Dielectric loss in RbCl crystals doped with Co2+ ions

HAL Id: jpa-00209162

https://hal.archives-ouvertes.fr/jpa-00209162

Submitted on 1 Jan 1979

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.

Dielectric loss in RbCl crystals doped with Co2+ ions

S.K. Gupta, S.D. Pandey

To cite this version:

S.K. Gupta, S.D. Pandey. Dielectric loss in RbCl crystals doped with Co2+ ions. Journal de Physique,

1979, 40 (8), pp.779-782. �10.1051/jphys:01979004008077900�. �jpa-00209162�

Dielectric loss in RbCl crystals doped ^{with Co2+ ions}

S. K. Gupta ^{and S. D.} Pandey

Physics Department, ^P.P.N. College, Kanpur, 208001, ^India

(Reçu ^{le 25} septembre 1978, ^{révisé le 29 mars} 1979, accepté ^{le 27} avril 1979)

Résumé. 2014 On a étudié les pertes diélectriques ^dans des cristaux de chlorure de rubidium dopés ^{avec des} impuretés

de Co2+. Dans ce cas, la couche externe de l’ion impureté ^est constituée par des électrons d ^et l’ion est beaucoup plus petit que le cation-hôte. Les pics dans les isothermes de tan 03B4 en fonction de la fréquence, obéissent à une rela- tion de type Arrhénius. Les valeurs de l’énergie d’activation pour le dipôle I-V et le facteur pré-exponentiel ^{ont été}

déterminés et comparés ^{aux valeurs} correspondantes dans d’autres systèmes apparentés. ^On envisage ^la possibilité

que les impuretés ^de Co2+ occupent ^{des sites} interstitiels dans le réseau et on propose ^{un nouveau} modèle de relaxa- tion de lacune.

Abstract.

Dielectric loss studies have been made in rubidium chloride crystals doped ^{with Co2+} impurities.

In this case the impurity ^{ion consists} of d-electrons in its outermost shell and is much smaller than the host cation.

The peaks ^{in the tan 03B4 vs.} frequency ^{isotherms follow an Arrhenius type relation.} The values of activation _energy for the I-V dipole ^{and the} pre-exponential factor have been determined and compared ^{with the} corresponding

values in other related systems. ^The Co2+ impurities ^are envisaged ^{to have} occupied interstitial sites in the lattice and a new vacancy relaxation model is proposed.

Classification Physics ^Abstracts

77.40

1. Introduction.

The dielectric behaviour of divalent impurity-vacancy ^{defects in alkali halide} crystals ^has generated much interest. At moderate temperatures, the divalent impurity-host ^cation

vacancy (I-V) dipoles ^{are in} significant concentration and under the action of external a.c. electric fields the I-V dipoles reorient themselves and the resulting

relaxation gives ^{rise to} dielectric loss which exhibits

a Debye peak. ^{The loss} peaks in alkali halides have been reported for various divalent impurities [1-7].

However, ^{the loss} peaks ^{in RbCI} crystals ^{have been} reported only ^{for a few} dopants (impurities) ^viz.

Pb2+ [5],Sr2+ [6], ^{and Bi2+} [7].

Our present ^{work concems} dielectric loss results

on CO2+ doped ^RbCI crystals. ^From optical ^studies Co2+ was assigned ^to interstitial sites in KCI by Washimiya [8] ^{and Musa} [9], ^{while Co2+ in RbCI}

was also assigned ^to interstitial sites by ^Musa [9].

But Trutia and Voda [10] ^from their detailed optical

studies have concluded that most of Co2+ in KCI exists in precipitate ^{form of} composition K2COCl4,

while they ^{have not ruled out the} possibility ^{of a small}

fraction of the total Co2+ existing ^at substitutional sites. We have interpreted ^our dielectric loss results in RbCI : Co2+ whilst correlating the diverse models

reported [3, 8-11] regarding ^the occupation ^{of Co2 +}

in RbCI and KCI. The activation _energy and _pre-

exponential factor for RbCI : Co2+ have been com-

pared ^{with the} corresponding results in other related systems.

