Aqueous phase self-assembly of nanoscale p-n heterojunctions

Article abstract of DOI:10.1021/jp020024n

Methods, adapted from photographic microcrystal growth technology, are used to assemble organized ternary organo-inorganic, nanoscale heterostructures. The resulting ensemble consists of free-standing, oriented AgBr microcrystals, upon the ???111??? surfaces of which is self-assembled a monolayer of spectrally sensitizing dye, and upon the corners of the hexagonally shaped AgBr substrates are epitaxially grown nanoscale p-type CuSCN nodules. EPR spectroscopy and photophysical measurements are employed to show that the ensembles are capable of separating photogenerated geminate pairs. One of the remarkable features of this approach is that it utilizes the ultrafast kinetics of aqueous precipitation and, thus, allows the assembly of heterostructures at rates of 10¹⁰/s?·L, or greater.

Full text of DOI:10.1021/jp020024n

5346
J. Phys. Chem. B 2002, 106, 5346-5350
Aqueous Phase Self-Assembly of Nanoscale p-n Heterojunctions
Joseph F. Bringley,* Raymond S. Eachus, and Alfred P. Marchetti
Imaging Materials and Media, Research & DeVelopment, Eastman Kodak Company,
Rochester, New York 14650-24708
ReceiVed: January 9, 2002
Methods, adapted from photographic microcrystal growth technology, are used to assemble organized ternary
organo-inorganic, nanoscale heterostructures. The resulting ensemble consists of free-standing, oriented AgBr
microcrystals, upon the 〈111〉 surfaces of which is self-assembled a monolayer of spectrally sensitizing dye,
and upon the corners of the hexagonally shaped AgBr substrates are epitaxially grown nanoscale p-type
CuSCN nodules. EPR spectroscopy and photophysical measurements are employed to show that the ensembles
are capable of separating photogenerated geminate pairs. One of the remarkable features of this approach is
that it utilizes the ultrafast kinetics of aqueous precipitation and, thus, allows the assembly of heterostructures
10
at rates of 10 /s‚L, or greater.
Introduction
discrete, free-standing “substrate” for further growth of hetero-
geneous materials. Onto the surfaces of the hexagonally shaped
microcrystals is assembled a monolayer of spectrally sensitizing
cyanine dye, capable of photoinduced electron transfer into the
Nanoscale and monomolecular electronics hold great promise
because of the possibility of reaching microelectronic device
scales on the order of single molecules. However, a significant
challenge lies in creating methods for assembling or fashioning
nanoparticles, or molecules, into “materials” capable of being
fabricated into devices. Langmuir-Blodgett and self-assembly
methods have been used to create organized two-dimensional
1
6
AgBr microcrystals with efficiencies near 1.0. The dye is
known to J-aggregate (see ref 6) on the 〈111〉 surfaces of AgBr
microcrystals, effectively blocking further growth on the
surfaces, and directing any subsequent growth onto the corners
5
of the microcrystals. We demonstrate the site-directed, epitaxial
2
molecular monolayers on a substrate. However, discrete three-
growth of the p-type semiconductor, CuSCN, onto the templated
AgBr substrates. CuSCN is a wide band-gap p-type semicon-
ductor and it has been shown that photographic cyanine dyes
are capable of injecting holes into CuSCN with very high
dimensional assemblies of nanoparticles or molecules are
generally much more difficult to “fabricate” and far fewer
examples and methodologies have been reported. A further
challenge of nanoscale and monomolecular technology is
establishing methods for creating electrical contact between
individual, or small groups, of atoms or molecules. Herein, we
describe a methodology adapted from photographic microcrystal
growth technology, and show the directed, aqueous phase, self-
assembly of heterogeneous materials into organized, organic-
inorganic, nanoscale heterostructures. p-n heterojunctions thus
created are capable of geminate pair separation, and outside of
applications in the photographic process, may represent a natural
bridge between conventional solid-state methods of device
fabrication and the fabrication of organized nanoscale or
monomolecular arrays.
7
,8
efficiencies. We use EPR spectroscopy to show that holes
generated by band-gap irradiation of the AgBr substrate may
be trapped in the CuSCN heterogrowths, and the intrinsic
photographic properties of the AgBr substrate to show that the
electrons are trapped in the substrate. The ensemble is, thus,
capable of geminate pair separation. The aqueous precipitation
method of assembling the composite is remarkably facile,
capable of fabricating heterojunctions at a rate on the order of
1
0 ‚
1
0 /s L, or greater.
