Interface modification for efficiency enhancement in silicon nanohole hybrid solar cells

Abstract

In this paper, the interface between Si nanoholes (SiNHs) and poly(3,4-ethylenedioxythiophene):poly(styrenesulfonate) (PEDOT:PSS) is investigated and improved to achieve high-efficiency SiNH/PEDOT:PSS hybrid solar cells. The high-density SiNHs are fabricated using short-time Ag deposition before metal-assisted chemical etching (MacEtch) method. Also, a polymer coverage method is explored to overcome the difficulty of PEDOT:PSS infiltration into SiNHs. PEDOT:PSS is mixed with co-solvent dimethylsulfoxide (DMSO) to have better polymer infiltrate into SiNHs via two-step coating process. This technique significantly improves the interface between SiNHs and PEDOT:PSS; the greatly reduced contact angle from 90° to 16° at the interface of Si and PEDOT:PSS has established this fact. In addition, the minority carrier lifetime is dramatically increased from 31.52 to 317.20 μs. The property improvement enables the SiNH/PEDOT:PSS hybrid solar cell with high J_sc of 36.80 mA cm⁻², V_oc of 0.524 V, FF of 66.50%, and thus PCE of 12.82%. Also, the SiNH structures have an excellent light-trapping effect, which contributes to very low average total reflectance of 3%, due to internal multiple reflections caused by subwavelength features. At an angle of incidence up to 60°, the specular reflectance maintains at as low as 1%; even at a large angle of 70°, the reflectance is still below 10%. This work provides a feasible solution process to fabricate SiNH structure and to improve organic/Si hybrid solar cells in energy and cost-effective manner.

---

Introduction

Among renewable energy sources, crystalline silicon (c-Si) solar cells still dominate the PV market, on account of their nontoxicity, high efficiency, and supply of abundant material for the fabrication.^1,2 However, one drawback is that their manufacture involves expensive, high-temperature dopant diffusion or ion implantation, and anti-reflection (AR) coating in vacuum. To increase the cost-effectiveness of Si-based solar cells, numerous attempts have been made to produce low-temperature processed Si solar cells. Hybrid photovoltaic devices incorporating organic and inorganic materials are thus receiving great attention^3–5 because of their easy fabrication and low cost compared to conventional p–n junction Si solar cells. Such Si-conjugated polymer hybrid solar cells combine the advantages of high carrier mobility of Si and low-cost processing of polymer materials. The polymer poly(3,4-ethylenedioxythiophene):polystyrenesulfonate (PEDOT:PSS) is a widely used organic material for hybrid devices, because it is transparent, highly conductive, and water-soluble in its metallic state.^6,7 On the other hand, to improve device optical absorption and to replace conventional AR coating, nanostructures, like nanowires, nanorods, or nanocones, are fabricated on a Si surface to form Si-nanostructure/organic hybrid solar cells.^8–13 Because light bounces back and forth inside the nanostructures, light absorption is increased and the light reflection is reduced therein.^14–16 Therefore, much effort has been made to fabricate hybrid solar cells using organic/Si-nanostructures.^17–21

Nevertheless, nanostructures often increase the surface defects and lead to the difficulty of infiltrating of hole transporting material PEDOT:PSS into the gap between nanostructures.^22,23 The performance of hybrid solar cells hence depends on interface and surface preparation. If surface defects increase and PEDOT:PSS cannot thoroughly attach onto Si nanostructures to suppress surface defects, the cell performance would be weakened. The interface properties are primarily influenced by the wettability and conductivity of PEDOT:PSS, spin coating parameters (spin casting speed and time), and annealing temperature. However, the commercially available PEDOT:PSS solution does not infiltrate easily into Si nanostructures due to hydrophobic effect and nanostructure spacing. To conquer this problem, other organic solvent or additive was added to ensure complete coverage on high density nanostructures.^24–26 Dimethylsulfoxide (DMSO) and ethylene glycol (EG) are frequently used co-solvents for improving conductivity and wettability, by adding relative amount into the PEDOT:PSS.²⁷ Nonetheless, investigation is still in progress of interfacial properties and conformal coating in the nanostructure gaps in Si/PEDOT:PSS hybrid solar cells and corresponding optimization.

