1: \documentclass[12pt]{article}
2: \usepackage{epsfig,amstex,amssymb}
3: \pagestyle{empty}
4:
5: \parindent0cm
6:
7: \setlength{\textwidth}{16cm}
8: \addtolength{\evensidemargin}{-1.5 cm}
9: \addtolength{\oddsidemargin}{-1.5 cm}
10:
11: \addtolength{\textheight}{9\baselineskip}
12: \addtolength{\topmargin}{-1.4 cm}
13:
14: \begin{document}
15:
16: \noindent \textit{Mat. Res. Soc. Symp. Proc. \textbf{651}: T5.3.1-T5.3.6 (2001)}\\
17:
18: \noindent \textbf{\Large Mechanical, adhesive and thermodynamic \\
19: properties of hollow nanoparticles}\\
20:
21: \noindent \textbf{U. S. Schwarz$^{1}$, S. A. Safran and S. Komura$^{2}$} \\
22: Department of Materials and Interfaces, Weizmann Institute, Rehovot 76100, Israel\\
23: ${}^{1}$Max-Planck-Institute of Colloids and Interfaces, 14424 Potsdam, Germany\\
24: ${}^{2}$Department of Chemistry, Tokyo Metropolitan University, Tokyo 192-0397, Japan
25:
26: \section*{Abstract}
27:
28: When sheets of layered material like C, WS$_2$ or BN are restricted to
29: finite sizes, they generally form single- and multi-walled hollow
30: nanoparticles in order to avoid dangling bonds. Using continuum
31: approaches to model elastic deformation and van der Waals interactions
32: of spherical nanoparticles, we predict the variation of mechanical
33: stability, adhesive properties and phase behavior with radius $R$ and
34: thickness $h$. We find that mechanical stability is limited by forces in
35: the nN range and pressures in the GPa range. Adhesion energies scale
36: linearly with $R$, but depend only weakly on $h$. Deformation due to
37: van der Waals adhesion occurs for single-walled particles for radii of
38: few nm, but is quickly suppressed for increasing thickness. As $R$
39: is increased, the gas-liquid coexistence disappears from the phase
40: diagram for particle radii in the range of 1-3 nm (depending on wall
41: thickness) since the interaction range decreases like $1/R$.
42:
43: \section*{Introduction}
44:
45: The prototypical hollow nanoparticle is the buckyball C$_{60}$, which
46: crystallizes into \textit{fullerite}. Many other morphologies for
47: carbon sheets have been found since the early 1990s, including
48: fullerenes C$_n$ of varying size, multi-walled fullerenes
49: (\textit{carbon onions}), and single- and multi-walled carbon
50: nanotubes. Closed structures of carbon are formed in order to avoid
51: dangling bonds at the edges of finite sized carbon sheets. This
52: mechanism is generic for anisotropic layered material of finite size,
53: and up to now more than 30 other materials have been prepared as
54: hollow nanoparticles (see the reviews \cite{n:tenn95} and references
55: therein). This includes the metal dichalcogenides MeX$_2$ (Me = W,
56: Se, X = S, Se), BN, and GaN. Usually \textit{inorganic
57: fullerenes} like WS$_2$ and MoS$_2$ are multi-walled. In
58: Fig.~\ref{tem} we show transmission electron micrographs of
59: MoS$_2$-particles which were synthesized by solid-gas reaction.
60: Control of size and shape is rather difficult, as evidenced by the
61: irregularly faceted shapes. However, the methods used for the
62: synthesis of hollow nanoparticles are developing very rapidly, and it
63: is to be expected that control of size, thickness and shape will
64: become much better in the future.
