Preparation Of Pure And Doped Silicon Carbide By Pyrolysis Of Silane Compounds

294

J. Electroehem. Soc.: S O L I D - S T A T E S C I E N C E A N D T E C H N O L O G Y

One point commonly overlooked is the question of how to define c u r r e n t density. The phosphor crystallites are exposed to an i n s t a n t a n e o u s c u r r e n t density given by (assuming a "pill-box" c u r r e n t distribution) j

4

_

--

d~

[14]

A common error is to call j' =-

A

[15]

(where A is the raster area) the c u r r e n t density, while in actuality it is the t i m e - a v e r a g e d c u r r e n t density. This is not the value to which the phosphor responds. The instantaneous value j, or the integral ~ = f j dr, are the only m e a n i n g f u l quantities if one wishes to study phosphor physics.

Conclusion Thus we have seen that in a raster excitation experiment, m u l t i e x c i t a t i o n can cause a considerable effect. A n y e x p e r i m e n t which permits multiexcitation to occur will yield misleading results, as the c u r r e n t density of the excitation pulse will be u n k n o w n and the n u m b e r of excitation pulses per frame will v a r y as the beam current, and hence beam diameter, varies.

February 1978

It s~hould be pointed out that a r a s t e r - s c a n n i n g exp e r i m e n t w h e r e i n only raster brightness m e a s u r e m e n t s are made cannot be properly used e v e n for crude evaluation experiments, for the degree of n o n l i n e a r i t y differs for various phosphors, p r e v e n t i n g a n y comparison, even if the same excitation means (electron gun, deflection yoke, etc.) are used. Hence raster brightness m e a s u r e m e n t s are a poor w a y to characterize cathodol u m i n e s c e n t phosphors: a m u c h more m e a n i n g f u l techn i q u e is a spot brightness m e a s u r e m e n t as a function of beam excitation density on a single line scan raster, or on a stationary pulsed b e a m (meeting the criteria discussed above) as a function of c u r r e n t density.

Acknowledgment The author acknowledges helpful discussions with H. N. Hersh and L. Ozawa, the assistance of G. Ban, and the e n c o u r a g e m e n t of G. E. Weibel i n the work presented here. Manuscript submitted J u l y 15, 1977; revised m a n u script received Oct. 5, 1977. A n y discussion of this p a p e r will appear in a Discussion Section to be published in the December 1978 JOURNAL. All discussions for the December 1978 Discussion Section should be submitted by Aug. 1, 1978.

Publication costs of this article were assisted by Zenith Radio Corporation.

Preparation of Pure and Doped Silicon Carbide by Pyrolysis of Silane Compounds W. von Muench* and E. Pettenpaul Institut A fur WerkstofJkunde, Technische Universit~t, D-3000 Hannover, Germany ABSTRACT Polycrystalline #-SIC is prepared from three different silane compounds (methyl-trichlorosilane, dimethyl-dichlorosilane, and trimethyl-chlorosilane) by the van Arkel process. The influence of the growth parameters (flow rates, vapor pressures, and deposition t e m p e r a t u r e ) on the growth rate and the stoichiometry of the deposit is investigated. The most satisfactory results in terms of stoichiometry are obtained i n the 1400~176 t e m p e r a t u r e range. Doping is accomplishect by adding t r i m e t h y l - a l u m i n u m and diborane. The 6H silicon carbide crystals are grown by the sublimation technique using doped and undoped polycrystalline material. The 6H polytype of silicon carbide is considered a potentially useful material for special semiconductor devices, e.g., for l i g h t - e m i t t i n g diodes, high t e m p e r a ture rectifiers, and avalanche diodes. The 6 H silicon carbide crystals for electronic applications are produced by s u b l i m a t i o n of polycrystalline silicon carbide or e l e m e n t a r y silicon and carbon at temperatures around 2500~ (Lely process). In most cases, so far, commercial (green or black) grinding powder has been used as a feed m a t e r i a l (1-3) ; the synthesis from the elements is generally less efficient in terms of crystal size and perfection. With the grinding powder, however, a large a m o u n t of various impurities is i n t r o duced into the growth system. The resulting ( n - or p - t y p e ) : c r y s t a l s , therefore, contain compensating (acceptor or donor) impurities and also deep traps: These crystals cannot be directly used for j u n c t i o n formation in most applications. They m a y serve as a substrate material for VPE or LPE processes; some i m p u r i t y transfer (autodoping) from the substrate to * Electrochemical Society Active Member. Key words: silicon carbide, van Arkel process, methyltrichloro$ilane, dimethyl-dlchlorosilane, trimethyl-chlorosilane.

