GRAIN BOUNDARY SEGREGATION IN SILICON HEAVILY DOPED WITH PHOSPHORUS AND ARSENIC

HAL Id: jpa-00221781

https://hal.archives-ouvertes.fr/jpa-00221781

Submitted on 1 Jan 1982

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.

GRAIN BOUNDARY SEGREGATION IN SILICON HEAVILY DOPED WITH PHOSPHORUS AND

ARSENIC

A. Carabelas, D. Nobili, S. Solmi

To cite this version:

A. Carabelas, D. Nobili, S. Solmi. GRAIN BOUNDARY SEGREGATION IN SILICON HEAVILY

DOPED WITH PHOSPHORUS AND ARSENIC. Journal de Physique Colloques, 1982, 43 (C1),

pp.C1-187-C1-192. �10.1051/jphyscol:1982125�. �jpa-00221781�

G R A I N BOUNDARY SEGREGATION I N S I L I C O N H E A V I L Y DOPED W I T H PHOSPHORUS AND A R S E N I C

A. Carabelas

* ,

D. Nobili and S. Solrni

CNR, I s t i t u t o LAMEL, V i a CastagnoZi, 1

-

40126 BoZogna, ItaZy

RESUME

Dans ce travail on reporte 19,determination de la ségrégation dans le joints de grains de couches de silicium lourdement dopé avec As ou P et recuit a tempéra- ture entre 700 et 1050°C. On a verifié un taux de ségrégation tres haut; le phéno- mène étant plus marqué avec l'As, probablement à cause de l'entropie en excès.La chaleur de ségrégation est 13 Kcal/mol, à peu près le même pour les deux dopants.

On discute l'applicabilité du modèle de ségrégation dans le joints des grains aux solutions concentrés.

ABSTRACT

Determinations of the grain boundary segregation in polycrystalline Silicon films heavily doped with As or P were performed in the temperature range 700-1050°C. It was verified that a significant segregation takes place; the pheno- menon is m o r e m a r k e d in the case of A s , probably due to the excess entropy term.

The heat of segregation

A h

results -13 kcal/mol, nearly the same for both dopants. The applicability

O?

the models for grain boundary segregation to con- centrated solutions is discussed.

INTRODUCTION

The electrical properties of polysilicon are markedly affected by the dopant segregation at the grain boundaries (1,Z). The occurrence of this phenomenon for P and As was recently analyzed by Mandurah et al. (3); they showed that, in spite of some crude approximations, the classical model of solute segregation (4) was applicable in the range of conditions explored.

The aim of this work is to study the grain boundary segregation of these dopants in a broader range of conditions, typical of device technology, and to verify whether the simplified model is still applicable although its approximations are justified only for dilute solutions. The fit obtained with the more realistic condition that the fraction of defects positions occupied by the solute is not negligible, is also considered.

*

Dernocritus University of Thrace, Faculty of Engineering, Xanthi, Greece

Article published online by EDP Sciences and available at http://dx.doi.org/10.1051/jphyscol:1982125

JOURNAL DE PHYSIQUE

EXPERIMENTAL

Poly-Silicon films were deposited in a low pressure chemical vapor deposition reactor ont0 thermally oxidized single crystal Silicon wafers. This process was carried out at a temperature of 57S°C, a silane pressure of 0.03 Torr, and a deposition rate of 5 nm/min. The thickness of the poly-Si films, of the order of 500 nm, was measured on each specimen with a Talystep; the doping concentration'was calculated from the implanted dose divided by the film thickness.

Arsenic and phosphorus ions with doses in the range from 5x10'~ to 5x10'~ cm-*

were implanted into the films at an energy of 150 keV. This process was performed with a Lintott III accelerator which provides a doping uniformity better than 2%.

Al1 the specimens were previously annealed at 1050°C in a nitrogen atmosphere for lh in order to recover the radiation damage, to redistribute the doping into the films, and to stabilize the grain size so that further, lower-temperature, annealing would not change the grain size appreciably.

The average grain size of each composition, after the heat treatment, was determined by transmission electron microscopy (TEM)

,

using the formula suggested by Wada et al. ( 5 ) .

The specimens were put into thermal equilibrium at different temperatures in the range of 1000-700°C, in steps of 50°C. The annealing experiments were performed in nitrogen atmosphere for times which increased as the temperature decreased, and exactly as follows: 3, 9, 27, 81, 243, 729, and 1500 hrs at 1000, 950, 900, 850, 800, 750, and 700°C respectively. We have verified that these times are sufficient to obtain the equilibrium values of carrier concentration.

