LETTER pubs.acs.org/NanoLett
Size- and Temperature-Dependent Charge Transport in PbSe Nanocrystal Thin Films Moon Sung Kang,† Ayaskanta Sahu,†,‡ David J. Norris,*,‡ and C. Daniel Frisbie*,† †
Department of Chemical Engineering and Materials Science, University of Minnesota, 421 Washington Avenue SE, Minneapolis, Minnesota 55455, United States ‡ Optical Materials Engineering Laboratory, ETH Z€urich, Universitaetstrasse 6, 8092 Z€urich, Switzerland
bS Supporting Information ABSTRACT: We report the size- and temperature-dependence of electron transport in thin films of PbSe nanocrystals. Upon increasing temperature over the range 28200 K, the electron transport underwent a transition in mechanism from Efros-Shklovskii-variable-rangehopping (ES-VRH) to nearest-neighbor-hopping (NNH). The transition occurred at higher temperatures for films with smaller particles. The electron localization length, estimated from the ES-VRH model, was comparable to the nanocrystal size and scaled systematically with nanocrystal diameter. The activation energy from the NNH regime was also size-dependent, which is attributed both to size-dependent Coulomb effects and the size-distribution of nanocrystals. KEYWORDS: PbSe nanocrystals, size-dependence, Efros-Shklovskii-variable-range-hopping, nearest-neighbor-hopping, localization length, activation energy
he size-tunable electronic properties1,2 and the solution processability3 of colloidal semiconductor nanocrystals (NCs) have made these materials promising candidates for thin-film optoelectronics, such as solar cells,4,5 light emitting devices,6 and photosensors.7 For these applications, significant progress has been made in understanding optical properties of NCs810 as well as in development of synthetic methods to prepare monodisperse NCs.3,11,12 Among many other NC systems, PbSe NCs have attracted great interest because PbSe NCs feature (i) sizetunable interband transitions in the near-infrared,13 (ii) multiple exciton generation,14,15 (iii) opportunities for hot-electron transfer,16 (iv) high thin-film charge carrier mobility,17 and (v) ambipolar (hole and electron) charge transport.18,19 To further exploit these properties in optoelectronic applications, more emphasis on fundamentals of the charge transport in NC assemblies is necessary, since many of these devices rely on electrical conduction between NCs. In particular, a basic understanding of the influence of particle size, which is the most characteristic parameter of nanomaterials, on the charge transport mechanism is essential. For PbSe NC assemblies, Coulomb-blockade transport20 and variablerange-hopping transport21 were observed at low temperatures, whereas Arrhenius-type thermally activated transport was observed at high temperatures.20,22 However, all these studies were performed on a film based on NCs with a single particle size and the influence of particle size on the charge transport mechanism has not been elucidated yet. Recently, the dependence of electron and hole mobility on particle size was demonstrated for chemically treated PbSe NC films.23 However, this study did not include any temperaturedependent transport measurements, which are critical to understand the origin of the size-dependent electrical conduction.
T
r 2011 American Chemical Society
Here, we describe the influence of particle size on the charge transport mechanisms, the transitions in the transport mechanisms, the carrier localization length, the conductivity, and the thermal activation energy in chemically treated PbSe NC thin films. We employed field effect transistors (FETs) based on films of different sized NCs, which allowed monitoring of electrical transport in these films at different temperatures (20028 K). We observed that electron transport exhibits nearest-neighborhopping (NNH) at higher temperatures and transitions to EfrosShklovskii variable-range-hopping (ES-VRH) at lower temperatures, consistent with previous reports by Guyot-Sionnest and co-workers on single-sized CdSe NC films.24,25 The transition temperature between these charge transport mechanisms was size-dependent with the transition taking place at higher temperature for films of smaller particles. From the temperature regime that exhibits ES-VRH transport, the electron localization length in the NC films was determined to scale linearly with particle diameter. The activation energy for charge transport from the NNH temperature regime was inversely proportional to the particle diameter. We attribute the origin of the NNH activation energy to the sum of (i) the size-dependent Coulomb penalty necessary for each hopping process between NCs and (ii) the inherent energy disorder due to the finite particle size distribution. These two contributions can be qualitatively estimated, and their sum is in good agreement with the measured activation energy values from experiments. Overall, the size- and Received: June 15, 2011 Revised: August 8, 2011 Published: August 16, 2011 3887
dx.doi.org/10.1021/nl2020153 | Nano Lett. 2011, 11, 3887–3892
Nano Letters
LETTER
Figure 1. (a) Optical absorbance spectra of four different sized PbSe NCs dispersed in tetrachloroethylene. (b) Schematic of cross section of PbSe NC FET (not to scale). (c) AFM height image of EDT treated 7.1 nm PbSe NC films.
