SpecialIssue Paper

J o u r n a l o fE l e c t r o n i c Materials, Vol. 24, No. 5, 1995

P-Type Doping of Double Layer Mercury Cadmium Telluride for Junction Formation L.O. BUBULAC, D.D. EDWALL, S.J.C. IRVINE, E.R. GERTNER, and S.H. SHIN Rockwell International Science Center, 1049 C a m i n o Dos Rios, T h o u s a n d Oaks, CA 9 1 3 6 0 Extrinsic p-type d o p i n g of long wavelength infrared-HgCdTe double l a y e r heterostructure for p-on-n device application requires good control of the p-type dopant, regardless of the d o p i n g technique. The approach is to place the electrical j u n c t i o n a h e a d of the compositional interface, thus avoiding q u a n t u m efficiency reduction. This research addresses the As and P d o p i n g of HgCdTe by a n implant/diffusion process. The data demonstrates a n enhanced atomic diffusion process for As and P from a n ion i m p l a n t e d source, w i t h i n the single p h a s e domain, with a diffusion rate orders of m a g n i t u d e h i g h e r t h a n the rate u n d e r Hg-saturated conditions a t the same temperature. This work also reveals a new phenomenon, namely, a t r a n s i t i o n i n the enhanced diffusion of both As and P from a n exponential to a G a u s s i a n redistribution. This t r a n s i t i o nis controlled by temperature a t a g i v e n PHg" G a u s s i a n diffusion dominates a t high temperat u r e s , T >400°C. The diffusion coefficient of the G a u s s i a n mechanism decreases as the P~ increases, from Dp - 2 × 10-11 cm2/s a t Pn - 0 . 0 2 atm to Dp - 3 × 10-14 cm2/ s u n d e r ~Ig-saturated conditions ( q u a r t z ampoule) a t 440°C. The difference in the diffusion coefficients between open tube and closed tube (quartz ampoule), u n d e r nominally Hg-saturated conditions, indicates t h a t PHg is undersaturated regardless of the Hg-source proximity. The deviation of PHg from s a t u r a t i o n is estimated from the a n n e a l i n g furnace temperature profile up to a m a x i m u m of 50%. Variation of the diffusion coefficient close to Hg s a t u r a t i o n appears to be s h a r p l y dependent on the a c t u a lPH- v a l u e (example: Dp - 1 × 10-12 cm2/s in opentube a n n e a l vs De - 3 × 10-14 CIn~/S in closed tube a t nominally the same temperature, T = 440°C). Comparative anneals of As and P showed faster diffusion r a t e s for P t h a n for As in both mechanisms.

K e y words: Annealing, diffusion, H g C d T e

INTRODUCTION Extrinsic p-type d o p i n gof long wavelength H g C d T e has received a wide i n t e r e s t for double l a y e r heteros t r u c t u r e p-on-n device applications. 1-4 This structure requires a n electrical j u n c t i o n located deeper t h a n the compositional j u n c t i o n to a v o i d the formation of any valence band barriers to the t r a n s p o r t of the photogenerated m i n o r i t y carriers across the j u n c t i o n . This implies, besides a stable interface, a controlled behavior of p-type d o p a n t s in any t h e r m a l t r e a t m e n t following t h e i r introduction into the layer. One w a y to p-type dope HgCdTe is to diffuse As or P (Received October 30, 1993; revisedAugus 20, 1994)

from a n ion i m p l a n t e dsource. Diffusion by a convent i o n a l technique (closedt u b e , Hg-saturated, T >400°C) w o u l d r e q u i r e long a n n e a l times, w h i c h c a n n o t be used because the interdiffusion in the heterointerface is f a s t e r than the d o p a n t diffusion. The r a n g e of CdHginterdiffusion coefficients across the pseudobinary a t 400°C are 2 × 10-12cm2/s to 5 × 10-1~ cm2/s5 w h i l e the diffusion coefficients for As ~ and p,9 are in the low 10-14 cm2/s range. On the o t h e r h a n d , the advantage of a heterostructure i n decreasing leakage c u r r e n t s m a y be diminished if the compositional difference between the c a p l a y e r and the absorber l a y e ris very small. An enhanced diffusion process c a n be applied to form p/n p l a n a r junctions, with the p u r p o s e of consistently locating the electrical j u n c t i o n b e y o n d the g r a d e d

