Physica status solidi / A.: Volume 70, Number 1 March 16 [Reprint 2021 ed.] 9783112494387, 9783112494370

Citation preview

plrysica status solidi (a)

ISSN 0031-8965 * VOL. 70

NO. 1

MARCH 1982

Classification Scheme 1. Structure of Crystalline Solids 1.1 Perfectly Periodic Structure 1.2 Solid-State Phase Transformations 1.3 Alloys. Metallurgy 1.4 Microstructure (Magnetic Domains See 18; Ferroelectric Domains See 14.4.1) 1.5 Films 1.6 Surfaces 2. Non-Crystalline State 3. Crystal Growth 4. Bonding Properties 5. Mossbauer Spectroscopy 6. Lattice Dynamics. Phonons 7. Acoustic Properties 8. Thermal Properties 9. Diffusion 10. Defect Properties (Irradiation Defects See 11) 10.1 Metals 10.2 Non-Metals 11. Irradiation Effects (X-Ray Diffraction Investigations See 1 and 10) 12. Mechanical Properties (Plastic Deformations See 10) 12.1 Metals 12.2 Non-Metals 13. Electron States 13.1 Band Structure 13.2 Fermi Surfaces 13.3 Surface and Interface States 13.4 I m p u r i t y and Defect States 13.5 Elementary Excitations (Phonons See 6) 13.5.1 Excitons 13.5.2 Plasmons 13.5.3 Polarons 13.5.4 Magnons 14. Electrical Properties. Transport Phenomena 14.1 Metals. Semi-Metals 14.2 Superconductivity. Superconducting Materials and Devices 14.3 Semiconductors 14.3.1 Films 14.3.2 Surfaces and Interfaces 14.3.3 Devices. Junctions (Contact Problems See 14.3.4) 14.3.4 High-Field Phenomena, Space-Charge Effects, Inhomogeneities, Injected Carriers (Electroluminescence See 20.3; Junctions See 14.3.3) 14.4 Dielectrics 14.4.1 Ferroelectrics 15. Thermoelectric and Thermomagnetic Properties 16. Photoconductivity. Photovoltaic Effects 17. Emission of Electrons and Ions 17.1 Field Emission Microscope Investigations 18. Magnetic Properties 18.1 Paramagnetic Properties 18.2 Ferromagnetic Properties 18.2.1 Ferromagnetic Films 18.3 Ferrimagnetic Properties 18.4 Antiferromagnetic Properties (Continued on cover three)

physica status solidi (a) applied research Board of Editors S. A M E L I N C K X , Mol-Donk, J . AUTH, Berlin, H. B E T H G E , Halle, K. W. B Ö E R , Newark, P. GÖRLICH, Jena, G. M. HATOYAMA, Tokyo, C. HILSUM, Malvern, B. T. K O L O M I E T S , Leningrad, W. J . MERZ, Zürich, A. S E E G E R , Stuttgart, C. M. VAN V L I E T , Montréal Editor-in-Chief P. GÖRLICH Advisory Board L. N. A L E K S A N D R O V , Novosibirsk, W. ANDRÄ, Jena, E. B A U E R , Clausthal-Zellerfeld, G. C H I A R O T T I , Rom, H. C U R I E N , Paris, R. G R I G O R O V I C I , Bucharest, F. B. H U M P H R E Y , Pasadena, E. K L I E R , Praha, Z. M A L E K , Praha, G. O. M Ü L L E R , Berlin, Y. NAK AMURA, Kyoto, T. N. R H O D I N , Ithaca, New York, R. SIZMANN, München, J . S T U K E , Marburg, J . T . W A L L M A R K , Göteborg, E. P. W O H L F A R T H , London

Volume 70 • Number 1 • Pages 1 to 342, K 1 to K92, and Al to A8 March 16, 1982 PSSA 70(1) 1 - 3 4 2 , K 1 - K 9 2 , A 1 - A 8 (1982) ISSN 0031-8965

A K A D E MI E - V E R L A G • B E R L I N

Subscriptions and orders for single copies should be directed in the G D R : to the Postzeitungsvertrieb or to the Akademie-Verlag, DDR-1086 Berlin, Leipziger Straße 3 - 4 ; in the other socialist countries: to a book-shop for foreign language literature or to t h e competent news-distributing agency; in the F R G and B E R L I N (WEST): to a book-shop or t o t h e wholesale distributing agency K u n s t und Wissen, Erich Bieber OHG, D-7000 Stuttgart 1, Wilhelmstr. 4 - 6 ; in the other Western European countries: to K u n s t und Wissen, Erich Bieber GmbH, CH-8008 Zürich, Dufourstr. 51; in USA and CANADA: to Verlag Chemie International, Inc., Plaza Centre, Suite E , 1020 N. W . 6th Street, Deerfield Beach, F L 33441, USA; in other countries: to t h e international book- and journal-selling trade, to Buchexport, Volkseigener Außenhandelsbetrieb der Deutschen Demokratischen Republik, DDR-7010 Leipzig, Postfach 160, or to the Akademie-Verlag, DDR-1086 Berlin, Leipziger Straße 3—4.

Editorial Note: "physica status solidi (a)" undertakes t h a t an original paper accepted for publication before the 23rd of any month will be published within 50 days of this date unless the author requests a postponement. I n special cases there m a y be some delay between receipt and acceptance of a paper due to the review and, if necessary, revision of the paper.

Schriftleiter u n d v e r a n t w o r t l i c h f ü r den I n h a l t : Professor D r . D r . h.c. P . GSrlich, DDR-1020 Berlin, Neue Schönhauser S t r . 20 bzw. DDR-6900 J e n a , Schilibachstr. 24. Verlag: Akademie-Verlag, D D R - 1 0 8 6 B e r l i n , Leipziger S t r . 3 - 4 : Fernruf 2 2 3 6 2 2 1 und 2 2 3 6 2 2 9 ; T e l e x - N r . : 114420; B a n k : S t a a t s b a n k der D D R , B e r l i n , K t o . - N r . : 6836-26-20712. Chefredakteur: D r . H . - J . Hänsch. Redaktionskollegium: P r o f . D r . E . Gutsche, D r . H . - J . H ä n s c h , D r . H . Lange, D r . S. Oberländer. Anschrift d e r R e d a k t i o n : DDR-1020 Berlin, Neue Schönhauser Str. 20; F e r n r u f : 282 3380. Veröffentlicht u n t e r der L i z e n z n u m m e r 1620 des Presseamtes beim Vorsitzenden des Ministerrates der Deutschen Demokratischen R e p u b l i k . Gesamtherstellung: V E B Druckerei „ T h o m a s M ü n t z e r " , DDR-5820 B a d Langensalza. Erscheinungsweise: Die Zeitschrift „ p h y s i c a s t a t u s solidi ( a ) " erscheint jeweils a m 16. eines) eden Monats. Jährlich erscheinen 6 B ä n d e zu je 2 H e f t e n . Bezugspreis: J e B a n d 165,— M zuzüglich VersandBpesen (Preis f ü r die D D R : 130,— M). Bestellnummer dieses B a n d e s : 1085/70. © 1982 h y Akademie-Verlag Berlin. Printed i n t h e German Democratic Republic. AN ( E D V ) 20 735

Contents Review Article A Review Article for Volume 70 will appear in No. 2 of this Volume.

Original Papers P . P U R E U R , J . V . KUNZLER, W . H . SCHREINER, a n d D . E . BRANDAO

Electrical Resistivity of the Cobalt-Rich Co-Fe Alloy

11

V . E . ARKHIPOV, V . I . VORONIN, A . E . R A S K I N , a n d A . V . MTRMEISHTEIN

Radiation Disordering in V3Si A . B . GERASIMOV, N . D . D O L I D Z E , R . M . D O N I N A , B . M . K O N O V A L E N K O , G . L . O F E N G E I M ,

17 and

A . A . TSERTSVADZE

On the Identification and Possible Space Orientation of "Light-Sensitive" Defects in Ge

23

A . V . MTTROMTSEV a n d E . A . R A I T M A N

Relaxation of Complex Susceptibility of Ferrites in AC Magnetic Field

29

P . KTJKK, Ö . P A I M R E , a n d E . MELLIKOV

The Structure of Recombination Centres in Activated ZnSe Phosphors. .

35

L . N . SKITJA a n d A . R . S I L I N

A Model for the Non-Bridging Oxygen Center in Fused Silica. The Dynamic Jahn-Teller Effect

43

H . BINCZYCKA, 0 . GZQWSKI, L . MURAWSKI, a n d J . SAWICKI

Mössbauer Effect and Electrical Conductivity in Te0 2 -Fe 2 0 3 Glasses. . .

51

F . M . M A N S Y , N . K . G O BRAN, a n d G . S A I D

High-Temperature Creep Mechanism and Dissolution Process in Al-9.5 a t % Zn Alloy

57

G . S . M U R T H Y a n d D . H . SASTRY

Impression Creep of Zinc and the Rate-Controlling Dislocation Mechanism of Plastic Flow at High Temperatures

63

Z . CZAPLA, L . S O B C Z Y K , a n d J . M R 6 Z

Isotopic Effect in Ferroelectric Rubidium Hydrogen Selenate Crystals

73

S. PARVIAINEN a n d M . LEHTINEN

Magnetic Contribution to the Thermal Expansion in Some SiFe^Co.^ Alloys l*

79

4

Contents

T . T R O E V , A . K R U S T A N O V , V . P . SHANTAROVICH, L . G . A R A V I N , a n d B . P . M O L I N

Doppler Broadening and Positron Lifetime Measurements in Al-49 a t % Zn at Different Temperatures

87

J . KOCKA, J . STUCHXIK, O . STIKA, M . LÄZNIÖKA, O . R E N N E R , a n d E . K R O U S K Y

Physical Properties of a-Si:H Prepared at about 1 Torr W.

GRAPE

93

Description of Internal Friction Background of Glasses by Means of Migration Processes

101

L . M . C A S P E R S , A . VAN V E E N , M . R . Y P M A , a n d G . J . VAN D E R K O L K

Helium Precipitation in a-Fe

109

R . H A M A S e t J . P . RIQTTET

Estimation de l'energie d'un joint de phases metalliques

Y u . S . BTTLYSHEV, I . M . K A S H I R S K I I , V . V . S I N I T S K H ,

127

G . F . MYATCHTNA, T . G . ERMAKOVA,

a n d V . A . LOPYREV

Trapping Centres in Poly-l-Vinyl-l,2,4-Triazole

139

G . VAN T E N D E L O O , J . VAN L A N D T J Y T , a n d S . A M E L I N C K X

High Resolution Electron Microscopic and Electron Diffraction Study of Gold-Manganese Alloys with a Composition in the Vicinity of Au a Mn. . .

145

J . X . ZHANG, T . S . K E , G . Y . L I , G . FANTOZZI, P . F . GOBIN, a n d M . W E L L E R

Low-Temperature Cold Work Internal Friction Peaks in Al-0.5 w t % (0.21 at%) Cu

159

W . BRÜCKNER, W . MOLDENHAUER, a n d K . - H . BÄTHER

Time Dependence of Degradation in ZnO Varistors

167

R . V E T T E R , R . H . J . FASTENATT, a n d A . VAN D E N B E U K E L

Dislocation Mean Free Path Calculated by Computer Simulations . M . RENNINGER

. .

