Elucidating chemical and morphological changes in tetrachloroauric solutions induced by X-ray photochemical reaction

Article abstract of DOI:10.1021/jp7104852

Chemical and morphological changes induced by an X-ray photochemical reaction in tetrachloroauric solutions leading to Au³⁺-to-Au ⁰ reduction are monitored in real time by X-ray absorption spectroscopy and X-ray small angle scatterin

Full text of DOI:10.1021/jp7104852

4568
J. Phys. Chem. A 2008, 112, 4568–4572
Elucidating Chemical and Morphological Changes in Tetrachloroauric Solutions Induced by
X-ray Photochemical Reaction
Qing Ma,*,† Ralu Divan,‡ Derrick C. Mancini,‡ and Denis T. Keane†
DuPont-Northwestern-Dow CollaboratiVe Access Team (DND-CAT), Northwestern UniVersity Synchrotron
Research Center, 9700 South Cass AVenue, Argonne, IL 60439 and Center for Nanoscale Materials, Argonne
National Laboratory, 9700 South Cass AVenue, Argonne, IL 60439
ReceiVed: October 30, 2007; ReVised Manuscript ReceiVed: February 11, 2008
Chemical and morphological changes induced by an X-ray photochemical reaction in tetrachloroauric solutions
leading to Au3+-to-Au0 reduction are monitored in real time by X-ray absorption spectroscopy and X-ray
small angle scattering. Prior to metal precipitation, the intermediate state, also observed by other techniques,
is unambiguously determined for the first time to be the reduction of Au3+ to Au1+, whose kinetics is strictly
of the zeroth order. The morphological changes occur simultaneously in the solutions, that is, the gold complexes
rearrange and aggregate, as unequivocally observed by the correlated changes in the Au L3 emission and
small angle scattering intensities. The experimental evidence indicates that the eventual metal precipitation is
strongly influenced by the changing solution acidity under X-ray irradiation. Detailed local structure changes
are also described.
AuCl2 during thermal reduction.15 In studying linear AuCl2
-
-
Introduction
ions in CH3CN using optical spectroscopy, Koutek and Mason
assigned a weak band (parity forbidden d f s transitions) and
an intense band (parity allowed d f p transitions) to these ions.16
However, both the band positions and shapes depend on the
solvents.17,11 The bands, if existing, are overshadowed in the
One of the important aspects for tailoring nanostructure
formation in solutions is to understand the reaction mechanism,
including the details about the intermediates generated in the
reactions.1,2 Tetrachloroauric (AuCl4-) solutions are broadly
used to produce gold nanoparticles chemically,3,4 thermally,5
or photochemically6 for applications in efficient catalysts,
advanced electronics, and medical diagnoses. Despite numerous
studies of particle synthesis and production,1,2 few studies were
carried out to elucidate the evolution in the solutions prior to
particle formation. Knowledge of the details of intermediates
and the interaction between gold complexes prior to particle
formation is critical for controlling the sizes, shapes, and thus
the properties of the nanoparticles.1,2,7,8
-
presence of AuCl4 ions, and no attempts have been reported
to differentiate them. In addition, the questions have rarely been
tackled regarding what kinetics this intermediate state follows,
when produced, and what the morphological consequences are
as it forms in the solutions.
In this paper, we address these issues using X-ray absorption
spectroscopy (XAS) and small angle X-ray scattering (SAXS)
simultaneously, which allows the study of chemical and
morphological changes in the solutions.18,19 Chemical reactions
are induced by synchrotron X-ray photons. The kinetics of
chemical reaction in the solutions that leads to the Au3+-to-
Au0 reduction is followed in real time.
