Performance dependence of SWCNT/n-silicon hybrid solar cells on the charge carrier concentration in silicon substrates

Abstract

In this work, single-walled carbon nanotube (SWCNT)/n-silicon based hybrid solar cells have been fabricated and characterized in order to investigate the effect of different doping levels/charge carrier concentrations in silicon. The SWCNT films were characterized by UV-Vis-NIR spectroscopy for optical absorption, photoluminescence spectroscopy for chirality determination, and resistivity measurements. The n-type silicon substrates were characterized by Hall-effect measurements for mobility of charges, charge carrier concentration, and resistivity measurements. Atomic force microscopy and scanning electron microscopy were employed to study the morphology of SWCNT films. The solar cells were characterized by current–voltage measurements under AM1.5 solar irradiance. We have found that a heavily doped silicon substrate does not result in a functional solar cell, whereas lightly doped silicon substrate produces much less photocurrent. The SWCNT/n-silicon solar cells with silicon substrate resistivity of 1.82 Ω cm and change carrier concentration of 2.33 × 10¹⁵ produced the best power conversion efficiency of 2.35% among the silicon substrates under investigation. These studies provide sufficient insight into material selection parameters for the SWCNT/n-silicon solar cell architectures to optimize device efficiencies.

---

Introduction

Inorganic solar cells, primarily silicon based, occupy more than 90% of the solar energy generation market.¹ The efficiency of common commercially available silicon based solar cells ranges from 15% to 20%. They have dominated the market due to their stability in various environments and the abundance of raw materials, namely silicon. The only drawback for such types of solar cells is their relatively high cost, and hence the energy produced per unit watt is higher than conventional means of energy generation techniques.¹ In order to further bring the fabrication cost down and reduce the production complexity, novel materials such as carbon nanostructures have been proposed as photoactive layers rather than atomic implantation of silicon substrates to produce junctions capable of photovoltaic activity. These carbon nanostructured materials have the advantage of easy deposition onto silicon substrates, using cost effective methods for processing and film formation.^2–12 These types of solar cells constitute a new class of devices: organic–inorganic hybrid solar cells. They are typically fabricated from organic materials (such as carbon nanomaterials) and inorganic materials (such as silicon). In this type of device, the carbon nanomaterials form a layered heterojunction with silicon substrate, creating a built-in potential at the junction which promotes the dissociation of excitons.^2–6 The development of this architecture promises to result in an economical and efficient solar cell due to its simple fabrication procedure, low production cost, chemical and environmental stability, and could result in a device that is comparable to the conventional inorganic or organic solar cells. The main drop in cost is due to the low junction formation temperature compared to the silicon p–n junction that requires high temperature during the diffusion process. These hybrid photovoltaic devices benefit from the advantageous properties of materials, such as single-walled carbon nanotubes (SWCNTs), that possess high aspect ratio,¹³ large surface area,¹⁴ wide range of direct bandgaps,^15,16 strong photoabsorption from infrared to the ultraviolet region,^17,18 high carrier mobility¹⁹ and reduced carrier transport scattering.²⁰ These characteristics make SWCNTs ideal materials for various components in electronic applications and solar cell devices.^21–23 The high aspect ratio and large surface area of SWCNTs are beneficial in exciton dissociation, whereas high carrier mobility and reduced carrier transport scattering helps to transport the photogenerated charge carriers away from the heterojunction. Semiconducting SWCNTs possess a wide range of direct bandgaps and strong photoabsorption, which is favorable in efficiently capturing a wide solar spectrum.^17,18 The maximum power conversion efficiencies reported thus far for these device architectures are up to 15%.^24–27 It is worth mentioning here that these solar cell devices are not robust and are prone to scratches. The carbon nanotube film can be wiped off the silicon surface by a gentle scratch or touch. A passivating scratch resistant encapsulation would be required for real world applications of these hybrid solar cells, similar to those found on commercial inorganic solar cells.

A fair amount of work has been performed on the development of such solar cell devices; including varying the type of carbon nanotubes (single-, double- and multi-walled carbon nanotubes),^12,28–30 adjusting the thickness of carbon nanotube films,^3,4,12 doping carbon nanotubes,^2–4,26 modulating the interfacial oxide layer,^31,32 using electrolytes^33,34 and modifying the electrodes.³⁵ However, to the best of our knowledge, no previous studies have reported the impact of doping levels in the silicon substrate on the overall performance of the SWCNT/n-silicon hybrid solar cells. Therefore, in this work SWCNT/n-silicon hybrid solar cells were investigated by varying doping concentrations (charge carrier concentration) of the silicon substrate. We have used six n-type silicon substrates with different doping level concentrations, and kept other parameters, such as the type of carbon nanotubes and device dimensions, constant throughout the studies. The SWCNT films were deposited on silicon substrates and their morphology, transparency and sheet resistivity were analyzed. The silicon substrates used for these studies were characterized by using Hall-effect measurements to determine the mobility of the charge carriers, charge carrier concentration, and resistivity of the substrate material. The characteristics of the SWCNT film and Si-substrates were correlated with the observed photovoltaic device properties.