177

Imaging of Lattice Defects by Double Diffractometric Transmission Topography

183

F . C A L L E N S , W . MAENHOTJT-VAN D E R V O R S T , a n d L . W . K E T E L L A P P E R

The Effect of the Solution pAg and p H on the Space Charge Characteristics of Silver Bromide Emulsion Grains

189

L . PASEMANN, H . BLUMTRITT, a n d R . GLEICHMANN

Interpretation of the EBIC Contrast of Dislocations in Silicon P. J. R A T C U E E E X-Ray Topography and Laser Tomography of Crystalline TGSe

197 211

Contents W. KLEINN

5 Restwiderstandsbestimmung aus Strom-Spannungs-Kurven kurzer freitragender Elektrotransportproben epitaktischer Goldschichten

219

J . METZDOM, R . NIES, a n d F . R . KESSLEB

Interband Faraday and Kerr Effects in Germanium. Interrelations and Dispersion Relations

233

W . WESCH, E . WILK, a n d K . H E H L

H . MÜLLER

Radiation Damage and Near Edge Optical Properties of Nitrogen Implanted Gallium Arsenide

243

Simulation of Oriented Crystallization in Evaporated Antimony Thin Films

249

H . MIZITBAYASHI a n d S . OKUDA

Low Temperature Internal Friction in Deformed Ta

257

A . D . R E D D Y , S . M . D . R A O , a n d G . S . SASTBY

Electrical Conductivity and Simulated Thermo-Current Studies on V 0 2 + Doped Li(N 2 H 6 )S0 4 Single Crystals

269

K . ISHIDA, Y . MATSUMOTO, a n d K . TAGTJCHI

Lattice Defects in L P E InP-InGaAsP-InGaAs Structure Epitaxial Layers on I n P Substrates

277

J . K I L M E B , E . R . C H E N E T T E , C . M . VAN V L I E T , a n d P . H . H A N D E L

Absence of Temperature Fluctuations in 1//Noise Correlation Experiments in Silicon

287

W . HERBEMASTS, M . DAVID, a n d R . GEVEBS

THAN CHOT

Crystal Potential near a Stacking Fault

295

Some New Experimental Results of a Study of Metal-n-Type Semiconductor Schottky Barrier Contacts

311

V . K A L Y ATTAR AM AN a n d V . KUMAR

Properties of the Gold Related Acceptor Level in Silicon

317

X . W . F A N a n d J . WOODS

Green Electroluminescence and Photoluminescence in CdS

325

K . N . R E D D Y , M . L . R A O , a n d V . H . BABTX

Thermoluminescence and Optical Absorption Studies of Z r Centres in NaCl Crystals Doped with Samarium

335

6

Contents

Short Notes

J . TEUHO a n d K . MAKINEN

On the Out-of-Step Period in Au3Zn(H)

K1

K . TENNAKONE a n d W . G . D . DHABMABATNE

Electrical Conduction in Lead Bromide

K5

I . KENNING, P . FRACH, E . HEGENBARTH, a n d V . I . FRITSBERG

Glass-Like Behaviour of PLZT Demonstrated by Heat Capacity Measurements

K7

R . BHAEATT a n d R . A . SINGH

Electrical Transport in Chromium(III) Tungstate

Kll

K . B . KADYRAKUNOV, E . V . NIDAEV, a n d L . S. SMIRNOV

Behaviour of a Built-in Charge in a Dielectric at Pulsed Annealing of Si-Si0 2 Structures K15

D . KOSTOPOXJLOS, S. KOTRREMENOTJ, P . VAEOTSOS, a n d S. MOURIKIS

Thermocurrents in Ferroelectric Materials

K19

H . KIDO, M . SHIMADA, a n d M . KOIZUMI

Synthesis and Magnetic Properties of GdCoSi and GdMnSi

K23

X . W . FAN a n d J . WOODS

Blue and Violet Electroluminescence Emission in ZnSe^Sj^ MIS Diodes K27

H. TAGTJCHI

Effect of CO Adsorption on ( L a ^ S r ^ F e O . , (0 ^ x ^ 0.5)

K31

L . BISCHOEF, T . GESSNER, a n d H . MORGENSTERN

Iron Donor Activity at Heat Treatment of High Resistivity S i l i c o n . . . .

K35

M. STOJI(273 K)/g(4.2 K ) ratios for Co

Table 2 Residual resistivity and resistivity ratio for Co q0 ((ifl cm) White and Woods [11] Price and Williams [13] Laubitz and Matsumura [12] this work [1st T.T.] this work [2nd T.T.]

0.0808») 0.1275 0.1053 0.0624

P 64») 49**) 120 50 85

Obs.: *) mean value of 3 samples. * * ) in this case p = q (300 K)/q (4.2 K).

for the two thermal treatments, as compared to other relevant measurements [11 to 13]. The cooling rate evidently has strong influence on the resistivity ratio. I t is known (Laubitz and Matsumura [12]) that the residual resistivity of pure Co is mainly due to little quantities of the f.c.c. phase which are retained at low temperature through the martensitic f.c.c. - » h.c.p. transition. Then, a strong cooling rate, favouring the formation of f.c.c. precipitates, increases q0. The ideal resistivity data, Qi{T), are in good agreement with White and Woods' [11] results. 3.2 Residual resistivity Fig. 1 shows the measured residual resistivities for the different alloys. I t looks clear that the anomalous variation of the resistivity with increasing Fe concentration is linked with the structure transformations: at 1 a t % F e the h.c.p.-h.c.p. transition occurs [7, 8]; between 4 and 5 a t % F e a two phase region initiates and between 6 and 7 a t % Fe the alloys become f.c.c. Also it can be seen, somewhat surprisingly, that the second heat treatment has systematically increased the residual resistivity values. This effect is remarkable at concentrations where a structural phase transition occurs. This behaviour has been confirmed by a third measurement on the 1 and 2 a t % samples, which were re-annealed at 800 °C for 24 h, followed by rapid cooling. I t is

2 3 1 5 6 78 9 xiotV —

Pig. 1. Residual resistivity at 4.2 K for Co-Fe alloys. • 24 h, O 48 h, A 72 h annealing

P. Pubextr, J. V. Ktjnzler, W. H. S c h r e i n e r , and D. E. Brandao

14

unlikely t h a t the observed trends are associated with texture always present in hexagonal polycrystals. Rather it suggests t h a t the original alloys configurational states were out of equilibrium, somewhat heterogeneous, and showing some solute ordering in distances as short as the electronic mean free path. I n a given alloy, disregarding its lattice structure, it seems that the random configurational state is reached only through long annealing times. Zubchenko et al. [14] ascribed similar effects to the complex resistivity variation in Cu-Al alloys, during heat treatments. Nevertheless, other experiments would be welcome to clarify this unexpected behaviour. 3.3 Temperature-dependent

resistivity

The temperature-dependent resistivity for the Co-Fe alloys is expressed through the quantity A(T,

x) =

QMOY(T,

x) -

Q0(X) -

QCo(T)

,

where £>aiioy is the total resistivity and Q0 is the residual resistivity of the alloy, while Qco is the ideal resistivity of pure Co. A(T, x) simply expresses the temperaturedependent alloy resistivities as compared to the matrix. For diluted alloys, this quantity is known as the deviation of Matthiessen's rule. Fig. 2 shows A as a function of the concentration for four different temperatures. As a function of T, A shows three qualitatively different behaviors, depicted in Fig. 3. For x 6 a t % , A{T) shows the characteristics of the deviations of Matthiessen's rule in ferromagnetic dilute alloys [1]. I n the cubic samples A(T) shows a relative maximum at intermediate temperatures. This must be connected with changes in the phonon spectrum and in the electronic structure which come through the lattice hexagonal-cubic transformation. For x = 10 a t % Fe, A{T) has negative values between 120 and 240 K.

Fig. 2

Pig. 3

Fig. 2. Deviations of Matthiessen's rule for the Co-Fe alloys as a function of the Fe concentration Fig. 3. Deviations of Matthiessen's rule for the Co-Fe 1, 8, and 10 at% alloys, as a function of temperature, after the first thermal treatment

15

Electrical Resistivity of the Cobalt-Rich Co-Fe Alloys

4. Final Remarks I n the rigid band model the alloying of iron to cobalt, in addition to the introduction of scattering centers, should modify the electronic band structure in augmenting the vacant states. This would enhance the effectivity of the s - d scattering, which according to Mott [15] is the relevant resistive mechanism in incomplete d-band metals. Following this idea one should expect the cobalt-rich Co-Fe resistivity to increase continuously with iron concentration. The experimental data do not corroborate this tendency, but reveal the strong influences of structural phase transformations and, even though, a significant resistivity drop a t 7 at % Fe. Thus the electrical resistivity of the cobalt-rich Co-Fe alloys is a complicated function of iron concentration and temperature. References [1] [2] [3] [4]

J . BASS, A d v . P h y s . 21, 4 3 1 (1972). J . DTTEAND a n d F . GAUTIER, J . P h y s . C h e m . S o l i d s 3 1 , 2 7 7 3 (1970). B . LOEGEL a n d F . GAUTIER, J . P h y s . C h e m . S o l i d s 3 2 , 2 7 2 3 (1971). I . A . CAMPBELL, A . FERT, a n d A . R . POMEROY, P h i l . M a g . 1 5 , 9 7 7 (1967)

[5] M. HANSEN, Contitution of Binary Alloys, McGraw-Hill 1958. [6] D. A. WOODFORD and H. J. BEATTIE, Metallurg. Trans. 2, 3223 (1971). [7] T. WAKIYAMA, Proc. 1972 Conf. Magnetism and Magnetic Materials, Ed. C. D. GRAHAM, JR., and J. J. RHYNE, AIP Conf. Proc. 10,921 (1973). [8] T. ONOZUKA, S. YAMAGUCHI, M. HDRABAYASHI, a n d T. WAKIYAMA , J . P h y s . Soc. J a p a n 3 3 , 8 5 7 ( 1 9 7 2 ) ; 37, 687 (1974).

[9] C. R. WHITEMORE, Rare Metals Handbook, Ed. C. A. HAMPEL, Reinhold, New York 1961 (p. 114). [10] International Tables for X-Ray Crystallography, The Kynoch Press, 1969. [11] G. K. WHITE and S. B. WOODS, Phil. Trans. Roy. Soc. (London) 251A, 273 (1959). [12] M. J . LATJBITZ a n d T. MATSUMTTRA, Canad. J . P h y s . 51, 1247 (1973).

[13] D. C. PRICE and G. WILLIAMS, J. Phys. F 3, 810 (1973). [ 1 4 ] V . S. ZUBCHENKO, N . P . KULISH, a n d P . V . PETRENKO, F i z . M e t a l l o v . i M e t a l o v e d e n i e 4 7

31 (1980). [15] N . F . MOTT, A d v . P h y s . 1 3 , 3 2 5 (1964). (Received

August

20,

1981)

17

V. E . ABKHIPOV et al. : Radiation Disordering in V3Si phys. stat. sol. (a) 70, 17 (1982) Subject classification: 1.3 and 11; 14.2; 21 Institute of Metal Physics, Sverdlovsk1)

Academy

of Sciences

of the USSR,

Ural Scientific

Centre,

Radiation Disordering in Y3Si By V . E . ABKHIPOV, V . I . VORONIN, A. E . KABKXN, a n d A . V . MIRMELSHTEIN Disordering in V3Si, irradiated by fast neutrons of fluence (0.1 to 4) x 10 20 neutrons/cm 2 , is investigated by neutron diffraction technique. I t is shown, that in the investigated samples irradiated by maximum fluence the long-range-order degree remains sufficiently high, while the superconducting transition temperature fall« down to ss 1.5 K. IlpOBeHeHbIHeflTpOHHO-HH(|)$paKI(HOHHHe HCCJiefl0BaHHHpa3yn0pHH0HeHHHBC0ej(HHeHHH V3Si, nocne oOjiyieniiH GHCTPHMH HeftipoHaMH (jwnoeHCOM (0,1 ho 4) x 10 2 0 HeyTpoirti/cm 2 . Il0Ka3aH0, HTO B HccjienoBaHHbix o6pa3iiax nocjie oSjiyieHiifi MaiiciiMajibHUM jnoeHCOM CTeneHb HajibHero nopnana ocTaeTcn SHaiHTejihuoil, Tor^a Kan TevmepaTypa nepexojia B CBepxnpoBOflHmee COCTOHHHC najiaeT ho ~ 1 , 5 K .

