Microwave synthesis of supported Au and Pd nanoparticle catalysts for CO oxidation

Article abstract of DOI:10.1021/jp0526849

We report the microwave synthesis and characterization of Au and Pd nanoparticle catalysts supported on CeO₂, CuO, and ZnO nanoparticles for CO oxidation. The results indicate that supported Au/CeO₂ catalysts exhibit excellent activity for low-temperature CO oxidation. The Pd/CeO₂ catalyst shows a uniform dispersion of Pd nanoparticles with a narrow size distribution within the ceria support. A remarkable enhancement of the catalytic activity is observed and directly correlated with the change in the morphology of the supported catalyst and the efficient dispersion of the active metal on the support achieved by using capping agents during the microwave synthesis. The significance of the current method lies mainly in its simplicity, flexibility, and the control of the different factors that determine the activity of the nanoparticle catalysts. ? 2005 American Chemical Society.

Full text of DOI:10.1021/jp0526849

17350
2005, 109, 17350-17355
Published on Web 08/25/2005
Microwave Synthesis of Supported Au and Pd Nanoparticle Catalysts for CO Oxidation
Garry Glaspell, Lindsay Fuoco, and M. Samy El-Shall*
Department of Chemistry, Virginia Commonwealth UniVersity, Richmond, Virginia 21284-2006
ReceiVed: May 22, 2005; In Final Form: July 28, 2005
We report the microwave synthesis and characterization of Au and Pd nanoparticle catalysts supported on
CeO2, CuO, and ZnO nanoparticles for CO oxidation. The results indicate that supported Au/CeO2 catalysts
exhibit excellent activity for low-temperature CO oxidation. The Pd/CeO2 catalyst shows a uniform dispersion
of Pd nanoparticles with a narrow size distribution within the ceria support. A remarkable enhancement of
the catalytic activity is observed and directly correlated with the change in the morphology of the supported
catalyst and the efficient dispersion of the active metal on the support achieved by using capping agents
during the microwave synthesis. The significance of the current method lies mainly in its simplicity, flexibility,
and the control of the different factors that determine the activity of the nanoparticle catalysts.
Nanophase metal and metal oxide catalysts, with controlled
particle size and shape, exhibit high surface area and densely
populated unsaturated surface coordination sites that can result
in significantly improved catalytic performance over conven-
tional catalysts.1-4 The large number of surface and edge atoms
provide active sites for catalyzing surface reactions. Research
in this area is motivated by the possibility of designing
nanostructured catalysts that possess novel catalytic properties
such as low-temperature activity, selectivity, stability, and
resistance to poisoning and degradation effects.1 Such catalysts
are essential for technological advances in environmental
protection, improving indoor air quality, and in chemical
synthesis and processing.
Among the current important environmental issues is the low-
temperature oxidation of carbon monoxide, since small exposure
(ppm) to this odorless, invisible gas can be lethal.5 Therefore,
there is a need to develop highly active CO oxidation catalysts
to remove even a small amount of CO from the local environ-
ment.
It has been demonstrated that nanoparticles of precious metals
such as Au, Pd, and Pt, when used as CO oxidation catalysts,
are not as susceptible to moisture and sulfur-containing com-
pounds which typically affect the performance of transition
metal oxide catalysts.6,7 Haruta and co-workers demonstrated
that the high surface area exhibited via Au nanoparticles makes
them particularly useful for the catalytic oxidation of CO to
CO2.8,9 The high activity of the Au catalysts is consistent with
the strong tendency of Au nanoparticles to efficiently adsorb
CO molecules. Surprisingly, the Au nanoparticles do not
strongly adsorb and activate oxygen molecules.10 Thus, it is now
well-accepted that the oxide support plays a key factor in the
activation of oxygen molecules during the CO oxidation.11-13
In this letter, we report a simple method to prepare Au and
Pd nanoparticle catalysts supported on CeO2, CuO, and ZnO
and compare their catalytic activities for CO oxidation. We also
demonstrate that the shape and morphology of the support
nanoparticles can have a significant effect on the activity of
the catalyst. The approach utilized in the present work is based
on microwave synthesis of nanoparticles from metal salts in
solutions. Microwave irradiation (MWI) has several advantages
over conventional methods, including short reaction time, small
particle size, narrow size distribution, and high purity.14-18
Synthesis of the nanoparticles of CeO2, CuO, or ZnO was
achieved by dissolving approximately 4 g of Ce(NO3)4, Zn-
(NO3)2, or Cu(NO3)2 (Alfa Aesar), respectively, in ethanol.
