In ohmic contacts to n-type GaAs

In ohmic contacts to n-type GaAs

Thin Solid Films, 165 (1988) 77-82 77 ELECTRONICS AND OPTICS Au43e/In OHMIC CONTACTS TO n-TYPE GaAs BARNARD* National Institute for Materials Resea...

265KB Sizes 0 Downloads 39 Views

Thin Solid Films, 165 (1988) 77-82

77

ELECTRONICS AND OPTICS

Au43e/In OHMIC CONTACTS TO n-TYPE GaAs BARNARD* National Institute for Materials Research, Council for Scientific and Industrial Research, P.O. Box 395, Pretoria 0001 (South Africa) A. J. WILLIS Department of Mathematics, Imperial College, 180 Queen's Gate, London S W 2BZ ( U.K.) (Received December 22, 1987; accepted May 3, 1988)

W . O.

In this study structural and electrical investigations of the Au-Ge/In/GaAs system were carried out. The S shape of the log I vs. Vprofile disappeared for the contacts annealed at temperatures higher than 350 °C and correlated well with both theoretical models for a dual-phase system and the Auger electron spectroscopy results. The high ideality factors for this metallization system suggested a recombination mechanism. l. INTRODUCTION Metallization systems based on Au-Ge have been used extensively to form ohmic contacts to n-type GaAs ~-4. The alloying step of this contact system causes inherent problems. The problem is typically circumvented by adding wetting agents such as platinum 1, indium 5 and nickel 6. Although a large amount of work has been carried out on nickel as a wetting agent 6-8, only a small number of papersS'9 have reported on the influence of indium. In this paper we report on an investigation into the properties of the In/Au-Ge metallization system, using both electrical and structural characterization techniques. 2. EXPERIMENTALDETAILS Contacts were prepared on GaAs(100) silicon-doped (10 ~8 cm-3) horizontal Bridgman-grown n-type material by evaporation. Prior to deposition the samples were etched in a 10:1:1 solution of H2SO4:H202:H20 for 3 min to remove any damaged surface layer that might be present. This was followed by rinsing and etching in concentrated HCI for 90 s to remove oxides. The metallization system consisted of a 500/~ indium layer deposited by resistive heating onto the cleaned substrate, followed by a 2700/~ Au-Ge layer (88%-12% composition by weight). The deposition took place in a vacuum of better than 10 -6 Torr after which the samples were furnace annealed for 5 min in forming gas 90% N2:10% H2). The Auger analysis subsequently performed on these samples and described * Presentaddress: Departmentof Physics,Universityof Pretoria0083,SouthAfrica. 0040-6090/88/$3.50

© ElsevierSequoia/Printedin The Netherlands

78

W. O. BARNARD, A. J. WILLIS

100

Au-Ge/In/GaAs As-dep0sited 80

cO

Au. 60

cO 0¢.. 0 0 0

E

1

//

\

/

Ga

,.,,,""~s

X,, I / ' -

40 k

<

,n

".........

~ .................

,

,

2

4

0 0

.....

\\ .--"/

! 20

.............

q/

.-'k

/~'---~\

\

...~.. ..........Z....~,, ..... 7 ,

(a)

./,~,,,~..

, 6

8

10

12

I

14

16

Sputter time (rain)

100

Au-Ge/In/GaAs 350°C - 5 min 80

o~ tO

Au 60

\\

t-

O ¢O O

",

40

;.

E o

<~ 20

Ga

,,,,

//

',,,//

_ ~,Ge

In

oo~/(/

"°°'--.. . . . . . . ..°

~,~

,

..........

0 0

2

4

6

8

10

Sputter time (min)

(b)

Fig. 1

.........

(continued).

12

14

16

Au-Ge/In/n-GaAs

OHMIC CONTACTS

79

100

Au-Ge/In/GaAs 400°C - 5 min 80

t-

._o

Au

60

E

/

O tO O

._o

Ge

/

40

E .9.0 <

,,

Ga

\

/

\\

\\

," i I

f~•.. 20

.#..,,,, ..... ,, ..-" As ..." . f - - -

x\

i/

11

,,,,.

X//

......~ . ~ , , . .

-~-

.......

- - - . . . . ~ . . ~ "~,"

":...'~...:...-~.. " ~ . . ~ . . . . ' . . . . 7 " < C

........................

f 0

...,/

\\

"

0

¢'1"

;;;,-,-,,; .........