618

Bubulac,Edwall, Irvine, Gertner, and Shin

I O1 9

A

region of the heterointerface. Fast As diffusion by a n atomically enhanced mechanism, from a n ion imp l a n t e d source2,7 as well as a g r o w n sourc,2,4 has been demonstrated. This p a p e r addresses the p-type d o p i n g of HgCdTe, in conjunction with p-on-n j u n c t i o n formation, by a n implant/anneal process of As and p.7,8,1o The object i v e s of this work are as follows: • T o control the p-type d o p i n g concentration for j u n c t i o n formation, by f i r s t identifying the different diffusion mechanisms t h a t m a y occur. The material and fabrication parameters could be chosen to minimize or suppress the undesirable components and develop a reproducible diffusion process. • T o develop a j u n c t i o n formation m e t h o d by w h i c h the electricalj u n c t i o n in a double l a y e rH g C d T e structure w o u l d be consistently located beyond the g r a d e d region of the heterointerface.

E"

As, 3 0 0 kev, 1 x 1014cm2 3 9 0 ° C / 2 0 min 1018 ~-/PHg 0 . 0 2 a t m

1017

,~ o< e-X

8 lo16 ?

EXPERIMENT

1015

0

1019

1

2 3 4 5 D e p t h (l~m)

6

As, 3 0 0 kev, 1 x 1 0 1 4 c m2 440°C/10min

,~_~ 1018

~

1017 C o~ e-X2/4t \ D - 2 x 10"11 c ~ 2 / s

8 lo16

I

<~ 1015

Data

~

- - -

\

As

The m a t e r i a lused for the i n i t i a l experiments was a long wavelength infrared (LWIR) H g C d T e double l a y e r structure, with a n m i d d l e wavelength infrared (MWIR) c a p layer, g r o w n by metalorganic chemical v a p o r deposiont (MOCVD).11 The i n i t i a l anneals were also performed in the M O C V D reactor. The confirmation of the initial r e s u l t s and f u r t h e r s u p p o r t i n g evidence were obtained u s i n g LWIR l i q u i d p h a s e epitaxy (LPE) single l a y e r m a t e r i a l g r o w n from the Te-rich corner of the p h a s e diagram. 12 The source for d o p i n g with As and P was ion implantation. The post i m p l a n t a n n e a l i n g was performed in a n open tube s e t u p , in a modified LPE g r o w t h furnace with a high pressure of H2, - 1 0 arm.13 We chose the open tube a n n e a l i n g technique because i t is more productionoriented. The a n n e a l conditions were chosen w i t h i n the single p h a s e d o m a i n of the p h a s e diagram. For reference, we used the conventional a n n e a l in a closed tube (quartz ampoule). The PHg was estimated from P-T tables. The main analytical techniques employed were secondary ion mass spectrometry (SIMS) 14,1~ for As and P concentration profiles and the electron beam i n d u c e dc u r r e n t (EBIC) technique for j u n c t i o n d e p t h determination. Whenever indicated, the v a l u e of the diffusion coefficient, D, refers only to the Gaussian component, and it was determined from a theoretical fit to the Gaussian model down to very low concentrations, ~1014/cm3.

M o d e l

As AND P DIFFUSION PHg Close to Te-Phase Boundary

I

0

i

i

2

4

6

8

Depth (gm) Fig. 1. As-redistribution (SIMS) in post-implant anneal in open-tube (MOCVD-reactor) at PH,~0.02 atm (within the existence domain but c l o s e to Te-phase boundary) for MOCVD, double layer HgCdTe: (a) at 390°C for 20 min,and (b) at 440°C for 10 min.

Annealing ofAs-ion implanted samples u n d e r a low p a r t i a l pressure of Hg (PHg - 0 . 0 2 atm) with one atmosphere pressure of H2, in a n M O C V D reactor, has been s h o w n2 to y i e l d a n enhanced diffusion of As compared with the a n n e a l in a s a t u r a t e d Hg overpressure in a closed tube ( q u a r t z ampoule) a t the same temperature. The material on w h i c h this effect was observed was LWIR-MCT (double l a y e r structure)

P-Type Doping of Double Layer MCT for Junction Formation

619

g r o w n by MOCVD. Diffusion data up to 350°C has been published. 2 This p a p e r extends the experiments on As diffusion in undersaturated Hg pressure by u s i n gh i g h e rsample

temperatures, 400 and 440°C, and w i t h i n experimental error, the same p a r t i a l pressure of Hg; PHg ~0.02 atm. The material and the annealing technique were the s a m e . Secondary ion mass spectrometry r e s u l t s on As redistribution are presented in Fig. l a for 390°C

1019 1019

As, 300k, 1 x 1014cm'2 400°C/20 min. PHg ~ 0.2 atm

1018

A

P, 300 key, 1 x 1014cm-2 390°C/20 min PHg ~ 0.2 atm.

