Available information defining the state of the art of encapsulation materials and processes for terrestrial photovoltaic devices and related applications were collected and analyzed. Based on criteria of properties, processability, availability, and cost, candidate materials were identified which have potential for use in encapsulation systems for low-cost, long-life terrestrial photovoltaic arrays manufactured by automated, high-volume processes. The study was in support of the Encapsulation Task of the ERDA Low-Cost Silicon Solar Array (LSSA) Project, managed by JPL. The criteria for consideration of the encapsulation systems were based on the LSSA goals for arrays with a lifetime of over 20 years high reliability, an efficiency greater than 10 percent, a total array price less than $500/kW, and a production capacity of 5 x 10-> kW/yr. Published and unpublished information relating to encapsulation systems and materials properties was collected by searching the literature and appropriate data bases (over 1300 documents were selected and reviewed) and by personal contacts including site and company visits. A data tabulation summarizing World experience with terrestrial photovoltaic arrays (50 installations) is presented in the report. None of the encapsulation materials used meets all of the LSSA criteria (particularly cost), but some have performed well. R. B. Bennett. F. A. Sliemers led the polymer-chemist effort with contributions from R. E. Sharpe, G. P. Nance, and A. R. Bunk. R. D. Igou led the information searching and retrieval efforts. E. L. Briich managed the information system for the study and J. J. Breslin helped with the reporting; their assistance is certainly appreciated.

Work Performed Under Contract No. NAS-7-100-954328 Battelle Columbus Laboratories Cofumbus, Ohio U.S. Department of Energy Solar Enery-DISCLAIMER This report was prepared as an account of work sponsored by an agency of the United States Government.

The thermoelectric force developed by the following junctions up to 2200°C has been measured: Re vs W, Re vs Mo, Re vs W-30Re, and Re vs Mo-50Re. The thermoelectric force for W vs W-30Re and Mo vs Mo-50Re has been calculated. The Re vs W junction develops about 15 mv at 1000°C and may be usable up to 2600°C. Re vs Mo develops about 17 mv at 1000°C; the thermoelectric power is low above 1600°C. The outputs of Re vs W-30Re and Re vs Mo-50Re couples are low and change signs at about 1200 and 1800°C, respectively. The calculated outputs of W vs W-30Re and Mo vs Mo-50Re are high; the W vs W-30Re output is nearly linear over a wide range in temperature, and the thermocouple should be useful to about 2600°C. All of these couples must be used in vacuum or in neutral or reducing atmospheres.

Rhenium, tungsten, and rhenium‐coated tungsten have been exposed to carburizing atmospheres and to water‐cycle attack, rhenium and tungsten to aluminum oxide at 1750 °C for 7000 hr, and rhenium to boron‐containing atmospheres. It was confirmed that rhenium, unlike most other refractory metals, does not form a carbide. Rhenium forms a boride, which forms a eutectic with another boride or with rhenium, and melts under 2250 °C. Thoriated rhenium maintains microstructural and mechanical integrity after long exposure in vacuum at 1750°C in contact with alumina. Rhenium and rhenium‐coated tungsten are highly resistant to water‐cycle effects.

Thir npoat wm p m p d u m lfromt d aurL qmwd by W UntW Stat81 Ckmmment. Wther the Uaitsil State!! nw the Unibcd Strtw hpartmnt of Enozgy* naa mu oZ HI AT wnplwce& IIOP. may Of thair contractan, aubcontnctox), oa tha,ir m p l o y~, ~m y-t y ,~w t a s r~,~t r r s s u m c r w m m t v o r~~f o r the ac6unoy. eomplegsnrm m kdblnm8 of my infonn8tion, qpetna, pduct ~a pmeew ~~r b p r w o n~t l u t i t r w r s D u l d a a t Z n f r i p e e~b s l y o~* t k Thir =port has beem reproduced directly fiom the bmt a-b mpy.