William Nemeth

Interdigitated Back Passivated Contact (IBPC) Solar Cells Formed by Ion Implantation

We describe work toward an interdigitated back passivated contact (IBPC) solar cell formed by patterned ionimplanted passivated contacts. Formation of electron and hole passivated contacts to n-type Cz wafers using a thin SiO₂ layer and ion-implanted amorphous silicon (a-Si) is described. P and B were ion implanted into intrinsic a-Si films, forming symmetric and IBPC test structures. The recombination parameter J₀, as measured by a Sinton lifetime tester after thermal annealing, was J₀ ~ 2.4 fA/cm² for Si:P and J₀ ~ 10 fA/cm² for Si:B contacts. The contact resistivity for the passivated contacts was found to be 0.46 Ω·cm² for the n-type contact and 0.04 Ω·cm² for the p-type contact. The IBPC solar cell test structure gave 1-sun V_oc values of 682 mV and pFF = 80%. The benefits of the ion-implanted IBPC cell structure are discussed.

Incorporation of a light and carrier collection management nano-element array into superstrate a-Si:H solar cells

Superstrate a-Si:H solar cells incorporating a nano-column array for light and photocarrier collection have been fabricated and evaluated. It is found that the short circuit current density (JSC) is significantly increased while the open circuit voltage and fill factor are not detrimentally affected by this architecture. Numerical analysis of JSC matches experiment and shows that the enhanced JSC observed is due to both effective absorber thickness and photonic-plasmonic effects. Further analysis shows that this nano-column architecture can lead to a 42% increase in conversion efficiency over that of the planar control for a 200 nm absorber thickness cell.

Carrier-selective, passivated contacts for high efficiency silicon solar cells based on transparent conducting oxides

We describe the design, fabrication and results of passivated contacts to n-type silicon utilizing thin SiO 2 and indium tin oxide. High-temperature silicon dioxide is grown on both surfaces on an n-type Si wafer to a thickness 0,contact , and a non-ohmic, high contact resistance. However, after a forming gas anneal, the passivation quality and the contact resistivity improve significantly. The contacts are characterized by measuring the recombination parameter current density of the contact (J 0,contact ) and the specific contact resistivity (ρ contact ) using a transmission line method (TLM) pattern. The best ITO/SiO 2 passivated contact in this study has J 0,contact = 93.5 fA/cm2 and ρ contact = 11.5 mOhm-cm2. These values are placed in context with other passivating contacts using an analysis that determines the ultimate efficiency and the optimal area fraction for contacts for a given set of (J 0,contact , ρ contact ) values. The ITO/SiO 2 contacts are found to have a higher J 0,contact , but a similar ρ contact compared to the best reported passivated contacts.

Hydrogenated amorphous si deposition for high efficiency a-Si/c-Si heterojunction solar cells

We study the differences in hydrogenated amorphous Si (a-Si:H) depositions between Hot-Wire Chemical Vapor Deposition (HWCVD) and Plasma Enhanced Chemical Vapor Deposition (PECVD) for high efficiency a-Si/c-Si heterojunction (HJ) solar cells. In HWCVD, process gases such as silane decompose from the high-temperature hot filament. The resulting deposition is thought to be gentle due to the lack of ion bombardment that may cause damage to c-Si surface. In PECVD, process gases decompose from a high frequency electric field and ion bombardment is expected during the a-Si:H deposition. We found that the initial minority carrier lifetime of a-Si:H passivated high-quality n-type wafer was higher (about a ms) with the HWCVD process, and the final minority carrier lifetime (after 250°C annealing) was higher (over a few ms) with the PECVD process. These findings suggest that the damage from the ion bombarding in PECVD is not as detrimental as we expected; or if there is damage, it can be repaired by the annealing. We also speculate that the lack of further increase of the lifetime after annealing with HWCVD intrinsic a-Si:H layer can be related to the direct substrate heating from the hot filament during the deposition. A high substrate temperature will promote epi-Si growth and drive hydrogen out of the a-Si/c-Si interface to decrease the quality of surface passivation. To reduce the heating effect, a shutter and a low filament temperature are preferred. With the optimized process, we were able to fabricate HJ solar cells with high open circuit voltage of 714 mV and efficiency greater than 19% on an un-textured n-type wafer using the PECVD process, and independently confirm best efficiency of 19.7% on textured n-type wafer with the HWCVD process.