1018

O3 I

E

1017

O

,.. 1017

O

Coc e"x

~ 1016

L. ,,Id

c 0 C 0

<¢ 1015

C~

1016

P

!

o. 1015

\

0

1

2 3 Depth (l~m)

4 1

0

2 3 Depth (l.Lrn)

a

1019

¢?, E

e "x

1019

As, 300 kev, 1 x 1014cm'2 440°C/10min PHg ~ 0.2 atm

1018

A

1018

P, 300 kev, 1 x 10-14cm-2 440°C/10min. PHg ~ 0.2 atm.

E

0

1017

C o: e-X2/4t D ~ 3.2 x 10-12 c m 2 / s

O ¢O . n

I 1017

(9 0

o

I 0

Data

---Model I 1

---

Data Model

,4,-¢

i._ ,4,.i

=o 1016 <~ 1015

4

1016

~,..l. C ~ e"x2/4t " ~ , ~ 1.63x 10"1 cm2/s

1015

,,, I , , ,

0 0

V ~ As

2 3 Depth (gm)

4

b Fig. 2. As-redistribution (SIMS) in post-implant anneal in open-tube (MOCVD-reactor) at PHg ~0.2 atm (within the existence domain) for LPE, single layer HgCdTe: (a) at 400°C for 20 rain, and (b) at 440°C for 10 rain.

0

2

,

'

4

~

6

8

Depth (gm) b Fig. 3. P-redistribution (SIMS) in post-implant anneal in open-tube (MOCVD-reactor) at PH ~0.2 atm (within the existence domain) for LPE, single layer HgCd~'e: (a) at 390°C for 20 rain. and (b) at 440°C for 10 min.

620

Bubulac, Edwall, Irvine, Gertner, and Shin I

. . . .

P,

]

. . . .

I

I . . . . .

50KeV, l x 1 0 1 4 cm -2 435°C/2.25 hrs PHg ~ 3 atm.

1018

I

I

i

I

P, 300 KeV, lx1014 cm-2 440°C/2.25 hrs PHg ~ 3 atm.

1018

C o~e-X2/4t 3.7 x 10"13 cm2/s

E

"~'1017 O

, w

o

L,if

3xl 0"14 cm2/s

m1016 o¢0

¢-

.2

I- D a t a I I - - - M°delI

~% -'~A

?

t._

Q.

=oc° ,01

I - - - Data I Model

1015

n = 5 x 1014 cm-3 ( H a l l )

h

0

,o

P i 0

,0,,

, ,

,,

II,

1

,

I

,

~4 Xj EBIC

Fig. 5. P-redistribution (SIMS) in open tube anneal (modified LPE furnace) atPHQ~3 atm (closeto Hg-phase boundary--sample placed as close to Hg source as possible) and 440°C for 2.25 h.

I 0

I r , 2 Depth (~m)

,

,,

I ,

2 Depth (~m)

,

, ,

3

Fig. 4. P-redistribution (SIMS) in conventional anneal (closed tube; quartz ampoule) at saturated Hg-overpressure, (on Hg-phase boundary, PHg~3 atm) at 435°C for 2.25 h for LPE-single layer HgCdTe.