.;

--....~,x., x

i

i

i

i

L

2

4

6

8

10

(c)

\

...... \

1"2"~

L

12

14

16

Sputter time (min) 1 O0

Au-Ge/In/GaAs 4 9 5 ° C - 5 min 80

o~ 0

Au % \

60

x\

cO tO (3 0

Ga

",,,\

40

F=

/7z

.....-" /

\

0

<

\

" A

s

I./

20

"°o..,."2~....................... ~.-.-°.'Eo,." ~ 0 0

(d)

2

\_ ".

,-,

I

I

I

I

~

4

6

8

10

12

%"

I

14

16

Sputter time (min)

Fig. 1. Auger depth profiles for the A u ~ e / I n / G a A s 350 °C, (c) 400 °C and (d) 495 °C.

contact: (a) as deposited; annealed for 5 m i n at (b)

80

W. O. BARNARD, A. J. WILLIS

below was carried out with a PHI 595 scanning Auger microprobe. The samples were excited by a 3 keV, 0.5 laA electron beam. Continuous sputtering with an Ar ÷ beam at 1 keV and an ion current density of 50 ~tA c m - 2 was used to obtain the depth profiles. Room temperature current-voltage (I-V) measurements were made by probing on each sample directly after annealing. 3.

RESULTS AND DISCUSSION

The Auger depth profiles of the In/Au-Ge metallization system were obtained for samples annealed at various temperatures. The as-deposited sample and samples annealed at 350, 400 and 495 °C will be discussed. From the Auger depth profile shown in Fig. l(a) it is clear that, for the asdeposited case, the interfaces between the layers are not sharp. According to the depth profile, diffusion within the films is observed, even at room temperature, which results in the formation of an Au-In mixture. For the as-deposited samples, the substrate was not intentionally heated; however, during the evaporation process the substrate temperature could be as high as about 120°C. Room temperature diffusion of gold and indium layers was studied by Simic and Marinkovic 1° and various intermetallic compounds were identified for a range of indium contents. According to the layer thicknesses used in this study, the Au:In ratio indicates that an Au4In phase could form. (Although no evidence exists for the formation of an Au4In phase, it is assumed according to the depth profile and ref. 10 that an Auxin phase formed; for simplicity it will be referred to as an Au-In phase.) The asdeposited films had a copper-like colour in comparison with the more gold-like colour of an Au-Ge metallization layer, which is an indication that diffusion occurred within the layer. Auger electron spectroscopy (AES) spot analysis reveals no indium on the contact surface. In Fig. l(b) a depth profile of a contact annealed at 350 °C is presented. The germanium surface concentration is higher than that of the bulk alloy, while some out-diffusion of gallium and arsenic into the contact layer is also noticed. The Au-In phase still exists, but it is spread out more than in the previous case. Auger spot analysis reveals some indium on the contact surface which is an indication that some indium diffused out towards the surface. Figure l(c) shows the elemental distribution on annealing of the contact at 400 °C. The indium is now distributed throughout the contact layer while it seems as if the Au-In phase is not located close to the interface as in the previous cases. Germanium accumulation is observed at the metal-semiconductor interface. The surface of the contact was uneven and consisted of small grains which had a metallic colour. The bad surface morphology is probably caused by the absence of indium at the interface, where it would have acted as a wetting agent. The outstanding feature of the Auger depth profile for the contact annealed at 495 °C is the germanium accumulation at the interface (see Fig. l(d)). The In-Au is now evenly distributed throughout the contact with more gallium, arsenic and indium present on the surface of the contact than in the previous annealed cases. It is believed that the presence of the gold at the substrate surface enhances the outdiffusion of gallium, while the germanium can sit on the substitutional gallium sites

Au~Je/In/n-GaAs

OHMIC CONTACTS

81

in order to form an n ÷ layer, as suggested by Yoder 6. The surface morphology of the contact annealed at this temperature was also non-uniform. Figure 2 schematically illustrates the I - V and log I - V characteristics for these contacts. Instead of showing the ordinary I-V curve for the as-deposited contact, the log I vs. Vcurve is shown in Fig. 2(a). The result is an S-shaped curve. This shape has recently 11 been related to the idea of multiple phase contacts proposed by Freeouf and Woodal112. This I - V profile is retained after annealing 6f the contact at temperatures up to 350 °C (see (Fig. 2(b)), but at higher annealing temperature the S characteristic disappears and a linear plot returns (see for example Fig. 2(c)) and according to the Auger depth profile (Fig. l(c)) the Au-In phase is much more evenly distributed throughout the contact layer. It seems as if the S-shaped characteristic is only present when the Au-In phase is situated close to the interface but disappears when the phase is spread out through the contact layer. The contact became ohmic when annealed at 495 °C, as is indicated by the linear I-Vplot in Fig. 2(d). It was Au-Ge/In/GaAs 5 min