1. Introduction The study of radiation disordering in superconductors with A-15 structure is of interest from the point of view of possible correlations between the change in critical parameters, which is often fairly appreciable under irradiation, and the degree of disordering. At present there are [1] similar studies on the Nb-base systems. As for V-base compounds, they practically have not been studied. The V 3 Si compound has been irradiated by fast neutrons of fluence 2.2 x 1019 neutrons/cm 2 [2] and it has been found that at this fluence the disordering is rather weaker than, for example, in Nb 3 Sn. In the present paper the results of neutron diffraction studies concerning disordering in the V 3 Si samples, irradiated by fast neutrons of fluence (0.1 to 4) X 10 20 neutrons/cm 2 (E 1 MeV) are presented. 2. Sample Preparation and the Research Technique The samples of V 3 Si were prepared by arc-melting in a furnace with a permanent electric vanadium electrode V E L - I and semiconducting silicon. According to the data of the chemical analysis the deviation in the composition of alloys from the stoichiometry was not more than 0.5 at%. The lattice parameter of the initial samples, according to the X-ray data, equals 4.729 A. Neutron diffraction studies have been carried out on powders (the size of a grain being 60 X 1 0 - s cm using a neutron diffractometer with the wavelength of monochromatized neutrons A = (1.29 + ± 0.05) A. Irradiation of the samples has been performed in the water cavity of the atomic reactor in a sealed aluminium capsule. The irradiation temperature, according to the test measurements, was not higher than 70 °C. !) ul. Kovalevskoi 18, GSP-170, 620219 Sverdlovsk, USSR. 2

physica (a) 70/1

18

V. E. Abkhipov, V. I. VOBONIN, A. E. Kabkin, and A. V. Mibmelshtein 3. Determination of the Long-Range-Order Parameter

I n V3Si with A-15 structure (space group P m 3n) atoms of Si occupy 2(a) sites (000), while atoms of Y occupy 6(c) sites (0 \ . For this structure the long-range-order parameter is introduced as in [3], S =

P?

Ck

where P a ' is the possibility of the A atom to occupy its usual site, C A the concentration of atoms of the A type, and v the relative concentration of sites of t y p e I. Expressions for the structural factors are as follows: •^110,220,310 = 2£(&Si — by) , ^200,211,321 = 4[6 + -i- S(bsi - 6 V )], F.210,320 = 4[b~\

S(bSi - by)] , ^222 = 4[-f- S(bSi — by) — 6] , Fiw> = 86 . Here and by are amplitudes of neutron scattering by the Si and V nuclei equal 0.42 x IO" 12 and 0.05 X 1 0 " 1 2 c m , respectively; b = Cgiòsi + Cyby. F r o m the structural factors it is seen t h a t the intensity of reflexes (110), (220), (310) is completely determined by the long-range-order parameter and these reflcetions are the most sensitive ones to disordering. Thus the degree of long-range-order has been calculated in terms of ratios of the intensity of these reflexes to those which depend upon the mean scattering amplitude, a p a r t from the long-range-order parameter. The values of the long-range-order parameter, obtained from different relations, have been averaged. During the calculations a correction for the Debye-Waller isotropic factor has been introduced. The Debye temperatures have been taken equal to 505 [2] and 540 K [2] for unirradiated and irradiated samples, respectively.

Fig. 1. Dependence of the intensity of the (110) reflex uponfluence. The upper curve (•) represents the initial sample, + / = 1 X 1019, o 3 X 1019, A 5 X 1018, V 10 X 1019, • 4 x 1020 neutrons/cm

19

Radiation Disordering in V3Si

Pig. 2. Dependence of Tc and the degree of the longrange order of the V3Si sample irradiated by afluence of 4 X 1020 neutrons/cm2 upon the temperature of isochronal annealing

Table 1 0 (1020 neutrons/cm2) 0

0.1

0.3

0.5

1.0

4.0

4.0 + annealing

0.89

0.83

0.79

0.74

0.62

0.68

0.296

0.176

0.144

0.102

0.074

—

0.99

K (10"2° cm2/neutron) —

Table 1 presents the mean values of the long-range-order parameter obtained for the samples under investigation. The maximum deviation from the mean value of 8 was not more than 0.05. I t is easily seen that in the compound under investigation the long-range-order parameter remains rather high even at the maximum fluence, as distinct from the Nb-base compounds with A-15 structure. I n isostructural Nb 3 Sn, irradiated under similar conditions* the long-range-order parameter was at the same fluence less than 0.1 [4], Qualitatively this slight variation of S in V3Si is illustrated by the change in the intensity of reflex (110) depending upon the fluence (Fig. 1). I t is also to be noticed that the shift of reflexes, indicating an appreciable change in the lattice parameter, has not been observed on neutron diffraction patterns of the irradiated samples (up to (400)). Fig. 2 shows curves of recovery of the long-range-order degree and T e for the sample irradiated by a fluence of 4 X 1020 neutrons/cm 2 depending upon the temperature of isochronous annealing (T a n n = 30 min). I t is seen that the curves have one strictly marked stage of annealing, which coincides with the annealing stage on [5]. 4. Small-Angle Neutron Scattering In all of the investigated samples the behaviour of the diffuse background at small angles has also been studied. Fig. 3 shows the dependence obtained for the small-angle scattering. The small-angle scattering is seen to be rather high even in the sample irradiated by a fluence of 1 X 1019 neutrons/cm 2 . I t increases with further increasing fluence, having its maximum at a fluence of 1 X l 2 0 neutrons/cm 2 . I t is to be mentioned that the angular dependence of the small-angle scattering is illustrated by a curve with a maximum. With increasing fluence there is a slight shift of the maximum towards zero angle. The annealing of the sample irradiated by a fluence of 4 X 1020 neutrons/cm 2 at 500 °C leads to an appreciable decrease in the intensity of the small-angle scattering. After the annealing at 600 °C the small-angle scattering in the sample disappears. The sizes of segregations, the presence of which can explain the small-angle scattering have been evaluated according to the decrease in the small-angle scattering towards larger angles by the normal technique [6], The mean radius of the segregations changes from 7 to 10 A depending upon the fluence. 2'

20

V. E. Abkhlpov, V. I. Vobontk, A. E. Kabkin, and A. V. Mirmelshtein

f

7

0 -1 4°

Fig. 3. Angular dependence of the small-angle scattering on the irradiation samples. (1) 0 = 0, (2) 1 x 10", (3) 5 x 101*, (4) 1 X 1020, (5) 4 X 1020 neutrons/cm2

I 6°

8'

W

28

5. Discussion At present there is a series of radiation disordering models for the ordered materials, which allow to give a satisfactory explanation for the available experimental data [7], They are: (i) Disordering in the regions of the displacement peaks; (ii) disordering owing to the chains of successive replacements ; (iii) disordering by virtue of non-correlated replacements. For all these models the dependence of the long-range-order degree upon the fluence is as follows: S =

S0

exp

( - K & ) ,

(1)

where if is a constant, which varies in different models, 0 is the fluence. I n this case, Kjcf (where a is the cross-section of the fast-neutron scattering) is the number of displacements per primarily knocked out atom (PKA). In the case of disordering which does not go up to the end, reaching at large fluences the saturation region, expression (1) is as follows: S -

S*

=

(S0

-

S*)

exp

( - K 0 ) ,

(2)

where S* is the value of the long-range-order degree in the region of saturation. Table 1 demonstrates the values of K, calculated from the data on the long-range-order degree, using (2). Evidently it is impossible to describe the long-range-order degree more or less satisfactorily using the single value of K or the number of displacements per PKA. Since there is no reason to suppose that the number of displacements per P K A changes with fluence, the results obtained can be explained by partial recovery of the long-range-order degree at the given irradiation temperature, due to the diffusion induced by radiation. In this case the degree of recovery is higher, the longer the irradiation time is. At a first sight the presence of the small-angle neutron scattering in the samples under investigation speaks in favour of an inhomogeneous mechanism of radiation disordering. However, in the case of inhomogeneous disordering in the samples there appear statistically distributed disordered microregions, the distribution character of which needs an exponential increase of the small-angle scattering in approaching the

21

Radiation Disordering in V3Si

zero angle. The experimental curves of the small-angle scattering (Fig. 3) have a maximum which evidences partial ordering in the distribution of microregions t h e presence of which, in its turn, can also lead to the small-angle scattering. Thus there is no reason to suggest t h a t the disordered microregions appearing under irradiation, have an ordered arrangement over the sample. I t is likely t h a t the existence of the small-angle scattering speaks in favour of ordered clusters of atoms. Since the amplitude of the neutron nuclear scattering in vanadium is very small and actually its clusters are not seen, we can suppose the small-angle scattering be caused by cluster of silicon interstitials. The problem concerning the configuration of the clusters cannot be solved on the basis of the experimental data available. The disappearance of the small-angle scattering during the annealing at 600 °C supports the suggestion about the existence of the clusters of interstitial atoms. Thus the experimental data obtained cannot lead to a universal conclusion about the mechanism of the radiation disordering in the system under investigation. I n order to obtain more specific information and to elucidate the nature of the smallangle scattering, experiments are to be carried out with the samples obtained at low temperatures when diffusion processes are suppressed. The results of these experiments will be reported elsewhere. Acknowledgement

The authors would like to thank B. N. Goshchitskii for his consideration of the study. References [1] A. R. SWEEDLEE, D. E. Cox, and L. NEWKIRK, J. Electronic Mater. 4, 885 (1975). [2] D. E. Cox and J. A. TABVIN, Phys. Rev. B 18, 15 (1978). [3] M. A. RRIVOGLAZ, Teoriya uproyadochivayushchikhsya splavov, Izd. Nauka, Moscow 1961. [ 4 ] A . E . KABKIN, B . N . GOSHCHITSKII, V . E . ABKHIPOV, E . Z . VALIEV, a n d S . K . SIDOBOV, p h y s .

stat. sol. (a) 46, K87 (1978). [ 5 ] A . E . KABKIN, V . E . ABKHIPOV, V . A . MABCHENKO, a n d B . N . GOSHCHITSKII, p h y s . s t a t . sol. (a) 5 4 , K 5 3 ( 1 9 7 9 ) .

[6] A. GUINIEB, Theorie et technique de la radiocristallographie, Dunod, Paris 1956. [7] V. C. POLENOK, Fiz. Metallov i Metallovedenie 36, 195 (1973).

(Received August 7, 1981)

A. B.

GERASIMOV

et al.: Identification and Space Orientation of Defects in Ge

23

phys. stat. sol. (a) 70, 23 (1982) Subject classification: 11; 20.1; 22.1.1 Institute of Physics, Academy of Sciences of the Georgian SSS,

Tblissi1)

On the Identification and Possible Space Orientation of „Light-Sensitive" Defects in Ge By A . B . GEBASIMOV, N . D . DOLIDZE, R . M . D O N I N A , B . M . G. L . OFENGEIM, a n d A . A .

KONOVALENKO,

TSEKTSVADZE

A radiation defect responsible for the 0.52 eV absorption band in the I R absorption spectrum of irradiated Ge is investigated, n-type samples cut in three main crystallographic directions , , and are irradiated by 3 to 5 MeV electrons at T ^ 77 K. During the measurements monochromatic polarized light and uniaxial stress are used. I t is shown that the radiation defect studied has the property of dichroism. Besides, depending on the direction of applied uniaxial stress, variations in the value of the 0.52 eV absorption band are revealed. The space orientation of the radiation defect giving rise to the 0.52 eV absorption band is determined. Based on the analysis of the data obtained as well as on the analogy with silicon, it is assumed that the defect is a divacancy whose axis is oriented close to the , < 1 1 0 > H < 1 0 0 > o6jiyHajiHCB 3jieKTpoHaMH c SHeprHeii 3 no 5 M e V npw T 7 7 K . B npouecce H 3 M e p e H H i i HcriojibaoBajicH M o n o x p o M a T M i e c K i i i i n0JiHpn30BaHHBiii CBeT H 0HH00CH0e « K a r a e . IIoKa3aHO, HTO HaynaeMHM pa^nauHoiiHLrii ne$eKT oSjianaeT CBOHCTBOM nHxpoH3Ma. KpoMe Toro, B 3 a B u c n M 0 C T H OT HanpaBjieHHH npnjiosKeHHH 0 H H 0 0 C H 0 r 0 cmaTHH o0HapvjKeHo H 3 M e H e H H e BCJIHTOHBI n o r j i o m e H H H nojiocti 0 , 5 2 e V . OnpeaejieHa npocTpaHCTBeHHan opneHTauHH p a n n a i i H O H H o r o neeKTa oTBeTCTBeHHoro 3a n o n o c y norjiomeHHH 0,52 eV. B pe3yjiBTaTe aHaJiH3a nojiyieHHHX aaHHbix, Hcnojib3yn aHajiornio c KpeMHHeM, nejiaeTCH npennojioHteHHe, h t o HCCJIeayeMtiM He$eKT HBJIHCTCH HimaKaHCHeii, ocb KOTOpoil opHeiiTHpoBaHa 6JIH3KO K KpHCTajuiorpacfiMiecKOMy HanpaBjieHHio < 1 1 0 ) .

1. Introduction I n the previous works [1 to 3] it has been shown t h a t after irradiation of Ge at t e m peratures approaching the liquid nitrogen temperature, an absorption b a n d is observed with the m a x i m u m in the region of 0.52 e V in t h e spectrum. I n [3] it is supposed t h a t the defects responsible for the given absorption band are "light-sensitive" defects [10] or as t h e authors of [4] term them, t w o - s t a t e defects. T h e performed investigat i o n s resulted in m a n y data on the defect behaviour responsible for the g i v e n absorpt i o n band. However, in spite of this, so far there is n o unified opinion as t o t h e inv o l v e d defect nature. T h e present work is d e v o t e d to further studies of "light-sensitive" radiation d e f e c t s in Ge, giving rise t o the absorption band in the 0.52 eV spectrum region. ul. Guramishvili 6, 380077 Tblissi, USSR.

24

A . B . GEBASIMOV e t al.

2. Experimental The study was carried out by infrared ( I R ) absorption spectrum measurements. During the measurements polarized light and uniaxial stress were used. Sb-doped n-Ge samples (n = 1 X 1014 to 1.1 X 1016 cm - 3 ) were cut in the form of parallelepipeds (2.5 X 2.5 X 12 mm3) strictly oriented in the three main directions of the Na). The identified centres and corresponding defects are given in Table 3. Table 3 Identification of recombination centres with crystal lattice defects in ZnSe: Cu: CI centre

structure of corresponding crystal lattice defect

N cu < 1 2 3 4 5 6 7

r-centre k-centre s-centre predominant Cu defect predominant CI defect predominant " + " defect predominant defect

Na

Ncu > Nei

(Cu Z n Cl S e) x

(CuznClse)

(Cuzn)i' or (CujCuZn) x Clse or VZ„(CIée)2 (CuiCuzn) x or (CuZnC]Se)2x

(Cuzn)â Cui (CujCuzn) x

Clse

(CuZnClse)

cià.

h-

e'

Cuzn

x

x

In the case of slow cooling (30 days) from 1000 °C to room temperature, the precipitation of Cu from ZnSe: Cu (jVCu Ncl) is observed [27], The initial copper concentration in ZnSe, Ncu = 10 19 cm - 3 , decreases to 10 18 cm - 3 . Especially intense precipitation appears in the temperature region from 350 to 300 °C. Fig. 2 a shows the 0 r > k = = f(Ncu) measurements at various stages of precipitation. Copper concentration was fixed by the cooling temperature. In order to investigate the precipitation of copper during slow cooling, the samples were quenched to fixed temperatures, then chemical analysis was done and the spectral distribution of luminescence was measured. I t can be seen that the experimental dependence ~ A 7 ^ 1 in Fig. 2 b holds under the condition Ncu > Na, assuming that the electroneutrality condition [CuZn] = p is replaced by [Cu Zn ] = [ClSe], see column 1, Table 2. Therefore, 0T ~ N0CajN^ = ^cu" and NCi), slowly cooled during 10 days from 1000 to 20 °C exhibited dependences 0 r ~ and 0k ~ (A'fju)' 1 ' 3 , where is the concentration of added Cu (Fig. 2c). That is why the copper precipitation during the slow cooling of the residual concentration of Cu is smaller than the initial concentration, particularly at high values NCa . Subsequently at more prolonged cooling the experimental points in Fig. 2 b are moved to smaller concentrations and the step of the shift is increased with increasing Cu concentration. In this way the obtained function i>r k = f(Nca) has slopes approximately — 1, which is in agreement with the curves in Fig. 2. The chemical nature of recombination centres in ZnSe identified by this independent method coincided with that determined in [27], In addition to this dominant defects of Cu and CI are clarified and predominant oppositely charged defects in ZnSe: Cu: CI as well (Table 3). The nature of the r-centre CuznClge in ZnSe (Table 3) is analogous to the G-Cu centre in zinc sulphide, see Section 1. On the basis of the dependence of the r-emission band in undoped ZnSe on thermal treatment in zinc or selenium atmosphere Sheinkman and Belenkii [30] assumed that the r-centre is an associate involving Cuzn as an acceptor and some native defect as donor (Zn i; for example). Even undoped A n B v r compounds always contain metallic impurities (Cu, Ag, Al) and non-metallic impurities (CI, Br, I) as well with concentrations of the order of 10 16 to 10 16 cm - 3 . Therefore, there is remarkable probability to form donor-acceptor associates of impurity ions. Especially this is possible in doped crystals, as in our experiments. Recently the associates of impurity donors and acceptors were found in doped ZnS crystals by infrared absorption measurements [31]. From our experiment it was found that the k-centre is composed of copper ions — the donor-acceptor associate CujCuZn or the molecular centre (CuZn)2. The former is in agreement with the structure of the B-Cu centre in zinc sulphide, see Section 1. The molecular centre (CuZn)2, see Section 1, is similar to the (SeAS)2 centre in GaAs [32] or the (Vua)2 centre in NaCl-type crystals [33], The bindings of these centres are analogous to those which form two-atomic molecules in gaseous atmosphere [33], The ionization energies of centres, formed in A n B V I compounds by Cu and CI doping, are summarized in [34], In zinc selenide there are the CI donor level at 0.33 eV below the bottom of conduction band [35] and the Cu acceptor levels at 0.35 and 0.72 eV above the top of the valence band [36]. According to the Williams-Prener transition model (cf. [1, 37, 38]), the photon energies calculated from the relation, hco = Eg — (Ex + Ejy) + e 2 /er AD (Eg band-gap energy, EA and ED acceptor and, donor ionization energies, respectively, e electron charge, e static dielectric constant r A D donor-acceptor separation; Eg = 2.7 eV [7], EA = 0.35 and 0.72 eV, /? D = = 0.33 eV, e 2 /er AD = 0.2 to 0.7 eV [39]) are in good accordance with those observed experimentally (A = 540 nm hoj = 2.26 eV; I = 630 nm hco = 1.94 eV). Cu; (N c u JWci) and Clse (-^ci -^Cu) a s s-centres are deep donors in zinc selenide. In accordance to [40] deep donors, having a large capture cross-section for electrons Sn = 10~ 14 to 1()~16 cm2 because of Coulomb attraction, may be the non-radiative recombination s-centres in A n B V I compounds. Hoshina and Kawai [41] (see also [17]) assumed that Cuj act as non-radiative recombination centres in uncompensated zinc sulphide crystals containing copper. The three-particle defect analogous to VZn(C]ge)2 was found in chlorine-doped CdSe platelets [42], Predominant copper defects in ZnSe: Cu: CI (JV C u > JVCI) are similar to copper defects found in CdS: Cu from high-temperature conductivity and Cu solubility measurements [43],

P. KUKK et al.: Structure of Recombination Centres in ZnSe

42

References [1] M. AVEN and J . S. PKENEK (Ed.), Physics and Chemistry of I I —VI Compounds, NorthHolland Publ. Co., Amsterdam 1967. [2] K. EBA, S. SHIONOYA, and Y. WASHIZAWA, J . Phys. Chem. Solids 29, 1827 (1968). [3] K . EBA, S. SHIONOYA, Y . WASHIZAWA, a n d H . OHMATSU, J . P h y s . Chem. Solids 29, 1843 (1968). [ 4 ] K . URABE, S. SHIONOYA, a n d A . SUZUKI, J . P h y s . S o c . J a p a n 2 5 , 1 6 1 1 (1968).

[5] F. A. KRÖGER, J . chem. Phys. 20, 345 (1952). [6] N. RIEHL and H. ORTMAN, Ann. Phys. (Leipzig) 4, 3 (1959). [7] A. M. GURVICH, IZV. A k a d . N a u k S S S R , Ser. fiz. 30, 644 (1966); 40, 1904 (1976).

[8] A. M. GURVICH, Uspekhi Khim. 35, 1495 (1966). [9] G . Y . B . BIRKLE, F . F . GAVRILOV, a n d G . A . KITAEV, IZV. VUZOV, S e r . f i z . 2 3 , 3 8 ( 1 9 7 9 ) . [10] C. S. KANG, P . BEVEBLEY, P . PHIPPS, a n d R . H . BUBE. P h y s . R e v . 1 5 6 , 9 9 8 (1967).

D. CUBIE, Luminescence in Crystals, Wiley, New York 1963. [11] S. SHIONOYA, K . ERA, a n d Y . WASHIZAWA, J . P h y s . S o c . J a p a n 21, 1 6 2 4 (1966). [12] A . SUZUKI a n d S. SHIONOYA, J . L u m . 3 , 7 4 (1970).

[13] M. TABEI, S. SHIONOYA, and H. OHMATSU, Japan J . appl. Phys. 14, 240 (1975). [14] T. HOSHINA a n d H . KAWAI, J a p a n J . appl. P h y s . 19, 267, 279 (1980). [15] F . J . BRYANT a n d P . S. MANNING, J . P h y s . Chem. Solids 35, 97 (1974). [16] K . MATSUURA a n d J . TSURUMI, J . P h y s . S o c . J a p a n 3 9 , 3 8 3 (1975). [17] H . KAWAI, S. KUBONIWA, a n d T . HOSHINA. J a p a n J . a p p l . P h y s . 1 3 , 1 5 9 3 (1974).

[18] G. H. BLOUNT, A. C. SANDERSON, and R. H. BUBE, J . appl. Phys. 38, 4409 (1967). [19] A. SUZUKI and S. SHIONOYA, J . Phys. Soc. Japan 31, 1455, 1462 (1971). [20] W . LEHMANN, J . E l e c t r o c h e m . S o c . 1 1 3 , 4 4 9 (1966).

[21] F. F. MOREHEAD, J . Phys. Chem. Solids 24, 37 (1963). [22] S. LABACH, J . chem. Phys. 21, 756 (1953). [23] R . M. DETWEILER a n d B. A. KULP, P h y s . R e v . 146, 513 (1966). [24] B . A . K U L P a n d R . M . DETWEILER, P h y s . R e v . 1 2 9 , 2 4 2 2 (1963). [25] F . J . BRYANT a n d A . F . J . C o x , P r o c . R o y S o c . A 3 1 0 , 3 1 9 (1969).

[26] P. L. KUKK, IZV. Akad. Nauk SSSR, Ser. neorg. Mater. 16, 1509 (1980). [27] P. L. KUKK and ö . V. PALMRE, IZV. Akad. Nauk SSSR, Ser. neorg. Mater. 16, 1916 (1980). [28] P. L. KUKK and A. J . ERM, phys. stat. sol. (a) 67, 395 (1981); Zh. neorg. Khim. 26, 2310 (1981). [29] M . K . SHEINKMAN,

N . E . KORSUNSKAYA,

I . V . MARKEVICH,

and

T . Y . TORSCHINSKAYA,

Izv. Akad. Nauk SSSR, Ser. fiz. 40, 2290 (1976). [30] M. K. SHEINKMAN and G. L. BELENKII, Fiz. Tekh. Poluprov. 2, 1635 (1968). [31] A . KRÖL, W . NAZAREWICZ, L . GLTJZINSKI, a n d M . J . KOZIELSKT, p h y s . s t a t . sol. ( b ) 1 0 6 , 4 8 9

(1981). [32] A. M. GURVICH and M. A. ILJINA, in: Problemöfizikisoedinenii A n B V I , Vilnius 1972 (p. 325). [33] F. A. KRÖGER, The Chemistry of Imperfect Crystals, Vol. 2, 2nd revised edition, NorthHolland Publ. Co./American Elsevier Publ. Co, Amsterdam/London/New York 1974. [34] R . H . BUBE a n d E . LIND, P h y s . R e v . 1 1 0 , 1 0 4 0 (1958). [35] G. JONES a n d J . WOODS, J . L u m . 9 , 3 8 9 ( 1 9 7 4 ) . [36] G . B . STBINGFELLOW a n d R . H . BUBE, P h y s . R e v . 1 7 1 , 9 0 3 (1968).

[37] E . F . APPLE a n d F . WILLIAMS, J . E l e c t r o c h e m . Soc. 106, 224 (1959). [38] J . S. PRENER a n d F . E . WILLIAMS, J . E l e c t r o c h e m . S o c . 1 0 3 , 3 4 3 (1956). [39] A . M . PAVELETS, I . B . ERMOLOVICH, G . A . FEDORUS, a n d M . K . SHEINKMAN, U k r . f i z . Z h . 19, 4 0 6 (1974). [40] V . F . GRIN, A . V . LYUBCHENKO, E . A . SALKOV, a n d M . K . SHEINKMAN, F i z . T e k h . P o l u p r o v .

9, 303 (1975). [41] T . HOSHINA a n d H . KAWAI, J . L u m . 1 2 / 1 3 , 4 5 3 ( 1 9 7 6 ) .

[42] B. M. AROBA and W. DALE COMPTON, J . appl. Phys. 43, 4499 (1972). [43] P . L. KUKK, H . A. AARNA, a n d M. VOOGNE, p h y s . s t a t . sol. (a) 63, 389 (1981). (Received

November

5,

1981)

L . N . SKTJJA

and

A . R . SILIN

: Non-Bridging Oxygen Center in Fused Silica

43

phys. stat. sol. (a) 70, 43 (1982) Subject classification: 10.2; 2; 20.1; 20.3; 22.6 Institute

of Solid State Physics,

Latvian

State University,

Riga1)

A Model for the Non-Bridging Oxygen Center in Fused Silica The Dynamic Jahn-Teller Effect By L . N . SKTJJA a n d A . R .

SILZN

A model for the electronic and geometric structure of the non-bridging oxygen defect center in fused silica is presented. The structure of the center is determined by the dynamic Jahn-Teller effect, which appears as a forming of an additional oxygen-oxygen bond between the non-bridging oxygen atom and another oxygen atom from the same silicon-oxygen tetrahedron. ripenjiOHteHa MOHejib ojieKTpoHHOii h reoMeTpHqecKoft CTpyKTypH ueHTpa HeMocTHKOBoro aTOMa KHCJiopona B cTenjioo6pa3Hoii jiByonwcH KpeivraHH. CTpyKTypa ueHTpa onpeaejiHeTCH jjHHaMHHecKHM 3(|)(J)eKTOM HHa-Tejuiepa, KOTopwii npoaBJineTca KaK oSpaaoBamie HOnOJIHHTejIbHOH 0-0 CBH3H MeiKHy HeMOCTHKOBbIM aTOMOM KHCJIOpOJia H HpyrHM aTOMOM KHCJiopoaa Toro Hte KpeMHHft-KHCJiopoAHoro Te.Tpaanpa.

1. Introduction Fused silica consists of a random network of silicon-oxygen tetrahedra joined together by the corner oxygen atoms (so-called bridging oxygen atoms). The non-bridging oxygen atom is a localized defect state in the glass structure, consisting of one half of the permanent broken S i - 0 bond — the oxygen atom bonded only to one Si0 4 tetrahedron. Such defects are formed as a result of permanent S i - 0 bond breaking under neutron irradiation in any fused silica sample, or by O - H bond breaking under X-ray irradiation in fused silica samples containing hydroxyl [1], The non-bridging oxygen atom defects give rise to optical absorption bands at 4.75 and 2.0 eV, a characteristic red luminescence band at 1.9 eV, which is excited in these absorption bands, and also to an ESR signal consisting of three lines [1, 2], The luminescence and optical absorption bands mentioned above have been observed in several papers [3 to 9], However, a detailed model for the respective defect center, consistent with the available experimental data, has not been offered in these works. I t is important to find out the origin of these bands also from the practical viewpoint, because they appear in fused silica optical fiber waveguides*[7, 8], in Si0 2 films on silicon [5], and in the fused silica windows, exposed to vacuum ultraviolet irradiation [6]. On the basis of an analysis of the electronic structure of vitreous silica and taking into account the unusually small Stokes shift when the red luminescence band is excited in the 2.0 eV absorption band, a suggestion has been made t h a t the absorption band at 2.0 eV and luminescence band at 1.9 eV are caused by electronic transitions between two split atomic-like 2p orbitals of the non-bridging oxygen atoms [2], This model is consistent with the small values ( « 10 -4 ) of the oscillator strengths of these transitions [10], but it leaves unexplained the origin of splitting of the two 2p orbitals Kengaraga 8, 226063 Riga, USSR.

44

L . N . SKTJJA a n d A . R . S I L I N

and also the relatively low value of the luminescence polarization degree (P = 12%, [2]). Here we present a model for the non-bridging oxygen defect center, in which the experimental properties are explained in terms of the dynamic Jahn-Teller effect. 2. Experimental Methods and Results The samples investigated were made of commercial high purity synthetic silica (Suprasil WI, Corning 7940, KSG) neutron irradiated at fluxes of 1016 to 1020 neutrons/cm 2 and neutron irradiated a-quartz (1018 neutrons/cm 2 ). Absorption and photoluminescence spectra were measured using grating monochromators VMS-I and MDR-23, photon counting techniques and on-line computer. The photoluminescence was excited by a He-Ne laser (hv = 1.96 eV). The luminescence spectra were corrected for the spectral sensitivity of the registration channel. The luminescence decay kinetics was measured using home-made time correlated single photon counting system. The luminescence spectrum and the low energy absorption band of the non-bridging oxygen center are shown on Fig. 1. A decrease of the sample temperature down to 6 K does not lead to an emergence of any structure in these spectra. The maximum of the luminescence band shifts from 1.91 to 1.89 eV when the sample temperature is

f\ f km - /HLJ V

^excitation

1.5

2.0

tight

-

0.20

0.10

2.5

Fig. 1. The photoluminescence spectrum (1) and the low energy absorption band (2) of the nonbridging oxygen center. T = 293 K, sample : type KSG fused silica irradiated by 1020 neutrons/cm2

3.0 hv(eV)—-

30 tips)—

Fig. 2. The dependence of the 1.9 eV luminescence band intensity on temperature Fig. 3. The decay kinetics of the 1.9 eV luminescence band intensity

(hvex

(hvex

= 2.1 eV)

=

1.96 eV)

45

A Model for the Non-Bridging Oxygen Center in Fused Silica

changed from 293 to 80 K . The dependence of the luminescence intensity on temperature is shown on Fig. 2 and the decay kinetics on Fig. 3. Lines in the E S R spectrum of the non-bridging oxygen center broadens strongly when the sample temperature is raised from 77 to 300 K . 3. Discussion On the basis of the results presented above and also of these published previously in [1, 2, 10] the origin of the splitting of two 2p orbitals of the non-bridging oxygen atom will be discussed. 3.1 The luminescence

polarisation

degree

and the symmetry

of the

center

The relatively low photoluminescence polarization degree ( ~ 12%) is one of the most interesting parameters. The value of the oscillator strength for transitions corresponding to 2.0 eV absorption and 1.9 eV luminescence bands indicate that these are partly forbidden electric dipole transitions, because magnetic dipole and electric quadrupole transitions have lower values of the oscillator strength [11], The luminescence polarization degree in the case of electric dipole transitions between non-degenerate states of randomly oriented centers theoretically can reach 5 0 % when the sample is excited by linearly polarized light [11], If the symmetry group of the center contains a threefold (or higher order) symmetry axis, i.e., if an orbitally degenerate state is present, the spatial orientations of the dipole momenta of absorbing and emitting transitions involving this state may differ (this problem is discussed in detail in [11]). The polarization degree then may fall to 1/7 (:=» 14%) which is close to the experimental value

(12%).

On the basis of the polarization data alone thus it may seem that the non-bridging oxygen atom centers are situated in sites with relatively high symmetry. I t is well known that high symmetry in the fused silica structure remains in the short range (within Si0 4 tetrahedron), but there is a random orientation of neighbouring tetrahedra: the Si-O-Si bond angle shows a distribution extending all the way from 120° to 180° [12], Therefore, high symmetry surroundings for the non-bridging oxygen atom may exist if it interacts effectively only with the atoms of its own silicon-oxygen tetrahedron, and all the other interactions are negligible. Thus it may seem that the non-bridging oxygen atom is situated at the corner of the Si0 4 tetrahedron and has the C3V symmetry of its surroundings (Fig. 4). B u t in such a case the atomic-like 2p orbitals of the non-bridging oxygen must remain degenerate, contrary to the model proposed previously [2, 10]. This discrepancy may be eliminated assuming that the non-bridging oxygen atom is displaced from the threefold symmetry axis of the tetrahedron and rotates around z

Q

0

Fig. 4. The non-bridging oxygen atom at the corner of a non-distorted Si0 4 tetrahedron, c-bonds are schematically denoted by heavy lines

46

L . N . SKUJA a n d A . R . S I L I N

it with a period, shorter than the lifetime of the excited state (T «J 20 ¡JLS, see Fig. 3). In such a case the symmetry of the center will be lower than C3V and the 2p orbitals will be split but the partial depolarization of luminescence will occur due to transitions from equally populated different minima of the upper adiabatic potential surface. The reasons of such a situation will be discussed in the next paragraph. 3.2 The dynamic Jahn- Teller effect and a model for the electronic of the center

structure

The non-bridging oxygen at the corner of the regular Si0 4 tetrahedron is shown on Fig. 4. In the symmetry group C3V the and 2pw orbitals of the non-bridging oxygen atom belong to the E type irreducible representation. The normal vibrational modes in the x and y directions (Fig. 4) of the non-bridging oxygen atom also belong to the same representation. I t follows from the Jahn-Teller theorem [13, 14] that the interaction of E type electronic states with E type vibrational modes makes the high symmetry configuration unstable and a lowering of symmetry occurs. In our case this means that the non-bridging oxygen atom will move away spontaneously from the C3 axis of the tetrahedron and the degeneracy of 2p orbitals will be lifted. The adiabatic potential surfaces for the case of interaction between E type electronic states and E type vibrational modes are well known (see for example [13]). I n the first approximation of the perturbation theory and taking into account only linear electron-phonon coupling the adiabatic potential surfaces have an axial symmetry (the so-called "Mexican hat", Fig. 5). Therefore, one may expect that the nonbridging oxygen should rotate around the C3 axis of the Si0 4 tetrahedron, i.e., that the dynamic Jahn-Teller effect takes place. A more accurate approximation gives three potential wells separated by saddles both on lower and upper sheets of the potential surfaces. Therefore, the rotation of the non-bridging oxygen atom may require a thermal activation energy, although a nuclear tunnelling between wells may take place at low temperatures. The dynamic effects give rise to pseudocentrifugal forces [14] which cause an additional term proportional to (x2 -f y2)'1 in the potential surfaces of Fig. 5a. Fig. 5 b shows an axial cut of potential surfaces including this term. Due to these forces a minimum appears on the upper sheet which prevents from a direct non-radiative multi-phonon transition to the lower sheet, hence the luminescence of an excited center can appear. On the basis of the above discussion the absorption band at 2.0 eV and luminescence band at 1.9 eV can be ascribed to electronic transitions between 2p type orbitals split

Fig. 5. a) The adiabatic potential surfaces for the case of E - E interaction without taking into account the dynamic effects (from [13]). b) A n axial cut of potential surfaces in the case of the dynamic Jahn-Teller effect (from [14]). The dashed lines correspond to the case when the contribution of the pseudocentrifugal forces is included

47

A Model for the Non-Bridging Oxygen Center in Fused Silica

by the dynamic Jahn-Teller effect. Rotation of the non-bridging oxygen atom in the excited state of the center causes partial depolarization of the luminescence but rotation in the ground state — a broadening of ESR lines when the sample temperature is increased from 77 to 300 K . The Jahn-Teller splitting energy in this model ( 2 eV) is relatively large as compared to the most investigated cases of transition metal ion impurity centers where the splitting is typically of order 0.1 eV [13], Nevertheless for the defects in covalent crystals the Jahn-Teller energies can be as high as 1.5 eV (vacancy in silicon) and 2.5 eV (vacancy in diamond) [14]. Therefore, our proposed splitting value of 2 eV seems not unrealistic. However, one must note t h a t distortions leading to so large splittings, cannot be regarded as small perturbations, therefore, the Jahn-Teller effect theory used above may be applied only qualitatively in our case. 3.3 The additional

oxygen—oxygen

bonding

From the viewpoint of the chemical bonding theory the reason for the Jahn-Teller effect is the formation of a chemical bond between the central atom and the ligands. I n our case this corresponds to the formation of the 0 - 0 bond between the nonbridging oxygen atom and one of the three oxygen atoms in the same Si0 4 tetrahedron, caused by an overlap of their 2p orbitals (Fig. 6 a). I t has been proposed [15, 16] that the formation of similar 0 - 0 bond occurs in the case of the hole selftrapping in the regular silica glass network with the bond energy 0.4 eV. The value of 0 - 0 bonding energy in our case is about 2 eV (see Fig. 6b). The stronger 0 - 0 bond between the non-bridging oxygen and oxygen ligands as compared to the hole selftrapping case is what may be expected because the presence of the second S i - 0 bond in the latter case should resist to the distortion of the tetrahedron and thus reduce the 0 - 0 bonding energy. The energy levels of the molecular orbitals for the additional 0 - 0 bonding are. depicted schematically on Fig. 6b. The symmetry group of the center is C3, the bonding and antibonding orbitals are denoted as a' and a'*, a " is the 2p lone pair "non-bonding" orbital of the non-bridging oxygen atom (2pa; in Fig. 6a). The lone pair orbital energy should slightly increase during bond formation because of Coulomb interelectronic repulsion. The optical absorption in the 2.0 eV band is then due to a transition a " -»• a'*. This corresponds to the hole transition from the ground state, partially localized on the three oxygen ligand atoms, to the excited state, almost completely localized on the non-bridging oxygen atom. Taking into account the discussion in the preceding paragraph the hole hops between the three oxygen ligands in the ground state of the center. I n this aspect the

2porbital /,) of ligand^ a'i |1 oxygen ~n—\

F? /

\ /

2pxand2py Iff orbitals of the non-bridging

*

b Fig. 6. The model for the non-bridging oxygen center, a) schematic illustration of the overlapping of the 2p orbitals in the distorted Si0 4 tetrahedron, b) energy level diagram corresponding to the 0 - 0 bond formation

L. N. S k t j j a and A. R. S i l i n

48

proposed model is similar to the well-known aluminium hole center in a-quartz, where the hole is localized on 2p orbitals of ligand oxygens and performs thermally stimulated hopping between them (see [17] and references therein). 3.4 The influence

of the surroundings

of the center on

luminescence

T h e model discussed a b o v e assumes t h a t t h e non-bridging oxygen a t o m i n t e r a c t s only with the a t o m s of its own S i 0 4 tetrahedron. T h e influence of strong e x t e r n a l low symmetry interactions should suppress t h e dynamic effects and t h e luminescence must disappear. T o avoid such interactions t h e a t o m s of other S i 0 4 tetrahedra must be sufficiently remote, i.e. there m u s t be large interstices. I n t h e case of a-quartz t h e free volume between t h e S i 0 4 tetrahedra reaches up t o 7 0 % of t h e t o t a l volume. T h e lowering of t h e density of glassy S i 0 2 (2.20 g/cm 3 ) as compared to a - q u a r t z (2.65 g/cm 3 ) is almost entirely due to an increase of interstices. Furthermore, t h e sizes of interstices = in t h e regular S i 0 2 glass network have some distribution and can be a s large as 4 A [18]. Therefore, one could expect t h a t t h e glassy s t a t e of S i 0 2 is essential for the existence of t h e non-bridging o x y g e n centers. Our experimental d a t a shows t h a t neutron irradiation (10 1 8 neutrons/cm 2 ) of a - q u a r t z produces a t least an order of magnitude smaller number of non-bridging o x y g e n centers t h a n in similarly irradiated fused silica. Geometric considerations exclude t h e stabilization of t h e non-bridging oxygen a t o m s in the crystalline l a t t i c e [10]. H e n c e it seems likely t h a t these centers are introduced in the amorphized regions of crystal caused b y t h e r m a l spikes. I t follows from t h e proposed model t h a t such centers m a y occur as surface defects. W e suppose t h a t experimental results of [19] support this n o t i o n : an e x c i t a t i o n of porous V y c o r glass (consisting of « 9 6 % S i 0 2 ) b y t h e argon laser (hv = 2 . 4 1 eV) produces a red luminescence b a n d peaking a t 1.92 eV. . Assuming t h a t t h e t h e r m a l quenching of t h e luminescence is caused by slight interactions with low s y m m e t r y surroundings, one could expect a wide range of t h e thermal a c t i v a t i o n energies because of the distribution [18] in t h e sizes of interstices. T h e non-exponential luminescence decay (Fig. 3) m a y be explained by e x i s t e n c e of such distribution in a c t i v a t i o n energies. T h e non-exponential luminescence decay is not caused b y t h e distribution in t h e radiative r a t e s because in this case t h e weak temperature dependence of the decay kinetics cannot be explained simultaneously with t h e strong temperature dependence of t h e luminescence intensity (compare F i g . 2 and 3). 3.5 The model of the non-bridging

oxygen center in alkali silicate

glasses

W e suppose t h a t in t h e framework of t h e proposed model the properties of nonbridging oxygen centers in alkali silicate glasses (the so-called HC Z centers [20]) can also be explained. I t has been assumed previously t h a t t h e H C j center is formed b y I hole trapping on t h e formation of t y p e — S i - O - N a followed b y drifting a w a y of the N a + ion [20]. T h e HC X center differs from t h e non-bridging oxygen center in pure fused silica b y t h e existence of the another close alkali ion located on t h e neighboring tetrahedron [20], T h e splitting of t h e 2p orbitals of t h e HC X center has been ascribed to t h e electrostatic field of t h a t ion. B o t h centers have a n u m b e r of similar properties: t h e optical absorption b a n d of the H C j center is also a t 2.0 eV [21], b o t h centers have a relatively high stability against optical or thermal bleaching, t h e E S R signals of both centers are placed in t h e same region of ^-factors.

49

A Model for the Non-Bridging Oxygen Center in Fused Silica

However, these centers sharply differ by a complete lack of luminescence of the HCj center. We propose the following explanation to this phenomenon. The splitting of 2p orbitals of the HCj center is also caused by the formation of the 0 - 0 bonding, similar to that for the non-bridging oxygen center in pure fused silica. The field of the alkali ion on the neighboring tetrahedron serves merely as a low symmetry perturbation which suppresses the dynamic effects. Therefore, minima on the upper sheet of the potential surfaces (Fig. 5 b) do not exist and the luminescence is completely quenched. This model is supported by the lack of temperature-dependent dynamic effects in the ESR spectrum of the HCX centers [22], Moreover, the ESR spectrum of HCj centers in samples enriched by isotope 1 7 0 contains hyperfine lines which may indicate an interaction of the unpaired spin simultaneously with two non-equivalent 1 7 0 nuclei [23], This finding is in accordance with our model. References [1] A. R. SILIN, L. N. SKUJA, and A. V. SHENDRIK, Fiz. i Khim. Stekla 4, 405 (1978). [2] L. N. SKUJA and A. R. SILIN, phys. stat. sol. (a) 56, K l l (1979). [3] G. H . SIGEL, JR., J . non-crystall. Solids 13, 372 (1973/74). [4] C. M . GEE a n d M . KASTNER, T h e P h y s i c s of M O S I n s u l a t o r s , E d . G . LUCIOVSKY, S. T . PANTE-

LIDES, and F. L. GALEENER, Pergamon Press, 1980 (p. 132.) [5] S. W . MCKNIGHT a n d E . D . PALIK, J . n o n - c r y s t a l l . S o l i d s 4 0 , 5 9 5 (1980). [6] S. LANGE a n d W . H . TURNER, A p p l . O p t i c s 1 2 , 1 7 3 3 (1973). [7] P . L . MATTERN, L . M . WATKINS, C. D . SKOOG, J . R . BRANDON, a n d E . H . BARSIS,

IEEE

Trans. Nuclear. Sci. 21, 81 (1974). [8] P. KAISER, J . Opt. Soc. Amer. 64, 475 (1974). [9] G . N . GREAVES, J . n o n - c r y s t a l l . S o l i d s 3 2 , 2 9 5 (1979). [10] A . R . SILIN, L . N . SKUJA, a n d A . N . TRUKHIN, J . n o n - c r y s t a l l . S o l i d s 3 8 & 3 9 , 1 9 5 ( 1 9 8 0 ) .

[11] P. P. FEOFILOV, The Polarized Luminescence of Atoms, Molecules, and Crystals, Gos. Izd. Fiz. Mat. Lit., Moskva 1959 (in Russian). [12] YE. A. PORAI-KOSHITZ, Fiz. K h i m . Stekla 3, 292 (1977).

[13] M. D. STURGE, Adv. Solid State Phys. 20, 91 (1967). [14] A. M. STONEHAM, The Theory of Defects in Solids, Izd. Mir, Moscow 1978 (in Russian). [15] N. F. MOTT a n d A. M. STONEHAM, J . P h y s . C 10, 3391 (1977).

