Photophysical Behaviors at Interfaces between Poly(3-Hexylthiophene) and Zinc Oxide Nanostructures

Abstract

Development of design rules through understanding exciton dissociation and charge generation dynamics is an important pathway to improve the performance of hybrid inorganic–organic solar cells. In this work, we study the photophysical behavior and dynamics of charge generation in organic/inorganic hybrid based on poly(3-hexylthiophene) (P3HT) and zinc oxide (ZnO) nanostructures by steady-state and time-resolved fluorescence spectroscopy. Both films of P3HT:ZnO nanoparticle (NPs) blend as the active layer and pristine P3HT as the active layer on ZnO nanorod (NRs) array buffer layer can provide efficient interfaces for exciton dissociation and charge transfer. P3HT can infiltrate into inter-rod space of ZnO NRs and cover them to enhance the 0-0 band against the 0-1 one in the steady-state photoluminescence spectrum. However, the configuration of P3HT: ZnO NPs blend as the active layer on ZnO NRs buffer layer poorly further improves the exciton dissociation, due to the decreased organic-inorganic interface area.

1. Introduction

Organic-inorganic hybrid nanocomposites have attracted intensive interest due to a great potential optoelectronic application including light emitting diodes (OLED), field effect transistor (FET) and organic solar cells (OSC). Such hybrid systems, composed of conjugated polymers and inorganic semiconductor nanostructures, can presumably offers multitudes of performance by integrating advantages of single-component counterparts^1–3). Poly(3-hexylthiophene) (P3HT) is a conventional conjugated polymer applied in OSC as donor in the active layer^4,5). As a candidate of inorganic semiconductor, ZnO is famous for its multiple nanostructures, which have high carrier mobility and ambient stability^6,7).

Photoinduced electron transfer occurs at the interface between the polymer (donor) and the inorganic semiconductor (acceptor) and results in efficient charge separation. In priori studies, generally there are two categories of P3HT and ZnO interfaces: I. ZnO nanostructures blended with P3HT as the active layer; II. P3HT as active layer on top of ZnO nanostructures as buffer layer^7,8). For applications in solar cells as well as others mentioned above, a detailed investigation of the photophysical behavior of P3HT upon excitation is a critical research point for optimizing device functions. Characterization of the fluorescence decay behavior is fundamentally important in understanding the charge generation process in conjugated polymer-based organic devices.

In the present work we report the use of femtosecond time-resolved photoluminescence (PL) spectroscopy to study the relaxation dynamics of the excited states and the decay mechanisms of photoinduced charge carriers for P3HT/ZnO hybrid systems. Several types of P3HT:ZnO interfaces were investigated, based on P3HT and P3HT:ZnO NPs blend as films on glass or on ZnO NRs array, respectively. The dynamics of charge-carrier relaxation at different P3HT: ZnO interfaces have been investigated in comparison with that in pristine P3HT, revealing that our incorporation of ZnO NPs into P3HT or ZnO NRs under P3HT are both able to increase the efficiency of exciton dissociation as long as the organic-inorganic phase separation is appropriate.

2. Experimental

ZnO NPs were fabricated according to Ref. 7). ZnO NRs array was fabricated according to Ref. 9). P3HT (>90% regioregular) was purchased from Sigma-Aldrich and dissolved in chloroform at a concentration of 5 mg/ml. P3HT solutions were stirred overnight. Then ZnO NPs were added into the previously prepared P3HT solutions. The blend solutions were stirred for five hours and then ultrasonicated (50 W, 43 kHz) for 1.5 hour in order to get a homogeneous solution and prevent any agglomeration taking place¹⁰⁾. Films of P3HT and ZnO NPs with a weight ratio of 1:0 and 1:0.3 were prepared by spin-coating the solutions on top of glass and ZnO NRs array, respectively, at a speed of 1000 rpm for 50 s. The optical absorption spectra were collected using a UV–visible dual-beam spectrophotometer (TU-1900, PG Instruments, Ltd., Beijing, China). Steady-state PL measurements were performed with a PG2000-Pro-EX spectrometer with the excitation wavelength of 400 nm. Time-resolved fluorescence characterization was performed using the fluorescence up-conversion technique^11–14).

