Sensitization of p-type NiO Using n-type Conducting Polymers - The

Oct 26, 2010 - ... of Technology, Julianalaan 136, 2628 BL Delft, The Netherlands, and Holst Centre/TNO, High Tech Campus 31, 5605 KN Eindhoven, The ...
0 downloads 0 Views 2MB Size
19496

J. Phys. Chem. C 2010, 114, 19496–19502

Sensitization of p-type NiO Using n-type Conducting Polymers Sudam D. Chavhan,†,§ Ruben D. Abellon,† Albert J. J. M. van Breemen,‡ Marc M. Koetse,‡ Jorgen Sweelssen,‡ and Tom J. Savenije*,† Optoelectronic Materials Section, Department of Chemical Engineering, Delft UniVersity of Technology, Julianalaan 136, 2628 BL Delft, The Netherlands, and Holst Centre/TNO, High Tech Campus 31, 5605 KN EindhoVen, The Netherlands ReceiVed: April 15, 2010; ReVised Manuscript ReceiVed: October 7, 2010

We report on the sensitization of a p-type inorganic semiconductor, NiO, by n-type conjugated polymers. NiO thin films were deposited using RF sputtering in pure Ar (NiO A) or in Ar + O2 (90% + 10%) (NiO B). XPS and Kelvin probe measurements indicate the incorporation of oxygen in NiO B leading to the formation of electron-accepting states. Photoconductance measurements using the time-resolved microwave conductance (TRMC) technique demonstrate that these electron-accepting states are required for the hopping-like charge transport within the metal oxide. Polymer/NiO hybrid bilayers were prepared by spin-coating poly(fluorenebis(1-cyanovinylenethienylene)phenylene) (PF1CVTP) or poly(fluorene-bis(2-cyanovinylenethienylene)phenylene) (PF2CVTP) on top of NiO B. By determining the photoluminescence quenching, long exciton diffusion lengths were found amounting to 25 ( 2 nm and to 17 ( 2 nm, respectively. From I-V measurements on prototype n-type polymer/NiO hybrid solar cells carried out in the dark and under illumination, hole transfer from the polymer to the NiO is confirmed. The photovoltaic effect in cells based on PF2CVTP/NiO B bilayers was much larger than in PF1CVTP/NiO bilayers, which is related to the higher photoconductance observed in pristine PF2CVTP. It is proposed that only those photoexcitations leading to the formation of charge carriers within the bulk of the polymer film can be collected, while excitons reaching the NiO interface are lost by fast (surface) recombination. Introduction Organic solar cells are promising alternatives to conventional inorganic solar cells because of their simplicity, flexibility, and low-cost of manufacture.1-3 Power conversion efficiencies surpassing 7% have been reported in different device configurations.4-6 Despite high conversion efficiency, these systems suffer from limited lifetime because of photo-oxidation of the organic materials, decrease of the conductive properties of the interfaces, and segregation of the donor and acceptor components in the bulk heterojunction structure.7-9 One promising alternative approach is to use an n-type nanostructured inorganic semiconductor material in conjunction with a light absorbing, conducting polymer to make hybrid solar cells.10-16 In organic/ inorganic hybrid solar cells, most commonly, transparent metal oxide semiconductors such as TiO2, ZnO, or SnO2 are used as electron acceptors because of their high electron mobility, transparency, and relatively high dielectric constant, in combination with a conjugated soluble polymer used as the electron donor. Figure 1A shows the photophysical processes involved in this type of hybrid solar cell. The mechanism of charge separation is based on the generation of neutral photoexcitations in the polymer film, which can reach the interface by exciton diffusion, indicated by processes I and III, respectively. At the interface with the metal oxide, electron injection into the conduction band of the semiconductor occurs (process IV) followed by charge collection by the electrodes. Processes II * Corresponding author. E-mail: [email protected]. † Delft University of Technology. ‡ Holst Centre/TNO. § Present address: Department of Physics and Electrical Engineering, Karlstad University, Karlstad, Sweden.

and V represent loss mechanisms, i.e., nonradiative decay of the excitons and interfacial charge recombination of electrons and holes. The reverse interfacial process, i.e., electron transfer from the metal oxide to the polymer is an interesting concept, which can be used to fabricate an alternative type of hybrid solar cell (Figure 1B). This process could be realized by using a p-type metal oxide semiconductor and an n-type conjugated polymer. NiO is a p-type material due to its Ni defective structure and has a large optical band gap of ca. 3.6 eV, and a high ionization potential of 5.0 eV vs vacuum, making this a potential interesting metal oxide semiconductor.17,18 Recently, thin films of NiO are applied as contact layers in light-emitting diodes and organic solar cells as substitutes for PEDOT:PSS.19,20 The class of n-type conjugated polymers is characterized by their high electron affinity caused by, for example, the presence of substituents containing a cyano group or by the introduction of benzothiadiazole moieties in the main chain of the polymer. There are few reports on p-type dye-sensitized NiO solar cells (pDSSCs).17,21-24 Borgstrom et al. studied the hole transfer from an excited phosphorus porphyrin analogue (dye) into NiO by using transient absorption spectroscopy.21 This study suggested that recombination of the excess electron on the dye and the hole in the NiO occurs on time scales ranging from picoseconds to nanoseconds. After 20 ns, only a few percent of the initial charge-transfer states remain. Recently, Gibson et al. fabricated successfully tandem DSSCs using a NiO photocathode in combination with a TiO2 photoanode showing increased open circuit voltages by using CoII/III electrolyte.24 Figure 1B shows the working principle of the reverse electron transfer process involved in the conversion of light energy into electric energy. Photoexcitation of the n-type polymer results in the formation