In this study, great improvement of polymer coverage is demonstrated by mixing a relative amount of DMSO co-solvent with PEDOT:PSS to fabricate SiNH/PEDOT:PSS hybrid solar cells. By adopting SiNH structure with appropriate polymer coating process and incorporation of co-solvent in the PEDOT:PSS, a remarkable efficiency of 12.82% is achieved without using high temperature process. We evaluate the related effects on the resulting hybrid solar cell properties. The contact angle between the Si and PEDOT:PSS interface is greatly reduced, from 90° to 16°, by adding relative amount of DMSO. In addition, PEDOT:PSS with DMSO greatly improves the minority carrier lifetime from 31.52 to 317.20 μs. Such a SiNH/PEDOT:PSS hybrid solar cell exhibits a high J_sc of 36.80 mA/cm², V_oc of 0.524 V, FF of 66.50%, and thus power conversion efficiency (PCE) of 12.82%. The high-density SiNH arrays are formed by two-step metal-assisted chemical etching (MacEtch) method. The SiNH structures have an excellent light-trapping effect, which contributes to very low average total reflectance of 3% over a wide range of wavelength from 300 to 1100 nm due to internal multiple reflections caused by subwavelength features. At an angle of incidence up to 60°, the reflectance is less than 1% and at the large angle of 70° the value is still less than 10%.

Experimental section

Fabrication of silicon nanohole (SiNH) arrays

As illustrated in Fig. 1(a), organic–inorganic hybrid heterojunction solar cells are fabricated in several steps. Firstly, n-type (100) double-sided polished CZ wafer was used (thickness 200 ± 25 μm, resistivity 1–10 Ω) to fabricate Si nanoholes by two-step metal assisted chemical etching (MacEtch) method. We cleaned the Si wafer by de-ionized water (DI water), acetone (ACE), piranha clean (H₂SO₄ : H₂O₂), and isopropanol (IPA) for 5–10 minutes. Next, the cleaned Si wafer was dipped into an aqueous solution containing HF (3.71 mol L⁻¹)/AgNO₃ (0.004 mol L⁻¹) for 5 seconds to form independently distributed silver nanoparticles. Then, the silver nanoparticle-coated wafer was immersed into an aqueous solution with HF (6.57 mol L⁻¹) and H₂O₂ (0.18 mol L⁻¹) to form vertically-aligned (100) Si nanoholes by etching the Si wafer. A high-density SiNH array was fabricated with the etching time of 25 s, which produced nanoholes with depths of 470 nm. During etching process, the rear side of the silicon substrate was protected by photoresist. Finally, the photoresist was removed by acetone. HNO₃ and HF were subsequently used to remove silver and the original silicon oxide, respectively.^17,28,29 The fabricated Si nanohole wafer was kept in ambient air for 3 to 4 hours to form a thin layer of native oxide. Afterwards, 50 nm of titanium (Ti) and 250 nm silver (Ag) were evaporated on the back side of Si substrate as the cathode electrode.

Fig. 1 Schematic illustration of the fabrication process for SiNH/PEDOT:PSS hybrid heterojunction solar cell. (a) Fabrication of SiNH arrays. (b) Coating PEDOT:PSS with 5 wt% DMSO on the SiNH surface. (c) Coating PEDOT:PSS PHCV4 on ITO substrate. (d) Finally the silicon wafer samples were stuck on ITO substrate.

Device fabrication and characterization

To fabricate Si nanohole/PEDOT:PSS hybrid solar cells, PEDOT:PSS solution was first spin-coated on top of silicon nanoholes by two-step coating process (600 and 4000 rpm for 10 s) and also on ITO to form good contact with Si nanoholes. Clevios™ PEDOT:PSS PH1000 (viscosity 50 mPa s) with 5 wt% of DMSO was spun on the silicon nanohole surface (Fig. 1(b)) and PHCV4 was spun (4000 rpm for 10 s) on ITO substrate (Fig. 1(c)) to have better contact because Clevios™ PEDOT:PSS PHCV4 has high viscosity (350 mPa s) so that silicon wafer can stick well onto the substrate (Fig. 1(d)). Finally, the fabricated devices were annealed under 140 °C for 10 minutes in atmosphere. The schematic structure of the device is depicted in Fig. 1

For JV characteristic measurement, the device was illuminated from the ITO side under 1 sun AM 1.5 G 100 mW cm⁻² by solar simulator SUN 2000, Abet Technologies, Inc, and measured by Keithley 2400 source meter. Scanning Electron Microscopic (SEM) pictures of Si nanoholes were observed using LEO 1530 FE-SEM. The optical reflectance of Si nanohole was measured using JASCO V-670 UV-VIS spectrophotometer with an integrating sphere. The effective minority carrier lifetime is measured using Quasi-Steady-State Photoconductance (QSSPC) lifetime measurement method (Sinton WCT-120 instrument).

