Shape Transformation and Surface Melting of Cubic and Tetrahedral Platinum Nanocrystals

Article abstract of DOI:10.1021/jp981594j

We report transmission electron microscopic studies of in-situ temperature-induced shape transformation and melting behavior of polymer-capped cubic and tetrahedral nanocrystals. Our results indicate that the surface-capping polymer is removed by annealing the specimen at temperatures between 180 and 250 deg C. The particle shapes show no change up to ca. 350 deg C. In the temperature range between 350 and 450 deg C, a small truncation occurs in the particle shapes but no major shape transformation is observed. The particle shapes experience a dramatic transformation into spherical-like shapes when the temperature is raised above ca. 500 deg C, where surface diffusion or surface premelting (softening) takes place. Above 600 deg C, surface melting becomes obvious leading to coalescence of the surfaces of neighboring nanocrystals and a decrease in the volume occupied by the assembled nanocrystals. The surface melting forms a liquid layer a few atomic layers deep around the still solid core of the nanocrystal. This temperature is much lower than the melting point of bulk metallic platinum (1769 deg C). The reduction in the melting temperature is discussed in terms of the surface tension of the solid-liquid interface (γ_SL). For an 8 nm diameter Pt nanocrystal, γ_SL is calculated to be 2.0 N m^-1 at 650 deg C, which is smaller than that of the bulk solid-vapor metal surface tension (γ_SV). This reduction is proposed to be due to the compensation of the increase in γ_SV of the nanocrystal by the wetting effect at the solid-liquid interface.

Full text of DOI:10.1021/jp981594j

© Copyright 1998 by the American Chemical Society
VOLUME 102, NUMBER 32, AUGUST 6, 1998
LETTERS
Shape Transformation and Surface Melting of Cubic and Tetrahedral Platinum
Nanocrystals
Zhong L. Wang*
School of Materials Science and Engineering, Georgia Institute of Technology, Atlanta, Georgia 30332-0245
Janet M. Petroski, Travis C. Green, and Mostafa A. El-Sayed*
Laser Dynamics Laboratory, School of Chemistry and Biochemistry, Georgia Institute of Technology,
Atlanta, Georgia 30332-0400
ReceiVed: March 24, 1998; In Final Form: June 2, 1998
We report transmission electron microscopic studies of in-situ temperature-induced shape transformation and
melting behavior of polymer-capped cubic and tetrahedral nanocrystals. Our results indicate that the surface-
capping polymer is removed by annealing the specimen at temperatures between 180 and 250 °C. The particle
shapes show no change up to ∼350 °C. In the temperature range between 350 and 450 °C, a small truncation
occurs in the particle shapes but no major shape transformation is observed. The particle shapes experience
a dramatic transformation into spherical-like shapes when the temperature is raised above ∼500 °C, where
surface diffusion or surface premelting (softening) takes place. Above 600 °C, surface melting becomes
obvious leading to coalescence of the surfaces of neighboring nanocrystals and a decrease in the volume
occupied by the assembled nanocrystals. The surface melting forms a liquid layer a few atomic layers deep
around the still solid core of the nanocrystal. This temperature is much lower than the melting point of bulk
metallic platinum (1769 °C). The reduction in the melting temperature is discussed in terms of the surface
tension of the solid-liquid interface (γSL). For an 8 nm diameter Pt nanocrystal, γSL is calculated to be 2.0
N m-1 at 650 °C, which is smaller than that of the bulk solid-vapor metal surface tension (γSV). This
reduction is proposed to be due to the compensation of the increase in γSV of the nanocrystal by the wetting
effect at the solid-liquid interface.
1. Introduction
catalytic activity dependent on the size3 but it can also depend
on the shape of the nanocrystals. However, for efficient
catalysis, the capping polymer needs to be removed. Further-
more, many catalytic reactions take place at elevated temper-
atures. Thus, future applications would require heating the
catalyzed reaction mixture. It is thus necessary to determine
the temperatures at which the capping polymer is removed and
the thermal stability of these nanocrystals with respect to shape
changes and melting. This is the focus of this study.
Platinum nanocrystals with a high percentage of cubic-,
tetrahedral-, or octahedral-like shapes have been synthesized
by our group using a colloidal method.1,2 In this synthesis, the
shapes of the Pt nanocrystals are controlled by changing the
ratio of the concentration of capping polymer (polyacrylate) to
that of K2PtCl4 being reduced by H2 at room temperature. This
has opened a new direction for catalysis. Not only is the
Several questions are raised in this present study: (1) How
high does the annealing temperature need to be to remove the
* Corresponding authors.
