1: \documentstyle[preprint,aps]{revtex}
2: \begin{document}
3:
4: \title{Surface melting of methane and methane film on magnesium oxide}
5: \draft
6: \tightenlines
7: \author{N. Klipa}
8:
9: \address{Department of Physics, University of Zagreb, Bijeni{\v c}ka 32, 10000 Zagreb, Croatia}
10:
11: \author{ G. Bilalbegovi\'c }
12:
13: \address{Department of Physics, University of Rijeka, Omladinska 14, 51000 Rijeka, Croatia}
14:
15:
16:
17: \date{to appear in Surface Science}
18:
19: \maketitle
20:
21:
22: \begin{abstract}
23:
24: Experiments on surface melting of several organic materials have
25: shown contradictory results. We study the Van der Waals interactions between interfaces in
26: surface melting of the bulk $CH_{4}$ and
27: interfacial melting of the $CH_{4}$ film on the $MgO$ substrate.
28: This analysis is based on the
29: theory of Dzyaloshinskii, Lifshitz, and Pitaevskii for dispersion forces
30: in materials characterized by the frequency dependent dielectric functions.
31: These functions for magnesium oxide and methane
32: are obtained from optical data using an oscillator
33: model of the dielectric response.
34: The results show that a repulsive interaction between the solid-liquid and
35: liquid-vapor interfaces exists for the bulk methane.
36: We also found that the van der Waals forces between two solid-liquid interfaces
37: are attractive for the $CH_4$ film on the $MgO$ substrate.
38: This implies that the van der Waals forces induce the presence of
39: complete surface melting for the bulk methane and
40: the absence of interfacial melting for $CH_4$ on the MgO substrate.
41:
42:
43:
44: {\it Keywords}: Construction and use of effective interatomic interactions;
45: Equilibrium thermodynamics and statistical mechanics; Dielectric phenomena;
46: Surface melting; Wetting; Magnesium oxides; Alkanes;
47: Solid-liquid interfaces;
48:
49: \end{abstract}
50:
51:
52:
53:
54:
55:
56: \section{Introduction}
57:
58:
59:
60:
61: As the temperature is increased from below the bulk triple point,
62: some solids begin to melt from their surfaces inward. If the thickness of the
63: melted layer increases without limit as the temperature approaches the bulk
64: triple point, then the solid is said to undergo surface melting
65: \cite{Amsterdam,Dash}. The melting is incomplete if the thickness of the
66: melted film remains finite as the temperature increases at all temperatures
67: including the triple point. A special case of incomplete surface melting
68: where the liquid film thickness is zero is sometimes referred to as surface
69: nonmelting. For example, it was found that the $Pb$(110) surface melts.
70: The $Pb$(111) surface does not melt, whereas $Pb$(100) exhibits
71: incomplete melting. Therefore, incomplete or blocked melting appears
72: on $(100)$ fcc metal surfaces with an intermediate packing density
73: between $(110)$ and $(111)$ \cite{Frenken,Goranka}. It is characterized
74: by a quasi-liquid film of the finite and small thickness.
75: Surface melting was studied for several metals and other
76: materials, such as ice, germanium, and multilayers
77: of gases adsorbed on graphite \cite{Amsterdam,Dash}.
78: High-temperature disordering for surfaces
79: of several organic materials was also studied.
80:
81: X-ray reflectivity measurements were used for high-temperature studies of
82: the $(100)$ face of caprolactam ($C_6H_{11}ON$) \cite{Chandavarkar,Wong}.
83: Initially a series of layering transition followed by
84: prewetting was suggested, but later
85: a surface hydration mechanism was proposed for temperature disordering of
86: this surface. Chernov and Yakovlev
87: studied surface melting for the $(001)$ and $(010)$ faces of biphenyl
88: ($C_{12}H_{10}$) using the ellipsometry technique \cite{Chernov}.
89: They found that the thickness of a quasi-liquid layer at $0.5$ K
90: below the bulk melting point reaches $100 \AA$ for the $(001)$ face
91: and $300 \AA $ for the $(010)$ face. In contrast, de Jeu and coworkers
92: using the synchrotron X-ray reflectivity measurements have shown the
93: absence of surface melting for the $(001)$ face of the bulk biphenyl
94: up to $2$ K below the melting point \cite{deJeu}. It was proposed that surface
95: melting observed
96: by Chernov and Yakovlev is induced by the
97: roughness of the glass used in their ellipsometry experiment \cite{deJeu}.
98: Surface properties of biphenyl, caprolactam, and other
99: organic materials and organic thin films at high temperatures
100: deserve further experimental and theoretical studies.