temperature-dependent charge transport properties of NC films described here provides more thorough understanding of electrical conduction in NC films. Various sizes of PbSe NCs (diameters ranging from 3.8 to 8.4 nm), passivated primarily with oleic acid, were prepared by injecting a mixture of tri-n-octylphosphine selenide and diphenylphosphine into a hot solution of lead oxide, oleic acid, and octadecene following a modified literature procedure.26,27 Throughout the particle synthesis and the postsynthesis procedure to prepare clean NC dispersions, exposure of the particles to ambient was carefully avoided. See Supporting Information for details. The optical absorbance spectra of these NCs dispersed in tetrachloroethylene are displayed in Figure 1a. Particle diameters were determined from a published correlation of size with the first absorbance peak.28 Figure 1b shows a schematic of the cross section of a typical PbSe NC FET that was prepared by the following procedure. Briefly, a layer of Al/Au (10 nm/75 nm) was deposited on the backside of a heavily doped Si wafer to work as a gate electrode. The front side of the wafer was covered with thermally grown 300 nm thick SiO2 (specific capacitance =11.5 nF/cm2). On top of the SiO2 layer, source and drain Cr/Au (2.5 nm/32.5 nm) electrodes were patterned by standard photolithography.29 The length (L) and the width (W) of the channel varied from 50 to 200 μm and from 1 to 2 mm, respectively. These wafers were treated with octadecyltrimethoxysilane (OTMS)30 to passivate electron traps at the SiO2/NC solid interface and then transferred into a nitrogen glovebox. Films of NCs were spin-coated on these wafers from dispersions of different sized PbSe NCs in anhydrous octane. To improve conduction, the films were treated with 0.05 M ethanedithiol (EDT) in acetonitrile.26,29 Cracks in the films, which resulted from the chemical treatment, were filled by a second round of spin-coating of NC dispersions. The resulting films were continuous and devoid of cracks, as shown in an atomic force microscopy (AFM) image in Figure 1c. Without any air exposure, these devices were then transferred into another glovebox equipped with a vacuum probe station. The devices were inserted into the vacuum probe station either at room temperature or at 235 K and stored under vacuum (∼106 Torr) for more than two hours before taking electrical measurements. See Supporting Information for details. The drain currentgate voltage (IDVG) characteristics of the PbSe NC FETs at a given drain voltage (VD) showed a typical
Figure 2. (a) IDVG characteristics of an FET based on 7.1 nm PbSe NCs measured at different temperatures from 200 to 28 K (VG sweep direction: from 50 to 70 V). (b) Semilog plot of σ vs 1/kBT for films of four different sized NCs. (c) Loglog plot of d(log σ)/d(log T) vs T to determine temperature dependence of conductivity. (d) Semilog plot of σ vs 1/T0.5 for films of four different sized NCs.
