High efficiency hybrid PEDOT:PSS/nanostructured silicon Schottky junction solar cells by doping-free rear contact

Abstract

A high doping technique has been widely used for record-efficiency crystalline silicon (Si) solar cells to minimize the series resistance losses and to form a back surface field. However, it requires high temperatures (up to 1000 °C) and involves toxic gases, which may not be compatible for hybrid organic–silicon solar cells. Here, we report that a high power conversion efficiency (PCE) of 13.7% with a device area of 0.8 cm² has been achieved for organic-nanostructured Si hybrid solar cells by inserting a cesium carbonate (Cs₂CO₃) layer between Si and the rear electrode aluminium (Al), which is realized by a solution process under low-temperature annealing (<150 °C). Transient and constant current–voltage, capacitance–voltage, and scanning Kelvin probe microscope measurements are used to characterize the effect of the Cs₂CO₃ layer on the device performance. The insertion of Cs₂CO₃ not only decreased the contact resistance, but also generated a built-in electric field on the rear electrode. The recombination rates are suppressed at the back surface due to the deflection of minority carriers. These findings show a promising strategy to achieve high performance organic–silicon solar cells with a simple, low temperature and cost effective process.

---

Broader context

Organic–inorganic hybrid photovoltaic devices exhibit significant potential as they offer a simple and low temperature fabrication process, which can combine the advantages of both organics and inorganics for low cost and high efficiency photovoltaics (PVs). The exploration of novel interfacial materials with the required charge selectivity and compatibility is critical for hybrid PVs, and a proper combination of the interfacial layer with the active layer is the key to improve the power conversion efficiency (PCE). Here, we report that nanostructured silicon–organic hybrid PVs with a PCE of up to 13.7% under air mass 1.5G irradiation was achieved through interfacial engineering of a solution-processed cesium carbonate layer. The cesium carbonate layer was inserted between silicon and aluminum without any high temperature doping process (<150 °C), which resulted in facilitating electron extraction and suppressing charge recombination. These results offer an important concept for achieving high performance organic–silicon photovoltaic devices with a simple, low temperature, and cost-effective process. In addition, this concept can be also used for improving the metal–semiconductor contact.

Introduction

Transparent conducting polymer–silicon (Si) hybrid Schottky solar cells have attracted wide research interest due to their simple device structure compared with the Si-based traditional p–n junction photovoltaics.^1–10 One of the most important advantages of organic–Si hybrid solar cells is that they allow the stacking of different materials without the strict requirements of lattice matching in Si heterojunctions. Poly(3,4-ethylenedioxythiophene)/poly(styrenesulfonate) (PEDOT:PSS) has been extensively investigated for hybrid solar cells due to its high transparency and conductivity.^9,11–13 Various approaches have been made to enhance the performance of this kind of the device, such as Si surface passivation,^10,14–16 surface morphology controlling,^{3,9,11,13,17–19} and property tuning of PEDOT:PSS.^4,7,20–22 A power conversion efficiency (PCE) of over 13% has been achieved for the PEDOT:PSS/Si solar cells.^8,23,24 In addition, carbon nanotubes (CNTS)–Si heterojunction photovoltaics with efficiencies of up to 15% by coating a TiO₂ antireflection layer and doping CNTs with oxidative chemicals as a transparent window have been demonstrated.²⁵ However, to the best of our knowledge, there has hardly been any investigation on the rear electrode, which plays a critical role in device performance.^2,3 Generally, the cathode is either Ga:In eutectic, aluminum (Al) or Ti/Pd/Ag.^8,10,11 The physical mechanism of the rear electrode has not been a popular topic of research for the hybrid solar cells.^2,3,18

Traditionally, the Si–metal ohmic contact is synthesized by high p- or n-doping at high temperatures (up to 1000 °C)²⁶ to form the back surface field. This process requires the use of hazardous doping gases, such as diborane (B₂H₆) and phosphine (PH₃), which poses operational and environmental issues. Moreover, doping has a detrimental effect on the quality of crystalline Si, because it results in the creation of additional Si dangling bonds. Furthermore, due to the high temperatures required and the complicated process, the doping method may be not compatible for organic–Si hybrid solar cells.

