I

, '

User's Manual
i\

FOUR PROBE SET-UP

Model: DFP-03

. i­
. ...
~

,
I r,

\t,
,"

',)\
Mllnllfacturod by:
w·w
SCIENTIFIC EQUIPMENT & SERVICES
16. Civil Unes, Roorkee-247667
Ph.: 01332·272852. Fax: 274831
EmaIl: [email protected]
®§t

ISO 9001 : 2000

c.rtIIItd Company
~
Website: YMW.sestechno.com
l
I
~'\

' '. " "

]
, I •
, ,'. '.
~ ~~.

'I,
\,' V.. r,<I'""',"'"'~""~
'...
\

.
'

.~
'oJ ;. ~.

~
.
•
~.~".~'''''"·"'._._,·u.'_ ",~"~,,,-,",
....._.••
,

' '
. ·e·..·.. '...-.. ·· ,.' ..... ," ...... ~ ... , .... -,_.,-.,. "'.,,"
... ...... '.' ' - ' . . - .... "."" .
, "':",-~.'~.(...,.." ....
~·.. ; r ' " :f

;_,""":'li""~'!I'!'!!~~~~'
 "

iNTRODUCTION

The properties uf the bulk material used lor the fabncatlOn of transistors and other
semiconductor devices are essential in determining the characteristics of the cOlllplct('d
devices. Resistivity anJ lifetime (of minority carriers) measurements are generally made
..I on germanium crystais to determine their suitability. The resistivity, in pnrticular. must he
measured accurately since its valu~ is critical in many devices. The value of some
transistor parameters. like the equivalent base resistance. are at least linearly related to
the resistivity.

ELECTRONIC CONDUCTION IN ~OLlDS

The el~ctrical properties o~ semiconductors involve the motion of charged

particles within them. Therefore, we must have an understanding of the forces which
control the motion of these particle~. It is of course; the physical structure of the solid
which exerts their control. This topic is very large, and hence only the high lights will be

I
covered. The reader is referred to ma'ny excellent sources which are listed at the end, for
more detai Is on speci fic aspects. i

Atoms, of which a solid is composed, consist of positively charged nuclei with

electron orbiting around them. The positive charge is compensated by negatively charged •
electrons, so that a complete att:)m is electrically neutral. Electrons are a.crangeq in shells.
and the closer they are to the nucleus the more strongly they are bound. If we take the
particular case of silicon, a well known semiconductor, we find that it hus J4 electrons
which are accommodated in the shells as (1S)2, (2S)2, (2P)6, (3S)2, (3P)2. Since the third
shell is not even half filled, the 4 electrons are available for' chemical binding giving
silicon a valency of four. (Germanium also has a chemical valency 9f 4, but from the
fourth shell). Fig. I shows an energy d~agram of an individual atom.
Let us now concentrate our at~ention on solids, if we bring many atoms close to
., one another, interatomic forces become quite strong as' electronic orbits begin to overlap.
The ol,lter ~neH e~~~tnlg~ play an important role, because their orbits are the most ".
i: disturbed. These electrons are no longer associated with a particular atom, the outer 'shell
electron may make an orbit around one ~tom and continue about another. In this fashion,
the outer shell or valency electrons are' continually traded among atoms and wander all
over the solid. The continuous intorchlmge of valence electrons between atoms holds the
solid together. This is the predominant type of bonding in silicon and germanium, and is
called the v~~encebonding.
In solids, atoms are~usually arral1ged in a regular way to achieve a dense packing
. and thereby form a crystal. The arrangement has very desirable' characteristics, i.e. the
trlltlSport of holes and free electrons is very smooth in these structures. When the
arrangement is not crystalline complication arise. Here we will be concerned oniy with
~{ the properti~s of perfect crystals. Silicon and germanium (and carbon) crystallize with <111
jd'cmti~al crystal structure, the called diamond structure. Such a structure is shown in fig.2
!

AS~~~

~
I. . . . . . . . . . .

"'tq~,·,

11.",41 Ii Qt·,. .- .
")
-- I.
;·''''J,\''l:Ii>~!R!Wt\!:.lI'

..; .~

J
ENERGY
f
.....--r r -_

_
rl 10 r,
()
~ WORK REQUIRED TO FREt
OUTER SHELL ELECTRON
E=""IOe V
AN UNFILLED SHELL
~; ." rh'J{ . -_ __ _______ 1______ _
OUTER SHELL POSITION

SHELL CLOSER TO NUCLEUS

NUCLEUS

FIG. 1 POTENTIAL IQISTRIBUTION AROUND AN INDIVIDUAL

ATOM

.
-----­ ---,y
.n r

/ I
/ I \.
/ I
I
I
I
I
I

I
I
100..
I /

--------------0 I I

FIG.2 STRUCTURE OF THE DIAMOND

LATTICE

"-~'""",*•. .\.." .•• - .....- -.. -.,......-- ...-~,-."....."... ,~';',>I" ~~

 ; ~ arr3ngcm~nt of atoms in the illustration form a Unit Cell, and the crystal is made up
(adjacent UOI t cells.
tJ
FiS. 3 shows :I potential dinsram of an array of atoms. All adual crystal IS of
ourse there dim~nsional. The most important difference between the potential plot of an
"
iSIllated atom and one dimensional array is the Spliltlne 0f energy It:vt:l:;. In fact. bringing
'