V-shaped ambipolar transport characteristic, where ID increases with the magnitude of applied VG (|VG|) (see Figure 2a or Supporting Information Figure S1). The increase in ID with positive gate voltage (right wing) indicates electron conduction while increased ID with negative gate voltage (left wing) results from hole conduction. The IDVG characteristics measured near room temperature showed significant hysteresis such that ID upon carrier injection (while |VG| increases) was higher than that upon carrier extraction (while |VG| decreases) for both electron and hole conduction. For example, see the gray curve in Supporting Information Figure S1. This type of hysteretic behavior is generally observed from a system where injected charge carriers become trapped and screen the gate bias.26,31 Unfortunately, the prominent hysteresis in the IDVG curve prevented extraction of meaningful electron and hole mobilities or conductivities at room temperature. This is because an IDVG curve with noticeable hysteresis yields two very distinct transconductance (dID/dVG) values (one upon carrier injection and the other upon carrier extraction) and thus two distinct mobility or conductivity values. Importantly, the hysteresis was significantly suppressed at lower temperatures, especially for electron conduction. As a consequence, the hysteresis for electron conduction (right wing) became essentially negligible below 200 K, whereas noticeable hysteresis for hole conduction (left wing) persisted even at lower temperatures. (See the blue and the red curves in Supporting Information Figure S1 that were measured at 80 and 200 K, respectively.) Such a suppression of hysteresis upon cooling suggests that carrier trapping is thermally activated. Also, the nearly complete disappearance of hysteresis for electron conduction at low temperatures implies that electron transport is not influenced by traps at these low temperatures. This leads us to 3888
dx.doi.org/10.1021/nl2020153 |Nano Lett. 2011, 11, 3887–3892
Nano Letters
LETTER
believe that electron transport in PbSe NC films below 200 K occurred through the quantum states of each particle, rather than particle surface trap states. More importantly, the disappearance of the hysteresis in IDVG characteristics allowed us to extract meaningful mobility and conductivity values for electron transport. The main discussion in this letter will be focused on electron transport, rather than hole transport, in PbSe NC films. Figure 2a displays IDVG characteristics (at VD = 10 V) that were obtained from a film of 7.1 nm PbSe NCs measured from 20028 K. The same IDVG curves plotted on a linear scale are displayed in Supporting Information Figure S2. Conductivity (σ) at a specific temperature was estimated from the equation σ = enμ = (L/W)(VG Vth)[(dID/dVG)/VD], where e is the elemental charge, n is the carrier concentration, μ is the mobility, and Vth is the threshold voltage of electron conduction in an IDVG characteristic. The slope of an IDVG curve (dID/dVG) was taken at (VG Vth) = 50 V, which corresponds to ∼4 1012 charges/cm2. Because the thermal coefficient of capacitance for the SiO2 gate dielectric is quite small ( VD was satisfied to ensure that only electrons, but not holes, contribute to electrical conduction. Note that if (VG Vth) is comparable to or lower than VD, both electrons and holes can be involved in charge transport (true ambipolar transport). Arrhenius plots of conductivity versus temperature (σ vs T) for four different sized PbSe NC films are displayed in Figure 2b. As described above, these conductivity values were obtained at a charge density of ∼4 1012 charges/cm2, which corresponds to inducing 0.7, 1.0, 1.3, and 1.8 charges per NC for 4.1, 5.1, 6.0, and 7.1 nm NCs, respectively. A linear relation of σ vs T in the Arrhenius plots, in general, suggests that the transport follows the nearest-neighbor-hopping mechanism within the given temperature range.20,32 However, as shown in Figure 2b, the temperature dependence of conductivity deviated from a simple linear relation over the given temperature range. This implies that the charge transport cannot be simply described as NNH over the range 28200 K. It is well-known that for hopping conduction, the temperature dependence of the conductivity takes the general form σ ¼ σ0 exp½ ðT =TÞz