the epitaxial layer t h e n must be t a k e n into account (4). The realization of blue l i g h t - e m i t t i n g diodes with a rather high efficiency has prompted the present investigation on the preparation of polycrystalline silicon carbide as a starting material i n the Lely process (5). Previous studies on vapor-phase reactions involving silicon carbide were performed with emphasis on single crystal growth of different polytypes (6-13). Knippenberg, Kapteyns, and Hagen were the first to introduce presynthesized (CVD) silicon carbide i n the Lely process (14, 15); this m a t e r i a l was obtained with a methyl-dichlorosilane source at a low deposition rate ( 3 g / h r ) . The m a i n purpose of this study is the optimization of the growth rate, the yield, and the stoichiometry of the v a p o r - g r o w n material. This has been accomplished by proper reactor design, choice of reactants, and deposition conditions. In addition, some observations on the surface morphology are included. A m a j o r problem in the Lely process is the reproducibility of the incorporation of acceptors ( m a i n l y a l u m i n u m and boron). Previously, these dopants have been added in solid form to the (undoped) feed m a t e -

Vol. 125, No. 2

PREPARATION

rial in the L e l y process. It is difficult, however, to o b tain u n i f o r m l y d o p e d crystals in this m a n n e r , because of large differences in the v a p o r p r e s s u r e s of the dopants and the dissociation products of silicon c a r bide. This p r o b l e m can be overcome b y i n t r o d u c t i o n of dopants d u r i n g the p y r o l y s i s of the silane compounds. Doping w i t h d i b o r a n e and t r i m e t h y l - a l u m i n u m has been i n v e s t i g a t e d in this study.

Experimental

REACTION CHAMBER-

COMPOUND

B2H6/Ar

103

/[i"/

Torr

T

CH3SiHCI2(HM)

uJ 102'(0 Hs)sS i CI(M 3 / ~ / ~

/

=

The p r i n c i p l e of the van A r k e l process is shown in Fig. la. Details of t h e r e a c t i o n c h a m b e r can be seen in Fig. lb. The ( w a t e r - c o o l e d ) base p l a t e is m a d e from s i l v e r - p l a t e d copper, the c u r r e n t leads and the gas inlet (nozzles) a r e of p u r e silver. It is i m p o r t a n t t h a t the gas s t r e a m s ejected f r o m the nozzles a r e hitting both carbon , h e a t e r rods u n i f o r m l y and w i t h high velocity. This can be accomplished b y a p p r o p r i a t e shaping and a d j u s t m e n t of the nozzles. The carbon rods (6 m m diam, 150-30.0 m m length) are h e a t e d w i t h a.c. u n d e r constant voltage conditions. Exact m e a s u r e ments of t h e t e m p e r a t u r e ( b y p y r o m e t r y ) are possible o n l y in the initial stage of the g r o w t h e x p e r i m e n t , since reaction b y - p r o d u c t s are deposited at the inside of the q u a r t z bell jar. It is e s t i m a t e d t h a t the surface t e m p e r a t u r e of the silicon carbide decreases b y about 100~ d u r i n g the 6 h r g r o w t h period. The reaction c h a m b e r is c a r e f u l l y e v a c u a t e d and flushed w i t h p u r e h y d r o g e n (50 min, at least) before each run. The deposition t a k e s place in h y d r o g e n at atmospheric pressure. The c a r r i e r gas can be s a t u r a t e d w i t h silane compounds k e p t at ( t h e r m o s t a t i c a l l y controlled) t e m p e r a tures in the 13~176 range. The following compounds have been used: m e t h y l - t r i c h l o r o s i l a n e , CHsSiC13

SILANE

295

O F SiC BY P Y R O L Y S I S

EX-

(CH3)3AI/ HAUST 610H22 (PUMP)

H2

Fig. la. Apparatus for the growth of doped and undoped silicon carbide by the van Arkel process.

n-uJ

,/ / ~(~H3SiCI 3 (M1)

rr 0a. 101 < >

C H3)2SilCI 2 (M 2)

,oo/// 200

240 280 320 K 360 TEMPERATURE "

Fig. 2. Vapor pressure of silane compounds. From Landolt-Biirnstein (16).