N atoms together we find 'N' times as many levels throughout the crystal. The spreading
o( energy level ~epends on the degree of in.t~r?cJjoQ ..Jh!~rc:tf.QLe.Jh..!!j.n.D~Mb.itlupJit into \
iCvels combined in a narrow energy than the outer ones .
...
As a result of the interaction between the tremendolls number of atoms in Cl yslal
(tQ2~_per .<;m3.). the energy level found in isol~ted atoms will be split and form bands of
:ifJowed energies which contain almost continJum of levds. Accordingly. el~,ctrons an:
located in energy bands. in crystalt"ine solid. The band which contains I'he valence
electrons is called the valence bane]. The unoccupied energy. levels also split up and form
another band called t~e conduction bang. -Their1t~~!1~tion b,!!tween tIn: unused ~hells is'
very farge 'arid they spreaci wi®Jy. Therefore, vJNlethere is a bandgap, Eg (or forbidden
region) be~ween the valence and conduction bands, splitting of higher orbit is so wide that
they usually overlap.
The bands below the energy gap Egare c0I!lplelely filled at absolute zero
temperature and the conduction band is empty., This is a very important point and has
direct consequences on the conduction properties, as we shall see soon. The fundamental
theory is that current conduction is not possible in empty and filled bands. The reasons
about the empty band is obvious since current is ,not possible without carriers. The reason
about the filled band is as follows: though the valence electrons move about the crystal
but they can not be accelerated because the acceleration means gain of energy and there
are ilO higher energy levels available to which they could rise.
We can now readily see that the crystal band structure shown in Fig. 3 does not
allow current conduction at, T=O. If we increase the temperature, however, thermal

~I agitation increases and some valence electron wi'll gain energy greater than Eg and jump
into the conduction band. The electron in the conduction band is called a free electron~

and its fonner place in the valence band is called a hole. Electrons in conduction qand can

• gain 'energy when a field is applied, becausel there are many higher energy states

I available. The fact that electrons left the valence, band leaves some empty energy levels,

' this allows conduction in the valence band as wep. Electrons ca.,:t ~ow, g~~n <!n.ergy, in the

" \l'al~nce. band also, and we observe a motion of hO,les in the direction of the ~eld. Because

ofthis we begin to speak ofa hote as a current cagying particles.

. " I

/ , According. to the proceeding theory, an insulator must have a large bandgap, so

1 that af room temperature the conduction band is practically empty and the valence band is
practically filled and a semiconductors must have a narrower band gaps so that
appreciable number of camers are present in the valence and conduction bands at room
temperature.
In metals, however, the valence and conduction bands overlap and application of
• an eiectric field can. therefore, accelerate a great sea of electrons. The non-existance of a

'"
:mS2EL It iU J H II j.t4 £O~'I!l(>'1'~'h"""'jf"
 'r, ,,- .. ."''''~i"<.,,,,~·Jtf' rrl*'1ittrt
f J

-........

'"

,\

:\
! t QISTANCE
COV::;TAL .. , .. " VACUUM

ENERGY

FREE E.LECTRONS

-fif=
__ CQNDUCTION BAND

.f19C,\ B~rDij1ff

Eg
HOLES =rJ{ 7 '

I
FIG, J UNUSED SHELLS FORM THE CONDUCTION
BAND
I\
I ~~~~'~~~l
0,( CONDUCTION
MPTV BANQ
I '
Eg INSUL ATOR
SEMI-CONDUCTOR .~

CONDUCTION BAND .
i

Ii...

L....r= ,c(511:;l<)<)~,(~5)~ METAL 1­

VALENCE

BAND

_1

FIG, ~ BAND STRUCTURES FOR INSULATORS,SEMICONDUCTOR

AND METALS

I
~;i"
,
!"
'.~.
, •. ! '
...
i . ~
. il'i:,,:
,- . •
,.1.·
~'\ ~
:L~',
.'
, ,,,,,'.
10 ,~,~, 'L ;",' ,", <'
- CONDUCTIVITY OF INTRINSIC SEMICONDUCTOR
The electrical conductivity will be the sum of the conlrihlltlnns of both eleClruTlS and
!loles .

(1 ;;: (ni.q.Lc + pi.e·llh) I' ;:

(5 )

Where e is the electron charge. Ile and ~th are the average velocities bv Ihe

;h.!t:lIUn!l J.IIJ ilolc~ III d ullit d~LlIll., liclJ <'IlIJ kllOWll a:; mobllltlcs.

