Strong electronic coupling in two-dimensional assemblies of colloidal PbSe quantum dots.

Abstract

Thin films of colloidal PbSe quantum dots can exhibit very high carrier mobilities when the surface ligands are removed or replaced by small molecules, such as hydrazine. Charge transport in such films is governed by the electronic exchange coupling energy (beta) between quantum dots. Here we show that two-dimensional quantum dot arrays assembled on a surface provide a powerful system for studying this electronic coupling. We combine optical spectroscopy with atomic force microscopy to examine the chemical, structural, and electronic changes that occur when a submonolayer of PbSe QDs is exposed to hydrazine. We find that this treatment leads to strong and tunable electronic coupling, with the beta value as large as 13 meV, which is 1 order of magnitude greater than that previously achieved in 3D QD solids with the same chemical treatment. We attribute this much enhanced electronic coupling to reduced geometric frustration in 2D films. The strongly coupled quantum dot assemblies serve as both charge and energy sinks. The existence of such coupling has serious implications for electronic devices, such as photovoltaic cells, that utilize quantum dots.

Full text

Page 1

Strong Electronic Coupling in Two-
Dimensional Assemblies of Colloidal
PbSe Quantum Dots
Kenrick J. Williams,†William A. Tisdale,‡Kurtis S. Leschkies,‡Greg Haugstad,§David J. Norris,‡Eray S. Aydil,‡
and X.-Y. Zhu†,*
†DepartmentofChemistry,UniversityofMinnesota,207PleasantStreetSE,Minneapolis,Minnesota55455,‡DepartmentofChemicalEngineeringandMaterialsScience,
UniversityofMinnesota,421WashingtonAvenueSE,Minneapolis,Minnesota55455,and§InstituteofTechnologyCharacterizationFacility,UniversityofMinnesota,100
UnionStreetSE,Minneapolis,Minnesota55455
A
ties.1Such materials provide a unique
model system for the study of fundamental
physical processes, such as charge carrier
transport, exciton diffusion, and energy
transfer. To fabricate these solids, colloidal
QDs are particularly attractive because their
size and surface chemistry can be easily
controlled.2Moreover, they can be pro-
cessed from solution to form high quality
ordered assemblies known as
superlattices.3,4Consequently, a number of
studies have explored charge transport in
thin films of colloidal QDs.5?10However,
the presence of long-chain surface ligands,
which are required for the synthesis, colloi-
dal stability, and surface passivation of the
QDs, leads to poor conductivity in the films
due to weak inter-QD electronic coupling.
Therefore, much recent effort has focused
rrays of semiconductor quantum
dots (QDs) can act as artificial sol-
ids with tunable electronic proper-
on increasing the electronic coupling in
such films to improve their conductivity.
A number of studies have explored the
removal or exchange of capping molecules
to increase inter-QD electronic coupling. Ex-
amples include the drying of quantum dot