Experimental Section
Common commercial photographic film may contain as many
as 10 silver halide semiconductor particles per square centi-
To 0.50 mol of the AgBr microcrystals contained in 300 mL
of water and 20.0 g of deionized gelatin, at a residual Ag
concentration of pAg ) 8.2 (where the pAg ) -log [Ag ])
9
+
meter, a density that is vastly greater than that currently found
in the most advanced solid-state semiconductor devices. Silver
halides share many features with solid-state semiconductors,
and a constant temperature of 40 °C, was added 25.00 mL of
0.1 M KI while the pAg was simultaneously adjusted to ca. 7.4
with 0.05 M AgNO3. After the addition, 6.50 mL of 0.25 M
NaSCN was added followed by 0.617 mmol of the magenta
dye 5-chloro-2-(2-((5-phenyl-3-(3-sulfobutyl)-2(3H)-benzox-
azolylidene)methyl)-1-butenyl)-3-(3-sulfopropyl)benzoxazoli-
um. The contents were then allowed to stir at 40 °C for 20 min.
The epitaxial growth was accomplished by initially double-
jetting 4.0 mmol of AgNO3 (0.25 M at 5.0 mL/min for 3.2 min)
with 0.25 M NaSCN at 5.0 mL/min and then switching to 0.14
M CuSO4 (which contained 0.55 mol ascorbic acid/mol Cu)
for 32 min, while the flow of NaSCN was maintained. In total,
3,4
such as doping to manipulate electronic properties, epitaxial
5
6
growth of secondary phases, and dye-spectral sensitization.
Silver halide microcrystals may be produced in large quantities
with unprecedented size monodispersity and morphological
control. We have adapted methods used in silver halide
microcrystal growth to develop an approach to the self-assembly
of ternary organic-inorganic heterostructures. The approach is
depicted in Figure 1 and utilizes monodisperse, morphologically
controlled AgBr microcrystals. Each microcrystal represents a
-
*
E-mail: joseph.bringley@kodak.com. Fax: (716) 477-9674.
21 mmol of Cu and 25 mmol of SCN were added, while a
1
0.1021/jp020024n CCC: $22.00 © 2002 American Chemical Society
Published on Web 05/04/2002

Self-Assembly of Nanoscale p-n Heterojunctions
J. Phys. Chem. B, Vol. 106, No. 21, 2002 5347
Figure 1. Schematic illustration of the templating method used to prepare ternary organic-inorganic heterostructures.
constant residual silver ion concentration was maintained, pAg
)
7.5. The contents were then cooled to 15 °C and kept
refrigerated. All procedures were performed in the dark.
Electron paramagnetic resonance (EPR) measurements were
made with a Bruker ESP300e spectrometer fitted with a helium
gas flow cryostat of conventional design. Irradiations were
performed in situ using a 200 W Hg/Xe source, the output of
1
which passed through a /8 m monochromator and a combination
of long-pass and band-pass filters. EPR experiments were carried
out on dispersions that were evaporated to dryness under a
stream of dry N2.
Results and Discussion
Materials Assembly. The experimental approach for prepar-
5
ing the composite structures was adapted from Maskasky who
showed that mixed silver halides and pseudohalides could be
grown epitaxially upon AgBr. The substrate consists of oriented
AgBr microcrystals with average particle dimensions of 4.2 µm
×
0.072 µm (the equivalent circular diameter of the hexagonal
shaped disks and the disk thickness, respectively). The geo-
metrical standard deviation of the equivalent circular diameter
was 1.31 µm. The microcrystals have their 〈111〉 surfaces
oriented perpendicular to the thickness dimension. The micro-
crystal substrates are prepared for the growth of a secondary
Figure 2. SEM image of the AgBr microcrystals after the epitaxial
growth procedure. The CuSCN epitaxy appears as capsule-like growths
localized near the corners of the host grains, bar indicates 2.0 µm.