0r (J = c ni (Ilnl- Ilh); Since ni = pi

'~(K) T3/ 2 (Iln + IlP). exp2kT (6)

using Eq. 4 where K is a constant.

The factor T3/2 and the mobilities change relatively slow with temperature compared

)lith the exponential term, and hence the logarithm of resistivity p lfa) varies linearly with

lrr. The width of the energy gap may be detennined from the slope of the curve.

Thus we have,

E . \

Log p = ~ -log K (7)

c 2kT c

:XPERIMENTAL CONSIDERATIONS
High resistance or rectification appears fairly often in electrical contacts to semiconductors

and in fact is one of the major problem.

In single crystal IIlaterial th~ resistivity may vary sII!oothly.from point to point. In fact this

is generally the case. The question is the amQ...l!I'!!_~_thi~..y~atiQIl rather than any question

of its presence. Often, however, it is conventionally stated that the resistivity is constant

within some percentage and when the variation does in fact fall within this tolerance, it is

ignor:ed.

f ';oldered probe contacts may disturb the current flbw shorting out part of the sample and

.: add to the ambiguity in the measurement of the probe spacing. Soldering directly to the

body of the sample can affect the sample properties by heating effect and by

. contamination unless care is taken. These proble~s can be avoided by using pressure ~

-
contacts. The principal draw backs of this kind of contacts are that
' ­,
--
they may be noisy:
. ~

~
The current through the sample should not be large enough to cause heating. A further

precaution is necessary to prevent 'injecting effect' from affecting the measured value of p

Even good contacts, to germanium, for example, may inject. This is minimized by keeping

the voltage drop at the .contacts low.(If the surface' near the current contacts is rough

(1apped 'sulface) and the electric flow in the crystal is low, these injected earners will

• recombine before reaching the measuring probes)