Cesium carbonate (Cs₂CO₃) has been used as an effective electron-injection (and hole-blocking) layer at the cathode interface in organic devices.^27,28 This is mainly ascribed to the decreased work function (WF) of Cs₂CO₃/Al due to Cs₂CO₃ decomposing into cesium oxide (Cs₂O) when annealed.²⁹ In addition, an Al–O–Cs structure is formed while thermally evaporating Al, which further lowers the WF of Al.²⁸ A solution fabrication process can be used to avoid any physical and chemical stresses at the interface, thus leading to better junction properties.³⁰

Here, we develop an alternative back contact method by inserting a Cs₂CO₃ layer between the rear side of n-Si and Al for a Si/PEDOT:PSS solar cell. Nanostructured Si is used to increase the light harvesting capability. When the Cs₂CO₃ layer is inserted between nanostructured Si and the rear electrode Al, a high PCE of 13.7% is achieved, which is a 27% improvement in comparison to a Si/Al direct contact device (PCE: 10.8%).

Experimental section

Preparation of nanostructured Si

The nanostructured Si was prepared by a metal ion-assisted electroless chemical etching process.³¹ Planar Si substrates (n-type Si (100)) with a resistivity of 0.05–0.1 Ω cm were immersed in a solution of hydrofluoric acid (HF) (4.8 M) and silver nitride (AgNO₃) (0.02 M) at room temperature for 5 min. Then, they were rinsed with deionized water and dipped in a nitric acid (HNO₃) solution. To reduce the surface/volume ratio, the substrates were immersed in a solution of HF (4.8 M) for 10 min and anisotropic tetramethyl ammonium hydroxide (TMAH)/DI water (v/v = 1 : 25) for 30 s.

Device fabrication

Cs₂CO₃ and polyethylenimine (PEI) were dissolved in 2-methoxyethanol with a concentration of 0.5 mg mL⁻¹ and 1 mg mL⁻¹, respectively. A Cs₂CO₃/PEI (v/v = 1 : 1) mixture solution was spin coated onto the rear of nanostructured Si substrates, and then annealed with infrared radiation for 20 min. A highly conductive PEDOT:PSS (CLEVIOS PH 1000) solution mixed with 5 wt% dimethyl sulfoxide (DMSO) and 1 wt% Triton (from Aldrich) was spin-coated onto the front side of the nanostructured Si substrates. Then, the substrates were annealed at 125 °C for 30 min in a nitrogen atmosphere. A silver grid electrode was deposited on top of the PEDOT:PSS layer through a shadow mask. A 150 nm-thick Al film was deposited onto the rear Si substrates by thermal evaporation (Mini-SPECTROS, Kurt J. Lesker Co.). The active area of the device was 1 cm × 0.8 cm.

Device characterization

The photovoltaic characterization was conducted under ambient conditions. A Newport 91160 solar simulator equipped with a 300 W xenon lamp and an air mass (AM) 1.5 filter were used to generate a simulated solar spectrum irradiation source. The irradiation intensity was 100 mW cm⁻², which was calibrated with a Newport Si solar cell 91150. A Newport monochromator 74125 and power meter 1918 with a Si detector 918D were used in the external quantum efficiency (EQE) measurements. All of the electrical data were recorded by a Keithley 2612 source meter. The capacitance versus voltage (C–V) measurements were carried out with a Wayne Kerr 6500B impedance analyzer. The reflection spectrum was measured by a spectrometer (Perkin-Elmer Lambda 750) with an integrating sphere. Morphology and scanning Kelvin probe microscopy (SKPM) were characterized with an atomic force microscope (AFM, Veeco, Multimode V). The nanostructured Si was characterized with a high-resolution scanning electron microscope (SEM) (Carl Zeiss, Supra 55). Optical microscopy images were obtained from a fluorescence optical microscope (Laica, DM4000M). Regarding the transient electric output characteristic setup, the devices were connected to a digital oscilloscope with an input impedance of 1 MΩ. The intensity of white light, which is referred to henceforth as a “light bias”, was used to control the open-circuit voltage (V_oc) of the devices. A laser with a wavelength of 532 nm was used as optical perturbation, pulse duration was set to 1 μs and frequency to 100 Hz, which resulted in a voltage transient with a peak value of 10 mV (≪V_oc). The frequency, light intensity and pulse duration were kept constant, with the photocurrent transient at an impedance of 50 Ω.

