received - NTRS - NASA

0 downloads 0 Views 3MB Size Report
This water and oxygen can cause oxide formation on the substrate and melt ... of the Ga.. Al As layer thickness, growth morphology and x. control. ... of Al needed per gram of Ga to form a melt which when saturated with GaAs .... 12. Leaching Experiments. During the course of this investigation it was determined that the.

NASA CR - 145008

FINAL REPORT PHASE I Optimization of Solar Cells for Air Mass Zero Operation and a Study of Solar Cells at High Temperatures.

NASA Contract

NASI-12812

Principal Investigators - H. Hovel, J. M. Woodall

Work Performed by Exploratory Materials and Devices Project IBM Research Laboratory Yorktown Heights, New York.

NASA National Aeronautics and Space Administration

RECEIVED FACILITY

_

THE OPTIMIZATION OF Ga., Al As-GaAs SOLAR CELLS FOR AIR MASS ZERO 1-x x OPERATION AND A STUDY OF Ga. Al As-GaAs SOLAR CELLS 1-x x AT HIGH TEMPERATURES

By H. J. Hovel and J. M. Woodall IBM Thomas J. Watson Research Center Yorktown Heights, New York

10598

Background'and General Objectives High efficiency solar cells based on a material such as GaAs have been sought for many years but have proved to be elusive in the past.

GaAs has

several advantages over Si for use as a solar cell; these include higher maximum efficiency, improved performance at high temperatures, higher output voltage, superior resistance to degradation from particle radiation, and a potentially better power to weight ratio. However, there are several major problems associated with GaAs which have prevented the realization of these advantages. The first problem is surface recombination.

GaAs has

much shorter lifetimes and higher surface recombination velocities than Si, while at the same time the optical absorption is high. Most of the charge carriers in GaAs solar cells are generated within 2 microns of the surface, and surface recombination losses therefore have a strong effect on the collection efficiency.

(By contrast, significant numbers of carriers are gen-

erated over 100 to 200 micron distances in Si, and surface processes have a smaller effect.)

In order to minimize the surface recombination problem

in GaAs, the junction depth (depth of the junction depletion region beneath the surface) must be made small; however, this introduces the second major problem, that of series resistance.

If the sheet resistance of the surface

region is high, the resulting high series resistance will reduce the available output power.

This can only be partly compensated by higher doping

levels, since higher doping levels mean shorter minority carrier diffusion lengths and reduced collection efficiency.

A compromise must be made be-

tween large junction depths and high doping levels to minimize series resistance and small junction depths and low doping levels to maximize col.lection efficiency.

The end result is that the power conversion efficiencies

of GaAs solar cells have been less than those of Si cells, and most of the work in solar cells in the past 10 years has therefore been concentrated on Si.

It has been recently demonstrated that the two major problems limiting GaAs cells can be minimized by the use of a thin, transparent layer of p-type Ga n _ Al As on the surface of the GaAs p-n junction. i"™X

X

The lattice of

the alloy layer matches that of the GaAs very closely, which greatly reduces the recombination velocity at the interface between the two materials (which is now the "surface" of the GaAs p-n junction) and therefore reduces the surface losses.

In addition, the p-type Ga1.L^X AlXAs forms an ohmic con-

tact to the p-type GaAs surface region, and since the sheet resistivity of this alloy layer can be made low, the series resistance in the device is considerably reduced for a given doping level in the GaAs.

The Ga-X~XAlXAs

is transparent to photons with energies below 2.1 eV and the absorption increases slowly with increasing energy above 2.1 eV.

The result is that

solar cells consisting of pGa.. AlXAs-pGaAs-nGaAs have exhibited efficiencies ^""X of over 16% in sunlight at Air Mass 1 and 20% at Air Mass 2 or higher at 123 4 room temperature ' ' and almost 6% at Air Mass 1 at 300°C. This new type of solar cell outlined above exhibits considerably higher efficiencies when measured in sunlight within the earth's atmosphere than any previous solar cells of Si or GaAs.

It is the purpose of this contract

to develope the techniques and procedures which are expected to result in the optimization of this new type of solar cell for operation at Air Mass 0 (outer space environment).

It is also the purpose of this contract to de-

velope fabrication techniques which should allow the devices to operate continuously at high temperatures (T = 300°C or higher). Three types of structures are to be investigated:

one consisting of a nGaAs substrate, a

Zn doped pGaAs region, and a Zn doped Ga- Al As layer, the second con• i™X

X

sisting of an nGaAs substrate, a Ge doped pGaAs region, and a pGa1

i—xAlxAs

upper layer, the third consisting of an n GaAs substrate, an nGa1

Al As

-L~X

region, a pGa,-i.*""XAlX As region, and a pGa.. Al As upper layer. i™"y y

X

In all three

cases, the upper alloy layer is thin and of high Al composition in order to obtain high spectral response over the widest possible range of photon energies.