thin films and the partial evaporation of
weakly bound capping molecules,11,12the
chemical removal of capping molecules via
reduction13or oxidation,14the thermal de-
sorption/decomposition of capping mol-
ecules under vacuum or inert conditions,15
and the exchange of bulky or long ligands
by short ones.16,17A significant increase in
the conductivity was demonstrated when
the surface ligands were replaced by
shorter molecules and charge carriers were
electrochemically injected into the QDs.16
Later it was shown that, even without the
injection of extra carriers, the conductivity
of films of PbSe QDs could be increased by
as much as 10 orders of magnitude if films
were exposed to a solution of hydrazine.17
This dramatic increase in conductivity has
been attributed to the partial removal of
oleic acid ligands from the surface of the
PbSe nanocrystals, resulting in decreased
inter-QD spacing and increased inter-QD
electronic exchange coupling. This was sup-
ported by the observation that the first ex-
citon absorption peak shifted red by ?1?
20 meV after immersion in 1 M hydrazine in
acetonitrile. Subsequent work has exam-
ined the structural, optical, and electronic
changes in PbSe QD thin films that are ex-
posed to a variety of thermal and chemical
treatments in more detail.18,19The ob-
served variations included red shifts in the
first exciton absorption peak of 12 and 27
meV following treatment with 1 M hydra-
zine in acetonitrile and in ethanol, respec-
*Address correspondence to
zhu@umn.edu.
Received for review February 22, 2009
and accepted May 13, 2009.
Published online May 20, 2009.
10.1021/nn9001819 CCC: $40.75
© 2009 American Chemical Society
ABSTRACT ThinfilmsofcolloidalPbSequantumdotscanexhibitveryhighcarriermobilitieswhenthe
surfaceligandsareremovedorreplacedbysmallmolecules,suchashydrazine.Chargetransportinsuchfilmsis
governed by the electronic exchange coupling energy (?) between quantum dots. Here we show that two-
dimensionalquantumdotarraysassembledonasurfaceprovideapowerfulsystemforstudyingthiselectronic
coupling.Wecombineopticalspectroscopywithatomicforcemicroscopytoexaminethechemical,structural,and
electronic changes that occur when a submonolayer of PbSe QDs is exposed to hydrazine. We find that this
treatmentleadstostrongandtunableelectroniccoupling,withthe?valueaslargeas13meV,whichis1order
of magnitude greater than that previously achieved in 3D QD solids with the same chemical treatment. We
attributethismuchenhancedelectroniccouplingtoreducedgeometricfrustrationin2Dfilms.Thestrongly
coupledquantumdotassembliesserveasbothchargeandenergysinks.Theexistenceofsuchcouplinghasserious
implicationsforelectronicdevices,suchasphotovoltaiccells,thatutilizequantumdots.
KEYWORDS: nanocrystal · PbSe · quantum dot solid · two-dimensional · electronic
coupling
ARTICLE
VOL. 3 ▪ NO. 6 ▪ WILLIAMS ET AL.www.acsnano.org
1532
Downloaded by UNIV OF MINNESOTA on August 19, 2009
Published on May 20, 2009 on http://pubs.acs.org | doi: 10.1021/nn9001819

