Mössbauer effect and perturbed angular correlation measurements on

Hyperfine Interactions 4 (1978) 727-731 North-Holland Publishing Company

~OSSBAUER EFFECT AND PERTURBED ANGULAR CORRELATION MEASUREMENTS M. Van Rossum,

ON 129mxe IMPLANTED

IN DIAMOND

J. De bruyn, G. Langouche,

M. Rots,

H. Ooms, J. Claes, F. Namavar and R. Coussement

Instituut voor Kern- en Stralingsfysika,

University of Leuven, Belgium.

1. INTRODUCTION In contrast

to most other elements of the 5s-p series,

haviour of Xe into diamond type semiconductors ed.

A few channeling

ed 1,2

In addition,

effect measurements ~6ssbauer

on Xe implanted

experiments

an unspecified

in Ge

effect and M~ss-

position. a search for the lattice location of Xe implanted

in diamond seemed quite interesting. in which a combinatlon

lar Correlations

Both channeling

agree on the fact that the Xe atoms occupy in these lattices

interstitial

In view of these results,

vestigation,

in Si have been report-

on 129mxe and 133Xe implanted

and Si have been performed by the authors 3. bauer measurements

the implantation be-

has only scarcely been investigat-

(IPAC)

We present here the conclusions of M~ssbauer

Effect

of this in-

(ME) and Perturbed Angu-

techniques has been used.

2. EXPERIMENTAL PROCEDURE 127i. The ;29mxe activity (tl/2 = 8d) was produced by neutron irradiation of The diamond samples were obtained

from Diamkaap N.V., Antwerp;

type IA(R > 1016~ cm) and IIB(R = 105~ cm). implantation;

in some cases a more elaborated

The implanted an accelerating 1014 at/cm2;

we used samples of

All chips were degreased prior to etching procedure 4 was applied.

sources were prepared

voltage of 70 kV.

all implantations

Most ME measurements

at the Leuven Isotope Separator using 13 The implanted doses ranged from 2.]0 to

were done at room temperature.

were performed

cryostat and a conventional

at LHe temperature using an exchange gas

electromechanical

drive system moving the source.

absorber we used Na4XeO 6 . 2H20 (obtained form P.C.R., The detector was a Xe filled proportional

counter;

Gainesville,

As

Florida).

the energy window was set on

the 10 keV and 6 keV escape peaks of the 39.6 keV M~ssbauer y-ray. Diamond

single crystals of type IIB were used for the IPAC measurement.

196 keV - 39 keV y-y cascade was observed by detecting of the 196 keV M 4 transition.

the conversion

The X-rays were separated

sition by using a Si(Li)detector.

The

elec[rons

from the 39 keV y-tran-

With a fast-slow type of coincidence

the correlation

of the cascade transitions was measured on five angles.

3. EXPERIMENTAL

RESULTS

set-up

3.1. ME results A typical M~ssbauer shown on fig.

I.

spectrum recorded from a 129mxe(diamond)

source is

In view of its complex nature, at least 3 Lorentzians

727

had to be

M. Van Rossum et al./129mxe in Diamond

728

included in the fitting procedure two situated around ~ 7mm/s. to be almost the same. ed samples

:

:

one peak near zero velocity and the other

The intensity of the three components turned out

Several measurements were performed on different implant-

the shape of the spectrum was found to be insensitive to the dia-

mond species (IA or IIB) or to the surface treatment.

Moreover,

a measurement

performed at 77 K did not show any change in the relative intensities of the components.

Table I shows the relevant parameters for a typical run.

In order to check for a possible graphitization of the diamond surface during implantation, was also recorded.

a M~ssbauer spectrum of 129Xe implanted in a graphite sheet The result shows a single line, the parameters of which have

been included in table I.

N+IO 3 60-

5958-

575655545352- 129m-~ .diam 51 5c

-~o

-~'o

6 - Fig.

host

positions

diam IA

(mm/s)

1

.Ib

~o v(mrn/s)

-

linewidth

(mm/s)

% effect

-7.27 • 0.42 -0.04 • 0.60

13.5 • 0.5

5.5

II.0 -+ 0.3

4.3

+6.92 -+ 0.38 graphite

-0.28 • 0.06 - Table I -

3.2. PAC results The experimental correlation pattern shows a significant deviation from the unperturbed case :

this means that a clear evidence for the existence of a qua-

drupole interaction at the Xe site is observed.

Assuming the electric field

gradient to be oriented and axially symmetric, which is the simplest case, one 5 can prove that the time integrated angular correlation function has the form :

M. Van Rossum et al./129mxe

in Diamond

729

2 W (@, ~) = ] + A22(aO + ~l cos

@ + ~2cos @ s i n @ )

In this expression, @is the angle between the detectors and the coefficients depend on the two orientation angles of the field gradient and on the perturbation factor G = (1 + (~OT)2) -I For the 129Xe case, we measured the angular correlation for two different positions of the diamond chips. from which G could be extracted

Then, two sets of ~. coefficients were obtained z :

G = 0.29 • 0.03 4. DISCUSSION The interpretation of the Mossbauer spectrum is by no means straighforward. To start with,

let us concentrate on the outer two lines.