Results and discussion

It has been found that SiNHs exhibit higher efficiency and superior mechanical stability than SiNWs, appearing to be an alternative structure with great potential for photovoltaics.^30,31 In addition, SiNH arrays have lower reflectance and stronger absorptance than silicon nanowire (SiNW) arrays in the Si absorption spectrum.^16,30 Furthermore, SiNW arrays formed by using a metal-assisted chemical etching (MacEtch) method tend to aggregate when the length of nanowire increases, so some of the gaps shrink, and some places have increased gaps between the nanowires.¹⁸ Therefore, it is beneficial to use SiNH arrays. Therefore, we explore the fabrication of SiNHs and their influences on the Si hybrid solar cells. Two-step etching method is commonly used to fabricate nanowire structures.³² Here in our investigation, the condition has been adjusted so that the Ag nano-particles deposited on the Si surface are tiny and separated from one another. Hence, the nanohole structure can be formed. From our investigation, it is very important that the silicon wafer samples are dipped into the solution for just 5 s, so independently distributed silver nanoparticles can be obtained. Then the later metal-assisted electro-less etching using HF/H₂O₂ solution will result in Si nanoholes. The density of the Ag nanoparticles is about 10¹¹/cm⁻² and the surface coverage is 57%, calculated by the statistical image analysis. If the etching time is increased, Ag nanoparticles become larger and coalesce with the adjacent nano-particles, leading to a high surface coverage. As a result, the following metal-assisted electro-less etching will give Si nanowires instead of nanoholes.

Fig. 2(a) shows the top-view SEM image of Ag nanoparticles with 5 s electro-less metal deposition. The SEM image of Fig. 2(b) shows the top view of high-density silicon nanohole arrays. Numerous small nanoholes are uniformly distributed on Si wafer with an average hole diameter of 25 to 30 nm and a density as high as 1.05 × 10¹¹ per cm². Fig. 2(c) shows the cross-sectional SEM pictures of SiNH arrays without PEDOT:PSS and Fig. 2(d) shows with PEDOT:PSS. Fig. 2(c) shows SEM images of high-density SiNH arrays with the etching time of 25 s, which produces nanoholes with depths of 470 nm. The depth of the nanohole can be controlled and varies linearly with the duration of catalyst etching process. SiNHs are also randomly distributed with a high aspect ratio.

Fig. 2 (a) Top-view SEM image of Ag nanoparticle with 5 s metal deposition. SEM images of SiNH arrays with the etching of 25 s (b) top-view, (c) cross-sectional and (d) SiNH covered with PEDOT:PSS mixed with 5 wt% DMSO.

Usually, PEDOT:PSS layer is coated with single-step process in Si hybrid solar cells, so the PEDOT:PSS layer is mainly coated just on top of the silicon nanostructures and not much permeation into the gap of nanostructures.^30,33 In particular, the SiNHs are surrounded by Si materials, so it is extremely difficult to make PEDOT:PSS infiltrate into the nanoholes. Here we develop a procedure to overcome this difficulty. First, PEDOT:PSS (PH1000, Clevios) is mixed with 5 wt% DMSO and then directly spin-coated onto SiNH surface. Mixing with co-solvent DMSO improves conductivity, reduces the solution surface tension and increases its wettability. Furthermore, we use two-step coating process at 600 and 4000 rpm. The low spin rate allows PEDOT:PSS to infiltrate through the silicon nanoholes, while the fast spin rate helps making thin PEDOT:PSS film. The cross-sectional SEM image in Fig. 2(d) shows high-density silicon array covered with PEDOT:PSS with 5 wt% DMSO. The sample for the SEM image is particularly detached from ITO substrate to check the morphology. The PEDOT:PSS film is observed to entirely cover the surface and also infiltrate into the nano-holes.

The surface of SiNHs provides more junction areas when they are completely covered with PEDOT:PSS. When PEDOT:PSS just covers the top portion of SiNHs, the depletion layer only exists near the top of SiNHs that PEDOT:PSS is in touch with. Then holes from the bulk silicon are not easy to diffuse across the interface, resulting in low collection efficiency of holes. While PEDOT:PSS entirely covers the SiNH along the nanohole surface, the carrier separation and collection from the SiNHs and bulk silicon are more efficient. Usually high density SiNH surface has a high hydrophobic effect with PEDOT:PSS, so PEDOT:PSS only partially covers the surface of SiNH, leading to a higher contact angle, less contact area and lower efficiency. Poor contact of SiNH with PEDOT:PSS also results in the decrease in shunt resistance (R_sh) due to high recombination of charge carriers^34,35 and also influence insufficient charge transport pathways.³⁴ Choosing suitable PEDOT:PSS with low viscosity and by adding relative amount of additives can reduce the contact angle between the SiNH/PEDOT:PSS interface and also improve the contact area.

