Spray CdS-polyacetylene thin films photovoltaic heterojunctions

Spray CdS-polyacetylene thin films photovoltaic heterojunctions

Solar Energy Materials 13 (1986) 307-318 North-Holland, Amsterdam 307 SPRAY C d S - P O L Y A C E T Y L E N E T H I N F I L M S P H O T O V O L T A ...

431KB Sizes 1 Downloads 54 Views

Solar Energy Materials 13 (1986) 307-318 North-Holland, Amsterdam

307

SPRAY C d S - P O L Y A C E T Y L E N E T H I N F I L M S P H O T O V O L T A I C HETEROJUNCTIONS

M. A B D - L E F D I L t), M. C A D E N E t), M. R O L L A N D I), j. B O U G N O T 2) and M.J.M. A B A D I E 3) l) Groupe de Dynamique des Phases Condensbes (L.A. 233), Universitb des Sciences et Techniques du Languedoc, 34060 Montpellier Cbdex, France e) Centre d'Electronique de Montpellier, Universitb des Sciences et Techniques du Languedoc, 34060 Montpellier Cbdex, France -~) Laboratoire d'Etudes des Mat~riaux Polymbres, Universitb des Sciences et Techniques du Languedoc, 34060 Montpellier Cbdex, France

Received 5 November 1985 We made thin films photovohaic cells by a direct polymerisation of (CH)x onto a sprayed CdS layer. Electrical and optical characteristics of this device were measured, both with undoped (CH)~ (p-n heterojunctions) and heavily doped (CH)x (Schottky diode). In spite of its low efficiency such a junction would lead, after improvement, to very cheap photovoltaic cells.

1. Introduction Since the synthesis by Shirakawa et al. [1] of large area free standing films of polyacetylene, a strong interest was focused on this cheap material. It may be doped by a number of chemical species such as 12, AsF s, H2SO 4, . . . (p-type), or Li, Na, K (n-type). Doping increases its electrical conductivity by more than twelve orders of magnitude (from 10 - 9 up to 10 +3 ( ~ c m ) - t ) . At low doping level, (CH)x has a semiconducting behaviour and reaches a quasi metallic state when heavily doped. On the other hand, the (CH)x optical properties are also of strong interest as it appears as a direct gap semiconductor with an optical bandgap lying from 1.5 eV (trans stable isomer) to 1.8 eV (cis isomer), which is well matched to the solar spectrum. At last, its absorptivity around the gap increases above l0 s cm-1. Concerning the field of energy storage or production, many works deal with electrochemistry to build up batteries [2,3]; photoelectrochemical cells [4], photovoltaic devices, both p - n heterojunctions [5-7], and Schottky devices [8-10]. Our goal is to build up very low cost solar cells even if the efficiency remains weak. Consequently, we chose the spray technique to obtain thin polycrystalline CdS layers. These cells have been characterised by SEM and by their electrical and optical behaviour, studied under dark and illuminated conditions. 0165-1633/86/$03.50 © Elsevier Science Publishers B.V. (North-Holland Physics Publishing Division)

3()~

Fig. 1. SEM micrographs of the device. (a) Cross section with as grown ('dS ~with bowls). (b) (Tro~, section with polished CdS. (c) Surface of the polyacetylene layer.

M. Abd-Lefdil et al. / CdS- polyacetylene thin films

309

2. Samples preparation and morphologies 2.1. CdS film

We used the typical method of Chamberlin et al. [11] to obtain thin sprayed CdS layers. Starting from an aqueous solution of cadmium chloride and thiourea we sprayed the melt solution on heated ITO conducting substrates (450°C). The chemical reaction which leads to CdS is written as: (T)

CdCI 2 + (NH2)2CS + 2 H 2 0 ~ CdS{ + 2NH4C--i1' +CO:1". The rate of CdS deposition is - 1 ~ m / h , the film thickness ranges from 3 up to 10 ~m. We used this technique as it is an economical one to obtain a large CdS area at low cost. 2.2. (CH)x films