101: Investigations of organic films are also important for applications,
102: such as tribology and lubrication, corrosion inhibition, modification of
103: the substrate optical properties, molecular electronics, and various
104: industrial chemical processes.
105:
106:
107:
108: Surface melting for a $\sim 10$ layer thick film of $CH_4$ adsorbed on
109: $MgO(100)$ was studied by neutron scattering
110: \cite{Bienfait,Gay}. It was found that the thickness of the quasi-liquid phase
111: changes from one layer at $72$ K to $\sim 6$ layers at $90.3$ K.
112: (At atmospheric pressure
113: the bulk melting point of methane is $90.66$ K, whereas the liquid-vapour
114: transition occurs at $111.56$ K.) The neutron-scattering measurements of the
115: translational diffusion coefficient above $72$ K
116: have shown that this quantity is in the
117: $10^{-5}$ $cm^2$ $s^{-1}$ range, i.e., it has a value typical for mobility
118: in quasi-liquid films.
119: It was found that for a methane thick
120: film ($\sim 20$ layers) at $T=94$ K (i.e., above the bulk melting temperature)
121: almost all layers are liquefied \cite{Gay}.
122: Recently the layering and melting properties of methane on $MgO(100)$
123: were studied between $70$ K and $96$ K using high-resolution
124: adsorption isotherms and for the film thickness of up to five layers.
125: \cite{Larese}. Melting transition was observed at $80$ K for the monolayer,
126: and at $85$ K for layers $2$ through $4$.
127:
128:
129:
130: Related to the experiments on surface melting of bulk
131: caprolactam and biphenyl, it would be interesting to study surface melting of
132: a bulk methane. These studies should determine whether surface melting of
133: the $CH_4$ film on $MgO$ is inherent feature of this organic material,
134: or it is induced by the substrate.
135: Interfacial melting is an appearance of a liquid film at the interface between
136: one solid material and the substrate of the other material as the bulk melting point
137: of the first material is approached \cite{Wilen,Beaglehole}.
138: Therefore, surface melting is a special case of
139: interfacial melting where the substrate is the vapor phase of the first material.
140: In this work long-range interactions important for surface melting of a bulk methane,
141: as well as for interfacial melting of a methane film on the MgO substrate
142: are analyzed using the theory of Dzyaloshinskii, Lifshitz, and Pitaevskii
143: (DLP) \cite{Dlp}. Our findings revealed that complete surface melting
144: exists for the bulk methane.
145: Because of an attractive interaction between interfaces
146: melting is absent at the boundary between
147: the $CH_4$ film and the $MgO$ substrate.
148:
149:
150:
151: In the following
152: the DLP theory of the van der Waals interaction between planar
153: interfaces and its consequences for surface melting are described in Sec. II.
154: The calculation of frequency dependent dielectric functions of $CH_4$ and $MgO$ is
155: discussed in Sec. III.
156: Results and discussion for the surface and interfacial melting problems
157: are presented in Sec. IV.
158: A summary and conclusions are given in Sec. V.
159:
160:
161:
162:
163:
164:
165: \section{Surface melting and Van der Waals forces}
166:
167: The surface free energy per unit area for a system with a quasi-liquid film
168: of the thickness $l$, at the temperature $T$ is
169: \cite{Amsterdam,Dash,Pluis}
170: \begin{equation}
171: {\cal F}(l,T)=\gamma_{sl} + \gamma_{lv} + L_m (1-T/T_m)l +
172: F_{sr} + F_{lr}.
173: \label{eq:1}
174: \end{equation}
175: In this equation $\gamma_{sl}$ and $\gamma_{lv}$ are the free energies of
176: the solid-liquid and liquid-vapor interfaces, respectively, $L_m$ is the
177: latent heat of melting per unit volume, and $T_m$ is the bulk melting
178: temperature. The fourth term $F_{sr}$ represents the gain in interfacial
179: energy associated with the wetting of the solid surface by a quasi-liquid
180: film. This term is the result of short-range interactions.
181: Their contribution is oscillatory and decreases exponentially to zero
182: for large $l$. Short-range forces produce a logarithmic temperature dependence
183: of the liquid film thickness as the bulk melting temperature is approached.
184: Long range van der Waals interactions are given by the
185: last term $F_{lr}$ in Eq. (1).
186: The contribution of the van der Waals forces for
187: non-retarded interactions (where $l$ is much smaller than the wavelength
188: of light for a characteristic excitations) is given by $H l^{-2}$.