(M1); d i m e t h y l - d i c h l o r o s i l a n e , (CH3)2SIC12 (M2); and t r i m e t h y l - c h l o r o s i l a n e , (CH~)3SiC1 (M3). The vapor pressures of these compounds are shown in Fig. 2. The use of m e t h y l - d i c h l o r o s i l a n e (H_M) a p p e a r s to be impractical, due to the high v a p o r p r e s s u r e at room temperature. Doping of p o l y c r y s t a l l i n e m a t e r i a l with b o r o n is accomplished b y adding a m e t e r e d flow of d i b o r a n e (diluted in argon) d u r i n g t h e deposition cycle. A s e p a r a t e b u b b l e r containing a 10-20% solution of t r i m e t h y l - a l u m i n u m in n - d e c a n e (C10H22) serves as the source for a l u m i n u m doping; the v a p o r p r e s s u r e of n - d e c a n e is negligible in the t e m p e r a t u r e r a n g e i n volved. The carbon rods and excess carbon (if any) are r e moved b y h e a t - t r e a t m e n t in oxygen. A n y excess of silicon is r e m o v e d b y e v a p o r a t i o n in vacuum. Thus, both an excess of carbon and an excess of silicon can be d e t e r m i n e d q u a n t i t a t i v e l y b y simple g r a v i m e t r i c means. The silicon c a r b i d e rods (doped and u n d o p e d ) are b r o k e n into l u m p s of suitable size and cleaned in h y drofluoric acid. These l u m p s serve as a feed m a t e r i a l for the modified L e l y process described p r e v i o u s l y (5, 14). The 6H silicon c a r b i d e crystals are obtained in the form of platelets with flat surface p e r p e n d i c u l a r to the c-axis. Hall and resistivity m e a s u r e m e n t s a r e p e r f o r m e d b y the van der P a u w method.

Growth of PolycrystallineSiC m

CARBON /HEATER QUARTZ~

The g r o w t h of p o l y c r y s t a l l i n e silicon carbide f r o m silane compounds can be described in t e r m s of the following reactions (17). A t low t e m p e r a t u r e s ( ~ 1200~ CH~SiC18 4- 2H2-~ Si + CH4 + 3HC1

i

(CH3)2SiC12 4- 2H2-> Si q- 2CH4+ 2HCI

NOZZLE~

(CI-I3)3SiCI4- 2H2-> Si 4- 3CH4 + HCI CH4-> C + 2H2 Si + C ~ SiC

GAS I N L E T s - -

A t high t e m p e r a t u r e s ( ~ 1800~

GAS OUTLET Fig. lb. Design of reaction chamber

2CH~SiCI3 + H2 ~ 2St + C2H2 -~- 6HC1 (CHs)2SiCl2 -F H2"-> Si + C2H2 + 2HCI 4- 2H2