a n n e a l a n d F i g . lb f o r 440°C a n n e a l . A l t h o u g h t h e diffusion a t t h e s e e l e v a t e d temperatures i n c r e a s e d with t e m p e r a t u r e , t h e c h a n g e in t h e type o f diffusion was u n e x p e c t e d . T h e m e c h a n i s m o f t h e atomic enh a n c e d diffusion switches from a n e x p o n e n t i a l d e p e n d e n c e o f t h e c o n c e n t r a t i o n vs d e p t h (C ~ e-x) to a G a u s s i a n r e d i s t r i b u t i o n (C o¢ e-x2/4t). This result was confirmed on o t h e r samples with t h e same s t r u c t u r e from a different wafer. T h e t r a n s i t i o n from e x p o n e n tial to G a u s s i a n b e h a v i o r was determined b y temp e r a t u r e . G a u s s i a n atomic diffusion a p p e a r e d to pred o m i n a t e a t h i g h t e m p e r a t u r e s , T >400°C. To avoid t h e p o t e n t i a l c o m p l i c a t i o n s o f t h e double l a y e r s t r u c t u r e , such as a c h a n g e in b a n d gap, defects in t h e h e t e r o i n t e r f a c e , a n d h i g h c o n t e n t o f dislocat i o n s ( e t c h pit density, EPD ~ l o w 107/cm2), t h e diffusion b e h a v i o r was i n v e s t i g a t e d u s i n g l o w dislocation d e n s i t y (105 c m2) LPE single l a y e r s . T h e experiments o f F i g . 1 were r e p e a t e d with P~g approximately one o r d e r o f m a g n i t u d e higher, b u t still b e l o w t h e Hg rich p h a s e b o u n d a r y , -0.2 atm., on LPE single layers u s i n g two different dopants, As a n d P. B o t h d o p a n t s were i n t r o d u c e d in t h e l a y e r via i o n implantation a t a f l u e n c e o f 1 × 1014/cm2 with a n e n e r g y o f 300 KeV. S e c o n d a r y i o n m a s s s p e c t r o m e t r y r e d i s t r i b u t i o n profiles f o r b o t h As a n d P o b t a i n e d a t 400 a n d 440°C are s h o w n in F i g . 2 a n d Fig. 3, respectively. T h e s e d a t a confirm t h e p r e v i o u s r e s u l t s r e g a r d l e s s o f t h e differ-

ence in t h e g r o w t h t e c h n i q u e , t h e s t r u c t u r a l d e f e c t d e n s i t y o f t h e material, a n d t h e implanted species. A t r a n s i t i o n in t h e diffusion m e c h a n i s m o c c u r s w h i c h is c o n t r o l l e d by t e m p e r a t u r e . T h e e x p o n e n t i a l atomic diffusion p r e d o m i n a t e s a t 400°C a n d G a u s s i a n atomic diffusion a p p e a r s t o p r e d o m i n a t e , o r p e r h a p s is init i a t e d a t h i g h t e m p e r a t u r e , T >400°C

PHgClose to Hg-Phase Boundary A n n e a l i n g a t e l e v a t e d P H g , close t o t h e H g - p h a s e b o u n d a r y was explored in parallel f o r As a n d P in a n open t u b e a p p a r a t u s .

Closed- Tube Anneal F o r c o m p a r i s o n to o u r open t u b e m e t h o d , we h a v e determined t h e P-diffusion coefficient u s i n g a convent i o n a l closed-tube a n n e a l in s a t u r a t e d Hg v a p o r . Figure 4 s h o w s r e s u l t s o f P a n n e a l e d in a q u a r t z a m p o u l e . T h e implant was performed a t 50 KeV with a f l u e n c e o f 1 × 1014/cm2. T h e p o s t - i m p l a n t a n n e a l was performed a t 435°C a n d t h e a n n e a l i n g time was l o n g , 2.25 h , to m i n i m i z e t h e initial effects o f d a m a g e on t h e r e d i s t r i b u t i o n o f t h e implanted species. S e c o n d a r y ion m a s s s p e c t r o m e t r y profiling s h o w e d G a u s s i a n r e d i s t r i b u t i o n with a d i f f u s i o n coefficient o f Dp = 3 × 10-1t/cm2/s o r l e s s . T h e s e p r o f i l e s , a s e x p e c t e d , are s i m i l a r to t h e profiles o b t a i n e d p r e v i o u s l y f o r As, fitting well with t h e p u b l i s h e d d a t a f o r t h e volume d i f f u s i o n coefficient. 7

Open-Tube Anneal T h e closest c o n d i t i o n to H g - s a t u r a t i o n in t h e open t u b e a p p a r a t u s was with a n a n n e a l i d e n t i c a l in temperature a n d time t o t h a t used f o r t h e closed t u b e