Results and discussion

The organic–inorganic hybrid device was generally assumed to be a Schottky junction between n-Si, PEDOT:PSS, and the rear electrode.³² PEDOT:PSS acted as a window electrode, which allowed light to reach Si. The PEDOT:PSS film displayed good light transparency (300–1100 nm, >90%) with a high conductivity (up to 1000 S cm⁻¹).³³ In order to improve the hole collection efficiency, a silver grid electrode was deposited onto the PEDOT:PSS film. PEDOT:PSS with a WF of ∼5.1 eV (ref. 33) was used for hole collection and Al as the rear metal for electron collection. Light was harvested by the Si substrate, and a built-in electric field was formed in Si to sweep the charges towards the proper direction. For the rear contact of the most hybrid organic–silicon devices, a metal layer (e.g., Al) is directly deposited on Si,⁷ as shown in Fig. 1(a). In an ideal metal–semiconductor Schottky barrier, the barrier height is mainly determined by the properties of the metal and the metal–semiconductor interface. For Si, the barrier height was nearly independent of the doping concentration. There was a large energy barrier between Al and Si, and electrons have to pass through this barrier to be collected by Al. In addition, there was an equal amount of electrons and holes at the Si interface, and thus charge recombination became a serious issue. If the rear contact was ohmic, then the electrons could be effectively collected by Al and the total charge recombination could be minimized. A few methods were used to tune the barrier height of the metal and silicon interface.²⁶ Traditionally, a thin highly doped layer (<10 nm) was incorporated, and the effective barrier height for a given metal–semiconductor interface was tuned, as shown in Fig. 1(c). For the contact with a thin n⁺⁺Si layer, the barrier was dramatically reduced due to the tunneling effect. In addition, an electric field was formed at the rear interface, which suppressed the holes that flow to the rear surface. The hole concentration was thus maintained at higher levels in the bulk of the device, and hence the recombination velocity decreased. Cs₂CO₃ was inserted between Si and Al in order to decrease the barrier height, as shown in Fig. 1(b). In addition, charge recombination was also suppressed, which was similar to the thin n⁺⁺Si layer. A device without Cs₂CO₃ was used as the reference.

Fig. 1 Energy alignment and electron/hole distribution of (a) Si/Al; (b) Si/Cs₂CO₃/Al; (c) Si/n⁺⁺Si/Al under illumination. For Si directly contacting with Al in (a), there is a large energy barrier for the electrons to be collected by the Al electrode. The electron and hole concentrations are equal at the rear interface of Al and Si. After inserting a Cs₂CO₃ layer to make (b) Si/Cs₂CO₃/Al, the effective barrier is reduced, due to the lowering of the work function of Al. In traditional (c) Si/n⁺⁺Si/Al devices, the barrier is only slightly reduced, and generally the tunnelling effect is dominant.