10.1021/jp1033883  2010 American Chemical Society Published on Web 10/26/2010

Sensitization of p-Type NiO

J. Phys. Chem. C, Vol. 114, No. 45, 2010 19497

Figure 1. Schematic representations of the (A) electron transfer mechanism in a hybrid TiO2/P3HT bilayer, (B) reverse electron transfer mechanism in a p-type metal oxide and PF2CVTP. For PF1CVTP, the energy levels of the HOMO and LUMO are -5.93 and -3.36 eV, respectively. In addition, molecular structures of poly(fluorene-bis(1-cyanovinylenethienylene)phenylene) (PF1CVTP) and poly(fluorene-bis(2-cyanovinylenethienylene)phenylene) PF2CVTP are shown.

of excitons, which are dissociated at the interface with the NiO, yielding an excess electron in the polymer layer and a hole in the valence band of NiO. Processes II and V indicate energy loss mechanisms through nonradiative decay and interfacial recombination of electron and holes at the interface, respectively. An idea behind the above concept is to fabricate tandem hybrid solar cells, allowing improvement of the power conversion efficiency of hybrid solar cells.22,24,25 In the present work, we have studied prototype solid-state hybrid solar cells using smooth, dense layers of p-type NiO as electron donor and n-type conjugated polymers PF1CVTP, PF2CVTP (see Figure 1 for molecular structures) as electron acceptors. The energy levels of the highest occupied molecular orbital (HOMO) and lowest unoccupied molecular orbital (LUMO) with respect to vacuum levels are -5.93 and -3.36 eV for PF1CVTP and -5.72 and -3.77 eV for PF2CVTP, respectively.26 In view of the energy levels of the HOMO and the ionization potential of NiO, the driving force for hole transfer from the polymer to the NiO is estimated to be 0.93 eV for PF1CVTP and 0.72 eV for PF2CVTP. These polymers have been investigated as electron acceptors in all polymer solar cells.27,28 NiO thin films were prepared by RF-sputtering, and layers were characterized by XRD, Kelvin probe, and XPS. In addition, the photoconductance of NiO layers prepared using different sputter conditions was studied using the contactless

time-resolved microwave conductivity technique (TRMC). The exciton diffusion lengths in PF1CVTP and PF2CVTP have been determined by photoluminescence quenching experiments. Combination of the results allows the derivation of a model to explain the photovoltaic properties observed for different hybrid cells. Experimental Section Sample Preparation. Pure (99.999%) NiO target used for sputtering was purchased from AJA International Inc. (North Scituate, MA). The deposition parameters such as substrate temperature, RF-power, deposition time, and gas pressure in the deposition chamber were maintained at 100 °C, 100 W, 6 h, and 3 mTorr, respectively. The sputtering atmosphere was pure Ar or Ar (90%) + O2 (10%) mixture gas. Smooth, dense NiO samples were deposited on quartz and on fluorine-doped tin oxide (FTO) substrates. Before the deposition, the vacuum pressure in the sputtering chamber was less than 1 × 10-6 Torr, while during the deposition the pressure was kept at 3 mTorr. The sample holder was rotating at ca. 60 rpm, resulting in smooth, dense layers of NiO of ca. 85 nm. As-deposited films were used for X-ray diffraction in order to find out the structural properties of the material. X-ray photoemission spectroscopy has been carried out using a takeoff angle of 90° corresponding to an information depth of approximately 5 nm. According to

19498

J. Phys. Chem. C, Vol. 114, No. 45, 2010

Chavhan et al.

