JApplPhys 54 5049

Oxidation enhanced and concentration dependent diffusions of
dopants in silicon
Satoru Matsumoto, Yutaka Ishikawa, and Tatsuya Niimi

Citation: J. Appl. Phys. 54, 5049 (1983); doi: 10.1063/1.332776
View online: http://dx.doi.org/10.1063/1.332776
View Table of Contents: http://jap.aip.org/resource/1/JAPIAU/v54/i9
Published by the American Institute of Physics.

Related Articles
Ion-beam mixing in crystalline and amorphous germanium isotope multilayers
J. Appl. Phys. 110, 093502 (2011)
Diffusion of Ag and Co in ultrafine-grained -Ti deformed by equal channel angular pressing
J. Appl. Phys. 110, 083514 (2011)
Diffusion of tin in germanium: A GGA+U approach
Appl. Phys. Lett. 99, 162103 (2011)
Ab initio calculations on Li-ion migration in Li2FeSiO4 cathode material with a P21 symmetry structure
Appl. Phys. Lett. 99, 141909 (2011)
Fluorine effect on As diffusion in Ge
J. Appl. Phys. 109, 113527 (2011)

Additional information on J. Appl. Phys.
Journal Homepage: http://jap.aip.org/
Journal Information: http://jap.aip.org/about/about_the_journal
Top downloads: http://jap.aip.org/features/most_downloaded
Information for Authors: http://jap.aip.org/authors
Downloaded 23 Nov 2011 to 203.106.62.29. Redistribution subject to AIP license or copyright; see http://jap.aip.org/about/rights_and_permissions
Oxidation enhanced and concentration dependent diffusions
of dopants in silicon
Satoru Matsumoto and Yutaka Ishikawa
Department of Electrical Engineering, Keio University, Hiyoshi, Yokohama 223, Japan
Tatsuya Niimi
Department of Applied Physics, Tokai University, Hiratsuka 259, Japan
(Received 1 February 1983; accepted for pUblication 12 May 1983)
Diffusions of impurities in silicon under oxidizing ambient and under extrinsic conditions have
been analyzed by developing the Hu's model. The general expression of diffusion coefficient,
which includes both the vacancy and interstitialcy mechanisms, is used. The temperature
dependences of a fraction ofinterstitialcy mechanism for arsenic/As' phosphorus/p , and boron/B
are obtained from the data of oxidation enhanced diffusion (OED) and oxidation induced stacking
faults. /As = 42 exp( - 0.542/kT),jp = 156 exp( - 0.666IkT),and/
B
= 860 exp( - 0.829IkT).
Time dependence of OED and the concentration dependence of the diffusion coefficient of arsenic
and boron under extrinsic conditions can be explained by this model.
PACS numbers: 66.30.Jt, 61.70.Bv, 61.70.Wp, 61.70.Bv
I. INTRODUCTION
It is well known that the diffusion coefficients ofimpur-
ities such as boron, phosphorus, and arsenic are enhanced
under oxidizing ambient, depending upon the crystal orien-
tation of the surface plane. 1-9 It is also known that stacking
faults grow during thermal oxidation of silicon, depending
upon crystal orientation.
1
0-
17
The similarity between the
phenomenon of oxidation enhanced diffusion (OED) and
that of oxidation induced stacking faults (OSE) has been
pointed out by Hu. 18 In order to explain both phenomena, he
proposed the model based on the generation of silicon self-
interstitials at the Si-SiO
z
interface due to the incomplete
oxidation of silicon. OED is considered to be attributed to
these excess silicon self-interstitials because impurities dif-
fuse via both vacancies and interstitials, i.e., dual mecha-
nism. Besides Hu's model, several models have also been
proposed. 19-21
From the study of OED, Antoniadis
22
,23 introduced a
factor of a fraction of interstitialcy mechanism and calculat-
ed the factor in phosphorus, arsenic, and antimony diffu-
sions, assuming a dual mechanism. Fair
24
25
also determined
the role of silicon self-interstitials in OED according to the
Hu's model and showed that 10-20% of impurity flux was
due to interstitialcy mechanism. Recently, Gosele et al. 26
reported the fraction of interstitialcy mechanism for some
substitutional group III and group V dopants at 1100 0c.
However, the lack of the detailed data of OED restricted the
discussion on the fraction ofinterstitialcy mechanism to the
limited diffusion temperatures and times.
Recently, Ishikawa et al. 27 investigated the time depen-
dence of OED of arsenic, phosphorus, and boron in detail. It
is found that the respective impurities have different time
dependences of OED.
In this study, we will try to apply the general equation
