Space-charge-limited current (SCLC) flow has been investigated as a function of applied potential, temperature and specimen thickness in amorphous silicon films prepared by the glow-discharge techn...

The density of states in amorphous silicon determined by space-charge-limited current measurements

▪ Abstract Hydrogenated amorphous silicon (a-Si:H) exhibits a metastable light-induced degradation of its optoelectronic properties that is called the Staebler-Wronski effect, after its discoverers. This degradation effect is associated with the relatively high diffusion coefficient of hydrogen and the changes in local bonding coordination promoted by hydrogen. Reviewed are the fundamental aspects of the interplay between hydrogen and electronic energy states that form the basis of competing microscopic models for explaining the degradation effect. These models are tested against the latest experimental observations, and material and preparation parameters that reduce the Staebler-Wronski effect are discussed.

Development in Understanding and Controlling the Staebler-Wronski Effect in a-Si:H

A microscopic model based on charge-induced breaking of weak Si-Si bonds is proposed as an explanation for the experimentally observed increase of the dangling-bond density in a-Si: H upon illumination, charge injection, and doping. Energetic aspects of the conversion between weak and dangling bonds and implications for the electronic density of states of a-Si: H are described, and the relation of the model to negative4 defects in this material is discussed.

Weak bond-dangling bond conversion in amorphous silicon

Recent developments in the investigation of light-induced metastable defects in a-Si:H are discussed. A kinetic model that quantitatively describes the creation of metastable dangling bonds in UHV deposited a-Si:H can be extended for a qualitative understanding of defect creation in SiGe alloys, samples containing common impurities (C,N,O), and in compensated a-Si:H. It is also shown that mechanical stress enhances the defect creation efficiency. Detailed information about the energy barrier distribution of the metastable dangling bonds can be obtained by ESR annealing studies. Finally, new results showing a light-induced, reversible quenching of low energy tunneling modes are presented.

Light-induced metastable defects in hydrogenated amorphous silicon

The density of states at the Fermi level N(EF) has been measured on hydrogenated polymorphous (pm-Si:H) silicon samples using both capacitance measurements on Schottky barriers and space-charge-limited current measurements on n+/i/n+ structures. From both techniques, N(EF) values of 7–8×1014 cm−3 eV−1 have been obtained, which is significantly lower than reported in the literature for hydrogenated amorphous silicon (a-Si:H). Such values demonstrate that pm-Si:H is a very low defect density material which should be able to replace a-Si:H in the field of applications like photovoltaics.

Very low densities of localized states at the Fermi level in hydrogenated polymorphous silicon from capacitance and space-charge-limited current measurements