Spectral response, capacitance-voltage, current-voltage, diffu-

sion length, sunlight (or the equivalent)-efficiency, and efficiencytemperature measurements will be made as a function of device parameters in order to analyze and optimize the solar cell behavior. Liquid Phase Epitaxial Growth Experiments At present, the most successful technique for growing Ga,

-L~X

Al As layers X

on GaAs is liquid phase epitaxy (LPE) .. This is a solution growth method in

which Ga, Al As deposits from Ga rich melts onto a GaAs substrate as a re1-x x suit of slowly cooling the melt'from an initial high temperature. A considerable amount of effort was devoted towards developing this method to produce uniform Ga.. Al As layers, 0.75 -^ _^ TJ "DA -»-»-» H A U _L + TJ H-^U,

j

Z.

_^3 ~* "*

t

PH -v 2.0 50 Volts

4 4 1) Procedure for substrate with no thin layer epi growth

M

2) Procedure for substrate with thin layer epi growth

N

R

Load«Wafer

-Q

Blow Dry

D. I. H00 Rinse 2

4 4

Blow Dry-> -* -» -» -

4 P HCl Rinse*" «^ -«-«-«-

0

Store

4 4

Table 2.

Melt Compositions, Layer Compositions, Layer Thicknesses for Ga-Al-As Melts Saturated with GaAs at 900°C

Weight Al (in gms.) per gm of Ga

X in Ga.. Al As Layer x

Thickness (in micron per:gm. Ga, °C eooling and cm of substrate area

0.27 x 10~3

0.1

3.0

0.6 x 10~3

0.2

2.6

1.02 x 10~3

0.3

2.3

1.56 x 10~3

0.4

1.9

2.31 x 10~3

0.5

1.6

3.42 x 10~3

0.6

1.2

5.26 x 10~3

0.7

0.93

8.97 x 10~3

0.8

0.62

20.4 x 10~3

0.9

0.31

Table 3.

Flow Diagram for LPE Growth

Fixture Preparation

I Load Substrate

Substrate Preparation

4 Load Melt Comp.

4 Assmeble Apparatus

Vacuum & H Backfill

LPE Growth Program

Prepare Melt Comp,

vo

ON

00

N>

ro o

Ui

5O W C O 3 OS

OO

OO

oo S co o

S3

U>

oo

oo

-o

•sj

Ul

Ul

O>

ON

00

Ul

b

S3 o

UI O

vo

J>

3 s:

VO

o

o

O



oo

VO U>

00 Ul

CD

co u>

00

O

U)

oo

.00

oo u>

03

Ui

u> ro

UJ

"CO

H S3

NJ

N3

X

T=

N3

SJ

*

*•

*

oo

00 0 P3 (-3

*

3fr

Ul 0

00 Ul O

O3 O U

N3

M

• S H

vo

O >

vo

O J M

| > N

O 3

*

vo

U

>

N C 1

O U

| O

1

*

*

vo 3

U

vo O

I

cr l-> re

TO N *•-. 3 00

NJ

vo N

' O U

vo

J V

1

I

JD

O

J>

M

-

Ui

j

•**J

vo O

"*

£ CO

H^

'

rt

0 0

O Hi

!« g 3

o> OH

Ul

CD CO

5

1

ON M CO Ul O —



o

*

Z

Z

C

Z

Z

ON O

^ ON

ON O

*.j

o

o

o^

o

«^j

o

*ij

o

o\

o

=

1

O

s*

0 ON 0

O ON 0 Ul O

=

O N ? CL _

H. 3 Hi Hi

H 0

SJ

3

1

rt f H-

re

11 &

f CD C/3 O

o n N3

o *

N)

N3

N3

''1 ^^ H- ar oo

o c I

•^ as c ro re "a

I

3 Oi •O rt

re o. C

V O

N3

UJ

N>

W

OJ

re co

C

If re

oo' oo

N3

o a H. • o 3 •a a- H1 re re ho if ^•* ^ rt O

CD rt rt X

C

n

if o O ffi 0 •

re

CL rt O *

Ui

z

3

ON O

oo

o c

^ H-*

^l O O

3

< o o

c

M

3 Ui rr re 2

re

3 o X •o HI re o o 3 re CD re UJ

3 a ?D

H rr

rt ^•u C rt

re

O O O

X

o



M Ui

re a.