S1089-5647(98)01594-6 CCC: $15.00 © 1998 American Chemical Society
Published on Web 07/17/1998

6146 J. Phys. Chem. B, Vol. 102, No. 32, 1998
Letters
Figure 1. In-situ TEM images recorded from the same area of capped Pt nanocrystals at various specimen temperatures, showing the evaporation
of capping polymer, coalescing, and macroscopic melting of Pt nanocrystal agglomeration (see text). The temperature for the images was sensed
by a thermocouple attached to the thermal cup of the specimen holder.
capping polymer without changing the particle shape? (2) What
is the temperature at which the shape changes? (3) Is there a
temperature-induced shape transformation due to surface dif-
fusion? (4) What is the temperature at which the metallic crystal
surface begins to melt, i.e., the temperature at which the liquid
and solid metal coexist? (5) What is the value of the surface
tension at the solid-liquid interface at this temperature?
The answers to these questions regarding the thermal proper-
ties of Pt nanocrystals are obtained by in-situ transmission
electron microscopy (TEM). The shapes of the individual
nanocrystals as well as their agglomerations are examined as
the temperature increases from room temperature to ∼700 °C.
The observed changes suggest that the capping polymer is
removed between 180 and 250 °C. Significant truncation to
nearly spherical shapes occurs near 500 °C. This temperature
is just before surface melting, which occurs above 600 °C. The
nanocrystals are transformed into a spherical-like shape by
melting of the first few atomic layers of the surface. This gives
a particle with a liquid layer surrounding a solid core. The large
reduction in the melting temperature of the nanocrystals as
compared to the bulk platinum is shown to result from the
formation of the solid-liquid interface surface tension (γSL).
The γSL is calculated at 650 °C for 8 nm diameter nanocrystals
as 2.0 N m-1. This value is lower than that of the bulk solid-
vapor surface tension (γSV), which is given in the literature4 as
between 2.6 and 3.5 N m-1. This decrease is attributed to the
wetting of the solid core surface atoms with the liquid layer,
thus increasing the number of Pt-Pt bonds in comparison to
the atoms at the crystal surface. Therefore, the wetting of the
solid surface has a greater affect on the surface tension than
the curvature of the nanocrystals.
2. Experimental Section
Pt nanocrystals with an average diameter of 8 nm were
prepared by using the technique of Rampino and Nord5 and
Henglein et al.6 Briefly, Ar gas was bubbled through the aged
solution of K2PtCl4, which was adjusted to an initial pH of ∼7,
and reduced by flowing H2 gas. The starting ratio of K2PtCl4
to polyacrylate (MW 2100) was 1:1, with the initial concentra-

Letters
J. Phys. Chem. B, Vol. 102, No. 32, 1998 6147
Figure 2. A plot of the agglomeration width of the TEM images shown in Figure 1 as a function of temperature. The first drop is observed in (b)
at 180 °C, which is attributed to the capping polymer dissociation. Another small drop is observed in (g) at 610 °C, and a much larger decrease is
observed in (h) at 660 °C, which is attributed to the melting of the agglomeration of the Pt nanocrystals. (Inset) In-situ TEM images recorded (a)
at room temperature and (b) after annealing to 270 °C, showing the desorption of the capping polymer on the Pt nanocrystal.
tion of K2PtCl4 being 8 × 10-5 M. The TEM grids were
prepared for imaging by placing a small drop of the specimen
solution on an ultrathin amorphous carbon film supported by a
copper grid and allowing it to dry completely in air at ambient
temperature.
because both surfaces have very different crystallographic as
well as electronic structure.7
3. Results
Tetrahedral, cubic, and truncated octahedral are the most
frequently observed1 particle shapes for Pt nanocrystals. Tet-
rahedral and cubic particles were chosen to examine the shape-
induced transformation by temperature increase in the present
study. The following sections are divided according to the
observed changes as the temperature increases.
Specimens with mixed cubic and tetrahedral shapes were
chosen for the studies reported here in order to observe the shape
transformation of Pt nanocrystals with different shapes in the
same image. The in-situ TEM experiments were carried out at
200 kV using a Hitachi HF-2000 TEM, equipped with a field
emission source. The specimen was annealed from room
temperature to ∼700 °C using a Gatan TEM specimen heating
stage, and the image was recorded using an CCD camera with
a short exposure time (0.5 s) to minimize the effect introduced
by thermal drift. The column pressure was kept at < 5 × 10-8
Torr during the experiments.