While stirring, 10 N NaOH (Alfa Aesar) was added dropwise
until the pH of the resulting solution was 10. The resulting
solution was then placed in a conventional microwave. The
microwave power was set to 33% of 650 W and operated in
30-s cycles (on for 10 s, off for 20 s) for 10 min. The resulting
powder was washed with distilled water and ethanol and left to
dry. M-doped oxide support nanoparticles (M ) Au or Pd) were
prepared as above, but with the addition of the appropriate
amounts of the metal salt (HAuCl4 or Pd(NO3)2) mixed with
the Ce(NO3)4, Zn(NO3)2, or Cu(NO3)2 solution to obtain the
desired dopant concentration (2%, 5%, or 10%). For capped
nanoparticles, the starting precursors were mixed with poly-
(ethylene glycol) (PEG, molecular weight ) 1450) or poly(N-
vinyl-2-pyrrolidone) (PVP, molecular weight ) 40 000) as a
protective polymer prior to microwaving.
The X-ray diffraction (XRD) patterns of the powder samples
were measured at room temperature with an X’Pert Philips
Materials Research Diffractometer, with CuKR radiation. The
samples were mounted on a silicon plate for X-ray measure-
ments. For the CO catalytic oxidation, the sample was placed
inside a Thermolyne 2100 programmable tube furnace reactor.
The sample temperature was measured by a thermocouple placed
near the sample. In a typical experiment, 4 wt % CO and 20 wt
% O2 in He was passed over the sample while the temperature
was ramped. The gas mixture was set to flow over the sample
at a rate of 100 cm3/min controlled via MKS digital flow meters.
The conversion of CO to CO2 was monitored using an infrared
gas analyzer (ACS, Automated Custom Systems, Inc.). All the
catalytic activities were measured (using 20 mg sample) after a
heat treatment of the catalyst at 300 °C in the reactant gas
mixture for 15 min in order to remove moisture and adsorbed
10.1021/jp0526849 CCC: $30.25 © 2005 American Chemical Society

Letters
J. Phys. Chem. B, Vol. 109, No. 37, 2005 17351
Figure 1. (A) TEM of CeO2 nanoparticles prepared via microwave irradiation (the scale bar is 50 nm). (B) XRD of the CeO2 nanoparticles as
prepared (top) and after the oxidation of CO (bottom). (C) Temperature dependence for the CO conversion over: (blue) bulk CeO2 (Aldrich,
particle size <5 µm); (green) CeO2 nanoparticles prepared by laser vaporization, particle size ∼20 nm; and (red) CeO2 nanoparticles prepared by
microwave irradiation, particle size ∼ 5 nm.
Figure 2. (A) TEM of 10% Au/CeO2 nanoparticles. The scale bar is 50 nm. (B) Temperature dependence for the CO conversion over 2% and 10%
Au/CeO2 coprecipitated and precipitated separately (PS).
impurities. Transmission electron microscopy (TEM) was carried
out using a Joel JEM-1230 electron microscope operated at 120
kV.
area of 79 m2 g-1 exhibits a 100% CO conversion at 450 °C.21
The higher catalytic activity of the sample prepared by MWI is
attributed to the small size of the nanoparticles. After the CO
oxidation reaction, the XRD still shows the typical CeO2 pattern
without any indication of a phase change, as shown in Figure
1B (bottom). However, the average particle size after the
catalysis test increases to ∼9 nm as determined via the peak
widths in the XRD pattern. It is interesting to note that repeated
catalysis tests after the first heat treatment result in reproducible
CO conversion curves. This indicates that the advantages of
the heat treatment resulting from removing moisture and
generating clean surfaces outweigh the disadvantage of increas-
ing particle size and subsequently decreasing surface area.
To examine the effect of the Au-ceria support interaction,
we prepared 2%, 5%, and 10% Au/CeO2. Figure 2A displays a
representative TEM image of the Au/CeO2 samples, specifically,
the 10% Au/CeO2 sample. The XRD patterns for 10% Au/CeO2
before and after catalysis (Supporting Information) indicate that
only CeO2 peaks are present with no indication of Au diffraction
peaks. This provides further support for the dispersion of the
small Au nanoparticles within the ceria support.
