" --

/.

",

-.. .'

..THE CHARACTERIZATION OF THIN HEAVILY DOPED LAYERS m SILiCOr~

•

,

•

"

/

•.~

• !'

•

THE CHARACTERIZ'y,ION-, THIN HEAVIL Y'-OOPED LAYERS IN SILICON

, .

,By" . ..

",

JOHN ERLE JONES B.f\.Sc .. (Toronto].. '~ , 4

. ...

A Thesis..Submitt~d to the SCh001 of Graduate Studies

in Partial Fulfilment of the Requi rements )

for the Degree

Doctor of Phi losopy .

)("

McMas ter Un; ve rs i ty

November. 1973

@ .John Er1 e Jones 1974 ii-

DOCTOR OF PHILOSOPHY (1973)(Electri ca1 En9i neeri n9) J

.'HCMASTER UNIVERSITY J.

Hamilton, Ontario

. , TITLE:

AUTHOR:

~

The Characterization of Thin Heavily Doped Layers in Silicon

"

-I,.

John Erle Jones, B.A.Sc. (Toronto)

SUPERVISOR: Dr. H.D. Barber,

NUMBER OF PAGES: 173, xiii

,,'

11

•

" .

,'". '

... ~~

-',. --

\

ABSTRACT

•The electrical characteristics of "a thin heavily doped layer in

silicon were determined.+" .A P 200 n![J:~ ~m junction depth boron diffu$ion in a 2 n em n-type

phosphorous. doped substrate was considered a typical thin heavily dopedII.

layer. This layer is of considerable practical importance beinga stan-

dard base a~d resistor diff:?ron in Planar' proceSS'i~9. •

•

To. completely characterize the heavily doped layer, it .was necess­

ary to study·the layers"~n either side; ~ubstrate and thermal o~idei and. .t~ei ri nterfaces with the heavily doped 1ayer. The resul tant structure

•

" ..was examinJd by mean~ 'of MOS C-V, oxide leakage and breakdown, Hall effect,

V. ,.p-n junction characterization, and differential sheet resistivity using, ""

anodic sectioning and" a four~point pro~~~ These tests yielded the i~

purity profile, th'e .transport and recombi.n~ion properties of the sub­

strate, the dielectric const~, thickness, breakdown field and leakage

'~urrent of the oxide and the. properties of the substrate diffused layer. .,interface. However, the properties of the heavily doped layer and its

interface with the oxide were still unknown. To find them, an experimental

pro~edurc utilizing a new test structure was.. introduced.

The test structure consista:l of an MOS capacitor formed over ~

heavily doped layer. The layer in turn formed one side of a p-n junction

diode. The resultant structure was an MOS ca~acitor-emitter transistor" I .(MOSCET). In operation, the p-n junction was reverse biased and became

a minority carrier collector, whi~ the MOS capacitor was biased into

i11\

l "

inversion for a.cbntrolled length of tire during wh.ich minori~y·

· carriers were generated and. forre~ the inversion layer. A voltage-pulse applied to the MOS capacitor caused it to 'release minority

, ,carriers from the i~ersion laye·r. The carriers were. transported. .

acrll,ss-the layer under the combined :influence af diffusion,a.nd internal. .fields to the p-n junction where the charge that had not recombined

·during transit was collected.

J. The theoretical behaviour of·the MOSCET was ~etermined. First

a MOS capacitance program was used to d~termfne the inversion charge

· as a function of the bi as voltage. Then a model of the devi ce was.•developed to explain the transient response when no inversion charge

was present. A finite difference formulation was used to account for

· the minority .carrief>transport across, and recombination in, the .

,,

..~.a boundary condition which included va1ues calculated in the non-·/;

inversion resp~nse. The result 'was a self consistent mode] of t~MOSCET which also agreed with the other expe~imental data available.

The MOSCET can be used t'a. study two other phenomena as wel1/~s. .

•diffused layer. A time dependent release mechanism was introduced as. .

i. ,the cha~acterization of thin heavily dop~ layers. This structure

. . - -..

following properties of-

,

,\.

supplies the only direct confirmation of the relationship ,betwee~I '-.