Implementation of tunneling pasivated contacts into industrially relevant n-Cz Si solar cells

We identify bottlenecks, and propose solutions, to implement a B-diffused front emitter and a backside pc-Si/SiO2 pasivated tunneling contact into high efficiency n-Cz Si cells in an industrially relevant way. We apply an O-precipitate dissolution treatment to make n-Cz wafers immune to bulk lifetime process degradation, enabling robust, passivated B front emitters with J0 <; 20fA/cm2. Adding ultralow recombination n+ pc-Si/SiO2 back contacts enables pre-metallized cells with iVoc=720 mV and J0=8.6 fA/cm2. However, metallization significantly degrades performance of these contacts due to pinholes and possibly, grain boundary diffusion of primary metal and source contaminates such as Cu. An intermediate, doped a-Si:H capping layer is found to significantly block the harmful metal penetration into pc-Si.

From amorphous to nanocrystalline: the effect of nanograins in amorphous matrix on the thermal conductivity of hot-wire chemical-vapor deposited silicon films

We have measured the thermal conductivity of amorphous and nanocrystalline silicon films with varying crystalline content from 85 K to room temperature. The films were prepared by the hot-wire chemical-vapor deposition, where the crystalline volume fraction is determined by the hydrogen (H2) dilution ratio to the processing silane gas (SiH4), R  =  H2/SiH4. We varied R from 1 to 10, where the films transform from amorphous for R  <  3 to mostly nanocrystalline for larger R. Structural analyses show that the nanograins, averaging from 2 to 9 nm in sizes with increasing R, are dispersed in the amorphous matrix. The crystalline volume fraction increases from 0 to 65% as R increases from 1 to 10. The thermal conductivities of the two amorphous silicon films are similar and consistent with the most previous reports with thicknesses no larger than a few μm deposited by a variety of techniques. The thermal conductivities of the three nanocrystalline silicon films are also similar, but are about 50-70% higher than those of their amorphous counterparts. The heat conduction in nanocrystalline silicon films can be understood as the combined contribution in both amorphous and nanocrystalline phases, where increased conduction through improved nanocrystalline percolation path outweighs increased interface scattering between silicon nanocrystals and the amorphous matrix.

Utilization of Tabula Rasa to stabilize bulk lifetimes in n-Cz silicon for high-performance solar cell processing

We investigate a high temperature, high cooling-rate anneal Tabula Rasa (TR) and report its implications on n-type Czochralski-grown silicon (n-Cz Si) for photovoltaic fabrication. Tabula Rasa aims at dissolving and homogenizing oxygen precipitate nuclei that can grow during the cell process steps and degrade the cell performance due to their high internal gettering and recombination activity. The Tabula Rasa thermal treatment is performed in a clean tube furnace with cooling rates >100°C/s. We characterize the bulk lifetime by Sinton lifetime and photoluminescence mapping just after Tabula Rasa, and after the subsequent cell processing. After TR, the bulk lifetime surprisingly degrades to <; 0.1ms, only to recover to values equal or higher than the initial non-treated wafer (several ms), after typical high temperature cell process steps. Those include boron diffusion and oxidation; phosphorus diffusion/oxidation; ambient annealing at 850°C; and crystallization annealing of tunneling-passivating contacts (doped polycrystalline silicon on 1.5 nm thermal oxide). The drastic lifetime improvement during high temperature cell processing is attributed to improved external gettering of metal impurities and annealing of intrinsic point defects. Time and injection dependent lifetime spectroscopy further reveals the mechanisms of lifetime improvement after Tabula Rasa treatment. Additionally, we report the efficacy of Tabula Rasa on n-type Cz-Si wafers and its dependence on oxygen concentration, correlated to position within the ingot.