P-Type Doping of Double Layer MCT for Junction Formation

621

experiment (440°C, 2.25 h) b u t with the sample placed as close to the Hg source as possible. This a i m e d to o b t a i n near s a t u r a t e d Hg v a p o r conditions in the open tube The estimated PH was - 3 atm The SIMS r e s u l t s on P-concentration and compositional profiles are s h o w n in Fig. 5 for a n i m p l a n t with 300KeV, and fluence of 1 x 10Wcm2. The effect of the different i m p l a n t energy from the closed tube experiment (Fig. 4) is negligible due to the long-term anneal. These data show a significant redistribution of P by a classical G a u s s i a n mechanism, with a diffusion coefficient a b o u t a n o r d e r of magnitude h i g h e r than the corresponding closed tube anneal, i.e., D -3.7 x 10-1~ cm2/s vs 3 x 10-14 cm2/s. There is a s t r i k i n g similarity of P-redistribution d u r i n g this a n n e a l with the previous Gaussian enhanced diffusion obtained a t low PHg but a t the same temperature (T > 400°C). This suggests t h a t the Pn_ conditions are undersaturated regardless of the proximity of the Hg source, causing a n enhanced G a u s s i a n diffusion to occur• This also suggests that a t very s m a l l variations of PHg, in the proximity of the s a t u r a t i o n (closed tube vs open tube), the diffusion coefficient varies s h a r p l y with the a c t u a lv a l u e of PHg" The electrical activity of P in this a n n e a lwas determined to be p-type by u s i n g the EBIC technique t h r o u g h j u n c t i o n d e p t h determination. The electrical j u n c t i o n was located, w i t h i n experimental errors, w h e r e Hall-determined n-type background e q u a l e d the P-SIMS concentration p r o f i l e , s u g g e s t i n g a h i g h a c t i v a t i o n efficiency, -100%. We then compared As and P a t different a n n e a l conditions• T o avoid variations in P_ from r u n to run, two samples were implanted u n d e r mmilar conditions (50 KeV and 10Wcm2), one with As and a n o t h e r one with P and were annealed in the same open tube. PHg and a n n e a l temperature were a b o u t the same as in the experiment s h o w n in Fig. 5 (PHg - 3 atm, 430°C), b u t the a n n e a l i n g time was shorter, 10 min. Secondary ion mass spectrometry data for the As and P redistribution are s h o w n in Fig. 6a and Fig. 6b, respectively• No noticeable redistribution of As was observed (Fig. 6a), except for the i n i t i a l fast redistrib u t i o n in a shallow r e g i o n due to the i n i t i a l l y unannealed i m p l a n t damage (which for As are heavier than for P). The As profile appears s h a r p a t a d e p t h of 0.8 ~tm. However, P showed a G a u s s i a n redistribution with a diffusion coefficient of 1 × 10-12 cm2/s, confirming, w i t h i n the experimental error, the previous result from Fig. 5. From this experiment, P appears to be a f a s t e r diffusant t h a n As a t high temperature (~430°C) and high (but not saturated) PH,. O t h e r experiments were performed a t temperat u r e s e q u a l or less than 400°C and a t lower PHg, to investigate the comparison of As and P in the atomic exponential regime. Results for As and P from ann e a l s in low P. , ( - 0 . 2 atm) and a t 390°C for 10 min are s h o w n m Fig. 7. Both As and P exhibit exponential extended components, and the diffusion rate for P was h i g h e r t h a n for As. These r e s u l t s show that for the exponential enhanced diffusion, P is also faster than •

.

g

1019 As, 5 0 kev,1 x 10"14 cm-2 4 3 0 ° C / 1 0 min ~ 3 atm.

•

.-.

1018

!

1017 Coc

~ 1016 8 <¢

As

1015

0

1

2 3 D e p t h (l~m)

4

1019 P, 50 key,1 x 1 0 - 1 4 c m"2 4 3 0 ° 0 / 1 0 min PHg ~ 3 atm.

n g

•

e"x

A ¢q

1018

E

I

o

~ p C

c 1017 O

m E

c~ e-X2/4t C ~ 1 x 10-12/cm-2/s

L -

c ¢

o cO o!

1016

I~ D a t a I I - - - M°del I

<¢ 1 015

0

1

2 3 D e p t h (gm)

4

P g

Fig. 6. Comparative redistribution profiles forAs and P (SIMS) in open tube anneal at P._ ~3.0 atm (close to Hg-phase boundary--sample placed as close to ~g source as possible) and 430°C for 10 min:(a) for As, and (b) for P.