189: The parameter $H$ is the Hamaker constant and gives information about the
190: magnitude of the van der Waals interactions in different media
191: \cite{Israelachivili}. (Sometimes $H'=12\pi H$ or $H''=-12\pi H$ are
192: called
193: the Hamaker constant.)
194: Surface melting occurs when $H>0$, i.e., when two interfaces
195: (solid-liquid and liquid-vapour) repel each other.
196: Then $F_{lr}(l)$ is monotonously decreasing function, or its minimum is
197: not deep enough to bind the interface.
198: The van der Waals contribution produces a power-law temperature dependence of the
199: liquid film thickness as $T\rightarrow T_m$
200: \begin{equation}
201: l=\left[\frac {L_m(T_m-T)}{2HT_m}\right]^{-1/3}.
202: \label{eq:2}
203: \end{equation}
204: For $H<0$ two interfaces attract each other and the melting is at most
205: incomplete. Then a deep minimum exists on the free energy
206: curve $F_{lr}$, and the interface is trapped there.
207:
208:
209:
210:
211: Dzyaloshinskii, Lifshitz, and Pitaevskii developed the theory of
212: long-range dispersion interaction between macroscopic bodies
213: \cite{Dlp,Israelachivili}.
214: The van der Waals interaction between interfaces
215: is a consequence of electromagnetic fluctuations in a polarizable medium
216: and the discontinuity of the dielectric function across the boundaries.
217: In the DLP theory the force between interfaces
218: (such as those shown in Fig. 1) is a function of the frequency-dependent
219: dielectric functions $\epsilon(\omega)$ of materials in contact.
220: In the following all media are taken as isotropic and continuous.
221: The free energy per unit area of two media (1 and 2) separated by
222: a film of thickness $l$ (media 3, see Fig. 1) is \cite{Dlp,Elbaum,Wilen}
223: \begin{eqnarray}
224: F_{lr}(l,T)= \frac {kT}{8\pi l^2}
225: \sum_{m=0}^{\infty} \ '
226: \int_{r_{m}}^{\infty}
227: dx \ x \ (ln (1- \frac {(x-x_1)(x-x_2)} {(x+x_1)(x+x_2)}
228: exp(-x)) \nonumber \\
229: +ln (1- \frac {(\epsilon _2 x -\epsilon _3 x_2)
230: (\epsilon _1 x - \epsilon _3 x_1)}
231: {(\epsilon _2 x + \epsilon _3 x_2)(\epsilon _1 x + \epsilon_3 x_1)} exp(-x))),
232: \end{eqnarray}
233: where
234: \begin{equation}
235: r_m=2l\xi _m (\epsilon _ 3)^{1/2}/c,
236: \label{eq:4}
237: \end{equation}
238: and
239: \begin{equation}
240: x_{1,2}= \left(x^2 - r_m^2(1- \frac {\epsilon_{1,2}}
241: {\epsilon_3})\right)^{1/2}.
242: \label{eq:5}
243: \end{equation}
244: The dielectric functions $\epsilon _1 (\omega)$,
245: $\epsilon _2 (\omega)$, and $\epsilon _3 (\omega)$ are evaluated at
246: imaginary frequencies $i\xi _m = i(2\pi k T/{\hbar})m$.
247: The functions $\epsilon(i\xi)$ are positive, real and
248: decrease monotonically from the static dielectric function $\epsilon _0$
249: for $\xi=0$, to $1$ as $\xi \to \infty$.
250: In the above expressions $k$, $\hbar$, and $c$ are
251: the Boltzmann and Planck constants and velocity of light, respectively.
252: The prime on the sum in Eq. (3)
253: means that the term $m=0$ has to be multiplied with
254: $\frac {1} {2}$.
255: In the derivation of Eq. (3) retarded interactions were assumed.
256: If retardation is ignored,
257: then $c \to \infty$, $r_m \to 0$, and Eq. (1) simplifies to
258: \begin{equation}
259: F_{lr}^{nr}(l,T)= \frac {kT} {8\pi l^2} \sum_{m=0}^{\infty}
260: \int_0^{\infty}
261: dx \ x \ ln \left(1- \frac {(\epsilon _2 (i\xi_m) -\epsilon _3(i\xi_m) )
262: (\epsilon _1 (i\xi_m) - \epsilon _3(i\xi_m) )}
263: {(\epsilon _2 (i\xi_m) + \epsilon _3(i\xi_m) )
264: (\epsilon _1(i\xi_m) + \epsilon_3(i\xi_m))} exp(-x) \right).