Experimental details

The CoMoCat SWCNTs used in this work were purchased from Sigma Aldrich (704148, SWeNT-SG65). According to the manufacturer specifications 90% of SWCNTs were semiconducting in nature with the majority being of (6,5) and (7,5) chirality having diameters of 0.8 ± 0.1 nm. The n-type silicon substrates were purchased from WRS Materials and were supplied with the specification of six different ranges of resistivities: <0.003 Ω cm, 0.01–0.02 Ω cm, 1–3 Ω cm, 10–20 Ω cm, 87.5–112.5 Ω cm and 200–1000 Ω cm. The exact impurity doping concentration was not provided by the manufacturer; therefore, Hall-effect measurements were employed to determine the charge carrier concentration as well as the exact value of resistivity. In order to fabricate SWCNT/n-silicon hybrid solar cells, the SWCNTs were dispersed in dimethylformamide (DMF) (0.1 mg mL⁻¹) by using bath sonication followed by centrifugation at 20 000 rpm for 30 min to remove any catalyst nanoparticles and large SWCNTs bundles. The top 80% of the supernatant was decanted and used for further processing. A small amount of SWCNT dispersion was then vacuum filtered through alumina membrane in the form of thin film. The alumina membrane was dissolved in aqueous sodium hydroxide (NaOH) solution (3 M) and the freely floating SWCNT film was washed with plenty of de-ionized water. Separately, a 300 nm silicon dioxide was grown on silicon samples by thermal oxidation. Next, 5 nm Titanium (Ti) film was deposited by e-beam deposition followed by sputter deposition of 100 nm gold (Au) film. Ti was used as an adhesion layer. The samples were patterned using lithography and Au was etched using aqua regia HCl : HNO₃ (3 : 1). A 0.8 cm × 0.8 cm window was etched in silicon dioxide by dipping samples in 3.36% hydrofluoric acid for 2 minutes. This is a complementary step which also removes Ti film before silicon dioxide can be etched away. After rinsing the samples in deionized water and removing the photoresist by using acetone, the silicon samples were ready for next step. For solar cell fabrication, the SWCNT film was lifted on silicon substrates with predeposited and patterned Au/Ti electrodes and then placed on a hot plate heater (90 °C). With the evaporation of water, the SWCNTs film comes in intimate contact with silicon substrate and creates numerous micro- and nanojunctions. The Au/Ti electrode was deposited to make electrical contact with SWCNT film whereas indium–gallium eutectic was used to make electrical contact with n-silicon substrate. Current–voltage measurements were performed to ensure the ohmic back side contact for all the silicon substrates. Fig. 1 shows the schematic diagram of SWCNTs film deposition on n-silicon substrate.

Fig. 1 Schematics showing, (a) the deposition process of SWCNT film on patterned n-silicon substrate, (b) the solar cell after deposition of SWCNT film and (c) side view of device architecture.

The SWCNT film was lifted on silicon and glass substrates simultaneously to achieve similar SWCNT film thickness. The silicon substrates deposited with SWCNT films were used for photovoltaic device fabrication and SEM imaging, whereas glass substrates with SWCNT films were used for sheet resistivity measurements and to perform optical transmission studies using UV-Vis-NIR spectrometer. The six silicon substrates used for these studies are herein identified as Si-1, Si-2, Si-3, Si-4, Si-5 and Si-6 (with resistivity specifications: <0.003 Ω cm, 0.01–0.02 Ω cm, 1–3 Ω cm, 10–20 Ω cm, 87.5–112.5 Ω cm and 200–1000 Ω cm, respectively).