of the diffusion coefficient in the dual mechanism for ar-
senic, phosphorus, and boron diffusions under oxidizing and
under extrinsic conditions by developing the Hu's model.
Assuming the dual mechanism, the temperature dependence
of the fraction ofinterstitialcy mechanism, for these impuri-
ties are determined. Using these values, the time dependence
of OED and the concentration dependence of diffusion coef-
ficient under extrinsic conditions are analyzed.
II. CALCULATION OF SILICON SELF-INTERSTITIAL
CONCENTRATION
We assume that impurities in silicon diffuse via both
vacancies and silicon self-interstitials as proposed by HU.
18
In such case, the flux of impurity can be expressed by the
superposition of vacancy and interstitialcy mechanisms. It
should be noted that this interstitialcy mechanism is differ-
ent from the so-called interstitial mechanism. In the intersti-
tial mechanism, an impurity atom on the interstitial site can
diffuse by hopping from one such site to another. On the
other hand, in the interstitialcy mechanism, an impurity
atom can diffuse by the following process.
28
A silicon self-
insterstitial next to a substitutional impurity atom displaces
its impurity atom into an interstitial position and occupies
the lattice site. The displaced impurity atom subsequently
pushes out a silicon atom on a neighboring site into an inter-
stitial site and occupies the lattice site. Thus, the impurity
atom moves by one interatomic distance.
The intrinsic diffusion coefficient of impurity D * is giv-
en by
D*=Da +D
b
, (1)
where Da and Db are the diffusion coefficients of impurity
related to vacancy and interstitialcy mechanisms, respec-
tively. Generally, the diffusion coefficient D can be written
22
C
v
C
j
D=D - +Db - (2)
a CifJ C
iO
'
where C
v
and C
i
are the concentrations of vacancy and sili-
con self-interstitial, respectively, and CifJ and C
jO
are the
concentrations of them at thermal equilibrium. If a fraction
of interstitialcy mechanism in the intrinsic diffusion coeffi-
cient is given by f, i.e.,f = DblD *, we have
D C
v
Cj
-=(1-/)-+/-. (3)
D* CifJ C
iO
5049 J. Appl. Phys. 54 (9). September 1983 0021-8979/83/095049-06$02.40 @ 1983 American Institute of Physics 5049
It is possible to estimate the diffusion coefficient D by
calculating the concentrations of C
v
and C
i
in various diffu-
sion conditions, such as diffusion under oxidizing ambient or
diffusion with high surface concentrations. We will apply
Eq. (3) for the calculation of time dependence of OED and
concentration dependence of diffusion coefficient under in-
ert ambient in extrinsic conditions.
According to Hu's model, the generation rate of silicon
self-interstitials Rgen is a fraction 0 of the silicon atoms oxi-
dized, that is,
o
Rgen = - (dxldt),
v
(4)
where v is the molecular volume of Si0
2
and x is the thick-
ness of Si0
2
Assuming a parabolic oxidation rate, Rgen is
expressed by
R = (OB 0.5/2v)t - 0.5
gen ,
(5)
where B is a parabolic rate constant.
As pointed out by HU,14.29 the linear relationship
between the generation rate of self-interstitials and the oxi-
dation rate cannot explain the experimental results of time
dependence of stacking faults. It is necessary to introduce
the nonlinear relationship to rationalize the experimental
results. Lin et al. 21 introduced the nonlinear relationship as a
form of (dxldt )04 from the discussion of chemical reaction
between Si and Si0
2
Recently, Tan et al.
30
proposed the
relationship with (dxl dt )05 by the consideration of the vis-
coelastic flow of Si0
2
. Taking account of above works, we
will use the relationship of (dxldt )05 in Eq. (4), instead of
(dxldt). Thus, we have
Rgen = Og (dxldt )05, (6)
v
where g is a complex equilibrium and proportionality con-
stant with a dimension of (cm
O
.
5
s - 0.5).
The rate equation for silicon self-interstitials is given by
ignoring the diffusion term, 18
dC Oa
w -' = _0_ (dxldt )05 - R (7)
dt v reg'
where w is the thickness of the wafer and Rreg is the interface
regrowth rate.
Rrcg = 1Tp aDi(C
i
- C.o), (8)
wherep is the surface density of kinks, a is the capture radius
of interstitials by kinks, and Di is the diffusion coefficient of
silicon self-interstitials. The second term of Eq. (7) shows
that excess self-interstitials generated at Si-Si0
2
interface
may be captured at surface kink.
Equation (7) can be solved numerically. It is confirmed