622

Bubulac, Edwall, Irvine,Gertner, and Shin

As; the diffusion d e p t h (defined conventionally a t a concentration o f l x 10Wcm3) for P was -3.5 vs -2.0 pm for As.

The investigation of As and P behavior i n a posti m p l a n t a n n e a l reveals two k i n d s of m e t h o d s t h a t c a n be developed to place the p-on-n junctions deep into the layer, b a s e don ion implantation and a n n e a l i n g in a n open tube setup. T h u s , e a c h individual mechan i s m , exponential or Gaussian, or both in combinat i o n , c a n be used to o b t a i n a desired profile compatible with the device double l a y e r architecture and processing. An example is g i v e n in Fig. 8 for P - i o ni m p l a n t e d in a single l a y e r of LPE-grown MCT. The electrical j u n c t i o n is located on the SIMS redistribution profile ( w i t h i n experimental error) as determined by EBIC measurements. A significant observation is that in the experiment from Fig. 8, the i n i t i a lu n i f o r m compos i t i o n (single layer) is not changed (i.e., no significant H g loss). (The material composition profile was determ i n e d by the Te-SIMS profiling technique developed in Ref. 10.) The surface decrease in Te125-SIMS is shallow e n o u g h (<1000,~) to be explained by a SIMS surface effect. A low temperature Hg-vacancy a n n e a l is performed a t the end to reveal the n - t y p e backg r o u n d and thus to form the extrinsic d o p e d p-on-n junction.

1019 As, 300 kev, 1 x lC 3 9 5 ° C / 8 min PHg ~ 0.2 atm.

1018 1017

E

Co¢ e "x

o= 10 6

,~

1015

As

\ \ \

0

1

2 3 D e p t h (pm)

4

1019

03

P, 300 kev,1 x 1 0 - 1 4 c m-2 3 9 5 ° C / 8 min i PHg ~ 0.2 atm.

1018

I

E o ,--

0

1017 C~

L -

¢

o ¢:

e -x

1016

0 0I

<:~ 1015 \

0

1

2 3 D e p t h (gm)

APPLICATION OF ENHANCED DIFFUSION TO J U N C T I O N F O R M A T I O N

4

b Fig. 7. Comparative redistribution profiles forAs and P ( S I M S ) in open tube anneal at ~0.2 atm and 3 9 5 ° C for 8 rain: (a) f o r As, and (b) for P.

DISCUSSION

The presented data demonstrate a n e n h a n c e d atomic diffusion process for As and P in post-implant anneals in Hg-undersaturated conditions w i t h i n the single p h a s e solidus. A t r a n s i t i o n in the atomic enhanced mechanism from exponential to Gaussian, w i t h i n the single p h a s e region, is revealed for both As and P. The t r a n s i t i o n appears to be controlled by t e m p e r a t u r e a t any g i v e n PHg" The exponential atomic mechanism appears to predominate a t T <400°C. The diffusion rate a p p e a r s to be h i g h e r as the temperature increases. At T >400°C a transition is revealed in the diffusion mechan i s m , and the Gaussian redistribution profile predominates. The diffusion coefficient appears to decrease as the PH increases; a n example s h o w s variat i o n s a t ~440°Cgfrom D = 2 x 10-11 cm2/s a t PH - 0 . 0 2 atm to D = 3 x 10-14 cm2/s a t Hg s a t u r a t e d conditions (conventional closed tube anneal, on the Hg-phase boundary). D s h o u l d also increase with temperature a t a g i v e n PHg, as suggested by the data from the s a t u r a t e d anneals. 7 The nominally s a t u r a t e d open tube conditions for PH are in fact undersaturated regardless of the proxlmzty of the sample to the m e r c u r y source; thus D m the Gaussian diffusion obtained in the open tube anneal, with the sample as close as possible to the H g source (Fig. 5), is more t h a n one o r d e r of m a g n i t u d e h i g h e r than D in the conventional s a t u r a t e d closed tube a n n e a l (Fig. 4); D - 5 × 10-13 cm2/s vs D - 3 x 10-14 cm%. The large difference i n the diffusion coefficient of the two experiments, closed tube (Fig. 4) and open tube (Fig, 5), run in nominally similar conditions of