265: \label{eq:6}
266: \end{equation}
267: A common approach
268: is to obtain an approximate value of $\epsilon (\omega)$
269: from the optical measurements
270: \cite{Wilen,Israelachivili,Elbaum,Chen,DalCorso,Parsegian,Hough}.
271: An oscillator model of the dielectric response is then used to
272: generate data, i.e.,
273: the dielectric function is represented by
274: \begin{equation}
275: \epsilon (\omega) = 1 + \sum_i \frac {f_i} {{e_i}^2 - i\hbar\omega g_i
276: -(\hbar\omega)^2},
277: \label{eq:7}
278: \end{equation}
279: where $f_i$, $e_i$, and $g _i$ are fitting parameters.
280: The DLP theory was recently applied to surface
281: melting of ice \cite{Elbaum,Wilen}, germanium, and several metals
282: \cite{Chen,DalCorso}.
283: Unfortunately, optical and dielectric properties of many organic materials
284: are less studied than, for example of metals, semiconductors, or
285: $H_2O$.
286: As a consequence, we were not able to find necessary optical data for
287: biphenyl and caprolactam, two organic materials for which
288: experiments on surface melting were performed. In contrast,
289: we found data for optical properties
290: of organic material $CH_4$ and dielectric substrate $MgO$ \cite{Martonchik,Palik}.
291:
292:
293:
294: \section{Dielectric functions of methane and MgO}
295:
296:
297:
298: The complex dielectric functions $\epsilon(i\xi)$ were constructed
299: using experimental data for optical constants of $CH_4$ and $MgO$
300: \cite{Martonchik,Palik}.
301: As suggested in Ref. \cite{Elazar} we minimized the objective function
302: \begin{equation}
303: E= \sum_{j=1}^N
304: {\left[\left |{\frac
305: {\epsilon_r(\omega_j)-\epsilon_r^{exp}(\omega_j)}
306: {\epsilon_r^{exp}(\omega_j)}}\right | +
307: \left | {\frac
308: {\epsilon_i (\omega_j) - \epsilon_i^{exp} (\omega_j)}
309: {\epsilon_i^{exp}(\omega_j)}}\right |\right]}^2,
310: \label{eq:8}
311: \end{equation}
312: where $\epsilon_r(\omega)$ and $\epsilon_i(\omega)$ are the real and
313: imaginary parts of a dielectric function given by Eq. (7), and '{\it exp}'
314: denotes experimental values taken from Refs. \cite{Martonchik,Palik}.
315: Therefore, the calculation was performed by finding a minimum of
316: an objective function based on
317: a simultaneous fitting of both real and imaginary parts of the
318: dielectric function. We found that, in comparison with some other functions,
319: the functional form in Eq. (8) is suitable for minimization.
320: In principle, the most important factor in the minimization process is the
321: quality of experimental data, i.e., the number of points and their
322: distribution. It is known that the main contribution belongs to data in the UV
323: region of the spectra \cite{Parsegian,Hough}.
324: The UV part of the spectrum for $MgO$ is represented with many points
325: \cite{Palik}. In contrast, in the UV spectra for the solid and liquid methane
326: few points exist \cite{Martonchik}.
327: A smaller number of points gives more freedom in the choice of the parameters
328: which minimize Eq. (8). Therefore, more local minima exist and they
329: produce similar values of the objective function.
330: This is a limitation for the calculation of dielectric properties of methane.
331: Fortunately previous calculations for other
332: materials have shown that the complex dielectric functions
333: $\epsilon(i\xi)$ which enter expressions in the DLP theory are not
334: very sensitive to the fitting procedure and
335: the detailed experimental information.
336: As suggested in Ref. \cite{Parsegian} and followed by others
337: \cite{Elbaum,Wilen}, we used the constraint $\epsilon_r=n_{vis}^2$,
338: where $n_{vis}$ is the refractive index in the visible part of the
339: spectrum. The value $n_{vis}$ is known with a great accuracy and
340: fitting with this constraint gives better results.
341: Objective functions were minimized using the simplex method
342: \cite{Numrec,Multisimplex}.
343: The accuracy of calculation was checked by changing minimization routines and
344: also by calculating dielectric functions for other materials
345: where extensive data are available
346: \cite{Wilen,Israelachivili,Elbaum,Chen,DalCorso,Parsegian,Hough}.
347:
348:
349: Martonchik and Orton,
350: motivated by astrophysical applications, recently
351: calculated the optical constants of liquid and solid methane using the
352: spectra from the literature \cite{Martonchik}. Liquid methane was
353: studied at the boiling and melting temperatures, whereas the solid phase I
354: of methane was studied at the melting point and at $T=30$ K.