the wide-scan measurements, the elements C, and N are also present at the surface due to contamination of the NiO by air. Surface morphology and roughness of the NiO films were studied by atomic force microscopy. The surface roughness of NiO on quartz and on ITO was found to be 3 and 46 nm, respectively (see Supporting Information). Poly (fluorene - bis (2 - cyanovinylenethienylene) phenylene) (PF2CVTP) (weight-average molecular weight) (Mw ) 139.3 kg/mol, PDI ) 2.7) and poly(fluorene-bis(1-cyanovinylenethienylene)phenylene) (PF1CVTP) (Mw )11.0 kg/mol, PDI ) 2.2) were received from TNO, The Netherlands. Fifteen mg/mL solutions of PF2CVTP and PF1CVTP were prepared in 1,2dichlorobenzene, respectively, and spin-coated on top of the NiO layer at 800 rpm for 30 s and 1000 rpm for 60 s in a glovebox filled with N2. The thicknesses of the PF2CVTP and PF1CVTP layers were 90 and 86 nm, respectively, as determined by using a step profilometer (Dektak 8, Veeco). The transmission and reflection spectra of the films were recorded on a Perkin-Elmer Lambda-900 spectrophotometer equipped with an integrating sphere (Labsphere). The optical densities of the films were recorded using

( )

OD ) -log

It I0 - Ir

(1)

in which I0, It, and Ir are the incident, transmitted, and reflected intensities, respectively. The exponential optical absorption coefficient was calculated by

R)

OD ln 10 L

(2)

in which L is the thickness of the sample. Photoluminescence (PL) spectra were recorded using a Photon Technology International fluorescence meter upon excitation at λ ) 500 and 530 nm for PF1CVTP and for PF2CVTP, respectively. A detailed description of the TRMC set up has been reported elsewhere.29 In short, a thin film applied on a quartz substrate was mounted in an X-band (8.2-12 GHz) microwave cavity and subsequently photoexcited with 3 ns laser pulses (Vibrant, Opotek). The pulse intensity was measured using a Labmaster powermeter in conjunction with a pyroelectric sensor. The photoinduced electron-hole pairs in the sample result in a temporary change of the photoconductance (∆G), which leads to a change in the normalized amount of reflected microwave power (∆P/P) according to

∆G(t) ) -k

∆P(t) P

(3)

in which k is sensitivity factor of the loaded microwave cavity. The product of the charge carrier generation quantum yield, η, and the sum of the electron and hole mobilities, ∑µ, is determined from the maximum in the time-dependent photoconductance (∆Gmax) according to

η∑µ )

∆Gmax βeI0FA

(4)

Figure 2. X-ray diffraction patterns of NiO thin films grown on quartz substrates using Ar or an Ar/O2 mixture during deposition.

in which β denotes the ratio between the broad and narrow inner dimensions of the waveguide, e is the elementary charge, and I0 is the number of incident photons per unit area per pulse. Current-voltage (I-V) characteristic were recorded using a potentiostat (Autolab PGSTAT 10 EcoChemie) controlled by GPES 3 software in combination with a xenon lamp (Eurosep Instruments) equipped with a monochromator (Triax). The incident power was 11 mW/cm2 at the 580 nm wavelength. IPCE spectra were measured using a Keithley 2700 multimeter. The incident photon to current efficiency (IPCE) was calculated from the wavelength-dependent short circuit current (Isc) using

IPCE )

Isc hc × 100% Pin eλ

(5)

where Pin represents the power of the incident light, h is Planck’s constant, and c is the velocity of light. Results and Discussion Structural Study. Figure 2 shows the X-ray diffraction patterns of NiO thin films, deposited by rf sputtering in pure argon (NiO A) or in a gas mixture of Ar + O2 (NiO B) using a 100 °C quartz layer as substrate. From the main diffraction peak it is inferred that both samples have a (200) preferred orientation. The peaks of NiO A corresponding to the (111), (200), and (220) planes confirm the cubic structure of NiO, which is in correspondence to the JCPDS card. no.78-0643. Note that the diffraction peak present at 2θ ) 39.12° represents the hexagonal phase of nickel trioxide (Ni2O3) (JCPDS card no. 14-0481), which suggests that Ni deficiencies are present. The diffraction peaks of NiO B are slightly shifted toward lower angles (Figure 2). The lattice constants were calculated from the X-ray diffraction patterns, and amounted to a ) b ) c ) 4.16 Å for NiO A and 4.22 Å for NiO B. Films prepared in the presence of oxygen are expected to contain more oxygen than the films prepared in Ar, which causes the formation of point defects within the film. Hence, the microcrystallites experience residual stress, shifting the peaks of sample B toward smaller angles as reported previously.30 To confirm the elemental composition of the NiO films, X-ray photoelectron spectroscopy was carried out (see Supporting Information). From the observed peaks, the binding energies of Ni 2p3/2 and Ni 2p1/2 levels in NiO were found to be 856.40 and 874.42 eV, respectively, and an O1s peak was detected at

Sensitization of p-Type NiO

J. Phys. Chem. C, Vol. 114, No. 45, 2010 19499

Figure 4. The luminescence quenching expressed in percentage in PF1CVTP/NiO B and PF2CVTP/NiO B bilayers for different polymer film thicknesses. The full lines represent fits to the experimental data points using exciton diffusion lengths of 25 and 17 nm for PF1CVTP and PF2CVTP polymers, respectively.