3.1. Vaporization of the Capping Polymer on the Pt
Nanocrystals. To observe the loss of the capping polymer,
two different types of experiments were carried out. In the first
experiment, we examined regions in which the nanocrystals were
agglomerated (the images shown in Figure 1). In comparison
with the image recorded at 25 °C (Figure 1a), the width of the
agglomeration decreased when the temperature reached ∼180
°C (Figure 1b). The shrinkage of the agglomeration width was
measured by plotting this cross-sectional width vs temperature
(25 to 660 °C) and is given in Figure 2. The letters correspond
to the images in Figure 1. The widths were measured and
plotted as five separate trials to account for any error in the
measurements. A drop in the width of ∼3 nm is observed on
going from (a) room temperature to (b) 180 °C. The shrinkage
of the agglomeration is most likely due to the vaporization of
the capping polymer between the particles and the rearrangement
of Pt nanocrystals located on the top of the first layer adjacent
to the substrate. Another apparent decrease that is observed in
Figure 1 is the interparticle distance between the two individual
nanocrystals, which is indicated by the arrowhead. These results
suggest that the capping polymer evaporates at temperatures as
low as 180 °C, thus bringing the particles closer together.
The temperatures given here were sensed using a thermo-
couple attached to the furnace. The local temperature at the
specimen region where the TEM observation was performed
can be slightly different when the decay of heating and the
beam-heating effect in the TEM are taken into consideration.
Thus, an estimated uncertainty of (15 °C should be allowed.
This study mainly concentrates on cubic and tetrahedral
nanocrystals. A cubic particle is bounded by six large {100}
faces, although truncated {111} corners and {110} faces have
been seen.2b The arrangement of the atoms on the surfaces and
in the particle can be imaged in profile using TEM.
A
tetrahedral particle is enclosed by four {111} faces, and some
truncation at the corners is also possible. The cubic and
tetrahedral nanocrystals are ideal for studying the {100} and
{111} face-dependent catalytic properties of Pt, respectively,

6148 J. Phys. Chem. B, Vol. 102, No. 32, 1998
Letters
1. The truncation occurring from ∼200-500 °C does not seem
to have a great effect on the size of the agglomeration in Figure
1, as is evidenced in the plot of Figure 2.
The tetrahedral particles could still be identified when the
temperature reached 525 °C (Figure 3g), while the cubic
particles became spherical-like when the temperature reached
above 500 °C (Figure 3g,h). Therefore, tetrahedral particles
seem to be more stable than the cubic ones. A striking feature
observed in Figure 3 is that the tetrahedral shape of the T3
particle was preserved without truncation when the specimen
temperature was as high as 525 °C. This is possibly due to the
contact of the apexes of the T3 particle with the adjacent
particles.
The small circles in Figure 4 show the changes in a group of
TEM images illustrating the shape transformation of some
individual Pt nanoparticles with temperature. The particle
shapes showed significant truncation when the temperature
reached 350 °C, as is shown in the region enclosed by the small
circles. Further heating to 500 °C resulted in a dramatic change
in particle shape (Figure 4c). Also shown in this figure by the
small circles is the shrinking in the size of the individual
particles.
The observed changes in the individual nanoparticle sizes in
Figures 1, 3, and 4 could be induced by surface diffusion, which
would lead to a three-dimensional atomic rearrangement, or
possibly by surface sublimation. Atomic surface diffusion
leading to changes in the observed image projection is energeti-
cally the most favorable.
3.3. Coalescence and Melting of the Agglomerated Pt
Nanocrystals. After the removal of the polymer shown in
Figure 1, the agglomeration of the nanocrystals preserved nearly
the same configuration from 270 to 545 °C (Figures 1c-f and
Figure 2). This was because the coalescence and significant
surface melting had not occurred although surface diffusion was
taking place (of course surface diffusion could be a result of
surface premelting). Between 545 and 610 °C, a large change
in the width as well as the total volume of the sample of the
agglomerated nanocrystals is observed, as shown in both the
images of Figure 1f,g and the plot of Figure 2. This is a strong
indication of surface melting of this group of nanocrystals. This
is further shown in the group of nanocrystals within the large
circles of Figure 4. Some interparticle coalescing occurs at a
temperature of 500 °C. At 610 °C, islands of connected particles
are formed, which is indicated by the larger areas of red in the
image. The clusters at this temperature retain a crystalline core
as shown in the diffraction patterns from the same aggregate of
nanocrystals before and after being annealed in-situ from 25 to
650 °C in Figure 5. Thus, at the high-temperature range,
coexistence between the core solid and the liquid outside shell
is observed.