Fig. 3(a)–(c) shows the different contact angles for various conditions of PEDOT:PSS polymer deposited on the Si nanohole samples corresponding to 25 s etching time. Fig. 3(a) shows the hydrophobicity effect of SiNH surface when using PHCV4 PEDOT:PSS solution, leading to higher contact angle 90°. The contact angle reduces from 90° to 65°, when using PH1000 PEDOT:PSS solution due to less viscosity property, as shown in Fig. 3(b). The contact angle is further reduces from 65° to 16° with the addition of DMSO into the PEDOT:PSS (PH1000) solution. PH1000 with DMSO shows very low contact angle, so silicon nanohole surface has better coverage, compared to other polymer conditions. To achieve uniform coating of PEDOT:PSS on the SiNH surface, it is important to reduce the contact angle between the SiNH and PEDOT:PSS film. Adding co-solvent DMSO offers the above function to produce a continuous film and improve the contact area.

Fig. 3 Photographs of (a) as purchased PHCV4, (b) as purchased PH1000 and (c) DMSO mixed PH1000 PEDOT:PSS solution on silicon nanohole arrays, including different wettability. Minority carrier lifetime of three kinds of polymer on (d) plain silicon surface and (e) etched silicon nanohole surface.

The condition of PEDOT:PSS polymer is also found to greatly influence the minority carrier lifetime, which is measured using the quasi-steady-state photo conductance (QSSPC) technique. The measured minority carrier lifetime of total recombination can be expressed as

where τ_meas is the measured lifetime, τ_bulk is the bulk recombination lifetime, τ_surf is the surface recombination lifetime, W is the wafer thickness and S is the surface recombination velocity. Here we assume that both the front and rear side of Si wafer surfaces have the same recombination velocities and neglect the time for carrier diffusion to the surface from the middle of the wafer. We first prepare four different samples in ambient conditions, the samples include (a) a planar silicon wafer as reference and three planar silicon wafers with, (b) spin-coated PHCV4 PEDOT:PSS, (c) spin-coated PH1000 PEDOT:PSS, and (d) spin-coated PEDOT:PSS PH1000 + DMSO. The minority carrier lifetime values of the four samples are shown in Fig. 3(d), which correspond to a lifetime of 31.52, 271.68, 286.92 and 317.20 μs, respectively. Fig. 3(d) reveals that the τ_meas greatly increases after PEDOT:PSS (PHCV4) being used. PEDOT:PSS and silicon form a heterojunction on the front surface, and also PEDOT:PSS can passivate the surface. Minority carrier lifetime significantly increases from 31.52 to 271.68 μs because the built-in-field prevents the carriers from recombination at the interface. The τ_meas increases to 286.92 μs after PEDOT:PSS (PH1000) being applied. The sample with PEDOT:PSS (PH1000) mixed with DMSO shows the longest lifetime of 317.20 μs. It also corresponds to very low contact angle and more surface coverage area. In addition, we prepare five more different samples in ambient conditions. The samples include (a) a planar silicon wafer, (b) etched silicon nanohole surface, (c) silicon nanohole sample with spin-coated PHCV4 PEDOT:PSS, (d) similar to (c), but spin-coated with PH1000 PEDOT:PSS, and (e) similar to (c), but spin-coated with PEDOT:PSS PH1000 + DMSO. The minority carrier lifetime values of those samples are shown in Fig. 3(e). They correspond to a lifetime of 31.52, 30.61, 39.08, 41.70 and 50.19 μs, respectively. The τ_meas value of high density silicon nanoholes is determined to be 30.61 μs, which is lower than that of plain silicon wafer mainly because of high surface defects. The τ_meas greatly increases after PEDOT:PSS being used as a result of the passivation for the silicon surface. Compared with other polymer conditions, the sample with PEDOT:PSS (PH1000) mixed with DMSO shows the longest lifetime of 50.19 μs owing to very low contact angle and more surface coverage area. The increase in τ_meas can be ascribed to the effective reduction of recombination rate at the surface, which greatly influences the PV characteristics.

Energy band diagram of a hybrid heterojunction solar cell based on silicon and PEDOT:PSS is shown in the Fig. 4(a). Here the organic material PEDOT:PSS has the highest occupied molecular orbital (HOMO) energy level at 5.1 eV and the lowest unoccupied molecular orbital (LUMO) energy level at 3.5 eV. The PEDOT:PSS forms a Schottky junction^36,37 with silicon. It can also function as the barrier for electron transport at the interface. The photo-generated holes in silicon move toward PEDOT:PSS and electrons move toward the metal contact at the back of silicon. Si and PEDOT:PSS interface has two major problems. First, Si and pristine PEDOT:PSS have micropore defects at the interface, as shown in Fig. 4(b), which acts as barriers for charge transport during the cell operation, marked by arrows. Second, the minority carriers of holes in the n-type silicon at the Si and PEDOT:PSS interface undergo recombination with electrons via the trap levels shown in Fig. 4(c). Pristine PEDOT:PSS PHCV4 and PH 1000 show very high contact angle at the Si interface, as shown in Fig. 3. Therefore, the resulting micropore defects in the interface of Si and PEDOT:PSS does not passivate the entire Si surface well.