Figure 3. (A) Absorption spectra of a 76 nm thick film of PF1CVTP (triangle) and a bilayer of PF1CVTP/NiO B (circle) and corresponding PL spectra recorded upon excitation at 500 nm. Absorption spectra of bare NiO A (star) and of NiO B (square) are included. (B) Absorption spectra of a 80 nm thick film of PF2CVTP (triangle) and a PF2CVTP/ NiO B bilayer (circle) and corresponding PL spectra recorded upon excitation at 530 nm.

527.98 eV, which is in accordance with reported data31 (see Supporting Information for spectra). The concentration depth profiles show a constant O/Ni ratio for both samples except for the first few nanometers, which is due to some surface contamination. Additionally, it is inferred that NiO B contains 1.35% more oxygen than that of the NiO A, which is in accordance with our expectations. To determine the Fermi energy level of both NiO samples, Kelvin probe measurements were performed, yielding levels of 4.75 and 5.01 eV for the NiO A and NiO B layer, respectively. This shows that the incorporation of oxygen as in the case of sample NiO B leads to the formation of acceptor levels, which lowers the Fermi level. Optical Study. The absorption spectra of PF1CVTP and PF2CVTP on quartz and on NiO B are shown in Figures 3A and 3B, respectively. The absorbance in the range of 400 to 600 nm is mainly due to the conjugated polymers PF1CVTP and PF2CVTP as reported previously.30 The absorption coefficients (R) of PF1CVTP, PF2CVTP were calculated from their optical densities and corresponding thicknesses by using eq 2 and were found to be 1.0 × 107 m-1 at 460 nm, 1.4 × 107 m-1 at 540 nm, respectively. The spectra of both bare NiO layers are shown in Figure 3A. For both samples, the absorption below approximately 380 nm increases sharply, which is attributed to the direct band gap transition of NiO. Note that the absorption in the visible region is much stronger for NiO B than for NiO A. This is explained by the fact that NiO B is prepared in the presence of O2 and hence contains intraband gap states as

discussed above. These states are responsible for the enhancement of the optical absorption of NiO B in the visible region.30,32-35 A direct relationship between the percentage of oxygen in the deposition chamber and the optical absorption in the visible region has been reported previously.30,34 For NiO A, only Ar gas is used in the deposition chamber; hence, the film contains less excess oxygen. As a result, the absorption of the NiO A in the visible region is low since only localized ‘d’ orbital electrons of Ni contribute to the absorption.33 The photoluminescence (PL) spectra of PF1CVTP and PF2CVTP spin-coated on a quartz substrate or on NiO B are shown in Figure 3A and 3B. Maximum PL was observed at 580 nm and at 630 nm for PF1CVTP and PF2CVTP, respectively. Clearly, the PL of the polymer in the bilayer structures is strongly quenched with respect to the PL of the polymers on the quartz substrates. Hence, the quenching of excitons within the polymer can be attributed to a process occurring at the interface with the NiO. It is important to note that a similar quenching was observed when the polymer was applied on a NiO A layer. From this observation it is inferred that the quenching is not due to energy transfer between polymer and the NiO layer but via another mechanism, e.g., charge transfer. The extent to which excitons are quenched is determined by the exciton diffusion length (process III in Figure 1B). To determine this distance, the maximum luminescence of a number of polymer samples differing in thickness applied on quartz and on a NiO substrate was determined. The quenching efficiency (Q) was calculated by,

Q)1-

PLpol/NiO PLpol/quartz

(6)

Figure 4 shows the Q values as function of the polymer thickness for both polymers. A fit to the Q data was made using eq 7 to obtain the exciton diffusion length (LD).36,37

Q) [R2LD2 + RLD tanh(L/LD)]exp(-RL) - R2LD2[cosh(L/LD)]-1 (1 - R2LD2)[1 - exp(-RL)]

(7)

19500

J. Phys. Chem. C, Vol. 114, No. 45, 2010

Figure 5. (A) Intensity-normalized photoconductance transients observed upon excitation at 580 nm of PF2CVTP film (red) and bare NiO A (black) (intensity: 2.2 × 1015 photons/cm2 pulse), and NiO B (blue) and PF2CVTP-coated NiO B film (green). (intensity: 2.1 × 1014 photons/cm2). (B) Intensity-normalized photoconductance transients for a PF1CVTP film upon excitation at 460 nm (blue), and for a PF2CVTP film (red) upon excitation at 580 nm (intensity: 2.2 × 1015 photons/ cm2 pulse).