the ES-VRH model, conductivity follows σ = σ0 exp[(T0/ T)0.5], where T0 is a fitting parameter with units of Kelvin, as a consequence of considering the soft Coulomb gap (EC).33,34 The Coulomb gap, which arises from electron correlations, can be considered as the minimum energy required for a hopping process to occur between localized states with finite energy and spatial distribution. The change in slope of the d(log σ)/d(log T) versus T plot implies that rather than following a single charge transport mechanism over the entire temperature span, a transition in the charge transport mechanism from ES-VRH to NNH occurs at 70100 K. Pinpointing an exact transition temperature may not be trivial from the plot but one can observe that the transition temperature increases with decreasing particle diameter. The origin of the size-dependence of the transition temperature will be discussed below. A transition in the transport mechanism can be understood as follows. Equation 1 can be derived from a more general expression for hopping between disordered states 2x ΔE ð2Þ σ ¼ σ 0 exp a kB T where σ0 is the temperature independent pre-exponential factor, x is the spatial distance between states that are involved in hopping (or the hopping distance), a is the localization length of a carrier, ΔE is the energy difference between these states, and kB is the Boltzmann constant. At low temperatures within the ESVRH regime, where ΔE = EC = 0.35e2/4πεε0x (ε0 is the vacuum permittivity, and ε is the dielectric constant), hopping does not take place between the nearest neighbor states because hopping to closer states (shorter x) often requires more energy (larger ΔE) on average. Instead, hopping occurs over an optimum distance (x*) that maximizes eq 2. Note that x* is a function of temperature and decreases with increasing temperature (x* = (0.35e2a/8πεε0kBT)0.5).33,34 Accordingly, eq 2 becomes 2x EC σ ¼ σ 0 exp ð3Þ a kB T which can be reduced to σ ¼ σ 0 exp½ ðT0 =TÞ0:5
ð1Þ
where σ0 is the conductivity pre-exponential factor, T* is a fitting parameter with units of Kelvin, and z is a parameter that describes the power of the temperature dependence. z can be determined from the slope of a d(log σ)/d(log T) versus T plot.25 As shown in Figure 2c, a clear transition in the slope of d(log σ)/d(log T) versus T plot was observed for all sizes of particles tested over 28200 K, such that z values close to 0.5 were observed at lower temperatures and close to 1 at higher temperatures.32 Temperature dependence with z = 1, or an Arrhenius relation is consistent with the NNH model, whereas z = 0.5 corresponds to the ESVRH model. If the charge transport follows Mott variable-rangehopping (M-VRH), z should be either 0.25 (three-dimensional transport) or 0.33 (two-dimensional transport). Distinguishing transport models can be difficult because of the similar values of z. However, our data are best fit with z values of 0.5, not 0.25 or 0.33. Further, it has been argued previously that M-VRH should only occur when the 1Se state is either nearly empty or completely occupied (8 electrons for PbSe NCs), where carrier correlations essential to ES-VRH will be deemphasized.25 In
ð4Þ
2
with T0 = 2.8e /4πεε0akB. However, as the temperature increases, the optimum hopping distance eventually becomes equal to the nearest neighbor distance (d), which is the sum of the particle diameter (D) and interparticle spacing (δ). Physically, the hopping distance cannot be any shorter than the nearest neighbor distance. Hence, the hopping distance above this specific temperature (Ttr), where x* becomes equal to d, is independent of temperature and becomes constant (x* = d = constant). Therefore, above Ttr eq 3 reduces to σ ¼ σ 00 exp½ EC =kB T
ð5Þ
with σ00 = σ0 exp[2d/a] which is a temperature independent constant. Equation 5 resembles an Arrhenius relation, which describes NNH transport, with its activation energy (EA) corresponding to EC using x = d (EC = 0.35e2/4πεε0d). Such a transition in charge transport mechanism was observed in electrochemically charged 5.4 nm CdSe NC films near 120 K.24 From the σ versus T behavior, several characteristic parameters describing charge transport in NCs can be extracted. First, 3889
dx.doi.org/10.1021/nl2020153 |Nano Lett. 2011, 11, 3887–3892
Nano Letters
LETTER
Figure 3. Transport parameters in the ES-VRH regime. (a) Size dependence of the localization length. Error bars represent one standard error computed from the linear regression on the fitting of σ vs T0.5. (b) Hopping distance at different temperatures for different sized particles. (c) Coulomb energy at different temperatures for different sized particles. Solid lines are guides to the eye. Thermal energy is displayed in black dashed line for comparison.