6
throughout the visible and near-infrared spectral region.5,6 Prior
to dye addition, a small amount of iodide ion is added to improve
the adsorption of the dye to the 〈111〉 substrate surface and also,
inorganic phase by the addition of an amount of cyanine dye
sufficient to form a monolayer on their surfaces. Several authors
have shown that various spectrally sensitizing dyes are strongly
adsorbed onto AgBr 〈111〉 surfaces.5 The monolayer of dye
thus formed serves two purposes. First, the J-aggregated dye is
capable of absorbing visible light and transferring an electron
into AgBr, thus “sensitizing” the AgBr (and CuSCN) to specific
spectral regions. Second, the J-aggregated dye essentially
strongly adsorbs to the 〈111〉 substrate surfaces and hence
prevents further growth on the substrate surfaces, directing any
subsequent growth onto the edges and corners of the AgBr
microcrystals. We chose the magenta-colored dye shown in
Figure 1 having an absorption in the green region of the visible
spectrum (490-580 nm) with an absorption maximum at 545
nm. There exist many such dyes capable of sensitizing
,6
5
to further aid the inhibition of growth on these surfaces. The
+
-
fast double jet addition of Cu and SCN ions then leads to
the growth of CuSCN nodules, most typically on the corners
of the grains. SEM images of the AgBr microcrystals after the
procedure are given in Figure 2, and indicate capsule-like
growths of CuSCN on the corners and edges of the AgBr
microcrystals. Growth at corners is preferred over growth at
edges, presumably because the corners contain a greater density
of uncoordinated ions and hence sticking coefficients are greater
there. We estimate the CuSCN nodule size to be on average
150 nm, which can be directly controlled by varying the amount
+
-
of Cu and SCN added. The microcrystal population, having

5348 J. Phys. Chem. B, Vol. 106, No. 21, 2002
Bringley et al.
Figure 3. Elemental EDAX map of a AgBr grain containing epitaxy prepared by the described method. The grain analyzed is shown in the top
left-hand corner of Figure 2a. The subsequent frames represent elemental maps of the image area as measured by examining the X-ray fluorescence
emerging from the image with (b) S, (c) Cu, (d) Br, and (e) Ag. The sulfur and copper mapping is largely limited to the epitaxial positions of the
grain, while the Ag and Br mapping essentially reflects only the host grain.
2
+
EPR signals typical of impurity Cu bound to gelatin (black
2
+
spectrum). A small amount of Cu impurity is nearly always
present in silver halide microcrystals prepared in gelatin. There
0
is also a weak line at 333 mT possibly from Agn centers also
9
commonly found in silver bromide microcystals.
Exposure to band-gap light (440 nm) at 8 K produces a
cyclotron resonance signal from conduction band electrons
1
0
(
indicated by the sloping baseline in the blue spectrum), and
3
-
the hole signal assigned to Br4 centers trapped at grain
11
surfaces. These signals are characteristic of a pure AgBr grain
with a low concentration of lattice dislocations and point defects.
Recombination destroys these intrinsic centers once the exciting
light is quenched. Exposure to 440 nm light at higher temper-
0
atures produces more Agn image centers and oxidized gelatin
radicals, the latter formed by hole transfer from the AgBr grain
surface to the peptizing polymer. The concentration of Cu2
impurity ions in gelatin is unaffected by irradiation between 8
K and RT (e.g., see red spectrum in Figure 4).
+
Figure 4. 9.3 GHz EPR spectra of the substrate AgBr emulsion
measured at 8 K. Black trace: background signal obtained before
exposure to light. Blue spectrum: signal obtained at 8 K during
exposure to 440 nm light. The background signal has been subtracted.
Red trace: signal obtained after exposure to 440 nm light at 160 K
EPR spectra of the emulsion containing the CuSCN epitaxy
and the magenta dye before and after exposure to 440 nm light
8
are given in Figure 5. CuSCN has a band gap of 3.6 eV, while
1
2
the band gap of AgBr is 2.68 eV, thus the 440 nm (2.82 eV)
exposure produces photogenerated geminate pairs in the sub-
strate AgBr microcrystals, but produces few, if any, electrons
or holes by direct light absorption in the CuSCN heterogrowths
or the J-aggregated dye. Prior to exposure, the sample shows a
2
+
showing no change in the signal assigned to Cu in gelatin.
all six corners with CuSCN epitaxy, is estimated at 30%. We
note that the capsule-like growths appear from SEM images to
be oriented crystalographically with respect to the host micro-
crystals. An elemental EDAX map of the epitaxy (see Figure
2+
moderately strong Cu spectrum (black spectrum). This signal
is quite well resolved and shows the characteristics expected
9
2
2
3), clearly show that the Cu and S are localized to the epitaxial
of a d complex with a dx -y ground state (i.e., g| > g⊥ g 2).
Three of the four parallel hyperfine features from 6 Cu
nuclides are observed on the low-field tail. The composition of
the ligand shell is not revealed in this spectrum and thus we
3/65
positions and that the Ag and Br maps are limited to the
substrate positions, clearly indicating growth of CuSCN upon
AgBr.