Tested solar cell devices were prepared by spin-coating P3HT (or blended P3HT: ZnO NPs) solutions directly onto ITO glass or to ZnO NRs array which grew on ITO glass. Then films were annealed at 150℃ for 10 min inside a glove box. The devices were tested right after preparation. Current density–voltage (J–V) characteristics of cells were measured at room temperature under AM1.5G illumination of 100 mW/cm² provided by a built-in solar simulator (SAN-EI Electric XEC-301S).

3. Results and Discussion

Four film samples were fabricated to design diverse interfaces between P3HT and ZnO as illustrated in Fig. 1(a): Sample A, film of pristine P3HT; Sample B, film of P3HT: ZnO NPs blend; Sample C, film of pristine P3HT on ZnO NRs array; and Sample D: film of P3HT: ZnO NPs blend on ZnO NRs array. The ZnO NRs is wurtzite and its growth orientation is (002). The size of the ZnO NPs is about 4 nm, and the diameter of the cross section is c.a. 30 nm for ZnO NRs^8,9). In Sample C and D, P3HT films were spun on ZnO NRs array and may infiltrate into the space between NRs (see Fig. 1(b)). P3HT films with and without blended ZnO NPs on top of ZnO NRs arrays are shown in Fig. 1(c). The layer thickness of the samples is in the range of 150–230 nm since the spin-coating speed is the same for them.

Fig. 1

(a) Schematic illustration of four samples. (b) Top-view SEM images of ZnO NRS array before (right) and after (middle) the coating of P3HT, and cross-sectional-view SEM image of ZnO NRS array with P3HT infiltrated into the space between NRSs (right). (c) Top-view and cross-sectional-view SEM images of Sample C (left) and D (right) films on top of ZnO NRS array.

The UV–visible absorption spectra of the four samples are depicted in Fig. 2(a). The spectrum of the pristine P3HT film exhibits the characteristic absorption peak at 520 nm due to the π-π* electronic transitions in P3HT. The two shoulders near 550 nm and 600 nm are ascribed to the vibronic absorption of the ordered P3HT crystalline regions in the films¹⁵⁾. The absorption performance of P3HT films depends on the highly textured and oriented crystallites formed by the self-organized region-regular polymer, which possesses highly ordered packing, alignment, strong interchain and interlayer interactions. The absorption peak shifts to 540 nm when 30 mass% ZnO NPs were added. Since the absorption of P3HT consists of the (0-1) transition with shorter-wavelength and the (0-0) transition with longer-wavelength, the variation of 20 nm can be regarded to result from the changed relative intensities of the two transitions. According to Nuzzo et al.¹⁶⁾, the red-shift of the absorption peak is associated with the reduction of torsional defects. Hence, the larger composition of the longer-wavelength sideband in the P3HT:ZnO NPs hybrid indicates that more ordered structures with less torsional defects are formed in the hybrid than in the pristine P3HT.

Fig. 2

UV–visible absorption spectra (a) and normalized steady-state PL spectra (b) for the four samples. The inset of (a) is the absorption spectra of ZnO NRS array and ZnO NPS. (c) Intensity ratio between 0-0 and 0-1 sideband for all the samples.

The films on ZnO NRs array show strong absorption in UV region. The significant absorption below 400 nm corresponds to the absorption of ZnO, showing a typical trailing edge of ZnO near 400 nm. The inset of Fig. 2(a) shows strong absorption tail in the short-wavelength range for ZnO NRs array and a weak one for ZnO NPs. The curves for Sample C and D are both superpositions of the absorption of P3HT and ZnO, indicating there is no electron transfer between P3HT and ZnO on ground state but only on excited state. The characteristics of ZnO is more prominent as its composition ratio increases since ZnO NPs and NRs both exist in Sample D.

