Dopant-Free and Carrier-Selective Heterocontacts for Silicon Solar Cells: Recent Advances and Perspectives

Abstract

By combining the most successful heterojunctions (HJ) with interdigitated back contacts, crystalline silicon (c-Si) solar cells (SCs) have recently demonstrated a record efficiency of 26.6%. However, such SCs still introduce optical/electrical losses and technological issues due to parasitic absorption/Auger recombination inherent to the doped films and the complex process of integrating discrete p⁺- and n⁺-HJ contacts. These issues have motivated the search for alternative new functional materials and simplified deposition technologies, whereby carrier-selective contacts (CSCs) can be formed directly with c-Si substrates, and thereafter form IBC cells, via a dopant-free method. Screening and modifying CSC materials in a wider context is beneficial for building dopant-free HJ contacts with better performance, shedding new light on the relatively mature Si photovoltaic field. In this review, a significant number of achievements in two representative dopant-free hole-selective CSCs, i.e., poly(3,4-ethylene dioxythiophene):poly(styrenesulfonate)/Si and transition metal oxides/Si, have been systemically presented and surveyed. The focus herein is on the latest advances in hole-selective materials modification, interfacial passivation, contact resistivity, light-trapping structure and device architecture design, etc. By analyzing the structure-property relationships of hole-selective materials and assessing their electrical transport properties, promising functional materials as well as important design concepts for such CSCs toward high-performance SCs have been highlighted.

Keywords: carrier‐selective contacts; dopant‐free; heterojunction solar cells; poly(3,4‐ethylene dioxythiophene):poly(styrenesulfonate) (PEDOT:PSS); transition metal oxides.

Figure 1

a) Schematic diagram of PEDOT:PSS/n‐Si HSCs; b) chemical structural formulae of the PEDOT and PSS; c) junction formation at the hybrid PEDOT:PSS/n‐Si interface. Reproduced with permission.4 Copyright 2015, Nature Publishing Group.

Figure 2

a) Conductivity and film thickness of PEDOT:PSS treated with various concentrations of PTSA and the number of the treatment cycle; b) J–V curves of PEDOT:PSS/n‐Si HSCs without and with PTSA/DMSO and PTSA/DMSO with ARC. Reproduced with permission.33

Figure 3

a) XPS spectra of the c‐Si surfaces and b) J–V curves of the PEDOT:PSS/Si HSCs treated with different oxidizing agents; c) XPS spectra in the Si 2p region collected from bare Si, HF treatment and TMAH treatment; d) J–V curves of the PEDOT:PSS/Si HSCs with HF treatment, TMAH treatment, and TMAH treatment & a capping layer of high work function CuI. Reproduced with permission.35 Copyright 2010, American Chemical Society.

Figure 4

Schematic models of the PEDOT:PSS layer formation on SiO_x/Si for a) EG7‐PEDOT:PSS and b) pristine PEDOT:PSS, added with 0.25 wt% surfactant (FS or TX); J–V curves of PEDOT:PSS/planar‐Si HSCs with 7 wt% EG and 0.10, 0.25, and 0.50 wt% c) TX and d) FS. Reproduced with permission.42

Figure 5

a) Device configuration of the PEDOT:PSS/Si HSC with a CuI or MoO₃ capping layer; b) schematic diagram of photogenerated carrier separation and transportation mechanism with a WO₃ interlayer between the Ag electrode and the PEDOT:PSS layer. Reproduced with permission.47 Copyright 2015, Elsevier.

Figure 6

a) J–V curves of the 20 µm PEDOT:PSS/c‐Si HSCs with a MACE‐reconstructed Si NW structure. The inset shows the structure configuration; Reproduced with permission.66 Copyright 2015, American Chemical Society. b) structure configuration of 20 µm PEDOT:PSS/n‐Si HSCs with the NC–NP dual‐structure as a light‐trapping scheme. Reproduced with permission.52

Figure 7

a) J–V characteristic curves of the core–shell hybrid SCs; Reproduced with permission.67 Copyright 2011, American Chemical Society. b) the molecular structures of TAPC, the device structure and cross‐sectional SEM image of Si‐NW/TAPC hybrid SC; ‘Reproduced with permission.68 Copyright 2013, American Chemical Society. c) J–V characteristic curves of the hybrid SCs with ALD‐processed aluminum oxide as passivation layer; Reproduced with permission.70 Copyright 2013, American Chemical Society. d) the structure configuration of PEDOT:PSS/structured‐Si hybrid SC with liquid‐phase processed TiO₂ as an interface passivation layer, the energy band diagram, the SEM image of the Si/PEDOT:PSS interface, and the J–V curve. Reproduced with permission.41 Copyright 2016, American Chemical Society.

Figure 8

a) Cross‐sectional scanning electron images of the as‐fabricated PEDOT:PSS/n‐Si junction on textured‐Si substrate without DEP coating and b) with DEP coating (the DEP layer was removed with an acetone solution to capture clear images). Scale bars, 500 nm in (a) and (b) and 250 nm for the insets, respectively. c) Light J–V characterization of the HSCs with a planar n‐Si substrate and the pyramidal n‐Si substrate with and without a DEP coating. Reproduced with permission.24

Figure 9

Efficiency evolution of PEDOT:PSS/n‐Si HSCs over time. FrontPEDOT:PSS means the PEDOT:PSS is placed upon the top‐surface of the SC where the active light shines, whereas the PEDOT:PSS in the BackPEDOT:PSS structure is only used as a back‐surface field (the main junction in front is still a high‐temperature diffused Si‐emitter). Reproduced with permission.24

Figure 10

a) Simulated PCE versus front/rear recombination velocities of the n‐Si interface (S _front/S _rear) for the PEDOT:PSS/n‐Si HSCs under N _d = 2 × 10¹⁵ cm⁻³; b) V _oc (blue line) and PCE (red line) as a function of N _d; c) FF and PCE versus R _c for the PEDOT:PSS/n‐Si HSCs. Reproduced with permission.60 Copyright 2017, American Chemical Society.