in which R is the absorption coefficient and L is the thickness of the polymer layer. The curves are fits to the data points using LD ) 17 ( 2 nm and LD ) 25 ( 2 nm for PF2CVTP and PF1CVTP, respectively. The found diffusion lengths are significantly larger than the values reported for alkyl-substituted poly(phenylvinylene) 7 nm,38 PPV39 12 nm, and ladder-type poly(p-phenylene) 14 nm.40 This might be due to the substantially higher electron affinity of the energy level of the LUMO of these n-type polymers, reducing the possibility for the excitons to decay via, for example, charge transfer with impurities, such as oxygen and moisture.41 Photoconductance Study. Figure 5A shows the conductance transients on photoexcitation at 580 nm for PF2CVTP, NiO A, NiO B, and PF2CVTP/NiO B bilayers normalized with incident intensity (I0). The photoconductance of the bare PF2CVTP layer is small because of the low dielectric constant of the organic material, resulting in a high exciton binding energy. Hence, lightinduced charge carrier generation within the pure polymer layer is not efficient explaining the low signal. In case of NiO A, the conductance on excitation at 580 nm is below the limit of detection, even on direct band gap excitation at 300 nm. In contrast, NiO B shows an appreciable conductance transient upon excitation using a 580 nm laser pulse. The striking difference can be explained by the higher oxygen concentration in the NiO B sample. As already mentioned above, to maintain charge neutrality, excess interstitial oxygen is maintained by

Chavhan et al. the formation of Ni3+ ions, which lead to the formation of electron-accepting, intraband gap states. The TRMC results for NiO sputtered using 10% oxygen (NiO B) confirms our statements that excess oxygen is needed for the hopping-like transport in NiO. For NiO prepared in 5% oxygen, a lower TRMC signal is found as compared with NiO B in agreement with the explanation above (see Supporting Information). A number of studies suggest that photoexcitation leads to the population of these electron-accepting states and that charge transport occurs by hopping of these electrons to neighboring acceptor sites. Our findings are in line with temperature- and wavelength-dependent conductivity measurements on various NiO samples prepared in 100% Ar atmosphere or in mixtures of Ar with O2.30,42 The absence of any detectable photoconductance signal as in case of NiO A makes this type of layer unsuitable for application in devices. The maximum TRMC signal obtained for the PF2CVTP/NiO B bilayer is 1.6 times larger than for NiO B as shown in Figure 5A. Both transients show an initial rise due to the 18 ns response time of the microwave cell. After reaching a maximum at approximately 30 ns after the laser pulse, the photoconductance signals start to decay due to the trapping and/or recombination of charge carriers. Both decays follow approximately a power law. The half-life times of the decays are found to be 0.6 and 0.23 µs for the PF2CVTP/NiO B bilayer and for NiO B, respectively. Upon comparison of the TRMC signals of PF2CVTP with the PF2CVTP/NiO B bilayer, it is apparent that for the latter, the maximum change in conductance is much larger. Note that the TRMC signal of the bilayer cannot be reconstituted by a summation of the TRMC signals of the individual layers. In short, the higher conductance and longer lifetime of the TRMC signal for the bilayer is indicative for interfacial hole transfer from the PF2CVTP to the NiO layer (process IV in Figure 1B). Hence, these experiments give direct evidence for sensitization of the metal oxide by the polymer. From the maximum of the photoconductance transient observed upon low excitation energies, a value of η∑µ of 7.2 × 10-3 cm2/(V s) was calculated using eq 4. The η∑µ value represents the product of quantum yield for free charge carrier generation per incident photon (η) and the sum of the mobilities of electrons and holes (∑µ). For NiO mobilities have been reported varying from 5 cm2/(V s) down to 0.2 cm2/(V s).43-46 The differences can be attributed to differences in crystallinity, temperature, and dopant concentration. However, it is clear from our measurements that the mobility in the prepared NiO samples is smaller and/or the charge carrier generation is less than unity. Figure 5B shows the TRMC signals normalized to the incident intensity observed upon excitation of PF1CVTP and PF2CVTP on a quartz substrate. Note the different scales as compared to Figure 5A. Clearly, the photoconductance in PF2CVTP is much larger than in PF1CVTP. Photovoltaic Study. Prototype hybrid solar cells, made of NiO B and n-type conjugated polymer, were prepared on FTO substrates while using mercury as a back contact. The I-V curves of FTO/NiO B/Hg and FTO/PF2CVTP/Hg are shown in Figure 6, and the symmetric nature of the curves indicate ohmic contacts for both electrodes. For FTO/NiO A/Hg, the observed current was more than 1 order of magnitude smaller, which is in agreement with the TRMC measurements. For FTO/ PF1CVTP/Hg devices, symmetric I-V curves were recorded similar to the FTO/PF2CVTP/Hg devices. The photovoltaic properties of bilayer devices consisting of FTO/NiO B(85 nm)/ PF2CVTP (90 nm)/Hg were studied in the dark and under illumination from monochromatic light at 580 nm and an