Fig. 4 (a) Energy band diagram of a hybrid heterojunction solar cell based on silicon and PEDOT:PSS. (b) Schematic diagram of the pristine PEDOT:PSS layer formation on the silicon. Energy band diagram of a hybrid heterojunction solar cell (c) with and (d) without surface traps.

Here we add DMSO into PEDOT:PSS to improve the minority carrier lifetime. By the presence of co-solvent DMSO in PEDOT:PSS (Clevios™ PH 1000), PSS is rearranged and the following surface chain networks of PSS are reduced. Therefore, the conductivity of PEDOT:PSS can be significantly enhanced. Moreover, DMSO can reduce the contact angle between the Si and PEDOT:PSS and so increase the contact area accordingly. As a result, Si has better polymer coverage and the non-contact micropore areas between the Si and PEDOT:PSS are reduced, hence enhancing the charge transport, as shown in Fig. 4(d). Consequently, the minority carrier lifetime is improved to 317.20 μs.

The optical reflectance spectra of different depth of silicon nanoholes have been measured in the wavelength range of 300 to 1100 nm, as shown in Fig. 5(a). Etching time of 25, 35, and 50 s leads to nanohole depths of 470, 570, and 720 nm, respectively. The plain silicon wafer shows average reflectance of 38% over a whole measured spectral region. In contrast, the samples with SiNHs shows very low reflectance, less than 5% in the spectral window between 300 to 700 nm; also in the long wavelength range still the reflectance is less than 10%. Even in just 25 s of etching time, the average reflectance is 3.84% in the wavelength range of 300 to 1100 nm. The reason of suppress in the reflectance is that the silicon nanoholes are closely packed to boost light scattering and prolong the optical path. When the etching time increases, reflectance is further suppressed. For an etching time of 50 s, the average total reflectance of about 3.30% is observed over a wide spectrum from 300 to 1100 nm. As a result, the wafers are black, and have been named as “black silicon”.^15,38 They thus make good antireflective surfaces and absorption resources for photovoltaic cells. To fabricate a hybrid solar cell device, we choose 25 s etching time of nanoholes with 470 nm depth. Longer etching time of Si nanoholes usually gives more surface defects and poor infiltration of PEDOT:PSS.²² The etching time of 25 s can still provide very low reflectance, which is nearly equal to average reflectance of 50 s etching time.

Fig. 5 (a) Optical measurement spectra of SiNH arrays with different etching time. (b) Optical measurement of wide range of AOI and wavelength for the etching time of 25 s.

An excellent antireflective surface should also exhibit low reflectance over a wide angle of incidence (AOI), which is necessary for solar cells. To investigate the omnidirectional light-trapping ability of SiNH structure, a reflectance spectrum of SiNH at different angle of incidence (AOI) is measured and shown in Fig. 5(b). A wide range of AOI is observed for a wide range of wavelengths for SiNHs on the silicon wafer, fabricated with the etching time of 25 s. At all AOI, even 60°, it shows pretty low reflectance at near infrared region. At an AOI up to 60°, the specular reflectance maintains at as low as 1%; even at 70°, it remains below 10%. These results reveal that one of the exclusive characteristics of the SiNH arrays is efficient light trapping irrespective of AOI.

Regarding the hybrid solar cell, SiNH arrays are subjected to cooperate with an organic conducting material PEDOT:PSS, to form heterojunction hybrid solar cells. Because the lowest unoccupied molecular orbital (LUMO) of PEDOT:PSS (−3.5 eV) is higher than the conduction band of Si, so the electron transportation direction can be regulated toward cathode. Moreover, the highest occupied molecular orbital (HOMO) of PEDOT:PSS (−5.2 eV) is near the valence band of Si, so the holes can transport from Si to PEDOT:PSS.^17,39 To investigate the influence of co-solvent on the interface and device performance, three different conditions of polymer have been used to fabricate hybrid solar cells. First, PEDOT:PSS (PH1000, Clevios) is mixed with 5 wt% DMSO and then directly spin-coated onto SiNH surface in two steps at 600 and 4000 rpm. The low spin rate allows PEDOT:PSS to infiltrate through the silicon nanoholes, while the fast spin rate helps to make thin PEDOT:PSS film. After the film is dry, another polymer solution, PEDOT:PSS (PHCV4) is spin-coated on ITO substrate to stick the SiNH surface. For comparison, PEDOT:PSS (PH1000) without DMSO is also spin-coated onto SiNH surface and then sticks on the ITO/glass substrate with the assistance of PHCV4 solution. To make an addition comparison, the PEDOT:PSS (PHCV4) solution is spin-coated onto the ITO substrate and then SiNH surface with thin native oxide stick onto the ITO/glass substrate.

