CHANGES OF SURFACES OF SOLAR BATTERIES ELEMENTS OF ORBITAL STATION MIR AS A RESULT OF THEIR PROLONGED EXPOSITION ON LOW- EARTH ORBIT (LEO) V. E. Skurat (1), I. O. Leipunsky (1), I. O. Volkov (1), P. A. Pshechenkov (1), N. G. Beriozkina (1), V. A. Letin (2), L. S. Gatsenko (2) (1) Institute for Energy Problems of Chemical Physics, Russian Academy of Sciences, 38, bldg 2, Leninsky prospect, Moscow, 119334, Russian Federation, email: skurat@chph.ras.ru (2) SPRE Kvant, 16, Tretja Mytischinskaya, Moscow, 129626, Russian Federation ABSTRACT Several fragments (front and back protective glasses) of solar battery (SB) were investigated after its return from the orbital station MIR to Earth in 1998 year after working in LEO conditions during 10.5 years. Three methods of surface analysis were employed: Scanning electron microscopy, local X-ray microanalysis and X-ray photoelectron spectroscopy (PES). Main changes of SB surfaces after prolonged exposition on LEO are caused by contamination layers (thickness up to 3 µm) that deposit from intrinsic outer atmosphere of the station. Besides contamination deposits, some surface damages are observed (dimensions 5 20 µm), caused by micrometeoroid and / or debris impacts. Outer layers of contamination deposits consist of spheroid particles with dimensions 0.3 0.5 µm. Main components of deposits are carbon (three chemical forms were observed by PES) and silicon (two chemical forms silicon dioxide and linear organosilicone). Silicon carbide presence is possible. The smoothness of initial glass plate substrate for contamination is sufficient for angleresolved measurements by PES. But the surface of contamination deposit is too rough for these measurements after exposition. During flight, one side of fragment was predominantly illuminated by sun, while opposite side was predominately in shadow. So, comparative studies of sun-illuminated and back sides give the opportunity to study the effect of solar illumination on contamination deposits. Quantitative estimates were made for losses of SB power as a result of surface contamination and local damages. Full optical losses of power during 10.5 years reach 8%. Contamination of optical system is responsible for 3%, and bombardment by micrometeoroids and debries for another 3%, other factors for remaining 2%. INTRODUCTION Numerous investigations and results which were obtained during working of solar batteries (SB) on low- Earth orbits (LEO) showed that in many cases, the decrease of SB power is caused by deterioration of parameters of optical system of SB (optical degradation). Optical system has many-layer structure, consisting of outer layers protecting glass plates (PGP) and inner layers organosilicon laquer and resin, antireflecting coating on silicon plate of SB, glass cloth layers. Optical degradation can be caused by several factors: - Contamination from intrinsic outer atmosphere of station, gas evolution from station volume, and products of incomplete combustion from plumes of station engines; - Destruction of SB materials by orbital atomic oxygen and solar UV radiation; - Erosion of PGP surfaces and their contamination by micrometeoroides and debris: - Light-induced darkening of organosilicon adhesives. Now, it is known [1] that one of the main mechanisms of contamination deposit (CD) formation on optical and thermo-control surfaces is photochemical deposition (besides condensation). Solar radiation, mainly UV and vacuum UV components, induces numerous photochemical reactions with participation of organic materials of spacecraft surfaces and organic components of intrinsic outer atmosphere. Products of these reactions have enhanced attitude to CD formation on spacecraft surfaces. Besides, contaminations can be 1
formed by destruction of organic materials in solid state and by diffusive transfer of formed products. So, study of CD on optical surfaces of SB of their chemical composition, morphologies and other properties has significant interest from the point of view of elucidation of mechanism of optical degradation of SB and development of methods of its decrease. Other points of interests are following: detection of photochemical effects by comparison of CD on illuminated and back sides of SB fragments; smoothness of glass surfaces which permit, in principle, to obtain angle-resolved X-ray photoelectron spectra of CD layers. Here, results are given for study of surfaces of SB from space station MIR after prolonged work on LEO (during 10.5 years) and returning to Earth in 1998 year. Mainly, three methods were used: Scanning electron microscopy (SEM) for study of surface morphology; Local X-ray microanalysis (LXMA) for study of element composition of surface parts with dimensions 1x1 µm 2 on the depths 0.5 1.0 µm. Method is sensitive to carbon, oxygen and other elements with atomic numbers above 11. X-ray photoelectron spectroscopy (XPES) or electron spectroscopy for chemical analysis (ESCA). Method determines element composition (except hydrogen and helium) of surface parts with areas up to 1 cm 2 on