355: (The solid phase I is the face centered cubic phase of methane which
356: at atmospheric pressure exists for $T>20$ K.)
357: Therefore, data obtained by Martonchik and Orton for the solid and
358: liquid methane at the melting point were used to model the solid and liquid
359: phase. For the vapor phase we put $\epsilon=1$.
360: The static dielectric function of the gas methane is $\epsilon=1.000944$ \cite{CRC}.
361: The fits for $CH_4$ were performed
362: to match dielectric functions at zero frequency. These
363: static dielectric functions
364: are $\epsilon_0=1.67$ for the liquid methane, and
365: $\epsilon_0=1.74$ for the solid methane \cite{Martonchik}.
366: The calculated complex dielectric functions of the solid and liquid methane
367: are shown in Fig. 2.
368: Corresponding fitting parameters are given in Tables I and II.
369: The curves $\epsilon (i\xi)$ for the solid and liquid methane are close to
370: each other, but do not cross. In similar calculation for surface melting
371: of ice it was found that the complex dielectric functions of
372: water and ice
373: cross at $10$ eV (i.e., for $m \sim 205$) \cite{Elbaum,Wilen}.
374:
375:
376:
377: The detailed review of optical properties of $MgO$ was
378: given by Roessler and Huffman \cite{Palik}.
379: The values of optical constants were tabulated in the range $(0.002-586)$ eV.
380: Because of a large set of data we use only selected points.
381: From these data
382: the function $\epsilon(i\xi)$ for $MgO$ was constructed.
383: The static dielectric function of $MgO$ is
384: $\epsilon_0 = 9.8$ \cite{Palik}.
385: Here, as in Ref. \cite{Sabisky} where the fitting
386: based on a double band was performed,
387: we did not match the static dielectric function.
388: The value of static dielectric function of $MgO$ is much larger than the
389: average value of $\epsilon_r$ in the UV part of the spectrum.
390: Additional problem is the existence of a close sharp peak at
391: $\hbar \omega = 7.8 $ eV. Therefore,
392: the inclusion of the static dielectric function into the fit for $MgO$
393: prevents a good behavior of the whole fitted curve.
394: Fitting parameters for $MgO$ are given in Table III.
395: The complex dielectric functions $\epsilon(i\xi)$ of
396: $MgO$ is shown in Fig. 2.
397: Our four-band result is almost the same as a double band calculation
398: from Ref. \cite{Sabisky}.
399:
400:
401:
402:
403: \section{Surface and interfacial melting}
404:
405: \subsection{Bulk methane}
406:
407: In surface melting problem
408: the frequency dependent dielectric functions of solid,
409: liquid (melt), and gas methane are respectively:
410: $\epsilon _1 = \epsilon _s (\omega)$,
411: $\epsilon _3 = \epsilon _l (\omega)$, and
412: $\epsilon _2 = 1$ (see Fig. 1).
413: The integral in Eq. (3) was calculated numerically using the Gaussian method.
414: The accuracy of the integration technique was checked by changing
415: numerical procedures and also by reproducing the results for ice
416: \cite{Elbaum}.
417: The calculated free energy as a function of liquid film thickness is
418: shown in Fig. 3. The minimum is reached for $l\rightarrow\infty$.
419: This functional form of $F(l)$ is typical for complete surface melting.
420: We found that the Hamaker constant is
421: $H=5.26 \times 10^{-22}$ J. Similar values were calculated for metals with
422: melted surfaces, such as $Pb$, $Al$, and $Au$ \cite{Chen}.
423:
424:
425:
426:
427:
428:
429:
430: \subsection{Methane film on MgO}
431:
432:
433: In interfacial melting problem,
434: and for interfaces shown in Fig. 1:
435: $\epsilon_1 = \epsilon_{MgO}(\omega)$
436: is dielectric function for $MgO$,
437: $\epsilon_2 = \epsilon_s (\omega)$ is for a solid phase,
438: and $\epsilon_3 = \epsilon_l (\omega)$ is for a liquid phase of methane.
439: Therefore, we consider the solid $CH_4$ - liquid $CH_4$ - $MgO$ substrate
440: system. The same numerical procedure as for the surface melting problem was applied.
441: The free energy is shown in Fig. 4. The Hamaker constant is
442: $H=-6.03 \times 10^{-22}$ J.
443: Therefore, the van der Waals forces produce an attractive interaction
444: between interfaces and prevent interfacial melting.
