APPLICATION OF CESIUM PRIMARY BEAM TO THE CHARACTERIZATION OF III.V SEMICONDUCTORS BY SIMS

HAL Id: jpa-00223940

https://hal.archives-ouvertes.fr/jpa-00223940

Submitted on 1 Jan 1984

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.

APPLICATION OF CESIUM PRIMARY BEAM TO THE CHARACTERIZATION OF III.V

SEMICONDUCTORS BY SIMS

M. Gauneau, R. Chaplain, A. Rupert

To cite this version:

M. Gauneau, R. Chaplain, A. Rupert. APPLICATION OF CESIUM PRIMARY BEAM TO THE

CHARACTERIZATION OF III.V SEMICONDUCTORS BY SIMS. Journal de Physique Colloques,

1984, 45 (C2), pp.C2-119-C2-123. �10.1051/jphyscol:1984227�. �jpa-00223940�

JOURNAL DE PHYSIQUE

Colloque C2, supplément au n"2, Tome 45, février 1984 page C2-119

APPLICATION OF CESIUM PRIMARY BEAM TO THE CHARACTERIZATION OF I I I . V SEMICONDUCTORS BY SIMS

M. Gauneau, R. Chaplain and A. Rupert

LAB ICM/MPA, CNET, B.P. 40, 22501 Lannion, France

Résumé : Ce travail décrit, d'une part, les profils en profondeur des élé- ments : C, Si, 0, S, Se, dans du phosphure d'indium non recuit, d'autre part le comportement après recuit, des dopants résiduels : Fe, Zn, dans des sub- strats implantés Si, Se, ou diffusés zinc. L'utilisation d'un faisceau d'ions césium permet d'obtenir une description précise des profils pour les éléments ayant une grande affinité électronique.

Abstract,: This paper reports depth profiles of unannealed C, Si, O, S, Se implants in indium phosphide, and describes the comportment of bulk dopants (Zn, Fe) in Si, Se- implanted, or Zn - diffused layers. The use of cesium primary beams allows to obtain well depicted profiles for these elements having a strong electron affinity.

In P is becoming under consideration for electro-optical applications such as lasers, LED'S and photo diodes which are used with low loss optical fibers in the 1.1 - 1.5pm wavelength region. Among previous papers concerning depth profiles of dopants, or residual impurities, in this material we can quote XD. OBERSTAR et al. [1-3] who studied depth distributions of Be, Mg, Fe and Si in as-implanted or annealed samples.

This paper, firstly, reports the depth profiles of unannealed C, Si, O, S, Se implants, secondly, describes the behaviour of bulk dopants (Zn, Fe) in implanted or diffused layers. Si, S and Se are of primarily importance as n-type dopants ; C, O and Si are the main residual impurities and we need to know their residual levels. Such a work has to be compared to the previous study of D.P. LETA et al. [4], but, here, we use a CAMECA IMS 3 f instrument equiped with a primary Cs ion source for elements having strong electron affinities and high negative ion yields [5].

Experimental

The samples were <100> crystals, implanted at fluences and energies ranging from 10 to 101 5 atom.cm ,and from 400 to 1000 KeV, respectively. They were tilted 7°

from the incoming ion beam to prevent any channeling effect. C,-Si, O, S, Se were analysed by using Cs+ primary ions, and recording negative secondary ions. The Cs ion beam of luA intensity, and of 14.5 KeV impact energy, was scanned over areas of about 500 urn2. The secondary ion extraction optics of the microanalyser was tuned either in the 150 pm mode (image field diameter equal to 60 um with the 750 um field aperture), or in the 25 pm mode (image field diameter equal to 10 pm with the 750 pm field aperture). To calculate, and to compare, the useful yields of the analyses, for both experimental mode, a conversion factor equal to 20 has been adopted after calculation by recording the same low dose profile in each case. Due to better focus conditions at the cross over, with the 25 pm transfer optics, the secondary ion beam intensity is not divided according to the ratio of the analysed areas. In all circumtances, the energy slit was fully opened for maximum sensitivity. Zinc was also analysed with Cs primary ions but recording CsZn positive molecular ions [5] [6]. Iron was determined by bombarding the sample with positive oxygen primary ions and recording positive Fe secondary ions.

