Engineering topics

Iles, P. A.

Publications and source records attributed to Iles, P. A..

At least 19 records

NASA NTRS2004

Performance Trade-Offs for MBG Using Different Bandgap Pairs

Present GAInP2/GAAs cell grown on Ge (or GaAs) substrates have demonstrates high efficiency, However, the individual bandgaps (1.85 eV, 1.42 eV respectively) are not the optimum match to convert the AM0 spectrum efficiently, although by reducing the thickness of the GaInP2 cell, good efficiencies have been achieved. Within the III-V alloys, modeling shows that several dual junction cells, comprising different bandgap pairs, could give higher efficiency. This paper outlines the modeling used, the bandgap pairs selected, and projections for the AM0 efficiency assuming that the effects of lattice mismatch associated with the bandgap can be reduced. We also present preliminary data on I-V performance for the selected top and bottom cells and for some full cascade cells. We have included some characterization results to check the bandgaps and lattice strain, to check how closely the fabricated cells conform to the model,

Iles, P. A.↗

NASA NTRS1993

Workshop summary: New silicon cells

The workshop on new silicon cells held during SPRAT12 is summarized. A smaller than average group attended this workshop reflecting the reduction in research dollars available to this portion of the photovoltaics community. Despite the maturity of the silicon technology, a core of the group maintained an excitement about new developments and potential opportunities. The group addressed both the implications and the applications of recent developments. Topics discussed include: light trapping and ultrathin silicon cells; different uses for silicon cells; new silicon cell developments; and radiation tolerant high efficiency cells.

Meulenberg, A.↗

NASA NTRS1991

Recent advances in GaAs/Ge solar cells

By growing the GaAs cell on a Ge substrate, the advantages of GaAs cells can be retained and the higher mechanical strength of the Ge makes larger, thinner GaAs cells possible. To conform to immediate user requirements, GaAs growth conditions were modified to eliminate the additional PV output at GaAs/Ge interface. To demonstrate acceptable cell manufacturing technology, the major areas in cell manufacture were analyzed and developed, and efficiency combined. Also the cells were successfully assembled on current lightweight arrays. The main areas of effort are discussed.

Chu, C. L.↗

NASA NTRS1991

Radiation resistance of Ge, Ge0.93Si0.07, GaAs and Al0.08Ga0.92 as solar cells

Solar cells made of Ge, Ge(0.93)Si(0.07) alloys, GaAs and Al(0.08)Ga(0.92)As were irradiated in two experiments with 1-meV electrons at fluences as great as 1 x 10(exp 16) cm(exp-2). Several general trends have emerged. Low-band-gap Ge and Ge(0.93)Si(0.07) cells show substantial resistance to radiation-induced damage. The two experiments showed that degradation is less for Al(0.08)Ga(0.92)As cells than for similarly irradiated GaAs cells. Compared to homojunctions, cells with graded-band-gap emitters did not show the additional resistance to damage in the second experiment that had been seen in the first. The thickness of the emitter is a key parameter to limit the degradation in GaAs devices.

Timmons, M. L.↗

NASA NTRS1991

Status of GaAs/Ge solar cells

With experience at increasing production levels, GaAs/Ge cells are proving their effectiveness for some demanding missions. The experience in producing inactive-Ge structures should benefit parallel work on forming monolithic cascade cells using AlGaAs or GaAs top cell layers grown on Ge or other bottom cell materials. The adjustments needed to ensure that the GaAs/Ge interface is inactive are more likely to preserve the properties of a separately optimized bottom cell.

Iles, P. A.↗

NASA NTRS1991

Graded-bandgap AlGaAs solar cells for AlGaAs/Ge cascade cells

Some p/n graded-bandgap Al(x)Ga(1-x)As solar cells were fabricated and show AMO conversion efficiencies in excess of 15 percent without antireflection (AR) coatings. The emitters of these cells are graded between 0.008 is less than or equal to x is less than or equal to 0.02 during growth of 0.25 to 0.30 micron thick layers. The keys to achieving this performance were careful selection of organometallic sources and scrubbing oxygen and water vapor from the AsH3 source. Source selection and growth were optimized using time-resolved photoluminescence. Preliminary radiation-resistance measurements show AlGaAs cells degraded less than GaAs cells at high 1 MeV electron fluences, and AlGaAs cells grown on GaAs and Ge substrates degrade comparably.