-
In his potentiometric titration experiments of AuCl4 solu-
tions, more than half a century ago, Bjerrum observed an
increase, instead of the normally expected decrease, in the
electrochemical potential at the onset of the titration process.9
This was confirmed by Lingane a decade later.10 This increase
-
-
Experimental Methods
was attributed to the reduction of AuCl4 to AuCl2 as an
intermediate state towards metal precipitation. In the UV- or
γ-ray-induced reduction experiments, initial irradiation does not
produce any particles detectable by the optical spectrometer,
X-ray experiments were carried out at the bending magnet
beamline (5-BM-D), operated by DND-CAT, at the Advanced
Photon Source. The beamline uses a Si(111) monochromator
for energy selection. The energy resolution at the Au L3 edge
(11918.7 eV) is ∼1.4 eV. The synchrotron storage ring was
operated in the “top-up” mode with the electron beam current
kept at ∼100 mA. The incident X-ray flux density at the sample
is estimated to be 8 × 109 s-1 mm-2. The X-ray beam size in
the experimental station is selected by two sets of Huber slits,
which is either 1 × 1 mm2 or 1 × 8 mm2. Therefore, the X-ray
deposit rate is 15 µW or 120 µW. The solutions (0.4-2 mM)
were prepared by dissolving hydrated tetrachloroauric acid
(HAuCl4 ·4H2O) in deionized water. The pH values are from
2.8 for the 2 mM solution to 3.5 for the 0.8 mM solution
(ORION-290A). A buffered HAuCl4 solution with pH ) 2.8 is
also prepared using 100 mL 0.1 M KC8H5O4 and 57.8 mL 0.1
M HCl for comparison.
-
while the optical band associated with AuCl4 decreases and
the solutions become discolored. Henglein has suggested that
this is due to a lower valence state being produced and reasoned
that it is likely to be Au1+ because it has a longer lifetime than
Au2+ 11 Gachard et al. used the Au1+ assumption in their
.
mechanism discussion in their γ-ray irradiation work.12
In an attempt to confirm this spectroscopically, two recent
Raman studies contradicted each other due to ambiguity in the
data interpretation.13,14 Subsequent X-ray absorption studies
supported one of these results that negated the existence of
* Corresponding author E-mail: q-ma@northwestern.edu. Phone: 630-
252-03229. Fax: 630-252-0226.
† DND-CAT.
‡ Argonne National Laboratory.
10.1021/jp7104852 CCC: $40.75
2008 American Chemical Society
Published on Web 04/24/2008

X-ray Induced Changes in Tetrachloroauric Solutions
J. Phys. Chem. A, Vol. 112, No. 20, 2008 4569
The incident X-ray intensity is measured by an ion chamber
(Oxford Danfysik). A 13-element Ge solid state detector
(Canberra) was used to collect the Au L3 fluorescence emission
for X-ray absorption near edge structure (XANES) and extended
X-ray absorption fine structure (EXAFS) spectra. The Fourier
transform of the EXAFS spectrum reveals the inter-atomic
distances and coordination numbers around Au atoms.18 The
detector has an energy resolution of ∼2% or ∼250 eV at the
Au L3 absorption edge, which is sufficient to collect the Au L3
fluorescence emission (at ∼9700 eV) with little background. A
MAR 165 (Mar USA) wide area detector was used for small
angle X-ray scattering. A Tungsten beam stop of 5 mm in
diameter is used to block the direct beam, from which the X-ray
induced current is measured through a SR570 amplifier to
determine the sample absorbance. The air scattering is mini-
mized by placing a vacuumed flight pipe between the sample
and the MAR 165 area detector. The scattering in the q range
from 0.005 to 0.2 Å-1 is measured. The solution samples are
held in Kapton-sealed polypropylene or Kapton-insulated and
-sealed Al cells. For XAS measurements, the incident beam size
is 1 × 8 mm2, for which ∼0.4 cc solutions are used in the 8 ×
15 × 3.5 cm3 cell. For simultaneous XAS and SAXS measure-
ments, the beam size is 1 × 1 mm2, for which ∼0.05 cc solutions
are used in the 6 mm-diameter and 2 mm-thick cell. For all
measurements, the sample is placed 45° both to the incident
X-ray beam and to the Ge detector, which are perpendicular to
each other. This introduces a negligible error in the SAXS angle
calculation.