Si is the foremost light absorber and the medium for excited electrons transport, while PEDOT:PSS is a transparent layer and acts as a hole transport and electron blocking layer. The hybrid solar cells are fabricated using high-density silicon nanoholes with the etching time of 25 s Fig. 6(a) and (b) show the current density vs. voltage (J–V) characteristics of best performing devices in light and dark conditions for three different polymer solution conditions. Table 1 shows the maximum record values and summarizes the corresponding photovoltaic properties, including short-circuit current density (J_sc), open-circuit potential (V_oc), fill factor (FF), and PCE. Fig. 7 shows the data and the overall device statistics based on three cells of various conditions, including average and standard deviation for J_sc, V_oc, FF and PCE. Interestingly, 5 wt% of DMSO mixed PEDOT:PSS solution under two-step coating process achieves the highest PCE of 12.82%. As shown in Fig. 6(a), the best hybrid heterojunction cell demonstrates a PCE of 12.82% with a high J_sc of 36.80 mA cm⁻², higher than the reference counterpart of 10.84% and 9.50%. We note that the enhancement also results from an improved V_oc (from 0.500 V to 0.524 V) and FF (56.49% to 66.50%). The J_sc value increases significantly to 33.63 mA cm⁻², 35.84 mA cm⁻², and 36.80 mA cm⁻² for the polymers such as PHCV4, PHCV4 + PH1000, and PHCV4 + PH1000 + DMSO, respectively. The J_sc value increases upon coating another highly conductive polymer PH1000 through a two-step coating process. The V_oc and FF values increase from 0.510 V to 0.524 V and 59.34% to 66.50%, respectively, due to the addition of DMSO with highly conductive polymer PH1000. The efficiency increases from 10.84% to 12.82% because of adding 5 wt% DMSO into the PEDOT:PSS (PH1000). The reference solar cells fabricated without the addition of DMSO shows low PCE value of 9.50% for PHCV4. Compared to the reference cell, with the addition of 5 wt% DMSO to PH1000, the efficiency increases from 9.50% to 12.82%, which gives about 35% of increment in the efficiency. This device shows J_sc, V_oc, and FF values of 33.63 mA cm⁻², 0.500 V, and 56.49%, respectively. The devices fabricated with PHCV4 and PHCV4 + PH1000 without addition of DMSO have large reverse current density, implying an increased leakage current. One can see that R_sh value decreases significantly due to poor contact of PEDOT:PSS with SiNH in the absence of DMSO. The samples fabricated without adding DMSO may give rise to the poor junction contact, which reflects the high series resistance and the low FF of 59.34%. The SiNH samples with PEDOT:PSS and addition of DMSO expose better junction contact and exhibit further reduce the series resistance and improve FF up to 66.50%.

Fig. 6 (a) Photo J–V characteristics of the SiNH/PEDOT:PSS for different polymers. (b) Dark J–V characteristics of hybrid solar cells for three different polymer conditions.

Table 1 Photovoltaic characteristics of SINH/PEDOT:PSS hybrid solar cell devices

SiNH with	PCE (%)	Jsc (mA cm^−2)	Voc (V)	FF (%)	Rs (Ω cm^2) @V = Voc	Rsh (Ω cm^2) @V = 0
PHCV4	9.50	33.63	0.500	56.49	3.55	207.94
PHCV4 + PH1000	10.84	35.84	0.510	59.34	3.43	354.25
PHCV4 + PH1000 + DMSO	12.82	36.80	0.524	66.50	1.92	956.32

---

Fig. 7 The dependence of PV parameters of SiNH/PEDOT:PSS heterojunction solar cells, plotted as a function of different polymers.