from the slope of σ vs 1/T0.5 plots in the ES-VRH temperature regime, as shown in Figure 2d, the fitting parameter T0 and the localization length (a) of electrons for different sized particles can be estimated according to a = 2.8e2/4πεε0kBT0.24,33,34 We find that the localization length of electrons in EDT-treated PbSe NC films is comparable to the diameter of a particle and thus scales with particle diameter, as shown in Figure 3a. For these calculations, values of ε for different sized PbSe NC films were taken directly from the literature.35 The scaling between the localization length and the particle diameter is expected because the localization length is essentially the decay length of an electron in NC assemblies and should be inversely proportional to the number of hops necessary to travel a fixed distance. The number of hops, in turn, is inversely proportional to the particle size. Thus, the localization length should increase uniformly with increasing particle diameter.36 To the best of our knowledge, Figure 3a is the first demonstration of the scaling of localization length with NC diameter. Using the localization length for each sized particle, the optimum hopping distance at a given temperature can be estimated from x* = (0.35e2a/8πεε0kBT)0.5. Circles in Figure 3b display the optimum hopping distance of electrons in different sized NC films. Dotted lines in Figure 3b display the nearest neighbor distance for different sized particles. An interparticle spacing of 0.5 nm was used to calculate the nearest neighbor distance.23 As expected, x* values at low temperatures are longer than the distance between nearest neighbors and decrease with increasing temperature. One can also estimate Ttr from the point in Figure 3b where x* becomes equal to d, which ranged from 40 to 75 K for different sized particles. Consistent with Figure 2c, we find that the transition in the charge transport mechanism takes place at higher temperature for smaller particles. Also, for a given optimum hopping distance at a given temperature in the ES-VRH regime, the Coulomb gap or the Coulomb penalty for hopping to occur was estimated from EC = 0.35e2/4πεε0x*. As shown in Figure 3c, these values are larger than the thermal energy at a given temperature throughout the entire ES-VRH temperature range for all sized particles tested, which satisfies the basic assumption of ES-VRH model. Additional characteristic parameters describing charge transport in NCs can be extracted from the NNH temperature regime above Ttr as well. In particular, conductivity pre-exponential factors and activation energies were obtained from the y-intercept and the slope of the Arrhenius plots of conductivity for different sized particles, respectively. In addition to the four
Figure 4. Transport parameters in the NNH regime. (a) Size dependence of σ at 200 K (closed) and σ0 (open) for electron conduction at VG Vth = 50 V. (b) Size dependence of EA (black squares) and its comparison to the sum of EC (blue) and ED (red). Error bars represent one standard deviation obtained from measurements on 35 devices.