445: The functional form of $F(l)$ shown in Fig. 4 was not found for interfacial
446: melting of ice on various substrates \cite{Wilen}, but two other forms
447: of $F(l)$ were calculated there. They correspond to complete and
448: incomplete interfacial melting of ice on the particular substrate.
449:
450:
451:
452: \section{Summary and Conclusions}
453:
454:
455: The interaction of interfaces present in
456: surface melting of the bulk methane and interfacial melting of
457: the $CH_4$ film on the $MgO$ substrate
458: is analyzed within the
459: Dzyaloshinskii, Lifshitz, and Pitaevskii theory of
460: the van der Waals interaction. Frequency dependent dielectric functions for
461: these materials are calculated from optical data.
462: It was found that complete surface melting exists for methane.
463: Our study is the first investigation of surface melting properties of the
464: bulk methane.
465: We also found that the van der Waals interactions do not produce
466: interfacial melting of the $CH_4$ film on the $MgO$ substrate.
467: Therefore, premelting observed in the experimental investigations
468: of the $CH_4$ films on the $MgO$ substrate \cite{Bienfait,Gay,Larese}
469: is inherent feature of the bulk methane.
470: Methane is till now the only organic material for which clear
471: complete surface melting was found.
472: The results obtained here depend on the quality of the
473: optical measurements used to model dielectric functions.
474: Nevertheless, these results shed light on the
475: interaction between interfaces in surface and interfacial melting involving
476: methane. These results suggest
477: experimental studies of surface melting of the bulk methane,
478: for example using the synchrotron X-ray reflectivity.
479: In addition, {\it ab initio} molecular dynamics study of
480: the high-temperature properties
481: of methane surfaces is feasible,
482: similar as a recent simulation of a high-pressure behavior for
483: bulk $CH_4$ \cite{Francesco}.
484: It is important to do all necessary optical measurements for biphenyl
485: and caprolactam, and then the calculation of the van der Waals interactions for
486: these materials.
487: There is a possibility that
488: surface melting observed for biphenyl on the glass surface \cite{Chernov}
489: is interfacial melting induced by the dielectric properties of the substrate.
490: Till now the roughness of the glass and impurity induced
491: interfacial melting were
492: proposed to explain the results of Chernov and Yakovlev.
493: A possibility of interfacial melting induced by the van der Waals forces
494: for biphenyl on the glass
495: also explains the absence of surface melting
496: for the bulk biphenyl found using the X-ray reflectivity \cite{deJeu}.
497: Recent atomic force microscopy study of the (001) surfaces of
498: $n-C_{23}H_{48}$ paraffin crystals \cite{Plomp} shows that this technique
499: gives a good opportunity to investigate melt growth on a molecular scale
500: for surfaces of methane and other organic materials.
501:
502:
503:
504: \acknowledgments
505: This work has been carried out under the HR-MZT project 119206 -
506: ``Dynamical Properties of Surfaces''.
507: We would like to thank I. Kup{\v c}i{\' c} and E. Ljubovi{\'c}
508: for their help.
509:
510:
511:
512:
513:
514: \begin{references}
515:
516:
517: \bibitem{Amsterdam}
518: J. F. van der Veen, B. Pluis, and A. W. Denier van der Gon,
519: in {\it Chemistry and Physics of Solid Surfaces VII}, edited by
520: R. Vanselow and R. F. Hove (Springer, Berlin, 1988), p.455.
521:
522: \bibitem{Dash}
523: J. G. Dash, Contemp. Phys. {\bf 30}, 89 (1989).
524:
525: \bibitem{Frenken}
526: H. M. van Pinxteren and J. W. M. Frenken,
527: Surf. Sci. {\bf 275}, 383 (1992).
528:
529: \bibitem{Goranka}
530: G. Bilalbegovi{\' c} and E. Tosatti,
531: Phys. Rev. B {\bf 48}, 11240 (1993).
532:
533: \bibitem{Chandavarkar}
534: S. Chandavarkar, R. M. Geertman, and W. H. de Jeu,
535: Phys. Rev. Lett. {\bf 69}, 2384 (1992).
536:
537: \bibitem{Wong}
538: G. C. L. Wong, W. H. de Jeu, G. B{\"o}gels, E. P. G. van den Berg,
539: and P. Bennema, Phys. Rev. B {\bf 55}, 1388 (1997).
540:
541:
542: \bibitem{Chernov}
543: A. A. Chernov and V. A. Yakovlev,
544: Langmuir {\bf 3}, 635 (1987);
545: A. A. Chernov and V. A. Yakovlev,
546: Pis'ma Zh. Eksp. Teor. Fiz. {\bf 45}, 130 (1987)
547: [JETP Lett. {\bf 45}, 160 (1987)].