Pristine Cs₂CO₃ was unable to form a continuous film on a Si substrate through a spin-coating process. As shown in the optical microscope images in Fig. 2(a), large particles were formed due to Cs₂CO₃ re-crystallization during the spin-coating process. The film displayed a discontinuous layer (shown in the AFM images in Fig. 2(c)), and therefore uncovered regions still existed where Al could still directly contact with Si. The film quality was dramatically improved when PEI was added to the Cs₂CO₃ solution, which is shown in Fig. 2(b) and (d). The surface roughness of both the films was characterized by AFM in terms of root-mean-square (RMS) roughness. PEI, which exhibited a high number of amine groups (the chemical structure of PEI is shown in Fig. S1†), displayed excellent dispersion properties in a methoxyethanol solution. The incorporation of PEI formed a uniform film with a RMS value of 0.817 nm, as extracted from the AFM image, while the film based on pristine Cs₂CO₃ displayed a rather rough surface with a RMS value of 1.88 nm. Moreover, it assured that Cs₂CO₃ formed a uniform dispersion on the Si substrate, which was confirmed by the energy-dispersive X-ray spectroscopy of cesium, as shown in Fig. 2(e) and (f). There was almost no difference between Fig. 2(e) and (f), which indicated that the dispersion of Cs₂CO₃ hardly changed after PEI addition. The energy-dispersive X-ray spectroscopy of carbon was also measured and is shown in Fig. S2.† The concentration of carbon increased with PEI addition, which was ascribed to the carbon source of PEI. When Al was deposited on the uniform Cs₂CO₃ layer, a mild annealing treatment led to the formation of a Cs–O–Al complex.²⁹

Fig. 2 Optical microscopy images (a) without PEI, (b) with PEI; AFM images of the Cs₂CO₃ film (c) without PEI, (d) with PEI on Si, the inset figures in (c) and (d) show the RMS values, respectively; and the element mapping images by energy-dispersive X-ray spectroscopy of cesium (e) without PEI, (f) with PEI.

The devices were fabricated to explore the effect of the Cs₂CO₃ interlayer on the rear contact. Compared to planar Si, nanostructured Si displayed a considerably lower light reflectance, which could dramatically improve the light harvesting capability. The reflectance spectra of planar Si and nanostructured Si coated with PEDOT:PSS are shown in Fig. S3,† where nanostructured Si exhibited an average reflection ratio of ∼15%, while the planar one displayed a reflection ratio of 25%. The morphology of the nanostructured Si (Fig. S4†) revealed that antireflection was provided by the graded index of refraction in the nanostructure.

Current versus voltage (J–V) curves of the hybrid solar cells with and without Cs₂CO₃ under simulated AM 1.5 illuminations at 100 mW cm⁻² are shown in Fig. 3(a). The photovoltaic parameters of the short circuit current density (J_sc), the V_oc, the fill factor (FF), and the PCE of the hybrid solar cells with and without the Cs₂CO₃ layer and their statistics (of seven cells for each type) are summarized in Table 1. The excellent device with the Cs₂CO₃ layer exhibited the highest FF of 68.4%, a V_oc of 0.621 V, a J_sc of 32.2 mA cm⁻² and a record PCE of 13.7%. The reference device (direct contact of Si and Al) with a V_oc of 0.586 V, a J_sc of 28.6 mA cm⁻², and a FF of 64.9% yielded a PCE of 10%. The inserted Cs₂CO₃ layer resulted in 27% PCE enhancement compared with the reference device. The EQE spectra were measured, as shown in Fig. 3(c). The device with the interlayer displayed higher EQE values in the visible and near infrared region when compared with the reference device, which was in line with the enhanced J_sc value.

Fig. 3 Current density versus voltage characteristic of the hybrid solar cells with and without Cs₂CO₃ under (a) 100 mW cm⁻² illumination (AM 1.5G), (b) dark; (c) EQE spectra of the hybrid solar cells with and without Cs₂CO₃.