sized particles described so far, IDVG curves for seven more sizes of particles were measured but within a temperature range of 80220 K. Absorbance spectra of these extra sets of particles and their sizes are displayed in Supporting Information Figure S3. The conductivity pre-exponential factor is displayed in open squares in Figure 4a. This value describes the hopping rate between NCs, which is determined by (i) the number of hops necessary for charge transport, and (ii) the interparticle coupling between NCs. If the conductivity pre-exponential factor is primarily determined by the number of hops, one would expect a monotonic increase in σ00 with increasing particle diameter, because the total number of hops should be less for films based on larger particles. Such behavior was observed from films of different sized CdSe NCs.37 However, the results on our PbSe NC films shows a slight deviation from a monotonic increase. This deviation from monotonicity probably originates from a nonlinear change in interparticle electronic coupling upon the variation of particle diameter. This is reasonable because the electronic coupling can vary with numerous factors including the packing density of the particles, the packing density of the ligands on the particle surface, and the size or the shape of the particles, which may not be controlled precisely for particles with different diameters.23 In addition to the conductivity pre-exponential factors, conductivities for different sized particles at 200 K are shown in Figure 4a with closed squares. The overall sizedependence of conductivity for electron conduction is consistent with the size-dependence of electron mobility from PbSe NCs measured at room temperature that was reported previously.23 3890
dx.doi.org/10.1021/nl2020153 |Nano Lett. 2011, 11, 3887–3892
Nano Letters
LETTER
Table 1. Summary of Characteristic Values Describing Charge Transport in PbSe NC Filmsa D (nm)
a
fwhm (meV)
ED (meV)
EC (meV)
ED + EC (meV)
EA (meV)
T0 (K)
a (nm)
3.8
80
7.3
19.0
26.3
24.0 ( 2.3
3.9
75
6.8
18.0
24.9
20.1 ( 0.8
4.2
69
6.3
15.2
21.5
23.2
1702
3.9
5.1
59
5.4
9.7
15.1
12.9
845
6.0
5.1
55
5.0
9.7
14.7
14.6 ( 0.5
6.0
54
4.9
6.7
11.6
10.4
527
7.6
6.0
45
4.1
6.7
10.8
10.8 ( 0.2
6.8 7.1
51 51
4.7 4.7
5.1 4.8
9.8 9.4
9.2 ( 0.5 7.8
367
9.2
7.3
53
4.8
4.5
9.3
7.9 ( 0.4
8.4
44
4.0
3.3
7.3
6.5 ( 0.5
Values in bold are from the sets of particles for which the electrical measurements were taken down to 28 K.
The activation energy obtained from the Arrhenius plot describes the average energy that is necessary for each hop. As displayed in Figure 4b, a monotonic increase in EA is observed with decreasing particle diameter. A similar trend was also observed in a previous letter for activation energies of electron transport in films of different sized CdSe NCs.37 In that letter, we claimed that the size-dependent activation energy scales with the size-dependent charging energy which scales with 1/d. The origin of the size-dependent activation energy can be primarily understood from a description based on the transition in the mechanism for charge transport that takes place upon increasing temperature. We discussed above that electron transport in our PbSe NC films follows the ES-VRH model at lower temperatures, but transitions to NNH above Ttr (that follows σ = σ00 exp[EC/kBT]). It is important to emphasize that above Ttr, EC is temperature-independent but is size-dependent (EC = 0.35e2/4πεε0d). This suggests that even in nearest-neighborhopping transport, the size-dependent Coulomb penalty or EC must be paid for each nearest neighbor hop to occur, unless the thermal energy is significantly high enough to ignore EC completely. Therefore, the Coulomb gap with fixed d contributes to the activation energy for a given size of particles in the NNH temperature regime. Note that this formula of EC = 0.35e2/ 4πεε0d resembles the formula for the charging energy of a spherical capacitor with a numerical coefficient; previous reports often used the concept of charging energy to describe the sizedependence of transport activation energy.24,37 As summarized in Table 1, estimated values of EC and experimentally obtained values of EA both scale inversely with particle diameter. However, the values of EC are always smaller than those of EA for all of the sizes of NCs tested. This implies that an additional component that has not yet been treated contributes to the activation energy. Consequently, the influence of disorder in the electronic states was considered, which is inherent for NCs with a finite sizedistribution. Assuming that energy states follow a Gaussian distribution, one can estimate the average spacing in energy between the states. Because charge transport occurs through the very first PbSe NC layer on top of a SiO2 gate dielectric layer, charge transport can be considered as taking place between closed packed spheres in a two-dimensional layer. This yields a 6-fold coordination between adjacent particles. Then, the average energy difference between the adjacent energy states, which is the minimum energy (ED) necessary for electrons to hop to neighboring states, can be calculated to be 0.43s1S. Here, a numerical coefficient of 0.43
was obtained from a standard normal distribution table that corresponds to a probability of one-sixth and s1S is the standard deviation in energy of the Gaussian distribution which can be estimated from the full width at half-maximum (fwhm) of the first excitonic peak in the absorbance spectrum according to s1S = fwhm/(4 (2 ln 2)0.5). As displayed in Figure 4b, good agreement between EA and the sum of EC (blue) and ED (red) is observed. This suggests that, although the thermal activation energy results primarily from the size-dependent Coulomb penalty for each hopping process, the inherent energy disorder in the NC films also contributes and should be considered. In summary, the size- and temperature-dependent electron transport in EDT-treated PbSe NC films was thoroughly investigated. At low temperatures, electron transport followed the ES-VRH model. From the model, the localization lengths of electrons for different sized particles were estimated and they scaled with particle diameter. The charge transport mechanism transitioned to NNH upon increasing the temperature. The transition temperature was higher for films of smaller particles. The activation energy for charge transport in the NNH model was size-dependent and closely matched the sum of the sizedependent Coulomb gap and an average energy disorder due to size distribution in NC films. We believe that this study provides fundamental understanding of charge transport in NC films and thus will promote further applications of NCs in optoelectronics.