296

J. E l e c t r o c h e m . Sac.: S O L I D - S T A T E

SCIENCE AND TECHNOLOGY

2(CH~)3SiCI -p H2-~ 2St + 3C2H2 -F 2HCI -F 61-12

2,

C~H~-* 2C + H2

wt?/o

I 2o z 0 rn 0

--~----o--

M1 M2 M3

~ /

/

/

0

U.l 0 X W 20 7 9 0 _..J

cO 40

I wt.% 60 1000

/

M2

20

In the i n t e r m e d i a t e t e m p e r a t u r e r a n g e b o t h methane and acetylene can be formed. The a m o u n t of excess silicon or carbon r e s u l t i n g from t h e above reactions is d e p e n d e n t on the silane composition and the g r o w t h t e m p e r a t u r e . As shown in Fig. 3, an essentially stoichiometric g r o w t h of silicon carbide is possible w i t h all t h r e e silane compounds in t h e 1450~176 t e m p e r a t u r e range. M e t h y l - t r i chlorosilane (M1) yields a r e l a t i v e l y large excess of silicon at t e m p e r a t u r e s b e l o w 1400~ The results obtained w i t h d i m e t h y l - d i c h l o r o s i l a n e (1Vi2) and t r i m e t h y l - c h l o r o s i l a n e (M3) a r e quite similar in the 1400~176 t e m p e r a t u r e range. The t e m p e r a t u r e d e p e n d e n c e of the o v e r - a l l g r o w t h r a t e is shown in Fig. 4. W i t h the process p a r a m e t e r s fixed as follows: b u b b l e r t e m p e r a t u r e , 12~ b u b b l e r flow rate, 60 l i t e r s / h r ; and total flow rate, 300 l i t e r s / h r ; one obtains the highest deposition r a t e w i t h silane M3 (15 g / h r a t 1600~ due to the high vapor p r e s s u r e of this compound (see Fig. 2). The deposition rates achieved w i t h the silane compounds M1 and M2 are quite similar (6 g / h r . a p p r o x i m a t e l y ) . These figures are valid for a simultaneous deposition on two c a r bon r o d s of 150 m m l e n g t h each. The y i e l d can be i n c r e a s e d b y 50%, a p p r o x i m a t e l y , b y using carbon rods of 300 m m length. A h i g h e r g r o w t h rate can also be achieved b y increasing the b u b b l e r flow rate, at the expense of the bypass.flow. A n e x a m p l e is shown in Fig. 4 (top curve) for silicon carbide g r o w t h using silane M2 ( b u b b l e r flow r a t e 100 l i t e r s / h r , no b y p a s s ) . F i g u r e 5 d e m o n s t r a t e s the influence of the g r o w t h t e m p e r a t u r e on the m o r p h o l o g y of silicon carbide (with excess silicon or carbon) p r o d u c e d from dim e t h y l - d i c h l o r o s i l a n e . High density m a t e r i a l with small g r a i n size a n d u n i f o r m l y d i s t r i b u t e d excess silicon is obtained at low temperatures. The g r a i n size increases with increasing g r o w t h t e m p e r a t u r e . The excess carbon which is f o r m e d at t e m p e r a t u r e s above 1600~ is concentrated m a i n l y b e t w e e n p a r -

/

// /

/ 1200 1400 1600 ~ GROWTH TEMPERATURE

1800

Fig. 3. Deviation from stoichlemetry (excess carbon or silicon) Ys. growth temperature. Arrows are pointing from the initial growth temperature to the (estimated)temperature at the end of the growth process.

1978

I

g/h

Si + C -~ SiC

February

/

w/o b y p a s s

16

"

//

b< ~2

/

~.~

(2_ co 9

M1

uJ 8

e, / 0 1200 GROWTH

1400

1600

~

1800

TEMPERATURE

Fig. 4. Temperature dependence of the over-all growth rate. (See text for other growth parameters). tides of silicon carbide. Similar results are obtained with the other silane compounds. The influence of various process parameters (silane temperature, gas flow) on the growth rate is shown in Fig. 6. As expected, there is a monotonic increase of the growth rate with increasing saturation temperature (silane vapor pressure, Fig. 6a). A higher growth rate can also be achieved by increasing the flow rate through the silane bubbler (Fig. 6b). At flow rates exceeding 60 liters/hr, however, there is only a small further increase of the growth rate. The dependence of the growth rate on the total hydrogen flow (bubbler plus bypass) is shown in Fig. 6c. There is a m a x i m u m at 300 liters/hr. Obviously, the deposition reaction will be incomplete if the gas velocity in the vicinity of the carbon rods is too high. Properties of Silicon Carbide Crystals Grown from the Van Arkel Material The 6H silicon carbide crystals have been g r o w n b y the sublimation technique ( L e l y process) using comm e r c i a l "green grit" a b r a s i v e and silicon carbide s y n thesized by the van A r k e l process. The electronic p r o p e r t i e s of these crystals are s u m m a r i z e d in Table I. Crystals g r o w n from the a b r a s i v e p o w d e r a r e u s u a l l y p - t y p e and highly compensated; the hole m o b i l i t y is in the 10-15 cmz/V sec range. A n addition of n i t r o g e n d u r i n g the L e l y process yields n - t y p e crystals w i t h a m o b i l i t y a r o u n d 120 cm2/V sec. Crystals g r o w n from undoped v a n A r k e l m a t e r i a l a r e n - t y p e , w i t h a m o b i l i t y up to 200 cm2/V sac. Doping w i t h a l u m i n u m and boron has been achieved by adding t r i m e t h y l a l u m i num and d i b o r a n e during the van A r k e l process. Using 10 and 20% solutions of (CH3)sA1 in CloHz2 ( n - d e c a n e ) , hole concentrations b e t w e e n 10 ~s and 10 is c m - 3 have been obtained. The m o b i l i t y is in the r a n g e f r o m 10 to 35 cm2/V see. The d e p e n d e n c e of the hole concentration on the p a r t i a l pressure of the dopant during the van A r k e l process is shown in Fig. 7. T h e r e is a l i n e a r r e l a t i o n -

Vol. 125, No. 2

P R E P A R A T I O N OF SiC BY P Y R O L Y S I S

297

Fig. 5. Morphology of silicon carbide grown by the van Arkel process at different temperatures (si!ane M2). (a) 1200~ (c) 1600~ (d) 1800~ (initia! temperatures).