2. Expérimental techniques.

^The crystals ^were

grown by ^the Stockbarger technique ^{from a rubidium}

chloride melt containing ^{about 1} % (by weight) ^of CoCl2. Plates of size 7 x 7 x 0.5 mm’ were cut from the bulk of the crystal ^{and were found to} contain blue

precipitate regions ^{visible to the naked} eye. The actual cobalt impurity concentration determination

was made by ^atomic absorption analysis ^{at the} Regional Sophisticated Instrumentation Centre, Madras, India and the RbCI crystals ^{were found to}

contain 2 500 _ppm of cobalt. A thin coating ^{of silver} paint ^{was made on two faces of} crystals ^and they

were then given ^{a heat} treatment at a temperature ^of

N

450 °C for about four hours to remove dipole aggregation. ^The crystals ^were still found to contain blue islands even after annealing.

The dielectric loss factor (tan ô) ^{of the} crystal ^was

measured using ^a three-terminal configuration ^with

the help ^{of a GR 1615-A} capacitance bridge ^combined

with a GR 1311-A Audio Oscillator. For higher ^fre- quencies ^{a PM 5100} Philips Oscillator was used in combination with the capacitance bridge. ^{The cons-}

tancy ^{in the} temperature ^{of the} crystal ^{was attained}

Article published online by EDP Sciences and available at http://dx.doi.org/10.1051/jphys:01979004008077900

780

by using ^a Philips Plastomatic temperature controller.

The bridge ^is extremely sensitive for measuring ^the

dielectric loss factor ^{to an} accuracy of the order of 10-6 and the capacitance could be read in the range 10-17 to 10-6 farad. All measurements were

made in vacuum (- ^10-2 torr) ^{with a suitable} crystal

mount.

3. Results.

The tan b vs. frequency ^isotherms

were obtained for RbCl : Co2+ samples ^{at different} temperatures ^{and were} plotted ^{on a double} log ^scale (see Fig. 1). However, figure ^{1 shows} only ^{three such}

isotherms for the sake of clarity. ^{One loss} peak ^is

found for each isotherm and this shifts to higher frequencies ^{as the} temperature ^is increased. This indicates that as the température ^increases higher frequencies ^{of a.c. fields are} required ^to orient the I-V dipoles ^{in order to} get the relaxation losses.

Further, ^it has been verified that the resulting peak

is not due to interfacial polarization ^{or Maxwell} Wagner losses which could also give ^{rise to a similar} type ^of peak. Figure ^{2 shows the} plot ^of logarithm

Fig. ^1.

Isotherms of tan ô vs. frequency ^{at different} tempera- tures.

Fig.- ^2.

^{Plot of} logarithm ^of peak frequency fm against 1/r.

of the peak frequency log fm ^{as a} function of 1 000/T,

where T is the temperature of measurement of tan ô in degree Kelvin. The plot ^is very nearly ^a straight

line and hence the peak frequency fm ^{can be} expressed

as

where fo ^{and E are the} pre-exponential factor and the activation _energy respectively ^{for the} migration ^{of the}

cation _vacancy bound to the impurity ^{ion. The} slope

and intercept ^{of the above} straight ^line plot give ^the

values of E and fo respectively. These values are given

in table 1 together with the data for other Co2+

doped alkali halides and rubidium chloride crystals doped with different divalent cations.

Table I.

Activation energies ^and pre-exponential factors for ^{the loss} peaks ⁱⁿ different ^{alkali halides} doped with divalent cation impurities.