’ ASSOCIATED CONTENT
bS
Supporting Information. Description of materials, detailed experimental methods, and supplementary figures. This material is available free of charge via the Internet at http://pubs.acs.org.
’ AUTHOR INFORMATION Corresponding Author
*E-mail: (D.J.N.)
[email protected]; (C.D.F)
[email protected].
’ ACKNOWLEDGMENT We thank Alexander L. Efros and Professor Boris I. Shklovskii for valuable discussions. This work was supported primarily by the MRSEC Program of the National Science Foundation, Award Number DMR-0819885. Additional support was provided by the NSF Materials World Network under Award Number DMR0908629 and by the Center for Nanostructure Applications at the 3891
dx.doi.org/10.1021/nl2020153 |Nano Lett. 2011, 11, 3887–3892
Nano Letters University of Minnesota (UMN). M.S.K. was supported partially by a University of Minnesota Doctoral Dissertation Fellowship during this project. A.S. acknowledges funding from the Industrial Partnership for Research in Interfacial and Materials Engineering (IPRIME). We utilized resources at the UMN Characterization Facility and the UMN Nanofabrication Center, which receive partial support from the NSF under the NNIN program.
’ REFERENCES
LETTER
(32) Yildiz, A.; Serin, N.; Serin, T.; Kasap, M. Jpn. J. Appl. Phys. 2009, 48, 111203. (33) Efros, A. L.; Shklovskii, B. I. J. Phys. C 1975, 8, L49. (34) Shklovskii, B. I.; Efros, A. L. Electronic Properties of Doped Semiconductors; Springer-Verlag: Berlin, 1984. (35) Luther, J. M.; Law, M.; Beard, M. C.; Song, Q.; Reese, M. O.; Ellingson, R. J.; Nozik, A. J. Nano Lett. 2008, 8, 3488. (36) Zhang, J.; Shklovskii, B. I. Phys. Rev. B 2004, 70, 115317. (37) Kang, M. S.; Sahu, A.; Norris, D. J.; Frisbie, C. D. Nano Lett. 2010, 10, 3727.