from the numerical calculation that the exact solution ofEq.
(7) can be approximated well by the following simple form
for t> 0.2 s as
C-C =
I '0
(9)
We will express the oxide thickness in terms of oxygen
partial pressure. The parabolic rate constant B is defined as
31
(10)
5050 J. Appl. Phys., Vol. 54, No.9, September 1983
where M is the diffusion coefficient of oxidizing species in
the oxide layer, H is the Henry's law constant, N is the num-
ber of oxidant molecules in a unit volume of oxide, and P O
2
is the oxygen partial pressure in the gas phase. For oxidation
in dry oxygen, M is given by the expression
32
of
2.88 X 10-
4
exp( - 1.171kT) andNis to be 2.3 X 10
22
cm-
3
.
From Eq. (10), the oxide thickness in the regime of parabolic
relationship is given by
x=(2MHP0
2
tIN)05. (11)
Using Eq. (11), Eq. (9) can be written as
KPO 0.25
C - C = 2 t -0.25 (12)
, '0 2
0
.
5
1Tp aDiv
where K = Og(2MH IN)0.25.
On the other hand, the rate of growth of stacking faults
is given by
dr 2 C C
dt = 1Ta Di ( i - .0)' (13)
where r is the radius of the stacking faults. By substituting
Eq. (12) into Eq. (13) and integrating, we have
4aKPO 0.25
r = 2 to.
75
(14)
3.2
0
.
5
pv
The value of K can be determined by comparing Eq. (14)
with the experimental data of the length of stacking faults. In
this case, experimental data within the temperature region
where the retrogrowth effect is negligible should be used
since Eq. (13) does not include the effect of retrogrowth of
stacking faults.
The time dependence of the growth of stacking faults
has been investigated intensively. It is known that the length
of stacking faults has a relationship of t m where m varies
between 0.5 and 1.0 (Ref. 33). In this work, we calculate the
value of K from one set of experimental data of80-min oxida-
tion in 100% dry O
2
by Murarka.
17
Thus, we have
K = 1.10X 10-
1
exp( - 2.29IkT). (15)
In these calculations, values of p = 10
12
cm-
2
,
a = 3.85 X 10-
8
cm, and v = 2x 10-
23
cm
3
are taken. IS We
also use the Leroy's estimation for the diffusion coefficient
Di and the equilibrium concentration of self-interstitia Is C.o.
That is, Di = 2.13 exp( - 2.0IkT) and C.o = 4.76x 10
25
exp( - 3.02lkT). 20 The present value of K is available for the
growth of OSF in dry O
2
ambient on (100) silicon. K will
generally depend on the orientation of the wafer and oxidiz-
ing ambient.
By using Eq. (14) with the now known K, the length of
stacking faults can be calculated for different oxidation
times and partial pressure of oxygen. An example of thus
calculated results of the length of stacking faults is shown in
Fig. 1 as a function of reciprocal absolute temperatures. In
Fig. 2, the calculations of the length of stacking faults are
shown as a function of partial pressure of oxygen as solid
lines. In both cases, the solid lines fit well with the experi-
mental data for all temperatures, times and pressures.
It is convenient to express C
i
and r as an numerical form
by substituting Eq. (15) into Eqs. (12) and (14), respectively.
We have
Matsumoto, Ishikawa, and Niimi 5050
1 0 2 ~ - - - - - - - - - - - - - - - - - - - - - - ~
100
%
dry02
o 3hr Hu
'4
A80min} 17
[] 30min Murarka
- Calcula tOd
7.0 4 17.5 8.0
10 / T (K- )
FIG. I. Temperature dependence oflength of stacking faults in 100% oxy-
gen for different oxidation times.
C;lC
iO
= 1 + 3.2X to-
IO
exp(2.73IkT)P0
2
025
t -0.25,
(16)
(17)
Equation (16) has a same expression with that derived
by Antoniadis,22 though there is a difference of about factor
2 in the estimation ofC;lC
iO
Equation (17) has also essen-
tially same dependence of time and partial pressure with Eq.
(30) of Ref. 20 except a somewhat different activation energy
and the preexponential factor.
III. TIME DEPENDENCE OF OED
We consider the diffusion of impurity under oxidizing
ambient in intrinsic conditions in which the concentration of
impurity is below the intrinsic carrier concentration at the
diffusion temperature. Previously, we
34
calculated the diffu-
sion coefficients under oxidizing ambient on the assumption
that the concentrations of self-interstitials were increased
and the concentrations of vacancies were maintained at ther-
mal equilibrium values during thermal oxidation of silicon.
This idea can explain the oxidation enhanced diffusion but
~ 1 0
E
::J..
lJ..
(/)
lJ.. v
o
I
t-
(!)
Z
W
-.J
- CalculatOd