4. Discussion.

Dreyfus [12] through ^{his studies} of NaCI doped ^{with a} number of divalent cations has found that, ^when the size of the divalent impurity

cation is smaller than the host alkali ion, ^{two dielectric}

loss peaks ^{are observed. He has} attributed one

of these peaks ^to the reorientation of the I-V dipole

when a vacancy is at a nn position whilst the other is that due to a vacancy ^{at a nnn} position. ^In conformity

with this theory Jain and Lal [2] ^{and Jain et al.} [3]

have also found two peaks ^{in NaCl : Co2+ and}

KCI : Co2+ crystals. ^{On the} other hand Varotsos and Miliotis [4] ^{who have made a} systematic study

in as many ^as nine systems ^have always ^observed only

one peak ^{and have} established the following ^{facts :} 1) ^{The loss curves in} alkali halide crystals doped

with Mg2 +, Ca2 +, ^{Sr2+ and Ba2+} impurities ^follow

the simple Debye Theory and the activation _energy of I-V complexes depends upon the radius of the

impurity ion concemed.

2) ^{In the case} of divalent impurity ^cations having

d-electrons on their outer shell _e.g. the C02+ ion, the loss curves are wider than those expected ^from Debye Theory and the activation _energy does not

depend ^on the ion concemed.

3) Further, only ^one peak is observed for _any system irrespective of the relative size of the impurity

ion and the host cation.

Thus the results obtained by Varotsos and Milio- tis [4] contradict the result of Jain and Lal [2] ⁱⁿ

NaCI : C02+.

In the present ^{case of RbCI :} Co2+ also only ^one

loss peak ^has been observed at any temperature.

This is again ^{too wide as observed} by Varotsos and Miliotis [4] ^{in the case of} impurities ^{with outer d-}

electrons.

A divalent cation impurity entering ^{an alkali halide} usually substitutes for the monovalent cation and creates a vacancy necessary for charge compensation.

In the present ^case too, ^if Co2+ ions are assumed to substitute for Rb+ sites, ^the possible ^{vacant cation} positions would be the nn or nnn Rb+ sites. Further,

as Co2+ is too small an ion (0.72 A) ^as compared

to Rb+ (1.48 A), ^{it could} acquire off-centre positions

with respect ^to regular ^Rb+ sites in the RbCI lattice.

Such off-centre _energy minimum positions ^have been experimentally established by Pandey [13] ^{and Kawano}

et al. [14] ^{in RbCI : Eu2+} and also in other _sys- tems [15-17]. Off-centre positions ^{for an} impurity ^ion

can also be visualized on the basis of a trivial classical model according ^{to which a vacant Rb+ site} which’

would behave as a negative charge, ^would try ^to pull

the Co2+ ion towards it. In the case of a vacancy

being ^{at an nnn} position ^this would, further, ^{cause the} dipole ^{to be more} stable. This thus _may be a reason

why ⁿⁿⁿ cation vacancies are favoured in an alkali halide host, if the divalent impurity cation radius is much smaller than the host cation radius.

However, if the initial site for the _vacancy is at an nn position, ^its pull ^on the divalent cation impurity

would shorten the Co2+ - _vacancy dipole ^and may

finally ^{lead to} impurity

^vacancy exchange ^{or else}

as an intermediate stage ^the impurity may prefer ^stable

interstitial sites.

In a RbCI lattice the interstitial sites are surrounded

by four CI- ions at the corners of a regular ^tetra-

hedron and four Rb+ ions at the corners of yet another regular tetrahedron, all in the first coordina- tion sphere of the site of the impurity (cobalt) ^ion.

The occupation of such sites by ^a Co2 + would, ^howe-

ver, ^cause two nearest Rb+ sites to remain vacant for local charge compensation requirement. ^Washi- miya [8] ^{in order to} explain ^the optical absorption ^of Co2+ in KCI has given ^{the reason that Co2+ is most}

stabilized in the Td ^field (compared ^to octahedral Oh).