(1) Alivisatos, A. P. Science 1996, 271, 933. (2) Klimov, V. I. Semiconductor and Metal Nanocrystals: Synthesis and Electronic and Optical Properties; Marcel Dekker: New York, 2004. (3) Murray, C. B.; Kagan, C. R.; Bawendi, M. G. Annu. Rev. Mater. Sci. 2000, 30, 545. (4) Nozik, A. J. Physica E 2002, 14, 115. (5) Kamat, P. V. J. Phys. Chem. C 2008, 112, 18737. (6) Colvin, V. L.; Schlamp, M. C.; Alivisatos, A. P. Nature 1994, 370, 354. (7) Konstantatos, G.; Howard, I.; Fischer, A.; Hoogland, S.; Clifford, J.; Klem, E.; Levina, L.; Sargent, E. H. Nature 2006, 442, 180. (8) Du, H.; Chen, C.; Krishnan, R.; Krauss, T. D.; Harbold, J. M.; Wise, F. W.; Thomas, M. G.; Silcox, J. Nano Lett. 2002, 2, 1321. (9) Nirmal, M.; Brus, L. Acc. Chem. Res. 1998, 32, 407. (10) Norris, D. J.; Bawendi, M. G. Phys. Rev. B 1996, 53, 16338. (11) Murray, C. B.; Sun, S.; Gaschler, W.; Doyle, H.; Betley, T. A.; Kagan, C. R. IBM J. Res. Dev. 2001, 45, 47. (12) Murphy, J. E.; Beard, M. C.; Nozik, A. J. J. Phys. Chem. B 2006, 110, 25455. (13) Wehrenberg, B. L.; Wang, C.; Guyot-Sionnest, P. J. Phys. Chem. B 2002, 106, 10634. (14) Ellingson, R. J.; Beard, M. C.; Johnson, J. C.; Yu, P.; Micic, O. I.; Nozik, A. J.; Shabaev, A.; Efros, A. L. Nano Lett. 2005, 5, 865. (15) Nozik, A. J.; Beard, M. C.; Luther, J. M.; Law, M.; Ellingson, R. J.; Johnson, J. C. Chem. Rev. 2010, 110, 6873. (16) Tisdale, W. A.; Williams, K. J.; Timp, B. A.; Norris, D. J.; Aydil, E. S.; Zhu, X. Y. Science 2010, 328, 1543. (17) Talapin, D. V.; Murray, C. B. Science 2005, 310, 86. (18) Wehrenberg, B. L.; Guyot-Sionnest, P. J. Am. Chem. Soc. 2003, 125, 7806. (19) Kang, M. S.; Lee, J.; Norris, D. J.; Frisbie, C. D. Nano Lett. 2009, 9, 3848. (20) Romero, H. E.; Drndic, M. Phys. Rev. Lett. 2005, 95, 156801. (21) Wehrenberg, B. L.; Yu, D.; Ma, J.; Guyot-Sionnest, P. J. Phys. Chem. B 2005, 109, 20192. (22) Mentzel, T. S.; Porter, V. J.; Geyer, S.; MacLean, K.; Bawendi, M. G.; Kastner, M. A. Phys. Rev. B 2008, 77, 075316. (23) Liu, Y.; Gibbs, M.; Puthussery, J.; Gaik, S.; Ihly, R.; Hillhouse, H. W.; Law, M. Nano Lett. 2010, 10, 1960. (24) Yu, D.; Wang, C.; Wehrenberg, B. L.; Guyot-Sionnest, P. Phys. Rev. Lett. 2004, 92, 216802. (25) Liu, H.; Pourret, A.; Guyot-Sionnest, P. ACS Nano 2011, 4, 5211. (26) Luther, J. M.; Law, M.; Song, Q.; Perkins, C. L.; Beard, M. C.; Nozik, A. J. ACS Nano 2008, 2, 271. (27) Steckel, J. S.; Yen, B. K. H.; Oertel, D. C.; Bawendi, M. G. J. Am. Chem. Soc. 2006, 128, 13032. (28) Moreels, I.; Lambert, K.; De Muynck, D.; Vanhaecke, F.; Poelman, D.; Martins, J. C.; Allan, G.; Hens, Z. Chem. Mater. 2007, 19, 6101. (29) Leschkies, K. S.; Kang, M. S.; Aydil, E. S.; Norris, D. J. J. Phys. Chem. C 2010, 114, 9988. (30) Ito, Y.; Virkar, A. A.; Mannsfeld, S.; Oh, J. H.; Toney, M.; Locklin, J.; Bao, Z. J. Am. Chem. Soc. 2009, 131, 9396. (31) Egginger, M.; Bauer, S.; Schwodiauer, R.; Neugebauer, H.; Sariciftci, N. S. Monatsh. Chem. 2009, 140, 735. 3892
dx.doi.org/10.1021/nl2020153 |Nano Lett. 2011, 11, 3887–3892
