@20K)/(50L@100K) (20L@20K)/(50L@100K) (50L@100K)

ASW II: H2O(20L@20K) / H2O(50L@100K) 18 57 ASW I: H2O(20L@20K) 18 57 HOPG 57

4

10

Intensity

Intensity (arb. units)

10

1811

3

10

2

10

1

10 100

120

140

160

180

200

T (K) 100

120

140

160

180

Fig. 2. TPD spectra of water (18 amu) and n-hexane (57 amu) desorbed from differently tailored ASW films and the HOPG substrate; n-hexane (1 L) was adsorbed on them at 20 K. The water films (20 L at 20 K) deposited onto the clean HOPG substrate and preexisting ASW films (50 L at 100 K) on HOPG are referred to as ASW I and II, respectively.

Fig. 3. TPD spectra of n-hexane (1 L at 20 K) adsorbed on the surfaces of differently tailored ASW films formed on the clean HOPG substrate.

20 K is referred to as ASW I. For comparison, a basal ASW film (50 L) is formed initially on HOPG at 100 K, and then the porous ASW layer (20 L) is deposited on it at 20 K. This film is referred to as ASW II. The n-hexane molecules (1 L) are adsorbed on the surface of these films at 20 K. From the HOPG substrate, desorbed n-hexane peaks in intensity at around 180 K. All n-hexane adspecies are in direct contact with HOPG, as evidenced by the absence of an n-hexane multilayer peak that is expected to occur at 130–140 K [38]. The nhexane forms a main TPD peak at around 160 K when deposited on ASW I, whereas a relatively broad peak occurs at around 150 K from ASW II. The n-hexane molecules desorb after being incorporated in the interior of both ASW films because of porosity. No clear bump is observable in the water TPD spectrum from ASW II. This is expected to occur when the film morphology changes gradually, as inferred from broadness of the n-hexane TPD peak. The result resembles that shown in Fig. 1, indicating that desorption of n-hexane additives is correlated to dewetting of the ASW film during crystallization. The crystallization kinetics of ASW II appears to be determined by interactions of the basal ASW layer with the HOPG substrate rather than the nature of the porous ASW moiety. The thickness dependence of the porous ASW moieties on crystallization kinetics is investigated in the ASW II configuration. Here we focus on the TPD peak of n-hexane rather than water TPD because the bump in the latter is rather vague in most cases for identification of the phase transition. The experimental results are displayed in Fig. 3. From the basal ASW film (exposure of 50 L H2 O at 100 K), n-hexane forms a peak at 130–135 K. The multilayer nhexane peak occurs at this temperature [38], suggesting that the molecules come from the ASW film surface. In reality, however, a considerably large n-hexane peak is identifiable at 175 K, which is ascribed to the molecules in direct contact with HOPG or those incorporated in crystalline water. Therefore, a part of the n-hexane adspecies should be incorporated in the interior of the basal ASW

film to reach the HOPG substrate. The main TPD peak of n-hexane occurs at 150 K from the porous ASW overlayer (20 L at 20 K) formed on the basal ASW film. The peak becomes broad when the overlayer has comparable thickness (50 L at 20 K) with the basal ASW film, and then the peak is shifted to higher temperature gradually with further increasing thickness of the overlayer. The peak temperature finally agrees with that of the ASW I film (50 L at 20 K), but a considerably thick overlayer (∼200 L at 20 K) is required to extinguish the effects of the basal ASW layer (50 L at 100 K) or the HOPG substrate. The intensity of the high-temperature n-hexane peak (170–180 K) from ASW II is lower than that from ASW I, and the peak shifts to higher temperature with increasing film thickness. The latter is correlated to the shift of high-temperature cutoffs of the water TPD peaks (see e.g., Fig. 2), indicating that a small amount of n-hexane molecules that survived the phase transition can be trapped in crystalline water. The high-temperature peak is also observed in TPD of other hydrophobic species like CCl4 [31]. Fig. 4 shows TOF-SIMS and TPD experimental results for the nhexane (1 L at 20 K) adsorbed HOPG substrate which is capped with the ASW film (20 L at 100 K). In this case, not only the water and n-hexane TPD spectra but also the TOF-SIMS C2 − intensity exhibit the phase transition at around 160 K. The fundamentally same TPD spectra are obtained when the capping ASW film is formed at 20 K and 120 K onto the n-hexane adsorbed HOPG substrate (not shown). The TPD spectra in Fig. 4 resemble those of the ASW I film on which n-hexane is adsorbed at 20 K (Fig. 2). Consequently, the deposition temperature of water has no effects on the phase transition of thin films when a submonolayer of the n-hexane adspecies is present on the HOPG substrate. This result clearly shows that the decrease in Tc with increasing water deposition temperature is not simply ascribable to hydrophobicity of HOPG because n-hexane is also hydrophobic in nature.