Since the spin se-

quence is 3/2 § ]/2, these might arise from a single quadrupole split level or from two levels separated by a large isomer shift (IS) difference.

We will first

examine the second possibility. The presence of two lines with different IS's should be interpreted in relation with the corresponding electronic densities. several chemical compounds containing

An energy scale based on

129Xe has been built up and is shown in

fig. 2.

129Xe

Isomer Shifts sp3 s~ps -r

-o~

6

,~

o2

o~

6s

6~

vetodty (minis) , electron density - Fig. 2

-

As one can see, all known IS's are situated between + 0.7 and - 0.2 mm/s our two peaks are lying far off this range.

:

The peak at the high-energy side

could eventually arise from a very compressed atomic state of the Xe impurity. However,

it is much more difficult to justify the position of the second peak in

the same frame : ty at the nucleus.

in this case, one has to explain a very low electronic densiA formula has been given by Perlow 6, which relates the ob-

served IS "6" of a given 129Xe compound to the number of s an p holes h in the outer shell of the Xe atom in that Particular configuration

:

s

and h

p

730

M. Van R o s s u m et al./129mxe

= K~hs+ with

u

in Diamond

+ hs)(2-hs)

P

~__~_6= -0.79 mm/s

~h S

-

= +0.13 mm/s

-

~h

P

From these numbers

it is clear that an IS of -7 mm/s would correspond with

a number of stripped electrons far too large than can be expected for an implanted atom.

We are therefore

for the observed M~ssbauer drupole interaction

inclined

to reject this hypothesis

spectrum.

The remaining

(QI) at the Xe location.

as an explanation

assumption

is that of a qua-

Following

this interpretation, the eQVzz splitting of the outer lines corresponds with an average QI strength of - ~ - - =

410 • 60 MHz. The presence of a QI is attested by the IPAC measurement, is ambiguous,

since the value one gets for the interaction

the assumed population of the sites with and without QI.

but this result

strength depends on Assuming a unique im-

plantation site, the perturbation factor G corresponds with a QI frequency of eQVzz 172 • 18 ~ z . This could only cause a line broadening in the Mossbauer 2h spectrum. However, assuming two sites respectively without any QI and a site with an interaction of 4]0 MHz, as given by the M~ssbauer relative population

of 20 to 30 % for the zero interaction

result, one obtains a site, in rough agree-

ment with the numbers deduced from ME experiments. In order now to explain this unusual QI, we should look for an implantation site with a non-cubic gonal interstitial site is unlikely

symmetry.

one.

The only regular site of this kind iN the hexa-

However,

it has been pointed out by Weiser 7 that this

to be populated by big impurities

hedral interstitial

: these will prefer the tetra-

site because of the larger volume available.

location of the Xe impurity,

the tetrahedral

e.g. by the presence of one or more vacancies

If this is the

symmetry should be somehow broken as nearest neighbours

to the Xe.

This assumption appears very plausible,

since the nn distance at the tetrahedral

site in diamond

the radius of the free Xe atom is about

2.2 ~.

is only 1.54 ~, whereas

A location at a vacancy associated

site should therefore be energetically

much more favorable in diamond than e.g. in Ge, where the nn distance and the Xe radius are of comparable magnitude. Finally, absorption

we should also discuss

line.

the location corresponding

precise value for the IS can be deduced, M~ssbauer

with the central

It should arise from a site with nearly cubic symmetry,

spectrum.

because of the poor resolution of the

This site could be the undisturbed

but we must then assume a considerable

but no

tetrahedral

interstitial,

distortion of the lattice surrounding.

Taking into account the damage created during ion implantation,

an unspecified

M. Van Rossum

et al./129mxe

in Diamond

73]

location in a disordered environment is therefore more probable. ACKNOWLEDGMENTS The authors wish to thank J. 0deurs, R. Vanautgaerden and H. Pattyn for the implantation work. They also wish to thank Prof. I. D@zsi for some helpful comments. REFERENCES I.

J.A. Davies, J. Denhartog, L. Eriksson and J. W. Mayer, (1967), Can. J. Phys. ~5, 4053.

2.

L. Eriksson, J. A. Davies, N.G.E. Johansson and J. W. Mayer, (1969), J. Appl. Phys. 40, 842.

3.

M. Van Rossum, J. De bruyn, G. Langouche, R. Coussement and P. Boolchand, (1976), J. Physique 37~ 66-889.

4.

L.A. Davidson, S. Chou, J. F. Gibsen and W.S. Johnson, (1971), Rad. Elf. ~, 35.

5.

M. Rots, H. 0oms, J. Claes and F. Namavar, to be published.

6.

G.J. Perlow, Chemical Applications of M~ssbauer Spectroscopy, eds. V.I. Goldanskii and R.H. Herber, (Academic Press, 1968) p. 422.

7.

K. Weiser,

(1962), Phys. Rev. 126, 1427.