o

o

CO

H-

^ O (D

o

o o

*

2

re

3

re CO

yer not removed Ing substrate

T-t

tJ h CU

*4

(0

rH

i 3

u w o &o

•o

to-

e rH cu cu .e s

rH

0

u

Q

M 0) i-l ^4 73 O O *W

C


ca

a ca cu

U

CO

o •o c

CU

X

r*_

-3" CM

CO

CO

cu eo C

C -H

^ O 33

cu o. cu o. . -a

periment

T CO 1

P-I

< -H

r*. ON ,^ I

CM •H -CT CO 1

*»3* CN

04

CM

ca « 0 vO

'

M



o

\O

3

• o vO

o

i

1

i

1

1

1

i

1

cn m co

cn m oo

o m co

1

rH

i O



d o u

o

0 oo — m CO iH

a

o m

CO r-l



-3>

cu

iH

ca

H

.

0 o

o in

S "*

CM

•-

cn

m•

1

CN

CM

CM

iH

rH

0 0

*^» o

CM

cn

m o\

iH o

ao ON co

ON

O ON

CM

0 ON

CO

&\

'-(

cn

0

ON

cn

CM

CM

CM

o

CM

CM

1

CM

vO

ON

ON

CM

O

cn

in cn

0\

O

1 rH

f>-

00

co

CM

en oo

m

m

m

o

o

O

0

0

o

o

oo

CM

oo

vO

ON

co

-3-

cn

CM i— 1

o cn

rH

rH CM

in

cn

*

rH

rH

rH

rH

O 1

1

1

1



rH

ON

CM

1

1

1

1



O ON

ON

0\

-T vO

-3vO

vr

I

°.

,

vO

0 vO

^ O *

CM

cn rH

O* rH-

rH

rH

CM rH

O

O

O

O

0

rH

rH

CM

en

cn

m

tn

m

m

VO

00

a\

0

rH CM

CM

cn

CM

CM

CM

rH

^_ rH

rH

rH

rH

CM

cn

^ CM

•s

CU

r-l iH

to o cn

iCO

CO O -rl

AJ

c

cd

•H 01 U O. cd O M

0)

'O CD

CO

C

c

LN1 •H

TJ C

cd

r-4

0) 0 ^4

a g

0

CO

O

^ CN

urce

t failed

*d

^•1

CO

•H

CO

(U

cu < cd .H O

ao

M

CU

w

J—I


v.

o 15 o

10 nGaAs(Base)

G aj . x Al x As

.05

.1

.2

.5

I

25

10

Thickness of the Ga,.xAlx As, Microns. Short circuit current contributions. P a r a m e t e r s of T a b l e 1. Xj - 0.5u. Figure 12

50

10 O

CO

o Q

ce: o

0.1 46

2

0.01 0.4

0.5 0.6 0.7 0.8 FORWARD B I A S VOLTAGE

Measured current-voltage characteristics Figure 13

0.9

1.0

I

.05

I

I I I II I

.1 .2 .5 I 2 5 Ga i-x A l x A s THICKNESS, microns.

Calculated AMO efficiencies versus GaAIAs thickness P a r a m e t e r s of T a b l e 1, Xj - 0.5u.

Figure 14

1.2

1.0

85,114,148,179, 2 o 0.8

1 55 22

D_ (/) LU

0.6 Q_

0.4

3 LU

0.2

335-1 0

1.4

1.6

1.8 PHOTON

2.0

2.2

2.4

ENERGY

VARIATION OF Gai_ x Al x As BANDGAP WITH TEMP.

Figure 15

2.6

0.7

o

T i i i i

i i

0.6 149 82 22

H—'

O

0.5 TO O

22 00

O

0.4

Q_ CO LU

0.3 o

scb46 0.1 1.2

I 1.6

2.0

2.4

PHOTON ENERGY,

I 2.8

3.2

eV

S p e c t r a l r e s p o n s e s of a d e v i c e w i t h t h i n G a A I A s layer, D - 0.6u.

Figure 16

1.1 12

1.0 11

32

0.9 10

28

24

8 20

0.6 7

16

298D1-2 0.5 6 0

I 50

I 100

I

150

200

250

TEMPERATURE, C DEVICE PARAMETERS AS A.FUNCTION OF TEMPERATURE

Figure 17

300

14

12

10

8 o

o

6

0

50

I

I

100

150

TEMPERATURE,

200

1 250

300

°C

Measured efficiencies ( contact area corrected )

Figure 18

2.49

Quantum efficiency variations at 6328A0 The n u m b e r s are in 10"1

Figure 19

miliiamps.