• Aucun résultat trouvé

LIGHT-INDUCED CHANGES IN a-Si : H ANALYSED BY FIELD EFFECT MEASUREMENTS

N/A
N/A
Protected

Academic year: 2021

Partager "LIGHT-INDUCED CHANGES IN a-Si : H ANALYSED BY FIELD EFFECT MEASUREMENTS"

Copied!
5
0
0

Texte intégral

(1)

HAL Id: jpa-00220945

https://hal.archives-ouvertes.fr/jpa-00220945

Submitted on 1 Jan 1981

HAL

is a multi-disciplinary open access archive for the deposit and dissemination of sci- entific research documents, whether they are pub- lished or not. The documents may come from teaching and research institutions in France or abroad, or from public or private research centers.

L’archive ouverte pluridisciplinaire

HAL, est

destinée au dépôt et à la diffusion de documents scientifiques de niveau recherche, publiés ou non, émanant des établissements d’enseignement et de recherche français ou étrangers, des laboratoires publics ou privés.

LIGHT-INDUCED CHANGES IN a-Si : H ANALYSED BY FIELD EFFECT MEASUREMENTS

T . Stoica

To cite this version:

T . Stoica. LIGHT-INDUCED CHANGES IN a-Si : H ANALYSED BY FIELD EF- FECT MEASUREMENTS. Journal de Physique Colloques, 1981, 42 (C4), pp.C4-407-C4-410.

�10.1051/jphyscol:1981488�. �jpa-00220945�

(2)

JOURNAL DE PHYSIQUE

CoZZoque C4, suppldment au nOIO, Tome 42, octobre 1981 page C4-407

LIGHT-INDUCED CHANGES IN a-Si:H ANALYSED BY FIELD EFFECT MEASUREMENTS T . Stoica

I n s t i t u t e of Physics and Materials TechnoZogy, Bucharest-MagureZe, Romania

Abstract.- Amorphous Si:H films deposited by glow discharge tech- nique are analysed by field effect in both annealed and light- changed states. The spatial distribution of potential through the sample and the state density convolution with the derivative of the Fermi function are extracted, by a new method, directly from the experimental data. It is found that, in the light-changed state, the higher conductivity activation energy associates to a density of states near the Fermi level which is lower than in the annealed state. The effect of light exposure appeares to be simi- lar to weakly "p" type doping the a-Si, without significantly changing the midgap state density.

Introduction.- The light-induced changes in a-Si:Hare well known (1) :

a thermal annealed sar~ple changes by long-time illumination in another state stable at room temperature. The light changed states have a higher activation energy and a smaller conductivity. From photoconductivity measurements it is obvious that recombination via localized centers is increased (1) after lightinduced changes, but that may be due to the simple shift of the Fermi level and without changes in the density of the localized centers, as in "p" type dopped material (2). In order to get some informations about the density of states near the Fermi level, after light induced changes, we have performed field effect measurements.

The processing of the experimental data was performed by using a new method. By this method it is possible to determine directly from the experimental data the convolution N(E,T) of the density of states N(E) with the derivative of the occupation function, and also the potential profile in the semiconducting sample. In order to evaluate the flat- band potential, the temperature dependence of the field effect has been used. The possibility to determine the state density out of field effect data is analysed through numerical calculation in (3). Our method offers an analytical basis to discuss this possibility.

1. The p r o c e s s i n g . - The well known stuc- ture for field effect method is sketched in figure 1

-

only electrons as carries are considered. For the normalized potential to the kT/e, U(X) = (EC

-

EC (x) ) /kT, we have,

where Ec, Ec(x) are the values of the mobility edge of the conduction band within the unperturbed material and near the surface ; ~ ( x ) is the electric field along the "x" direction ; k , T, e have their usual mea- ning.

Using equation 1, the equations which describe the field effect are changed from variable "x" to variable "u". Thus, with the space charge given by the expression :

Article published online by EDP Sciences and available at http://dx.doi.org/10.1051/jphyscol:1981488

(3)

JOURNAL DE PHYSIQUE

+w

P ( 1 = -w / N E [ f )

-

f

(%

-LI)]

,

the Poisson equation becomes +w

E~ (u)=z

(5)'

[du*r / N(~TU~)[~(UI-U) -5 (ul

)J

( 3 )

-w

where f)&( is the occupation Fermi function of the electron states. The thickness " a "

of the sample is assumed to be large enough so that one should consider u+O at x+d.

eamiconduc tor The equation describing the relative varia- tion of the conductance "G" of the semicon- ductor films for a certain gate potential V is :

Fig. 1 : Usual field ef-

fect structure. (V) G =

E

e

- 1

d '0

yo &!zL

E (u) du (4 where uo=u(O) and GO=G/ is related to the gate potential V through the equation : Uo=O

In this equation the first two terms represent the voltage on the insu- lator (VFB- the flat band potential), and the last the voltage on the space charge region.

In the following, we will use a gate potential shifted with the flat-band potential,

The field effect experiment gives the correlation i(v) and it is requi- red that the state density N(E) to be deduced out of it (supposing VFB known, we will come later to its determination). But what can be directly determined is the convolution N(E,T) which is closely related to the density of states :

Indeed, by writingequation 3 at u=uo and differentiating it twice with respect to ug, we may write :

If i (v) is known the functions v(u0)

,

uo (v)

,

i (uo) are determined from

which is obtained by differentiating equation 4 with respect to uo, using equations 5, 6 and finally reintegrating.

When the dependence uo(v) is known the equations 5 , 6 give the e(uO) dependence, and in conjunction with the equation 1 determine the spatial profile of the potential within the sample. But this spatial profile is not needed when it is intended to determine only the state density N(E). Through the equation 8,the dependencev(u0) relates direc tly

W(E,T)

to the experimental data for a limited range of the energy.