The fluorescence excited by the laser at the wavelength of 400 nm is only from electronic transitions of P3HT not from ZnO, because 400 nm (3.1 eV) is lower than the energy band gap of ZnO (3.37 eV at room temperature). The higher energy peak at 650 nm and the lower energy one at 710 nm in the steady-state PL spectra plotted in Fig. 2(b) are attributed to (0-0) and (0-1) transitions, respectively, according to the Franck-Condon progression model¹⁷⁾ (see the inset of Fig. 2(b)). From the PL normalized to the 0-1 peak, the intensity ratio between the (0-0) and (0-1) sideband, as demonstrated in Fig. 2(c), increases from 0.82 for pristine P3HT film (Sample A) to 0.83 for P3HT:ZnO NPs blend film (Sample B). The ratio increases significantly to 0.94 when pristine P3HT film is on top of ZnO NRs array (Sample C). This number goes down to 0.85 with respect to P3HT: ZnO NPs blend film on ZnO NRs array (Sample D).

There are H-type (face-to-face) and J-type (head-to-tail) aggregates for P3HT^18,19). While only 0-1 sideband appears in the former type due to the constraint of symmetry, the transition corresponding to 0-0 sideband is strongly allowed in the latter type. Baghgar proposed a theory that such 0-0 to 0-1 sideband ratio would rise from H- to J- aggregates as the coil chain conformation of polymer becomes more prevailing compared to the rod-like chain²⁰⁾. In our study, as illustrated in Fig. 3, P3HT chain remains its rod-like shape when it is blended with ZnO NPs, but it becomes increasingly coiled when infiltrating into the accessible inter-rod space between individual ZnO NRs. P3HT self-organizes into a coiled chain and covers ZnO NRs, which results in the conspicuous rise in the intensity ratio I_0-0/I_0-1 in Sample C. Similar phenomena were found in blended nanocomposites of P3HT and multi-walled carbon nanotubes^21,22). The incorporation of ZnO NPs in Sample D is hypothesized to hamper the contact of P3HT and ZnO NRs, ruining part of the coverage conformations and imparting the fall of I_0-0/I_0-1 ratio.

Fig. 3

Schematic representation of micro morphological structures in the four samples.

The un-normalized emission spectra of pure P3HT and its blend films with ZnO nanostructures are presented in Fig. 4(a) to show the P3HT fluorescence quenching effect by ZnO as a result of charge carrier transfer from P3HT to ZnO nanostructures. The dynamic photophysical processes were investigated by femtosecond fluorescence upconversion technique (see Fig. 4(b)). The samples were excited at 400 nm and probed at 650 nm. Fluorescence decay profiles of pristine P3HT and P3HT/ZnO hybrid systems are shown in Fig. 4(c). The instrumental response function (IRF) was measured to be 400 fs, as shown in the inset. The blending of P3HT with ZnO NPs and P3HT (or P3HT: ZnO NPs blend) on ZnO NRs array all make excitons dissociate more effectively, by showing a shorter fluorescence lifetime in Sample B, C and D compared to that of Sample A. The decreased lifetimes demonstrate the charge transfer indeed occurs from P3HT to ZnO. Exciton dissociation and charge transfer is more efficient when the donor P3HT is in contact of the acceptor ZnO. The fitted decay times are shown in Table 1. The average lifetime of pristine P3HT film is 226 ps, which is in agreement with previous work^23,24). The average fluorescence lifetime of the P3HT/ZnO hybrid decreased to 67 ps for Sample B and 65 ps for Sample C. The compositional ratio of ZnO NPs of Sample B was optimized to be 30%, which is limited by the solubility of ZnO NPs in P3HT. It was observed that the fluorescence decayed faster as the concentration of ZnO NPs increases in Sample B, consistent with the results in Ref. 7).

Fig. 4

(a) Un-normalized steady-state PL spectra of the four samples. (b) Schematic diagram of the experimental setup for fluorescence up-conversion measurement. (c) Fluorescence decay profiles of pristine P3HT and P3HT:ZnO NPS blend films on glass and on ZnO NRS array respectively. The inset is the IRF of the experimental setup. (d) Normalized kinetic fluorescence decays for a P3HT thin film following excitation at 400 nm probing at 650 nm as a function of excitation laser power. Light excitation was at ～1.4 × 10²¹, 2.6 × 10²¹ and 5 × 10²¹ incident photons per cm² at 400 nm for the blue, red and black lines respectively. The left inset shows normalized fluorescence decay curves from 0 to 400 ps. The right inset exhibits un-normalized kinetic decays.