[J 1050t} 16
v 1100"(: Murarka
<> 1150C
B
10 100
PERCENT OXYGEN(%)
FIG. 2. Partial pressure dependence oflength of stacking faults for SO-min
oxidation.
cannot explain the oxidation retarded diffusion of antimo-
ny.35
It is rational to assume that the dynamical equilibrium
between vacancies and silicon self-interstitials is estab-
lished. 19,22,36-38 That is,
CvC; = CvaC
iO
(18)
Using Eq. (18), the diffusion coefficient of Eq. (3) can be re-
written as
D C
iO
C
i
- =(1-1)- +1-
D* C
i
C
iO
(19)
The effective diffusion coefficient in a specified oxidation
time (D ) is given by
ll'
(D) = - Ddt.
t 0
(20)
By substituting Eqs. (12) and (19) into Eq. (20), D can be
calculated numerically if the diffusion times and tempera-
tures are given.
Values ofl for arsenic, phosphorus, and boron can be
estimated by comparing Eq. (20) with the experimental data
of the diffusion coefficient obtained under oxidizing ambient
in intrinsic conditions with the known value of K determined
in the previous section. Thus, values ofl for arsenic, phos-
phorus, and boron determined from one set of experimental
data for fixed diffusion time by Ishikawa et al.
27
are plotted
in Fig. 3 as a function of the reciprocal absolute tempera-
tures. From the plots in Fig. 3, the temperature dependence
of I for arsenic, phosphorus, and boron are expressed by,
respectively,
lAs = 42 exp( - 0.542IkT),
Ip = 156 exp( - 0.666IkT),
IB = 860 exp( - 0.829IkT).
(2Ia)
(2Ib)
(2Ic)
The contribution of interstitialcy mechanism increases as
the temperature increases for all impurities in this work. The
degree of interstitialcy mechanism is largest in boron diffu-
sion.
Fair
24
estimated 10--20% fraction of interstitialcy
mechanism based on the Hu's model, though the diffusion
coefficients were not averaged by diffusion time. Antonia-
dis
23
also obtained the interstitialcy fraction for phosphorus
TEMPERATURE (oC)
1100 1000 900
1.0 ,...------,:....:.....-----.------'-'=T:....:.....----r-----=-r=,
0.1
7.0
o ARSENIC
A PHOSPHORUS
[J BORON
7.5 4 1 8.0
10/T(K- )
8.5
FIG. 3. Temperature dependence of a fraction of interstitiaIcy diffusion of
arsenic, phosphorus, and boron.
diffusion by using a nonlinear relationship of oxide growth
rate. His estimation of/at 1100 C for phosphorus is very
similar to that of the present work.
Thus, the diffusion coefficients of arsenic, phosphorus,
and boron can be calculated by using Eq. (20) with values of
K and/. which are given by Eqs. (15) and (21), respectively,
for different diffusion times. Solid curves in Fig. 4 show the
calculations of the diffusion coefficients of arsenic as a func-
tion of diffusion time with the experimental plots obtained
by Ishikawa et al. In this figure, data obtained by Antoniadis
et al.
4
are also shown. Calculated results fit well with the
experimental data for all diffusion temperatures and times
performed. Figures 5 and 6 also show the calculations and
the experimental data for phosphorus and boron, respective-
ly. At the estimation of (D )/D * in Figs. 4, 5, and 6, the
following intrinsic diffusion coefficients of arsenic, 27,39 phos-
phorus,27,40 and boron 27,41 were used.
Ishikawa et al.
27
gave the empirical equations of the
diffusion coefficients of arsenic and phosphorus as a func-
tion of diffusion time. In these equations, time dependence of
OED was expressed as a form of t m. However, time depen-
dence of OED is a complex function of/as shown in Eq. (20).
The different time dependence of OED for respective impur-
ities originates in the different temperature dependence of/
for respective impurities, Superficially, the diffusion coeffi-
cients of arsenic and phosphorus can be approximated by the
empirical equations given by Ishikawa et al.
IV. CONCENTRATION DEPENDENCE OF DIFFUSION
COEFFICIENT IN EXTRINSIC CONDITIONS
Next, we consider the diffusion of impurity under inert
ambient in extrinsic conditions where the doping concentra-
tions exceed the intrinsic carrier concentration at the process
temperature, n i' In this case, we will also utilize the general
equation of Eq. (3), assuming a dual mechanism.
(18) of dynamical equilibrium, Eq. (3) can be rewntten In
terms of the vacancy concentration as
*'
o
......
'"
o
v
D C
u
C
vIl
- =(1-/)- +/-.
D* C
vIl
C
u
... 1000:C
1100 C
ARSENIC DIFFUSION
d rOo 950 }
Anton.iadiS Y 2 t::. 1000 C Ishikawa
.1 all, D 10SOC .1 01.27