Scanning electron microscopy (SEM) was performed using JEOL 7000F under an accelerating voltage of 15 keV. The samples were imaged without any special surface treatment and the as-prepared silicon substrates were placed on aluminum stubs with double sided carbon tape for SEM analysis. Photoluminescence spectroscopy is a very reliable technique to determine the chirality of semiconducting SWCNTs³⁶ and was utilized for the characterization of the SWCNTs in this study. The SWCNTs were dispersed in aqueous sodium cholate solution and characterized using a NanoLog fluorescence spectrometer (HORIBA Jobin Yvon) to determine their chirality. The NanoLog is equipped with a Xenon lamp excitation source (300–800 nm) and a liquid nitrogen cooled InGaAs array detector to detect emission wavelengths from 900 nm to 1500 nm. The transmission spectra of SWCNT films deposited on glass substrates were recorded in the UV-Vis-NIR range using a Shimadzu UV-3600 double-beam spectrophotometer with three detectors. Veeco Dimension Nanoscope 3100 atomic force microscopy (AFM) was used to carry out a surface roughness analysis and a cross-sectional step-height analysis for determination of thickness of the SWCNT film. AFM measurements were taken at six different places on the SWCNT film. Hall effect measurements provide important information concerning the conductivity type of charge carriers, charge carrier mobility, charge carrier concentration, and the sheet resistivity of silicon substrates. The Hall-effect measurements were performed on the silicon substrates using an Ecopia HMS-5000 Hall-effect measurement system with magnetic flux density of 0.55 T at room temperature. For Hall-effect measurements the silicon substrates were cut into 1 cm × 1 cm squares, and the contacts were made using indium in a van der Pauw configuration. The current–voltage characteristics were measured under a Class B AM1.5 solar simulator with an incident illumination power of 100 mW cm⁻² (PV Measurements, Inc.), and a Keithley 2400 Source Meter was used for electrical measurements. The incident light power was calibrated to 100 mW cm⁻² using NREL certified reference cell.

Results and discussion

In order to explain the effects of charge carrier concentrations in silicon substrates on the performance of SWCNT/n-silicon solar cells, it is important to investigate the characteristics of the actual materials (SWCNTs and silicon) used for the fabrication of these devices. First, the SWCNTs used in these studies were characterized by photoluminescence and Raman spectroscopy with the results shown in Fig. 2.

Fig. 2 (a) Photoluminescence of SWCNTs demonstrating that the majority of them consist of the (6,5) chirality. (b) Raman spectrum of SWCNTs obtained with 633 nm laser (red line) and 785 nm laser (black line) excitation. Inset shows the RBM modes observed with 633 nm laser (red line) and 785 nm laser (black line) excitation, respectively.

The 2D contour plot of the excitation–emission matrix scan revealed the photoluminescence characteristics of the SWCNTs as shown in Fig. 2a. The chirality of the SWCNTs (assigned by using HORIBA Jobin Yvon's Nanosizer software) is given by their (n,m) values, with the emission peak wavelength related to the diameter and chiral angle of the nanotubes.³⁷ The majority of the SWCNT species were found to be of (6,5) chirality, while a few species of (7,5) and (8,3) were also found, and a fraction of (8,4) were present in the SWCNT samples. Raman analysis of the SWCNTs used in these studies is shown in Fig. 2b with 633 nm and 785 nm as the laser excitation wavelength. The spectra consist of a Radial Breathing Mode (RBM) (100 cm⁻¹ to 400 cm⁻¹), a D-band (1300 cm⁻¹) due to the presence of defects, longitudinal and tangential G-band (1585 cm⁻¹) arising from graphite like structures, and a 2D-band (2600 cm⁻¹) that is a second harmonic of the D-band.³⁸ The ratio of the intensity of the G-band and D-band (I_G/I_D value) provides an insight in the density of defects present in the SWCNTs.³⁹ It was found that I_G/I_D is 17.82 under 633 nm laser excitation and 15.79 under 785 nm laser excitation. This confirms that the SWCNTs used in these studies had low defect density along with high crystallinity. The inset in Fig. 2b shows the radial breathing mode (RBM) of SWCNTs observed with 633 nm and 785 nm laser excitation. The RBM of SWCNTs is useful to determine the tube diameter using the following relation:³⁶

d_t = C₁/(ω_RBM − C₂) (1)

where d_t is the diameter of SWCNTs, ω_RBM is the wavenumber of laser radiation used, and C₁ and C₂ are 223.5 nm cm⁻¹ and 12.5 cm⁻¹, respectively.³⁶ The 633 nm laser excitation generates three major RBM modes which are assigned to SWCNTs having diameters 1.11 nm, 0.926 nm, and 0.833 nm. The 785 nm laser excitation generates two major RBM modes that are assigned to SWCNTs with diameters of 0.877 and 0.761 nm. As shown in the inset of Fig. 2b, the intensity of the RBM modes generated using a 785 nm laser is much higher than those generated using a 633 nm laser. The theoretically calculated diameter of a (6,5) chiral SWCNT is 0.753 nm and that of a (7,5) chiral SWCNT is 0.823 nm.⁴⁰ These diameters are very close to the SWCNT diameter calculated using a 785 nm laser excitation in the RBM mode of the Raman spectrum. Since, photoluminescence excitation only determine semiconducting SWCNT species and Raman analysis was available only from lasers excitations in resonance with semiconducting SWCNTs, we performed optical absorption studies on SWCNT dispersion, before the film fabrication, in order to estimate the concentration of the metallic species present in our samples (see ESI Fig. S1†). It was found that about 8.66% SWCNT species in the sample were metallic in nature. The metallic content estimated is very close to the manufacturer specifications (90% semiconducting).