"
~~~~~
 I.' (Z::3 J
jt
(5)
. ~i"!"
,"._~_~_u,,,,,,,,,, ....\:ilIiIIiiiii~ I ".

I. I

(!:J-] ~ {JPJ :t - ~
~ (.g)
15­ I? - Q.... (/'r'j - 4..)

( )~)"'b)

bandgap nuke conduction in metal almost independent of temperature, as compared to

9 ....
semiconducturs. Fig. 4 summarizes the above points,
1-­
.:3­ The that det.:trulls call be fuulld ill energy bands withill a crystal alld ullll:l

r,lf':!

1/ 'Z ) concillSIl)I1" mentioned so far can be shown by quantum mec.hanical calculations also

(/f) whIch we :,11;111 not carry out here. but will use the results for I./uanlllativc J.nalysis.

~-,i
tf') CONCENTRATION OF INTRINSIC CARRIERS
Il-)_

llle l'oncentration of intrinsic carriers ie. the number of electrons in conduction band

per unit volume is given by the expression :*

n:: ( me kT) 2 exp (/l - Eg) I kT

? (I)
2n h­
and the concentration of holes in valence band is given by the expressions·
"
;

P =o!- (mn k T)2

21th2 exp(-/l/kT) ~ (2)

If we mUltiply together the expression for ~ and p to obtain equilibrium relation:

3 I

kT ) 3/2 '

(3)
np=4 ( 21tX' 2 (memh) exp(-Eg/kT)

This does not involve the Fermi level /l and is known as the expression of iaw of mass

nction.

Where me = Effective mass of an electron

mh = Effective mass of a hole

j
k = Boltzamann's constant, /l = Fdrmi level

I
i Eg == Bandgap, T == Temperature in K
f In case of intrinsic (highly purified) crystals, the ~urnber' of electrons is equal to the
; mlmoer of holes, -because the thermal excitation of an electron leave behind a hole infhe
'" valence band. Thus, from(3) we have, letting the subscript i denote intrinsic
/
kT
( 314 )~ (4)
I ni = Pi =2 2n )i2 (me mh) exp (-E g 12kT)

I
I
!
E/2kT.
Thus, we see that the concentration of intrinsic carrier depend exponentially on

· 'Introduction to Solid State Physics' I Forth addition, chapter Xl, C. ~~el

I
I
I

I 3

1
I
IlI.:
'.
 ~,

y
1$;(
Cif

111111
,lIld

Foun Pi

• SEMICONDUCTOR scm i C II Il\!

AI:;,) 111,'1
I
An t~YI:I'S.

t modulate

\ / 'C!
\ I offcrs sc\
" '- a wide"

--- . //
/
of S(,nllCl
determine
Thj~ Illl't
FIG.5 MODEL FOR THE, FOUR PROBE RESISTIVITY I m3tt:li~d;.-,
MEASUREMENTS

BASIC S

II
prob·:s ,I

i! II ( ,;;!:: t I
pair. 11 t!:
the S\;llli c
carrier
mechanic:
Til
GALVANOMETER
i 6. A !lOll
POTENTIOMETER'
distance (
currenl 0 f
MILLIAMMETER I bLi.IVOLTMETER
01 REC T ;::J-.(7)f----r-r-, In
CURRENT
SOURCE
- f
PROBtS
necessary
i. Thl~ rc
2. If ther
> ciectrc
condw
FIG.6 CIRCUIT USED FOR RESISTIVITY r.-~EASURE II1ENTS

3. The ftl

i
~-~~.

*
~~ I

CONDUCTIVITY OF INTRINSIC SEMICONDUCTOR

1'It: electrical conductivity will he the slim of the cnnlrihlltlnns of both electrons and
: 11\,1\"
r I.'
(5)
(ni.qlc + pi.c.ph) -
Where e is the electron charge. pe and ph are the aver~ge velocities acquired h;.; the
elL'Ltllll!~ holes 111 it unit cll.!L:tlll,; fidd amI known as mobilities.
.i
,.: (j = c ni (~.m !- ph); Since ni = pi
~
. (K) T3 12 (fln + pp). exp2kT

"sing Eq. 4 where K is a constant.

,
~ (6)

T~1e factor T3/2 and the mobilities change relatively slow WIth temperature compared
"I.·ith ',h'.~xponential term, and hence the logarithm of resistivity p (=1/0) varies linearly with
1T The width of the energy gap may be determined from the slope of the curve.
Thus we have,
.E
Log p = -.-g- -log K (7)
_e 2kT e

EXPERIMENTAL CONSIDERATIONS
1. High resistance or rectification appears fairly often in electrical contacts to semiconductors
and in fact is one of the major problem.
I 2. In smgle crystal material thl! resistivity may vary s~oothly Jrom point to point. In fact this
is generally the case. The question is the am~'!Ilt of!f1~_~~jQJ1 rather than any question
of its presence. Often, however, it is conventionally stated that the resistivity is constant

II..{ within some percentage and when the variation does in fact fall within this tolerance, it is
ignored.
,;oldered probe contacts may disturb the current flow shorting out part of the sample and
add to the ambiguity in the measurement of the probe spacing. Soldering directly to the
body of the sample can affect the sample properties by heating effect and by
contamination unless care is taken. These problems can be avoided by using pressure
cOJltacts. The principal draw backs of this kind of contacts are that they may b~, !loisy.
4. The cllrrent through the sample should not be large enough to cause heating. A further
precaution is necessary to prevent 'irijecting effect' from affecting the measured value of p