•invers ion cha r.ge and bias voltage on the MOS capaci;;O'r;' ,A study"of

the avalanche properties of p-n junctions can be made, at very low. .multiplication values, with the MOSCET since the number of injected

carriers ;s so well known.

Measurements on the MOSJtT Yiel~ the

,iv

r

thin heavily doped layers.

f (l) The minority carrier generation lifetime near the inter-" ."

"fac~.

"(2) "The generation centre energy level.

(3) The faverage minorit arrier recO{Jlb~nation lifetime

""~

..

es across the layer.(4) T

(5) The

(6) The fixed oxide charge density.

(7) 'The insulator s\ability.

Thes~ properties combined with" those determined by the other

techniques provi~e the "most complete electrical characterization of the" " .

" th\n heavily doped layer to date.

\ ,," r

\'" .

" .

I

v

•

ACKNOWLEDGMENTS. ,

for his guidance, encouragement and assistance throughout this project':'-

l>,I would like to express my t~ank~ to my supervisor, Dr. H.D. Barber..

,.

and.tD the other members of my superv.isory cOlmii,ttee, Dr. C.K. Campbell.......... ~

and Dr. J. Sti~wchun.fo.r.J~eirassi stance.......

/. . ' ..

. I am grateful. to ~V;A.K~·temPle·for. the use of his computer

pr~gram for cal~ulating MOS capa,:itance ~urves 'an~ to Mr. G.T.Bereznai

and Mr. N.D. Mar;kettos 'for their computer programs ,for.sys.tem modeling.

The ass is tance of the Soli d-Sta.te Devi ce~ .g1'oup of Wes ti nghouse

Canada ltd. in fabricating the devices and inal]owing the use of their

test facilities ts gratefully' acknowledged.

'This work would not have been possirrle without the fin~ncial

assistance provided by the National Research Council of Canada and McMaster

Uni vers,i tV.\ ,/

. Finally, thanks are due to Mrs .. lillian Mogensen for her care andr

patience in typing from the manuscript.,

I

vi

\

"~.

TABLE OF CONTENTII

CHAPTER 1 INTRODUCTI ON 1\

CHAPTER 2 DEVICE FABRICATION 62.1 Processing 62.2 Preliminary Tests 11- ,

CHAPTER 3 CONVENTIONAL CHARACTERIZATION 14

I---3.1 Introduction 143.2 The MOS Capacitor 143.2.1 Theory of Operation q- 143.2.2 Experimental 'Equipment 193.2.3 The Impurity Concentration at the

Surface of the Silicon 223.2.4 , Non-Ideal Behaviour 22

~3.2.4'.1 Work. Function Difference 233.2.4.2 Oxfde Charge ,," 263.2.4,3 Surface States 273.2.5 The Measurement of Surface States 273.2.5.1 Introduction 273.2.5.2 The High Frequency Method 283.2.5.3 The Low Frequency Method' 283.2.5.4 The Conductance Method • 293.2.5.5 The Static Method 303.2.5.6 Conc1us ions 303.~6 Impurity Band Broadening 313.2',7 The Inversion Layer Response 32

0# 3.2.7.1. The Small Signal Response ,323.2.7.2 The Transient Response • 333.2.8 Conclusions 333.3 The Die1ectri c 343.3.1 Introduction ' 34-.3.3.2- Oxi de Charge 343.3.2.1 Fixed Oxide Charge / 353.3.2.2 Mobile Oxide ~harge 35

{i) Mobile Ions 35i i) Po1ari zati-o!l 36iii) Trapping 37

3.3.3 Oxide Thickness 37,3.3.4 Diele~tric Breakdown 39 •3.3.5 Conduction Through the Dielectrtc 403.3.6 Conclusions 44

• 3.4 The Impurity Profi le , 44. , 3.4.1 Introducti on ' , 44': ,3.4.2 Theory of Ditfus i on' ~ _ 453.4.3 Approximate ChecK on the Distri~on 45 ~