Calculated
10 10
2
DIFFUSION TIME (hr)
(22)
FIG. 4. Enhancement of diffusion coefficient of arsenic as a function of
diffusion time for different temperatures.

PHOSPHORUS DIFFUSION 0 9500C}
dry 02 '" 1000:C IShlkow.a
DIOSOC .10127
V IIOOC

10 1 1
DIFFUSION TIME (hr)
FIG. 5. Enhancement of diffusion coefficient of phosphorus as a function of
diffusion time for different temperatures.
It is generally known that the negatively charged va-
cancy C v- is responsible for arsenic diffusion whereas positi-
vely charged vacancy C v+ is responsible for boron diffusion
under inert ambient in extrinsic conditions. Considering the
Fermi-level effect, negatively and positively charged va-
cancy concentrations in extrinsic conditions are expressed
by, respectively,42.43
(23)
C v+ = C vi L, (24)
n
i
where C -: and C + are the concentrations of C v- and C u+
in nand p are electron and hole concen-
trations, respectively. In addition to the Fermi-level effect,
the electric-field enhancement effect is also included in im-
purity diffusion in extrinsic conditions. The field enhance-
ment factor h is given by42

I
*01
......
'" o
v
BORON DI FFUSION
dry O
2

o 950C
D Ishl kawa
v VII oooe et al.
27
... 1000
0
e 8
.1050oe Lin et al
Calculated
16
1