Figure 1. Kinetics of X-ray induced Au reduction from solution
illustrated by the measured Au L3 fluorescence vs time. The inset shows
the XANES spectra as a function of X-ray exposure time for 2 mM
solutions as examples: (a) 0 min, ∆µ ) 0.079; (b) 220 min, ∆µ )
0.076; (c) 300 min, ∆µ ) 0.080; (d) 420 min, ∆µ ) 0.095; (e) 680
min, ∆µ ) 0.137; where ∆µ is the absorption edge step.
triple oxidation state (5d8) for halide compounds or Au2O3 and
is assigned to a 2p3/2 f 5d transition.23,24 The intensity and
position of this peak varies with the ion charge and the electron
affinity of the ligand (influenced by both initial and final state
effects).23,25 It is used as a benchmark for the Au3+ state here.
The absorption edge defined by the deflection point of the rising
step is found at ∼0.5 eV below that of the gold metal, which is
somewhat smaller than the value for Au2O3.24 As the X-ray dose
increases this peak attenuates and disappears, and a distinct
structure that is typical of the gold metal appears. The data are
thus compounded, at least with these two components.
To decompose the spectra, the “zero” dose spectrum is
combined either with an AuI or a gold metal spectrum to fit
the sample spectra using a linear combination analysis (LCA)
routine.26 No shifts in spectral energy are allowed because all
the spectra are energy referenced to that of gold metal that is
simultaneously measured. The spectra can be reasonably fitted
in both cases with similar weighting factors. However, analysis
of the fitting merits indicates that the fits that involve the AuI
spectrum are better. The merit is defined by eq 1,26
Results and Discussion
Under thermal or chemical treatment or light irradiation, the
-
AuCl4 aqueous solutions undergo reduction reaction and
generate nanoparticles.3–6 The cause of the reduction detailed
for γ-ray irradiation should readily apply to X-ray irradiation.20,21
In brief, high energy X-ray photons generate strong reducing
agents, such as hydrated electrons,22 in aqueous solutions that
undergo chemical reactions, such as Au3+ + 3ee-q f Au. The
gold deposits on the Kapton windows are the end products.
A. Chemical Changes in the Aqueous Solutions under
X-ray Irradiation. Figure 1 illustrates the X-ray-induced gold
reduction kinetics by the integrated Au L3 fluorescence emission
intensity for three HAuCl4- solutions (see Supporting Informa-
tion). Note that the time is plotted logarithmically to make room
for the inset. The time resolution is ∼5 min, which is adequate
for the slow kinetics shown here. The kinetics is essentially
very similar to γ-ray-induced gold reduction, reported by
Gachard et al. using the optical spectroscopy technique.12 The
zero amplitude in Figure 1 is obtained after subtracting out the
emission intensity at the time “zero”. There are three distinct
regions. The first region is before any significant increases in
the Au L3 emission intensity. The second region includes the
linear rise in the magnitude, and the third region is the plateau.
Unlike the optical data, however, the data in Figure 1 also carry
the information on the evolution prior to metal precipitation,
such as the initial drop in the emission intensity seen clearly
for the slower kinetics, as will be discussed later on. It is in
this region that the current understanding lacks concrete
experimental support. The inset in Figure 1 shows the XANES
spectra for the 2 mM solution collected at the times indicated.
Note that the absorption edge steps for these spectra can be
read in the figure caption and give exactly the same kinetic
pattern as the integrated fluorescence intensity.
(data - fit)2 ⁄
data2
(1)
∑
∑
i
i
i
i
which ranges from 3.87 × 10-3 to 7.2 × 10-3 for the 2.0 mM
data in the first region. Figure 2 shows the differential merits
obtained by subtracting the fitting merits for HAuCl4 and Au
from those for HAuCl4 and AuI, as model compounds,
respectively. The quick rises indicate that gold precipitation
starts to dominate the system at the onset of the second region,
whereas the general downward trends in the first region suggest
that the monovalent Au state is preferred. Poor data quality for
1.3 mM is simply due to lower concentration, and the results
are even worse for 0.8 mM (not shown). Fits using all three
compounds, that is, HAuCl4, Au, and AuI, in the LCA routine
were also carried out. The results are shown in Figure 3, where
the weights for three components are presented as a function
of time. It is seen that the percentage of Au does not rise up to
The XAS spectrum at time “zero” (curve a in Figure 1)
displays a strong, narrow peak that is characteristic of a gold

4570 J. Phys. Chem. A, Vol. 112, No. 20, 2008
Ma et al.