2,

11

,2__

g/h

g/h

I I

20

I

(b) 1400~

/

lo--

16--

LLI I'-,< rr" Z12 O I.-CO O (3LU 8 - - I

/

.

./

k-, [12 0

6

CO n

~

,

i.U

FLOW RATES

TOTAL FLOW RATE 300 I / h

~

BUBBLER: 60 I/h TOTAL: 300 I/h

/

4

/ 2

./

BUBBLER TEMP. 18~

I 10 20 BUBBLER TEMP.

30 ~

Fig. 6a. Dependence of the growth rate on the saturation tern* perature (silane M2).

0

20 40 60 I/h BUBBLER FLOW RATE

t00

Fig. 6b. Dependence of the growth rate on the bubbler flow rate (silane M2).

J. E[ectrochem. Sac.: S O L I D - S T A T E S C I E N C E A N D T E C H N O L O G Y

298

February 1978

1019 12 g/h - 10

w I< Ix

8

/

/

l cm-3

/

Z 10 is 0 Ix bz OaJ 1017 z 9 0 w rr 10~ Ix < 0

0

/

f

rr

z O 6 t-.-

uJ c~

(CH3)3AI

/

Q/ /

B2H6

[]

4

B U B B L E R FLOW R A T E 60 I / h 2

0

1015 10-3

BUBBLER TEME 18~

100 200 300 I/h TOTAL FLOW RATE

10 -2 104 PARTIAL PRESSURE

Torr ----,'-

10

Fig. 7. Hole concentration in Lely-grown silicon carbide crystals vs. partial pressure of doping compound in the van Arkel process.

500 "-

Fig. 6c. Dependence of the growth rate on the total hydrogen flow rate (silane M2). ship ( a p p r o x i m a t e l y ) for the doping with aluminum. The o v e r - a l l doping efficiency is m u c h smaller t h a n unity, because of losses of a l u m i n u m during the v a n A r k e l process and during th e sublimation of silicon carbide (i.e., high v a p o r pressure of a l u m i n u m as compared to silicon carbide). The incorporation of boron exhibits a sublinear behavior. A visual i n spection of the polycrystalline m a t e r i a l indicates t h a t a large part of the boron is deposited as e l e m e n t a r y boron particles. Thus, most of the boron is p r o b a b l y lost during the oxidation and degassing cycles. Th e 77~ photoluminescence spectra of silicon carbide crystals g r o w n f r o m different types of p o l y e r y s talline m a t e r i a l are shown in Fig. 8. Crystals p r o duced from abrasive p o w d e r e x h ib it a broad emission spectrum w i t h a m a x i m u m around 640 n m and additional peaks at 505 and 535 rim. T h e emission s p e c t r u m of crystals grown f r o m undoped v a n A r k e l m a t e r i a l is d o m i n a te d by the 640 nm peak. A r e l a t i v e l y w e a k emission peak is found at 475 nm; this peak m a y be due to a " m e m o r y effect" resulting f r o m previous doping e x p e r i m e n t s w i t h aluminum. Crystals g r o w n f r o m A l - d o p e d van A r k e l m a t e r i a l exhibit a strong emission peak at 490 nm; the 640 nm peak is also present.

3

-,~ eV

~ ,

'

1

PHOTON 2.5

,

t

'f,~

', /i j

3,

>tz u.J t-..-

.'

z 0.5

ENERGY 2

. . . .

~ f \

i

/

1

. _

,

uJ > I< ,,._] u.,I Ix

.,, /

r

,

2. / \

/

i

11/i

400

/

/

/

~

~

,

\

/

,

i i

!

-~

"\

l

~./ V

/ /

0

/\ /--

"~

.