Ivii

u

3.4.4 C~V Profil i ng 46. 3.4.5 Spreading ~sistance Profiling 463.4.6 Profiling by Sectioning 473.4.7 Concl us ions 48.3.5 Determination of the' Substrate

Parameters ~ 503.5.1 Introduction 503.5.2 Theory 503.5.3· Experimenta1 Procedure 51 d3.5.4 . Experimental Results 51 ,3.5.5 Discussion 523.5.6 Concl us ions 523.6 The P-N Junction 533.6.1 Introduction 53 .3.6.2 The Diode Capacitance-Voltage

. Re1ationshi p . 533.6.2.1 I"t,"d"'ti'"~ . 533.6.2.2 Th~ory" • 53

, 3.6.2.3 Experimental esu1ts 54

~3.6.2.4 Conclusions . 573.6.3 The Diode·Fo ard Current-Voltage •

Relationship 573.6.3.1 Theory 573.6.3.2 Determination of Parameters 59

~ i') Material Properties 60i1) Trap Characteristics 61

3.6.3.3 Experimental Results and Calculations 683.6.3.4 Conclusions 683.} Conclusions 71

CHAPTER 4 THE HOSCH 724.1 Introduction 724.2 The Measurement System 73

\4.3 A Typical Response 804.4 Cone1us ions 84

CHAPTER 5 THE TRANSIENT RESPONSE 855.1 Iniroduction .-.. 85........ 5.2 +. The Non-Ir;lVersion Response '"""\; 855.2.1 The External Cir~uitry '~ 865.2.2 The Applied Voltage Pu1.se "- 865.2.3 The Devi ce Moder 005.2.4 Detenrri rrtl ti on of the Non- Inversion

Response 925.3 The Inversion Charge Response 945.3.1 Introduction 945.3.2 The Transport Equation 955.3.3 The Release Mechanism 1035.3.4 ,The Circuit Response • 107

.,viii ,~

'" .5.3.5 • Determinatlon of the Inversion Charge

Respons~ 1095.4 'Conclusions III

CHAPTER 6 THE MOSCET RESULTS 1126.1 The Non-Inversion ,Response 1126.1.1 Experimental Results and Discussion 1126.1 .2 Conclusions . 1186.2 Generation 1206.2.1 Theory 120

• 6.2.2 Experimental Results and Discussion 1256.2.3 Concl us ions 1326.3 The Invers ion Charge 1326.3.1 Experimental Results and ·Discussion 1336.3.2 C~lusions 141 ~6.4 Tr nsport 1426.4.1 . Exp ritnental Resul ts and Discussion . 1426.4.2 Concl usi ons 1476.5 Conclusions 147

..

~CHAPTER 7 CONCLUSIONS AND RECOMMENDATIONS 150 ,.

APPENDIX A IMPROVEMENTS IN THE EQUIPMENT 155

REFERENCES 156

--"\

," ....

• 1x

FIGURE

,,-

• LIST OF FIGURES'

. '

2.1 The mask used to define'the boror diffusion 7

2.2 The mask used to locate the base contact 7

2.3 The mask used to define the' metal i zatioR pattern ' 9

2.4 Pho~ograph of the completed device 9

2.5 A cross-section of the test'structure 10

3.1 Energy bands and charge distribution in a HOS structureunder various bias conditions 16

3.2

3.3

3.4

3.5

3.6

3.7

3.8

MOS capacitance-voltage .curyes 17>

Experimental and ideal high frequency MOS C-V curves 20

,The energy band diagram of a HOS structure illustrating'the metal-semiconductOr work function difference 24....The charge distribution.of'a HaS s\ructore due ,to ,a,charge in the oxide 24

•The current-voltage characteristics of the thermal, oxide plotted to demonstrate ,the conduction mechanism 43

'The impurity profile of the p+'dtffused layer 49I

Theoretical and experimental C-V curves for a reversebiased p-n junction dio~e

"

"3.9 The'variation of the reverse current of a p-n junctiopdiode with temperature ' 65

3.10

4.1

4.2

4.3

The theoretical and experimental I-V curves of a for­ward biased p-n junction diode .