Page 2

tively. After these hydrazine treatments, it was
also verified that the size of the PbSe QDs did not
change, but the inter-QD distance decreased.
Thus, these results support the initial claim that
the red shift in the exciton transition occurs due
to a decrease in the inter-QD spacing. This can
cause (i) an increase in the average dielectric con-
stant of the film, (ii) an increase in the inter-QD ra-
diative coupling, and (iii) an increase in the
inter-QD electronic coupling. However, the actual
extent of the ligand exchange during the hydra-
zine exposure and how this impacts the accompa-
nying optical, electronic, and structural changes
is a complicated issue. Under the chemical condi-
tions used initially (exposure to 1 M hydrazine in
acetonitrile), only 2?7% of the oleic acid surface
ligands are lost. A much higher percentage of
ligand removal (?85?90%) of oleic acid was ob-
served when 1 M hydrazine in ethanol solution
was used instead.18Surprisingly, these differences
in the surface coverage did not translate into dra-
matic changes in the conductivity. Therefore, ad-
ditional experiments are needed to clarify the
electronic coupling of hydrazine-treated QD sol-
ids. In particular, this would be helpful for the de-
velopment of QD photovoltaic devices, for which
PbSe QDs are becoming a heavily studied model
system.20,21
To address questions about the electronic cou-
pling, 3D QD solids may not be the best experi-
mental system. When the surface ligands are re-
moved from the QDs in such a film, geometric
frustration may arise as the QDs try to pack more
tightly or fuse. In other words, drastic structural re-
arrangements in the solid will be required. This
can lead not only to cracks on a long length scale
(?100 nm), as has been observed,18but also to
variations in the local inter-QD spacing or the extent of
aggregation and fusing. These complications due to
geometric frustration can be avoided when the assem-
bly of QDs occurs in the solution phase. Kotov and co-
workers showed that, when the surface protective
ligands are removed in a slow and controlled fashion,
CdTe QDs in the solution can self-organize into one-
dimensional pearl necklace type aggregates that subse-
quently recrystallize into uniform nanowires with diam-
eters determined by those of the original QDs.22
Fluorescence spectroscopy showed a gradual red shift
of the first exciton transition by as much as ?30?60
meV in the aggregation?recrystallization process.
These authors have also extended the approach to the
self-assembly of CdTe QDs into free-floating two-
dimensional sheets with single QD thickness.23
Since most applications of strongly coupled QD sol-
ids require a solid support, here we explore whether
strong electronic coupling due to reduced geometric
frustration is possible with chemical treatment of two-
dimensional (2D) assemblies of QDs on flat surfaces.
We study 2D PbSe QD solids because of the high con-
ductivity demonstrated in this system and the extensive
chemical, structural, and electronic information avail-
able from recent studies. Compared to other well-
studied II?VI QD systems, such as CdSe, PbSe pos-
sesses better properties for transport. For example, it
has a much higher dielectric constant (?), lower effec-
tive masses for both the electron and the hole (meand
mh), and a higher density-of-states (DOS) due to a high
degeneracy in the electronic energy levels.24High ?
leads to small Coulomb charging energies (EC) while
low meand mhincrease the spatial extent of the elec-
tronandholewavefunctionsoutsidetheQD.Thesefac-
tors can explain the particularly high conductivity and
charge carrier mobility in films of PbSe QDs. By combin-
ing attenuated total reflection Fourier transform infra-
red (ATR-FTIR) spectroscopy and atomic force micros-
copy (AFM), we probe the chemical, structural, and
electronic properties of these assemblies. We find that
Figure 1. Attractive-regime AC-mode atomic force microscopy (AFM) images of
submonolayers of PbSe QDs assembled on TiO2(110): (a,b) as-deposited QDs with
oleic acid surface ligands; (c,d) after exposure to 1 M hydrazine in acetonitrile
for 3 min. Panels (b) and (d) are magnified scans of (a) and (c), respectively. All
green scale bars are 100 nm. (e) Height histograms of images (a) and (c). The car-
toon illustrates proposed morphology changes with reaction time. Note that the
surface coverage of QDs in (a) and (c) is slightly different due to variations is the
dip-coating process.
ARTICLE
www.acsnano.orgVOL. 3 ▪ NO. 6 ▪ 1532–1538 ▪ 2009
1533
Downloaded by UNIV OF MINNESOTA on August 19, 2009
Published on May 20, 2009 on http://pubs.acs.org | doi: 10.1021/nn9001819