T (K)

R. Souda / Applied Surface Science 357 (2015) 1809–1815

10

5

10

4

10

3

SIMS -

OH C2

10

4000

TPD 18 amu 57 amu

Intensity (arb. units)

Intensity (arb. units)

1812

-

CH3COCH3 / HOPG CH3COCH3 / H2O (20L @20K) CH3COCH3 / H2O (20L @20K) / H2O (50L @100K) H2O (20L @20K) / CH3COCH3 / H2O (50L @100K)

3000

2000

2

1000

1

10 100

120

140

160

180

200

T (K) Fig. 4. Temperature-programmed TOF-SIMS anion intensities and TPD spectra obtained from the ASW film (20 L) deposited onto the n-hexane (1 L at 20 K) preadsorbed HOPG substrate at 100 K.

The interactions of formaldehyde with water and the HOPG substrate, as well as the effect of the phase transition of ASW, are investigated based on the formaldehyde TPD spectra. Fig. 5 displays experimental results for 1 L H2 CO deposited onto the HOPG substrate and differently tailored ASW films at 20 K. A doublet peak

H2CO / HOPG H2CO / H2O(50L @100K) H2CO / H2O(20L @20K) H2CO / H2O(20L @20K) / H2O(50L @100K) H2O(20L @20K) / H2CO / H2O(50L @100K)

Intensity (arb. units)

6000

4000

2000

0 80

100

120

140

160

180

200

T (K) Fig. 6. TPD spectra (43 amu) of acetone additives (1 L at 20 K) from differently tailored ASW films and the HOPG substrate.

3.2. Polar molecules

8000

0

100

120

140

160

180

T (K) Fig. 5. TPD spectra (30 amu) of formaldehyde additives (1 L at 20 K) from differently tailored ASW films and the HOPG substrate.

is observed from the HOPG substrate; they are ascribable to monolayer (103 K) and multilayer (96 K) of the H2 CO adspecies. On the basal ASW film (50 L at 100 K), the monolayer peak is broadened considerably with a leading edge at 123 K. The multilayer peak of H2 CO disappears from the ASW I film (20 L H2 O at 20 K); the molecules are trapped in the film interior and form a sharp peak at 160 K in addition to those desorbed from the film surface below 123 K. From the ASW II film, the trapped molecules desorb at 152 K. When the H2 CO molecules are present between two water layers in the ASW II configuration, a larger number of molecules are trapped, as evidenced by increase in intensity of the peak at 154 K. The H2 CO molecules desorbed from the crystalline water (175 K) and the ASW film surface (116 K) are also identified. The experimental results using H2 CO additives on ASW I and II are fundamentally same as those using n-hexane additives in terms of the trapped molecules forming peaks at ca. 150 and 160 K. In contrast to nhexane, the H2 CO molecules tend to be bound more tightly to the ASW film surface than the HOPG surface, suggesting the contribution of hydrogen bonds. Fig. 6 shows TPD spectra of acetone (1 L at 20 K) deposited onto the HOPG substrate and the ASW films. The acetone molecules desorb from HOPG at 145 K, forming a monolayer peak. From the ASW I film (20 L at 20 K), acetone trapped in the film interior forms a peak at 160 K, in addition to conspicuous tails to lower and higher temperature sides. The acetone TPD peak is expected to occur at 150 K from the ASW II film (20 L at 20 K/50 L at 100 K), but it appears to be depressed. This behavior is confirmed more clearly when acetone is placed between two water layers in the ASW II configuration, where a steep onset of a broad component occurs at 155 K instead of the characteristic peak at 150 K. The leading edge of the acetone TPD shifts to higher temperatures with increasing ASW film thickness along with the shift of the high-temperature cutoff of the water TPD spectra (not shown). These behaviors imply that acetone is bound tightly to water via hydrogen bonds. On the other hand, the acetone TPD peak is observed rather intensively at 160 K when