PH1000 with the addition of DMSO greatly improves the minority carrier lifetime and also reduces the contact angle between silicon and PEDOT:PSS. PH1000 with DMSO gives better polymer coverage into the gap of SiNHs, so the corresponding device exhibits higher FF of 66.50%, higher R_sh of 956.32 Ohm cm², V_oc of 0.524 V, and efficiency of 12.82%. The improvement in the performance of the device with DMSO addition is justified by the saturation current in the dark J–V curve. It is almost an order of magnitude lower than the reference devices. The enhancement in efficiency of the device is mainly due to the improved contact of PEDOT:PSS/SiNH at the interface and the increased effective minority carrier lifetime. First, with the addition of DMSO into the PEDOT:PSS, results show that the contact angle is greatly reduced from 90° to 16° at the interface of Si nanoholes and PEDOT:PSS. In addition, the minority carrier lifetime is increased from 30.69 to 50.19 μs for SiNH surface. The property improvement gives SiNH hybrid solar cells high PCE. Second, the two-step coating process helps infiltration and forms more junction area, as shown from the SEM Fig. 2(d). On the other hand, after coating PEDOT:PSS on the surface of silicon nanohole, the samples were stuck on the ITO substrate with the help of another polymer PEDOT:PSS PHCV4, which is described in the schematic diagram Fig. 1. Here the whole layer of PEDOT:PSS contacts ITO after the sample is placed onto ITO substrate and the capillary effect helps PEDOT:PSS fill into the micropores or gap occurs. Furthermore, the coated PEDOT:PSS completely contacts ITO substrate, so hole transport can be efficient to improve the device performance. Moreover, DMSO added into the PEDOT:PSS is comprehensively considered as a dopant to improve the conductivity of the thin films,^40,41 which altogether increases the efficiency of hybrid solar cell. Fig. 8 shows the external quantum efficiency (EQE) of hybrid solar cells with different polymer of PEDOT:PSS. Clearly, improvement in the device performance can be attributed to the introduction of DMSO into PEDOT:PSS. The improvement in EQE is markedly observed in the visible and near-infrared region for the PH1000 with DMSO devices, with reference to the EQE of PH1000 with the absence of DMSO and PHCV4 in this region. The EQE of the device with PH1000 + DMSO is 83.4% at 620 nm. Large improvement is observed from 350 nm to 900 nm. In contrast, EQE without DMSO is lower for all regions of the wavelengths. The improvement in the EQE spectra is due to the improved contact between PEDOT:PSS and SiNH at the interface. The substantial improvement of high J_sc is obtained due to more junction area, efficient carrier transport and strong light trapping effect.

Fig. 8 EQE spectra of hybrid solar cells for three for different polymers.

Conclusions

In conclusion, SiNHs are explored to improve the Si hybrid solar cells. Also, an excellent polymer coverage coating method for the fabrication of SiNH/PEDOT:PSS hybrid solar cell is demonstrated. The high-density SiNHs are fabricated using a simple solution process method. PEDOT:PSS is mixed with co-solvent DMSO for better polymer coverage. The contact angle between the Si and PEDOT:PSS interface is greatly reduced from 90° to 16°. In addition, PEDOT:PSS combined with DMSO greatly improves the minority carrier lifetime from 31.52 to 317.20 μs. The hybrid solar cell exhibits high J_sc of 36.80 mA cm⁻², V_oc of 0.524 V, FF of 66.50%, and thus PCE of 12.82%. The SiNHs with sub-wavelength structures provide excellent light-trapping effect, which contributes to a very low average total reflectance of 3% due to internal multiple reflections caused by subwavelength features. At an AOI up to 60°, the reflectance is as low as 1%; even at a large angle of 70°, the reflectance value is still below 10%. This unique SiNH nanostructure is highly attractive for the development of cost-effective photovoltaic devices.

Acknowledgements

We thank the Ministry of Science and Technology of Taiwan and National Taiwan University for the financial support under grant numbers NSC 99-2221-E-002-104-MY3, NSC 100-2221-E-002-158-MY3, NSC 100-2923-E-002-005-MY3, NSC 102-3113-P-002-027, NSC 103-2623-E-002-015-ET, MOST 103-3113-E-003-011, MOST 103-2221-E-002-132-MY3, NTU-ICRP-102R7558, NTU ICRP-103R7558, NTU-CESRP-102R7607-1, NTU CESRP-103R7607-1, MOST 104-3113-E-002-010, and MOST 104-3113-E-002-019. We also thank Prof. Chung-Chih Wu's group at the Department of Electrical Engineering, National Taiwan University, for the permission to use optical measurement instrument.