65:
66: \begin{figure}[tb]
67: \begin{center}
68: \begin{tabular}{cc}
69: \epsfig{figure=TEM.eps,width=0.5\textwidth} &
70: \epsfig{figure=Interaction.eps,width=0.5\textwidth}
71: \end{tabular}
72: \end{center}
73: \caption{Left: Transmission electron micrographs of hollow MoS$_2$-nanoparticles
74: (courtesy of R. Tenne). Van der Waals interactions leads to strong
75: adhesion. The faceted shape is due to grain boundaries. Right: (a)
76: Model for two spherical nanoparticles of radius $R$ and thickness $h$
77: each, interacting over a distance $r$. (b) The effective interaction
78: potential $V$ as a function of distance $r$ is characterized by the hard
79: core diameter $2 R$, potential depth $\epsilon \sim R$ and interaction
80: range $\delta \sim 1/R$.}
81: \label{tem}
82: \end{figure}
83:
84: Hollow nanoparticles combine covalent in-plane strength with flexible
85: out-of-plane bending of thin films, which results in high mechanical
86: stability. They interact by van der Waals (vdW) forces, which for
87: example determine the material properties of fullerite \cite{n:giri92}
88: and the phase behavior of buckyballs \cite{n:cacc97}. VdW interactions
89: become even more important for larger fullerenes, carbon onions and
90: inorganic fullerenes; these materials, therefore, feature a large
91: degree of non-specific adhesion. In particular, vdW adhesion can lead
92: to deformation of the hollow nanoparticles. For example, it has been
93: found that carbon nanotubes adhering to each other or to flat surfaces
94: show observable deformations for radii as small as 1 nm
95: \cite{n:ruof93}. In the presence of local surface features
96: like crystal steps, adhering nanoparticles will have spacially varying
97: electronic properties due to spacially varying deformation.
98:
99: Hollow nanoparticles may lead to new applications in nanoelectronics
100: and -optics. In contrast to filled nanoparticles, they can also be
101: used for storage and delivery systems. They show superior mechanical
102: properties, like high flexibility, high tensile strength and light
103: weight, which will lead to applications as ultra-strong fibers.
104: Larger fibers can be produced using the cohesiveness provided by vdW
105: interactions. Inorganic fullerenes have also been shown to offer
106: favorable tribological properties \cite{n:rapo97}. Detailed
107: investigations with the Surface Force Apparatus demonstrated that low
108: friction and wear is caused by material transfer onto the sliding
109: surfaces \cite{n:gola99}.
110:
111: For these applications, a good theoretical understanding of the
112: physical properties of hollow nanoparticles is important. Here we
113: discuss structural properties of spherical hollow nanoparticles in a
114: continuum approach that allows us to focus on unusual generic
115: properties which result mainly from geometrical effects. In
116: particular, we address mechanical properties, van der Waals
117: interaction, thermodynamic behavior and deformation by adhesion. In
118: this paper, we focus on the main results of our work; more details can
119: be found elsewhere \cite{uss:schw00d,uss:schw00e}.
120:
121: \section*{Mechanical properties}
122:
123: During recent years, mechanical properties of fullerenes and carbon
124: nanotubes have been extensively studied at the level of molecular
125: calculations, that is by first principles, tight binding and force
126: field techniques. The advantage of molecular models is that they can
127: provide detailed quantitative predictions. For larger systems like
128: multi-walled nanoparticles, their implementation becomes difficult due
129: to computer time requirements. Here we use classical continuum
130: elasticity theory, which is asymptotically correct for large systems
131: and allows us to treat all these systems using the same framework. The
132: large advantage of this approach is that it provides insight into the
133: generic properties of these systems. Molecular calculations for carbon
134: nanotubes have shown that the predictions of continuum elasticity
135: theory persist well into the limit of radii smaller than 1 nm
136: \cite{n:robe92}. Recently this observation could be
137: verified also for BN and MoS$_2$ nanotubes \cite{n:hern98}.
138:
139: The layered material discussed here usually has a hexagonal lattice
140: structure, for which elasticity theory \cite{b:land70} predicts that there
141: are only two in-plane elastic constants and the sheet is elastically
142: isotropic. Its deformation energy consists of stretching and bending
143: terms. For single-walled carbon sheets, the in-plane stretching
144: modulus $G \approx 3.6 \times 10^5$ erg/cm$^2$ and the out-of-plane
145: bending rigidity $\kappa \approx 3 \times 10^{-12}$ erg. For both
146: MoS$_2$ and WS$_2$, $G$ is smaller by a factor $4$ and $\kappa$ is
147: larger by a factor $5$. For multi-walled nanoparticles, the elastic
148: constants scale with thickness $h$ as $G
149: \sim C_{11} h$ and $\kappa \sim C_{11} h^3$, where $C_{11}$ is
150: the largest in-plane elastic constant of the corresponding layered
151: material \cite{n:srol94}. The values for $C_{11}$ are $1060$, $238$ and $150
152: \times 10^{10}$ erg/cm$^3$ for C, MoS$_2$ and WS$_2$, respectively.