Intensity (arb. units)

R. Souda / Applied Surface Science 357 (2015) 1809–1815

10

5

10

4

10

3

10

2

10

1

10

5

10

4

10

3

10

2

10

1

(a) CH 3COCH3

18 43

(b) CH 3OH

18 31

140

160

180

T (K) Fig. 7. TPD spectra of water (18 amu), acetone (43 amu), and methanol (31 amu) from ASW films (20 L at 20 K) deposited onto (a) acetone (1 L at 20 K) and (b) methanol (1 L at 20 K) precovered HOPG substrates.

the acetone (1 L) adsorbed HOPG substrate is capped with the ASW film (20 L at 20 K) as shown in Fig. 7(a), where the water TPD spectrum exhibits the bump clearly. Consequently, the 160 K peak of adspecies fundamentally occurs for the n-hexane, formaldehyde, and acetone adspecies interacting with ASW I, although the 150 K peak of acetone from ASW II is missing for some reasons. As shown in Fig. 7(b), when methanol (1 L) is used as adspecies on HOPG instead of acetone, no TPD peak of methanol occurs at 160 K and the bump of water TPD shifts to lower temperature. The methanol forms a monolayer peak at ca. 145 K when adsorbed on the HOPG substrate (not shown). Therefore, the methanol TPD spectrum is broadened considerably because of the interaction with water. Note that its shape resembles that of the multilayer water spectrum, indicating that methanol mixes with water and is liberated via breakage of hydrogen bonds with surrounding water molecules. Such a broad component is also evident in the acetone TPD spectrum in Fig. 7(a) although its intensity is small relative to the 160 K peak. 4. Discussion The crystallization temperature of thin ASW films decreases by increasing deposition temperature of water on HOPG. The interfacial structure of water plays a role because the submonolayer of n-hexane adspecies on HOPG quenches this phenomenon. The n-hexane adspecies on the free surface of ASW has no appreciable effects on the crystallization kinetics of water. The interaction between n-hexane and water is governed by the microscopic structure of the ASW film which is determined by the water deposition temperature. The porous ASW film is formed at 20 K, and adspecies trapped in the film interior desorb during water crystallization. The hydrogen bond network is incomplete in porous ASW, so that the n-hexane species can be trapped in the film interior during the heating process. This phenomenon is associated with