Figure 2. Linear combination simulation merit analysis results for two
solutions. The analyses use the HAuCl4 spectrum and either the AuI
or the Au spectrum. ∆Merit ) Merit(1) - Merit(2), (1) HAuCl4/AuI
and (2) HAuCl4/Au.
Figure 4. Difference spectra from the measurements at various
exposure times for the 2 mM solution are compared with the Au and
AuI spectra: (a) 140 min, (b) 220 min, (c) 300 min, and (d) 420 min.
The inset is the absorption edge shift versus time relative to the Au0 L3
edge, determined from the difference spectra and compared with the
Au L3 edges of AuI (cross) and HAuCl4 (cross and circle).
2). In the second region, the metallic XANES structure becomes
distinct, which marks the metal precipitation in the solutions
(see the inset in Figure 1).
B. Morphological Changes in the Aqueous Solutions
under X-ray Irradiation. The step towards metal precipitation,
however, may involve more than chemical change. Figure 5
displays in the top panel for the 2 mM solution, the Au L3 X-ray
emission increase ∆If, and small angle scattering intensity Isaxs
versus X-ray exposure time. In the experiments, a 30-second
SAXS exposure is executed before each 20 min Au L3 emission
measurement. The integrated data are presented. The inset
displays the SAXS profiles as a function of time. For a diluted,
spherical particulate system the scattering intensity is propor-
tional to the difference square of the electron densities of the
Figure 3. Fractions of three gold species as a function of time, obtained
by the LCA analysis using three components, HAuCl4, Au, and AuI.
any statistic significance in the first region. These results are
consistent with that described in Figure 2. Therefore, no direct
conversion of Au3+ to Au0 is observed; rather, a conversion to
the Au1+state is strongly indicated. From the two-component
analysis it is possible to obtain the difference spectra: ∆St ) St
- wSt)0, where w is the weight of the “zero” dose spectrum.
Note that only the w factor for HAlCl4 is used in the calculation.
Figure 4 shows ∆St for the 2 mM solution compared to the
spectra of AuI and Au. The inset plots the absorption edge
positions of these ∆St values, relative to that of gold metal,
versus time compared to those of HAuCl4 and AuI. Despite the
noise, the difference spectra in the first region remarkably
reproduce the features associated with the monovalent gold, in
particular, the edge feature (d10). Therefore it can be concluded
with great certainty that, prior to gold precipitation in these
solutions, the Au1+ state forms as an intermediate under X-ray
illumination. It increases to ∼12% at the juncture between the
first and second regions (see Figure 3), strictly following the
zero-order kinetics C - C0 ) -kt (k2mM ) 10-5 s-1), typical
for a photochemical reaction. This explicitly shows that the Au3+
complexes are reduced to the Au1+ ones as one of the
2
particulate and matrix: I(q) RδFe , where q ) 4π sin θ/λ. The
absolute intensity, Iabs (q) ) [I(q) - Im (q)e-µx]/[µxI0e-µ x],
m
where m denotes matrix, that is , H2O + Kapton in this case.