Sensitization of p-Type NiO

J. Phys. Chem. C, Vol. 114, No. 45, 2010 19501

Figure 6. I-V characteristics for FTO/NiO B/Hg, FTO/PF2CVTP/Hg in the the dark and FTO/NiO B/PF2CVTP/Hg in the dark and under monochromatic excitation at 580 nm and an intensity of 11 mW/cm2. Inset shows the same I-V curves of FTO/NiO B/PF2CVTP/Hg in lin-lin representation.

incident power of 11 mW/cm2 (Figure 6). The dark curve has a symmetric nature, which indicates that no heterojunction is formed at the interface between NiO B and the PF2CVTP layers. This is in contrast with previous I-V measurements of p-type polymers on n-type TiO2.16 Lack of free mobile charge carriers in the bilayer could be the cause of the absence of a depletion layer. Upon illumination, an open-circuit voltage (Voc) of 198 mV, a short-circuit current density of 0.023 mA cm-2 and a fill factor of 0.33 were found. Note that the FTO is positive under illumination, confirming hole transfer from the polymer to the NiO. Upon lowering the thickness of the polymer layer, the Voc decreases gradually, while increasing the polymer layer results in a lower fill factor. The low open circuit voltage might be explained by the fact that no heterojunction is formed between the n-type polymer and NiO B, and the resulting potential is solely determined by the differences in workfunction of both contacts (Hg: 4.5 eV; FTO: 4.7 eV).47 However, the low Voc may also originate from other causes such as fast recombination reactions at the NiO/polymer interface or a low injection efficiency. By changing the metal backcontact, it is possible to elucidate the actual cause for the low Voc and improve this Voc value.48 An even lower photovoltaic effect was observed using cells based on PF1CVTP/NiO B. The incident photon to current conversion efficiency (IPCE) spectra of an FTO/NiO B/PF2CVTP/Hg shows two maxima amounting to 1.5% (380 nm) and ca. 0.5% (580 nm) corresponding to the absorption of the NiO and the polymer layer, respectively. In the next section, we analyze the results from the measurements described above in more detail to elucidate the mechanism involved in the sensitization of NiO. From the PL quenching experiments, an LD of 17 and of 25 nm for PF2CVTP and PF1CVTP, respectively, were found. Furthermore, the transmission of the glass/FTO/NiO substrate (IT(FTO,NiO)) was determined. The fraction of the incident light absorbed by the polymer within a distance LD from the interface with the NiO can be approximated by:

ILD ) IT(FTO,NiO)(1 - 10-RLD)

(8)

in which R is the absorption coefficient of the polymer. For PF2CVTP, upon excitation at 580 nm ILD is 0.23 and for

PF1CVTP on excitation at 460 nm ILD is 0.30. If at the interface between polymer and NiO excitons decay by hole transfer with 100% efficiency, these fractions represent the maximum attainable IPCE values. The observed values, however, are much smaller and for the device based on PF2CVTP around 1%. For the other polymer/NiO bilayer, this value is even lower. The conclusion we can draw from these large discrepancies is that most of the excitons formed within LD from the NiO interface do not yield long-lived charge carriers, which can be collected. This means that instead of decay by hole transfer, excitons recombine mediated by, for example, surface states on the NiO. Alternatively, it is possible that directly after hole transfer, the charge carriers recombine (process V in Figure 1B). Both explanations are plausible in view of the fact that NiO B contains many intraband gap states and are in agreement with conclusions drawn by Borgstrom et al.21 Recently, also another type of sensitization for P3HT/ZnO was reported, starting with the generation of free charge carriers within the bulk of the polymer layer.49 Subsequently, the mobile electrons are collected at the interface by the ZnO, resulting in the photovoltaic effect observed in those devices. For the PF2CVTP/NiO B devices, a similar mechanism could be applicable. Upon excitation of the pure polymer layers, mobile charge carriers are to some extent formed. Upon diffusion of these photogenerated charge carriers, holes can be collected by the NiO at the interface. This model is consistent with the fact that free mobile charge carriers in pure layers of PF2CVTP can be generated as demonstrated in Figure 5B. However, this explanation might seem incompatible with the fact that the difference between the TRMC signal of the PF2CVTP/NiO B bilayer and of the NiO B is much larger than that of the bare PF2CVTP layer. We explain this difference by assuming that the mobility of charge carriers in the NiO layer is substantially larger than in the polymer layer. The model presented above provides an explanation for the much smaller photovoltaic effect found for the PF1CVTP/NiO B devices. As shown in Figure 5B, the photoconductance observed in a pure PF1CVTP layer is much smaller than in PF2CVTP. Hence, less positive charge carriers can be collected by the NiO, yielding a smaller photovoltaic effect.