548:
549:
550: \bibitem{deJeu}
551: W. H. de Jeu, J. D. Shindler, S. Chandavarkar, R. M. Gertman, and K. Liang,
552: Surf. Sci. {\bf 342}, 341 (1995).
553:
554:
555: \bibitem{Bienfait}
556: M. Bienfait, Europhys. Lett. {\bf 4}, 79 (1987).
557:
558:
559: \bibitem{Gay}
560: M. Bienfait, J. M. Gay, and H. Blank,
561: Surf. Sci. {\bf 204}, 331 (1988).
562:
563:
564: \bibitem{Larese}
565: A. Freitag and J. Z. Larese, Phys. Rev. B {\bf 62}, 8360 (2000).
566:
567:
568: \bibitem{Wilen}
569: L. A. Wilen, J. S. Wettlaufer, M. Elbaum, and M. Schick,
570: Phys. Rev. B {\bf 52}, 12426 (1995).
571:
572:
573: \bibitem{Beaglehole}
574: D. Beaglehole and P. Wilson,
575: J. Phys. Chem. {\bf 98}, 8096 (1994).
576:
577: \bibitem{Dlp}
578: I. E. Dzyaloshinskii, E. M. Lifshitz, and L. P. Pitaevskii,
579: Adv. Phys. {\bf 10}, 165 (1961);
580: V. L. Ginzburg, {\it Theoretical Physics and Astrophysics},
581: (in russian) (Nauka, Moskow, 1981), p. 350;
582: Y. S. Barash and V. L. Ginzburg,
583: Usp. Fiz. Nauk {\bf 143}, 345 (1984)
584: [Sov. Phys. Usp. {\bf 27}, 467 (1984)].
585:
586: \bibitem{Pluis}
587: B. Pluis, D. Frenkel, and J. F. van der Veen,
588: Surf. Sci. {\bf 239}, 282 (1990);
589: D. E. Sullivan and M. M. Telo da Gama,
590: in: {\it Fluid Interfacial Phenomena}, edited by C. A. Croxton
591: (Wiley, New York, 1986).
592:
593: \bibitem{Israelachivili}
594: J. N. Israelachivili,
595: {\it Intermolecular and Surface Forces},
596: (Academic Press, London, 1985).
597:
598:
599:
600: \bibitem{Elbaum}
601: M. Elbaum and M. Schick,
602: Phys. Rev. Lett. {\bf 66}, 1713 (1991).
603:
604:
605: \bibitem{Chen}
606: X. J. Chen, A. C. Levi, and E. Tosatti,
607: Surf. Sci. {\bf 251/252}, 641 (1991);
608: Nuovo Cimento D {\bf 13}, 919 (1991).
609:
610: \bibitem{DalCorso}
611: A. DalCorso and E. Tosatti,
612: Phys. Rev. B {\bf 47}, 9742 (1993).
613:
614: \bibitem{Parsegian}
615: V. A. Parsegian and G. H. Weiss,
616: J. Colloid Interface Sci. {\bf 81}, 285 (1981).
617:
618: \bibitem{Hough}
619: D. B. Hough and L. R. White,
620: Adv. Colloid Interface Sci. {\bf 14}, 3 (1980).
621:
622:
623: \bibitem{Martonchik}
624: J. V. Martonchik and G. S. Orton,
625: Appl. Optics {\bf 33}, 8306 (1994).
626:
627: \bibitem{Palik}
628: D. M. Roessler and D. R. Huffman, in ``{\it Handbook of optical constants of
629: solids II}'', ed. by E. D. Palik, p. 919 (Academic Press, Boston, 1991).
630:
631:
632: \bibitem{Elazar}
633: A. B. Djuri{\v s}i\'c, A. D. Raki\'c, and J. M. Elazar,
634: Phys. Rev. E {\bf 55}, 4797 (1997).
635:
636:
637:
638: \bibitem{Numrec}
639: W. H. Press, S. A. Teukolsky, W. T. Vetterling, and B. P. Flannery,
640: {\it Numerical Recipes in Fortran} and
641: {\it Numerical Recipes in C},
642: (Cambridge University Press, Cambridge, 1992/93).
643:
644: \bibitem{Multisimplex}
645: F. H. Walters, L. R. Parker, S. L. Morgan, and S. H. Deming,
646: {\it Sequential Simplex Optimization},
647: http://www.multisimplex.com/simplexbook/index.htm,
648: (MultiSimplex AB, 1997-2000).