Figure 1A displays a typical TEM image for CeO2 nanopar-
ticles prepared via MWI. The individual particles have spherical
shapes with average diameters of 4-5 nm with a significant
degree of aggregation. The XRD pattern of the as-prepared
nanoparticles, shown in Figure 1B (top), matches well with CeO2
(ICCD 00-034-0394) from the database with no evidence of
impurity peaks. Volume-weighted average crystalline size
calculated from the XRD peak width using Scherrer’s equation
indicates that the average particle size of the ceria nanoparticles
is smaller than 5 nm.19
Figure 1C compares the catalytic activities of the ceria
nanoparticles prepared in our laboratory by MWI and by the
Laser Vaporization-Controlled Condensation (LVCC) method20
with the commercial ceria powder consisting of micron-size
particles (Aldrich, <5 µm). The CeO2 nanoparticles prepared
by MWI exhibit 50% and 97% conversions of CO to CO2 at
temperatures of 295 and 387 °C, respectively. For comparison,
a commercial ceria powder (Aldrich) with a specific surface

17352 J. Phys. Chem. B, Vol. 109, No. 37, 2005
Letters
TABLE 1: Temperatures for the Corresponding
Conversions of CO over CeO2, Au/CeO2, and Pd/CeO2
Nanoparticle Catalysts Synthesized by Microwave
Irradiation
ticles can significantly decrease the dispersion of the particles
within the support. Interestingly, in the SP samples, the 5% Au
is the most active sample, indicating that agglomeration of the
Au nanoparticles to a larger size is more effective if the Au
nanoparticles are present on the surface of the support. It is
also clear that the coprecipitated catalysts exhibit significantly
higher activities than the SP catalysts. For example, the best
SP sample (5% Au) is still not as good as the worst coprecipi-
tated sample (2% Au). This merely stresses the point that having
Au nanoparticles uniformly dispersed within the oxide support
is an important factor in determining the activity of the catalyst.
sample
T3% (°C)
T50% (°C)
% max (T °C)
CeO2
2% Au/CeO2
5% Au/CeO2
10% Au/CeO2
2% Au/CeO2 (SP)a
5% Au/CeO2 (SP)
10% Au/CeO2 (SP)
2% Pd/CeO2
5% Pd/CeO2
10% Pd/CeO2
170
45
<20
<20
120
36
49
137
70
295
129
67
97% (387)
94% (245)
93% (211)
97% (156)
94% (424)
98% (301)
91% (274)
100% (206)
100% (151)
100% (173)
59
239
132
172
190
140
141
Figure 3 compares the CO conversions over 10% Au
supported CeO2, CuO, or ZnO nanoparticle catalysts prepared
by the MWI method. It is clear that the ceria-supported catalyst
exhibits the highest activity, and this is most likely due to the
high oxygen storage capacity of ceria.22 In this case, oxygen is
assumed to adsorb on the ceria nanoparticles where it may or
may not dissociate before reacting with CO molecules adsorbed
on Au nanoparticles.21,23 We also note that the activity of the
10% Au/CeO2 catalyst prepared here by the MWI method is
higher than that prepared by the coprecipitation method, but
lower than similar catalysts prepared by deposition-precipitation
(DP) and solvated metal atom dispersion methods.21,24 This may
suggest that the presence of small Au nanoparticles, as provided
by the MWI method, is not the determining factor in achieving
the highest CO conversion.21 It appears that the presence of
ionic gold well-dispersed within ceria, as provided by the DP
method, is necessary to achieve the 100% room-temperature
CO conversion.21
85
a (SP): Separately precipitated.
To force the gold nanoparticles onto the surface of the
support, we modified the original synthesis procedure by
precipitating the support with NaOH before the Au solution was
added. The XRD patterns for the 10% sample (see Supporting
Information) show that Au peaks are present before and after
the catalysis reaction, indicating that the Au nanoparticles are
concentrated on the surface of the ceria particles. Figure 2B
compares the effect of the Au loading on the catalytic activity
of the coprecipitated and separately precipitated (SP) samples
containing 5% and 10% Au, and the overall results for the
catalysts supported on ceria are presented in Table 1. Increasing
the metal loading increases the catalytic activity as a result of
increasing the overlap between the metal and the support.