1813

“hydrophobic hydration” [38]. In this respect, crystalline hydrates are known to be formed during deposition of water and guest molecules at cryogenic temperatures [40]. In the framework of polyamorphism [41], it is established that low density amorphous solid (LDA), which is identical to ASW, is a distinct phase of normal liquid water. The hydrogen bond structure of ASW (LDA) resembles that of crystalline water [42]. Therefore, additives trapped in porous ASW are thought to be caged by water molecules upon heating, leading to hydrophobic hydration of nonpolar solute species. The hydrated species are released at 160 K because the cage collapses during the water phase transition. In contrast, the hydrogen-bond network of ASW is completed when water molecules are deposited at 100 K, so that n-hexane is not hydrated in the basal ASW film (see Fig. 3). However, the n-hexane adspecies can penetrate the film interior, as manifested by the presence of residues at higher temperature. The adspecies might diffuse through “cracks” of ASW without being hydrated. On the other hand, the crystallization of water has been discussed based on a “molecular volcano” picture [26,31] or explosive desorption of adspecies embedded underneath the ASW film. May et al. [31] demonstrated that crystallization kinetics and morphology of the initial ASW films are controllable by the water deposition temperature and film thickness on the CCl4 multilayer film deposited onto graphene. With increasing deposition temperature, mobility of the adsorbing water molecules increases on the hydrophobic CCl4 film surface, so that crystallization (or the film morphology change) occurs at lower temperature. In contrast, the present study reveals that the Tc reduction to 150 K is rather quenched by the presence of the n-hexane adspecies at the interface of HOPG. Moreover, the initial ASW films deposited at 100 K do not dewet the HOPG substrate, as revealed from evolutions of TOFSIMS intensities as functions of temperature (Fig. 1) and exposure [21]. Apparently, the situation is not so simple for crystallization of ASW on HOPG compared to that on the multilayer CCl4 films. In previous papers [21,37], we have investigated the interaction of ASW films with HOPG by using TOF-SIMS. It was revealed that the dewetting temperature of ASW deposited at 20 K depends on the film thickness. The morphology of the water monolayer changes at 145 K whereas thicker ASW films dewet at 160 K [21]. This phenomenon is induced by reorientation of the water monolayer at the interface with increasing temperature [37]. The OH group of water tends to be directed toward the HOPG substrate to reduce the number of non-bonding OH groups at the interface. This is thermodynamically most stable form of the water monolayer on graphite [43,44]. The reorientation of water at the interface is quenched when thicker ASW films are deposited at 20 K, thereby increasing Tc from 145 K to the bulk value of 160 K with increasing film thickness [21,37]. Such a specifically oriented layer can be created at the interface of thicker ASW films when the water molecules are deposited at 100 K. This is confirmed from the initial evolution curves of the TOF-SIMS H+ yield relative to the H3 O+ yield [21]. From these experimental facts, it is concluded that crystallization initiated in the bulk (160 K) is switched to that from the substrate interface (150 K) when the reoriented water layer is formed at the interface with HOPG. The n-hexane adspecies on HOPG bothers the formation of such a layer, thereby resulting in the occurrence of crystallization in the bulk irrespective of the water deposition temperature on the n-hexane adsorbed HOPG substrate. The crystallization kinetics is influenced by the interfacial interactions even for a considerably thick ASW film, as revealed by a gradual shift of the main n-hexane TPD peaks in Fig. 3. The crystallization kinetics of ASW is also influenced by methanol adspecies, as manifested by a shift of the bump in water TPD spectrum to the low temperature side (see Fig. 7(b)). The effect of methanol on water crystallization has been investigated based on RAIRS measurements [39]. It was revealed that Tc of ASW on the

1814

R. Souda / Applied Surface Science 357 (2015) 1809–1815

hydrophilic Ni(1 1 1) substrate is depressed to 150 K by methanol additives. This phenomenon is ascribable to the free surface effect on crystallization kinetics of ASW as described below. The methanol adspecies tend to stay at the free surface of ASW as a surfactant, as evidenced by that the water droplet formation is suppressed by surface segregated methanol adspecies [18]. This phenomenon and Tc depression are correlated to each other because they are thought to be associated with reduction of surface tension by methanol. The role of the surfactant methanol at the free surface resembles the effect of the reoriented water layer formed at the interface with the HOPG substrate because Tc depression is associated with the reduction in number of the free OH groups at both interfaces. As regards formaldehyde additives, no specific features are observed in comparison with hydrophobic hydration of n-hexane, indicating that the polar carbonyl group does not influence the water cage formation. In contrast, no indications of hydrophobic hydration are identifiable in interactions between methanol and water. The broadened TPD spectrum of methanol resembling that of water multilayer TPD spectrum is indicative of “hydrophilic hydration” or mixing of molecules via the formation of hydrogen bonds. The methanol can enter the hydrogen-bond network of water because it has the abilities of both donor and accepter of hydrogen bonds similar to water. In contrast, the hydrogen-bond acceptors like formaldehyde tend to be trapped in the cage although formaldehyde can form hydrogen bonds with free OH groups on the interior surface of the cage. The acetone–water interactions exhibit features of both hydrophilic and hydrophobic hydrations, as manifested by coexistence of the peak at 160 K and the broad component in TPD spectra. The branching ratio of molecules into these two states appears to be dependent on the ways of the ASW deposition and the acetone supply. In any case, the role of hydrophobic hydration of acetone is suppressed considerably in comparison with formaldehyde. In this respect, it should be noticed that the 150 K peak of acetone disappears in the ASW II configuration. The acetone adsorbed on the basal ASW film can form hydrogen bonds with water, so that the hydrophobic hydration by the porous ASW moiety can be suppressed. The same appears to occur when acetone interacts initially with the porous ASW moiety. The reason for this behavior is not very clear, but uptake of acetone in the interior of the porous ASW moiety of ASW II is apparently suppressed, as manifested by that the TPD peak at 145 K is enhanced considerably relative to that from the ASW I film formed directly on HOPG. This result strongly suggests that the initial structure and reorganization of water hydrogen bonds are different in the porous ASW moiety between the ASW I and II configurations. The 160 K peak of acetone is evident when the molecules are present at the interface with HOPG (Fig. 7(a)). It might be presumed that these acetone species are released directly from the substrate interface during the droplet formation without interactions with water (i.e., the molecular-volcano mechanism). However, the confinement of adspecies at the interface without mixing with water is apparently too simple in the framework of the present study. For example, mixing of water and methanol is known to occur at T > Tg = 136 K [18]. Even for nonpolar species, a submonolayer of n-hexane on HOPG tends to detach from the interface to be hydrated during the water deposition or the ASW film heating, as evidenced by that a relatively intense 160 K peak occurs in Fig. 4 despite that the n-hexane molecules in direct contact with HOPG should form a higher temperature TPD peak at 180 K. The attractive force of acetone on HOPG is weaker than n-hexane, as inferred from the lower acetone TPD peak temperature of 145 K. In any case, the acetone adspecies on HOPG is expected to be hydrated in some form. Consequently, the branching ratio of the hydrophilic and hydrophobic hydrations of acetone depends not only on the ASW film preparation methods but also on properties of the substrate.