649:
650:
651: \bibitem{CRC}
652: {\it Handbook of Chemistry and Physics}, editied by R. C. Weast and M. J. Astle,
653: (CRC Press, Boca Raton, 1979).
654:
655:
656:
657: \bibitem{Sabisky}
658: E. S. Sabisky and G. H. Anderson,
659: Phys. Rev. A {\bf 7}, 790 (1973).
660:
661:
662:
663: \bibitem{Francesco}
664: F. Ancilotto, G. L. Chiarotti, S. Scandolo, and E. Tosatti,
665: Science {\bf 275}, 1288 (1997).
666:
667:
668: \bibitem{Plomp}
669: M. Plomp, P. J. C. M. van Hoof, and W. J. P. van Enckevort,
670: Surf. Sci. {\bf 448}, 231 (2000).
671:
672:
673:
674: \end{references}
675:
676:
677:
678: \begin{figure}
679: \caption{Schematic representation for configuration of interfaces between
680: media characterized by the frequency
681: dependent dielectric functions.}
682: \label{fig1}
683: \end{figure}
684:
685:
686:
687:
688: \begin{figure}
689: \caption{Complex dielectric functions. Dashed line for $MgO$ is the fit
690: from Ref. [28]. The index $m$ is defined in Eqs. (3) and (6)
691: in the text, and $m=1000$ corresponds to $48.8$ eV. The contribution of
692: various $m$-dependent terms in Eqs. (3) and (6) is obvious.}
693: \label{fig2}
694: \end{figure}
695:
696:
697:
698:
699: \begin{figure}
700: \caption{The free energy per unit area ($a_0$ is the Bohr radius)
701: as a function of the film thickness $l$
702: for surface melting of the bulk methane.
703: The dashed line shows the free energy without retardation,
704: whereas the full line represents the result with retardation.}
705: \label{fig3}
706: \end{figure}
707:
708:
709:
710:
711: \begin{figure}
712: \caption{The free energy vs film thickness
713: for interfacial melting of a methane film on $MgO$.
714: Details as in Fig. 3.}
715: \label{fig4}
716: \end{figure}
717:
718:
719:
720:
721: \begin{table}
722: \caption{Parameters used for an oscillator model of dielectric response
723: for the solid methane using data from Ref. [22].}
724: \label{table1}
725: \begin{tabular}{l l l }
726: $e_i$ (eV) & $f_i$ ($eV^2$) & $g_i$ (eV)\\
727: \hline
728: $10.0$ & $10.0$ & $1.8$ \\
729:
730: $12.7$ & $44.0$ & $3.6$ \\
731:
732: $14.4$ & $64.0$ & $3.5$ \\
733:
734: $22.3$ & $10.0$ & $2.4$ \\
735:
736: $23.1$ & $23.0$ & $6.8$ \\
737: \end{tabular}
738: \end{table}
739:
740:
741: \begin{table}
742: \caption{Parameters used for an oscillator model of dielectric response
743: for the liquid methane using data from Ref. [22].}
744: \label{table2}
745: \begin{tabular}{l l l }
746: $e_i$ (eV) & $f_i$ ($eV^2$) & $g_i$ (eV) \\
747: \hline
748: $10.0$ & $8.0$ & $1.5$ \\
749:
750: $12.8$ & $42.0$ & $3.5$ \\
751:
752: $14.8$ & $62.0$ & $4.0$ \\
753:
754: $22.7$ & $8.0$ & $2.4$ \\
755:
756: $23.8$ & $21.0$ & $6.5$ \\
757: \end{tabular}
758: \end{table}
759:
760:
761:
762:
763:
764:
765: \begin{table}
766: \caption{Parameters used for an oscillator model of dielectric response
767: for solid MgO using data from Ref. [23].}
768: \label{table3}
769: \begin{tabular}{l l l }
770: $e_i$ (eV) & $f_i$ ($eV^2$) & $g_i$ (eV) \\
771: \hline
772: $7.6$ & $3.0$ & $0.1$ \\
773:
774: $10.8$ & $150.0$ & $3.5$ \\
775:
776: $13.2$ & $40.0$ & $1.0$ \\
777:
778: $17.0$ & $130.0$ & $4.0$ \\
779: \end{tabular}
780: \end{table}
781:
782:
783:
784:
785: \end{document}
786:
787:
788:
789:
790:
791:
792:
793:
794:
795:
796:
797:
798:
799:
800:
801:
802:
803:
804:
805:
806:
807:
808:
809:
810:
811:
812:
813:
814:
815:
816:
817:
818:
819:
820:
821:
822:
823:
824:
825:
826:
827:
828:
829:
830: