Electron paramagnetic resonance in low-temperature electron-irradiated diamond

HAL Id: jpa-00208607

https://hal.archives-ouvertes.fr/jpa-00208607

Submitted on 1 Jan 1977

HAL

is a multi-disciplinary open access archive for the deposit and dissemination of sci- entific research documents, whether they are pub- lished or not. The documents may come from teaching and research institutions in France or abroad, or from public or private research centers.

L’archive ouverte pluridisciplinaire

HAL, est

destinée au dépôt et à la diffusion de documents scientifiques de niveau recherche, publiés ou non, émanant des établissements d’enseignement et de recherche français ou étrangers, des laboratoires publics ou privés.

Electron paramagnetic resonance in low-temperature electron-irradiated diamond

P.R. Brosious, J.W. Corbett, J.C. Bourgoin

To cite this version:

P.R. Brosious, J.W. Corbett, J.C. Bourgoin. Electron paramagnetic resonance in low- temperature electron-irradiated diamond. Journal de Physique, 1977, 38 (5), pp.459-462.

�10.1051/jphys:01977003805045900�. �jpa-00208607�

ELECTRON PARAMAGNETIC RESONANCE IN LOW-TEMPERATURE

ELECTRON-IRRADIATED DIAMOND (*)

P. R. BROSIOUS

(**),

J. W. CORBETT

(***)

Physics Department,

^State

University

^{of New York at}

Albany, Albany,

^{N. Y.}

12222,

^U.S.A.

and

J. C. BOURGOIN

Groupe

^de

Physique

des Solides de l’E.N.S.

(****),

Université Paris

VII,

^Tour

23, 2, place Jussieu,

75221 Paris Cedex

05,

^France

(Reçu

^{le 4 octobre}

1976, accepté

^{le 20}

janvier 1977)

Résumé. ²⁰¹⁴ Un diamant synthétique dopé ^{au bore a} été irradié à 77 K avec des électrons de

1,5 ^MeV. Après irradiation, ^{on observe un}

signal isotrope composé

d’une résonance centrale à g ⁼ 2,00 ^{et d’un} spectre (A 7) à ^deux composantes décalées de 60 G de chaque côté de la résonance centrale.

Après

guérison ^il apparait un nouveau spectre

(A

8) à ^deux composantes décalées de 33 G de chaque côté de la résonance centrale. On a étudié la variation de l’amplitude ^{de ces} spectres ^{avec la} dose et le recuit

(jusqu’à

³³⁰ K). ^Le comportement

pendant

^la guérison ^est

comparé

^avec des résultats obtenus par des ^mesures de conductivité déjà

publiées.

^{On en} déduit que les spectres ^{A 7 et A 8}

sont associés à la désexcitation thermique ^des

pièges

présents ^avant irradiation et que l’étape ^de guérison de la résonance centrale à 250-300 K est associée avec la

guérison

des défauts introduits par irradiation. Les défauts associés à ces spectres n’ont pas pu être identifiés.

Abstract. ²⁰¹⁴ A man-made boron

doped

diamond has been irradiated at 77 K with 1.5 MeV elec- trons. After irradiation an

isotropic

signal is observed which is

composed

^{of a central resonance at}

g ⁼ 2.00 and of a two line spectrum (A 7)

split symmetrically

^{60 G from} the central line. After suffi- cient

annealing

^an additional two line spectrum (A 8) split

symmetrically

^{33 G} from the central line appears. The variation with dose and with isochronal annealing

(up

^{to 330} K) ^{of these} spectra ^is measured. Their

annealing

behaviour is correlated with previously

published

conductivity ^measure- ments,

demonstrating

that the A 7 and A 8 spectra ^are associated with the thermal deexcitation of traps present before irradiation and that an

annealing

stage ^of the central line at 250-300 K is asso-

ciated with the recovery of irradiation-induced defects. The defects associated with these spectra have not been identified.

Classification

Physics ^Abstracts

8.630 ^- 8.634

1. Introduction. ^- There have been extensive studies on

diamond,

^but

relatively

little has been done on

samples

irradiated at low

temperatures.

Conductivity

measurements

performed

ⁱⁿ

synthetic boron-doped diamonds, following high

energy elec- tron irradiation at - 12

K,

have shown that most of the defects which are created recover around 270 K

[1].

Because the defect creation rate is of the order of

(*) ^Research supported ⁱⁿ part by the Office of Naval Research under contract N00014-70-C-0296.

(**) Present address : IBM Research Laboratories, ^Yorktown Heights, ^{N. Y. 10598.}

(***) John ^Simon Guggenheim Memorial Fellow.

(****) Laboratoire associé au C.N.R.S.

the calculated one and because no defect _recovery is observed below 270

K,

it has been

proposed

^that

these defects are

vacancy-interstitial pairs [1] (the

interstitial

being

associated with ^a donor level situated at 0.05 6V below the conduction band

[2]). ^Stages

in the _recovery of the

conductivity (at 50, 160,

²³⁰

and 300

K)

associated with the thermal deexcitation of carriers from traps ^{which are}

present

before irra- diation or which are created

by

the irradiation

(50

^K

stage)

^are also observed

[1, 2], indicating

^{that these}

traps

undergo charge

^state

changes during annealing.

Electron

paramagnetic

^resonance

(EPR)

^measure-

ments in low

temperature

electron-irradiated diamond have been

reported by

^{Lomer and Wild}

[3]

^{and Lomer}

and Welboum

[4]

ⁱⁿ type IIa diamonds. Their measu- rements demonstrated that

charge

redistribution was

Article published online by EDP Sciences and available at http://dx.doi.org/10.1051/jphys:01977003805045900

460

induced

by

irradiation

(because

of the variation of

intensity

^with irradiation and

annealing

^{of EPR}

lines present ^before

irradiation); they

also found a new spectrum ^after

annealing

^{at 80}

K,

^which

disap- peared

^{at 140 K.}

This _paper reports ^an attempt ^to correlate the

conductivity

measurements

previously performed

with electron

paramagnetic

^resonance measurements,

during

isochronal

annealing.

^{The electron} irradiation

was

performed

^at

liquid nitrogen temperature,

^well below the temperature ^{at which the} recovery of the defects

(observed using conductivity measurements)

occurs.

In the present

study

^the

sample

^{used is a semi-}

conducting boron-doped diamond,

grown

by

^General

Electric,

^{similar to} the diamonds used for conducti-

.

vity

measurements. This

sample,

ⁱⁿ which the boron concentration

[5]

^{is 3.5 x}

¹⁰¹⁷ cm-3,

^was

synthe-

sized from

a l3C

enriched

carbon powder

^{so that it}

contains a

percentage

^of

13C (spin 1/2) greater

^than the natural abundance

(1.1 %);

^{this was done in}

order to search for

possible hyperfine

interaction.

Prior to the

reported experiment

^the

sample

^under-

went an electron irradiation

(with

^{a dose of}

1.3 x

1016

electrons

cm-2

at 1

MeV)

^{at 77 K and was}

subsequently

^{annealed to room}

temperature;

^{the elec-} trical

conductivity

after irradiation and

annealing (10-3

^Q-1

^cm-’

^{at room}

temperature)

^returned

practically

^{to its}

original

^value

(0.6

^x

10- 3 Q - 1 cm - l).

2.

Experiment.

^{- The}

sample

^{was studied at 77 K}

with an X-band

superheterodyne

^EPR

spectrometer using

^{a TEO 11 mode}

cylindrical

^brass

cavity

^mounted

in a

cryostat.

^Low

frequency (87.5 Hz), pole-face

modulation was used. The

sample

^was

glued

^with

epoxy

[6]

^on the end of a

long

stainless steel rod around which was a nichrome heater wound used for

annealing.

^{The outer tail}

piece

and the central tube of the cryostat ^were fitted with

beryllium

^windows

so that it was

possible

^{to move the}

sample

^{from the}

cavity

up ^to the central tube

(in

front of the Be win-

dows)

for irradiation and back down to the

cavity

for EPR measurements, ^{while the}

sample

^{was main-}

tained at 77 K.

The SUNYA

Dynamitron

accelerator was used to irradiate the

sample

^{with 1.5} MeV electrons to a

total fluence of 7.5 x

1017

electrons

cm- 2.

The EPR

intensity

^{was measured at 77 K} after doses of

2,

5.5 and 7.5 x

1017

electrons

cm-2

and then the

sample

^{was annealed in 8 K} steps

(for

^{10 min.}

each)

from 77 to 300

K;

^{the EPR}

intensity

^{was measured}

at 77 K after each step.

3. Results. ^- 3 .1 DESCRIPTION OF THE SIGNALS. ^-

Prior to irradiation no EPR

signal

^{was observed in the}

sample.

After irradiation a

signal

^{was observed}

(Fig. 1), composed

^{of a central resonance} at g ^{= 2.00}

(having

a linewidth of 36

G)

^{and a two-line} spectrum

(desi- gnated

^{here as the A 7}

spectrum) split symmetrically

FIG. 1. ^- Central line and A 7 resonances observed in first deri- vative dispersion ^mode.

FIG. 2. ^- Microwave _power saturation of the central line measured at 77 K.

FIG. 3. ^- Central line, A 7 and A 8 resonances observed in first derivative absorption ^{mode after} annealing around 180 K.

60 G from the central

line ;

^each _spectrum ^is

isotropic.

Figure

² shows the result of a microwave _power saturation

study

of the central line after irradiation

with 5.5 x

1017

electrons

cm-2,

^{with saturation}

beginning

^{at about 1 mW.}

After sufficient

annealing (around

¹⁸⁰

K)

^{an addi-}

tional two-line spectrum

(designated

^{as the A 8}

spectrum) split symmetrically

^{33 G} from the central line was observed

(Fig. 3).

^{We do not} have sufficient information on the A 7 and A 8 spectra ^{to determine} for certain if each consists of a

pair

of lines which are

due to

hyperfine

^{structure or if}

they

^{are due to two}

spin-one

centers; since there are a number of

spin-

one centers in diamond we

suspect

^{A 7 and A 8 are}

also,

in which case the

spin

Hamiltonian parameters for the spectra ^{are : A}

7 - g - 2.00,

^{D - 40}

G ; A 8 - g - 2.00,D - 22 G.

3. 2 INTRODUCTION RATE. - The variation of the

signal intensity

with the dose of irradiation for the central line and for the A 7 resonance is

given

ⁱⁿ

figure 4;

note, ^no central line was

present

^{in our}

sample prior

^to irradiation.

Figure

⁴ shows that the

intensity

of the A 7 resonance seems to increase

linearly

^{with the}

dose,

while the

intensity

^{of the}

central line seems to vary

sublinearly.

3. 3 ANNEALING. - After irradiation with the total dose of 7.5 x

1017

electrons

cm-2

the

sample

^was

isochronally

^{annealed from 77 K to room}

tempera-

ture ; measurements after each

annealing

step ^are

FIG. 4. ^- Derivative signal intensity ^versus dose of irradiation for the central line and the A 7 resonance.

FIG. 5. ^- Variation of the intensity of the central line, the A 7 and A 8 resonances versliv temperature during isochronal annealing.

performed

^{at 77 K.} The results for the central line and the A 7 and A 8 resonances are shown in

figure

^5.

The

intensity

^{of the A 8} resonance was too small to be followed

accurately

^{at each} temperature; ^{it is}

only

noticed that the A 8 resonance appears around 150 K and

disappears

^{around 220 K. The A 7 resonance}

disappears completely

^{near 220}

K;

^{the central line} anneals out between 250 and 300 K.

4. Discussion. ^- The irradiation has introduced at least two kinds of

paramagnetic

^{centers since the}

central line and the A 7 resonance do not have the

same

dependence

^on fluence. We must also note that the central line _may

correspond

^{to the}

superposition

of several spectra associated with different types ^of centers ; sublinear

growth

^{of the}

intensity

of the central line with the dose indicates a saturation in the concen-

tration of ^some of the centers.

These

paramagnetic

^{centers could be new centers}

introduced

by

the irradiation ^or centers

which

are

present before irradiation

(traps [7])

^{in a non-para-}

magnetic

^state and which become

paramagnetic

_ because of a

change

^{in their}

charge

state, ^{due to the} ionization which

accompanies

^the electron irradiation.

The reason the central line saturates with the dose could be the consequence of the fact that all the centers present before irradiation are ionized for small doses of irradiation.

Lomer and Wild

[3]

^{observed a}

growth

^{of a similar}

central line in

type

^{IIa diamond}

(irradiated

^{at 17}

K)

but it was also

present

^before

irradiation,

then increas- ed

by

^{a factor of} 1.5 upon irradiation with

1016

elec- trons

cm- 2

and did not increase anymore upon further irradiation with

1017

electrons

cm-2 ;

^the

production

rate is

quite

different from the

production

^{rate we}

observe in this

experiment. They

observed the inten-

sity

₈₀ of the central line to increase on

annealing to

K;

the resultant amount of

damage

^is

comparable

to our

damage

^rate

(’). They

also observed a reso- nance similar to the A 7 resonance, with a linear

production

rate; ^{but their resonance}

disappears

upon

annealing

^{at 140 K. This} argues that their resonance is different from the A 7 resonance which anneals around 230 K.

We do not have sufficient information to

identify

any of these centers. There have been a number of EPR spectra ^observed

[8-18]

ⁱⁿ irradiated diamond.

A number of tentative models for these centers have been

proposed,

^{but none is}

firmly

established. A number of these spectra ^are

spin-one

centers; it is

generally thought [9]

^{that their D values are determined}

by

^the

magnetic dipole-dipole interaction,

^the

magni-

tude of D is related

[17]

^{to the}

separation

^{of the two}

dipoles. Following

that line of argument,

if A

^{7 and A 8}

are

spin-one

centers, ^{their D values would}

imply

^a

mean

separation

between the

interacting dipoles

^of

- 7.5 and - 9

A respectively.

(1) ^{We are indebted to E.} W. J. Mitchell for pointing ^{this out.}

462

We can also _compare the

annealing experiments reported

^{in this} paper with

annealing experiments performed using conductivity

measurements after low and

high

energy electron irradiation

[1, 2].

Conductivity

measurements have demonstrated the existence of the thermal deexcitation

of

traps ^{at 160} and 230

K;

steps ^{in the EPR}

signal intensity

^are

observed at these temperatures

(they

^{are not} very well resolved on the

annealing

^{curve of}

figure

⁵

because the

temperature

rise time from 77 K was not

negligible

^with

respect

^{to the}

annealing

^{time at a}

given temperature).

Around 160 K the A 8 resonance

appears; around 230 K the A 7 and A 8 resonances

disappear.

^{Lomer and Wild}

(Thesis, Reading,

unpu-

blished)

^{observed two lines}

split

^{30 G} from the central

peak (their

^{R 14}

center)

which behaves

similarly

^to

A 8

(split

³³

G)

but anneals at a different tempera- ture ; ^the measurements are difficult

and,

^{while we} believe there are substantial

differences,

^{one should}

bear in mind that

subsequent

measurements may reveal that both resonances

belong

^{to the same defect.}

Since the variations of the

amplitude

^{of these reso-}

nances correlate with stages ^observed

using

^conduc-

tivity

measurements and attributed to the thermal deexcitation of carriers from traps

[2],

^{we can con-}

clude that

they

^{are also due to} carrier deexcitation.

Conductivity

measurements . have shown that the defects created

by

irradiation anneal in the _range 250-300 K

[1];

^a step ^is also observed in this tempe-

rature range in the EPR

signal intensity. Finally,

the

regular

decrease in the EPR

intensity

^versus

anneal

temperature

^can be attributed to the thermal deexcitation of a distribution of other traps ^which

are not detected

by conductivity.

5. Conclusion. ^- The results obtained

using

^EPR

measurements in diamond irradiated ^at low tempe-

rature are in agreement with the results obtained

using conductivity

measurements.

They

^reveal spectra associated with

traps already

present ^{in the}

sample (which

^become

paramagnetic

upon

ionization) and,

most

probably,

^a spectrum associated with the defects which anneal in the _range 250-300 K. Identification of these defects is not

yet possible.

^{An ENDOR}

study

with

special emphasis placed

^{on the}

frequency

^to

excite nuclear

alignment

^{of the}

13C and "B isotopes

could _prove fruitful to infer the

geometric configu-

ration of the defects.

Acknowledgements.

^- The authors are indebted to R. M. Chrenko and the General Electric Research Center

(Schenectady,

^N.

Y.)

^for

providing

^the

sample

used in this

study.

References

[1] MASSARANI, B. and BOURGOIN, ^J. C., Phys. ^{Rev. 14} (1976)

3682 + 3690.

[2] MASSARANI, B., BOURGOIN, J. C. and VISOCEKAS, R., to be

published.

[3] LOMER, J. N. and WILD, ^{A. M.} A., ^Phil. Mag. ²⁴ (1971) ^273.

[4] LOMER, ^J. N. and WELBOURN, C. M., Diamond Conference

(Cambridge, 1975), unpublished.

[5] This boron concentration is measured from the absorption

coefficient at 2 800 cm-1.

[6] ^The glue, ^once irradiated, gives ^{rise to an} isotropic ^EPR spectra ; ^{but its} intensity ^{is small in} comparison ^{with the} spectra studied and moreover it can be easily substracted from the combined spectra.

[7] Some of these traps ^are possibly those detected using ^ther-

moluminescence measurements and studied in ref. [2].

[8] GRIFFITH, ^{J. H.} E., OWEN, ^{J. and} WARD, I. M., ⁱⁿ Report of

the Conf. ^on Defects ⁱⁿ Crystalline ^Solids (Phys. Soc., London) 1955, p. 81.

[9] FAULKNER, ^{E. A. and} LOMER, J. N., Phil. Mag. ⁷ (1962) ^1995.

[10] FAULKNER, ^E. A., MITCHELL, ^{E. W. J. and} WHIPPEY, ^P. W., Nature 198 (1963) ^981.

[11] BALDWIN, ^{J. A.} Jr., Phys. ^{Rev. Lett. 10} (1963) ^220.

[12] CLARK, ^C. D., DUNCAN, I., LOMER, J. N. and WHIPPEY, P. W., Proc. British Ceramic Soc., No. 1 (1964) p. 85.

[13] OWEN, J., in Physical Properties of Diamonds, Ed. R. Berman

(Clarendon Press, Oxford) 1965, Chap. ^10.

[14] KIM, ^{Y. M. and} WATKINS, ^G. D., ^J. Appl. Phys. ⁴² (1971) ^722.

[15] LOMER, J. N. and WILD, A. M. A., Phil. Mag. ²⁴ (1971) ^273.

[16] WHIPPEY, P. W., Can. J. Phys. ⁵⁰ (1972) ^803.

[17] KIM, ^Y. M., LEE, ^Y. H., BROSIOUS, ^{P. and} CORBETT, ^J. W., ⁱⁿ Radiation Damage ^and Defects ⁱⁿ Semiconductors (The Institute of Physics, London) 1973, Conf. Series 16,

p. 202.

[18] LOMER, J. N. and WILD, ^{A. M.} A., Rad. Effects ¹⁷ (1973)

37.