19502

J. Phys. Chem. C, Vol. 114, No. 45, 2010

Conclusions In this work we studied the sensitization of NiO using n-type polymers. From our measurements we can conclude the following issues: • Photoconductance can only be observed in NiO samples prepared in the presence of O2 which is responsible for the formation of Ni3+ ions. These lead to the formation of intraband gap states near the valence band and a lowering of the Fermi level. • We observe long exciton diffusion lengths within both n-type polymers, which can be related to the high electron affinity of these materials. However, decay of excitons at the interface with conductive NiO does not lead to long-lived charge separation. • Sensitization of NiO by PF2CVTP likely starts with the photogeneration of mobile charge carriers within the bulk of the polymer followed by hole transfer at the interface. To prevent exciton decay back to the ground state at the interface of the polymer with NiO it might be of interest to use only a nanometer sized thin intrinsic NiO layer on top of a doped NiO layer. We believe that this work will motivate for the fabrication of hybrid tandem solar cells in conjunction with a subcell consisting of an n-type metal oxide to achieve high conversion efficiencies, which will help to commercialize hybrid solar cells. Acknowledgment. This work was supported by Senter Novem. D. H. Murthy is acknowledged for recording part of the TRMC measurements. Supporting Information Available: Additional information as noted in the text. This material is available free of charge via the Internet at http://pubs.acs.org. References and Notes (1) Halls, J. J. M.; Walsh, C. A.; Greenham, N. C.; Marseglia, E. A.; Friend, R. H.; Moratti, S. C.; Holmes, A. B. Nature 1995, 376, 498–500. (2) Yu, G.; Gao, J.; Hummelen, J. C.; Wudl, F.; Heeger, A. J. Science 1995, 270, 1789–1791. (3) Mayer, A. C.; Scully, S. R.; Hardin, B. E.; Rowell, M. W.; McGehee, M. D. Mater. Today 2007, 10, 28–33. (4) Green, M. A.; Emery, K.; Hishikawa, Y.; Warta, W. Prog. PhotoVoltaics 2009, 17, 320–326. (5) Kim, J. Y.; Lee, K.; Coates, N. E.; Moses, D.; Nguyen, T. Q.; Dante, M.; Heeger, A. J. Science 2007, 317, 222–225. (6) Liang, Y.; Xu, Z.; Xia, J.; Tsai, S.-T.; Wu, Y.; Li, G.; Ray, C.; Yu, L. AdV. Mater. 2010, 22, E135-E138. (7) Pacios, R.; Chatten, A. J.; Kawano, K.; Durrant, J. R.; Bradley, D. D. C.; Nelson, J. AdV. Funct. Mater. 2006, 16, 2117–2126. (8) Jeranko, T.; Tributsch, H.; Sariciftci, N. S.; Hummelen, J. C. Sol. Energy Mater. Sol. Cells 2004, 83, 247–262. (9) Schuller, S.; Schilinsky, P.; Hauch, J.; Brabec, C. J. Appl. Phys. A: Mater. Sci. Process. 2004, 79, 37–40. (10) Salafsky, J. S. Phys. ReV. B 1999, 59, 10885–10894. (11) Arango, A. C.; Carter, S. A.; Brock, P. J. Appl. Phys. Lett. 1999, 74, 1698–1700. (12) Beek, W. J. E.; Wienk, M. M.; Janssen, R. A. J. AdV. Mater. 2004, 16, 1009–+. (13) Coakley, K. M.; Liu, Y. X.; McGehee, M. D.; Frindell, K. L.; Stucky, G. D. AdV. Funct. Mater. 2003, 13, 301–306.