Scanning electron microscopy is a direct method to visually obtain the morphology of SWCNT films. Fig. 3 shows the side and top view SEM images of SWCNT film. As can be observed from the SEM images, the SWCNTs are connected to each other in a random manner and forms percolation length. The SWCNTs are also free of large bundles, catalyst metal nanoparticles and amorphous carbon impurities. The SWCNTs are in intimate contact with underlying silicon substrate (attached with underlying silicon by van der Waals force), thereby creating SWCNT–silicon heterojunctions. The SWCNT films were observed to have a rough surface with relatively high porosity, which is believed to be useful for SWCNT/n-silicon solar cells, since the light can pass through the SWCNT mats and reach the SWCNT–silicon heterojunction with less interference.

Fig. 3 (a) Side view and (b) top view SEM images of SWCNT film deposited over silicon substrate demonstrating the SWCNT film morphology and SWCNT/n-silicon heterojunction.

Fig. 4a shows the optical transmission spectrum of SWNT films in the wavelength range of 350 nm and 1650 nm. As previously observed in photoluminescence studies, it is evident from the optical transmission spectrum that the majority of SWNTs in the films are of (6,5) and (7,5) chirality. The Van Hove singularities calculated for (6,5) chiral SWNTs is S₁₁ – 1.207 eV and S₂₂ – 2.12 eV, whereas, for (7,5) chiral SWNTs, it is calculated as S₁₁ – 1.026 eV and S₂₂ – 1.832 eV which is in good agreement with the literature.⁴¹ The transparency (at 550 nm wavelength) and resistivity of the SWCNT film was 66% and 2.15 kΩ, respectively. The resistivity of SWNT film is dependent on its transparency (or thickness), and as the transparency of the SWNT film decreases (or the thickness increases), the resistivity of the SWNT film also decreases. The thickness of SWNT film was determined via a cross sectional step height analysis using AFM. The step height measurements were taken at six different places on the film. The average step height and RMS roughness were found to be 56.13 nm and 23.38 nm, respectively. A typical step height image is shown in Fig. 4b.

Fig. 4 (a) Optical transmission spectrum of SWCNT films with 66% transparency at 550 nm wavelength. (b) AFM image of SWCNT film and line scans of the same SWCNT film for thickness measurement.

The results obtained from Hall effect measurements for the six different silicon substrates are shown in Fig. 5. The reproducibility of Hall effect results shown in Fig. 5 were tested for each sample by taking ten measurements. It was observed that all the measurements resulted with similar numbers. Therefore, the accuracy of Hall-effect results was testified. The negative Hall coefficient values (not shown here) for all of the silicon substrates indicated they were n-type.

Fig. 5 (a) Charge carrier concentration, (b) resistivity, and (c) mobility of charge carrier in silicon substrates calculated using a Hall effect measurement system.

The SWCNT/n-Si heterojunction solar cells were characterized by performing current–voltage measurements in the dark and under AM1.5 illumination. The measurements were performed using a Keithley 2400 source meter by sourcing voltage and measuring current. When light is shone on SWCNT/n-Si heterojunction solar cells, the photons are absorbed and excitons are produced in semiconducting SWCNTs and n-silicon.^2–4 Currently, it is a matter of debate whether excitons produced in SWCNTs take part in current generation or not.^28,42,43 The excitons diffuse to the SWCNT/n-silicon heterojunctions due to their high mobility⁴⁴ and diffusion length.^45,46

At the SWCNT/n-silicon heterojunction, the excitons dissociate into free electrons and holes due to built-in potential at the junction (Fig. 6a). Next, the electrons are transported to the cathode through n-silicon, and holes are transported through the SWCNT network to anode as illustrated in Fig. 6b. Fig. 6c compares the open-circuit voltage (V_OC), short-circuit current density (J_SC), fill factor (FF) and power conversion efficiency of various SWCNT/n-silicon solar cells. Fig. 6d and e presents the current–voltage response of these devices under illumination and dark, respectively. The current–voltage response revealed that devices fabricated on substrate Si-1 exhibited no photovoltaic response, whereas all of the other devices produced photocurrent. The Si-3 device demonstrated the best results due to high open-circuit voltage (0.405 V), fill factor (0.57) and efficiency (2.35%). Whereas Si-5 device exhibits the maximum short-circuit current density (14.90 mA cm⁻²) but the overall power conversion efficiency (1.43%) was lower due to low open-circuit voltage (0.297 V) and fill factor (0.33). All the solar cells figure of merit parameters are tabulated in Table S1 in the ESI.†