Table 1 Fitted decay times and relative intensities of P3HT film and P3HT/ZnO blend film on glass and ZnO NRS array.

Film	Fitted decay times (ps)
	τ1	τ2	τ3	τavg
P3HT	3.48 (33.4%)	40.2 (29.4%)	575 (37.2%)	227
P3HT:ZnO NPS	2.18(46.2%)	22.9 (36.1%)	315 (17.7%)	65
ZnO NRS/P3HT	2.44(49.1%)	26.6 (32.6%)	308(18.4%)	67
ZnO NRS/P3HT:ZnO NPS	2.16 (48.5%)	38.5 (25.8%)	524 (25.6%)	145

For organic polymer blend films, the formation of an interpenetrating donor/acceptor network is essential to enable efficient separation of photogenerated excitons at the donor/acceptor interface. The separated charges may recombine in two alternative pathways which can compete with charge collection and photocurrent generation — germinate recombination of the initially generated charges at the donor/acceptor interface and bimolecular recombination of dissociated charges during their transport to the device electrodes²³⁾. In order to find out whether bimolecular recombination plays a significant or trivial role in the charge transfer processes, the excitation laser power dependent fluorescence decays were measured. The un-normalized fluorescence decay curves are shown in the right inset of Fig. 4(d). To make the comparison and analysis more accurate, normalized decay data are presented in Fig. 4(d) and they almost overlap with one another under three different excitation laser intensities. In fluorescent systems, bimolecular exciton-excition annihilation contributes to a rapid intensity dependent decay at times shortly after the excitation. Therefore the decay trace is also zoomed in to the initial part to observe whether bimolecular recombination exists. As shown in the inset of Fig. 4(d) on the left hand side, taking Sample A as an example, the normalized fluorescence decay traces under three different excitation laser intensities almost overlap even in the short time range from 0 to 400 ps. In another word, the signal amplitude increases linearly with the excitation density, whilst the decay dynamics are invariant²³⁾. Therefore, bimolecular recombination was trivial in the samples under study.

All decay profiles from the four films were fitted by a triple exponential function indicating three lifetime components. The first two short lifetimes ranging from hundreds of femtoseconds to tens of picoseconds are attributed to exciton hopping to sites with lower energy and to a slow conformational planarization of the polymer backbones²⁴⁾. The third component is correlated to the relaxed interchain singlet excitons which are formed on a 100–200 ps time scale and decay with a several hundred picosecond time constant by nonradiative recombination²⁵⁾. All the three lifetimes τ₁, τ₂ and τ₃ become shorter in Sample B and C compared to Sample A, indicating that the presence of ZnO NPs or NRs facilitates exciton hopping and relaxation at the interfaces with P3HT.

However, the coexistence of ZnO NRs and NPs in Sample D did not enhance but hampered the exciton dissociation. A possible cause is that ZnO NPs intended to accumulate on ZnO NRs as evidenced by the results of the steady-state PL characterization. The prolonged PL decay lifetime in the ZnO NRs/P3HT:ZnO NPs hybrid compared to that in Sample B and C is mainly due to the fall of charge separation efficiency as a result of decreased interfacial areas by the addition of ZnO NPs in P3HT on ZnO NRs array. The presumable effect of area enhancement by the co-presence of ZnO NPs and NRs is surpassed by the effect that ZnO NPs block the contact of ZnO NRs and P3HT. This interpretation is consistent with the steady-state PL results and can also be demonstrated with Fig. 3.