Table 1 Electron output characteristics of the hybrid solar cells with and without Cs₂CO₃ layer

Rear electrode^a	V oc (V)	J sc (mA cm^−2)	FF^b (%)	PCE^b (%)	R s (Ω cm^2)
a Data and statistics based on seven cells of each condition. b Numbers in bold are the maximum record values.
Al	0.586 0.585(±0.007)	28.6 28.5(±0.7)	64.9 63.5 (±1.28)	10.8 10.5 (±0.3)	3.46 3.42(±0.3)
Cs2CO3/Al	0.621 0.624(±0.003)	32.2 31.4 (±0.9)	68.4 68.8 (±1.14)	13.7 13.4 (±0.6)	2.16 2.03(±0.7)

---

In order to quantify the improvements in device performance after inserting the Cs₂CO₃ layer between the nanostructured Si and Al, detailed measurements, including the constant and transient output characteristics and Kelvin scanning microscopy were conducted in the following sections.

Fig. 3(b) shows the J–V characteristics of the hybrid solar cells with and without Cs₂CO₃ in the dark. It was observed that the saturation current density (J_s) was suppressed significantly after the Cs₂CO₃ layer was inserted between Si and Al. The dark J–V curves were simulated according to the thermionic emission model.³⁴ The J_s was obtained by fitting.

where J is the current density value, V is the applied voltage, T is the absolute temperature (298 K), k is the Boltzmann constant (1.38 × 10⁻²³ m² kg s⁻² K⁻¹), and q is the electronic charge (1.6 × 10⁻¹⁹ C)). The device without Cs₂CO₃ displayed a J_s value of 1.8 × 10⁻⁸ A cm⁻². The device with Cs₂CO₃ exhibited a J_s of 1.4 × 10⁻⁹ A cm⁻². The J_s decreased ∼15 times and the V_oc increased ∼40 mV when the Cs₂CO₃ layer was incorporated. The suppressed J_s was ascribed to a reduced contact barrier occurring by inserting the Cs₂CO₃ layer. The device with a lower J_s led to a larger V_oc according to the photovoltaic general relation:

The contact between Al and the rear side of Si was dramatically improved using the Cs₂CO₃ layer. Al pads were deposited onto Si through thermal evaporation. According to the J–V measurements, as shown in Fig. 4(a), the contact between Si and Al was non-ohmic. A transmission line measurement (TLM) was conducted to measure the contact resistance.³⁵ The detailed TLM method is described in the ESI.† A schematic diagram for measuring the contact resistance value is shown in Fig. S5.† The contact resistance was 3.97 × 10⁻¹ Ω cm² for the Al/Si direct contact. A non-ohmic contact resulted in a high contact resistance between Si and the rear electrode Al, which led to a charge accumulation at the rear electrode interface. A highly doped layer was used for an improved contact in the traditional Si solar cell, where the contact resistance was reduced to 10⁻²–10⁻³ Ω cm².^2,35 The highly doped layer also significantly lowered the saturation current density (to as low as 10⁻¹³ A cm⁻²) and enhanced the V_oc.³⁶ Here, with the insertion of the Cs₂CO₃ layer between Si and Al, the current–voltage curve was linear, as shown in Fig. 4(a), which indicated an improved contact. The contact resistance was 6.18 × 10⁻² Ω cm² when the Cs₂CO₃ layer was inserted, which was similar in value to the contact resistance when the high temperature and highly doped process was used. The series resistance (R_s) was extracted from the plot of dV per d (lnJ) versus J (the detailed derivation process is described in the ESI†). Fig. 4(b) shows the dV per day (lnJ) versus J curves of the hybrid solar cell, with and without the Cs₂CO₃ layer. The device with and without the Cs₂CO₃ layer displayed R_s values of 2.16 Ω cm² and 3.46 Ω cm², respectively, as shown in Table 1. The value of R_s decreased after inserting the Cs₂CO₃ layer, which was consistent with the improved back contact.

Fig. 4 (a) Current–voltage measurements between the T-shaped Al pads deposited by thermal evaporation at the rear side of Si with and without the Cs₂CO₃ layer; (b) series resistance values were extrapolated from the dV per day (lnJ) versus J curves of the hybrid solar cells with and without the Cs₂CO₃ layers. (c) C⁻²–V plots of the hybrid solar cells with and without the Cs₂CO₃ layer; (d) photovoltage decay measurements on the devices with and without the Cs₂CO₃ layer.