P-Type Doping ofDouble Layer MCT for Junction Formation

623

temperature and t i m e , suggests t h a t the diffusion coefficient of the G a u s s i a n component decreases s h a r p l y as the PH~ approaches the Hg-phase b o u n d ary. This implies that the associated p-type electrical activation m a y increase sharply, depending on site occupancy. The experiments r u n on samples with various EPD (Fig. 1 and Fig. 2) show t h a t the behavior of both components is independent of EPD for two o r d e r s of m a g n i t u d e difference (from low 105 to low 10V/cm2). The n u m b e r ofa t o m s u n d e r the G a u s s i a n enhanced profile in Fig. l b (integrated dose is 1.3 x 1014 cm-2) is close, w i t h i n experimental error, to the n u m b e r of i m p l a n t e d a t o m s (1 × 1014 cm-2). This data suggests t h a t i m p l a n t e da t o m s u n d e r g o a site transfer u n d e r low PH and t h a t all the i m p l a n t e da t o m s participate m enhanced G a u s s m n dlffusmn. The electrical activation of p - t y p e d o p a n t s appears to be dependent on the p a r t i a l pressure of rig. At one extreme of PH close to the Te-phase b o u n d a r y (PHg - 0 . 0 2 atm), an~'at a low diffusion temperature, -350°C, the electrical activation of p-type d o p a n t s in the p r e d o m i n a n t exponential redistribution is undetectable.2Therefore, the exponential component requires electrical activation to p - t y p e by a separate site t r a n s fer, w h i c h is observed a t T >415°C. 1° At the o t h e r extreme of PHg close to the Hg-phase b o u n d a r y ( h i g h PHg b u t s t i l l undersaturated, - 3 atm) and a t a high temperature, 430-450°C, the activation temperature of P in the predominant G a u s s i a n redistribution appeared to be the s a m e , w i t h i n experimental error, as the diffusion temperature. The activation efficiency is - 1 0 0 % a t s a t u r a t i o n conditions as observed a t 440°C in the open tube a n n e a l(Fig. 5) and a t 400°C and a t 450°C in closed t u b e .1° G i v e n the dependence of the activation efficiency on the PHg of c o l u m n V elements (higher as the Hg-overpressure approaches the Hg-phase boundary) the activation efficiency s h o u l d be h i g h e s t precisely on the p h a s e boundary, and it s h o u l d decrease as Png decreases. The data i m p l y t h a t a t slightly undersaturated conditions elect r i c a l activity m a y vary as much as the diffusion coefficient varies. F u r t h e r work is needed to elucidate this point. From the comparative a n n e a l s of As and P, P a p p e a r s to have a f a s t e r diffusion rate in both exponential and Gaussian enhanced diffusion mechanisms. L a c k of s t r i n g e n t control of p a r t i a l pressure of Hg (PHg) a t any step in the a n n e a l i n g procedure m a y cause a g r e a t variance in the redistribution profiles of p-type dopants, regardless of the m a t e r i a l defect structure. As a consequence of this study, two atomic enh a n c e d mechanisms c a n be u s e d , individually or in combination, to diffuse AS and P from a n ion imp l a n t e d source for p-on-n device applications. The choice of e i t h e r mechanism for j u n c t i o n formation s h o u l d d e p e n d on t h e d e v i c e a r c h i t e c t u r e . The f i n a l choice s h o u l d be b a s e d on t h e device performance. .

g

.

.

.

SUMMARY We have revealed the existence of a t r a n s i t i o n in the atomic enhanced mechanism from exponential to G a u s s i a nw i t h i n the single p h a s e region. The t r a n s i tion of the mechanism from exponential to G a u s s i a n is controlled by temperature a t any g i v e n PHg, T >400°C. The diffusion coefficient, D, of the G a u s s i a n mechanism appears to be s h a r p l y dependent on PHg close to Hg-phase b o u n d a r y (De~ 1 × 10-12cm2/s in open tube and Dp-3 × 1@14cm2/s in closed tube a t nominally similar conditions of temperature and t i m e , 440°C and 10 m i n ) . The electrical activation of As and P, ~ 100% a t saturation, m a y also vary significantly with PT close to the Hg-phase boundary. The P diffusion mechanisms a p p e a rto be f a s t e rt h a n As in any atomic enhanced mechanism. The t r a n s i t i o n process in the diffusion mechanism and the high sensitivity of the diffusion coefficient of the Gaussian component on PHg in the proximity of saturation c a n provide a n explanation for the wide r a n g e of r e s u l t s for the d o p a n t redistribution from any source. It is i m p o r t a n t to identify t h e s e phenomena in o r d e r to be able to control the atomic diffusion, and by t h a t to control the device performance. rig