the depths 3 5 nm. Different chemical forms of elements can be discovered on the basis of chemical shifts of spectral bands. Their contents can be determined in surface layers with averaging on depths 3 5 nm. These methods were used for investigation of other samples from space station MIR in [2 4]. EXPERIMENTAL Front (sun-illuminated) and back (shadow) sides of three fragments from SB of orbital space station MIR were studied. Reference sample, which had not been exposed in space, was also studied. These sides belong to surfaces of two different protective glass plates (PGP) on mutually opposite sides of a fragment. During space station flight, PGP on front side was predominantly illuminated by sun, while PGP on back side was predominantly in shadow conditions (without direct illumination). Solar elements were attached by threads to substrate manufactured from glass net. This glass net covered the back side of PGP, leaving free access of contamination during space exposition. Measurements by SEM and LXMA were made on instrument Camebax MBX-1 (Cameca, France). Thin copper layers were deposited on sample surfaces for studying of their morphologies by SEM. This deposition was absent when carbon content was measured by LXMA. Measurements by XPES were made on instrument XSAM-800 (Kratos Analytical Instruments, UK). Excitation of spectra was made by characteristic X-rays of Mg with photon energy 1253.6 ev. Spectra were detected at two values of angle of photoelectron take off 90 0 and 35 0 (angle between direction of photoelectron exit and sample surface) which correspond to probing depths 5 and 3 nm respectively. RESULTS Fig. 1 3 show morphologies of sample surfaces reference sample and samples from front and back sides of SB which have been obtained by SEM in regime of secondary electron registration at magnifications from 400 to 50000. In difference with smooth surface of reference sample, exposed samples of both types show deposits consisted of particles of oval form with typical dimensions 0.2 0.5 µm. Front and back sides of SB have marked differences in morphologies. Particles of front side have flattened forms. On some areas, particles of irregular forms are observed with dimensions higher than 1 µm imbedded into CD. These particles appear to be debris particles. Their surface density is equal to about 10 7 cm -2. CD particles on back side have more rounded forms in comparison with front side. They are packed less densely and have other alignment. Their long axes align mainly along surface plane while on front side, they align predominantly at normal to surface. 2
Back side contains also flakes with dimensions up to 15 µm and thickness about 1 2 µm. They can be formed by scaled up parts of CD layer. Both sides contain local surface damages with dimensions 5 20 µm. Most likely, they were caused by micrometeoroid impacts. Front side contains surface parts where upper layer of CD is absent. Such parts were not found on back side. May be, some parts of CD upper layer were removed from front side by erosion with orbital atomic oxygen. LXMA data evidence the presence of carbon, silicon, sodium, calcium, chlorine, and phosphorus in CD. Carbon content in near-surface layer of CD on front side is not less than 10% by weight, while on back side, it is not less than 25% (on virgin glass sample, carbon content is less than 0.5%). Results of analysis by PES are given in Table 1. For virgin glass surface, there are significant differences in C1s-spectra at take-off angles 90 0 and 35 0. Relative carbon content increases significantly from 90 0 (probing depth d = 5 nm) to 35 0 (d = 3 nm). This means that registered carbon is localized in thin surface layer (thickness less than 5 nm). For exposed samples, the differences in carbon contents at various take-off angles are not marked. This can be explained both by great thickness of CD (much higher than probing depth), and by its grain structure. PE-spectra of virgin glass surface show presence of carbon, oxygen, silicon, boron, sodium, and potassium. Spectral bands of O, Si, B, Na, and K are typical for boron-silicon glass K-208. Band C1s can be resolved on four Gaussian components that correspond to four chemical forms of carbon. Three of them can be related to surface contaminations from air (bond energies 285.0; 286.5; 287.9 ev), but the form with bond energy 283.3 ev can be related with carbon in more deep layers. Surfaces of exposed samples contain carbon, oxygen and silicon. Half-width of Si2p-band is higher in comparison with virgin glass. This band can be resolved on two Gaussian components corresponding to SiO 2 (103.7 ev) and to linear chains - O Si O -, typical for organosilicon compounds (Fig. 4). Decomposition of organosilicon compounds can, in principle, gives not only silicon oxides SiO x, but silicon carbide, too. Its search by PES did not give positive results. But it is not possible to completely exclude its presence in CD on the basis of the absence of component with energy bond 100.4 ev in Si2p signal. According to [5], SiC