However, in the coprecipitated catalysts, the 10% sample shows
only slight improvement over the 5% sample, which may indi-
cate that this is the maximum load that can be achieved with
the MWI technique before the aggregation of the Au nanopar-
The TEM images of the Au/CuO and Au/ZnO catalysts,
shown in Figure 3, reveal different morphologies from that
observed in the Au/CeO2 nanoparticle catalyst (Figure 2). Of
Figure 3. Temperature dependence for the CO conversion 10% Au/CeO2, 10% Au/CuO, and 10% Au/ZnO.

Letters
J. Phys. Chem. B, Vol. 109, No. 37, 2005 17353
TABLE 2: Temperatures for the Corresponding
Conversions of CO over CuO, Au/CuO, and Pd/CuO
Nanoparticle Catalysts Synthesized by Microwave
Irradiation
followed by a slow growth process, which prevents particles
from growing during the time scale of the MW irradiation. It is
well-known that, when nucleation and growth can be separated,
particles with a narrow size distribution are obtained.25 There-
fore, the current results also indicate that the nucleation and
growth processes are very different in the Au/CeO2 and Pd/
CeO2 systems. This could be related to the difference between
interfacial free energies of the Au-CeO2 and Pd-CeO2
interfaces.
The high activities of the Au/CeO2 and Pd/CeO2 nanoparticle
catalysts prepared by MWI are attributed to the strong interaction
between Au or Pd and CeO2 and to the oxygen storage capacity
and redox properties of CeO2 nanoparticles.21-24 However, the
low activity of the catalysts supported on the ZnO is attributed
to the nanorod morphology, which appears to result in poor
dispersion and aggregation of the metal nanoparticles. To
stabilize the metal nanoparticles against agglomeration, the MW
synthesis was carried out in the presence of PEG or PVP as a
solvent for the starting materials. Figure 5A demonstrates that
the MW-synthesized ceria nanoparticles in the presence of PVP
have significantly smaller diameters and narrower size distribu-
tion than those prepared in the absence of PVP.
sample
T3% (°C)
T50% (°C)
% max(T °C)
CuO
235
130
<20
<20
94
75
63
335
177
112
135
160
131
124
91% (550)
97% (310)
100% (150)
100% (211)
98% (215)
98% (186)
97% (192)
2% Au/CuO
5% Au/CuO
10% Au/CuO
2% Pd/CuO
5% Pd/CuO
10% Pd/CuO
TABLE 3: Temperatures for the Corresponding
Conversions of CO over ZnO, Au/ZnO, and Pd/ZnO
Nanoparticle Catalysts Synthesized by Microwave
Irradiation
sample
T3% (°C)
T50% (°C)
% max(T °C)
ZnO
489
206
212
175
177
126
162
8% (535)
99% (420)
99% (418)
100% (331)
99% (223)
98% (200)
99% (194)
2% Au/ZnO
5% Au/ZnO
10% Au/ZnO
2% Pd/ZnO
5% Pd/ZnO
10% Pd/ZnO
305
305
213
210
194
189
Figure 5B compares the CO conversions over the Au/ZnO
catalyst prepared in the absence and presence of PVP. The TEM
images displayed in Figure 5B clearly demonstrate the disap-
pearance of the nanorod morphology of ZnO and the appearance
of well-dispersed Au nanoparticles in the catalyst sample
prepared in the presence of PVP. It is evident that PVP acts as
a stabilizing agent for the growth of the Au nanoparticles and
protects them from aggregation and increase in size. The
enhanced catalytic activity of the Au/ZnO/PVP catalyst observed
at lower temperatures is remarkable, since it directly correlates
with the change in the catalyst morphology.
Similar effects are observed by using PEG as a capping agent
during the MWI synthesis of the Pd/CeO2 and Pd/ZnO catalysts
as shown in the Figure 5C. Again, a significant increase in the
CO conversion is observed for the Pd/CeO2 catalyst prepared
in the presence of PEG, as shown in Figure 5C (left). Also, a
remarkable change in the morphology of the catalyst ac-
companied by enhanced activity is observed for the Pd/ZnO
catalyst prepared in the presence of PEG, as shown in Figure
5C (right). The CO conversion results for the catalysts prepared
in the presence of PVP or PEG are presented in Table 4.
special interest is the formation of long ZnO nanorods with
lengths of several hundred nanometers and diameters of 10-
15 nm in the Au/ZnO catalyst. It appears that the nanorod
morphology does not allow for a uniform dispersion of the Au
nanoparticles. This may explain the low activity of the Au/ZnO
catalyst as indicated by the high temperatures needed to achieve
high CO conversions. The CO conversion results over the
catalysts supported on CuO and ZnO nanoparticles are presented
in Tables 2 and 3, respectively.