Chavhan et al. (14) Ravirajan, P.; Haque, S. A.; Durrant, J. R.; Bradley, D. D. C.; Nelson, J. AdV. Funct. Mater. 2005, 15, 609–618. (15) Oosterhout, S. D.; Wienk, M. M.; van Bavel, S. S.; Thiedmann, R.; Koster, L. J. A.; Gilot, J.; Loos, J.; Schmidt, V.; Janssen, R. A. J. Nat. Mater. 2009, 8, 818–824. (16) Savenije, T. J.; Warman, J. M.; Goossens, A. Chem. Phys. Lett. 1998, 287, 148–153. (17) He, J. J.; Lindstrom, H.; Hagfeldt, A.; Lindquist, S. E. J. Phys. Chem. B 1999, 103, 8940–8943. (18) Odobel, F.; Le Pleux, L.; Pellegrin, Y.; Blart, E. Acc. Chem. Res. 2010, 43, 1063–1071. (19) Irwin, M. D.; Buchholz, B.; Hains, A. W.; Chang, R. P. H.; Marks, T. J. Proc. Natl. Acad. Sci. U.S.A. 2008, 105, 2783–2787. (20) Long, H.; Fang, G. J.; Huang, H. H.; Mo, X. M.; Xia, W.; Dong, B. Z.; Meng, X. Q.; Zhao, X. Z. Appl. Phys. Lett. 2009, 95, 013509. (21) Borgstrom, M.; Blart, E.; Boschloo, G.; Mukhtar, E.; Hagfeldt, A.; Hammarstrom, L.; Odobel, F. J. Phys. Chem. B 2005, 109, 22928–22934. (22) Nakasa, A.; Usami, H.; Sumikura, S.; Hasegawa, S.; Koyama, T.; Suzuki, E. Chem. Lett. 2005, 34, 500–501. (23) Bandara, J.; Weerasinghe, H. Sol. Energy Mater. Sol. Cells 2005, 85, 385–390. (24) Gibson, E. A.; Smeigh, A. L.; Le Pleux, L.; Fortage, J.; Boschloo, G.; Blart, E.; Pellegrin, Y.; Odobel, F.; Hagfeldt, A.; Hammarstrom, L. Angew. Chem., Int. Ed. 2009, 48, 4402–4405. (25) He, J. J.; Lindstrom, H.; Hagfeldt, A.; Lindquist, S. E. Sol. Energy Mater. Sol. Cells 2000, 62, 265–273. (26) Veldman, D.; Meskers, S. C. J.; Janssen, R. A. J. AdV. Funct. Mater. 2009, 19, 1939–1948. (27) Koetse, M. M.; Sweelssen, J.; Hoekerd, K. T.; Schoo, H. F. M.; Veenstra, S. C.; Kroon, J. M.; Yang, X. N.; Loos, J. Appl. Phys. Lett. 2006, 88, 083504. (28) Quist, P. A. C.; Sweelssen, J.; Koetse, M. M.; Savenije, T. J.; Siebbeles, L. D. A. J. Phys. Chem. C 2007, 111, 4452–4457. (29) Kroeze, J. E.; Savenije, T. J.; Warman, J. M. J. Am. Chem. Soc. 2004, 126, 7608–7618. (30) Nandy, S.; Saha, B.; Mitra, M. K.; Chattopadhyay, K. K. J. Mater. Sci. 2007, 42, 5766–5772. (31) Mansour, A. N. Surf. Sci. Spectra 1994, 3, 231–238. (32) Newman, R.; Chrenko, R. M. Phys. ReV. 1959, 114, 1507–1513. (33) Adler, D.; Feinleib, J. Phys. ReV. B 1970, 2, 3112–3134. (34) Lu, Y. M.; Hwang, W. S.; Yang, J. S.; Chuang, H. C. Thin Solid Films 2002, 420, 54–61. (35) Jang, W. L.; Lu, Y. M.; Hwang, W. S.; Hsiung, T. L.; Wang, H. P. Appl. Phys. Lett. 2009, 94, 062103. (36) Markov, D. E.; Amsterdam, E.; Blom, P. W. M.; Sieval, A. B.; Hummelen, J. C. J. Phys. Chem. A 2005, 109, 5266–5274. (37) Markov, D. E.; Hummelen, J. C.; Blom, P. W. M.; Sieval, A. B. Phys. ReV. B 2005, 72, 045216. (38) Markov, D. E.; Blom, P. W. M. Phys. ReV. B 2006, 74, 085206. (39) Stubinger, T.; Brutting, W. J. Appl. Phys. 2001, 90, 3632–3641. (40) Haugeneder, A.; Neges, M.; Kallinger, C.; Spirkl, W.; Lemmer, U.; Feldmann, J.; Scherf, U.; Harth, E.; Gugel, A.; Mullen, K. Phys. ReV. B 1999, 59, 15346–15351. (41) McCulloch, I.; Heeney, M.; Chabinyc, M. L.; DeLongchamp, D.; Kline, R. J.; Coelle, M.; Duffy, W.; Fischer, D.; Gundlach, D.; Hamadani, B.; Hamilton, R.; Richter, L.; Salleo, A.; Shkunov, M.; Sporrowe, D.; Tierney, S.; Zhong, W. AdV. Mater. 2009, 21, 1091–1109. (42) Tsu, R.; Esaki, L.; Ludeke, R. Phys. ReV. Lett. 1969, 23, 977–&. (43) Roilos, M.; Nagels, P. Solid State Commun. 1964, 2, 285–290. (44) Bosman, A. J.; Crevecoe, C. Phys. ReV. 1966, 144, 763–&. (45) Tallan, N. M.; Tannhaus, Ds. Phys. Lett. A 1968, A 26, 131–&. (46) Spear, W. E.; Tannhaus, Ds. Phys. ReV. B 1973, 7, 831–833. (47) Sze, S. M. Physics of semiconductor deVices, 2nd ed.; John Wiley &Sons: New York, 1981; Vol. 2. (48) Blom, P. W. M.; Mihailetchi, V. D.; Koster, L. J. A.; Markov, D. E. AdV. Mater. 2007, 19, 1551–1566. (49) Piris, J.; Kopidakis, N.; Olson, D. C.; Shaheen, S. E.; Ginley, D. S.; Rumbles, G. AdV. Funct. Mater. 2007, 17, 3849–3857.

JP1033883