Photophysical Properties. An EPR spectrum of the substrate
AgBr microcrystals is shown in Figure 4. The sample was
examined prior to the adsorption of the magenta dye. Before
light exposure, evaporated samples of this dispersion show weak
2
+
cannot distinguish between Cu as an impurity in CuSCN or
gelatin. There is also a weak signal at 333 mT, as found in the
0
substrate spectrum and likely from Ag clusters or gelatin
radicals. The exposure of this dispersion to 440 nm light at 8 K

Self-Assembly of Nanoscale p-n Heterojunctions
J. Phys. Chem. B, Vol. 106, No. 21, 2002 5349
Figure 6. Photographic response (density versus exposure) curve for
the AgBr/CuSCN emulsion exposed by 365 nm light, each step
represents a 0.15 log E increase in exposure.
The potential utility of photographic heterojunctions was first
1
5
pointed out by Levy et al., who showed that when semicon-
ductor materials such as CdS, ZnO and PbO were physically
mixed with silver halide emulsion grains, spectral sensitization,
and increases in intrinsic photographic activity could be
Figure 5. 9.3 GHz EPR spectra measured at 8 K from the AgBr
emulsion containing site-directed CuSCN epitaxy. Black trace: before
light exposure. Red trace: after exposure to 440 nm light at 160 K,
with the low-field region expanded in the insert. Background spectrum
15
observed. The methods employed by Levy et al., however,
proved ultimately unreliable or impractical from a processing
standpoint. A more amenable approach requires direct aqueous
precipitation of a semiconductor material onto the silver
halide grains. The criteria for such heterogeneous growth upon
AgBr are, indeed, enormously limiting. The more obvious
criteria include aqueous compatibility, a solubility near that of
AgBr, structural similarity, and/or a commensurate lattice
parameter. In fact, perhaps 99% of all semiconductor ma-
terials can be eliminated based on these criteria alone. CuSCN
meets many of these requirements and it has recently been
shown that photographic cyanine dyes are capable of injecting
6
3/65
has been subtracted. The arrows indicate probable
Cu HF features.
produces no substantial change in the EPR spectrum. However,
irradiation at 160 K generates a new set of features in the
spectral region between 250 and 330 mT. The insert (red
spectrum) in Figure 5 shows the change caused by light (i.e.,
the background signal has been subtracted). The position of the
new EPR signal and the detection of poorly resolved HF features
in the low-field tail (marked by arrows in the insert) suggest an
2
+
7
assignment to Cu . We conclude that there is a substantial
generation of a new Cu2 center as the result of a thermally
activated hole transfer mechanism. In the course of this process,
a small population of holes is also transferred to the magenta
dye and to the gelatin peptizer. The resultant oxidized dye and
gelatin radicals produce new signals convoluted near g ) 2 in
Figure 5. Because the Cu2 spectrum is not that observed from
this ion bound to gelatin, we propose that the photosignals
between 260 and 320 mT in Figure 5 are from holes transferred
from AgBr to the CuSCN epitaxy.
holes into CuSCN with efficiencies near 0.90. CuSCN has a
+
pKsp ) 14.3, comparable to that of AgBr (pKsp ) 12.3), and is
soluble in thiosulfate allowing it to be fixed out of (i.e., removed
from) an imaging system (a basic requirement of color
photography).
The photographic response (density versus exposure) curve
for the AgBr/CuSCN emulsion coated onto a cellulose acetate
support in gelatin is given in Figure 6. The data indicate a rather
typical photographic response with a high D-min level likely
resulting from the ascorbic acid reducing agent used to produce
+
+
Photographic Properties. The photographic process involves
the photogeneration of electron-hole pairs within silver halide
particles, and the subsequent detection and amplification of long-
Cu in the preparation. Although the photographic response is
by no means comparable to that of state-of-the-art commercial
systems, the data do indicate that the method generates a viable
imaging system, and the occurrence of normal photographic
behavior indicates that conduction-band electrons generated by
actinic irradiation are effectively trapped in the silver halide
substrate. It is anticipated that considerable optimization of the
system would be required to realize the full potential of the
approach.
The EPR spectroscopy data show that holes generated by
band-gap irradiation of the AgBr substrate may be trapped in
the CuSCN heterogrowths, and the photographic properties of
the ensemble indicate that the electrons are trapped in the
substrate. These data, taken together, indicate that the ensemble
is capable of geminate pair separation and provide indirect
experimental evidence for the presence of p-n heterojunctions.