Article published online by EDP Sciences and available at http://dx.doi.org/10.1051/jphyscol:1984227

C2-120 JOURNAL DE _PHYSIQUE

The SIMS technique, performed from implanted samples, allows t o determine : i/ t h e a c t u a l maximum concentrations and t h e background levels of t h e profiles by setting t h e integrals of t h e depth distributions equal t o t h e implant fluences, a c c u r a t e t o within about 5 % [7] ; ii/the useful yields of t h e analyses, defined a s t h e ratios of t h e number of collected ions, for t h e implants under investigation, respect t o t h e a t o m s sputtered from t h e a c t u a l sampling areas.

The X-axes w e r e calibrated by measuring t h e depths of the c r a t e r s with a stylus type s u r f a c e profiler a f t e r analyses. Anneals w e r e carried o u t in a 0.4 % P H +K ambient,

a t 750°c f o r 10 minutes. 3 2

Results and discussion

The useful yields calculated either f o r mono a t o m i c ions or for molecular ion species a r e displayed in table 1 and 2.

( MONITORED

i

i

i

( CpS(b) . LIMIT ATOM-~II-~

b-:

.

.

( InP : C : 1 3 ~ : 7.5 1W3 : 250

.*

( : Si : 2asi- : 1.6 10-3 : 80 : 2: 1015

( : 0 : 160- : 9.5

i

( 32r : '0.7 35-: 2 . 1 0 1 5

(

i i

:

I

I

( : S i

I

( : S

.

( : Se

I

( a )

-

-

xl5Oclm t r a n s f e r optics, xx2511m m f e r o p t i c s ,

xxxl5Oclm t r a n s f e r optics + hi& resolution conditi?ns,

( c ) these detection l i m i t s must be m l t i p l i e d by t h e reverse of t h e so topic r a t i o . (dl Sulfur is m n i t o r e d with high resolution conditions, s o t h e useful y i e l d determination

is not very accurate.

TABLE 1 : USEFUL YIELDS AND DEl'ECTION W SFOR NEGATIVE SECONDARY IOXS

SUBSPRATE

i

i

i

i

(

- :

I

'

( (

: 2 9 ~ W (e) 1.8

'

_:

:

( 3.5 3 380 xx

:

1 s 6 3 ~ s - ~ P Z T i n t e r f e r h g ions (

(

:

:

.

;

C

( s ^{: lo'&} s : 5.1 lo-z : 120 xx : 3. 1015 )

( : Se : a 2 ~ d 1 2 (e): 2 . 8 1 0 - ~ : 5 0 m : 1 . 1 0 1 5 )

(a)-(b)-(c)-(d)-saz m k s as in t a b l e 1

(el

-

,

TABLE 2 : USEFUL YIELDS IIND DEl'ECTION LIMITS FOR SOME MOLECUIAR NEGATIE IONS

--

The background levels, in counts per second, and t h e corresponding detection - limits, in atom. ~ m - ~ , a r e also mentioned. The useful yields have been normalized for t h e s a m e experimental conditions (150 pm transfer optics with a 750 pm field aperture) while t h e background levels a r e given without any correction. Carefull examination of table 1 and 2 results suggests t h e following remarks :

I ) The useful yields for t h e 4b - 6b group elements (C, Si, 0, S, Se) a r e very high, in t h e range - lo-', several orders of magnitude higher than those calculated by D.P. LETA e t al. [4] f o r oxygen primary ion bombardment. Therefore, b e t t e r sensitivities and, consequent1 y, lower detection limits c a n b e expected. Mono a t o m i c useful yields (table 1) increase accordingly with electron affinities of t h e elements, e x c e p t for carbon.

The electron a f f i n i t i e s a r e e q u a l t o l . 2 7 , 1.39, 1.46, 2.02, 2.07 for C, Si, 0 , Se, S, respectively. In t h e c a s e of carbon i t c a n be suspected t h a t t h e useful yield calculation is not a c c u r a t e due t o t h e high background level ; t h e s a m e remark holds for oxygen.