/

/'

/

'~,

/

]

'

i

\. ., X.., / I

500 600 WAVELENGTH

i ~ q q I nm 700

Fig. 8. Photoluminescence spectra (77~ of silicon carbide crystals grown from different kinds of polycrystolline material: curve 1, "green grit" commercial silicon carbide abrasive; curve 2, undoped silicon carbide produced by the van Arkel process; curve 3, AI-doped silicon carbide produced by the van Arkel process. The 640 n m p e a k is also observed for all b o r o n doped crystals. It is not clear, h o w e v e r , w h e t h e r boron

Table I. Properties of silicon carbide crystals Carrier concentration (cm-~)

Starting material

Dopant

Abrasive Abrasive van Arkel mat. van Arkel mat.

None N~* None ( CHa) aA]**

* Doping during Lely process. ** Doping during van A r k e l process.

Room temperature p n n p

= = = =

1 . . . 5 x 101~ 1...5 x 1016 1 x 10 l~ 1.5 X 10 le

High

temperature 2 2 3 6

x x x x

1 0 TM 1017 10 ~ I 0 is

Mobility sec)

(cmr

Activation energy (meV)

10... 15 120 200 35

270 70 70

270

p -- 8 x IO TM

1 • I0 TM

28

270

p = 3 X 1017 p = 1 x 10 TM

2 • i0 TM 2 . 5 x 1019

22 10

270 270

Vol. 125, No. 2

PREPARATION

i n c o r p o r a t i o n is the only reason for the occurrence of this peak; c r y s t a l defects m a y also c o n t r i b u t e to t h e 640 n m emission. The l a t t e r process is suggested b y the results obtained w i t h crystals of e x t r e m e l y high purity.

Conclusion P u r e and d o p e d p o l y c r y s t a l l i n e silicon c a r b i d e c a n be p r o d u c e d b y the v a n A r k e l process using s i l a n e compounds. A r e a s o n a b l e g r o w t h r a t e (in t h e o r d e r of 40 g / h r ) can be achieved b y p r o p e r choice of the g r o w t h conditions; a stoichiometric deposit is o b t a i n e d in the 1400~176 t e m p e r a t u r e range. The p o l y c r y s t a l l i n e SiC m a t e r i a l can serve as feed m a t e rial for t h e L e l y process, i.e., for the g r o w t h of 6H silicon c a r b i d e crystals. A s u b s t a n t i a l i m p r o v e m e n t of the electrical p r o p e r t i e s (i.e., h i g h e r c a r r i e r m o b i l i t y ) has been achieved. The process is p a r t i c u l a r l y useful for doping w i t h a l u m i n u m ; hole concentrations in the 1016-10 is cm -3 r a n g e a r e obtained r e p r o d u c i b l y in this manner. M a n u s c r i p t s u b m i t t e d J u n e 17, 1977; revised m a n u script received Sept. 7, 1977. A n y discussion of this p a p e r will a p p e a r in a Discussion Section to be p u b l i s h e d in the D e c e m b e r 1978 JOURnaL. A l l discussions for the D e c e m b e r 1978 Discussion Section should be s u b m i t t e d by Aug. 1, 1978.

Publication costs o] this article were assisted by the Institut A ]~r Werkstof]kunde der Technische Universit~Lt Hannover.

O F SiC B Y P Y R O L Y S I S

299

REFERENCES 1. J. A. Lely, Bet. Deutsch. Keram. Ges., 32, 229 (1955). 2. W. F. K n i p p e n b e r g , Philips Res. Rep., 18, 205 (1963). 3. R. M. P o t t e r and J. H. Sattele, J. Cryst. Growth, 12, 245 (1972). 4. W. von Muench, W. K u e r z i n g e r , and I. Pfaffeneder, Solid-State Electron., 19, 871 (1976). 5. W. von Muench, J. Electron. Mater., 6, 449 (1977). 6. J. T. Kendall, J. Chem. Phys., 21, 821 (1953). 7. R. F. A d a m s k y and K. M. Merz, Z. Kristallogr., 111, 350 (1959). 8. V. E. S t r a u g h a n and E. F. Mayer, in "Silicon C a r bide, J. R. O'Connor and J. Smiltens, Editors, p. 94, P e r g a m o n Press, New Y o r k (1960). 9. S. Susman, R. A. Spriggs, and H. S. Weber, in ibid., p. 94. 10. H: W. Huggins a n d C. H. Pitt, Am. Ceram. Soc. Bull., 46, 266 (1967). 11. I. B e r m a n and C. E. Ryan, J. Cryst. Growth, 9, 314 (1971). 12. A. W. C. v a n K e m e n a d e and C. F. Stemfoort, ibid., 12, 13 (1972). 13. F. Kobayashi, K. I k a w a , and K. Iwamoto, ibid., 28, 395 (1975). 14. C. 3. K a p t e y n s and W. F. K n i p p e n b e r g , ibid., 7, 20 (1970). 15. S. H. Hagen a n d C. J. K a p t e y n s , Philips Res. Rep., 25, 1 (1970). 16. " L a n d o l t - B o e r n s t e i n , Z a h l e n w e r t e und F u n k tionen," Vol. IL 2a, p. 44 S p r i n g e r - V e r l a g , Berlin (1960). 17. P. R a i - C h o u d h u r y and N. P. Formigoni, This Journal, 116, 1440 (1969).