153:
154: Stretching can be avoided completely by bending the sheet into a
155: cylindrical nanotube. However, this is not possible for spherical
156: nanoparticles, which we discuss here. The same holds true for
157: saddle-like structures, and in general for all surfaces with non-zero
158: Gaussian curvature. On a microscopic level, the requirement for
159: non-zero Gaussian curvature also means that a flat sheet cannot be
160: bent into a sphere without introducing defects. Continuum elasticity
161: approaches to fullerenes therefore explicitly treated the 12
162: pentagonal defects necessary to close a sheet of carbon hexagons
163: \cite{n:ters92}. In order to avoid this complication, here we assume
164: that the curvature generating defects are distributed in a homogeneous
165: manner and result in a vanishing internal strain (like for a ping-pong
166: ball). In this case, the hollow nanospheres can be modeled as elastic
167: shells with thickness $h$ and prefered radius $R$.
168:
169: Mechanical collapse in tribological applications is likely to be
170: caused by either direct forces or high pressure. In the latter case,
171: an inward buckling instability occurs, since the pressure energy
172: scales with a higher power of indentation $H$ than the restoring
173: elastic energy ($H^2$ versus $H^{3/2}$) \cite{uss:schw00d}. Mechanical
174: collapse is expected when indentation $H$ becomes of the order of
175: radius $R$. The critical pressure for collapse can then be estimated
176: to be $p_c \sim G^{1/4} \kappa^{3/4} / R^{5/2}$. For buckyballs and
177: typical inorganic fullerenes, this yields values of 15 and 2 GPa,
178: respectively. For nanoparticles, the corresponding forces are in the
179: nN-range.
180:
181: \section*{Van der Waals interaction and phase behavior}
182:
183: Since vdW interactions are very difficult to treat on a molecular
184: level, a continuum approach is even more useful in this case. It
185: is well known from colloid science that vdW interactions strongly
186: depend on the geometry of the system under investigation.
187: Geometrical aspects of vdW interactions are well
188: treated in the Hamaker approach, in which an atomic Lennard-Jones
189: interaction is summed in a pairwise manner over the relevant geometry
190: and the molecular details of the interaction are lumped into the
191: effective Lennard-Jones parameters $\sigma$ and $\epsilon$. For
192: carbon, $\sigma = 3.5 \AA$ and $\epsilon = 4.6 \times 10^{-15}$ erg
193: \cite{n:giri92}. The corresponding Hamaker constant $A$
194: is in the order of $10^{-12}$ erg. For two buckyballs, the Hamaker
195: approach for thin sheets leads to the \textit{Girifalco potential}
196: \cite{n:giri92}, which is considerably shorter ranged than a
197: Lennard-Jones potential. Since a small range of attraction leads to
198: the disappearance of the gas-liquid coexistence from the phase diagram
199: \cite{n:gast83}, it has been suggested that buckyballs constitute the
200: first non-colloidal system for which the gas-liquid critical point
201: does not appear in the phase diagram. However, large scale
202: simulations have shown that buckyballs indeed do show a gas-liquid
203: coexistence, which extends roughly from 1900 to 2000$^{\circ}$C
204: \cite{n:cacc97}.
205:
206: For hollow nanoparticles in general, the effect of a finite thickness
207: $h$ becomes important. However, for nanoparticles in a crystal or
208: adhering to a flat substrate or to each other, the distance of closest
209: approach $D$ between interacting particles is on an atomic
210: scale. Under such conditions, the vdW interaction saturates on the
211: atomic scale $D$ with increasing thickness $h$. For example, the
212: surface energy for the interaction with a halfspace scales as
213: $\epsilon h / D^3$ and $\epsilon / D^2$ for thin films and halfspaces,
214: respectively, with the crossover occuring for $h = D$. Since $h
215: \gtrsim D$ in our case, the surface energy is $\epsilon / D^2$
216: independently of $h$ and no significant scaling with thickness is
217: expected.