The insertion of the Cs₂CO₃ layer not only decreased the contact resistance but also generated a larger built-in electric field on the rear electrode. This deflected the minority carriers and recombination rates were suppressed at the back surface. The C⁻²–V plots at a signal frequency of 1 KHz of the hybrid solar cells with and without Cs₂CO₃ are shown in Fig. 4(c). C was the capacitance of the device. Here, C⁻² was linear with the applied voltage, and the intercept corresponds to the built-in potential. The devices with the Cs₂CO₃ layer yielded a higher built-in potential than the reference one. A higher built-in potential effectively deflected electrons from migrating towards the cathode.

The improved rear contact and increased built-in potential were ascribed to the Al WF dropping in the presence of the Cs₂CO₃ layer. The WF change was measured by SKPM under ambient conditions.^37,38 Since our devices were characterized in ambient air, the SKPM technique should be suitable to interpret the WF change. The same conductive tip was used during the entire measurement to ensure that its WF remained constant. The relationship between the WF of the conductive tip, Φ_t, and the samples, Φ_s is given by

Φ_s = Φ_t − e V_sp

where e is the elementary charge, and V_sp is the surface potential measured by SKPM. A higher surface potential corresponds to a lower WF of the measured sample. As shown in Fig. 5, the surface potential of Cs₂CO₃/Al decreased, compared with bare Al. This result is consistent with previous measurements made by ultraviolet photoemission spectrum (UPS) in a high vacuum,^28,29 where a strong chemical reaction occurred between the Cs₂CO₃ layer and thermally evaporated Al. In the thermal evaporation process, the first layer of Al atoms formed chemical bonds with the underlying Cs₂CO₃ layer. Metallic Al could only be detected after deposing a second Al layer, as determined by X-ray photoelectron spectroscopy. Many O atoms bonded to both Al and Cs atoms in the Al–O–Cs complexes, which was critical in reducing the WF of Al. Moreover, the decreased WF of Al reduced the Schottky barrier height, as shown in Fig. 1(b). As a result, the built-in potential increased, which would further reduce charge carrier recombination.

Fig. 5 Using SKPM to probe the potential difference of Cs₂CO₃ and Al. (a) Surface potential images. (b) Cross-section line profile of the surface potential image.

The charge carrier recombination kinetics under open circuit conditions can be explored by transient photovoltaic measurements according to previous reports.^39–41 In order to be consistent with the J–V test, the background light intensity was set to 100 mW cm⁻². The transient photovoltage (TPV) decay was fitted with a mono-exponential decay equation:

δV = ΔV₀ exp(−t/τ)

where ΔV₀ is a constant that fits to the peak height (10 mV), t is the decay time, andτ is the carrier lifetime. A steady state of electron and hole concentration and a stable V_oc were achieved under illumination. The concentration of electrons and holes started to decrease as a result of recombination when the illumination was turned off. The decay of V_oc revealed charge carriers recombination in the hybrid device. Fig. 4(d) shows the transient photo-voltage as a function of time for the devices with and without Cs₂CO₃. The reference device showed a charge carrier lifetime of 53.7 μs, while the excellent device with Cs₂CO₃ displayed a longer carrier lifetime of 73.6 μs. This result indicated that the insertion of the interlayer between Si and cathode suppressed the carrier recombination and lower saturation current, and both the J_sc and the V_oc were improved.

Conclusion

In conclusion, we developed a dopant-free rear contact for high performance organic–inorganic polymer/nanostructured Si solar cells. The rear contact was fabricated by inserting a solution processed Cs₂CO₃ layer between nanostructured Si and Al. An excellent PCE of 13.7% was achieved by this simple and low temperature process (<150 °C). The superior device performance was ascribed to the reduced WF of Al with the Cs₂CO₃ layer, which resulted in the suppression of the saturation current density, reduced the contact resistance, and enhanced the built-in potential. The improvement of the rear contact allowed the minority carrier to be deflected, leading to the suppression of carrier recombination. This simple, low temperature fabrication process for the hybrid solar cells is promising for low cost photovoltaics.