The MOSCET test circuit schematic

The experimental equipment employed

Typical transient respon~ of the HOSCET"

xr

70 '

74I ',76

81

,

82

87

89

The ~C e~uivalent test circuit,

TtM! experimentn1 voltage pul se and'the analyti c fit .

The"tnYersion charge Q-V characteristics of a MOSCET,~

4.4

5; 1

5.2"

5:3 Discrete component model,S of the MOSCET , ,90 0

5.4 The theoretical and experimental non~inversion responseof the MOSCET 93

5.5 The internal field across thedtffused 'layer ' 96'-

5.6 The calculated collected charge with and withoutinternal fields ' '100

5.7

5.8

The calculated voltage pulse across the MOS capacitorand the pi ece.wi se li n~ar model used in progralll11ing J06. ~ ~

The collector circuit. including the voltage dependentcurrent source 108

The variation of the'nop-inversion response with temp-erature ' 116

5.9

6.1

The calculated and experimental inversion response 110

6.2 The effect of'the reverse 'biased p-n junction capaci-t.ance on 'the ITOn-inversion response " ,... . .

", - f·

117

6.3

6.4

6.5

6.6

The' non-inversion response as a function of theapplie~ voltage,~ulse amplitude

The inversion charge generation curves for threedifferent values of MOS bias voltage

The inversion charge generation time for ,three 'different values of MaS bias 'voltage

',The inversion charge, generation time at 22'C

, 119

121

126

127

6.7 The inve~sion charge generation time for different~ applied,voltage pulse amplitudes 129 '

6.8

&.9

The variation of the generation time constant with'temperature

The effect of MOS.bias voltage,on t~e' c~llectedcharge '

130 '

134

xi

6.10

6.11

6.12

The collected charge versus MOS bias voltage fordifferent applied voltage pulse amplitudes

The .collected charge at various reverse biased diodevoltages

The multiplication factor as a function of thereverse biased diode voltage .

137

138

140

6.13 The variation of the collected charge with temperature 146

<;. .,

, "c•. .'

.'

. !

.~ .

•

xii~ .. '

2.1

3.1

3.2

3.3

3.4

3.5

3.6

5.1

6.1

.---.

(J

LIST OF TABLES

Initial Test Resuhs

MOS Capacitor Data

Metal Work Function (Volts)'

Basic Conduction Processes in Insulators

Reverse Biased Diode C-V

Reverse Biased Diode I-V

Calculated Forward Diode Current

Transit Times Across the Base

Experimental Results from the MOSCET

-.

f

•

xiii

12

21-,

25

42

55.

67

69

102

,113

,

•CHAPTER 1

INTRODUCTION

"

Since the discovery of the transistor'by Bardeen and Brattainl

in 1948 the electrical properties of semiconductor layers have beeri of

concern ,to the device designers and process engineers. Initially the,

work concentrated on producing better material of higher pl,lrity. witht ' ,

fewer crystal defects and more controlled means of introducing impurities;

first by alloying •.then diffusion and now epitaxy and ion implantation.

At first the efforts were concentrated on characterizing the lightly/

doped regions which dominated the d~vice operation. With closer specifica-'

tions ,find much better .cont.rol of the '~roperties of the lightly doped regions ,

the heavily doped regions have become much more significant. At present

the 'properties of t~in heaVily doped layers are of great importance

~hroughout the semiconductor ind.l,ls'try; in large scale integra~ion the shallow

heavily doped emitter is critical; in power devices the lifetimes in the" ,

heavily doped l!yers limit the operation and in FET's the heavily doped

source and drain are of major significance.o ..

,To completely characterize a thin heavily doped layer is not a

simple matter. It is ~ecessary to know not only the properties of the. .

layer itself but ~lso the properties of its interfaces Wit~ the nejgh-,

bouring layers and to detennine these, i~iS necessary to e~aluate the

properties of these layers aS~l!:ll .. With a semiconductor, three different

types of neighbouring layers.~re possible; another semiconductor, an,

Lt' •1' .......

1.'