218:
219: \begin{figure}[tb]
220: \begin{center}
221: \epsfig{figure=InteractionRange.eps}
222: \end{center}
223: \caption{Numerical results for the interaction range $\delta$
224: (a) for single-walled nanoparticles as a function of radius $R$ and
225: (b) for multi-walled nanoparticles as a function of radius $R$ and
226: thickness $h$. $\sigma = 3.5\ \AA$ is the atomic length of the
227: Lennard-Jones potential. The gas-liquid coexistence is expected to
228: disappear from the phase diagram for $\delta < 0.4$ (dashed lines).}
229: \label{InteractionRange}
230: \end{figure}
231:
232: For a more detailed investigation, one has to consider the general
233: case of two nanospheres of finite thickness interacting in a pairwise
234: manner by a Lennard-Jones potential \cite{uss:schw00e}. The
235: interaction energy can be written as $V = V_{R,R} + V_{R-h,R-h} - 2
236: V_{R,R-h}$, where $V_{R_1,R_2}$ is the interaction energy between two
237: filled balls of radii $R_1$ and $R_2$, respectively, which can be
238: calculated analytically for a Lennard-Jones potential. In order to
239: predict phase behavior, three quantities characterizing the resulting
240: potential are especially important: the hard core diameter $2 R$ sets
241: the density scale, the potential depth sets the temperature scale, and
242: the interaction range determines if a gas-liquid coexistence occurs
243: (compare Fig.~\ref{tem}). Our full analysis has shown that for hollow
244: nanoparticles of radius $R$ and thickness $h$, the potential depth and
245: interaction range scale linearly and inversely with $R$. Both
246: quantities depend only weakly on $h$. In
247: Fig.~\ref{InteractionRange} we show our numerical results for the
248: interaction range as a function of $R$ and $h$. Recent work for the
249: Double-Yukawa potential suggests that the gas-liquid coexistence
250: disappears for an interaction range smaller than $\delta = 0.4$
251: \cite{n:teje94}. Our analysis predicts that this happens for $R = 12
252: \AA$ and $R = 35 \AA$ for single- and multi-walled nanoparticles,
253: respectively. For larger $R$, the critical point will disappear
254: from the phase diagram.
255:
256: \section*{Deformation due to adhesion}
257:
258: We consider the case where a spherical hollow nanoparticle is indented
259: a distance $H$ due to adhesion to a flat substrate. The vdW energy
260: gained on adhesion can be estimated to be $E_{vdW} \sim (A / D^2)
261: D H$ where $A$ is the Hamaker constant and $D$ an atomic cutoff
262: \cite{comment}. The energy of deformation can be estimated to be
263: $E_{el} \sim G^{1/2} \kappa^{1/2} H^2 / R$. With increasing $R$,
264: deformation becomes more likely, since both the vdW energy and
265: deformability increase. However, deformability decreases rapidly with
266: increasing thickness $h$, since $G^{1/2} \kappa^{1/2} \sim
267: h^2$. Solving for the indentation $H$ and using the values given
268: above, it is found that for hollow carbon nanoparticles with few
269: walls, $H$ can be in the nm-range \cite{uss:schw00d}. However, for
270: typical inorganic fullerenes deformations are suppressed by the larger
271: shell thickness $h$.
272:
273: Finally we briefly discuss a nanotube adhering to a flat substrate.
274: As explained above, nanotubes can be bent without being stretched. On
275: deformation, one has to pay a bending energy per length of $E_{el} / L \sim
276: \kappa H^2 / R^3$. The gain in vdW energy scales as $E_{vdW} / L \sim
277: A H / D^{3/2} R^{1/2}$. It follows that for carbon nanotubes
278: with small thickness $h$, the indentation $H$ can be in the nm-range
279: for even smaller values of radius $R$ than for hollow carbon
280: nanoparticles.
281:
282: \section*{Acknowledgments}
283:
284: It is a pleasure to thank R. Tenne for many stimulating
285: discussions. USS was supported by the Minerva Foundation and SK by the
286: Ministry of Education, Science and Culture of Japan.
287:
288: \begin{thebibliography}{10}
289:
290: \bibitem{n:tenn95}
291: R.~{Tenne}, {\em Adv. Mater.}, \textbf{7}, 965 (1995) and
292: {\em Progress in Inorg. Chem}, \textbf{50}, 269 (2001).
293:
294: \bibitem{n:giri92}
295: L.~A. Girifalco, {\em J. Phys. Chem.}, \textbf{96}, 858 (1992).