Figure 3. In-situ TEM images recorded from an area of the capped Pt
nanocrystals at various specimen temperatures, showing the shape
transformation and melting of individual Pt nanocrystals (see text).
To carefully observe the polymer dissociation on individual
nanocrystals, the second experiment was carried out in which
the Pt particles were dispersed on surfaces of large carbon
spheres supported by a holey carbon film. This also enabled
us to determine the temperature at which the surface-capping
polymer was completely removed. The particles on the surfaces
of the carbon spheres can be imaged in profile in TEM, so that
the passivating surface polymer can be seen by phase contrast
imaging without the influence of the substrate. The insert in
Figure 2 shows an as-prepared Pt particle at room temperature
before annealing (a), where a thin organic molecular layer with
a thickness ∼1-1.3 nm is seen. After the specimen was heated
to 270 °C, the surface-capping layer had been removed (b).
3.2. Shape Transformation of the Pt Nanocrystals. Figure
3 shows a series of TEM images recorded from the same region
when the specimen temperature was increased from 25 to 660
°C. A few cubic (C1 to C4) and tetrahedral (T1 to T5)
nanocrystals are labeled to trace their shape transformation
behavior. Most of the particle shapes showed no significant
change when the specimen temperature was below 300 °C
(Figures 3a-c). Truncated cubic (C1-C4) and tetrahedral (T1,
T2, T4, and T5) particles were formed when the temperature
was increased to 350 °C (Figure 3d). This is evident by the
disappearance of corners and edges of the particles. This is
also consistent with the individually labeled particles in Figure
4. Discussion and Conclusions
The following is a summary of the in-situ TEM experiments,
depicting the four distinct stages of nanocrystal shape transfor-
mation and surface melting. The as-synthesized Pt nanocrystals
are capped with polymer, which can be removed when the
specimen temperature is in the range of 180-250 °C. The
particle shape shows no truncation until the temperature reaches
∼350 °C. Above 500 °C, the particle shape is dramatically
changed to a spherical-like shape. The surface atoms can either
diffuse on the surface and between different nanocrystals, or
they can sublime. Energetically, surface diffusion requires far
less energy. Between 600 and 660 °C, surface melting occurs
at the first few atomic layers of the nanocrystals causing the

Letters
J. Phys. Chem. B, Vol. 102, No. 32, 1998 6149
Figure 4. In-situ TEM images recorded from a region of capped Pt nanocrystals at various specimen temperatures. The images are color-enhanced
to show the distinctive changes in particle shapes in the region encircled by a small circle and the melting for the group of nanocrystals encircled
with the large circles. The blue color represents the substrate background, red depicts the approximate highest projected platinum atomic density,
and yellow is the approximate lowest project platinum atomic density (see text).
Figure 5. Diffraction pattern of the Pt nanocrystal sample at 25 and 650 °C.
formation of a spherical-like shape and/or the coalescence with
close neighbors.
the cubic ones (T ) 500 °C). This may be because the {111}
surfaces are lower in energy than the {100}.6 This might also
have a kinetic origin as the tetrahedral geometry is more difficult
to transform into spheres (the shape with the least surface
tension) than the cubic shape. These studies also show that
macroscopic melting of agglomerated nanocrystals starts at 600
°C, which is much lower than the melting point of bulk Pt (1769
°C at 1 atm pressure).9
Our results provide answers to the questions raised in section
1. First, by dispersing Pt particles evenly on the surface of a
substrate without the presence of agglomerated regions, the
capping polymer can be removed by annealing the specimen at
Our observations of the changes occurring in the Pt nano-
crystals upon increasing temperature show that shape and
relative orientation of neighboring particles affect truncation and
surface melting. Nanoparticles that have their apexes stabilized
by interactions with other nanocrystals retain their shapes at
higher temperatures (see T3 in Figure 3). This interaction
reduces their surface tension, and the added stabilization
prevents them from undergoing surface diffusion. The higher
temperatures at which truncation is observed for tetrahedral
particles (T ) 525 °C) indicates that they are more stable than

6150 J. Phys. Chem. B, Vol. 102, No. 32, 1998
Letters
a temperature of 180-250 °C. This procedure is likely to
produce activated Pt particles with well-preserved particle
shapes, allowing for efficient catalytic activity. Second, the
particle shape shows almost no change if the specimen tem-
perature is lower than 350 °C. At a temperature between 350
and 450 °C, slight truncation takes place. Therefore, the
resulting Pt particles can still be useful for studying shape-
dependent properties if the reaction temperature is below ∼450
°C. The particle shape experiences a dramatic transformation
into a spherical-like shape when the temperature is higher than
∼500 °C, possibly due to surface diffusion and/or sublimation.