gives this component. But absence of this signal can be explained by its shift to the region 102.5 ev where it is covered by SiO x signal. DISCUSSION Main source of contamination, observed on glass surfaces of fragments of SB, is intrinsic outer atmosphere of space station. It originates from both SB, and other station parts. Contamination sources from SB can be its organic materials: organosilicon compounds (resins, silicon lacquers), hydrocarbons and others. Carbon content in near-surface layer of contamination was higher for back side in comparison with front (sunilluminated) side. This result appears to be in contrary to expectation on the basis of mechanism of photochemically enhanced contamination deposition from intrinsic outer atmosphere. Because of this, we suppose that excessive carbon on back side is supplied by hydrocarbons from glass-net supporting solar elements from back side (glass-net is impregnated with paraffin oil). The other possible carbon source is organosilicon compounds and their volatile components. These compounds are also most likely source of silicon that has been discovered in two forms: organosilicon chains Si O Si O and SiO 2 - product of full oxidative degradation of organic parts of these molecules. At present time, the quantitative estimate of optical degradation of SB is rather difficult task, because of too many factors must be included in the calculations. Nevertheless, some approximate calculations give results in correlation with real data. For SB panel that was returned from space station MIR after 10.5 years of work in LEO conditions, optical losses were equal to about 8 %. About 3 % were caused by contamination of optical system, about 3 % - by erosion of protective 3
glass plates by micrometeoroides, and remaining 2 % - by some other factors. CONCLUSIONS 1. Overwhelming part of surfaces from both front and back sides of SB is covered by contamination deposit (CD) with thickness up to 3 µm. Surface of CD has granular structure. It consists of grains with oval form (from near spherical to extended with dimensions 0.3 0.5 µm. Structures of CD (forms and alignment of extended grains) are different on front and back sides of SB. 2. Local X-ray microanalysis data (element contents to depths up to 1 µm) showed that CD contain Si, C, K, Ca, Cl, and P. Carbon content in CD on back side is at least two times higher in comparison with front side. This result appears to be in contradiction with expectation on the basis of photochemical model of contamination deposition from intrinsic outer atmosphere of space station. It can be explained by presence one additional path of carbonaceous contamination deposition for back side, namely, by paraffin oil from glass-net. 3. XPES data for CB showed (probing depth up to 5 nm) that CD surface consists of C, Si, and O as main elements. Three carbon chemical structures and two silicon chemical structures were discovered. Grain morphology of CD surfaces gives them the roughness that impedes obtaining of quantitative data on depth dependence of CD chemical composition by angle-resolved XPES. 4. On CD surface, nitrogen was not found (its content is smaller than 1%). This result suggests the absence of contamination by products of incomplete combustion from plumes of space station engines (one of rather extended sources of contaminations). 5. Results of quantitative estimates are given for SB power loss because of contamination and local damages of SB surfaces. 2. Skurat V. E. et al., High Performance Polymers, Vol. 13, 337 353, 2001. 3. Naumov S. F. et al., Proc of the 9-th Intern. Symp. on Materials in a Space Environment, The Netherlands, Noordwijk, 16 20 June, 2003, 603 608, 2003. 4. Naumov S. F. et al., Proc. of the 9-th Intern. Symp. on Materials in a Space Environment, The Netherlands, Noordwijk, 16 20 June, 2003, 595 602, 2003. 5. Smith K. L. and Black K. M., J. Vacuum Sci. Technol., Vol. A2, 744 745, 1984. REFERENCES 1. Tribble A. C., J. Spacecraft and Rockets, Vol. 25, 114 116, 1998. Fig. 1. SEM images of reference surfaces 4
Fig,2. SEM images of various surface parts on front side 5
Fig.3. SEM images of various surface parts on back side Fig.4. Si 2p XPE spectra оf virgin glass (left) and contamination deposit on front side SB (right). 6
Table 1. Results of analyses by PES Element Concentration, at.% Bond energy, ev Line Intensity, % Assignment Virgin glass, θ =90 Cls(l) 25.7 283.3 37 Cls(2) 285.0 39 С-С/С-Н Cls(3) 286.5 19 С-О Cls(4) 287.9 5 С=О Ols 45.0 532.8 Nals 2.8 1072.8 K2p 2.0 293.8 Si2p 20.3 103.2 B1s 4.2 193.1 Virgin glass, θ =35 Cls(l) 42.6 283.3 33 Cls(2) 285.0 52 С-С/С-Н Cls(3) 286.5 10 С-О Cls(4) 287.9 5 С=О Ols 34.1 532.9 K2p Si2p 1.1 17.7 293.6 103.2 B1s 4.5 193.1 Glass from front side of SB, θ =90 Cls(l) 47.5 285.0 94 С-С/С-Н Cls(2) 286.5 4 С-О Cls(3) 288.0 2 С=О Ols 30.0 532.9 Si2p(l) 21.6 102.5 47 Si2p(2) 103.7 53 SiO 2 Fls 0.9 689.7 Glass from back side of SB, θ =90 C1s(1) 285,0 С-С/С-Н C1s(2) 45,4 286,5 С-О Cls(3) Ols 29.8 532.9 288.0 3 С-О Si2p(l) 22.1 102.5 40 Si2p(2) 103.7 60 SiO 2 7