[16] N. F. MOTT, Physics of SiO, and Its Interfaces, Ed. S. T. PANTELIDES, Pergamon Press, Oxford 1978 (p. 1.). [17] N. KOUMVAKALIS, J . appl. Phys. 51, 5528 (1980). [18] J . F . SHACKELFORD a n d J . S. MASARYK, J . n o n - c r y s t a l l . S o l i d s 3 0 , 127 (1978).

[19] C. A. MURRAY and T. J. GREYTAK, Phys. Rev. B 20, 3368 (1979). [20] D. L. GRISCOM, J . non-crystall. Solids 31, 241 (1978). [21] ,T. H . MACKEY, H . L . SMITH, a n d A . HALPERIN, J . P h y s . C h e m . S o l i d s 2 7 , 1 7 5 9 ( 1 9 6 6 ) . [22] E . A . ZAMOTRYNSKAYA, L . A . TORGASHINOVA, a n d V . F . ANUFRIENKO, I z v . A k a d . N a u k

SSSR, Ser. neorg. Khim. 8, 1136 (1972). [23] D. L. GRISCOM, J . non-crystall. Solids 40, 211 (1980). (Received

4

phvsiea( a) 70/1

November

12,

1981)

H . BINCZYCKA et al.: Mössbauer Effect and Electrical Conducti v y

51

phys. stat. sol. (a) 70, 51 (1982) Subject classification: 5 and 14.3; 2 ; 22.6.1 Institute Institute

of Physics, of Physics,

Jagiellonian University, Cracow (a) Technical University, Gdatf.sk1) (b)

and

Mössbauer Effect and Electrical Conductivity in Te0 2 -Fe 2 0 3 Glasses By H . BINCZYCKA ( a ) , 0 . GZOWSKI ( b ) , L . MTTEAWSKI ( b ) , a n d J . SAWICKI (a) Mössbauer spectra of " F e in iron-tellurium glasses are studied. The contributions of F e (II) a n d Fe(III) are determined and a comparison between Mössbauer and chemical analyses is given. The high electrical conductivity of these glasses is discussed and a mechanism of conductivity through tellurium ions in the iron-tellurium glasses is suggested. Mößbauerspektren von " F e in Eisen-Tellurgläsern werden untersucht. Die Beiträge v o n Fe(II) undFe(IIX) werden bestimmt, und es wird ein Vergleich zwischen Mößbauer- und chemischer Analyse durchgeführt. Die hohe elektrische Leitfähigkeit dieser Gläser wird diskutiert und ein Mechanismus für die Leitfähigkeit infolge von Tellurionen in den Eisen-Tellur-Gläsern vorgeschlagen.

I. Introduction Mössbauer spectroscopy is recognised as a useful method of identifying the oxidation state and the location of iron in glasses. The iron (III) state and iron(II) state in glasses can easily be identified on the basis of different isomer shifts and quadrupole splittings in Mössbauer spectra [1 to 5], The glasses with iron oxides form a rather large group of semiconductive oxide glasses. The majority of investigations conducted concern oxide glasses produced from P 2 0 3 , Si0 2 , and B 2 0 3 [6, 7], Recently, there has been a wide interest in semiconductive glasses in which Te0 2 is the glass-forming factor. The distinguishing factor about the matrix of this glass is that tellurium atoms have unshared pairs of electrons which do not take part in bonding. Lately, the production of binary glasses T e 0 2 - F e 2 0 3 has been found possible [8 to 10], Also some structural characteristics of these glasses have already been studied using neutron diffraction [11] (samples with 8 and 12.5% Fe 2 0 3 ) and Mössbauer spectroscopy [12] samples with 12.5 and 20 % Fe 2 0 3 . I n the present paper the Mössbauer transmission analysis of T e 0 2 - F e 2 0 3 glasses has been extended to the series of samples in the range of composition from 5 to 20 mol% Fe 2 0 3 . The electrical conductivity of these glasses has been measured in the temperature range from t h a t of liquid nitrogen to about 600 K . 2. Experimental The melting of glass was conducted in air at about 1300 K in an alumina crucible using the reagent grade chemicals Te0 2 and Fe 2 0 3 . The glass was poured in the form of button-shaped samples on a brass slab. After the annealing of glass at the temperature of 520 K and polishing, silver electrodes with a guard ring were evaporated on the samples in vacuum for electrical conductivity measurements. The content of x

4*

) Majakowskiego 11/12, 80-952 Gdansk, Poland.

52

H. BiriczYCKA, 0 . Gzowski, L. Mubawski, and J . Sawicki -

F»2*

t ^ h

-1

96 >

i

i

i

l

l l 2 3 velocity (mm Is)

7

2 3 velocity (mmls)

Fig. 2

Fig. 1

Fig. 1. Typical transmission Mossbauer spectra of Fe in iron-tellurium glass (20 F e 2 0 3 (1) at room temperature, (2) at liquid nitrogen temperature 57

80 TeO,).

Fig. 2. Mossbauer transmission spectra of the compositions in mol% (1) (5 F e 2 0 3 + 95 Te0 2 ) at 293 K, (2) (12.5 F e 2 0 3 + 87.5 TeO,) at 293 K, and (3) (12.5 F e 2 0 3 + 87.5Te0 2 ) at 80 K

F e ( I I ) and F e ( I I I ) ions was determined from Mossbauer effect investigations as well as the chemical analysis method. The results of chemical and Mossbauer analysis of glasses are compared in Table 1. Table 1 The results of chemical and Mossbauer analysis of the tested glasses composition before melting (mol%)

results of chemical analysiis of the glass (mol%) Te Fe2+ Fe3+ Fe 2 + /Fetot

95 TeO,-5 Fe.,0 3 92.5 TeO.,-7.5 F e , 0 3 90 T e 0 2 - 1 0 F e , 0 3 87.5 Te0 2 -12.5~ Fe„0 3 85 TeO,-15 F e 2 0 3 ' 82.5 T e 0 2 - 1 7 . 5 Fe,O a 80 TeO,-20 Fe,O s "

95.23 92.85 89.75 87.53 85.78 83.02 78.06

0.24 0.65 0.96 0.94 1.42 1.08 4.38

4.53 6.50 9.29 11.53 12.80 15.90 17.56

0.05 0.09 0.09 0.075 0.10 0.06 0.19

Mossbauer results Fe 2 + /Fetot a* 0 « 0 as 0 g 0.02 g 0.02 g 0.02 0.065

Mossbauer transmission analysis was performed for powder samples of 40 mg/cm 2 thickness encapsulated uniformly in organic glass pellets. A source of 67 Co in Cr with an activity of 20 mCi was used. The Mossbauer spectra measured at 300 and 80 K presented very similar quadrupole-split doublets. Examples of the spectra are presented in Fig. 1 and 2. 3. Discussion of Results The Mossbauer spectra measurements indicate that in the glasses under study iron exists predominantly in the F e ( I I I ) valency state, the contribution of the F e ( I I ) valency state being fairly small. The assignment of both states in Mossbauer spectra is presented in Fig. 1, curve 1. An admixture of F e ( I I ) has been observed only in the sample with 20 mol% F e 2 0 3 , i.e. at the upper limit of the glass-forming range. Comparison of the absorption peak areas for F e ( I I ) and F e ( I I I ) shows that the fractional

Mössbauer Effect and Electrical Conductivity in Te0 2 -Fe 2 0 3 Glasses

53

percentage Fe(II)/Fe t o t amounts to (6.5 4; 2.3)%. The contribution of Fe(II) in samples with smaller mole percentages of F e 2 0 3 , as determined by the statistical accuracy of experimental points in Mossbauer spectra, is less than about 2 % . The results of chemical analysis given in Table 1 indicate a much larger contribution of Fe(II) in all samples, varying between 5 and 2 0 % . The reason for the difference in Fe(II) concentration between chemical analysis and Mossbauer spectroscopy is not clear. However, these two methods show that in all samples the number of Fe(II) ions is small. Mossbauer spectra parameters of Fe(III) are plotted versus iron concentration in Fig. 3 (see also Table 2). The isomer shift data fall in the narrow range IS = 0.37 to 0.39 mm/s (relative to metallic iron) which coincides with the range of isomer shifts of the octahedrally coordinated ferric ion in glasses [4], An increase of the quadrupole splitting with increased F e 2 0 3 content is believed to indicate a gradual lowering of the octahedral symmetry at iron ions along the series of samples. The width of absorption lines is typically two or three times the natural linewidth. This together with the nonLorentzian lineshape can be attributed to the distribution of electric field gradients Table 2 Mössbauer spectra parameters of Fe(III) in Te0 2 -Fe 2 0 3 glasses A composition pellet (mol%) thickness (%) Te0 2 Fe,0„ (mg Fe/cm2)

A ,

1

80 80*) 82.5 85 87.5 87.5*) 90 92.5 95

20 20 17.5 15 12.5 12.5 10 7.5 5

6.32 6.19 5.35 5.37 4.30 5.53 3.04 2.06 1.53

4.76 4.76 4.17 3.61 2.80 2.80 2.07 1.49 0.98

(5) (4) (5) (5) (4) (4) (3) (2) (2)

(mm/s)

(mm/s)

IS (mm/s)

QS (mm/s)

0.440 0.495 0.436 0.431 0.429 0.504 0.434 0.464 0.477

0.407 0.424 0.403 0.392 0.380 0.428 0.399 0.417 0.438

0.371 0.473 0.375 0.380 0.388 0.478 0.383 0.382 0.388

0.895 0.929 0.851 0.827 0.756 0.782 0.772 0.778 0.720

W

(%)

6.15 5.81 5.40 5.42 4.41 5.48 3.08 2.08 1.57

(5) (5) (5) (5) (4) (4) (3) (2) (2)

l

(7) (8) (8) (7) (7) (8) (5) (8) (9)

W

2

(7) (8) (7) (7) (6) (7) (7) (8) (8)

(2) (2) (2) (2) (2) (2) (2) (2) (3)

(4) (5) (5) (4) (4) (4) (4) (5) (5)

S J S

1.11 1.24 1.07 1.08 1.10 1.19 1.07 1.10 1.06

2

(5) (6) (5) (5) (5) (6) (5) (5) (6)

Parameters obtained at room temperature: A amplitude, W linewidth reduced to the zero absorber thickness, IS isomer shift relative to metallic iron value, QS quadrupole splitting, S J S line intensity ratio. Errors (in brackets) do not include the error of the calibration. *) Data obtained at liquid nitrogen temperature. 2

-

i I

i

:

I

5

i

I

I

I

I

I

I

i

i

B

-

a

_ • -

i

1

5

•

'

D

s

f

S

1

10

a 1

ft

15 Fe2 0i contenflmd

Fig. 3. Parameters of the quadrupole doublets of Fe(III) ions vs. concentration of Fe 2 0 3

I

1

20 %)—

54

H . B I X C Z Y C K A , O . GZOWSKI, L . MURAWSKI, a n d J . SAWICKI

at 5 7 Fe nuclei in the amorphous environment. I t is to be noted that the linewidth shows a slight minimum at 12.5 mol% F e 2 0 3 which corresponds to the eutectic concentration. For the Fe(II) state observed in 2 0 % F e 2 0 3 sample, the relevant Mossbauer parameters are: IS = (1.18 ± 0.01) and QS = (2.20 ± 0.03) mm/s. On the basis of comparison of I S and QS data in glasses [4] one can assume that Fe(II) occurs here more probably in octahedral coordination than in tetrahedral. Referring to the data of Kozhukharov et al. [12] it should be mentioned that the values of quadrupole splitting for 12.5% F e 2 0 3 are in full agreement but their value QS for 2 0 % F e 2 0 3 was drastically lower. This discrepancy seems to indicate that the sample was more crystalline than in the present work. The results of electric conduction measurements at 100 °C in the form of a graph of the function of F e 2 0 3 content are shown in Fig. 4. On the same graph corresponding data for the iron-phosphate glasses [13] is presented for comparison.The conduction of iron-tellurium glasses is more than three orders of magnitude greater than that of the equivalent iron-phosphate glasses. As one can see from the analysis of the values of activation energies obtained Table 3) it cannot be said that the increase in conduction is only due to the lowering Table 3 Compositions, activation energy, and the ratio C of the tested glasses glass composition (mol%) 80 T e 0 2 - 2 0 F e 2 0 3 82.5 TeO.,-17.5 F e 2 0 3 85 T e 0 2 - 1 5 F e 2 0 3 87.5 T e 0 2 - 1 2 . 5 F e 2 0 3 90 T e 0 2 - 1 0 F e 2 0 3 92.5 T e 0 2 - 7 . 5 F e 2 0 3 95 T e 0 2 - 5 F e „ 0 3 " 64.5 P 2 0 5 - 3 5 . 5 F e , 0 3 69 P 2 0 5 - 3 1 F e 2 0 3 " 73 P 2 0 5 - 2 7 Fe„0 3 77 P , 0 5 - 2 3 F e 2 0 3 64 P l 0 5 - 2 3 F e 2 0 3 - 1 3 CaO 60 P 2 0 6 - 1 6 F e 2 0 3 - 2 4 CaO 58 P 2 0 5 - 1 3 F e j 0 3 - 2 9 CaO 53 P 2 0 5 - 5 F e 2 0 3 - 4 2 CaO

W (eV) 0.48 0.54 0.55 0.62 0.63 0.66 0.72 0.58 0.61 0.62 0.63 0.64 0.65 0.67 0.69

C

0.19 0.06 0.10 0.075 0.09 0.09 0.05 0.30 0.33 0.31 0.31 0.28 0.25 0.28 0.24

of activation energy of the iron-tellurium glasses. Therefore, we must expect that the reason for such a great increase in conductivity is the pre-exponential factor — more so as the value C = F(II)/Fe t o t is very small for tellurium glasses. Therefore, it can be concluded that the concentration of charge carriers in iron-tellurium glasses may be much greater than one would expect considering the Mott theory [14] of electron hopping between ions of different valence states. According to this theory a = cr0 e x p ( - W f k T ) ,

where c 0 ~ 0(1 — C). The results of Mossbauer experiments performed in this work enable us to reconsider the mechanism of conductivity in the iron-tellurium glasses. The small or zero concentration of Fe(II) ions, as compared with other glasses (Table 3), seems to ex-

Mössbauer Effect and Electrica] Conductivity in Te0,-Fe,0 3 Glasses

55

Fig. 4. The dependence of electrical conductivity on Fe 2 0 3 content (measured at 393 K) for glasses cited in Table 3

10

5

10 .15

20

25

SO

FejOj content imol'h)

elude the possibility of such high conductivity resulting from electron hopping between Fe(II) and Fe(III) ions. Therefore, the Te(IV) network must play the decisive role. The change of the coordination of Te(IV) in the series of compounds Fe 2 Te0 5 , Fe 2 Te 3 0 9 , and Fe 2 Te 4 O u by surrounding oxygen atoms has been discussed in the literature [15]. The authors indicated t h a t in compounds in this system the lone electron pair of the tellurium atom is stereo-chemically active. This means t h a t the introduction of Fe 2 0 3 in the network of Te0 4 polyhedra leads to the formation of new bonds and defects. The number of nonbridging oxygen atoms is increased and micro-holes and new unshared electron pairs are created [12], We suggest t h a t the presence of unshared electron pairs probably influences the conductivity and the mechanism of conductivity through the tellurium ions. References [ 1 ] C. R . KURKJIAN, J . n o n - c r y s t a l l . S o l i d s 3 , 157 ( 1 9 7 0 ) . [2] S. L . RUBY, J . n o n - c r y s t a l l . S o l i d s 8 / 1 0 , 7 8 ( 1 9 7 2 ) .

[3] S. P. TANEJA, C. W. KIMBALL, and J. C. SCHAFFER, in: Mossbauer Effect Methodology, Vol.

8, Ed. I. J. GRUVERMAN, Plenum Press, New York 1973 (p. 41). [4] J. M. D. COEY, J. Physique (Suppl. 12) 35, C6-89 (1974). [ 5 ] J . SAWICIU, B . S A WICK A, a n d 0 . GZOWSKI, p h y s . s t a t . sol. (a) 4 1 , 1 7 3 ( 1 9 7 7 ) .

[6] J. D. MACKENZIE, in: Modern Aspects of the Virteous State, Vol. 3, Ed. J. D. MACKENZIE Butterworths, London 1964 (p. 126). [7] C. H. CHUNG, J. D. MACKENZIE, and L. MURAWSKI, Rev. Chim. miner. 16, 308 (1979). [ 8 ] M. MABINOV, V . KOZHUKHABOV, a n d J . PAVLOVA, C. R . A c a d . B u l g . S c i . 2 6 , 3 4 3 ( 1 9 7 3 ) .

[9] V. KOZHUKHABOV, M. MARINOV, and G. GRIGOROVA, J. non-crystall. Solids 28, 429 (1978).

[10] V. KOZHUKHABOV and M. MARINOV, C. R. Acad. Bulg. Sci. 28, 245 (1975). [ 1 1 ] S. NEOV, I . GERASSIMOVA, K . KREZHOV, B . SYDZHIMOV, a n d V . KOZHUKHABOV, p h y s . s t a t . sol. (a) 47, 7 3 4 ( 1 9 7 8 ) . [ 1 2 ] V . KOZHUKHABOV, S. NIKOLOV, M. MABINOV, a n d T . TBOEV, M a t e r . R e s . B u l l . 1 4 , 7 3 5 ( 1 9 7 9 ) .

[14] N. F. MOTT, J. non-crystall. Solids 1, 1 (1968). [ 1 3 ] L . MURAWSKI a n d O. GZOWSKI, p h y s . s t a t . sol. (a) 1 9 , K 1 2 5 ( 1 9 7 3 ) . [ 1 5 ] R . ASTIER, E . PHILIPPOT, J . MORET, a n d M. MATJR .N, R e v . C h i m . m i n e r . 1 3 , 3 5 9 ( 1 9 7 6 ) . (Received

October 2,

1981)

F . M.

MANSY

et al.: Creep Mechanism and Dissolution Process in Al-Zn

57

phys. stat. sol. (a) 70, 57 (1982) Subject classification: 10.1; 14.1; 21.1 Physics Department, Faculty of Science, University of Cairo, Oiza1) (a) and Physics Department, Faculty of Education, University of Cairo, Fayoum2) (b)

High-Temperature Creep Mechanism and Dissolution Process in Al-9.5 at% Zn Alloy By F . M. M A K S Y (a), N . K . GOBKAN (a), a n d G. SAID (b)

Steady state creep of Al-9.5 at% Zn solid solution is investigated in the temperature range from 600 to 670 K and under applied stresses ranging from 5 to 30 MPa. The energy activating the creep process is found to be 1.6 eV in good agreement with the activation energy for the diffusion of Zn in Al-matrix. Mean internal stresses are measured and the results indicate that the creep deformation is controlled by the viscous glide of dislocations dragging the Cottrell-type solute atmosphere. From the dependence of the internal and effective stress on the applied stress, a value of "n" is estimated and is found to be in good agreement with the measured value. The dissolution process of the ß-phase is studied by employing fracture stress and resistivity measurements. Apeak value is observed and is attributed to dislocation-solute atom interaction. The activation energy involved in this process is found to be 1.3 eV. Stationäres Kriechen von Al-9,5 At% Zn-Festkörperlösungen wird im Temperaturbereich von 600 bis 670 K und unter äußeren Spannungen von 5 bis 30 MPa untersucht. Die Aktivierungsenergie für den Kriechprozeß beträgt 1,6 eV in guter Übereinstimmung mit der Aktivierungsenergie für die Diffusion von Zn in der AI-Matrix. Die mittleren inneren Spannungen werden gemessen und die Ergebnisse zeigen, daß die Kriechdeformation durch viskoses Gleiten von Versetzungen gesteuert wird, die eine Cottrell-Atmosphäre besitzen. Aus der Abhängigkeit der inneren und der effektiven Spannung von der angelegten Spannung wird ein Wert von " » " berechnet und es wird gefunden, daß er sich in guter Übereinstimmung mit dem Meßwert befindet. Der Auflösungsprozeß der ß-Phase wird durch Anwendung von Bruchspannungs- und Widerstandsmessungen untersucht. Ein Maximum wird beobachtet und der Wechselwirkung Versetzungs-Lösungsatom zugeordnet. Die beteiligte Aktivierungsenergie in diesem Prozeß wird zu 1,3 eV gefunden. 1. Introduction Creep behaviour of solid solutions is usually classified into two classes according to the value of the stress exponent n. Creep in class I (n = 3) is believed to be controlled by viscous glide of dislocations [1], On the other hand, the creep in class I I (n = 5) is thought to be controlled by a recovery process [2]. Kucharova and Cadek [3] studied the creep behaviour of Al-9.5 a t % Zn solid solution for which they suggested that non-conservative motion of jogs on screw dislocations was the mechanism controlling the creep rate. However, it was found that climb and glide play concomitantly an important role for Al-Zn alloys of different atomic percentage of Zn in Al matrix [4], In the meantime, for the composition of Al-20 a t % Zn, glide was considered as the main controlling process due to its slower rate as compared to the climb motion. In this investigation the creep characteristics of Al-9.5 a t % Zn alloy have been studied in order to check experimentally and theoretically the correspondence between the value of n and the rate controlling mechanism. I t is also believed that studies of dissolution strengthening characteristics in Al-Zn alloys might lead to better understanding of the dissolution process. !) Giza, Egypt, Fayoum, Egypt.

s)

58

P . M . MANSY, N . K . GO BRAN, a n d G . SAID

2. Experimental Al-9.5 a t % Zn alloy was made from spec-pure metals. The ingots were hot rolled to wires of 1 mm diameter. The spectroscopic analysis of this alloy showed the following impurities in atomic percentage: Cd Cu Pb Fe Mg Sn 0.004 0.002 0.003 0.002 0.002 0.001 Before testing, the specimens were annealed under vacuum at 720 K for a time of 3 h. The mean grain size reached was found to be (40 ± 2) p . In order to study the creep process of the a-solid solution of Al-9.5 a t % Zn alloy, the specimen was clamped in a creep machine and was heated in situ for 1 h at 720 K during which Zn solute atoms dissolve completely in the A1 matrix. The furnace was then cooled slowly to the temperature at which the creep measurement was performed. The precipitated samples were prepared by annealing at 610 K for 2 h and then cooled slowly to room temperature or otherwise stated henceforth. Dissolution of the p-phase was controlled by giving the test samples prescribed heat pulses at the temperatures 520, 540, and 570 K and then quenched to room temperature. Fracture stress measurements were carried out at room temperature under constant strain rate of 3 X 1 0 - 3 s - 1 using a strength tester machine. Moreover, using a standard potentiometer method [5], resistivity measurements were performed on these samples. 3. Results Fig. 1 shows the dependence of the steady-state creep rate on the applied stress at a temperature of 623 K . The stress exponent n was found to be 3.8. At a constant applied stress of 15 MPa, the steady-state creep rate was studied at different temperatures. From this study the activation energy of the creep process was calculated and found to be 1.6 eV. This value is in good agreement with the energy required for activating the diffusion of Zn in A1 matrix [6]. Instantaneous plastic strain associated with sudden stress changes was measured at 623 K and under a base applied stress of 10 MPa. For the same stress change Atf, the elongation A L + accompanying stress increments A