296:
297: \bibitem{n:cacc97}
298: C.~Caccamo, D.~Costa, and A.~Fucile, {\em J. Chem. Phys.}, \textbf{106}, 255 (1997).
299: M.~Hasegawa and K.~Ohno, {\em J. Chem. Phys.}, \textbf{111}, 5955 (1999).
300:
301: \bibitem{n:ruof93}
302: R.~S. Ruoff, J.~Tersoff, D.~C. Lorents, S.~Subramoney, and B.~Chan,
303: {\em Nature}, \textbf{364}, 514 (1993).
304: T.~Hertel, R.~E. Walkup, and P.~Avouris,
305: {\em Phys. Rev. B}, \textbf{58}, 13870 (1998).
306:
307: \bibitem{n:rapo97}
308: L.~Rapoport, Y.~Bilik, Y.~Feldman, M.~Homyonfer, S.~R. Cohen, and R.~Tenne,
309: {\em Nature}, \textbf{387}, 791 (1997).
310: M.~Chhowalla and G.~A.~J. Amaratunga, {\em Nature}, \textbf{407}, 164 (2000).
311:
312: \bibitem{n:gola99}
313: Y.~Golan, C.~Drummond, M.~Homyonfer, Y.~Feldman, R.~Tenne, and
314: J.~Israelachvili, {\em Adv. Mat.}, \textbf{11}, 934 (1999).
315: Y.~Golan, C.~Drummond, J.~Israelachvili, and R.~Tenne,
316: {\em Wear}, \textbf{245}, 190 (2000).
317:
318: \bibitem{uss:schw00d}
319: U.~S. Schwarz, S.~Komura, and S.~A. Safran,
320: {\em Europhys. Lett.}, \textbf{50}, 762 (2000).
321:
322: \bibitem{uss:schw00e}
323: U.~S. Schwarz and S.~A. Safran, {\em Phys. Rev. E}, \textbf{62}, 6957 (2000).
324:
325: \bibitem{n:robe92}
326: D.~H. Robertson, D.~W. Brenner, and J.~W. Mintmire,
327: {\em Phys. Rev. B}, \textbf{45}, 12592 (1992).
328: J.~P. Lu, {\em Phys. Rev. Lett.}, \textbf{79}, 1297 (1997).
329:
330: \bibitem{n:hern98}
331: E.~Hernandez, C.~Goze, P.~Bernier, and A.~Rubio,
332: {\em Phys. Rev. Lett.}, \textbf{80}, 4502 (1998).
333: G.~Seifert, H.~Terrones, M.~Terrones, G.~Jungnickel, and T.~Frauenheim,
334: {\em Phys. Rev. Lett.}, \textbf{85}, 146 (2000).
335:
336: \bibitem{b:land70}
337: L.~D. Landau and E.~M. Lifshitz,
338: {\em Theory of elasticity} (Pergamon Press, 1970).
339:
340: \bibitem{n:srol94}
341: D.~J. {Srolovitz}, S.~A. {Safran}, and R.~{Tenne},
342: {\em Phys. Rev. E}, \textbf{49}, 5260 (1994).
343:
344: \bibitem{n:ters92}
345: J.~{Tersoff}, {\em Phys. Rev. B}, \textbf{46}, 15546 (1992).
346: T.~A. {Witten} and H.~{Li}, {\em Europhys. Lett.}, \textbf{23}, 51 (1993).
347:
348: \bibitem{n:gast83}
349: A.~P. Gast, C.~K. Hall, and W.~B. Russel,
350: {\em J. Coll. Inter. Sci.}, \textbf{96}, 251 (1983).
351:
352: \bibitem{n:teje94}
353: C.~F. Tejero, A.~Daanoun, H.~N.~W. Lekkerkerker, and M.~Baus,
354: {\em Phys. Rev. Lett.}, \textbf{73}, 752 (1994) and
355: {\em Phys. Rev. E}, \textbf{51}, 558 (1995).
356:
357: \bibitem{comment}
358: This estimate is valid for weak adhesion. In \cite{uss:schw00d}, we
359: used an estimate for strong adhesion, $E_{vdW} \sim (A / D^2) R H$.
360: Then the predicted deformation is larger by a factor of $R/D$, yet the
361: general conclusions are the same.
362:
363: \end{thebibliography}
364:
365:
366: \end{document}