Page 3

the electronic coupling is 10 times greater than that re-
ported previously for 3D solids. This has implications
for understanding PbSe films for transport and photo-
voltaic devices. Further, it demonstrates that such 2D
assemblies can provide a powerful approach for study-
ing electronic coupling in a variety of QD systems.
RESULTS AND DISCUSSION
We form 2D assemblies of PbSe QDs with oleic acid
capping via the dip-coating method and subsequently
carry out exchange reactions in 1 M hydrazine/acetoni-
trile solution. Figure 1 shows atomic force microscopy
(AFM) images of a submonolayer of PbSe QDs on a
(110) TiO2surface taken before and after the hydrazine
exposure (3 min). Before exposure (Figure 1a), the sur-
face is characterized by a network of two-dimensional
islands. The magnified image (Figure 1b) clearly reveals
hexagonally close-packed domains with an inter-QD
distance of 10 ? 1 nm. This is also the height of the is-
lands, as shown by the blue histogram in Figure 1e. The
inter-QD distance and the island height are consistent
with the diameter (? ? 6 nm ? 5%) of the QDs plus a
2 nm surface layer due to the oleic acid ligands. After
hydrazine exposure, the AFM image (Figure 1c) reveals
two effects. First, as shown by the red histogram in Fig-
ure 1e, the island height decreases to 6 ? 1 nm, which
is identical to the size of bare QDs (i.e., without their sur-
face ligands). Second, individual QDs can no longer be
resolved in the images. Both effects would be expected
if the oleic acid ligands are removed and the inter-QD
separation decreased. The removal of the oleic acid is
also verified by vibrational spectroscopy (see below). In-
stead of individual QDs, the magnified image (Figure
1d) shows larger domains (lateral size 20?40 nm) that
are separated by boundaries. This behavior is consistent
with two-dimensional densely packed islands of QDs
that suddenly lose their surface ligands while exhibit-
ing limited mobility on the substrate. In this case, the
large islands would break up into domains. Our data
would then suggest that these domains correspond to
2D aggregates of 10?100 closely packed bare QDs. We
estimate that the surface coverage of these 2D aggre-
gates in our sample is ?20?30% of a monolayer (ML).
Note that, after hydrazine treatment, we also see a
broadening of the height distribution (Figure 1e). This
is consistent with the presence of more disorder in the
2D assemblies and the breakup of large islands as oleic
acid capping molecules are removed. For small do-
mains on the order of or smaller than the tip size, con-
volution of tip shape with the topography of the 2D ag-
gregates (including some individual QDs) tends to
smear out and broaden the height distribution in a
topographical AFM image.
Additional information can be obtained from ATR-
FTIR spectra on similar films. The hydrazine exposure in-
duces two major changes. First, the C?H stretch at
?2920 cm?1(362 meV) disappears (Figure 2a), consis-
tent with removal of oleic acid. The residual and poorly
resolved C?H stretch that remains may be attributed
to background contaminants, as it is also present in the
spectrum from a clean TiO2surface. Second, a system-
atic red shift in the first exciton peak with increasing hy-
drazine exposure time is observed (Figure 2b). For ?0.2
ML of 6 nm PbSe QDs on TiO2, this peak shifts from
5650 (700 meV) to 5000 cm?1(620 meV) after 10 min
of exposure to 1 M hydrazine in acetonitrile. This shift
is independent of the nature of the substrate surface, as
a nearly identical shift is observed for the same cover-
age of PbSe QDs assembled on the SiO2-terminated Si
surface. However, the red shift is not observed when
the starting coverage of PbSe QDs is significantly lower,
as shown in Figure 2c for 0.02 ML after hydrazine expo-
Figure 2. (a) ATR-FTIR spectra in the C?H stretch region of
?0.2 ML film of oleic-acid-capped PbSe QDs on TiO2before
(solid line) and after (dashed line) exposure to 1 M hydrazine
in acetonitrile for 10 min. (b) ATR-FTIR spectra in the exciton re-
gion (EX1and EX2label the first and second exciton transi-
tions, respectively) of the same surface as a function of reac-
tion time in 1 M hydrazine in acetonitrile. (c) ATR-FTIR spectra
of ?0.02 ML oleic-acid-capped PbSe QDs on TiO2(solid) and af-
ter exposure to 1 M hydrazine in acetonitrile for 70 min (dot-
ted line).
Figure 3. Right: Width of the first exciton absorption peak as a
function of the exposure time to 1 M hydrazine for 0.2 ML of 6
nm QDs on TiO2. Left: Energy of the first exciton transition as a
function of the exposure time to 1 M hydrazine in acetonitrile for
the indicated surface coverage and QD diameter (red, 0.2 ML of 6
nm QDs; blue, 0.02 ML of 6 nm QD; gray, 0.2 ML of 4 nm QDs) on
SiO2(green) and TiO2(all other colors).
ARTICLE
VOL. 3 ▪ NO. 6 ▪ WILLIAMS ET AL.www.acsnano.org
1534
Downloaded by UNIV OF MINNESOTA on August 19, 2009
Published on May 20, 2009 on http://pubs.acs.org | doi: 10.1021/nn9001819