Figure 4A displays a typical TEM image of the 10% Pd/
CeO2 nanoparticles. It is clear that the particles have a narrow
size distribution with no evidence for agglomeration (the average
particle diameter is 3-4 nm). XRD indicates that the Pd (similar
to the Au) remains dispersed within the CeO2 even after the
catalysis reaction. The CO conversions over the 10% Pd catalyst
supported on CeO2, CuO, and ZnO nanoparticles are shown in
Figure 4B, and the results for other catalysts with 2% and 5%
Pd loadings are reported in Tables 1-3.
The small particle size and the narrow size distribution of
the Pd/CeO2 nanoparticles may be related to rapid nucleation
Figure 4. (A) TEM of 10% Pd/CeO2 nanoparticles (the scale bar is 50 nm). (B) Temperature dependence for the CO conversion 10% Pd/CeO2,
10% Pd/CuO, and 10% Pd/ZnO.

17354 J. Phys. Chem. B, Vol. 109, No. 37, 2005
Letters
Figure 5. (A) TEM of CeO2 nanoparticles prepared in the presence of PVP (the scale bar is 50 nm). (B) Comparison of CO conversion over 10%
Au/ZnO catalyst prepared in the presence and absence of PVP. (C) (left) Comparison of CO conversion over 10% Pd/CeO2 catalyst prepared in the
presence and absence of PEG. (right) Comparison of CO conversion over 10% Pd/ZnO catalyst prepared in the presence and absence of PEG.
TABLE 4: Temperatures for the Corresponding
Because of the strong dependence of nucleation rate on
temperature and supersaturation, uniform heating results in a
narrow particle size distribution, which increases the activity
of the catalyst.
Conversions of CO over Nanoparticle Catalysts Synthesized
by Microwave Irradiation in the Presence of PVP or PEG
sample
T3% (°C)
T50% (°C)
% max(T °C)
Finally, the high activity and stability of the nanoparticle
catalysts prepared using the MWI method in the presence of
capping agents method are remarkable and imply that a variety
of efficient catalysts can be designed and tested using this
approach. The significance of the current method lies mainly
in its simplicity, flexibility, and the control of the different
factors that determine the activity of the nanoparticle catalysts.
The method allows the incorporation of one or more types of
active metals such as Pd, Pt, Au, and Cu, or bimetallic alloys
such CuAu, PdAu, and CuPd, as well as one or more type of
oxide supports. We are currently exploring the effects of MW
frequency, duration, and solvent polarity on the morphology,
particle size, and size distribution of the nanoparticle catalysts.
In conclusion, a simple method has been developed for the
synthesis of supported nanoparticle catalysts via microwave
irradiation. This method offers extremely short reaction times
and produces high purity and high yield of efficient nanoparticle
catalysts.
CeO2 (PVP)
CeO2 (PEG)
ZnO (PVP)
ZnO (PEG)
10% Au/CeO2 (PVP)
10% Au/CeO2 (PEG)
10% Au/ZnO (PVP)
10% Au/ZnO (PEG)
10% Pd/CeO2 (PVP)
10% Pd/CeO2 (PEG)
10% Pd/ZnO (PVP)
10% Pd/ZnO (PEG)
215
175
241
227
35
<20
<20
132
134
62
332
309
333
308
103
109
176
324
179
114
219
127
99% (407)
99% (333)
99% (377)
98% (348)
98% (267)
99% (353)
99% (252)
99% (387)
98% (254)
99% (138)
99% (223)
99% (136)
172
68
Because of the high dielectric constant of PEG, it is an
efficient absorber of MW radiation, and hence, rapid heating
occurs easily under MW irradiation. In this case, the fast heating
achieved by the presence of PEG accelerates the reduction of
the metal precursor and leads to the formation of supersaturated
metal solutions, which undergo rapid nucleation resulting in the
formation of metal nanoparticles. The higher the supersaturation,
the higher the nucleation rate and the smaller the critical sizes
of the nuclei necessary to overcome the barrier to nucleation.26
In addition, the rapid MW heating provides uniform temperature
and concentration conditions for the nucleation and growth.