Acknowledgements

This work was supported by the National Basic Research Program of China (973 Program) (2012CB932402), National Natural Science Foundation of China (91123005, 61176057, 61211130358), Jiangsu Key Laboratory for Carbon-Based Functional Materials & Devices, the Priority Academic Program Development of Jiangsu Higher Education Institutions and Collaborative Innovation Center of Suzhou Nano Science and Technology. We would appreciate Mr Tam Jasper for the very detailed paper reading and correction.

Notes and references

1. M. J. Sailor, E. J. Ginsburg, C. B. Gorman, A. Kumar, R. H. Grubbs and N. S. Lewis, Science, 1990, 249, 1146–1149.
2. Y. Zhang, F. Zu, S. T. Lee, L. Liao, N. Zhao and B. Sun, Adv. Energy Mater., 2014, 4, 1300923.
3. S. Jeong, E. C. Garnett, S. Wang, Z. Yu, S. Fan, M. L. Brongersma, M. D. McGehee and Y. Cui, Nano Lett., 2012, 12, 2971–2976.
4. I. Khatri, Z. Tang, Q. Liu, R. Ishikawa, K. Ueno and H. Shirai, Appl. Phys. Lett., 2013, 102, 063508.
5. S. Avasthi, S. Lee, Y. L. Loo and J. C. Sturm, Adv. Mater., 2011, 23, 5762–5766.
6. S. C. Shiu, J. J. Chao, S. C. Hung, C. L. Yeh and C. F. Lin, Chem. Mater., 2010, 22, 3108–3113.
7. J. P. Thomas, L. Y. Zhao, D. McGillivray and K. T. Leung, J. Mater. Chem. A, 2014, 2, 2383–2389.
8. P. Yu, C. Y. Tsai, J. K. Chang, C. C. Lai, P. H. Chen, Y. C. Lai, P. T. Tsai, M. C. Li, H. T. Pan, Y. Y. Huang, C. I. Wu, Y. L. Chueh, S. W. Chen, C. H. Du, S. F. Horng and H. F. Meng, ACS Nano, 2013, 7, 10780–10787.
9. L. He, D. Lai, H. Wang, C. Jiang and Rusli, Small, 2012, 8, 1664–1668.
10. L. He, C. Jiang, H. Wang, D. Lai and Rusli, Appl. Phys. Lett., 2012, 100, 073503.
11. W. R. Wei, M. L. Tsai, S. T. Ho, S. H. Tai, C. R. Ho, S. H. Tsai, C. W. Liu, R. J. Chung and J. H. He, Nano Lett., 2013, 13, 3658–3663.
12. T. G. Chen, B. Y. Huang, H. W. Liu, Y. Y. Huang, H. T. Pan, H. F. Meng and P. Yu, ACS Appl. Mater. Interfaces, 2012, 4, 6857–6864.
13. T. G. Chen, B. Y. Huang, E. C. Chen, P. Yu and H. F. Meng, Appl. Phys. Lett., 2012, 101, 033301.
14. F. Zhang, B. Sun, T. Song, X. Zhu and S. Lee, Chem. Mater., 2011, 23, 2084–2090.
15. D. Liu, Y. Zhang, X. Fang, F. Zhang, T. Song and B. Sun, IEEE Electron Device Lett., 2013, 34, 345–347.
16. M. Y. Bashouti, M. Pietsch, G. Brönstrup, V. Sivakov, J. Ristein and S. Christiansen, Prog. Photovoltaics, 2013, 22, 1050–1061.
17. H. Jeong, H. Song, Y. Pak, I. K. Kwon, K. Jo, H. Lee and G. Y. Jung, Adv. Mater., 2014, 26, 3445–3450.
18. J. Schmidt, V. Titova and D. Zielke, Appl. Phys. Lett., 2013, 103, 183901.
19. K. Sato, M. Dutta and N. Fukate, Nanoscale, 2014, 6, 6092–6101.