Because, to our knowledge, values for the electron affinities
and thermal work function of CuSCN are not yet known, it is
impossible to precisely discuss the energetics of the junction
interface. However, the experimental data herein suggests that,
1
2
lived trapped electron states, by developer chemistry. Belloni
13
et al. have recently brought attention to the fact that the salient
loss mechanism of this process results from the fast recombina-
tion of some fraction of photogenerated geminate pairs. They
further described an approach to minimizing recombination,
which involves doping the silver halide grains with hole
scavenger molecules, which upon trapping a hole, in turn,
decompose to produce a secondary photoelectron. A similar
approach has been described by Gould et al.14 The method
described herein represents a unique approach to enhancing the
yield of photoinduced electrons. The p-n heterojunctions
created are capable of geminate pair separation, and therefore
potential utilization of both the electron and hole as photo-
graphically active species. The heterojunctions are fabricated
in standard photographic emulsion processing equipment, at
rates compatible with silver halide precipitation itself, and in a
directed manner at targeted sites on the silver halide grains.

5350 J. Phys. Chem. B, Vol. 106, No. 21, 2002
Bringley et al.
since holes are transferred to the CuSCN heterogrowths, the
valence band of the CuSCN lies above that of AgBr, and would
further suggest that, while electron transfer from the adsorbed
dye molecules to AgBr occurs readily, electron transfer from
dye molecule to CuSCN is higly unlikely.
References and Notes
(1) Joachim, C.; Gimzewski, J. K.; Aviram, A. Nature 2000, 408, 541.
(2) Geddes, N. J.; Sambles, J. R.; Davis, D. J.; Parker, W. G.; Sandman,
D. J. Appl. Phys. Lett. 1990, 56, 1916.
(3) Eachus, R. S.; Olm, M. T. J. Soc. Photogr. Sci. Technol. Jpn. 1991,
5
4, 294.
4) Eachus, R. S.; Marchetti, A. P.; Muenter, A. A. Annu. ReV. Phys.
Chem. 1990, 50, 117.
(
Conclusions
(
5) Maskasky, J. E. J. Imaging. Sci. 1988, 32, 160.
We have used methods, adapted from photographic micro-
crystal growth technology, to assemble organized ternary
organo-inorganic, nanoscale heterostructures. EPR spectroscopy
and photophysical measurements are employed to show that the
ensembles are capable of separating photogenerated geminate
pairs. This approach may be applied, in photography, to enhance
the yield of photogenerated electrons, and in principle, might
allow both the electron and hole to be utilized as photographi-
cally active species. One of the remarkable features of this
approach is that it utilizes the ultrafast kinetics of aqueous
precipitation and thus, allows the assembly of heterostructures
at rates of 10 /s, or greater. In this manner, molecular mono-
layers are assembled upon nanoscale features that are “con-
structed” upon micron scale crystals. The ensembles are three-
dimensional and free-standing and may represent a natural link
between conventional methods of device fabrication and mo-
lecular approaches. The general approach depicted here may
be extended to include other electronic materials, based upon
the criteria outlined above, to assemble ternary organo-inorganic
systems in unprecedented ways.
(6) Sturmer, D. M.; Heseltine, D. W. In The Theory of the Photographic
Process, 4th ed.; James, T. H., Ed.; Eastman Kodak Co.: Rochester, NY,
977. (b) Tani, T.; Suzumoto, T.; Ohzeki, K. J. Phys. Chem. 1990, 94,
1298. J-aggregation refers to the ordering of dye molecules into a two-
dimensional array on a surface. The absorption spectrum is shifted to lower
energy versus the solution spectrum, indicating strong coupling between
dye molecules.
(7) O’Regan, B.; Schwartz, D. T. Chem Mater. 1995, 7, 1349.
(8) Fernando, C. A. N.; Kitagawa, A.; Suzuki, M.; Takahashi, K.;
Komura, T. Sol. Energy Mater. Sol. Cells 1994, 33, 301. (b) de Taccomi,
N. R.; Son, Y; Rajeshwar, K. J. Phys. Chem. 1993, 97, 1042.
1
(9) Stevens, A. D; Symons, M. C. R.; Eachus, R. S. Phys. Status Solidi
(B) 1989, 154, 483.
(10) Eachus, R. S.; Olm, M. T. Cryst. Latt. Def. Amorph. Mat. 1989,
10
1
8, 297.
11) Eachus, R. S.; Pawlik, T. D.; Baetzold, R. C. J. Phys.: Condens.
Matter 2000, 12, 8893.
12) Hamilton, J. F. AdV. Phys. 1988, 37, 359. (b) Marchetti, A. P.;
Eachus, R. S. AdV. Photophys. 1991, 17, 145.
13) Belloni, J.; Treguer, M.; Remita, H.; De Keyzer, R. Nature 1999,
02, 865.
14) Gould, I. R.; Lenhard, J. R.; Muenter, A. A.; Godleski, S. A.; Farid,
(
(
(
4
(
S. J. Am. Chem. Soc. 2000, 122, 11,934.
(15) Levy, B. Photogr. Sci. Eng. 1983, 27, 204.