Timmons, M. L.↗

NASA NTRS1987

Heterostructure solar cells

The performance of gallium arsenide solar cells grown on Ge substrates is discussed. In some cases the substrate was thinned to reduce overall cell weight with good ruggedness. The conversion efficiency of 2 by 2 cm cells under AMO reached 17.1 percent with a cell thickness of 6 mils. The work described forms the basis for future cascade cell structures, where similar interconnecting problems between the top cell and the bottom cell must be solved. Applications of the GaAs/Ge solar cell in space and the expected payoffs are discussed.

Chang, K. I.↗

NASA NTRS1985

Heavily doped polysilicon-contact solar cells

The first use of a (silicon)/heavily doped polysilicon)/(metal) structure to replace the conventional high-low junction or back-surface-field (BSF) structure of silicon solar cells is reported. Compared with BSF and back-ohmic-contact (BOC) control samples, the polysilicon-back solar cells show improvements in red spectral response (RSR) and open-circuit voltage. Measurement reveals that a decrease in effective surface recombination velocity S is responsible for this improvement. Decreased S results for n-type (Si:As) polysilicon, consistent with past findings for bipolar transistors, and for p-type (Si:B) polysilicon, reported here for the first time. Though the present polysilicon-back solar cells are far from optimal, the results suggest a new class of designs for high efficiency silicon solar cells. Detailed technical reasons are advanced to support this view.

Lindholm, F. A.↗

NASA NTRS1985

High efficiency solar cell processing

At the time of writing, cells made by several groups are approaching 19% efficiency. General aspects of the processing required for such cells are discussed. Most processing used for high efficiency cells is derived from space-cell or concentrator cell technology, and recent advances have been obtained from improved techniques rather than from better understanding of the limiting mechanisms. Theory and modeling are fairly well developed, and adequate to guide further asymptotic increases in performance of near conventional cells. There are several competitive cell designs with promise of higher performance ( 20%) but for these designs further improvements are required. The available cell processing technology to fabricate high efficiency cells is examined.

Ho, F.↗

NASA NTRS1985

Experience with OMCVD production of GaAs solar cells

The projected promise of the OMCVD approach, i.e., to make high efficiency GaAs space cells, has been demonstrated. The properties and control of the deposited GaAs and AlGaAs layers and the uniformity of the post layer processing have been most satisfactory. In particular the control of the critical thin layers (p-GaAs, p-AlGaAs) has been impressive. Experience has also been gained in routine areas, connected with continuous operation at high capacity. There are still a few areas for improvement, to further increase capacity, and to anticipate and prevent mechanical equipment problems.

Yeh, Y. C. M.↗

NASA NTRS1985

Front and back polysilicon-contacted silicon solar cells

The first use of a (silicon)/ (heavily doped polysilicon)/ (metal) structure to replace the conventional high-low junction (or back-surface-field, BSF) structure of silicon solar cells is reported. Compared with BSF and back-ohmic-contact control samples, the polysilicon-back solar cells show improvements in red spectral response and open-circuit voltage. Measurement reveals that a decrease in effective surface recombination velocity S is responsible for this improvement. Decreased S results for n-type (Si:As) polysilicon, consistent with past findings for bipolar transistors, and for p-type (Si:B) polysilicon, reported here for the first time. Though the present polysilicon-back solar cells are far from optimal, the results suggest a new class of designs for high efficiency silicon solar cells. Detailed technical reasons are advanced to support this view.

Lindholm, F. A.↗

NASA NTRS1984

Microcrystalline silicon growth for heterojunction solar cells

Microcrystalline Si (m-Si) films with a 1.7eV energy bandgap and crystal size of several hundred A were e-beam evaporated on single crystalline Si (c-Si) to form a heterojunction with the substrate, or a window layer to a single crystalline p-n junction (heteroface structure). The goal was to enhance Voc by such uses of the larger bandgap m-Si, with the intriguing prospect of forming heterostructures with exact lattice match on each layer. The heterojunction structure was affected by interface and shunting problems and the best Voc achieved was only 482mV, well below that of single crystal Si homojunctions. The heteroface structure showed promise for some of the samples with p m-Si/p-n structure (the complementary structure did not show any improvement). Although several runs with different deposition conditions were run, the results were inconsistent. Any Voc enhancement obtained was too small to compensate for the current loss due to the extra absorption and poor carrier transport properties of the m-Si film.