Page 4

sure for 70 min. Finally, the magnitude of the red shift
depends on the size of the QDs. For smaller QDs of ? ?
4 nm, the maximum red shift is almost twice as large
as that of ? ? 6 nm. Note that hydrazine treatment of
the 0.2 ML samples also increases the absorbance at en-
ergies higher than the first exciton transition (Figure
2b). This has been observed before for multilayer PdSe
thin films and was attributed to enhanced electronic ex-
change coupling between QDs,18,19although changes
in light scattering may also play a role.
The left panel in Figure 3 summarizes the observed
shift of the first exciton transition energy as a function
of reaction time for ?0.2 ML of 4 nm QDs on TiO2(gray
triangles), ?0.2 ML of 6 nm QDs (red circles on TiO2
and green diamonds on SiO2-terminated silicon), and
?0.02 ML of 6 nm QDs on SiO2(squares). For the low
surface coverage (0.02 ML), the exciton transition en-
ergy does not change with hydrazine exposure time.
For the high surface coverages (0.2 ML), the exciton
transition energy as a function of exposure time can
be described well by an exponential decay (solid
curves)forbothsizesandmostoftheshiftsoccurwithin
the first 10 min of reaction time. In contrast to peak po-
sition, width of the first exciton absorption peak in-
creases only slightly following hydrazine treatment
(?10% increase), as shown in the right panel for 6 nm
PbSe QDs.
For comparison, the red shift in the first exciton tran-
sition energy is much smaller when the thickness of the
QD film is more than one monolayer. Figure 4 shows
FTIR spectra of a 20 nm thick film of PbSe QDs (diam-
eter slightly less than 6 nm) deposited on the SiO2sur-
face. With increasing time of hydrazine exposure, we
see up to 80% loss of oleic acid ligands for a total reac-
tion time of 60 min (left panel). The first exciton transi-
tion red shifts by a total of 204 cm?1(25 meV), right
panel. This red shift is a factor of 3?4 smaller than that
for the submonolayer coverage.
Threepossibleexplanationsexistforthemuchlarger
red shift in the first exciton transition upon exposure
to hydrazine for monolayer QDs than that for multilay-
ers: (i) growth in the size of the QDs, (ii) an increase in
electronic interactions between the QDs and the sub-
strate,or(iii)anincreaseininter-QDelectroniccoupling.
The first possibility can be eliminated for the following
reasons. The PbSe QDs used in this study had a narrow
size distribution (?5%). Growth of some PbSe QDs can
only occur if others in the distribution shrink. This ripen-
ing of the distribution would lead to a broadening of
the exciton transition, rather than a systematic red shift.
Growth of the PbSe QDs is also unlikely since the height
of the films after hydrazine exposure (Figure 1e) quan-
titatively matched what would be expected for our bare
QDs.(AsimilarconclusionwasreachedbyNozikandco-
workers during their detailed studies on the influence
of hydrazine on PbSe QDs.) The second explanation can
also be eliminated as the systematic red shift was ob-
served for 2D PbSe QD assemblies not only on TiO2
but also on SiO2, which has a much wider band gap.
Both the electron and hole levels of the PbSe QDs
would be located in the band gap of SiO2. Thus, strong
electronic coupling would not be expected. In contrast,
an increase in the inter-QD electronic coupling can ex-
plain all of our observations. In this case, the red shift re-
sults when the QDs move closer to each other after hy-
drazine exposure and 2D aggregates form. This
conclusion is consistent with (i) the disappearance of
the oleic acid in ATR-FTIR spectroscopy, (ii) the inability
to resolve individual QDs in AFM, (iii) the decrease in is-
land height from 10 to 6 nm, and (iv) the absence of a
red shift at very low surface coverages when the ad-
sorbed QDs are isolated. There are two ways to under-
stand the increased inter-QD electronic coupling.
The first picture considers the uniform decrease of
inter-QD separation as capping molecules are removed
and the molecular capping layer gets thinner. With de-
creasing inter-QD separation, the electronic exchange
coupling energy, ? ??a|Hˆ|b? (between two neighboring
quantum dots a and b), would be expected to increase
due to more spatial overlap of the wave functions. This
would lead to a red shift in the first exciton transition, as
illustrated in Figure 5. In particular, for a hexagonally
close-packed 2D assembly of s-orbitals, tight-binding
theory predicts that a new electronic band would form
Figure 4. ATR-FTIR spectra for a 20 nm thick film of oleic-acid-
capped PbSe QDs (?6 nm diameter) on SiO2as a function expo-
sure time to 1 M hydrazine in acetonitrile for 0?60 min.
Figure 5. Schematic illustration of band formation in a QD
solid as the inter-QD distance (D) decreases; ?Eis the exci-
ton transition energy.
ARTICLE
www.acsnano.org VOL. 3 ▪ NO. 6 ▪ 1532–1538 ▪ 2009
1535
Downloaded by UNIV OF MINNESOTA on August 19, 2009
Published on May 20, 2009 on http://pubs.acs.org | doi: 10.1021/nn9001819