2) The useful yields calculated for molecular ions a r e also very high (table 2) even higher f o r PC- and PSI-, AsC- and AsSi- than those found for t h e corresponding single a t o m i c ions. The molecular 4b group elements have higher useful yields than t h e 6b group ones. As a consequence, t h e molecular ions can b e monitored t o depth profiling analyses ; they don't correspond t o higher background levels and allow a s low detection limits a s single a t o m i c ions. This c a s e is well illustrated by carbon whose detection limit is in t h e range of 2 . 1 0 ' ~ a t ~ m . c m - ~ when 44 PC- ions a r e monitored (fig. 1).

3) A t l e a s t for monoatomic ions, t h e useful yields obtained from GaAs substrates a r e higher than those calculated for InP. This result is in agreement with

V.R. DELINE e t al. measurements [7] who described a general matrix e f f e c t .

As examples C , Si, S and Se depth profiles a r e shown in fig. 1-4. For carbon a well depicted profile and a lower detection limit a r e obtained when 4 4 ~ ~ - molecular

16 -

ions a r e monitored. For oxygen, 0 or 4 7 ~ 0 - profiles have not high dynamic ranges, due t o t h e background level, and t h e detection limit is not very low, in t h e range 5-6.10 16 atom. cm-3 ; experiments should be performed from 1 8 0 implants by r e g i s t e r i n g 4 9 ~ ~ - molecular ions. Sulfur, in In P, is a special c a s e due t o t h e f a c t t h a t e a c h peak, in t h e mass

spectrum, which could b e monitored t o represent t h e sulfur signal, is overshadowed by inter- fering ~ o n s either at t h e mass 32(PH-) or a t t h e inass 63(P,H-). As shown in t h e inset of fig. 3, high mass resolving power facilities of t h e CAMECA ims 3 f instrument allow t o s e p a r a t e t h e two peaks, a t mass 32, without significant loss of sensitivy.

Finally t h e detection limits, or t h e background levels, in InP, for C, Si, 0 , S, Se a r e equal t o 2 . 1 0 ~ ~ ( 2 . 1 0 ' ~ x 100/1.107), 6.10i4, 5.1016, 2.1015, 1 . 1 0 ' ~ atom. c m -3

,

respectively. These limits a r e t w o or t h r e e orders of magnitude lower than previous d a t a [41.

The detection limits in GaAs a r e not really lower than in InP e x c e p t for selenium.

The first interest of these experiments is t o determine t h e background levels of impurities in semi-conducting substrates. Secondly, a s we a r e able t o chose t h e n a t u r e of t h e primary beam t o optimize t h e secondary ion yields according t o t h e electronic e l e m e n t properties 151, i t is possible t o well know the behaviour of dopants o r bulk impurities during thermal processing. Figures 5 , 6, 7, 8 concern Fe-doped or Zn-doped InP substrates. These substrates a r e Si or Se- implanted (fig. 5, 6 , 81, or Zn-diffused (fig. 7). A f t e r anneal under Pi-Ij ambient, iron piles-up a t t h e interface between t h e damage and t h e virgin crystal.

This gettering process is due t o residual point defects. The same f e a t u r e is observed in Zn-doped, p-type, substrates (fig. 8) : zinc migrates towards t h e surface and is g e t t e r e d a t t h e limit of t h e implanted layer. If t h e layer is strongly amorphyzed, a s in t h e c a s e of a 5 . 1 0 ' ~ ~ e . c m - ~ implantation (fig. 6), t h e gettering peak position depends on t h e fluence accordingly with t h e tickness of t h e amorphized layer. In a Zn- diffused layer (fig. 71, two important results a r e evidenced : i/a f a s t diffusion process takes place and a deep zinc plateau is observed ; i d w h e r e zinc diffuses, iron a t o m s a r e depleted. These results a r e in agreement with those of J.D. OBERSTAR e t al. [2] concerning Be-implant in Fe-doped InP, and with those of P.N. FAVENNEC e t al. 191. These experiments show clearly t h a t a f t e r anneal redistributions depend on t h e n a t u r e of dopants, and t h a t inter-element reactions may occur.