The carbon substrate has a large effect on the mobility of the
nanocrystals on the surface, largely prohibiting the coalescence
of the individual nanocrystals.
The issue of small clusters melting has been discussed
previously in the literature.8-26 Solid-liquid coexistence in
these clusters is discussed using the capillarity approximation
by Reiss, et al.8 and by Buffat and Borel.11 In our study, surface
melting occurs and a surface solid-liquid core coexistence starts
after the particle shape becomes near spherical and is best
observed as the surface of neighboring particles begins to
coalesce (e.g., the large circle in Figure 4) at 500 °C. At 610
°C, more surface melting has taken place as evidenced by the
presence of more atoms found between the particles than at
lower temperatures (500 °C). The core in these particles remains
crystalline as shown in the diffraction patterns at different
temperatures in Figure 5. In this figure, two diffraction patterns
were recorded from the same aggregate of nanocrystals before
and after being annealed in-situ from 25 to 650 °C. The
intensities of the high index reflections (at high scattering angles)
drop dramatically. This is due to the increased Debye-Waller
factor from the atomic vibrations and the disordering of the
atoms in the nanocrystals. Also, there are sharp spots in the
pattern recorded at 650 °C, suggesting the formation of larger-
size crystals due to the combination of nanocrystals. Finally,
the total diffraction intensity at 650 °C is significantly lower
than that at 25 °C, indicating the decrease in crystallinity of
the nanocrystals. All of these facts support that the nanocrystals
are experiencing a surface melting process at 650 °C.
nanocrystals in these samples average 4 nm in radius, δ is equal
to 1 nm, corresponding to the first few atomic layers of the
spherical surface. From this equation, γSL is estimated to be
2.0 N m-1. As the temperature continues to increase, δ also
increases (and r decreases), and γSL becomes 0 as the entire
particle turns into liquid and the surface interface disappears.
Owing to the increase in the curvature of the nanoparticle,
the solid-vapor surface tension, γSV, would normally increase
when compared to that of the bulk4 γSV, which is between 2.6
and 3.5 N m-1
. However, the calculation is not for the
outermost surface, but that of the interface between the two
phases present. Therefore, the decrease of the γSL can be
attributed to the increase in the total number of bonds formed
by the solid surface atoms (in comparison to the atoms on the
surface) owing to the wetting with the surrounding liquid layer.
This wetting, and subsequent bond formation, more than
compensates for the increase of γSV due to the large curvature
of the nanoparticle.
Finally, two points need to be mentioned. First, the macro-
scopic values were used in eq 1 for the heat of fusion and the
density. Upon heating the nanoparticle, the density and the heat
of fusion are expected to decrease. The change in these terms
would further decrease the value of γSL from that calculated
above. Second, as was shown recently by Petroski et al.,28 the
particle shape at room temperature is determined by the kinetics
of their formation (and not by the thermodynamics). Above
200 °C, each shape strives to be as near spherical in shape as
possible in order to decrease the surface tension and attain the
most thermodynamically stable configuration. This can be
accomplished initially by surface diffusion and finally by surface
melting.
Acknowledgment. This work was supported by the National
Science Foundation Grants CHE-9727633 and DMR-9733160.
References and Notes
(1) Ahmadi, T.; Wang, Z. L.; Green, T. C.; Henglein, A.; El-Sayed,
M. A. Science 1996, 272, 1924.
It is clear that surface melting at 650 °C occurs at much lower
temperatures than the melting point of the bulk (1769 °C).9
According to Buffat and Borel,11 the ratio of the melting
temperature of the cluster (T) with a radius r to that of the bulk
(T0) is given by
(2) (a) Ahmadi, T.; Wang, Z. L.; Henglein, A.; El-Sayed, M. A. Chem.