The system is essentially noninteractive. Unlike the emission
intensity, the scattering intensity increases monotonically, and
the increase rate starts to top off before the onset of the third
region. In the first region alone, the scattering intensity increased
by ∼37%, whereas the emission intensity remained unchanged
or decreased somewhat. These results unequivocally show that
under X-ray irradiation the gold complexes rearrange, likely
aggregate, resulting in the increased density contrast δF and,
thus, increased Isaxs, even before the occurrence of any metal
precipitations. The reason for the initial drop in ∆If could be
complex. The effect is obvious for the 2 mM solution due to
the slower kinetics (seen better in Figure 1). It may result from
some of the aggregates moving out of the beam in the first
region. However, this is inconsistent with what happens in the
second region, as will be discussed below, where the gold clearly
accumulates. Another likely cause may be the increased
intergold-complex shadowing of the exiting Au L3 emission
photons as they aggregate. This assertion seems to better
reconcile with the SAXS data. Of course, it is also possible for
-
-
intermediate steps and validates the reaction AuCl4 + eaq
f
-
AuCl2 + 2Cl-.
The changes in the first region are also quantifiable by the
changes in the local coordination, N, because the Fourier
transforms of the EXAFS spectra (not shown here) can be
3+
1+
3+
modeled well by N ) wNAu + (1 - w)NAu (NAu ) 4 and

X-ray Induced Changes in Tetrachloroauric Solutions
J. Phys. Chem. A, Vol. 112, No. 20, 2008 4571
Figure 6. Fourier transforms of the differential EXAFS spectra
measured for the 2.0 mM solution at (a) 300 min, (b) 420 min, and (c)
800 min, compared to the gold foil (d).
to the low concentrations. In fact, it will be misleading to use
the emission data to describe the reaction kinetics because they
mostly reflect the gold accumulation in the X-ray beam path.
The true kinetics can only be revealed by separating the reactant
and products. This was accomplished by the LCA analysis using
all three gold species, as mentioned above and presented in
Figure 5. Top panel: the circular- and q-integrated SAXS intensity
and the L3 emission intensity vs X-ray exposure time. The inset shows
the circular-integrated SAXS data. Bottom panel: a cartoon illustration
of the evolution of the solution under X-ray illumination and corre-
sponding mathematic description.
Figure 3 for the 2.0 mM solution. The reaction is a priori first
order for Au3+ and Au, that is , CAu ) C0e-kt and CAu
)
3+
C0(1 - e-kt) with k ≈ 4 × 10-5 sec-1 and that the fraction of
Au1+ decreases linearly from ∼12%. The intrinsic relation
between Au3+ and Au may imply that the reduction is
perpetuated by other mechanisms as well besides photochemical.
The pH measurements at the juncture between the first and
second regions indicate that the solution acidity departed from
the initial pH value of 2.8 towards 4 for the 2 mM solution.
This is consistent with the recent report by C-H Wang et al.
who used a polychromatic X-ray beam.28 Therefore, the
hydrolysis equilibrium has been changed under X-ray irradiation.
It is also noticed that bubbles slowly form near the solution top
as X-ray exposure proceeds. One of the reactions leading to
the pH change could be e-aq + H f H2 + OH-.22 The excessive
Cl- ions may exit as Cl2. There is an inverse linear relationship
between the onsets of the metal precipitation versus the pH
values (see Figure 1). For the given X-ray dosing rate, it takes
longer for gold to precipitate in a more acidic solution.
Experiments with a buffered 2 mM solution show that the
process in the first region is drastically slowed down. This
suggests that the solution pH conditions play an influential role
in the eventual metal precipitation. At pH g 4, the AuCl4- and/
or AuCl4-complexes are no longer stable. The experiments with
pH-adjusted HAuCl4 solutions (adding NaOH) show an im-
mediate metal precipitation under X-ray irradiation for pH >
7. The XANES spectra (not shown) bear the clear signature of
the substitution of Cl- by OH-,24 being a reductant itself. It is
also likely that hydrate electrons are more efficient in an alkaline
solution.22
both of these mechanisms to coexist. Therefore, in addition to
-
the reaction, Au3+ + 2eaq f Au1+, a process, such as nAuCl2
1-
1-
1-
f [AuCl2
]
and/or nAuCl4
f [AuCl41-]n, proceeds
n
simultaneously.