Even good contacts, to germanium, for example, may inject. This is minimized by keeping
the voltage drop at the .contacts low.erf the surface' near the current c;ontacts is rough
(lapped surface) and the electric flow in the crystal is low, these injected carriers will
recombine before reaching the measuring probes) I

.\;..
. i

'" I
'; JJ
--.~.~.'
' I"':"' "

TT
\

t, ,
,,:n

The four probes used for resistivity measurements contact the ~n.lrrace at points Ihat lie
in a straight line.

, The diameter of the contact between the metallic probes and the semiconductor should
be small compared to the distance between probe!:> '
II
6. The boundary uctwccn tbe currcnl-c;lrrying electrodes ,H1(1 the bulk matenal I~
hemispherical and small in diameter.!
:7. The surfaces of the semiconductor crystal llIay be' cltilcl lUllduCll (If III \ II

conduct1llg.
a) A conducting boundary is one on which a material of much lower resistivity
than semiconductor (such as Copper) has been plated.
b) A non-conducting boundary IS produced when the surface of the crystal is in
contact with an insulator.

CA.SE 1 - RESISTIVITY MEASUREMENTS ON A LARGE SAMPLE

One added boundary condition is required to treat this case namely, that the probes
are far from any of the other surfaces of the sample and the sample can thus be considered
a semi-infinite volume of unifonn resistivity material. Fig. 5 shows the geometry of this
case. Four probes are spaced SI, S2 and SJ apart. Current r is passed through the ollter
probes (1 and 4) and the floating potential V is measured across the inner pair of probes 2
and 3.
The floating potential V f a distance r from an electrode can-ying a current I in a
material of resistivity PO is given by

Vr = - ­
Pol
21t r
In the model shown in Fig. 5 there are two current-carrying electrodes, numbered I
and 4, and the floating potential V f, at any Y point in the semiconductor is the difference
between the potential induced by each of the electrodes, since they carry currents of equal
magnitude but in opposite directions Thus:

PoI(1 1)
V( = - - - - - (9)
21t rl r4

Where r1 = distance from probe number 1.

r4 = distance from probe n~ber 4.

The floating potentials at probe 2, Vfl. and at probe 3, Vf3 can be calculated from (9)
by substituting the proper distances as follows :

.V Po I ( 1 1)
f.l. = 21t ,g;-- 8 2
+8 3

,~ 1­
 .,

Since p is independent of current, it IS possible to detemline whCill!.!r or not any of

these effects are interfering with the measuring of p at several values of I II should be kept
III Illind rh:ll th(':;c point:; of '~'(pcrimcnt;ll Icrhninllr. .lff,·ct ~·!;$~Ilti.dlv..ll! i!'" mt::lsurcmcnts
,U}d !lot tbe resistivity measurements only.

FOUR PROBE METHOD

;\Iany conventional methods for measuring resistivity are un:",tisfactory for
:;,:miconductors because metal-semiconductor contacts are usually recti I:. ing in nature .
.\hu there is generally minority carrier injection by one of rhe current c;wying contacts.
,\1\ I!xcess concentration of mmority carriers will affect the pult:utiai uf Ulllt..'r contacts and
;r:odulate the resistance of the material.
The method described here overcomes the difficulties mentioned1bove and also
. i

() i'fers several other advantages. It penni ts measurements 0 f resistivity In ',lmples having

a wide variety of shapes, including the resistivity of small volumes withln bigger pieces
Ill' ~cnllconductor. In this manner the resistivity of bo~h sides of p-n Junction can be
detcnnined with good accuracy before the material is c~t into bars for making devices.
method of measurement is also applicable to silicon and other semiconductor
materials.

DJ\SIC SCHEME
The basic model for all these measurements is indicated in Fig. 5. four sharp
prohe's nrc plnced on a flat surface of the material to be measured, current is passed
the two outer electrodes, and the floating potential is measured across the inner
[I' the flat surface on which the probes rest is adequately large and the crystal is big
the semiconductor may be considered to be a semi-infinite volume. To prevent minority
carrier injecti<?n and make good contacts, the surface 011 which the probes rest, maybe
mechanically lapped.
I The experimental circuit used tor measurement is illustrated schematically in Fig.
j 6. ,\n'Jrninal value of probe ~pacing which h~s been. found satisfactor~ is an equal
, dlslailce of 2.0 mm retween adjacent probes. ThIS pennlt measurement with reasonable
I current of n-type or p-type semic~md!lctor from 0.00 I to 50 ohm. em.