Fig. 6 (a) Schematic diagram showing dissociation of excitons at the SWCNT/n-silicon interface. (b) Holes transport through SWCNTs, and electrons transport through n-silicon substrate to respective electrodes. (c) Comparison of V_OC, J_SC, FF and efficiency of devices with varying charge carrier concentration in silicon. (d) and (e) Current–voltage response of SWCNT/n-Si solar cell devices under illumination and dark, respectively.

The photovoltaic device characteristics are dependent on the properties of the materials, such as the optical absorption/transparency, electrical resistivity, etc., in fabricating the device. Although open circuit voltage depends upon the series and shunt resistances of a solar cell, it can be also approximated as the difference between the Fermi levels of the two materials:

V_OC = E_fSWCNT − E_fSi (2)

Ideally, carbon nanotubes are expected to have a Fermi level of around 4.5 eV; Y. R. Park, W. J. Kim, M. J. Ko, N. K. Min and C. J. Lee, Nanoscale, 2012, 4, 6532–6536 however, these nanostructures behave more like a p-type material in air, and the Fermi level would shift towards the valence band for SWCNTs. The intrinsic properties of the silicon greatly influence the figures of merit in a photovoltaic device, as explained earlier with the relation between V_OC and Fermi levels of materials. The charge carrier concentration, mobility of electrons, and resistivity of the best-performing silicon substrate (Si-3) were 2.33 × 10¹⁵ cm⁻³, 1470 cm² V⁻¹ s⁻¹, and 1.82 Ω cm, respectively. To understand the effect of silicon charge carrier concentration on the solar cell's performance, we calculated the Fermi levels of the silicon substrates using eqn (3):⁴⁷

where N_D is the bulk concentration calculated from Hall-effect measurements; n_i is the intrinsic carrier concentration and given as follows:⁴⁷

The calculated Fermi-level values are tabulated in Table 1. The bandgap of silicon was considered to be 1.12 eV and the electron affinity to be 4.05 eV at room temperature.

Table 1 Charge carrier concentrations of silicon substrates, calculated Fermi energy levels of silicon substrates, and resulting depletion region width in solar cell devices

Silicon substrate	Charge carrier concentration (cm^−3)	Fermi level (eV)	Depletion region width (μm)	Solar cell efficiency (%)
Si-1	1.32 × 10^19	4.07	0	0
Si-2	1.75 × 10^18	4.12	0.002	0.02
Si-3	2.33 × 10^15	4.29	0.478	2.35
Si-4	2.61 × 10^14	4.35	1.217	1.57
Si-5	4.46 × 10^13	4.40	2.955	1.43
Si-6	1.1 × 10^13	4.43	2.723	0.004

---

The Fermi energy level of the silicon substrates used in these studies varied from 4.07 eV to 4.43 eV. By using the values of open-circuit voltage, the silicon Fermi level from Table 1, and eqn (2), the Fermi level for SWCNTs can be approximated. Devices Si-3, Si-4 and Si-5 exhibit significant open-circuit voltage and power conversion efficiencies among all devices, therefore the Fermi energy level of SWCNTs was estimated to be 4.69 eV, 4.64 eV and 4.70 eV, which are rather close to each other.

Since the SWCNT films used in these studies originate from the same source and variation in Fermi level of SWCNT is small, the variation in device characteristics can be attributed largely to the charge carrier concentration in the silicon substrates. According to p–n junction theory, the charge carrier concentration, N, is given by the following:^48,49

where ρ, q, and μ are the resistivity, electronic charge, and carrier mobility, respectively. This equation reveals that an increase in charge carrier concentration will result in a decrease in the resistivity.

It is also possible to probe the depletion region width in silicon, since it is affected by the SWCNT films, resulting in the generation of built-in potential at the SWCNT/n-silicon junction. Due to this phenomenon, the photovoltaic effect may occur when light is shone on the junction. The built-in potential can be described as the maximum open-circuit voltage that can be obtained from a solar cell. The relationship between built-in potential (V_b) and open-circuit voltage (V_OC) is expressed as follows:⁵⁰

where n is diode factor, k is Boltzman constant, T is temperature, and x is a constant. At room temperature, V_b is directly dependent on V_OC. The depletion region width (W) can be expressed as follows:⁵¹where V_b is the built-in potential, q is electronic charge, ε_Si is silicon dielectric constant, and N is the donor density (approximately equals to charge carrier concentration at room temperature). Since V_b is the maximum V_OC obtained from a solar cell at room temperature, by determining the charge carrier concentration from Hall-effect measurements and assuming that the V_OC is equal to V_b, we can approximate the depletion region width from the eqn (7). The calculated depletion region widths for all silicon substrates are shown in Table 1.