20. Q. Liu, M. Ono, Z. Tang, R. Ishikawa, K. Ueno and H. Shirai, Appl. Phys. Lett., 2012, 100, 183901–183904.
21. M. Pietsch, M. Y. Bashouti and S. Christiansen, J. Phys. Chem. C, 2013, 117, 9049–9055.
22. D. Chi, B. Qi, J. Wang, S. Qu and Z. Wang, Appl. Phys. Lett., 2014, 104, 193903.
23. J. P. Thomas and K. T. Leung, Adv. Funct. Mater., 2014, 24, 4978–4985.
24. R. Liu, S. T. Lee and B. Sun, Adv. Mater., 2014, 26, 6007–6012.
25. E. Shi, L. Zhang, Z. Li, P. Li, Y. Shang, Y. Jia, J. Wei, K. Wang, H. Zhu and D. Wu, Sci. Rep., 2012, 2, 884.
26. S. M. Sze and K. K. Ng, Physics of Semiconductor Devices, Wiley-interscience, 2006.
27. J. Huang, G. Li, E. Wu, Q. Xu and Y. Yang, Adv. Mater., 2006, 18, 114–117.
28. J. Huang, Z. Xu and Y. Yang, Adv. Funct. Mater., 2007, 17, 1966–1973.
29. H. H. Liao, L. M. Chen, Z. Xu, G. Li and Y. Yang, Appl. Phys. Lett., 2008, 92, 173303.
30. M. Nakano, A. Tsukazaki, R. Gunji, K. Ueno, A. Ohtomo, T. Fukumura and M. Kawasaki, Appl. Phys. Lett., 2007, 91, 142113.
31. K. Q. Peng, Y. J. Yan, S. P. Gao and J. Zhu, Adv. Mater., 2002, 14, 1164–1167.
32. M. J. Price, J. M. Foley, R. A. May and S. Maldonado, Appl. Phys. Lett., 2010, 97, 083503.
33. Y. Xia, K. Sun and J. Ouyang, Adv. Mater., 2012, 24, 2436–2440.
34. X. Li, H. Zhu, K. Wang, A. Cao, J. Wei, C. Li, Y. Jia, Z. Li and D. Wu, Adv. Mater., 2010, 22, 2743–2748.
35. G. Reeves and H. Harrison, IEEE Electron Device Lett., 1982, 3, 111–113.
36. J. Nelson, The physics of Solar Cell, Imperial College Press, London, 2003.
37. L. Sun, J. Wang and E. Bonaccurso, J. Phys. Chem. C, 2010, 114, 7161–7168.
38. V. Palermo, M. Palma and P. Samorì, Adv. Mater., 2006, 18, 145–164.
39. A. H. Ip, S. M. Thon, S. Hoogland, O. Voznyy, D. Zhitomirsky, R. Debnath, L. Levina, L. R. Rollny, G. H. Carey and A. Fischer, Nat. Nanotechnol., 2012, 7, 577–582.
40. Z. Li, F. Gao, N. C. Greenham and C. R. McNeill, Adv. Funct. Mater., 2011, 21, 1419–1431.
41. C. Shuttle, B. ORegan, A. Ballantyne, J. Nelson, D. Bradley, J. De Mello and J. Durrant, Appl. Phys. Lett., 2008, 92, 093311–093313.

---

Footnote

† Electronic supplementary information (ESI) available: Method for extraction of contact resistance, method for extracting series resistance, PEI chemical structure, element mapping images by energy-dispersive X-ray spectroscopy of carbon with and without PEI, reflectance of planar Si and nanostructured Si coated with PEDOT:PSS, SEM image of the nanostructured Si, schematic diagram for measuring the contact resistance value. See DOI: 10.1039/c4ee02282c

---