DIFFUSION TIME(hr)
FIG. 6. Enhancement of diffusion coefficient of boron as a function of diffu
sion time for different temperatures.
h = 1 + [1 + 4(n;lC)2] -0.5, (25)
where C is the impurity concentration.
By substituting Eqs. (23) and (24) into Eq. (22), we have
D [ n n
i
]
- = (1-/)- +/- h,
D* n
i
n
(26)
for arsenic diffusion. In the case of boron diffusion, n is re-
placed by p in Eq. (26).
Then, the diffusion coefficient can be calculated by Eq.
(26) with the value off, which is determined for respective
impurities in the previous section, as a function of impurity
concentration. We use the general expression ofEq. (3) to the
case of impurity diffusion under inert ambient in extrinsic
conditions, too. It should be noted that vacancies substan-
tially dominate the impurity diffusion under inert ambient in
extrinsic conditions because of the presence of excess vacan-
cies, though the concept of a dual mechanism still holds.
Calculated results of diffusion coefficient of arsenic and
boron in silicon are shown in Figs. 7 and 8 as solid curves
with the experimental data, respectively. For arsenic diffu-
sion, calculations fit well with the experimental data of Ken-
nedy and Murley,44 Fair and Tsai,45 and Murota et al.
46
ex-
cept the concentration region above 2 X 10
20
cm - 3 where the
other effect
42
such as vacancy-arsenic complex formation
occurs. In case of boron diffusion, concentration dependence
of the diffusion coefficient obtained by the experiment be-
comes weak at higher temperature. This tendency also can
be explained well by the calculation, which comes from the
fact that interstitialcy fraction ofboron/
B
becomes large at
higher temperature. In phosphorus diffusion under inert
_ -13
11110
~
u
-
>
~
-
>
iii
::>
~
~ -14
0
10
-1
10
ARSENIC 01 FFUSION
I ~ r t Ambient
c
0
o950C }
I:;. 1 OOO"C Muroto
C 1 0 50C et 01.
45
.1000C 44
.1050C Fair&Tsai
v 1000C Kennedy
&Murl ey43
-- Calculated
,d
9
10
20
CONCENTRATION (em
3
)
FIG. 7. Concentration dependence of diffusion coefficient of arsenic in inert
ambient for different temperatures.
VI
N-
E
~
>
f-
>
-12
10
If) -13
~ 1 0
l.L.
o
-14
10
BORON DIFFUSION
Inert Ambient

