Dependence of Performance of Si Nanowire Solar Cells on Geometry of the Nanowires

The dependence of performance of silicon nanowires (SiNWs) solar cells on the growth condition of the SiNWs has been described. Metal-assisted electroless etching (MAE) technique has been used to grow SiNWs array. Different concentration of aqueous solution containing AgNO3 and HF for Ag deposition is used. The diameter and density of SiNWs are found to be dependent on concentration of solution used for Ag deposition. The diameter and density of SiNWs have been used to calculate the filling ratio of the SINWs arrays. The filling ratio is increased with increase in AgNO3 concentration, whereas it is decreased with increase in HF concentration. The minimum reflectance value achieved is ~1% for SiNWs of length of ~1.2 μm in the wavelength range of 300–1000 nm. The performance and diode parameters strongly depend on the geometry of SiNWs. The maximum short circuit current density achieved is 35.6 mA/cm2. The conversion efficiency of solar cell is 9.73% for SiNWs with length, diameter, and wire density of ~1.2 μm, ~75 nm, and 90 μm−2, respectively.


Introduction
The increasing demand for energy and the effect of global warming are two related issues attracting more and more attention from human being. Photovoltaic technology is the most promising approach to carbon-free energy production and it has received more attention [1]. Nanostructures can be used as solution for better photovoltaic application [2]. The photovoltaic based on silicon nanowires (SiNWs) are becoming more attractive due to their optical and electrical properties compared to bulk Si [3,4]. The important role of SiNWs arrays is to reduce reflection losses in photovoltaic devices. The reduction in the reflectance of SiNWs arrays is due to improved optical impedance matching between air and Si [5,6], multiple incident light scattering inside the arrays [7], and the role of individual SiNWs as nanoscale resonators with the antenna effect [8]. Several techniques have been reported in the literature to synthesize the SiNWs [9][10][11].
Several studies have been made to design the structural parameters, namely, filling ratio, wire density, and diameter, to enhance light absorption in SiNWs solar cells [12,13]. Some of the researchers [14,15] have done the simulation for diameter and periodicity ratio dependence on the optical absorption in the SiNWs arrays [16][17][18]. Previously, some of the groups [19][20][21] investigated that randomly distributed SiNWs array gives better absorption than the regularly distributed one due to stronger scattering. Metal-assisted electroless etching (MAE) technique has been recently developed for fabrication of randomly distributed SiNWs arrays [22][23][24]. It is important to design particular diameter and wire density (periodicity) to achieve minimum reflectance and hence to get the highest current density for SiNWs based solar cells. Secondly, reduction of the recombination and resistive losses are equally important to get the better conversion efficiency. The reflectance can be easily reduced by increasing the length of SiNWs but recombination and resistive losses also increase with the increase in SiNWs length. Thus, the SiNWs of smaller length with ultralow reflectance are potentially feasible for antireflection applications in next-generation solar cells [22][23][24][25][26]. Hence, it is important to optimize the geometrical parameters for obtaining SiNWs arrays as superior antireflective characteristics with minimum recombination and resistive losses. Recently, Jung et al. [27] have optimized Ag deposition time to get maximum conversion efficiency.

2
The Scientific World Journal In this paper, we have tried to achieve minimum reflectance ∼1% in the wavelength range 300-1000 nm by engineering the diameter and density of SiNWs of length ∼1.2 m to get the minimum reflection, recombination and resistive losses.

Synthesis of SiNWs.
SiNWs arrays were synthesized on Cz grown, boron doped p-type, 0.2-0.6 Ω cm resistivity, 550 mm thickness, and <100> oriented polished Si wafers using two steps metal-assisted wet etching (metal deposition and etching) method. First, the substrates were cleaned in hot solution of H 2 O 2 and H 2 SO 4 (1 : 3 by volume) for 10 min. Then, the substrates were dipped in dilute HF to remove oxide grown in the cleaning process. The silver nanoparticles (NPs) were uniformly deposited onto the Si substrate using an aqueous solution of AgNO 3 and HF. Two types of solutions (A and B) were used separately for Ag deposition for two sets of experiments. In set A, the concentration of AgNO 3 was taken as 0.01, 0.02, 0.03, and 0.04 M (i.e., set A 1 , A 2 , A 3 , and A 4 ) and the concentration of HF was constant (4.6 M). In set B, the concentration of AgNO 3 was constant (0.03 M) and the concentration of HF was taken as 2.3, 4.6, 6.9, and 9.2 M (i.e., set B 1 , B 2 , B 3 , and B 4 ). Henceforth, we shall refer to the samples simply as A 1 , A 2 , A 3 , A 4 , B 1 , B 2 , B 3 , and B 4 in the text. It can be noted that the samples A 3 and B 2 were identical because the solution concentration was same. In the Ag deposition process, the Ag ions captured electrons from Si and reduced to Ag metal NPs. These nanoparticles are deposited on the silicon surfaces. The areal density of Ag NPs depends upon solution concentration, deposition time, and temperature. The etching was performed at room temperature in a mixed solution of deionized water (DIW), HF (4.6 M), and H 2 O 2 (0.5 M) to obtain a vertically aligned SiNWs array. After completing the electroless etching, SiNW samples were dipped in concentrated HNO 3 for 10 min to completely remove the remaining Ag NPs, then rinsed in DIW, and dried in air. The filling ratio was calculated using method of Jung et al. [27].