I In order to use this four probe method in semiconductor crystals or slides it is

necessary to assume that :
1. The resistivity of the material is uniform in the area of measurement.
2. [f there is minority carrier injection into the semiconductor by the current· can-ying
electrodes most of the camers recombine near the electrodes so that their effect on the
/ conductivity ..is negligible. (This means that the measurements should be made on
surface which have a high recombination rate, such as.mechanicallapped surfaces).
J rhe slIJfa(;c on which the probes rcst is flat with no surface leakage.

~
5
.f
I
1
 Po I ( 1 __
1)
V(3=~ SI+S2 S3

'The potent.inl difference V between probes 2 and 3 IS then

t. V::: Vf2 - Vo = -Allll
-+--
2n: SI
I I
S3
---­
S2+S) SI+S2

I)
aIl d the resistivity Po is computable as
tc
, ro =,-
V
I I
27t

SI + S) - SI +S2 ;- S2 +S3

I I)

( 10l

When the point spacing equal"that is, $1 = S2 = S3 = S the above simplifies to .

~
'S /·~v~----~ .
Po = - X 2n:S, ( II)
,-----1 •
1\:
e 1- RESISTIVITY MEASUREMENTS ON A THIN SLICE·CONDuCn,Y(1
Z 1. 'AS BOTTOM SURFACE.
I

& TWO boundary conditions must be met in this case; the top surface of the slice'
be a reflecting (non-conducting) surface and the bottom surface must be ~1Il
· he I

ust bing (conducting) surface. Since the two boundaries are parallel. a solution by the
bs or d of images required for each current source an infinite series of images along it
ling I

, d of

. eth°ortllal to the places and passing through the current source.

me I1
·muaJ
The model for this case is shown in Fig. 7. The side surface of the slice an:
J'Iled to be far from the ,area of measurement and, therefore, only the effect of the
: The
sU til surface needs to be considered. In this analysis equal probe spacing S shall be
!d to
: I lIut.
ott~ed. The width of the slice is W. The array of images needed is indicated in Fig. 7.
sU .. the polarity and spacing of the first few images are as shown.
~d.1 ber",

!tbe p The floating potential V t2 at electrodes 2 is

· The ,I
:=.eJ.[ ~ (_1)"
I
,
v
f2 21t "=-aJ JS 2
1
+ (2nW) 2
- I
n=-<lO
(_l)n
J(2S) 2
1
+ (2n W) 2
1 (12)
. : Va .
~

Likewise, the floating potential at electrode (3) can be obtained and

~ikelij!
!: r ~~4 p~
V == E..!l.!.+ 2::<-1)" I • 2 - L(-I)" 4 ] (13)
{,J 21t S n-l ""S2 +:(2nW) n=1
r
The resistivity then becomes
i he fI

7
.
J . !I

ijj )..t-

 ':Fl,;'':'.-
,
, I~I:W:. ,_

0,

"

.
.....
\
' +1 -/
n _ +2 0

n '" ... 1
-I
III
+~+ 2W
Vr:

Tht
I 2 3 " fOP SURFACE

,., .. 0 .'0 0 0 -10 } - (NON.CONDUCfIN(,)

{ S ~ s 4 s J S LlC E ' v,
• _ BOTTOM SURFACE
(C ONOUCTlNG)
,., .- \
-I
o
+1
o-f-­ ann
1-.-- H --1 2W
f
P.J:
n =! 2
+1
(I -~~
,,..
FIG.7IMAGES FOR THE CASE OF THE; RESI~)TlVITY PROBES
Wh:::>
ON A THIN SLICE WITH A CONDUCTI 'lJG BOTTOM

SURFACE
Po'

!C4SE 1­
'-0 --- r-- 01

0·7 t--· ---

.. -
~ --. Tw(
ust be a
0-5 v
L bsorbing I
elhod of
0-' 1
0·)
I ,' - t---
le norm
V I The
0-2
I
I sumed to
I, SIS I S.~
Gs(W/S) V ~ ~ ~:~
I OUom sun
surned. T
0- ,
I L , here the p

0·07
/ T "i- ,,,,, '.' .,,'
'II
~~ I
The I
••
0'05
II ONDUCTING BOUNDARY
I _ I
V12
0·04

0'0) ,.
I
/
,
I Like!
0'02

I V~I
0·0'
0·' 0·2 0-3 0',(,0-5 0·7 "0 2 .I
- " 5 7 10
, ~

W/S The!':

FIG. 8 CORRECTION DIVISOR FOR PROB;:~; ON A SLICE

WITH A CONDUCTING BOTTOM SURF ACE

,
 -
T..
-,:I,;,I
,

Po
( i -l \
P=06(W/S)

Where resistivity po j)) l:omputablc from t, 10, and 11) can be used if 11lL' 1")11;: sp;lCing
e di~ fare different, but approximately cqll:J1. The function Gb (W/S) is \.()IIlPlItCti fhllil

'W') S n~'~
I

-j( ir :~~;,
G 6(S) = I +4W L:>-I)1I l i
n.. 1
S ):1 + (211)2
"( W 2

which is tabulated in Table I and plotted in Fig. 8,

\.St, ITABLE I
- --­

S. No. W/S G6 (W/S) G 7 (\VIS)

f
1 0.100 0.0000019 13,863
r 2
\

0.141 0.00018 9,704

3 0.200 0.00342 () 'nl
CH2..~ $/S'O
4 9.)3 0.0604 4 159
I
)! 5 0.500 0.228 2,780
C I
6 1.000 0.683 I )0<-1
);

, I
7 1.414 0.848 1.223
:, 8 2.00p 0.933 1.094

'1

9 3.333 0.9838 1.0228

.,"
;. itl . 10 5.000 0.9948 1.0070
i

'1 11 10.000 0.9993 -_ ­ 1.00045

... _.. I

CASE 3 RESISTIVITY MEASUREMENTS ON A THIN SLICE-NON-CONDUCTING

ISEI
BOTTOM SURFACE.
\
i
The mOdel for these measu~ements is like the case 2, except that the bottom surface
of the slice is nonconducting. This means that all the images of Fig. 7 have the same eharge as
;' j:he~
. Ji, cu~
the current source. Thus all the images on a row have equal charges and Eq. 13 describes the