Figure 11

a) Normalized photovoltaic parameters and b) efficiency degradation of the HSCs with and without the DEP coating; c) stability in the conductivity of the PEDOT:PSS films with and without the DEP coating. ‘Ref' represents cells without the DEP coating and ‘Coating' represents cells with the DEP protection; d) UPS was used to measure the WF of the as‐prepared PEDOT:PSS and the stored PEDOT:PSS with and without DEP coating. The shaded area in (d) represents the bandgap of n‐Si. Reproduced with permission.24

Figure 12

a) XPS spectra in the Si 2p region for the interfacial Si oxide tunneling layers of the as‐prepared PEDOT:PSS/n‐Si and long‐term stored samples with and without the DEP coating; b) AFM morphology and phase images of the fresh as‐prepared PEDOT:PSS film and the film stored over a long term without the DEP coating. Scale bars, 500 nm; c) C 1s and d) O 1s XPS spectra of the as‐prepared PEDOT:PSS film and the film without the DEP coating stored over a long term; e) schematic illustration of the PEDOT:PSS film on the performance degradation after long‐term storage. The yellow parts represent the PEDOT and the long chain of the PSS. Reproduced with permission.24

Figure 13

a) Energy‐level diagram for MoO_x on an n‐Si substrate; b) C–V curve of a device with the Ag/MoO_x/n‐Si/Al structure; Reproduced with permission.90 energy band diagrams for n‐Si HSC with a hole‐selective c) MoO_x contact and d) a standard p‐type a‐Si:H emitter. Reproduced with permission.91 Copyright 2014, AIP Publishing LLC.

Figure 14

a) WF, b) electronegativity contribution, and c) donor state contribution of MoO_x versus oxygen deficiency/x. The circles in (a) are the experimental data points, and the dashed curve is the predicted trend from Equation (2). Reproduced with permission.92

Figure 15

a) Annealing temperature dependence of the O/Mo ratio and conductivity of MoO_x film in N₂; b) transmittance spectra of MoO₃ films annealed at various temperatures in N₂; c) variation of the optical band gap of MoO₃ films with annealing temperatures in N₂ and O₂. Reproduced with permission.93 Copyright 2008, Elsevier.

Figure 16

HR‐TEM images of the ITO/TMOs/n‐Si heterostructures showing an interlayer between n‐Si and a) MoO₃, b) WO₃, and c) V₂O₅. Reproduced with permission.3 Copyright 2016, Materials Research Society.

Figure 17

ToF‐SIMS depth profile for the ITO/TMOs/n‐Si heterostructures showing an SiO₂ ⁻ signal at the interface between n‐Si and a,b) MoO₃, c) WO₃, and d) V₂O₅. For each TMO, the related reduced species (MoO⁻, WO₂ ⁻, VO₂ ⁻) are also detected near the interface. Reproduced with permission.3 Copyright 2016, Materials Research Society.

Figure 18

UPS spectra of the incremental MoO_x deposition thickness on an Si substrate. a) A close‐up of the valence band region with respect to the Fermi level; b) secondary electron cutoff region, the vacuum level difference between pristine Si and the thick layer of MoO_x is noted; Reproduced with permission.90 c) relative content of Mo⁺⁶ and Mo⁺⁵ oxidation states for different MoO₃ film thicknesses and postdeposition treatments. Reproduced with permission.96 Copyright 2015, MDPI.

Figure 19

a) The ρ_c values of Au/TMO/n‐Si systems extracted from the C–V responses in a TLM array; b) effective carrier lifetime of TMO/n‐Si/TMO configurations as a function of excess carrier density, where the implied‐V _oc values of three samples were also marked in the figure. The thicknesses of the three TMOs are 20 nm. Reproduced with permission.3 Copyright 2016, AIP Publishing LLC.

Figure 20

Minority carrier lifetime measurements before and after MoO_x thermal evaporation and after IOH sputtering. Reproduced with permission.91 Copyright 2014, AIP Publishing LLC.

Figure 21

a) Micrometer (LHS)‐ and 100 nm (RHS)‐scale cross‐sectional scanning electron microscopy images of the textured front (sunward side) and back surfaces of the DASH cell. The 100 nm scale image colors are inverted to highlight the different films on each surface; b) light J–V behavior and cell characteristics of the DASH cell measured under standard 1‐sun conditions; c) EQE (black) and IQE (purple) alongside the measured reflectance (blue) for the DASH cells. The J _sc obtained from the EQE, shown above a photograph of the DASH cell, agrees well with that measured from the light J–V analysis. Reproduced with permission.17 Copyright 2016, Nature Publishing Group.

Figure 22

Top) Schematic illustrating the MLBC HSC patterning fabrication process; below 1) images of two metal patterning masks; below 2) the structure of a completed MLBC HSC; below 3) SEM image of the results of metal mask patterning. Reproduced with permission.82 Copyright 2017, The Royal Society of Chemistry.

Figure 23

a) Comparative behaviors (time‐dependence of iV _oc under air exposure) of the V₂O_x/Si/V₂O_x samples (40 nm thick V₂O_x) with and without nickel capping; Reproduced with permission.98 Copyright 2017, Royal Society of Chemical. b) dependence of the selectivity of MoO_x/n‐Si heterojunction on the induced c‐Si band bending for different annealing temperatures. Here, V _bb is the equilibrium band bending, and ΔV = iV _oc − V _oc. Reproduced with permission.103 Copyright 2015, Elsevier.