Page 5

withanenergywidthof12|?|.25Ifsuchabandisformed
both by the 1s electron and hole levels of the QDs, we
can estimate the first exciton transition energy (?E) at a
particular interparticle separation D as
where ?E
lated QDs, and ?eand ?hare the electronic exchange
coupling energies for the electron and hole levels, re-
spectively. Equation 1 neglects the exciton binding en-
ergy and assumes that ?eand ?hare close in value. The
latter is justified since meand mhin PbSe are very simi-
lar.24With this simple model, the red shift can be esti-
mated to be 12|?|. Note that this tight-binding model
predicts that the electronic excitations can become
more delocalized as |?| increases. Also note that, in an
fcc close-packed 3D QD solid, the red shift should be
?24|?| as the number of nearest neighbors is 12.
The second picture is more localized. As the QDs are
starved of capping molecules, the remaining capping
molecules have more freedom to rearrange on each QD
surface. This rearrangement may expose bare PbSe sur-
faces and allow the direct contact (without the inter-
vening molecular layer) of neighboring quantum dots.
Due to the high surface energy of the bare QD surface,
we expect the direct contact or aggregation process to
be thermodynamically favorable. This leads to a 2D net-
work of strongly coupled QDs, resulting in exciton delo-
calization and a red shift in exciton energy. In fact, such
a localized coupling mechanism is similar to those dis-
covered by Kotov et al. for the nearly frustration-free
formation of 1D nanowires or 2D nanosheets from the
aggregation of QDs in the solution phase due to the
slow and controlled removal of capping molecules.22,23
In the present case on a 2D solid surface, the end re-
sult for the QD assemblies when all capping molecules
are removed is similar to the delocalized mechanism in
the previous paragraph. The difference lies in the extent
of order/disorder: the delocalized coupling mechanism
should yield a more ordered 2D superlattice, while the
localized mechanism should give more disorder.
In reality, both localized and delocalized coupling
mechanisms may operate to increase inter-QD elec-
tronic coupling. We arrive at two important conclu-
sions regarding the strong inter-QD electronic coupling.
First, the |?| value for 2D assemblies of PbSe QDs are
as large as 7 and 13 meV (for 6 and 4 nm QDs), which
is 1 order of magnitude larger than that for 3D solids
with the same chemical treatment. We believe this is
due to more significant geometric frustration in 3D as-
semblies than in 2D. Each QD in a 3D solid has 12 neigh-
bors,ascomparedto6inthe2Dassembly.Thegeomet-
ric frustration inhibits the close packing of QDs after
theremovalofthesurfaceligands.Thelarge?valueob-
served here for 2D PbSe assemblies is also consistent
with a scanning tunneling microscopy and spectros-
0is the first exciton transition energy for iso-
copy study of thermally annealed 2D assemblies of
PbSe QDs by Liljeroth et al.15These authors performed
calculations based on the effective mass model and
showed that |?| could be as large as 25 meV for an
inter-QD separation of 0.5 nm. Second, our results show
that the inter-QD electronic coupling is continuously
tunable for the 2D QD assemblies and that the coupling
can be controlled simply by changing the hydrazine ex-
posure time. Small changes in separation lead to large
changes in the electronic coupling because the effec-
tive masses in PbSe are small. In other words, PbSe QDs
have an increased spatial extension of the electron
and hole wave functions outside the QD boundary.
This leakage of the electron and hole will also increase
with decreasing QD size. This is indeed observed experi-
mentally: for 4 nm PbSe QDs the maximum |?| value is
13 meV while that for 6 nm QDs is 7 meV.
We now address the implications of the strong elec-
tronic coupling in QD solids on their applications, par-
ticularly photovoltaic devices. In a thin disordered film
of PbSe QDs, depending on how the film was formed
and treated, there may be domains of QDs with strong
electronic coupling separated by boundaries where the
coupling is weaker. These islands with strong elec-
tronic coupling may be viewed as “super” quantum
dots where the electron and hole wave functions will
be delocalized across the whole domain (island). We
can identify at least four important roles that these
strongly coupled domains can play in a photovoltaic
device. First, the optical absorption cross section above
the first exciton transition can increase with |?| such
that the strongly coupled domains serve as enhanced
light absorption centers. Second, energy transfer to the
strongly coupled domains from the surrounding QDs
can be a downhill process and serve as a pathway to
channel excitation energy. These strongly coupled re-
gions can act as basins for both charge and energy and
influence how carriers and excitons move within the
film. Charge transport in disordered QD solids occurs
through strongly coupled percolation pathways or via
hopping among these super dots or islands. Third, elec-
tronic delocalization within each strongly coupled do-
main decreases exciton binding energy and increases
the probability of exciton dissociation. Fourth, elec-
tronic coupling and delocalization decrease the QD
charging energy and increase charge mobility. One of
the challenges in the design of future QD solar cells is to
determine how to take advantage of these effects that
arise due to strong electronic coupling. Another chal-
lenge is to find ways to assemble nanostructures that
exploit such effects in order to increase the overall
charge collection and power conversion efficiencies of
these solar cells.
CONCLUSIONS
We find that hydrazine exposure of a two-
dimensional PbSe QD array assembled on a solid sur-
∆E) ∆E
0- (6|?e| + 6|?h|) ≈ ∆E
0- 12|?|(1)
ARTICLE
VOL. 3 ▪ NO. 6 ▪ WILLIAMS ET AL.www.acsnano.org
1536
Downloaded by UNIV OF MINNESOTA on August 19, 2009
Published on May 20, 2009 on http://pubs.acs.org | doi: 10.1021/nn9001819