potential difference across the inner pair of probes, if (·1)" is removed from the equation.
,enti1 Then,
~ ,,1 ~n,.. I
Po· 'f;/I
( 16)
vV P =a;-(Vi1S)

ihc

ij , ~

,f!4,
 "iff SNliih'!t

F"
aret

,,­

100· ,

10,
1
,

,SI,SI S ,
50'0

40·0
,-­

lo·n
o 0 0 0 .1 TAlIl

20·

G7 (W/S)
if ," T
NON.,CONOUCTfNG
w
-
~
80U~OARY

, 0·0

7·0

5,0,

4-01

)·0I

"" -
~ ,.
",
,

lIit
r

-/
.
-­
~
2'0I

'·0
~ .. '"" j~--
"
'
i

.'
I1
I

I,
I
·1
0·1 0·2 0·3 0-'0·5 0-7 1·0 2 3 , 5, 7 10·0

W/S

CASEI
,j
FIG.9 CORRECTION DIVISOR FOR PR)BES ON' A THIN ~
SLICE WITH A NONCONDUCTIIIG BOTTOM SURFACE
oftht
the~1
pot !
Tnen,;
'}
'~

,
I'

//

t~ ·
 Where,

G 7 (W/S)-1+-4-L/'
s ,,-J I
W I
(S'\ ,-
" '1 ) S'J ~ I
VI w J +(11)2 \1 - \'.'
,y + (111)' :
I
This fum;tioll G7(W/Sl is tabulatt:4 III Tabk I alld plOllt:d III Fig. <) hll ~Ii, '. ,1111'-'~
1 of W/S the function Ul (W/S) approacl!\!sth\! case tOI all Illlillltl!ly tlllli ',lilT, \11
f WISt
w'l ')S
.>'./ G 7 l-I:":'-tO!!c 2 ( .,
. --" •. S) W. - .
BRIEF DESCRIPTION OF THE APpARATUS REQUIRED
RlfFI
1. Probes Arrangement: !l has four individually SprIng loaded probes. Thl: ,!( Jiles ;m:
Pro! collinear and equally spaced. The probes ar'\! mOllnted in a tetlon bush, ,,\ hl<.'!l ','nSllre a
coiF good electrical insulation between the probes. A tetlon spacer n,7ar the ii;'> I~ ;tiso
good provided to keep the probes at equal distanc(;. 'the probe arrangement is nJuJ.:lled in a
prOYi suitable stand, which also holds the sample plate and RTD sensor. This st:wd aiSl' serves
suiti as the lid of Controlled Oven. Proper leads are provided for the current ;mJ voltage
as~
measurement.
meat
2. Sample: Ge crystal in the form of a chip.
SallJ
3. Oven : It is a small oven for the variation of temperature of the crystal from room
Ova
1 temperature to about 200°C.
te~ I 4. Four Probes Set-up : (Measuring Unit) - It has three subunits all enclused III Olll'
, FOill cabinet.
',l cahill
I
(i) Oven Controller
. n (i) 0 Platinum RTD (A class) has been Ilsed for sensing the tempera~ure. A wheatstom;
bridge and an instrumentation ampliiier are used for signal conditioning, Feedback
'1 ensures offset and linearity trimming and a fast accurate control of the oven temperate,
,I
. ·1 SPECIFICA TIONS
Temperature Range : Ambient to 473K

Resolution
: 1K
Stability
: ±O,5K
Measurement Accuracy : ±1 K (typical)

Oven
Specially designed for Four Probe Set-Up

Sensor RTD (A class)

Display 3Yz digit, 7 segment LED with autopotarity and decimal indication
power 150W

)i

 r'
~
;m

.
•• '
! .
,;
- .'

H"
• .'
II

'·0
I

-~ ........... .
........
-······ '
,:"
m...... :
.:. ~ l!:;;;::.
'.:.... .;...&:::::;.:
_ .... - .
 <. .....1

,I,

i
J. Place the Four Prohl.! Arr:lllt;l~!IlC-'1i In the oven and (,':lIleet theL'Il:;or ICln! 10 Ill ...' l~
connecter on the panel.
4. SWlt<.;h on the mains supply of I:llir Probe Sct-up ,\: j Plit th~ dlglt;.!l panel Illeler
the current measuring mode thrllllgh the sc\ector s\\ rch III tlk; r'Jsitioll LED ,.,ICI11~
rnA would glow Adjust the ctlrr:';;! to 0. deSired valuc Say:'i m. '. I
,.
5. Now put the dl~I!;]1 panel nH:kr In voltage nh':;15111: Ie IIlUJl' III this P,)"lliulI 1.1 I J
:I facmg mY would glow and the r;',::er would read the' )!tagt' OI.'!·,lccn the prohes
"I
'; I 6, (i) Switch ON the temperature Coptlt111er and o.pprox adJl:st the SCI-temperature, Tht: ~rel'lI
n LED would light up indic[lting tlk ,wen is 'ON' and th,~ 1l'l1lper;!I\I1'c \"Gulli ~t;lrt l'1SH1L'
t
,Ii
~ I Temperature of the oven in K would he indicat'cd by till' 11Pi\.!
Oi) The controller of ~the oven would switch off til;,!, power corresponding 10. :iC!

I temperature. In case, it is less then the desired temperattl'c, the sct tcmperat'yle silolllJ bl.:'
increased till the power is 'OELat :he desired temperatun', I f the sening of set-(emperJtur(.>~­
is higher than desired value, the s~t-temperature knob \\ould be moved back so tll;.!l rhl.'
LED is offat the de$ired temperature.

\, Because of thermal inertia of oven, there would be ~ome over shoot and under shun;
before a steady sct-temperature is <ltlHltlCO ami nwy take I" Il1ljllllc~ fill caeh Icadil1,I'
(iv) To save time, it is recommended to under adjust the sct-temperature, Example say it I~j
.
desired to set at 330K. Set the temDerature
, set l<11ob so that thel1-::D is Qrr at 3251( i il,:
temperature due to the thermal ihertia "Yould continue to rrse. At 330K adjl\.st the set·
temperature so that the LED is 'ON'."'l:t would automatically be 'OFF' when telllpcratlln: I',