Heterojuncion solar cells were fabricated with the sample solutions under the same active layer preparation condition. As shown in Fig. 5, the active layers of solar cells are composed of P3HT and P3HT:ZnO NPs on top of ITO glass or ZnO NRs array, respectively. From the J-V curve (Fig. 5) and the power conversion efficiency (PCE) of the solar cells (Table 2), it can be seen that the device performance is better when P3HT:ZnO NPs blend (corresponding to Sample B) and ZnO NRs:P3HT (corresponding to Sample C) serve as the active layers, compared to pure P3HT (corresponding to Sample A) and P3HT:ZnO NPs on ZnO NRs array (corresponding to Sample D). The PCE of Sample C becomes lower than that of Sample B, probably because the NRs deteriorate the surface morphology of the film. The device performances are consistent with the photophysical characterizations in which Sample B and C possess the most prominent fluorescence decay in the time-resolved PL spectroscopy. Our work provides an experimental foundation to understand the physics underlying the phenomenon that the active layer on top of ZnO NRs array which composed of high ZnO NPs concentration in P3HT deteriorated the power conversion efficiency of photovoltaic devices^26,27).

Fig. 5

Top: Schematic illustration of inverted-type bulk heterojunction solar cells fabricated with P3HT and P3HT:ZnO NPS as the active layer on ITO glass and ZnO NRS array respectively. There is a thin ZnO film between the active layer and ITO as the electron extraction layer and a thin MoO_x film between the active layer and Ag as the hole extraction layer. Bottom: Current density-voltage (J-V) characteristics of the four inverted-type solar cells.

Table 2 Solar cell performance of the films.

	JSC (mA/cm2)	Voc (V)	FF	PCE (%)
P3HT	-	-	-	-
P3HT: ZnO NPS	4.68	0.36	0.34	0.57
ZnO NRS array/P3HT	1.55	0.39	0.28	0.17
ZnO NRS array/P3HT:ZnO NPS	0.42	0.30	0.29	0.04

4. Conclusions

In summary, photophysical investigations were performed on several types of P3HT: ZnO interfaces. P3HT intends to form coiled chain around ZnO NRs, providing adequate interfacial areas to facilitate exciton dissociation and electron transfer by showing PL quenching in the time-resolved fluorescence spectrum. Almost the same effect can be achieved at P3HT: ZnO NPS interfaces at a concentration of 30 mass%. The coexistence of ZnO NPs and NRs in P3HT do not improve the effect further because ZnO NPs hampers the effective contact of ZnO NRs and P3HT. It has been shown higher charge separation efficiency of these P3HT: ZnO interfaces leads to better performances of heterojunction solar cells prepared under the same condition.

Acknowledgements

This work was financially supported by National Natural Science Foundation of China (No. 11404190).