The dens i ty of s t a t e s i n glow discharge deposi ted amorphous silicon has been deduced from space-charge-limited (SCL) c u r r e n t measurements on samples with an n+-i-nf sandwich configurat ion. The temperature and filmthickness dependenceoftheScL cur ren t a r e found t o be cons i s ten t with the theor y of one-carr ier s teady-state SCL cur ren t flow. The r e s u l t i n g midgap dens i ty of s t a t e s i s 1 0 ~ ~ 4 . 1 0 ~ ~ cm-3 .V-'in t h e energy region 0.75 t o 0.6 eV below t h e conduction band edge. This f i g u r e i s lower than t h a t usua l ly found by t h e f i e l d e f f e c t technique (-1017 cm-3 ev-l) , probably due t o t h e inc lus ion of int e r f a c e s t a t e s i n f i e l d e f f e c t measurements. Introduction.Much information on t h e q u a l i t y and proper t i es of hydrogenated amor-phous s i l i c o n (a-Si:H) can be obtained from i ts d i s t r i b u t i o n of loca l ized s t a t e s . The method most widely used t o measure t h e dens i ty of s t a t e s (DOS) i n a-Si a l l o y s is t h e f i e l d e f f e c t technique [1,2] . This method does not d i s t i n g u i s h between i n t e r f a c e s t a t e s a t t h e semiconductor-insulator i n t e r f a c e and bulk s t a t e s and consequently prov ides an upper l i m i t f o r t h e bulk DOS (1017 cm-3 ev-' f o r a-Si:H). Deep l e v e l t rans i e n t spectroscopy, on t h e o ther hand, i s assumed t o measure t h e t r u e bulk DOS and app l i c a t i o n of t h i s technique on a-Si:H has resu l ted i n a midgnp DOS below 1 0 l ~ c m ~ ev-l [31. Recently dark forward current-vol tage measurements on a-Si:H Schottky diodes have been in te rpre ted a s giving evidence f o r space-charge-limited (SCL)current flow [41. A quasi-exponential DOS was deduced from these experiments. The SCL cur ren t technique i s a well-known method f o r determining t r a p d i s t r i b u t i o n s i n high r e s i s t i v i t y s o l i d s and has been appl ied t o a l a r g e v a r i e t y of mate r ia l s [5]. We propose t o use a sandwich s t r u c t u r e with two ohmic contac t s t o measure SCL curr e n t flow. In t h i s configurat ion t h e I-V-character is t ic is not obscured by exponent i a l behaviour f o r low vol tages a s i n t h e case of Schottky diodes. Consequently t h e SCL cur ren t regime can be extended t o lower vol tages. Theory.Since undoped a-Si:H is s l i g h t l y n-type and t h e con tac t s i n an n+-i-n+ s t r u c t u r e a r e blocking f o r holes , t h e SCL cur ren t i s predominantly c a r r i e d by elect rons . Then, under t h e add i t iona l assumptions t h a t quasi-equilibrium e x i s t s , t h a t d i f fusion cur ren ts a r e neg l ig ib le and t h a t t h e ohmic contact i s an i n f i n i t e rese rvo i r of e lec t rons , t h e theory f o r one-carrier s teady-state SCL cur ren t flow i s appl icable [5]. F i r s t we consider a quasi-exponential d i s t r i b u t i o n of s t a t e s : E-Ec N (E) = !k eXp (P) t t kTt kT where Nt and Tt a r e parameters charac te r iz ing t h e t r a p d i s t r i b u t i o n and E i s t h e energy of t h e conduction band edge. For t h i s d i s t r i b u t i o n t h e current-density-voltage (j-V) c h a r a c t e r i s t i c i s l i n e a r on a log-log-scale, i . e . j--Vm, with m>2. N, and T, a r e deduced from t h e j-V-characteristic 151: where N i s t h e e f f e c t i v e dens i ty of s t a t e s a t t h e conduction band edge, pn i s the Article published online by EDP Sciences and available at http://dx.doi.org/10.1051/jphyscol:1981494 C4-452 JOURNAL DE PHYSIQUE f r e e e lec t ron mobil i ty , L is t h e sample thickness , s s is t h e semiconductor permitt i v i t y and T 1 = t . T I n general a s t r a i g h t l i n e f i t t o t h e log j-log V-character is t ic i s not pos,sible. The DOS can then be estimated by using the following ana lys i s . The concentrat ion of f i l l e d e lec t ron t r a p s i n equi l ibr ium i s n . When a vol tage i s t , O applied t h e Fermi-level i s r a i s e d from i t s equilibrium value E t o EFn due t o space F0 charge i n j e c t i o n from t h e ohmic contact . The concentrat ion of f i l l e d e lec t ron t r a p s then becomes nt. Assuming t h a t t h e t r a p d i s t r i b u t i o n N (E) i s continuous and only t slowly varying, n -n can be approximated by t t,a n t -n t ,o z F Nt(E) dB. (3) E ~ o To simplify t h e a n a l y s i s we assume t h a t t h e in jec ted charge i s uniformly d i s t r i b u t e d across t h e f i lmand t h a t t h e e l e c t r i c f i e l d is constant and equal t o throughout t h e film. L These two assumptions a r e contradictory, b u t it can be shown [ 5 ! t h a t t h i s s impl i f i ca t ion overest imates t h e DOS by not more than a f a c t o r of 2. The i n j e c t e d charge Q per u n i t area i s then equal t o W. L Since t h e t r a p d e n s i t y is much l a r g e r than t h e f r e e e lec t ron d e n s i t y most of t h e injected charge i s trapped, i . e . Taking two p o i n t s ( j l ,V1) and(j2,Vz) on t h e measured j-V-curve we c a l c u l a t e t h e Fermi-level s h i f t AEF=EF2-EF1 corresponding t o a vol tage change AV=V -V 2 1 AE is deduced from t h e current-densi ty equations F -where nl and n2 a r e t h e f r e e c a r r i e r d e n s i t i e s a t appl ied vol tages V1 and V 2' respect ively: E c E ~ l ) n1 = Nc exp (7 E -E c F2 n2 = Nc exp ( kT ) Then, from (5) and ( 6 ) , Using (3) and (4 ) , we can wri te: The i n t e g r a l i n (8) can be approximated by W