In principle, N ( E ) is given by the deconvolution of N(E,T), for which it is necessary an arbitrary extrapolation of N (E) or N (E,T)

,

and so

N(E) is not uniquely determined from the experimental data, even in the experimentally explored energy range. However, there are many cases of low enough temperature or relatively smooth energy dependence of N(E), where we can approximate,

(4)

For example, N[E,T) at room temperature calculated for the densitv N(E)

-

given in (4) for the a-Si:H corresponds to the approximation 10 for a11 the gap energies, except the range of energies where a minimum of N ( E ) near the conduction band-is considered. N(E) can not be deduced in the range of energies where N(E,T) i s not essentially determined by the lo- cal state densig for the same energy range.

In these cases N(E,T) behaves approximatively as where "c" is energy independent.

Up to now we have assumed a known flat band potential VFB. Further we propose an approximative method to get the flat band potential from the temperature dependence of the field effect at various constant gate vol- tages. The experiment gives the dependence of the activation energy Eo of the conductance of the sample on the corresponding "i" or "G" values at room temperature. By using equations 4-6, an expression for this ac- tivation energy can be deduced

,

if an approximative expression for the temperature dependence of ug is used,

which corresponds to the approximation 10.

Equations 9 and 12 determine Ea(i) provided E: and VFB are given.

A two points fit of the calculated curve to the experimental one, gives E

: and VFB.

Fig. 2 : The experimental data G (V-VFB) and calculated Eo ( G ) for "initial" (a)

,

annealed

(b) and light-changed states

(c) by ( A ) are represented the

experimental values for E, (G) ,

(5)

C4-410 JOURNAL DE PHYSIQUE

2. Experiment and interpretation.- The a-Si:H samples (0.6 pm) were prepared by the RF plasma decomposition of SiHq onto crystalline sili- con substrates coated with Si02 (1 ym)

.

The substrates temperature during deposition was 230°C. In order to measure the films conductance Mo electrodes were used. The field effect measurements on a-Si:H samples were performed for three states : a) "initial staten- just after depo-

sition which suffered only weak illuminations ; b) "annealed statew- at temperature 180-190°C for 2 hours ; c) "light-changed state"

-

the sample was irradiated 2 hours at room temperature with integral light of O.l~l/cm~. All the experimental results were processed through the above mentioned method. In this way the values of VFB and E,were dedu- ced.

In figure 2 experimental data are shown for G(V-VFB) and also the calculated correlations E,(G) in accordance with experimental values.

Tab. 1 : The fitted parameters of a-Si:H at T = 300°K

State E' (eV) VFB (V) a ( ~ c m ) -1 NF (e~'lcm-~) Le (m) a) initial 0.80 +5.75 1.3 x 10-9 1017 0.1 b) annealed 0.72 +2.00 5.1 x 3 x 1 0 ~ ~ 0.07 c) light-changed 0.80 +5.50 2.5 x 10-lo 1017 0.1

The table 1 gives the optimal values of E,, 0 VKBI NF, which corres- pond with experimental data in figure 2, for the t ree states of the sample. o o is the conductivity of the sample at V=VFB and NF is the average of the density of states given by approximation 10 and this average is taken over the energy range experimentally explored. This energy range was in all cases 0.2eV above the Fermi level. From the table 1 it can be seen that NF is higher for the sample in the "annea- led state" than that for the samples in the "light-changed states". The NF values correlated with the position of the Fermi level in the three states are in good agreement with the unique state distribution N(E) in the middle of the gap forthepure and doped a-Si:H (4). We can say now that state distribution N(E) in the middle of the gap is not essential- ly changed by light-induced transformations. The Fermi level shift un- der such conditions, as well as light-induced changes in photoconducti- vity (1) can be understood in a similar way as the transformations cau- sed by weak "p" type doping (2). It appears thus that light induces some acceptor states.

Our method described in this paper would permit also from the field effect data the determination of the special potential profile. In all three states this potential profile is approximately given by the for- mula V = Vo exp(-x/Le). See the table 1 for the Le values experimentaly determined. These features are very important in the solar energy ap- plications of a-Si:H describing the profile of the potential in the junctions with this material.

References

1. D.L. Staebler, C.R. Wronski, Appl. Phys. Lett. 31 (1977) 292.

2. D.A. Anderson, W.E. Spear, Phil. Mag. B

36

(197- 695.

3. N.B. Goodman, H. Fritzche, Phil. Mag. B 42 (1980) 149.

4. P.G. Le Comber, W.E. Spear, Topics in ~ p s , Phys.

2

(1979) 252.

Références

Documents relatifs

The main properties of this inner magnetospheric region are briefly recalled. The dynamical response of the plasma s phere to enhancements of geomagnetic activity

The transit of Venus in June 2012 provided a unique case study of the Venus' atmosphere transiting in front of the Sun, while at the same time ESA's Venus Express

applied field, and the corresponding charge and discharge of deep centers in the bulk, at the frequency of the applied field, contribute to the complex conductivity

Particle density in a clayey soil sample : validity of the method utilising the shrinkage properties of the sample remoulded by air pressure equipment.. Agronomie, EDP Sciences, 1989,

Although RNA density in fine processes could also be secondary to the increase in GFAP filament density linked to astrocyte reactivity, our observations of elevated mRNA density

Based on data for selected events from earthquake catalogues (time, geographical position, depth, magnitude, eventually also focal mechanism), geophysical parameters

The compressibility property of the Fermi gas in the ground state (i.e. 0 is due to the finite magnetization of Landau subbands at rational flux. An explicit and

We address the problem of uniform asymptotic approximations of the ground state η ε of the stationary Gross–Pitaevskii equation (1.2) using the Hasting–McLeod solution of