Notes and references

1. S. W. Glunz, R. Preu and D. Biro, Crystalline silicon solar cells: State-of-the-art and future developments, ed. W. van Sark, Comprehensive Renewable Energy, Elsevier, Oxford, 2012, pp. 1–63.
2. R. F. Service, Science, 2008, 319, 718.
3. J. Weickert, R. B. Dunbar, H. C. Hesse, W. Wiedemann and L. S. Mende, Adv. Mater., 2011, 23, 1810.
4. S. Shiu, J. J. Chao, S. C. Hung, C. L. Yeh and C. F. Lin, Chem. Mater., 2010, 22, 3108.
5. Y. Ding, R. Gresback, Q. Liu, S. Zhou, X. Pi and T. Nozakia, Nano Energy, 2014, 9, 25.
6. S. Kirchmeyer and K. Reuter, J. Mater. Chem., 2005, 15, 2077.
7. S. Garreau, J. L. Duvail and G. Louarn, Synth. Met., 2002, 125, 325.
8. T. Song, S. T. Lee and B. Sun, Nano Energy, 2012, 1, 654.
9. Y. Li, Q. Chen, D. He and J. Li, Nano Energy, 2014, 7, 10.
10. S. Thiyagu, C. C. Hsueh, C. T. Liu, H. J. Syu, C. T. Lin and C. F. Lin, Nanoscale, 2014, 6, 3361.
11. D. Sammito, P. Zilioa, G. Zaccoa, J. Janusonise and F. Romanatoa, Nano Energy, 2013, 2, 337.
12. 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.
13. L. Hong, X. Wang, H. Zheng, L. He, H. Wang, H. Yu and E. Rusli, Appl. Phys. Lett., 2014, 104, 053104.
14. J. Zhu, C. M. Hsu, Z. Yu, S. Fan and Y. Cui, Nano Lett., 2010, 10, 1979.
15. S. Thiyagu, B. P. Devi and Z. Pei, Nano Res., 2011, 4, 1136.
16. S. E. Han and G. Chen, Nano Lett., 2010, 10, 1012.
17. S. Thiyagu, B. P. Devi, Z. Pei, Y. H. Chen and J. C. Liu, Nanoscale Res. Lett., 2012, 7, 1.
18. H. J. Syu, S. C. Shiu, Y. J. Hung, C. C. Hsueh, T. C. Lin, S. Thiyagu, S. L. Lee and C. F. Lin, Prog. Photovoltaics, 2013, 21, 1400.
19. 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.
20. Y. Zhang, F. Zu, S. T. Lee, L. Liao, N. Zhao and B. Sun, Adv. Energy Mater., 2014, 4, 1300923.
21. X. Shen, B. Sun, D. Liu and S. T. Lee, J. Am. Chem. Soc., 2011, 133, 19408.
22. H. P. Wang, T. Y. Lin, C. W. Hsu, M. L. Tsai, C. H. Huang, W. R. Wei, M. Y. Huang, Y. J. Chien, P. C. Yang, C. W. Liu, L. J. Chou and J. H. He, ACS Nano, 2013, 7, 9325.
23. H. J. Syu, S. C. Shiu and C. F. Lin, Sol. Energy Mater. Sol. Cells, 2012, 98, 267.
24. 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.
25. L. He, C. Jiang, H. Wang, D. Lai and E. Rusli, ACS Appl. Mater. Interfaces, 2012, 4, 1704.
26. Q. Liu, M. Ono, Z. Tang, R. Ishikawa and K. Ueno, Appl. Phys. Lett., 2012, 100, 183901.
27. J. P. Thomas, L. Zhao, D. McGillivray and K. T. Leung, J. Mater. Chem. A, 2014, 2, 2383.
28. P. R. Pudasaini and A. A. Ayon, Microelectron. Eng., 2013, 110, 126.
29. C. Chartier, S. Bastide and C. Lévy-Clément, Electrochim. Acta, 2008, 53, 5509.
30. S. Thiyagu, H. J. Syu, C. C. Hsueh, C. T. Liu, C. T. Lin and C. F. Lin, RSC Adv., 2015, 5, 13224.
31. K. Q. Peng, X. Wang, L. Li, X. L Wu and S. T. Lee, J. Am. Chem. Soc., 2010, 132, 6872.
32. Z. Guo, J. Y. Jung, K. Zhou, Y. Xiao, S. W. Jee, S. A. Moiz and J. H Lee, Proc. SPIE, 2010, 7772, 77721C–1.
33. K. Sato, M. Dutta and N. Fukata, Nanoscale, 2014, 6, 6092.
34. T. G. Chen, B. Y. Huang, E. C. Chen, P. Yu and H. F. Meng, Appl. Phys. Lett., 2012, 101, 033301–033305.
35. F. Zhang, T. Song and B. Sun, Nanotechnology, 2012, 23, 194006–194009.
36. P. R. Pudasaini, F. R. Zepeda, M. Sharma, D. Elam, A. Ponce and A. A. Ayon, ACS Appl. Mater. Interfaces, 2013, 5, 9620.
37. Y. Zhang, W. Cui, Y. Zhu, F. Zu, L. Liao, S. T Lee and B. Sun, Energy Environ. Sci., 2015, 8, 297–302.
38. P. R. Pudasaini, D. Elam and A. A. Ayon, J. Phys. D: Appl. Phys., 2013, 46, 235104.
39. M. Sharma, P. R. Pudasaini, F. R. Zepeda, D. Elam and A. A. Ayon, ACS Appl. Mater. Interfaces, 2014, 6, 4356.
40. Y. Xiaa and J. Ouyang, J. Mater. Chem., 2011, 21, 4927.
41. Q. Liu, T. Imamura, T. Hiate, I. Khatri, Z. Tang, R. Ishikawa, K. Ueno and H. Shirai, Appl. Phys. Lett., 2013, 102, 243902.

---