An increase in charge carrier concentration and a decrease in depletion region width indicate that there is more charge accumulation at the edge of the depletion region in the silicon. It has been shown in a previous report⁴ that the charge transportation between SWCNTs and silicon follows the tunneling model. Since the depletion region width in Si-2 was only 0.002 μm, the charge carriers attained enough energy at room temperature to “hop over” or “tunnel through” the thin junction without any external bias. This could be the reason that J–V curve of Si-2 (Fig. 6d) exhibits very low shunt resistance and did not display significant open-circuit voltage nor power conversion efficiency (Table S1†).

Fig. 6c displays the short-circuit current density increase from Si-1 to Si-5 based devices. The increased short-circuit current density has been attributed to the decreased charge carrier concentration and wider depletion region width, which resulted in dissociation of larger number of photoexcited charge carriers. Si-6 shows poor short-circuit current density and high series resistance due to low conductivity of the silicon substrate resulting in high recombination and reduced photo-induced charge carriers (Table S1†). The Si-3 based device exhibited higher power conversion efficiency compared to any other device due to a favorable depletion region width and charge carrier concentration, giving rise to high open circuit voltage and fill factor with decent short-circuit current density.

Conclusions

After studying the results obtained from all of the devices, it was determined that the silicon substrate plays a very important role in the design of SWCNT/n-Si heterojunction solar cells. p-type SWCNTs and n-type silicon creates numerous heterojunctions which can act as photo-generated charge dissociation centers. We limit the type of SWCNTs used in these studies mainly to (6,5) and (7,5) chiralities and varied the charge carrier concentration in the silicon substrate. The resistivity of silicon substrates is affected by the doping concentration which changes the silicon Fermi energy level. The variation in charge carrier concentration (substrate resistivity) was correlated with the change in Fermi energy level, as well as the power conversion efficiency of SWCNT/n-Si heterojunction solar cells. A silicon substrate resistivity of ∼1.82 Ω cm was found to operate at maximum power conversion efficiency. A high charge carrier concentration in silicon substrates leads to a narrowing of the depletion regions, which results in the tunneling or hopping of charge carriers over the built-in electric field, and hence a non-functional SWCNT/n-Si solar cell. Decrease in charge carrier concentration reduces the photo-induced charge carriers and increase the series resistance of SWCNT/n-Si solar cells, resulting in reduced open-circuit voltage and short-circuit current density, hence low power conversion efficiency.

Investigations of SWCNT/silicon solar cells have yielded results comparable to organic photovoltaics and dye-sensitized solar cells. Likewise organic photovoltaics and dye sensitized solar cells; these hybrid solar cells also have the potential to produce high performing photovoltaic devices. These studies demonstrate that consideration should be given to the properties of the silicon substrate when performing scientific inquiries in order to further improve the performance of silicon-based heterojunction solar cells.

Acknowledgements

The financial support from the Arkansas Science & Technology Authority (Grant # 08-CAT-03), and the Department of Energy (DE-FG36-06GO86072) and National Science Foundation (NSF/EPS-1003970) is greatly appreciated. The editorial assistance of Dr Marinelle Ringer is acknowledged. We also appreciate Mr Zeid Nima for his help in designing the graphical abstract.