Diffractive light trapping in crystal-silicon films: experiment and electromagnetic modeling.

Diffractive light trapping in 1.5 μm thick crystal silicon films is studied experimentally through hemispherical reflection measurements and theoretically through rigorous coupled-wave analysis modeling. The gratings were fabricated by nanoimprinting of dielectric precursor films. The model data, which match the experimental results well without the use of any fitting parameters, are used to extract the light trapping efficiency. Diffractive light trapping is studied as a function of incidence angle, and an enhancement of light absorption is found for incidence angles up to 50° for both TE and TM polarizations.

Hydrogen passivation of poly-Si/SiOx contacts for Si solar cells using Al2O3 studied with deuterium

The interplay between hydrogenation and passivation of poly-Si/SiOx contacts to n-type Si wafers is studied using atomic layer deposited Al2O3 and anneals in forming gas and nitrogen. The poly-Si/SiOx stacks are prepared by thermal oxidation followed by thermal crystallization of a-Si:H films deposited by plasma-enhanced chemical vapor deposition. Implied open-circuit voltages as high as 710 mV are achieved for p-type poly-Si/SiOx contacts to n-type Si after hydrogenation. Correlating minority carrier lifetime data and secondary ion mass spectrometry profiles reveals that the main benefit of Al2O3 is derived from its role as a hydrogen source for chemically passivating defects at SiOx; Al2O3 layers are found to hydrogenate poly-Si/SiOx much better than a forming gas anneal. By labelling Al2O3 and the subsequent anneal with different hydrogen isotopes, it is found that Al2O3 exchanges most of its hydrogen with the ambient upon annealing at 400 °C for 1 h even though there is no significant net change in its total hydrogen content.The interplay between hydrogenation and passivation of poly-Si/SiOx contacts to n-type Si wafers is studied using atomic layer deposited Al2O3 and anneals in forming gas and nitrogen. The poly-Si/SiOx stacks are prepared by thermal oxidation followed by thermal crystallization of a-Si:H films deposited by plasma-enhanced chemical vapor deposition. Implied open-circuit voltages as high as 710 mV are achieved for p-type poly-Si/SiOx contacts to n-type Si after hydrogenation. Correlating minority carrier lifetime data and secondary ion mass spectrometry profiles reveals that the main benefit of Al2O3 is derived from its role as a hydrogen source for chemically passivating defects at SiOx; Al2O3 layers are found to hydrogenate poly-Si/SiOx much better than a forming gas anneal. By labelling Al2O3 and the subsequent anneal with different hydrogen isotopes, it is found that Al2O3 exchanges most of its hydrogen with the ambient upon annealing at 400 °C for 1 h even though there is no significant net change in it...

Plasma immersion ion implantation for interdigitated back passivated contact (IBPC) solar cells

We present progress to develop low-cost interdigitated back contact solar cells with pc-Si/SiO<inf>2</inf>/c-Si passivated contacts formed by plasma immersion ion implantation (PIII). PIII is a lower-cost implantation technique than traditional beam-line implantation due to its simpler design, lower operating costs, and ability to run high doses (1E14–1E18 cm⁻²) at low ion energies (20 eV–10 keV). These benefits make PIII ideal for high throughput production of patterned passivated contacts, where high-dose, low-energy implantations are made into thin (20–200 nm) a-Si layers instead of into the wafer itself. For this work symmetric passivated contact test structures grown on n-Cz wafers with PH<inf>3</inf> PIII doping gave implied open circuit voltage (iV<inf>oc</inf>) values of 730 mV with J<inf>o</inf> values of 2 fA/cm². Samples doped with B<inf>2</inf>H<inf>6</inf> gave iV<inf>oc</inf> values of 690 mV and J<inf>o</inf> values of 24 fA/cm², outperforming BF<inf>3</inf> doping, which gave iV<inf>oc</inf> values in the 660–680 mV range. Samples were further characterized by photoluminescence and SIMS depth profiles. Initial IBPC cell results are presented.