5. Conclusion The interactions of polar and nonpolar adspecies with nanoconfined water, as well as their effects on the crystallization kinetics of water, are investigated on the HOPG substrate. The crystallization temperature of ASW decreases from bulk Tc of 160 K to 150 K when the water deposition temperature increases from 20 K to 100 K. This behavior is associated with the formation of a specifically oriented water layer at the interface to reduce the number of free OH groups. Because of such a layer, water crystallization is thought to be initiated at the interface, resulting in the reduction of Tc to 150 K. This is manifested by the fact that Tc becomes independent of the water deposition temperature when n-hexane adspecies exists on the HOPG surface. On the other hand, the methanol adspecies reduces Tc to 150 K as well because the crystallization starts at the free surface: the number of free OH groups at the ASW film surface is also reduced by the methanol adspecies because of the surfactant effect. The n-hexane adspecies are trapped in the ASW film interior because hydrogen bonds of the water molecules can be reorganized to form hydration cages upon heating. The trapped additives are released during the phase transition of water because of the cage collapse, thereby forming a TPD peak at 160 K. The methanol–ASW interactions are characterized by hydrophilic hydration or the direct hydrogen bond formation between them, as evidenced by the broadened TPD spectrum of methanol resembling that of multilayer water. The polar carbonyl group is not active enough to realize hydrophilic hydration of formaldehyde although the contribution of hydrogen bonds between formaldehyde adspecies and water is recognizable. In contrast, features of both hydrophobic and hydrophilic hydrations are clearly identified in the acetone–water interaction. The branching ratio of these two hydration forms depends on the ASW film preparation method and substrate. In any case, not only polar groups but also nonpolar ligands are thought to play a role in hydration of polar aprotic molecules at cryogenic temperatures. References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29]