Acknowledgment. Acknowledgment is made to the Donors
of the American Chemical Society Petroleum Research Fund
(PRF #41602-AEF) for financial support of this research.

Letters
Supporting Information Available: XRD data of the Au/
CeO2 nanoparticle catalysts before and after the CO catalysis
test. This material is available free of charge via the Internet at
http://pubs.acs.org.
J. Phys. Chem. B, Vol. 109, No. 37, 2005 17355
(12) Schubert, M. M.; Hackenberg, S.; van Veen, A. C.; Muhler, M.;
Plazak, V.; Behm, R. J. J. Catal. 2001, 197, 113.
(13) Xu, Y.; Mavrikakis, M. J. Phys. Chem. B 2003, 107, 238.
(14) Kleinwechter, H.; Janzen, C.; Knipping, J.; Wiggers, H.; Roth, P.
J. Mater. Sci. 2002, 37, 4349-4360.
(15) Zhu, J.; Palchik, O.; Chen, S.; Gedanken, A. J. Phys. Chem. B
2000, 104, 7344.
References and Notes
(16) Boxall, D.; Lukehart, C. Chem. Mater. 2001, 13, 806, 891.
(17) Gallis, K.; Landry C. AdV. Mater. 2001, 13(1), 23.
(18) Liang, J.; Deng, Z. X.; Jiang, X.; Li, F.; Li, Y. Inorg. Chem. 2002,
41, 3602.
(19) Klug, H. P.; Alexander, L. E. X-ray Diffraction Procedures for
Polycrystalline and Amorphous Materilas, 2nd ed.; Wiley: New York, 1974.
(20) El-Shall, M. S.; Abdelsayed, V.; Pithawalla, Y. B.; Alsharach, E.;
Deevi, S. C. J. Phys. Chem. B 2003, 107, 2882.
(21) Venezia, A. M.; Pantaleo, G.; Longo, A.; Di Carlo, G.; Casaletto,
M. P.; Liotta, F. L.; Deganello, G. J. Phys. Chem. B 2005, 109, 2821.
(22) Travarelli, A.; Leitenburg, C.; Dolcetti, G.; Boaro, M. Catal. Today
1999, 50, 353.
(1) Moser, W. R., Ed. AdVanced Catalysts and Nanostructured
Materials; Academic Press: New York, 1996.
(2) Valden, M.; Lai, X.; Goodman, D. W. Science 1998, 281, 1647-
1650.
(3) McCrea, K. R.; Parker, J. S.; Somorjai, G. A. J. Phys. Chem. B
2002, 106, 10854-10863.
(4) Ahmadi, I. S.; Wang, Z. L.; Green, T. C.; Henglein, A.; El-Sayed,
M. A. Science 1996, 272, 1924.
(5) Carbon Monoxide. EnVironmental Health Criteria; World Health
Organization: Geneva, 1999; p 213.
(6) Flytzani-Stephanopoulous, M. MRS Bull. 2001, 885.
(7) Sinfelt, J. H.; Via, H.; Lythe, F. W. J. Chem. Phys. 1980, 72,
4832.
(23) Frost, J. C. Nature (London) 1988, 18, 577.
(24) Stoeve, S. L.; Prasad, B. L. V.; Uma, S.; Stoimenov, P. K.;
Zaikovski, V.; Sorensen, C. M.; Klabunde, K. A. J. J. Phys. Chem. B 2003,
107, 7441.
(8) Haruta, M. Chem. Rec. 2003, 3, 75.
(9) Haruta, M.; Tsubota, S.; Kobayashi, T.; Kageyama, H.; Genet, M.
J.; Delmon, B. J. Catal. 1993, 144, 175.
(10) Liu, Z. P.; Hu, P.; Alavi, A. J. Am. Chem. Soc. 2002, 124, 14770.
(11) Haruta, M. Catal. Today 1997, 36, 153.
(25) Ramaswamy, V.; Haynes, T. E.; White, W.; Chan, W. J. M.;
Roorda, S.; Aziz, M. J. Nano Lett. 2005, 5, 373-377.
(26) Kashchiev, D. Nucleation: Basic Theory with Applications; But-
terworth Heinemann: Oxford, 2000.