The electron concentration and mobility were determined as a function of annealing temperature by resistivity and Hall effects measurements, using the Van der Pauw geometry defined with photolitographic process.

RESULTS AND DISCUSSION

The average grain sizes, after annealing at 1050°C the P and As doped poly-Si films at various implanted doses, are reported on the following table:

A t low c o n c e n t r a t i o n s t h e g r a i n s i z e i s p r a c t i c a l l y t h e same f o r P and A s doped f i l m s , w h i l e a t t h e h i g h e s t dose a marked i n c r e a s e i s observed on t h e g r a i n s i z e of t h e P doped f i l m s . T h i s behaviour i s i n agreement w i t h p r e v i o u s l y r e p o r t e d d a t a ( 6 ) .

The c a r r i e r d e n s i t y vs. a n n e a l i n g temperature f o r t h e v a r i o u s implanted d o s e s i s r e p o r t e d i n F i g s . 1 and 2 f o r P and As-doped poly-Si f i l m s , r e s p e c t i v e l y . These d a t a c l e a r l y show t h a t only a f r a c t i o n of t h e dopant i s e l e c t r i c a l l y a c t i v e , and i t s amount d e c r e a s e s with d e c r e a s i n g temperature. Furthermore, we performed, on a s e t of samples, a n n e a l i n g a t h i g h e r t e m p e r a t u r e s , and v e r i f i e d t h a t t h e s e v a l u e s a r e r e v e r s i b l e and only temperature dependent.

For t h e lower implanted doses ( c u r v e s c and d of F i g s . 1 and 21, it can be assumed t h a t t h e e l e c t r i c a l l y i n a c t i v e dopant was s e g r e g a t e d a t t h e g r a i n bounda- r i e s .

The phenomenon of t h e c a r r i e r t r a p p i n g a t t h e g r a i n boundaries (7) can be n e g l e c t e d when we r e f e r t o high dopant c o n c e n t r a t i o n s . I n f a c t , c o n s i d e r i n g a d e n s i t y of t r a p p i n g S t a t e s o f about 3 x 1012 cm-2

,

t h e d e n s i t y of c a r r i e r s t r a p p e d

1000 900

10 21 700

1 . 1 * I

PHOSPHORUS a . 9 . 4 * 1 0 ~ c r n - ~

O 6 b 0 6.0.10~crn-~

F i g . 1

-

C a r r i e r d e n s i t y a s a f u n c t i o n of a n n e a l i n g temperature f o r poly S i l i c o n f i l m s d i f f e r e n t l y doped w i t h P.

Fig. 1

-

Concentration d e s por- t e u r s en f o u n c t i o n de l a tempe- r a t u r e pour couches dopé avec P à d i f f e r e n t e s c o n c e n t r a t i o n s .

JOURNAL DE PHYSIQUE

T

[Cl

1021, XH30 900 800 700

1 8 1 8 I I J

ARSENIC

.

^{1.3 x} 1 0 ~ ' c r n - ~

:

b 6.7 . 1 0 ~ c r n - ~ -

c A 2.8 . 1 0 ' ~ c r n - ~ '

d rn 1.2 1 0 ~ ~ ~ r n - ~ -

Fig. 2

-

Carrier density as a fun- ction of annealing temperature for poly Silicon films differently doped with As.

Fig. 2

-

Concentration des por- teurs en fonction de la tempera- ture pour couches dopé avec As à

differentes concentrations.

at the grain boundaries is very small (<1 x 10 18 cm-3

) compared to the total carrier concentration. On the other hand, alternative mechanisms proposed to explain the electrically inactive dopant, like E-centers for P doped samples (8) or clusters for As doped samples are excluded considering Our recent results on the solubility and precipitation of these dopants (10-11).

We notice that the carrier concentrations of the heavily doped samples (curves a and b) coincide within the precision of the measurements and are independent of the implanted dose; this phenomenon is due to the occurrence of precipitation (10-11). Moreover the carrier densities correspond, in both cases, to the solubili- ty values previously determined on single crystal specimens (11-12). This finding confirms that the chemical potential of the dopant in the bulk lattice states is not altered by the defect states and by the segregation process.

For the observations of segregation, therefore, we have to take into account only the data of curves c and d, since carrier concentrations are in both cases below the solid solubility in the whole temperature range.