ACKNOWLEDGMENT We t h a n k D. Stephenson for performing the ion implantation. We appreciate the help of Dr. G. L~x (Charles E v a n s and Associates) for performing SIMS measurements. We also t h a n k Dr. W.E. T e n n a n t and Dr. E.R. Blazejewski for t h e i r support. 1018

=

i

I

I

--

6 x 107

125Te P, 300 Kev, 1 x 1014/cm2

TM

395°C/3 min, PHg ~ 0.2 atm. 435°C/3 min, PHg ~ 3 atm. 250°C/15 hrs

1017

4 x 107

:¢

o c

.o

g

1016

g o

¢J

d.

2 x 101Scm"3 (HALL)

2 x 107

~P"

1015

Xj(EBIC) II

t,,JI| U

!

1014 i

I 1

I 2

Depth

A

IU

L,,I I

p

I 3

R~

(~m) I

Xj(EBIC) Fig. 8. P-redistribution profile (SIMS) in a combined a t o m i c enhanced diffusion togetherwith compositional profile ( S I M S - 125Te) and EBIC determined junction depth: anneal at 3 8 5 ° C / 3 min and ~0.2 atm, followed by an anneal at 4 3 5 ° C / 6 min and ~3 atm and followed by an anneal at 2 5 0 ° C / 1 5 h.

624

B u b u l a c , E d w a l l , I r v i n e , G e r t n e r , a n d Shin

REFERENCES 1. T . T u n g , M.H. Kalisher,A.P. Stevens and P.E. Herning, Mat. Res. Soc. S y m p . Proc. 90, 321 (1987). 2 . L.O. Bubulac, S.J.C. Irvine, E.R. Gertner, J . Bajaj, W.P. Lin and R . Zucca, Semicond. Sci. Technol. 8 , $270 (1993). 3 . J.M. Arias, J.G. Pasko, M . Zandian, S.H. S h i n , G.M. Williams, L.O. Bubulac, R.E. D e W a m e s and W.E. Tennant,Appl. Phys. L e t t . 62, 976 (1993). 4 . D. Chandra, M.W. Goodwin, M.C. Chen and J.A. Dodge, Proc. of the 1992 MCT Workshop, J. Electron. Mater. (8) (1993). 5 . Mei-Fan S. Tang and David A. Stevenson,Appl. Phys. L e t t . 50 (18), (4 May 1987). 6 . J.E. Falconer and H.D. Palfrey, J. Cryst. Growth 100, 275 (1990) N o r t h Holland. 7 . L.O. Bubulac, D.D. Edwall and C.R. Viswanathan, J. Vac. Sci. Technol. B 9 1695 (1991). 8 . L.O. Bubulac and C.R. Viswanathan, J . Cryst. Growth 123, 555 (1992).

9. A.V. Gorshkov, F.A. Zaitov, S.B. Shangin, G.M. Shalyapina, I.N. Petrov and I.S. Asaturova, Sov. Phys. Solid S t a t e 26 (19), October (1984). 10. L.O. Bubulac, D.S. Lo, W.E. Tennant, D.D. Edwall, J.C. Chen, J . Ratusnik, J.C. Robinson and G. Bostrup,Appl.P h y s . L e t t . 50 (22), 1 June (1987). 11. S.J.C. Irvine, E.R. Gertner, L.O. Bubulac, R.V. Gil and D.D. Edwall, Semicond. Sci. Technol. 6 (1991) C15. 12. H.R. Vydynath, J.B. Parkinson, R . S . Stapp, C.L. W e s t m a r k and J . Bajaj, M t g . IRIS Specialty Group on IR Detector, 1990. 13. S.H. S h i n , J.M. Arias, M . Zandian, J.G. Pasko and R.E. DeWames, to be published in J. Electron. Mater. (1993). 14. R.J. Blattner and C.A. Evans, Jr., Scanning Electron Microscopy/1990/1V-Sem. Inc., (Chicago: AMF O'Hare, 1990), p . 200. 15. L.O. Bubulac and C.R. Viswanathan,Appl. Phys. L e t t . 60 (2), 13 J a n u a r y (1992). 16. H.G. Vydyanath, R.C. Abbot and D.A. Nelson, L . Appl. P h y s . 54, 1323 (1983).