Notes and references

1. M. A. Green, Sol. Energy, 2003, 74, 181–192.
2. Z. Li, V. P. Kunets, V. Saini, Y. Xu, E. Dervishi, G. J. Salamo, A. R. Biris and A. S. Biris, Appl. Phys. Lett., 2008, 93, 243117.
3. Z. Li, V. P. Kunets, V. Saini, Y. Xu, E. Dervishi, G. J. Salamo, A. R. Biris and A. S. Biris, ACS Nano, 2009, 3, 1407–1414.
4. V. Saini, Z. Li, S. Bourdo, V. P. Kunets, S. Trigwell, A. Couraud, J. Rioux, C. Boyer, V. Nteziyaremye, E. Dervishi, A. R. Biris, G. J. Salamo, T. Viswanathan and A. S. Biris, J. Appl. Phys., 2011, 109, 014321.
5. S. E. Bourdo, V. Saini, J. Piron, I. Al-Brahim, C. Boyer, J. Rioux, V. Bairi, A. S. Biris and T. Viswanathan, ACS Appl. Mater. Interfaces, 2012, 4, 363–368.
6. Z. Li, V. Saini, E. Dervishi, V. P. Kunets, J. Zhang, Y. Xu, A. R. Biris, G. J. Salamo and A. S. Biris, Appl. Phys. Lett., 2010, 96, 033110.
7. V. Švrček, S. Cook, S. Kazaoui and M. Kondo, J. Phys. Chem. Lett., 2011, 2, 1646–1650.
8. H. Zhou, H. E. Unalan, P. Hiralal, A. Colli, S. C. Tan, L. Wang, F. Kong and G. Amaratunga, 33rd IEEE Photovoltaic Specialists Conference, 11–16 May 2008, pp. 1–5,  DOI:10.1109/PVSC.2008.4922871.
9. D. A. Straus, M. Tzolov, T. F. Kuo, A. Yin and J. M. Xu, IET Circuits, Devices and Systems, 2007, 1, 200–204.
10. Z. Li, Y. Jia, J. Wei, K. Wang, Q. Shu, X. Gui, H. Zhu, A. Cao and D. Wu, J. Mater. Chem., 2010, 20, 7236–7240.
11. V. L. Borgne, L. A. Gautier, P. Castrucci, S. D. Gobbo, M. D. Crescenzi and M. A. E. Khakani, Nanotechnology, 2012, 23, 215206.
12. Y. Jia, P. Li, J. Wei, A. Cao, K. Wang, C. Li, D. Zhuang, H. Zhu and D. Wu, Mater. Res. Bull., 2010, 45, 1401–1405.
13. J. Chen, A. M. Rao, S. Lyuksyutov, M. E. Itkis, M. A. Hamon, H. Hu, R. W. Cohn, P. C. Eklund, D. T. Colbert, R. E. Smalley and R. C. Haddon, J. Phys. Chem. B, 2001, 105, 2525–2528.
14. A. C. Dillon, K. M. Jones, T. A. Bekkedahl, C. H. Kiang, D. S. Bethune and M. J. Heben, Nature, 1997, 386, 377–379.
15. M. J. O'Connell, S. M. Bachilo, C. B. Huffman, V. C. Moore, M. S. Strano, E. H. Haroz, K. L. Rialon, P. J. Boul, W. H. Noon, C. Kittrell, J. Ma, R. H. Hauge, R. B. Weisman and R. E. Smalley, Science, 2002, 297, 593–596.
16. A. Hagen and T. Hertel, Nano Lett., 2003, 3, 383–388.
17. T. G. Pedersen, Phys. Rev. B: Condens. Matter Mater. Phys., 2003, 67, 073401.
18. T. W. Odom, J. L. Huang, P. Kim and C. M. Lieber, Nature, 1998, 391, 62–64.
19. M. S. Fuhrer, B. M. Kim, T. Durkop and T. Brintlinger, Nano Lett., 2002, 2, 755–759.
20. M. Freitag, V. Perebeinos, J. Chen, A. Stein, J. C. Tsang, J. A. Misewich, R. Martel and P. Avouris, Nano Lett., 2004, 4, 1063–1066.
21. E. Kymakis and G. A. J. Amaratunga, J. Appl. Phys., 2006, 99, 084302.
22. V. Saini, Z. Li, S. Bourdo, E. Dervishi, Y. Xu, X. Ma, V. P. Kunets, G. J. Salamo, T. Viswanathan, A. R. Biris, D. Saini and A. S. Biris, J. Phys. Chem. C, 2009, 113, 8023–8029.
23. O. A. Abdulrazzaq, V. Saini, S. Bourdo, E. Dervishi, A. R. Biris and A. S. Biris, Part. Sci. Technol., 2013, 31, 427–442.
24. E. Shi, L. Zhang, Z. Li, P. Li, Y. Shang, Y. Jia, J. Wei, K. Wang, H. Zhu, D. Wu, S. Zhang and A. Cao, Sci. Rep., 2012, 2, 884.
25. X. Li, Y. Jung, K. Sakimoto, T. H. Goh, M. A. Reed and A. D. Taylor, Energy Environ. Sci., 2013, 6, 879–887.
26. Y. Jia, A. Cao, X. Bai, Z. Li, L. Zhang, N. Guo, J. Wei, K. Wang, H. Zhu, D. Wu and P. M. Ajayan, Nano Lett., 2011, 11, 1901–1905.