Calculated
FIG. 8. Concentration dependence of diffusion coefficient of boron in inert
ambient for different temperatures.
ambient in extrinsic conditions, diffusion coefficients de-
pend on the surface concentration
47
as well as the impurity
doping concentration, and vacancies generated at the sur-
face extend inside the bulk deeply at least up to 25 flm.
48
These complicated phenomena cannot be explained by the
Fermi-level effect only. Thus, phosphorus diffusion in ex-
trinsic conditions is beyond the present work.
v. CONCLUSIONS
General equation of the diffusion coefficient, which in-
cludes both vacancy and interstitialcy mechanisms, is ap-
plied for arsenic, phosphorus, and boron diffusions in var-
ious diffusion conditions by developing the Hu's model. The
temperature dependences of a fraction of interstitialcy
mechanism for arsenic, phosphorus, and boron are obtained
from the data of OED and OSF, assuming the dual mecha-
nism. Time dependence of OED and the concentration de-
pendence of the diffusion coefficient of arsenic and boron
under inert ambient in extrinsic conditions can be explained
by this model.
ACKNOWLEDGMENT
The authors wish to express their gratitude to the
Kawakami Memorial Foundation for financial support of
this work.
'G. N. Wills, SolidState Electron. 12, 133 (1969).
2W. G. Allen and K. V. Anand, SolidState Electron. 14,397 (1971).
3G. Masetti, S. Solmi, and G. Soncini, Solid-State Electron. 16, 1419 (1973).
4D. A. Antoniadis, A. M. Lin, and R. W. Dutton, Appl. Phys. Lett. 33,
1030 (1978).
'D. A. Antonoadis, A. G. Gonzales, and R. W. Dutton, J. Electrochem.
Soc. 125, 814 (1978).
6R. Francis and D. S. Dobson, J. Appl. Phys. 50, 280 (1979).
Matsumoto. Ishikawa. and Niimi 5053
7K. Taniguchi, K. Kurosawa, and M. Kashiwagi, J. Electrochem. Soc. 127,
2243 (1980).
A. M. Lin, D. A. Antoniadis, and R. W. Dutton, J. Electrochem. Soc. 128,
1131 (1981).
"c. Hill, Silicon Semiconductor 1981, edited by H. R. Huff, R. J. Kriegler,
and Y.Takeishi (Electrochemical Society, Pennington, 1981), p. 988.
IOD. J. D. Thomas, Phys. Status Solidi 3, 226 (1963).
IIR. H. Finch, J. J. Queisser, G. Thomas, and J. Washburn, J. App\. Phys.
34, 406 (1963).
I'G. R. Booker and R. Stickler, Philos. Mag. 12, 1303 (1965).
Dc. M. Hsieh and D. M. Maher, J. App\. Phys. 44,1302 (1973).
14S. M. Hu, App\. Phys. Lett. 27,165 (1975).
1'1'. Y. Tan, L. L. Wu, and W. K. Tice, App\. Phys. Lett. 29, 765 (1976).
lOS. P. Murarka, J. App\. Phys. 48, 5020 (1977).
17S. P. Murarka, J. App\. Phys. 49, 2513 (1978).
I"S. M. Hu, J. App\. Phys. 45, 1567 (1974).
19S. P. Murarka, Phys. Rev. B 16, 2849 (1977).
2B. Leroy, J. App\. Phys. 50, 7996 (1979).
21 A. M. Lin, R. W. Dutton, D. A. Antoniadis, and W. A. Tiller, J. Electro-
chern. Soc. 128, 1121 (1981).
22D. A. Antoniadis, Semiconductor Silicon 1981, edited by H. R. Huff, R. J.
Kriegler, and Y. Takeishi (Electrochemical Society, Pennington, 1981), p.
947.
23D. A. Antoniadis and I. Moskowitz, J. App\. Phys. 53, 6788 (1982).
24R. B. Fair, J. App\. Phys. 51, 5828 (1980).
"R. B. Fair, J. Electrochem. Soc. 128,1360 (1981).
20U. Giisele and T. Y. Tan, Mater. Res. Soc. Meeting (Boston, November 1-
4,1982).
ny. Ishikawa, Y. Sakina, H. Tanaka, S. Matsumoto, and T. Niimi, J. Elec-
trochem. Soc. 129, 644 (1982).
'"A. Seeger and K. P. Chik, Phys. Status Solidi 29, 455 (1968).
29S. M. Hu, Defects in Semiconductor, edited by J. Narayan and T. Y. Tan
(North-Holland, New York, 1981), p. 333.
"'T. Y. Tan and U. Giisele, App\. Phys. Lett. 39, 86 (1981).
liA. S. Grove, Physics and Technology of Semiconductor Devices (Wiley,
New York, 1967), p. 22.
"F. J. Norton, Nature 191, 701 (1961).
"S. P. Murarka and G. Quintana, J. App\. Phys. 48, 46 (1977).
34S. Matsumoto, Y. Ishikawa, and T. Niimi, Mater. Res. Soc. Meeting (Bos-
ton, November 1-4, 1982).
35S. Mizuo and Higuchi, Jpn. J. App\. Phys. 20, 739 (1981).
l6E. Sirtl, Semiconductor Silicon 1977, edited by H. R. Huff and E. Sirtl
(Electrochemical Society, Princeton, 1977), p. 968.
nS. P. Murarka, Phys. Rev. B 21,692 (1980).
'"T. Y. Tan and U. Giisele, App\. Phys. Lett. 40, 616 (1982).
19T. L. Chiu and H. N. Ghosh, IBM J. Res. Dev. 15,472 (1971).
40S. Maekawa, J. Phys. Soc. lpn. 17, 1592 (1962).
41A. D. Kurtz and R. Yee, J. App\. Phys. 31, 303 (1960).
4'S. M. Hu, Atomic Diffusion in Semiconductors, edited by D. Shaw (Ple-
num, New York, 1973), Chap. 5.
43R. B. Fair, J. Electrochem. Soc. 122, 800 (1975).
44D. P. Kennedy and P. C. Murley, Proc. IEEE 59,335 (1971).
"R. B. Fair and J. C. C. Tsai, J. Electrochem. Soc. 122, 168911975).
46J. Murata, E. Arai, K. Kobayashi, and K. Kudo, J. App\. Phys. 50, 804
(1979).
47M. Yoshida, E. Arai, H. Nakamura, and Y. Terunuma, J. App\. Phys. 45,
149811974).
4S. Matsumoto, E. Arai, H. Nakamura, and T. Niimi, Jpn. J. App\. Phys.
16,117711977).
4"S. Matsumoto, Y. Ishikawa, Y. Shirai, S. Sekine, and T. Niimi, Jpn. J.
App!. Phys. 19, 21711980).

JApplPhys 54 5049

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

JApplPhys 54 5049

Uploaded by

Copyright:

Available Formats

Oxidation enhanced and concentration dependent diffusions of

You might also like