In conclusion. i t has been shown t h a t SIMS detection limits could b e very low, in t h e range of 1014 - 1016 atom. cm- depending on t h e elements and on t h e choice of t h e 3 primary beam. In particular 1 3 3 ~ s + primary beam will b e used for elements having strong e l e c t r o n affinities. Accurate a f t e r anneal depth profiles should make i t possible t o understand t h e various phenomena which occur in semi-conducting materials.

C2-122 JOURNAL DE PHYSIQUE

0- 13 C- l$ LOO W b - a c- 10" LOO W

C -23 C- Ma LOO k U d...lL PC- lo" 400 keV

Flgure 1

13C . bmplanl depth profder monitored nonr : a- 1015 C. cm.' b. 1014 C. em-' c. 10'3 c cm.' momtored tons 4 4 p ~ d. l o T 3 c cm.'

3 2 ~ Implant depth profoler a- 1 0 ~ ~ s . , m - ~ b- 1014 s.cm.'

c- I O ~ ~ S . ~ ~ . ~ d. 1 0 ~ ~ S . c m

After anneal (at 150°C tor 1 0 mln) 2 8 ~ m - and 5 6 ~ e + depth profller tn Fe-doped rernl-tnrulatmg rubstrate.

SI-~mplant. 150 KeV. 8 x S , . C ~ . ~

Flgure 7

After Zn dnfurlon lat 450' for 2 H I 1 9 7 ~ s Z n and 5 6 ~ e + depth profmler

~n Fe-doped remn-nnrulattng substrate.

Flgure 2

28s1 . Implant depth proftler a. 1015 ~n.cm.2 b- 1 0 1 4 ~ 1 . c m ~ 2 e- 1 0 1 3 ~ ~ . c m . 2

"Se ,mplant depth profller a- 2.6 x 1012~e.cm 2 b- 1013 ~ e . c m

'

e- 1014 ~e.cm.2

A f t w anneal lat 750% for 10 mtn) 8 0 ~ e - and 5 6 ~ e + depth profmler tn Fe-doped remo-inrulatong substrate.

Se-omplant, 400 KeV.5 x 1 0 ~ ~ s e . c m - ~

After anneal (at 750°C tor 1 5 mnn) 1 9 7 ~ s ~ n + and 2 8 ~ ~ depth proftler tn Zn doped substrate

St omplant, 150 KeV. 5 x l o q 4 SI cm

REFERENCES :

1) J.D. OBERSTAR, K. SHICHIJO, M. KEEVEK, B.G. STEETMAN, J.E. BAKER, P. WILLIAMS, Rad. e f f e c t s , 6 1 (1982) 109

2) J.D. OBERSTAK, B.G. STREETMAN, J.E. BAKER, P. WILLIAMS.

J. Electro. chem. soc. 129 (1982) 1312

3) J.D. OBERSTAR, B.G. STREETMAN, J.E. BAKER, P. WILLIAMS ibid., 129 (1982) '1320

4)

D.P.

5) L A . STORMS, K.F. BRCWN,'J.U. STEIN, Anal. chem., 49 (1977) 2023 6) M. GAUNEAU, A. RUPERT, P.N. FAVENNEC, L. HENRY, ti. L'HARIDOh,

H.N. MIGEON, J. Microsc. spectrosc. electron., 6 (1981) 213

7) V.R. DELINE, W. KATZ, C.A. EVANS, P. WILLIAMS, Appl. phys. Lett., 33 (1978) 832

8) M. GAUNEAU, ti. L'HARIDON, A. RUPERT, M. SALVI, Nuclear Instrum. and methods, 209/210 (1983) 671

9) P.N. FAVENNEC, M. SALVI, 1. HENRY, A.M. HUBER, G. MORILLCT, Semi-insulating 111

-

MAKRAM-EBEID and BRIAN TUCK, 1952, p. 318.