The classical treatment of segregation (13) leads to the formula:

n N - N

ALB

-

B -- - e x p ( - ~ ) N ' n - n

B B

where N is the density of bulk lattice states (in Silicon % 5 x 1 0 cm-3), ~ ~ N that B occupied by the solute B, n is the density of defect states and nB that occupied by the so1ute;Ah is the difference in entalpy between the solute in the defect state

B

n N - N

B B

AsB AhB

-

_NB ,

-

_{n - n} ^{= exp}

- _.

exp(- -)

B kT

where

AgB

=

ASB-~ASB

assumes the meaning of a free entalpy of binding of the solute B in the defect state. A negative value of

AgB

leads to grain boundary segregation.

This model involves several assumptions and approximations, mainly a single type of defect states and an ideal behaviour of the solvent, which are justified only for very dilute solutions, i.e. for NB<<N.

Furthermore, if the occupancy of both bulk and defect lattice states is neglected, we obtain the simplified very practical form:

which was used by (3). We notice that this simplified formula can be better ap- proximated by a straight line.

We report in Fig. 3 Our experimental data for As and P according to the above

Fiq. 3

-

^{nB/NB as} a function of annealinq temperature for different P or AS dopant concentrations.

Fiq. 3

-

^{n /N} en function de la temperatureB de? recuit pour di£ f e- rents concentrations de P ou d'As.

0.1 ^{1 , I I I I .}

7 8 9 10 11

IO 4~~

[.KI

simplified form. It can be noticed that in both cases the fitting with a straight line is better for the dilute solution (curve d). An increase of the solute concen- tration results in a positive curvature. The occurrence of this phenomenon, which depends on the inadequacy of the model, is clearly evidenced owing to the broader

JOURNAL DE PHYSIQUE

temperature range explored in Our experiments. Due to this effect the absolute values of the binding entaipyAhB turn out to be higher than those detennined by

(3) in a narrower temperature range.

On the other hand, if the occupancy of the defect States is considered, n becomes an adjustable parameter. It was observed that a fit, comparable to that in Fig. 3, is obtained by using a value of n which corresponds to a density of 4-5 defect positions for each lattice position on the boundary. This value is in prin- ciple acceptable but its physical significance has to be considered with caution due to the limits of the model. In addition, by using this more realistic formula, the resulting values ~ f n h ~ turn out to be higher than those reported in Fiq. 3; in the case of the more dilute solutions they are 0.57 and 0.56 eV for As and P

respectively. These binding entalpies do not differ appreciably, as it was formerly pointed out by ( 3 )

,

but the preexponential factor is higher in the case of As, if al1 other conditions are equal. This is an indication that the larger grain boun- dary segregation of As is due to the excess entropy term, which is included in the preexponential factor.

This work was partially supported by CNR, Progetto Finalizzato Chimica Fine e Secondaria.

REFERENCES

1. COWHER M.E. and SEDGWICK T.O., J. Electrochem. SOC.

119,

(1972) 1565.

2. MANDURAH M.M., SARASWAT K.C. and KAMINS T.I., Appl. Phys. Lett.

2,

⁽¹⁹⁸⁰⁾

683.

3. MANDURAH M.M., SARASWAT K.C., HELMS C.R. and KAMINS T.I., J. Appl. Phys.

2,

(1980) 5755.

4. MC LEAN D., "Grain boundaries in Metals" Oxford London (1957).

5. WADA Y., NISHIMATSU S. and HASHIMOTO N., J. Electrochem. Soc.,

127,

⁽¹⁹⁸⁰⁾

206.

SOLMI S., SEVERI M., ANGELUCCI R., BALDI L. and BILENCHI A., J. Electrochem.

Soc.

128

(1982) 1811.

BACCARANI G., RICCO' C. and SPADINI G., J. Appl. Phys.

s,

(1978) 5565.

FAIR R.B. and TÇAI J.C.C., J. Electrochem. Soc.

124,

(1977) 1102.

FAIR R.B. and WEBER G.R., J. Appl. Phys.

s,

⁽¹⁹⁷³⁾ 2 7 3 .

NOBILI D., ARMIGLIATO A., FINETTI M. and SOLMI S., J. Appl. Phys.

z,

⁽¹⁹⁸²⁾

1484.

NOBILI D., CARABELAS A., CEMTTI G. and SOLMI S., (t0 be published), see ais0 NOBILI D., 4th E.C. ~hotovoltaic Solar Energy Conference, Stresa-Italy (1981) MASETTI G., NOBILI D. and SOLMI S., in "Semiconductor Silicon" edited by H.R.

Huff and E. Shirtl (Electrichemical Society, Princeton, 1977) 648.

SWALIN R.A.

,

"Thermodynamics of Solids" J. Wiley (1971)

.