where N i s t h e average t r a p d e n s i t y between EF1 and EF2. t Then, from (8) and (9) , Experimental.The f i lms were grown i n an r f glow discharge of s i l a n e a t a subs t ra te temperature of 300°C. For t h e -50 m th ick n+ layer 1% phosphine was added t o t h e silane. With these f i lms so la r c e l l s with e f f i c i e n c i e s exceeding 4% have been produced [6]. Quartz o r g l a s s p l a t e s with evaporated chromium s t r i p e s were used a s subs t ra tes . A s upper e lec t rodes chromium-gold s t r i p e s were evaporated perpendicular t o t h e lower e lec t rodes , t h e overlapping a r e a ( l mm2 o r 4 mm2 ) def in ing t h e device. Xesults and discussion.The cur ren t dens i ty was found t o be independent of device a rea (1 mmL o r 4 mmL ) , which el iminates edge e f f e c t s . The appl ied e l e c t r i c f i e l d should not exceed -5.10~ V/cm. A t higher f i e l d s t h e f r e e c a r r i e r dens i ty i s not only increased by space-charge i n j e c t i o n , but a l s o by f i e l d ion iza t ion out of t r a p s [7] . The temperature dependence of t h e j-V-curves is i l l u s t r a t e d i n Fig. 1 f o r a re p resen ta t ive sample with an undoped layer thickness of 0.55 pm. The t r a p dens i ty was ca lcu la ted f o r t h e curves measured a t T=251 K, 290 K and 375 K, using equations (7) and (10) . The pos i t ion of t h e Fermi-levels was ca lcu la ted from (5) and ( 6 ) , taking a value of 1021 (cm V sec)' f o r t h e pnNc product [a] . The r e s u l t i n g dens i ty of s t a t e s i s represented i n Fig. 2. For energies between 0.75 and 0.6 eV below t h e conduction band edge t h e ca lcu la ted N+(E) i s i n t h e range Fig. 1 : j-V-curves a t t h e indicat e d temperatures f o r a sample with undoped layer thickness L=0.55 pm. C4-454 JOURNAL DE PHYSIQUE -The DOS has a l s o been ca lcu la ted by taking t h e bes t s t r a i g h t l i n e f i t s f o r t h e log jlog V-curves a t 251 K, 290 K and 375 K and using equation ( 2 ) . The r e s u l t i n g exponenti a l d i s t r i b u t i o n s a r e shown i n Fig. 3 and a r e i n reasonable agreement with t h e re s u l t s of Fig. 2. Equation (2) shows t h a t t h e SCL cur ren t s t rongly depends on t h e undoped layer thickness L. A s an add i t iona l check on t h e v a l i d i t y of t h e method t h e DOS has been calcul a t e d f o r samples with d i f f e r e n t values of L. Fig. 4 shows t h e DOS ca lcu la ted f o r samples with L=0.55, 1.1 and 2.45 pm, measured a t T=290 K. The DOS is v i r t u a l l y independent of sample thickness . Fig. 3 : Densi t ies of s t a t e s ca lcu l Fig. 4 : Densi t ies of s t a t e s calcula ted f o r t h r e e curves of Fig. 1 a t a ted f o r t h r e e d i f f e r e n t samples with T=251, 290 and 375 K. The ca lcu la t undoped layer thickness L=0.55, 1.1 ions a r e based on equat ions (1) and ( 2 ) . and 2.45 pm. The measuring temperat u r e was 290 K. Calculat ions a r e based on equations (7) and (10) . I n summary, a midgap dens i ty of loca l ized s t a t e s i n a-Si:H of -1016 e ~ l c m ~ h a s been deduced from SCL currentmeasurements. The dependenceoftheSCLcurrent onsample thickness and temperature i s i n reasonable agreement with theory. The ca lcu la ted DOS i s supposed t o represen t t h e t r u e bulk DOS. Therefore, we argue t h a t t h e SCL cur ren t technique provides a more r e l i a b l e est imate of t h e bulk DOS i n a-Si:H than t h e f i e ld e f f e c t r e s u l t s i n which i n t e r f a c e s t a t e s a r e l i k e l y t o dominate. References [ l ] Madan, A., Le Comber, P.G. and Spear, W.E., J. Noncryst.Solids 20 (1976) 239. [2] Madan, A., So la r Ce l l s 2 (1980) 277. [3] Cohen, J . D . , Lang, D.V., Harbison, J.P., Bean, J . C . , Solar C e l l s Z (1980) 331. [4] Ashok, S., Lester , A. and Fonash, S .J . , IEEE E1.Dev.Lett. EDL-1 (1980) 200. [5] Lampert, M.A. and Mark , P , Current I n j e c t i o n i n So l ids , New York, Academic Press (1970). [6] Ondris, M. and Den Boer, N., Proceedings of t h e Third E.C. Photovoltaic Solar Energy Conference, Cannes 1980, p. 809. 171 Carlson, D.E. and ~7ronski , C.R., Amorphous Semiconductors, ed i ted by M.H. Brodsky, Springer, Berl in (19791, p. 287. [8] Wronski, C.R., Carlson, D.E. and Daniel, R.E., Appl.Phys.Lett. 29 (1976) 602.

DETERMINATION OF MIDGAP DENSITY OF STATES IN a-Si : H USING SPACE-CHARGE-LIMITED CURRENT MEASUREMENTS

The far forward current in α-Si:H Schottky diodes has been compared with the current in nin devices with identical i-layers. The former is about one order of magnitude larger and has to be described by double injection, while the latter is governed by single-carrier space-charge-limited conduction. The transient behaviour of the diodes is also different from that in nin devices and confirms the occurrence of double injection. Metastable changes in α-Si:H (the Staebler-Wronski effect) are obtained after current-soaking in pip samples, but not after current-soaking in nin samples. To explain this, a model is suggested which is based on SiSi bond weakening by hole trapping in valence band tail states.

W. den Boer

Papers

DETERMINATION OF MIDGAP DENSITY OF STATES IN a-Si : H USING SPACE-CHARGE-LIMITED CURRENT MEASUREMENTS

Single and double carrier injection in A-Si:H

Computer simulation of the optical behaviour of amorphous silicon solar cells

Interference-enhanced long-wavelength response in amorphous silicon solar cells

Alternating space-charge-limited currents in hydrogenated amorphous silicon