Nature of Record Efficiency Fluid-Processed Nanotube–Silicon Heterojunctions

Abstract

SWCNT Purification and Sample Preparation: Nanotube fractions were prepared using a number of approaches to yield equivalent results. Initially, type purification was achieved using density-gradient ultracentrifugation (DGU) 1 and (6,5) enrichment was achieved using nonlinear DGU. 2 These techniques are well documented and details can be found elsewhere. [1][2][3] Length-enriched CoMoCat fractions were prepared by dispersing the SWCNTs (Southwest nanotechnologies lot SG65EX or SG65i lot 046) in water via ultrasonication with 2 % sodium deoxycholate (DOC). The suspensions were coarsely centrifuged to remove impurities and bundles and then purified by length using transient DGU. 4,5 For AFM characterization of the fractions, silicon wafers were functionalized with 3-(ethoxydimethylsilyl)-propylamine (APDMES, Sigma-Aldrich) by 15 min UVO followed by immersion in 1 % APDMES/isopropanol for 20 min and subsequent rinsing with isopropanol and DI water. The substrates were dried at 70 o C for 20 min. SWCNT samples in 10 g/L DOC were diluted 100× with an aqueous solution of 0.2 wt. % sodium cholate and 20 mM sodium thiocyanate and deposited on a small piece of silicon wafer

Advances in Carbon Nanotube–Silicon Heterojunction Solar Cells

“…Before our 2012 report on the early progress in this field, and the recent efficiency and active area records, the first report of carbon nanotube–silicon heterojunctions being used specifically for solar cells was by Wei et al in 2007 . The overall power conversion efficiency (PCE) of Wei et al's device was around 1.3% but, as can be seen in Figure 1 and in Table S1 of the Supporting Information, very rapid efficiency gains have been made since then, as might be expected for an architecture based on the already mature field of silicon photovoltaics (PV). However, there is still significant room for improvement in the maximum efficiency of proof‐of‐principle laboratory cells.…”

“…Despite the influence of other factors, such as the use of antireflection (AR) strategies in the higher performing devices, there is still a clear dependence of the power conversion efficiency on the active area. Second, it is well known that the photoresponse of the junction is markedly different near the electrodes due to various edge effects such as light biasing of the emitter by the illuminated region outside of the active area . This can have the effect of artificially enhancing the calculated performance for small active areas (less than ≈1 cm 2 ).…”

Improving Carbon Nanotube‐Silicon Solar Cells by Solution Processable Metal Chlorides

“…For bare CNT‐Si structures, the stability is possibly influenced by the growth of interfacial oxide under the porous CNT network, resulting in continuous degradation of cell efficiency over air storage (Table S4 and Figure S3, Supporting Information ). The result of solar cells with metal chlorides optimization indicates that the adsorption of metal cations on the surface of CNTs is very stable, which overcomes the volatility or unstability in previously used doping agents, such as SOCl 2 , AuCl3, and H 2 O 2 . Moreover, two or more metal chlorides are easily mixed to utilize their respective advantages given their processability in common solvents.…”

Direct‐Patterning SWCNTs Using Dip Pen Nanolithography for SWCNT/Silicon Solar Cells

“…Replacing the p‐type Si with CNT films is advantageous for a variety of reasons. CNT films are cheaper, and quicker to manufacture, have much greater flexibility, and have excellent charge carrier mobility (exceeding 10 5 cm 2 V −1 s −1 for individual CNTs at room temperature) . CNT films also still retain high solar spectrum transmittance, good stability under heat and light exposure, low light reflectance, low resistivity (electrical conductivity 10 4 –10 6 S cm −1 ), and tuneable light absorption properties .…”

“…The first CNT/Si cell reported was by Wei et al in 2007 and since then, different methods of manufacturing and applying the CNT film, and different compositions and treatments of the CNT film have been investigated. Improvements to the design of CNT/Si cells have increased their power conversion efficiency (PCE) including doping of the CNT film with an electron withdrawing/p‐doping chemical/nanoparticle/nanowire, using single‐walled and metallic or (6,5) semiconducting CNTs, creating a 1–2 nm Si oxide layer or a thin conductive polymer layer between the n‐type Si and the CNT film, using an antireflective layer, using a thinner n‐type Si layer, using an aligned CNT film, and optimizing the structure of the p‐type CNT film and front contact electrodes . Cells with PCE of up to 17% have been reported, which is becoming comparable to commercially available p–n silicon solar cells (20–30% PCE).…”