27. Y. Jung, X. Li, N. K. Rajan, A. D. Taylor and M. A. Reed, Nano Lett., 2013, 13, 95–99.
28. P. L. Ong, W. B. Euler and I. A. Levitsky, Nanotechnology, 2010, 21, 105203.
29. P. Castrucci, C. Scilletta, S. D. Gobbo, M. Scarselli, L. Camilli, M. Simeoni, B. Delley, A. Continenza and M. D. Crescenzi, Nanotechnology, 2011, 22, 115701.
30. Y. Jia, J. Wei, K. Wang, A. Cao, Q. Shu, X. Gui, Y. Zhu, D. Zhuang, G. Zhang, B. Ma, L. Wang, W. Liu, Z. Wang, J. Luo and D. Wu, Adv. Mater., 2008, 20, 4594–4598.
31. Y. Jia, P. Li, X. Gui, J. Wei, K. Wang, H. Zhu, D. Wu, L. Zhang, A. Cao and Y. Xu, Appl. Phys. Lett., 2011, 98, 133115.
32. Y. Jia, A. Cao, F. Kang, P. Li, X. Gui, L. Zhang, E. Shi, J. Wei, K. Wang, H. Zhu and D. Wu, Phys. Chem. Chem. Phys., 2012, 14, 8391–8396.
33. P. Wadhwa, B. Liu, M. A. McCarthy, Z. Wu and A. G. Rinzler, Nano Lett., 2010, 10, 5001–5005.
34. P. Wadhwa, G. Seol, M. K. Petterson, J. Guo and A. G. Rinzler, Nano Lett., 2011, 11, 2419–2423.
35. D. D. Tune, B. S. Flavel, R. Krupke and J. G. Shapter, Adv. Energy Mater., 2012, 2, 1043–1055.
36. S. M. Bachilo, M. S. Strano, C. Kittrell, R. H. Hauge, R. E. Smalley and R. B. Weisman, Science, 2002, 298, 2361–2366.
37. Y. Xu, E. Dervishi, A. R. Biris and A. S. Biris, Mater. Lett., 2011, 65, 1878–1881.
38. E. Dervishi, Z. Li, F. Watanabe, A. Courte, A. Biswas, A. R. Biris, V. Saini, Y. Xu and A. S. Biris, Chem. Mater., 2009, 21, 5491–5498.
39. E. Dervishi, Z. Li, Y. Xu, V. Saini, F. Watanabe, A. R. Biris, A. Bonpain, J. J. Garbay, A. Meriet, M. Richard and A. S. Biris, Particul. Sci. Technol., 2009, 27, 222–237.
40. A. R. T. Nugraha, R. Saito, K. Sato, P. T. Araujo, A. Jorio and M. S. Dresselhaus, Appl. Phys. Lett., 2010, 97, 091905.
41. G. Lolli, L. Zhang, L. Balzano, N. Sakulchaicharoen, Y. Tan and D. E. Resasco, J. Phys. Chem. B, 2006, 110, 2108–2115.
42. D. D. Tune, F. Hennrich, S. Dehm, M. F. G. Klein, K. Glaser, A. Colsmann, J. G. Shapter, U. Lemmer, M. M. Kappes, R. Krupke and B. S. Flavel, Adv. Energy Mater., 2013, 3, 1091–1097.
43. S. D. Gobbo, P. Castrucci, S. Dedele, L. Riele, A. Convertino, M. Morbidoni, F. D. Nicola, M. Scarselli, L. Camilli and M. D. Crescenzi, J. Mater. Chem. C, 2013, 1, 6752–6758.
44. M. S. Arnold, J. D. Zimmerman, C. K. Renshaw, X. Xu, R. R. Lunt, C. M. Austin and S. R. Forrest, Nano Lett., 2009, 9, 3354–3358.
45. L. Cognet, D. A. Tsyboulski, J. D. R. Rocha, C. D. Doyle, J. M. Tour and R. B. Weisman, Science, 2007, 316, 1465–1468.
46. J. Xie, T. Inaba, R. Sugiyama and Y. Homma, Phys. Rev. B: Condens. Matter Mater. Phys., 2012, 85, 085434.
47. B. E. Kareh, Silicon Devices and Process Integration: Deep Submicron and Nano-Scale Technologies, Springer, New York, 2009.
48. M. S. Hammer, D. Rauh, V. Lorrmann, C. Deibel and V. Dyakonov, Nanotechnology, 2008, 19, 485701.
49. B. G. Streetman, Solid State Electronic Devices, Prentice-Hall, New Jersey, 6th edn, 2005.
50. B. Kaplan, Phys. B, 2004, 351, 90–95.
51. K. Vanheusden, C. H. Seager, W. L. Warren, D. R. Tallant and J. A. Voigt, Appl. Phys. Lett., 1996, 68, 403–405.

---

Footnotes

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c4ra10735g

‡ Present address: Center for Integrated Nanotechnologies, Materials Physics and Application Division, Los Alamos National Laboratory, Los Alamos, NM, USA.

---