A. Verdaguer, G.M. Sacha, M. Salmeron, Chem. Rev. 106 (2006) 1478. U. Raviv, P. Laurat, J. Klein, Nature 413 (2001) 51. A. Hodgson, S. Haq, Surf. Sci. Rep. 64 (2009) 381. R.J. Mashl, S. Joseph, N.R. Aluru, E. Jakobsson, Nano Lett. 3 (2003) 589. M.C. Gordillo, J. Marti, Chem. Phys. Lett. 329 (2000) 341. J.C. Rasaiah, C. Jayendran, S. Garde, G. Hummer, Annu. Rev. Phys. Chem. 59 (2008) 713. O. Byl, J.-C. Liu, Y. Wang, W.-L. Yim, J.K. Johnson, J.T. Yate Jr., J. Am. Chem. Soc. 128 (2006) 12090. H.-J. Wang, X.-K. Xi, A. Kleinhammes, Y. Wu, Science 322 (2008) 80. N. Giovambattista, P.J. Rossky, P.G. Debenedetti, Phys. Rev. Lett. 102 (2009) 050603. G.A. Kimmel, J. Matthiesen, M. Baer, C.J. Mundy, N.G. Petrik, R.S. Smith, Z. Dohnalek, B.D. Kay, J. Am. Chem. Soc. 131 (2009) 12838. J. Tomasi, B. Mennucci, R. Cammi, Chem. Rev. 105 (2005) 2999. A.V. Marenich, C.J. Cramer, D.G. Truhlar, J. Phys. Chem. B 113 (2009) 6378. N.T. Southall, K.A. Dill, A.D.J. Haymet, J. Phys. Chem. B 106 (2002) 521. B. Hribar, N.T. Southall, V. Vlachy, K.A. Dill, J. Am. Chem. Soc. 124 (2002) 12302. M.F. Kropman, H.J. Bakker, Science 291 (2001) 2118. E. Mayer, R. Pletzer, Nature 319 (1986) 298. K.P. Stevenson, G.A. Kimmel, Z. Dohnalek, R.S. Smith, B.D. Kay, Science 283 (1999) 1505. R. Souda, Phys. Rev. Lett. 93 (2004) 235502. R. Souda, J. Phys. Chem. C 115 (2011) 8136. R. Souda, J. Chem. Phys. 129 (2008) 124707. R. Souda, J. Phys. Chem. B 110 (2006) 17524. P. Lofgren, P. Ahlstrom, D.V. Chakarov, J. Lausmaa, B.J. Kasemo, Surf. Sci. 367 (1996) L19. P. Lofgren, P. Ahlstrom, J. Lausma, B. Kasemo, D. Chakarov, Langmuir 19 (2003) 265. R.S. Smith, C. Huang, E.K.L. Wong, B.D. Kay, Surf. Sci. 367 (1996) L13. S.M. McClure, E.T. Barlow, M.C. Akin, D.J. Safarik, T.M. Truskett, C.B. Mullins, J. Phys. Chem. B 110 (2006) 17987. R.S. Smith, C. Huang, E.K. Wong, B.D. Kay, Phys. Rev. Lett. 79 (1997) 909. R.S. Smith, C. Huang, B.D. Kay, J. Phys. Chem. B 101 (1997) 6123. A.S. Bolina, A.J. Wolff, W.A. Brown, J. Phys. Chem. B 109 (2005) 16836. J. Günster, R. Souda, Phys. Rev. B 71 (2005) 041407.

R. Souda / Applied Surface Science 357 (2015) 1809–1815 [30] J. Günster, R. Souda, Langmuir 22 (2006) 6939. [31] R.A. May, R.S. Smith, B.D. Kay, Phys. Chem. Chem. Phys. 13 (2011) 19848. [32] P. Jenniskens, S.F. Banham, D.F. Blake, M.R.S. McCoustra, J. Chem. Phys. 107 (1997) 1232. [33] M. Ostblom, J. Ekeroth, P. Konradsson, B. Liedberg, J. Phys. Chem. B 110 (2006) 1695. [34] E.H.G. Backus, M.L. Grecea, A.W. Kleyn, M. Bonn, Phys. Rev. Lett. 92 (2004) 236101. [35] R. Souda, J. Phys. Chem. B 112 (2008) 11976. [36] L.H. Dubois, B.R. Zagarski, R.G. Nuzzo, J. Am. Chem. Soc. 112 (1990) 570.

[37] [38] [39] [40] [41] [42]

1815

R. Souda, J. Phys. Chem. C 116 (2012) 20895. R. Souda, J. Phys. Chem. B 109 (2005) 21879. R. Souda, Phys. Rev. B 75 (2007) 184116. F. Fleyfel, J.P. Devlin, J. Phys. Chem. 95 (1991) 3811. O. Mishima, H.E. Stanley, Nature 396 (1998) 329. J.L. Finney, A. Hallbrucker, I. Kohl, A.K. Soper, D.T. Bowron, Phys. Rev. Lett. 88 (2002) 225503. [43] G. Cicero, J.C. Grossman, E. Schwegler, F. Gygi, G. Galli, J. Am. Chem. Soc. 130 (2008) 1871. [44] D. Feller, K.D. Jordan, J. Phys. Chem. A 104 (2000) 9971.