Mat. 1996, 8, 1161. (b) Wang, Z. L.; Ahmadi, T.; El-Sayed, M. A. Surf.
Sci. 1997, 380, 302.
(3) (a) Clint, J. H.; Collins, I. R.; Williams, J. A.; Robinson, B. H.;
Towey, T. F.; Cajean, P.; Khan-Lodhi, A. Faraday Discuss. Chem. Soc.
1993, 95, 219. (b) Freund, P. L.; Spiro, M. J. Phys. Chem. 1985, 89, 1074.
(4) Solliard, C.; Flueli, M. Surf. Sci, 1985, 156, 487.
(5) Rampino, L. D.; Nord, F. F. J. Am. Chem. Soc. 1942, 63, 2745.
(6) Henglein, A.; Ershov, B. G.; Malow, M. J. Phys. Chem. 1995, 99,
14129.
(7) (a) Follstaedt, D. M. Appl. Phys. Lett., 1993, 62, 1116. (b)
Eaglesham, D. J.; White, A. E.; Feldman, L. C.; Moriya, N.; Jacobson, D.
C. Phys. ReV. Lett. 1993, 70, 1643.
(8) Reiss, H.; Mirabel, P.; Whetten, R. L. J. Phys. Chem. 1988, 92,
7241.
(9) Handbook of Chemistry and Physics, 52nd ed.; The Chemical
Rubber Co.: Boca Raton, FL, 1971.
(10) Reiss, H.; Wilson, I. B. J. Colloid Sci. 1948, 3, 551.
(11) Buffat, P.; Borel, J. P. Phys. ReV. A 1976, 13, 2287.
(12) Goldstein, A. N.; Echer, C. M.; Alivasatos, A. P. Science 1992,
256, 1425.
γSL
FS
FL
T
T0
2
) -
+ γL 1 -
(1)
(
)
]
[
FSλ r - δ
where γSL is the solid-liquid surface tension and γL is the
surface tension of the liquid, δ is the liquid layer thickness, FS
and FL are the densities of the bulk solid and liquid, respectively,
and λ is the heat of fusion. The equation derived by Reiss et
al.8 is essentially the same, but the liquid layer thickness term
does not appear. Since the difference in the liquid and solid
densities27 is less than 2%, the last term in eq 1 is neglected.
Using this approximation, one gets an expression for γSL from
eq 1 to be
(13) Peppiat, S. J.; Sambles, J. R. Proc. R. Soc. London, Ser. 1975, A345,
387.
(14) Hahn, M. Y.; Whetten, R. L. Phys. ReV. Lett. 1988, 61, 1190.
(15) Sheng, P.; et al. J. Phys. C 1981, 14, L565.
(16) Solliard, C.; Surf. Sci 1981, 106, 58.
(17) Coombes, C. J. J. Phys. F 1972, 2, 441.
(18) Pawlow, P. Z. Phys. 1909, 66, 549.
(19) Brian, C. L.; Burton, J. J. J. Chem. Phys., 1975, 63, 2045.
(20) Berry, R. S.; Jellinek, J.; Natanson, G. Phys. ReV. A, 1984, 30,
919.
FSλ
2
T
T0
γSL ) 1 - r - δ
(2)
(
)
]
[
As the temperature increases, the liquid layer begins to form
on the surface of the nanocrystal. At 650 °C, the liquid layer
thickness is estimated to be approximately 25%. Since the Pt

Letters
J. Phys. Chem. B, Vol. 102, No. 32, 1998 6151
(21) Kristensen, W. D.; Jensen, E. J.; Cotterill, R. M. J. Chem. Phys.
1974, 11, 4161.
(26) Cleveland, C. L.; Landmann, U.; Luedtke, W. D. J. Phys. Chem.
1994, 98, 6272.
(22) Honeycut, J. D.; Andersen, H. C. J. Chem. Phys. 1987, 91, 4950.
(23) Hanzen, K. J. Z. Phys. 1960, 157, 523.
(24) Hasegawa, M.; Hoshino, K.; Watabe, M. J. Phys. 1980, 10, 619.
(25) Wronski, C. R. J. Appl. Phys. 1967, 18, 1731.
(27) Zinov’ev, V. E.; Handbook of Thermophysical Properties of Metals
at High Temperatures; (Nova Science Publishers; New York, 1996; p 345.
(28) Petroski, J. M.; Wang, Z. L.; Green, T. C.; M. A. El-Sayed, J.
Phys. Chem. B 1998, 102, 3316.