REFERENCES

• 1)  I. Gur, N.A. Fromer, C.P. Chen, A.G. Kanaras and A.P. Alivisatos: Nano Lett. 7 (2007) 409. 10.1021/nl062660t
• 2)  J. Mawyin, I. Shupyk, M. Wang, G. Poize, P. Atienzar and I. Thilini: J. Phys. Chem. C 115 (2011) 10881. 10.1021/jp112369t
• 3)  W.J. Beek, M.M. Wienk, M. Kemerink, X. Yang and R.A. Janssen: J. Phys. Chem. B 109 (2005) 9505. 10.1021/jp050745x
• 4)  Y.Y. Lin, Y.Y. Lee, L. Chang, J.J. Wu and C.W. Chen: Appl. Phys. Lett. 94 (2009) 063308. 10.1063/1.3080203
• 5)  T. Xu, S. Venkatesan, D. Galipeau and Q. Qiao: Solar Energy Mater. Solar Cells 108 (2013) 246. 10.1016/j.solmat.2012.08.013
• 6)  P. Ruankham, S. Yoshikawa and T. Sagawa: Phys. Chem. Chem. Phys. 15 (2013) 9516. 10.1039/c3cp50266j
• 7)  P. Ruankham, S. Yoshikawa and T. Sagawa: Appl. Phys., A Mater. Sci. Process. 121 (2015) 301. 10.1007/s00339-015-9444-1
• 8)  P.Y. Ho, S. Thiyagu, S.H. Kao, C.Y. Kaoa and C.F. Lin: Nanoscale 6 (2014) 466. 10.1039/C3NR04418A
• 9)  Z.Y. Wang, B.B. Huang, X.Y. Qin, X.Y. Zhang, P. Wang, J.Y. Wei, J. Zhan, X.Y. Jing, H.X. Liu, Z.H. Xu, H.F. Cheng, X.N. Wang and Z.K. Zheng: Mater. Lett. 63 (2009) 130. 10.1016/j.matlet.2008.09.042
• 10)  Y. Dias and R. Yerushalmi-Rozen: Polymer 54 (2013) 6399. 10.1016/j.polymer.2013.09.057
• 11)  B. Ferreira, P.F. Silva, J.S.S. Melo, J. Pina and A. Maçanita: J. Phys. Chem. B 116 (2012) 2347. 10.1021/jp207418q
• 12)  W.L. Xu, B. Wu, F. Zheng, X.Y. Yang, H.D. Jin, F. Zhu and X.T. Hao: J. Phys. Chem. C 119 (2015) 21913. 10.1021/acs.jpcc.5b06931
• 13)  W.L. Xu, B. Wu, F. Zheng, H.B. Wang, Y.Z. Wang, F.G. Bian and X.T. Hao: Org. Electron. 25 (2015) 266. 10.1016/j.orgel.2015.06.042
• 14)  I.W. Hwang, D. Moses and A.J. Heeger: J. Phys. Chem. C 112 (2008) 4350. 10.1021/jp075565x
• 15)  Y. Sun, C. Cui, H. Wang and Y. Li: Adv. Energy Mater. 1 (2011) 1058. 10.1002/aenm.201100378
• 16)  D. Di Nuzzo, A. Aguirre, M. Shahid, V.S. Gevaerts, S.C. Meskers and R.A. Janssen: Adv. Mater. 22 (2010) 4321. 10.1002/adma.201001452
• 17)  J. Clark, C. Silva, R.H. Friend and F.C. Spano: Phys. Rev. Lett. 98 (2007) 206406. 10.1103/PhysRevLett.98.206406
• 18)  M. Baghgar and M.D. Barnes: ACS Nano 9 (2015) 7105. 10.1021/acsnano.5b03422
• 19)  T.P. Martin, A.J. Wise, E. Busby, J. Gao, J.D. Roehling, M.J. Ford, D.S. Larsen, A.J. Moule and J.K. Grey: J. Phys. Chem. B 117 (2013) 4478. 10.1021/jp308586k
• 20)  M. Baghgar, J.A. Labastide, F. Bokel, R.C. Hayward and M.D. Barnes: J. Phys. Chem. C 118 (2014) 2229. 10.1021/jp411668g
• 21)  A. Bakour, G. Geschier, M. Baitoul, M. Mbarek, K. El-Hadj, J.L. Duvail, S. Lefrant, E. Faulques, F. Massuyeau and J. Wery-Venturini: Mater. Chem. Phys. 143 (2014) 1102. 10.1016/j.matchemphys.2013.11.010
• 22)  H.D. Jin, F. Zheng, W.L. Xu, W.H. Yuan, M.Q. Zhu and X.T. Hao: J. Phys. D Appl. Phys. 47 (2014) 505502. 10.1088/0022-3727/47/50/505502
• 23)  S. Shoaee, M.P. Eng, E. Espíldora, J.L. Delgado, B. Campo, N. Martín, D. Vanderzanded and J.R. Durrant: Energy Environ. Sci. 3 (2010) 971. 10.1039/c003394d
• 24)  N. Banerji, S. Cowan, E. Vauthey and A.J. Heeger: J. Phys. Chem. C 115 (2011) 9726. 10.1021/jp1119348
• 25)  H. Wang, H.Y. Wang, B.R. Gao, L. Wang, Z.Y. Yang, X.B. Du, Q.D. Chen, J.F. Song and H.B. Sun: Nanoscale 3 (2011) 2280. 10.1039/c0nr01002b
• 26)  S. Cook, A. Furube and R. Katoh: Energy Environ. Sci. 1 (2008) 294–299. 10.1039/b805643a
• 27)  M.J. Wang, Y. Li, H.H. Huang, E.D. Peterson, W. Nie, W. Zhou, W. Zeng, W. Huang, G. Fang and N. Sun: Appl. Phys. Lett. 98 (2011) 103305. 10.1063/1.3554381