Solar Cell Fabrication.
The SiNWs samples were subjected to RCA cleaning before the fabrication process. After RCA cleaning, the front and back sides were simultaneously diffused by phosphorous and boron using phosphorussilicate and borosilicate glass to create n + -p-p + structure by spin on dopant (SOD) technique. Phosphorus and boron silicate precursors (P509 for P and B200 for B, supplied by Filmtronics) were spin-coated on different dummy wafers. Then, the SiNWs samples were loaded between the dummy wafers in a conventional quartz-tube furnace. The diffusion was carried out at 1000 ∘ C for 5 min, while the target samples were kept at a closely spaced distance.
After diffusion, the phosphosilica glass (PSG) and borosilica glass (BSG) were removed simply by immersing the prepared specimens in buffer oxide etchant (BOE) for 10 min. The patterned front gridded metal contacts were made by Ti/Ag using metal mask. The back contacts were made using Ti/Ag on the full surface. The contacts were sintered in rapid thermal processing unit (RTP) at 600 ∘ C. Finally, edges of the solar cells were scribed for isolation of junction.

Characterization Techniques.
The field emission scanning electron microscopy (FE-SEM) images were taken for front and cross-sectional views of SiNWs samples using Hitachi FE-SEM Model S-4800. The reflectance of SiNWs samples was measured in the wavelength range of 300-1200 nm with Perkin-Elmer Model Lambda 750 using an integrated sphere.
The illuminated current density-voltage ( -) characteristics of solar cells were measured under 100 mW/cm 2 intensity with AM1.5 Global spectrum using a Keithley sourcemeter Model 2400 (with basic accuracy 0.012%) and Newport 91192 solar simulator system (equipped with 1KW Xenon arc lamp from Oriel). The illumination intensity was measured using a reference silicon solar cell obtained from PV Measurements, USA. All the measurements were carried out at 25 ∘ C.  Figure 3. The increase in wire density may be due to increase in number of Ag nanoparticles with increase in AgNO 3 concentration. The reduction in wire diameter may be due to decrease in the Ag nanoparticles size with increase in AgNO 3 concentration. The variation of wire diameter and density with HF concentrations is shown in Figure 4. Both the wire diameter and density are decreased with increase in HF concentration. The diameter of the wires is initially decreased with fast rate then becomes slower. The wire density is decreased linearly with increase in HF concentration. The value of diameters and densities of wires has been used to calculate the filling ratio of SiNWs arrays. The variation of filling ratio of SiNWs with AgNO 3 and HF concentration are shown in Figure 5. The filling ratio is linearly increased with increase in AgNO 3 concentration, whereas it is exponentially decreased with the increase in HF concentration.

Dependence of Reflectance of SiNWs Arrays on Growth
Condition. The reflectance spectra of all the samples are shown in Figure 6. The minimum reflectance of ∼1% is observed for samples A 3 and B 2 , B 3 , and B 4 in the wavelength range 600-800 nm, whereas, in the wavelength range of 300-1000 nm, sample B 3 has the lowest reflectance value. It can be noted that the minimum reflectance is obtained for filling ratio of ∼40%. In case of samples B 3    The Scientific World Journal  Figure 3: Variation of average diameters and density of SiNWs with AgNO 3 concentration. (∼1.5 m) of nanowires is greater than in the all others samples (∼1.2 m) for the same Ag deposition time and the same etching conditions. The highest (∼2.43%) and the lowest (∼ 0.93%) values of reflectance are obtained for the samples A 1 and B 3 , respectively.

Analysis of Performance Parameters.
The illuminated -curves obtained at 25 ∘ C under illumination condition of 100 mW/cm 2 of simulated AM1.5G solar radiation of the above samples are shown in Figure 7. It can be seen from Figure 7 the maximum value of conversion efficiency obtained for samples A 3 and B 2 . The performance parameters, namely, short circuit current density sc , open circuit voltage oc , curve factor CF, and conversion efficiency of the solar cells, are listed in Table 1. The current densities of samples B 3 and B 4 are higher than those of the samples A 3 and B 2 due to lowest reflection in the wavelength range of 300-1200 nm but conversion efficiencies are very low due to low CF values. The highest and the lowest sc values are obtained as 35.6 and 25.0 mA/cm 2 for the samples A 1 and B 4 , respectively. The maximum achieved oc value is 0.534 V for samples A 3 and B 2 . The minimum and maximum CF

Conclusions
Dependence of SiNWs solar cells performance on the concentration of Ag deposition solution has been investigated. The filling ratio of SiNWs array changed with AgNO 3 and 6 The Scientific World Journal HF concentration. The diameter of SiNWs is decreased, whereas wire density is increased linearly with increase in AgNO 3 concentration. Both the diameter and wire density are decreased with increase in HF concentration. The filling ratio of the SiNWs array is increased with increase in AgNO 3 concentration, whereas it is decreased with increase in HF concentration. The minimum reflectance value of ∼1% is achieved for the filling ratio of ∼40%. It has been observed that the performance and diode parameters strongly depend on the solution concentration used for Ag nanoparticle deposition. A maximum short circuit current density of 35.6 mA/cm 2 is obtained for the sample B 4 . The highest conversion efficiency of 9.73% is achieved with CF of 0.531 for aqueous solution of HF (4.6 M) and AgNO 3 (0.03 M). The efficiency can be improved with improving the CF values using better process techniques.