This cis isomer polyacetylene films were prepared by Shirakawa's technique [1]. The molar ratio of catalysts was A1/Ti = 4 with 0 . 1 M / I of Ti; the gaseous acetylene pressure was 160 m m Hg. The cadmium sulfide layers were wetted by the catalytic solution. After polymerisation, the films were washed with toluene under vacuum. Simultaneously (CH)~ films were polymerised on glass substrates for measurements of the electrical conductivity, thickness, optical transmission etc. before and after doping. Such work is needed as the physical properties of (CH) x strongly depend on the synthesis parameters. To obtain the trans stable isomer the samples were heated, under dynamic pumping, at 140°C during 10 min. The film morphologies were studied by means of scanning electron microscopy (JEOL JSM 35 apparatus). The successive layers of this device can be observed in figs. l a and lb: glass with a conducting ITO layer (0.5 ~m), CdS sprayed film and fibrillar polyacetylene. The (CH)x surface aspect is shown in fig. lc. Electrical electrodag -I- 502

_\

f

ITO

I

- 1 " - - glass

//h, Fig. 2, Schematicdiagram of a cell.

311)

Jsc (~A cma+

lOOq

10(

I

,01

I

Fig. 3. Short circuit current dependence of HzSO a doped (CH), thickness.

contacts on (CH), are fastened with electrodag + 502 paste. The schematic diagram of the whole cell is shown in fig. 2. To strongly reduce the (CH)~ electrical resistivity we doped it with sulfuric acid (aqueous solution 6M). The samples were then washed with cyclohexene and dynamically pumped in order to remove water and interfibrillar nonfixed H 2SO4 molecules. We studied the influence of (CH)~ thickness on the photovoltaic parameters in ( C H A , ) , . - " s p r a y " CdS [12]. It appears that a thickness of - 6 ~m gives the best results (fig. 3).

3. Experimental results 3.1. Dark studies with undoped (CH), Fig. 4 exhibits log J versus the direct voltage at various temperatures. We observe a similar behaviour in the whole studied range (100-300 K); the saturation current density J, is deduced from these curves and its typical values are reported in table 1.

Table 1 Saturation current density versus temperature T (K) J~ (~A cm 2)

190 0.3

230 0.4

273 0.8

300 2

311

J cA~ 1~

300K

,'

,,~..~

273K

J

l(i6lJr ( A . c m - z )

230K

31111K

190K 273 K

23O K 190 K

q

1 li71

"

"

I

..~

168~_ . . . . . . . . . .

v (volt)

u

lJ

=.o

v

(,~)

Fig. 4. Direct log J = f ( V ) characteristics at various T for an undoped cell, Fig. 5. Reverse log J = f ( V ) characteristics at various T for an undoped cell.

~3KHz

T.3OOK

eSKHz

S=4mm~

I/C2*IO n

(F-2)

S KHz 8

• 12 KHz

i

.

.

.

.

.

.

.

. .

.

.

.

.

.

.

.

.

.

= |

-2

.



!

I

I

I

I

I

t

I

~

-t

!

t

:

:

:

:

;

:

~

~

. . . . . . . . .

0

i

-

I

;

~

:

:

:

:

~

i

2 V (VOLTS)

Fig. 6. Capacitance versus reverse voltage of an undoped cell.

312

,'~f 4t~d. Letdil et al

(dS

p,)/va~etv/enc thin /~lm~

Fig. 5 shows log J versus the reverse voltage for the same temperatures: ~ c simultaneously observe the variations of J with voltage and temperature. Fig. 6 shows the C 2 = f ( V ) dependence at various frequencies for an undoped ( C H ) , - C d S heterojunction. We always found linear curves, corresponding to a diffusion voltage of - 1 . 0 V, from which we can determine the carrier densi b variations and the depletion zone width. We deduced N.~ = 2 x 10 t'~ cm 3, which is similar to those given by other authors [8,9]. Elsewhere we used another device (undoped trans ( C H ) , Si n +n Schottky diode) [12] to determine the NA value. Taking into account that in such a device the space charge only lies in (CH)~, as the n ' region o f Si contains N D-- 10 2o cm ~, we deduced the mean value N.~ to be 3 x 10 ~7 cm ~. which is in good agreement with the above results.

J

tA) _,

163}-

I

300 K

,.

273 K

'

.f 190 K

, " x 230 K ,~,

,Z

1Jo~ r~ (A' cm-2)

~:" 140 K

¢.,+ 100 K

300 K

p

/

~

273K

/ d"

", /,7 , /,

165

;'¢/ j

230 i

,,,

i

/

'

" :'' '+

S= 10 mm

2

I+/ [ /

V/ 1071 Q,2

1,0

_-,i VlVOLT)

0.2

1.0

Fig. 7. Direct log J = f ( V ) characteristics at various temperatures for a H 2 SO4 doped cell. Fig. 8. Reverse log J = f ( V )

characteristics at various temperatures for a H 2 S O 4 doped cell.