The Au L3 emission intensity is largely determined by the
mass. Therefore, the rise in the emission intensity in the second
region indicates an accumulation of gold atoms in the X-ray
beam path, once metal precipitation occurs. The rise in the X-ray
scattering intensity in this region is thus weighted by metal
precipitation and increased gold density. The top-off of the
scattering intensity before ∆If is likely due to the bulk texture
formation, and indeed, the gold deposition on the Kapton
window starts to be visible. The bottom panel in Figure 5 gives
a cartoon illustration of the evolution in the solution.
Figure 6 shows the Fourier transforms of the difference
EXAFS spectra, calculated in a way described above, as a
function of X-ray exposure time, which are compared to that
of the gold foil. It is seen that at 300 min the metallic feature
dominates the spectrum and increases as a function of time. At
800 min, it is essentially undistinguishable from that of bulk.
The bulk gold has a close-packed cubic structure, and each atom
is neighbored by 12 others. For small gold particles N is smaller
than 12. Assuming a globular shape for the produced gold
particles, it is possible to appreciate the particle sizes using η
27
Another possible influence is from the formed Au particles.
The previous works concluded that the reduction of Au3+ in
the presence of pre-formed particles is catalytic in nature, which
may lead to direct reduction of Au3+ to Au through, say, a
surface mediated mechanism.29 It is shown that the reduction
continues after irradiation.6,12
) N/Nbulk
,
which can be obtained from the data in Figure 6.
It is estimated that the particle size is ∼0.5 nm (η ) 0.55) at
300 min, ∼1.0 nm (η ) 0.83) at 420 min, and >8.0 nm (η )
1) at 800 min.
C. Gold Nanoparticle Generation Kinetics and Mecha-
nism. The second region shows a rate that is nearly independent
of the concentration: ∆If/∆t ≈ constant. This may indicate that
the diffusion is not a limiting factor for the process, likely due
The driving force for gold ions to migrate into the X-ray beam
path is not clear. Fick’s first law may play a role for AuCl4- to

4572 J. Phys. Chem. A, Vol. 112, No. 20, 2008
Ma et al.
migrate into the beam, because these anions are consumed by
reduction in the beam and a concentration gradient may exist.
We also speculate that under high energy X-ray irradiation a
charge imbalance in the system may be created because the
generated electrons are more mobile than the ions. With this
argument, the X-ray illuminated volume should be charged more
positively than the nonilluminated volume. The complexes such
as AuCl4- will be driven towards the illuminated area through
the Columbic interaction and will thus accumulate. The gold
deposition on the Kapton windows is also believed to be related
to the X-ray induced charge effect.30
of the integrated fluorescence intensity is also given. This
information is available free of charge via the Internet at http://
pubs.acs.org.
References and Notes
(1) Burda, C.; Chen, X.; Narayanan, R.; El-Sayed, M. A. Chem. ReV.
2005, 105, 1025.
(2) Nanoscale Materials in Chemistry, Klabunde, K. J., Ed.; John Wiley
& Sons: New York, 2001.
(3) Paclawski, K.; Fitzner, K. Metall. Mater. Trans. 2006, B37, 703.
(4) Skibsted, L. H. AdV. Inorg. Bioinor. Mech. 1986, 4, 137.
(5) Zhao, C.; Qu, S.; Zhu, C. J. Mater. Res. 2003, 18, 1710.
(6) (a) Eustis, S.; El-Sayed, M. A. J. Phys. Chem. 2006, B110, 14014.
(b) Eustis, S.; Hsu, H.-Y.; El-Sayed, M. A abid 2005, 109, 4811.
(7) Murphy, C. J.; Jana, N. R. AdV. Mater. 2002, 14, 80.
(8) Chen, S.; Wang, Z. L.; Ballato, J.; Foulger, S. H.; Crroll, D. L.
J. Am. Chem. Soc. 2003, 125, 16186.
Conclusions
The X-ray induced gold reduction process is essentially
similar to those induced by UV light or γ-rays. The combination
of X-ray absorption spectroscopy and small angle X-ray
scattering has provided unprecedented details about chemical
and morphological changes in the solutions under irradiation.