-7

Au-Ge/In/GaAs

-9,

350°C -8

-1(] J

o) o _J

C~ O -J

-9

-11

-12

,

,

,

1,0

(a)

~ 1,5

,

~

~

,

~

V (volts) -6

-10

2,0

i

i

0,6

i

i

i

i

0,9

(b)

1,2

V (volts)

Au-Ge/In/GaAs 4 0 0 ° C - 5 min

Au-Ge/In/GaAs 4 9 5 ° C - 5 min

-7

m

O3 0 ...J

-8

I

I

- 6 x 10 3

-9

'

-10

0,2

0,4

0.6

V (volts)

0,8

I

I

I

I ,~

,

I 6

-2~x 10 3I

I x

10

V (volts)

1,0

(d)

(c) Fig. 2. Schematic illustrations of the I - V and log I - V characteristics for the Au-Ge/In/GaAs system.

`3

82

w.o.

BARNARD, A. J. WILLIS

shown recently by Murakami e t al. 13 that the transition from Schottky to ohmic behaviour for an M o G e W metallization system annealed in an InAs overpressure is strongly related to the formation of InGaAs phases at the metal-GaAs interface. However, no such evidence could be found from the AES results in this study. It is also interesting to note that the ideality factors for the Au-Ge/In contacts are on the whole too high for conventional tunnelling to account for these results with, for example, n = 8.7 for the contact annealed at 400 °C (Fig. 2(c)). It has been pointed out by Borrego e t al. ~4 that tunnelling assisted by the presence of recombination centres can lead to ideality factors twice as large as those expected from conventional tunnelling. It is possible that the bad surface morphology of these contacts could well act as a source of recombination centres at the interface.l 5 4. CONCLUSION In this study structural and electrical techniques have been used to investigate the AuGe/In/GaAs system. The shape of the current-voltage characteristics and the Auger depth profiles obtained indicate the existence of a dual-phase contact. However, the S-shaped characteristics disappeared for the contacts annealed at temperatures higher than 350 °C and the resultant spreading of the phase through the contact layer. Additionally, both structural measurements and ideality factors obtained suggest a recombination mechanism at the interface arising from the presence of defect centres, possibly resulting from the bad interfacial morphology of this system. REFERENCES l 2 3 4 5 6 7 8 9 10 11 12 13 14 15

V.L. Rideout, Solid State Electron., 18 (1975) 541. G.Y. Robinson, SolidState Electron., 18 (1975) 331. N. Braslau, J. Vac. Sci. Technol., 19 (1981) 803. C.J. Palmstrom and D. V. Morgan, in M. J. Howes and D. V. Morgan (eds.), Gallium Arsenide Materials, Devices and Circuits, Wiley, Brisbane, 1985, Chapter 6. A. Christou and K. Sleger, Proc. 6th Bienn. Cornell Conf., 1977, p. 169. M.N. Yoder, Solid State Electron., 23 (1980) 117. M. Ogawa, J. Appl. Phys., 51 (1980) 406. A . P . Botha and E. Relling, MRS Symposia Proceedings, Vol. 54, Materials Research Society, Pittsburgh, PA, 1986, p. 421. A. Christou, Solid State Electron., 22 (1979) 141. V. Simic and Z. Marinkovic, Thin Solid Films, 41 (1977) 57. F. Chekir, G. N. Lu and C. Barret, Solid State Electron., 29 (1985) 519. J.L. Freeoufand J. M. Woodall, Appl. Phys. Lett., 39 (1981) 727. M. Murakami, W. H. Price, Y. Shih, N. Braslau, K. D. Childs and C. C. Parks, J. Appl. Phys., 62 (1987) 3295. J.M. Borrego, R.J. GutmannandS. Ashok, IEEETrans. Nuc. Sci.,23(1976) 1671. A. Zussman, J. Appl. Phys., 54 (1986) 3844.