M. Abd-Lefdil et al. / CdS- polyacetylene thin films

313

3.2. Dark studies of doped (CH)x with H, SO 4 In fig. 7 we show log J versus direct voltage. We essentially note a higher J, saturation current and a much lower series resistance (typically R~ = 13 ~2 at 300 K) than the ones obtained with undoped samples. However the general behaviour is the same. We show in fig. 8 the log J variations versus reverse voltage; we only note an enhancement of JR (compared to undoped ( C H ) , ) .

3.3. Photovohaic behaviour of undoped (CH)x Fig. 9 shows J = f ( V ) curves under various illuminations. The reverse curves are characteristics of weak shunt resistance; the fill factor lies around 0.35, Due to the high series resistance ( - 3 . 2 kf~ for 1 cm 2 of surface) the short circuit current remains very weak (18 ~A cm -z under 100 mW c m - 2 ) . On the various samples we observed a large dispersion of Voc values up to 325 mV,

•hA cr~ 2)

(12 2O

50mwcm -z

130

/ 2

100

Fig. 9. I l l u m i n a t e d J - V characteristics for an u n d o p e d d o p e d cell.

¢,r~

--100 SO

FF - 0.43 0.35 o.37

J

.~cm "z)

/ -2

/

f

20

/

z

~

~

jj /

©

~J

J [email protected]

Fig. 10. I l l u m i n a t e d

J

Vcharacteristics

for an H 2 S O 4 cell.

0.3

M. Abd-Lefdil et al. / CdS- polyacetylene thin films

315

10

//

,o,

/

i0 -I

10

10 1~

10

10-,

~n'vNcm J)

Voc(,,~1

02

f~

Fig. 11. Energy dependence of J~¢ and V~,¢for (a) an H2SO 4 doped cell, (b) an undoped cell.

3.4. Photovoltaic behaviour of doped (CH)x The results of doped (CH)x are similar to the results of undoped (CH,) (fig. 10), however, the strong decrease of the series resistance leads to a higher J,,c ( × 100), 1.5 mA cm -2 under 1 sun. Elsewhere, we have plotted in fig. 11 J~ =f(q,) and Vo~ = f ( l o g ~) for undoped and doped (CH),., where ~ is the illumination density. We always found the classical laws Jsc=Aq~ and Vo~=B log q~. Lastly, fig. 12 exhibits the spectral response of CdS-undoped (CH)x and CdS-[CH(HzSO4)y] ~ cells. Although the best response lies around the CdS band gap (Eg -- 2.40 eV), we note a strong widening at the lower energy side of the transpolyacetylene bandgap. -

316

M. Ahd l.e#hl c't al.

( d.',;

po/)'ac'e/.vlen(' Hun ti/m,

,/ .u

10

./£." /

I

I

,.s

2

E(ev)

2.5

Fig. 12. Spectral response of undoped ( - - ) and doped (ll) cells.

We also notice an amelioration of the spectral response, after doping, in the low energy region. 4. Discussion 4.1. Dark studies

The J

=f(V)

characteristics of the cells, under forward polarization, follow the

current-voltage relationship:

J=J~(expIa(V- R~J)] -1}, where

a is t h e s l o p e o f t h e l o g J = f ( V )

c u r v e s , R~ is t h e s e r i e s r e s i s t a n c e a n d

J~ is

Table 2 Reverse saturation current density and series resistance versus temperature for undoped and doped diodes T (K)

Undoped Doped

J~ ~A c m - 2 R~ k[2 3~, ~A cm -2 R s f~

300

273

230

190

140

100

2 3.2 10 13

0.8 6.4 7.2 28

0.4 15 5 54

0.3 32 3.5 140

2 300

1.2 ,430

M. Abd-Lefdil et aL / CdS- polyacetylene thin films

317

the saturation current density. We observed that a is temperature independent. At "enough" weak current density J, we have qV>> R~J and log J versus V is linear, however the term RsJ becomes important and produces the departure of the linear law in the curves of fig. 3, Values of Js and R s versus temperature are summarized in the table 2 for doped and undoped samples (results for a 1 cm 2 cell surface). If we plot R s = f ( l O 3 / T ) we find, for undoped samples, the temperature dependence of (CH),. raw material. This proves that all series resistance lies in the polyacetylene. With doped samples, although R~ is much lower, the law is similar. This last result indicates that we need to improve the electrical conductivity of the doped polyacetylene. We also observed a variation of Rs with the current at has been already noted by Peebles [10]. We think that the correction of I,V curves only with the series resistance effect is not sufficient, but that the "two exponential" model will explain more correctly our I , V dark characteristics of undoped (CH)x-CdS cells. Concerning the doped cells, we can assume that they are semiconductor-metal junctions due to the metallic behaviour of polyacetylene. The theory which seems to explain correctly the conduction mechanism in our doped solar cells is the "multisteps tunnel effect". Consequently we found that:

J = J~ exp( a V )

and

J~ = A e x p ( y T ) ,

so J = A

e x p ( a V + "/T).

The multisteps tunnel effect gives J as: J = C exp[aR-l/2(VD -- V)],

where

4 [ m * , s ] 1/2 a='~

ND j

.

R, the number of steps in the tunneling effect, is defined by R = ( ~ / a ) 2. If we suppose that the diffusion length varies linearly with the temperature we find

J = C e x p [ - a R - ' / 2 ( V o o - T - V)]. By identification, we find R = ( a / a ) 2, A = C e x p ( - a R results give: R = 170

and

1/2VDo) and h = 7/a. Our

h = 1, 2 mK -1.

These values of R and h are in good agreement with those obtained by Martinuzzi et al. in C u 2 S / C d S cells.

5. Conclusion

We made, with sprayed cadmium sulfide, photovoltaic cells using trans (CH)~; both kinds of junctions were obtained: p - n heterojunction used undoped (CH)~

318

M. A bd-Le/dd et al. / ('dS- polvace(vlene thin !th~l~

a n d m e t a l - s e m i c o n d u c t o r j u n c t i o n w h e n the p o l y a c e t y l e n e is heavily d o p e d with s u l f u r i c acid. This last j u n c t i o n exhibits a c o n v e r s i o n efficiency of - 0.3~. W e ca~ c o n c l u d e that this efficiency is rather p o o r however it does n o t arise from a Io~ p h o t o v o l t a g e b u t r a t h e r f r o m a weak p h o t o c u r r e n t . T o r e m o v e this p r o b l e m we a c t u a l l y try to i m p r o v e the d o p i n g of p o l y a c e t y l e n e a n d also to r e d u c e the c o n t a c t resistance o n p o l y a c e t y l e n e in o r d e r to s t r o n g l y r e d u c e the d i o d e series resistance.

Acknowledgements W e g r a t e f u l l y a c k n o w l e d g e Mr. J.L. R i b e t for the p o l y a c e t y l e n e p r e p a r a t i o n . T h i s w o r k was s u p p o r t e d b y P I R S E M c o n t r a c t no. 95-34-03.

References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12]

T. Ito, H. Shirakawa and S. Ikeda, J. Polym, Sci. 12 (1974) 11. T, Skothein, O. Inganas, J. Prejza and I. Lundstrom, Mol. Cryst. Liq. Cryst. 83 (1982) 329. H. Shirakawa, S. Ikeda, M. Aigawa, J. Yoshitake and S. Suzuki, Synth. Meth. 4 (1981) 43. S.N. Chen, A.J. Heeger, Z. Kiss, A.G. MacDiarmid, S.C. Gan and D.L. Peebles, Appl. Phys. Lett. 36 (1980) 96. M. Cadene, M. Rolland and M.J.M. Abadie, Rev. Phys. Appl. 18 (1983) 691. M. Ozaki, D. Peebles, B.R. Weinberger, A.J. Heeger, A.G. MacDiarmid, J. Appl. Phys. 51 (1980) 4252. M. Ozaki, D.L. Peebles, B.R. Weinberger, C.K. Chiang, S.C. Gau, A.J. Heeger and A.G. MacDiarmid, Appl. Phys. Lett. 35 (1979) 83. J. Kanicki, P. Fedorko, S. Boue and E. Vander Donckt, Fourth EC Photovoltaic Solar Energy Conference, Stresa, Italy (Mai 1982)p. 562. B.R. Weinberger, M. Akhtar and S.C. Gau, Synth. Meth. 4 (1982) 177. D.L. Peebles, J.S. Murdoy, D.C. Weber and J. Milliken, J. Phys. C3 (1983) 44. R.R. Chamberlin, J.S. Sharman, J. Electrochem. Soc. 113 (1966) 86. M. Abd-Lefdil, Thesis, Montpellier (1984).