It is a powerful approach for studies of liquid or colloidal
systems in real time. Prior to metal precipitation, the intermedi-
ate state, also observed by other techniques,4,9–11 is unambigu-
ously determined for the first time to be the reduction of Au3+
to Au1+, whose kinetics is strictly of the zeroth order. The
morphological changes occur simultaneously in the solutions,
that is, the gold complexes rearrange and aggregate, as
unequivocally observed by the correlated changes in the Au L3
emission and small angle scattering intensities. The experimental
evidence indicates that the eventual metal precipitation is
strongly influenced by the changing solution acidity under X-ray
irradiation.
(9) Bjerrum, N. Bull. Soc. Chim. Belges 1948, 57, 432.
(10) (a) Lingane, J. J. J. Electroanal. Chem. 1958, 19, 394. (b) Lingane,
J. J abid. 1962, 4, 332.
(11) Henglein, A. Langmuir 1999, 15, 6738.
(12) Gachard, E.; Remita, H.; Khatouri, J.; Keita, B.; Nadjo, L.; Belloni,
J. New J. Chem. 1998, 22, 1257.
(13) Pan, P.; Wood, S. A. Geochim. Cosmochim. Acta 1991, 26, 671.
(14) Murphy, P. J.; Stevens, G.; LaGrange, M. S. Geochi. Cosmochim.
Acta 2000, 64, 479.
(15) Baily, E. H.; Scofield, P. F.; Mosselmans, J. F. Proceedings of the
7th Annual V. M. Goldschmidt Conference, Lunar and Planetary Institute,
Houston, TX, 1997; Vol. 921; p 2264.
(16) Koutek, M. E.; Mason, W. R. Inorg. Chem. 1980, 19, 648.
(17) Kunkely, H.; Vogler, A. Inorg. Chem. 1992, 31, 4539.
(18) Sayers, D. E.; Stern, E. A.; Lytle, F. W. Phys. ReV. Lett. 1971, 27,
1204.
(19) Guinier, A.; Fournet, G. Small-Angle Scattering of X-rays; Wiley
& Sons: New York, 1955.
(20) Henglein, A. Langmuir, 1998, 14, 7393.
(21) Ghosh-Mazumdar, A. S.; Hart, E. J. AdV. Chem. Ser. 1968, 81,
193.
(22) Hart, E. J. Science 1964, 146, 19.
(23) Choy, J.-H.; Kim, Y.-I. J. Phys. Chem. 2003, B107, 3348.
(24) Berrodier, I.; Farges, F.; Benedetti, M.; Winterer, M.; Brown, M. G.,
Jr.; Deveughèle, M. Geochim. Cosmochim. Acta 2004, 68 (14), 3019.
(25) Brown, N. M. D.; McMonagle, J. B.; Greaves, G. N. J. Chem. Soc.,
Faraday Trans. I 1984, 80, 589.
Acknowledgment. DND-CAT is supported by E.I. DuPont
de Nemours & Co., The Dow Chemical Company, and State
of Illinois funding to Northwestern University. The Advanced
Photon Source and The Center for Nanoscale Materials at
Argonne National Laboratory are supported by the U. S.
Department of Energy, Office of Science, Office of Basic Energy
Sciences, under Contract No. DE-AC02-06CH11357.
(26) Ravel, B. The Consortium for Advanced Radiation Sources;http://
cars9.uchicago.edu/∼ravel/software/, 2005.
(27) Greegor, R. B.; Lytle, F. W. J. Catal. 1980, 63, 476.
(28) Wang, C.-H.; Hau, T.-E.; Chien, C.-C.; et al. Mater. Chem. Phys.
2007, 106, 323.
(29) Henglein, A. Langmuir 1998, 14, 7393.
(30) Ma, Q.; Moldovan, N.; Mancini, D. C.; Rosenberg, R. A. Appl.
Phys. Lett. 2000, 76, 2015.
Supporting Information Available: Two figures are pro-
vided that show the raw XANES spectra and the normalized
XANES spectra, as a function of time. The detailed description
JP7104852