Leung, D. C.↗

NASA NTRS1984

Microcrystalline silicon growth for heterojunction solar cells

A single source of evaporation with B mixed with highly doped Si is used instead of the coevaporation of separate Si and B sources to reduce possible carbon contamination. The results of both the heterojunction or heteroface structures, however, are similar when evaporation is used. The best Voc of the heterojunction is about 460mV and no improvement in Voc in the heteroface structure is observed. Slight Voc degradation occurred. A study of the p m-Si/p c-Si structure showed a negative Voc in many cases. The interface properties between the two materials are such that instead of repelling minority carriers from the substrate carrier, collection actually occurred. Another study of cells made in the part of substrates not covered by n-Si results in performance lower than the controls. This indicates possible substrate degradation in the process.

Iles, P. A.↗

NASA NTRS1984

Effect of localized polycrystalline silicon properties on solar cell performance

Several forms of polycrystalline silicon, mostly from cast ingots, (including UCP, SILSO and HEM) were studied. On typical slices, localized properties were studied in two ways. Small area (about 2.5 sq mm) mesa diodes were formed, and localized photovoltaic properties were measured. Also a small area (about .015 sq mm) light spot was scanned across the cells; the light spot response was calibrated to measure local diffusion length directly. Using these methods, the effects of grain boundaries, or of intragrain imperfections were correlated with cell performance. Except for the fine grain portion of SILSO, grain boundaries played only a secondary role in determining cell performance. The major factor was intra-grain material quality and it varied with position in ingots and probably related to solidification procedure.

Leung, D.↗

NASA NTRS1983

Metallization problems with concentrator cells

Cells used with concentrators have similar contact requirements to other cells, but operation at high intensity imposes more than the usual demands on the metallization. Overall contact requirements are listed and concentrator cell requirements are discussed.

Iles, P. A.↗

NASA NTRS1983

Silicon solar cell process development, fabrication and analysis

Measurements of minority carrier diffusion lengths were made on the small mesa diodes from HEM Si and SILSO Si. The results were consistent with previous Voc and Isc measurements. Only the medium grain SILSO had a distinct advantage for the non grain boundary diodes. Substantial variations were observed for the HEM ingot 4141C. Also a quantitatively scaled light spot scan was being developed for localized diffusion length measurements in polycrystalline silicon solar cells. A change to a more monochromatic input for the light spot scan results in greater sensitivity and in principle, quantitative measurement of local material qualities is now possible.

Leung, D. C.↗

NASA NTRS1983

Microcrystalline silicon growth for heterojunction solar cells

A total of sixteen runs of e-beam vacuum deposition of p type microcrystalline Si (m-Si) films were attempted on n type or p-n junction single crystalline Si (C-Si) substrates. The m-Si film thickness varied from .15 to .7 um and metal contacts were deposited after plasma hydrogenation. The p-m-Si on n-c-Si structure had a Voc of up to 490 m V while no Voc improvements were observed in the p-m-Si on p-n C-Si structure against p-n controls. Both CFF and Jsc were lower than control. Possible problem areas were interfaced between m-Si and C-si and the back contacts due to lack of sintering for fear of dehydrogenation.

Iles, P. A.↗

NASA NTRS1982

Silicon solar cell process development, fabrication and analysis

For UCP Si, randomly selected wafers and wafers cut from two specific ingots were studied. For the randomly selected wafers, a moderate gettering diffusion had little effect. Moreover, an efficiency up to 14% AMI was achieved with advanced processes. For the two specific UCP ingots, ingot #5848-13C displayed severe impurity effects as shown by lower 3sc in the middle of the ingot and low CFF in the top of the ingot. Also the middle portions of this ingot responded to a series of progressively more severe gettering diffusion. Unexplained was the fact that severely gettered samples of this ingot displayed a negative light biased effect on the minority carrier diffusion length while the nongettered or moderately gettered ones had the more conventional positive light biased effect on diffusion length. On the other hand, ingot C-4-21A did not have the problem of ingot 5848-13C and behaved like to the randomly selected wafers. The top half of the ingot was shown to be slightly superior to the bottom half, but moderate gettering helped to narrow the gap.

Iles, P. A.↗