Ii above ,:YOK. It may go upto I or 2, degree but would quickly settle down at 330K. "Since
I change in temperature at this stage is very slow and response time of RTD sensor and
semiconductor is comparatively fast. The reading can also be taken Ht any tcmperatlilc
without waiting for an steady state:

OBSERVATION & TABULATION

I' S.No. Temperature
S
Voltage / p
. rl x 103 LOg 1ll P
(T in K) (volts) (~~.)
.;.'

l 1<- ll
,-.-
'"
,
J
,I
. }
Distance between probes (S) "" '2- mm
Thickness of the crystal (W) ;::; o,~ mm

11
01" ~
'"

cia­ 2 L~'Y7
~ ,.;.

" .~.
 ~

0) ;\lullir::IIIJ!c f)jgital V"lrmctcr

Tn this unit, intcrsil 3y~ digl! single chip AID COrlvertcl leL 7107 ha::, k:.:n
Ilsl.!d. It b.ls m;curacy. auto zero to kss than 10 ~IV. zero linn·less than I pVi -'(',
input bias current of 10 pA and rollover error of less than one COllnt SII1CC the li';e
nf Internal rcfcn::ncc causes the ucgradal10n 'in performance dut.: to Internal heall ~

,Ill e\(ernal reference lI;lS been u::,t.:d

SPEC/FICA nONS
Range XI (O·200mV) & XI 0 (0·2V)
Resolution 100uV at X 1 range
Accur.1CY ! 0.1 % of reading ± 1 digit
Impodance 3YJ digit, 7 segment LED with autopolanty and deCIma/IndicatIOn
Ovur/odd Indicator Sign Of 1 on the left and blanking of other dIgitS.
Display . Sign of 1 on the: left & blanking of other digits.

(ii) Constant Current Generator

It is an IC regulated current generator to provide a constant current to the
outer probes irrespective of the changing resistance of the sample due to change
in temperatures.
The basic scheme is to use the feedback principle to limit the load current of
the supply to preset maximum value. Variations in the current are achieved by a
potentiometer included for that purpose. The supply is a highly regulated--and
practically ripple free d.c. source. The current is measured by the digital panel
meter.
SPEC/FICA TION
Open circuit volt~ge 18 V
Current range 0- 20 mA
Resolution 10 J.lA
± 0.25% of the reading ± 1 digit
'"
Accuracy
Load regulation 0.05% for 0 to full load
Line regulation 0.05% for 10% changes

EXPERIMENTAL PROCEDURES
1. Put the sample on the base plate of the four probe arrangement. Unscrew the pipe
holding the four probes and let the four probes rest in the middle of the sample. Apply
a very gentle pressure on the probes and tighten the pipe "in this position, Check the
continuity b~tween tlte probes for proper electrical contacts. .
CAUTION: The Gecrystal is very brittle. Therefore, use only the minimum pressure
required for proper electrical contacts. ..

2. Connect the outer pair of probes (redlblack) leads to the constant current power
supply and the inner pair (green/yellow) to the probe voltage terminals. :;.

~a .-.. -. *'

 /. J
<J
r,i

..,~
t,
C
/
\. )­
\ (':
'0
(",
c c
:r.
C
JJ

'..J
Z
=

o
.~
r-
C
':J .
'lJ;.f o
.- '""
.......

0:"
.:L.
0>­
c5 z u
"0 = :.J
c­ :J
:-:l 0
.... u
UJ
 !' .",,-zk..jo\fal~~~"'" '"' "'-~-""""""'-"'.,~ •.,." ~ ~- ,~'''''

.. J'
,~
If':

..

CALCUL·' TJON
From Eq. (II)
v/

rn = - x 2,'l'S

I '"
Since the thickness or the cryS[;]/ is small compared 10 the I JDe distance J correction
f:letor for it has to be applied. Further the bottom surface is non )nducting in the present
casc. Eq ( 1 7) will be appllcd.

r=Gd\\,lS)
t';' ,

The runction G1 (WiS) maybe obtained from Table-J or g 9 for the appropriate
\ alII\.: l)( (\\",'S)' Thus I) may bc calculated for various temperature
Plot a graph for Logl@P. vs rt X I
. E~

Using Eq. (7) Logcp = --- - loge K

. , ?
- k T ·I
\
\.. - - - 1'0 "'.;'- '--i. ­ .
the slope of the curve is given by -L- c ~ =::::$' (i 9)
1 2k!
I

T
Thus Eg may be obtained from the slope of the graph. Note th~lt log~ = 2.3026 Jogla and
thc Eq. (7) is applicabie only in the intrinsic region of the: semiconductor. A typical graph is
~
shown in Fig. 10. .

QUESTIONS 11

I. What is the advantage of Four Probe method ov~r the other convelltional methods?
2. Can we use an ordinary millivoltmeter instead of electronic millivoltmeter or
potentiometer to measure the inner probe voltag,e? Why?
3. Explain the behaviour ofthe IOglOP vs. lIT curVe.
4. Why a semiconductor behaves as an insulator at zero degree kelvin?

MATERIALS GIVEN HERE IS BASED ON

1. Introduction of Solid State Physics, by C. Kittel Nth edition.
2. Fundamental of Semiconductor Devices by J. Lindmayer and C Y. Wriegley, 1071 East
West Press.
1. 'Resistivity Measurements on Gennanium for Transistor' by L.a. Valdes, Proceedings of
the IRE. Vol. 42, p. 420 (1954),

~ L "­
12

!.: ~ ...