Optical Reflectance Method for Determining the Surface Quality of Sapphire (AI O ) P. J. Zanzucchi,* M. T. Duffy,* and R. C. Alig RCA Laboratories, Princeton, New Jersey 08540 ABSTRACT Polished s a p p h i r e wafers are used as substrates for the g r o w t h of crystalline silicon. The q u a l i t y of the s u b s t r a t e surface is an i m p o r t a n t factor in d e t e r m i n i n g the c r y s t a l l i n e q u a l i t y of the silicon film and, in turn, r e l a t e s to silicon device p e r f o r m a n c e and r e l i a b i l i t y . To i m p r o v e s i l i c o n - o n - s a p p h i r e device m a n u f a c t u r e , a sensitive and n o n a e s t r u c t i v e Gptical reflectance t e c h nique has been developed to d e t e r m i n e the q u a l i t y of polished s a p p h i r e s u r faces. The correlation b e t w e e n surface d a m a g e and single or m u l t i p l e specular reflectance of s a p p h i r e in the lattice mode region, 900-300 cm -1, is reported. As a result of surface damage, the A1203 v i b r a t i o n a l modes of s a p p h i r e a r e d i s t o r t e d and the optical constants associated w i t h these modes change. To i n t e r p r e t the single and m u l t i p l e reflectance spectra the optical constants of u n d a m a g e d and d a m a g e d (1102) s a p p h i r e surfaces have been calculated from reflectance d a t a using the K r a m e r s - K r o n i g method. F r o m this analysis, the reflectance of s a p p h i r e at about 600 c m -1 is found to be v e r y sensitive to surface quality. In this spectral region m u l t i p l e reflectance can be used to m e a s u r e the surface q u a l i t y of s a p p h i r e w i t h a high degree of sensitivity.. The technique can, in principle, be used to m e a s u r e the q u a l i t y of a n y semiconductor or dielectric surface in spectral regions of high reflectance, such as regions of lattice b a n d or b a n d g a p absorption. The c r y s t a l l i n e p e r f e c t i o n of thin, e p i t a x i a l l a y e r s is s t r o n g l y influenced b y the c r y s t a l l i n e perfection of the s u b s t r a t e surface. F r o m studies of the g r o w t h of silicon films (1, 2) on silicon s u b s t r a t e s it is w e l l k n o w n t h a t defects on the s u b s t r a t e surface l e a d to crystal g r o w t h defects in the e p i t a x i a l layer. By the use of L a n g t o p o g r a p h i c techniques, M c F a r l a n e and W a n g (3) have shown that the c r y s t a l l i n e p e r f e c t i o n of I I I - V semicon* Electrochemical Society Active Member. Key words: quality control, (1~02) sapphire, multiple reflec-

tance, Kramers-Kronig method.

ductor films on s a p p h i r e or spinel s u b s t r a t e s is signific a n t l y a l t e r e d b y defects on the s u b s t r a t e surface. These defects, which often a r e not visible, are u s u a l l y scratches i n t r o d u c e d in the polishing of t h e substrate. Defects can cause ~he e p i t a x i a l film to be m i s o r i e n t e d in localized regions. As a consequence, this introduces l a t tice imperfections in the e p i t a x i a l film and these i m p e r fections a r e p a r t i c u l a r l y e v i d e n t w h e n the e p i t a x i a l film is r e l a t i v e l y thin, i.e., a m i c r o m e t e r or less in thickness. W i t h the increased use of semiconductors and dielectrics p r e p a r e d as thin e p i t a x i a l films b y