Trutia and Voda [10] have made detailed optical absorption studies in alkali chlorides MCI doped ^with Co2+ impurities. ^In each of these samples, ^{over and}

above the observation of crystal ^field spectra, they

report ^an absorption band in the uv region ^{due to the}

presence of regions ^of precipitates ^of M2CoCl4

or other combinations. Structurally M2COCI4 ^is thought ^to consist of (CoCl4)2- ions surrounded by

M + ions in the second coordination sphere ^{of C02 + .}

Trutia and Voda have, accordingly, explained ^the absorption band in the ultraviolet spectrum ^of

K2CoCl4 ^to originate ^{from a} transfer of an electron from a (CoCl4)2- complex ^{to a K+ ion and} similarly

for other systems. They ^have reported blue island like structures of K2CoCl4 in their KCI : Co2+ crystals.

Our RbCI : Co2+ crystals also consisted of similar blue structures ; therefore the existence of Rb2COCI4 precipitate ^{in our} crystals ^is expected. The dielectric relaxation phenomenon, ^{however, cannot be envi-} saged ^{with a} Rb2COCI4 compound ^without dipoles.

In order to explain ^{our loss results, we, therefore,}

invoke the need for the existence of interstitial Co2 + sites associated with the two vacancies in RbCI : Co2+.

The simultaneous observation of two different

phenomena ^{in two similar} systems i.e. the existence of dielectric loss peaks ^{in our} crystals ^and optical absorp-

tion in RbCI : Co2+ as observed by Trutia and Voda, lead one to believe that there probably ^{exists an} equilibrium ratio of interstitial cobalt and precipitated Rb2CoCl4 complex ^{in RbCl : C02+. An} interstitial Co2+ associated with two cation vacancies makes the

resulting ^ionic arrangement ^around Co2+ empirically equivalent ^{to a} Rb2COCI4 molecule. The crystal

structure data is not known for Rb2CoCl4. However, the Co2+ situated at interstitial sites _may suitably

relax so as to lead to the formation of Rb2COCI4

or vice versa. Nevertheless, ^{under an} equilibrium ^the

fraction of Co2+ situated at interstitial sites should be small enough ^to preclude their detection through optical absorption ^studies.

With the above interstitial model for Co2+ the dielectric loss peak ^{in our RbCI : Co2+} system may be attributed to the jump ^{of one of the nn} vacancies in the first coordination sphere ^{of Co2+ to Rb+} sites in the second coordination sphere. ^{Further, as per Trutia}

and Voda one may ^not rule out a small _presence of substitutional cobalt ions. But the quantity ^{of such}

cobalt should be sufficiently ^{small not to lead to} any additional features in the dielectric loss phenomenon

at least in the temperature range of our study.

Substitutional sites for cobalt have been reported by

Jain and Lal [2] ^{and Jain et al.} [3] ^{in NaCl : C02+ and}

KCI : Co2+ respectively. ^The proposal ^{of an intersti-}

tial model thus apparently ^{seems to} contradict their conclusions. The impurity concentration in their

samples was, however, small and about 5-7 _ppm.

It thus _appears that Co2+ impurities ^{in low concen-}

trations (10 ppm range) occupy only the substitu- tional sites in an alkali halide but an increased dose of

impurity ^at least in RbCI : Co 2 + leads to preferred

interstitial sites and finally ^to precipitate ^formation

of the type M2CoCI4.

Varotsos and Miliotis [4] have observed one loss

peak in each of their systems. They ^{have not} reported

the concentrations of impurities ^{in their} samples.

It seems that in their samples impurity concentration

was large ^{as in} ours so as to lead to broad peaks.

782

As shown in table I, the activation _energy of dipoles corresponding ^to different divalent cations in RbCI shows a variation in the _range 0.51-0.83 eV. The values of pre-exponential ^factors fo ^{in different}

systems, however, ^{show an} appreciable ^variation

from one system ^{to the other} e.g. for Co2+ in RbCI it is less than the corresponding value for the NaCI host by three orders of magnitude. ^The reciprocal ^of fo determines the relaxation time of the _{vacancy r,} and is characteristic of the impurity-vacancy arrange- ment. A small value of fo ^{and a} subsequent large ro for Co2 + in RbCI, accordingly, indicates that the relaxation mechanism for the _vacancy associated with the divalent ion in this system ^{would not be as} simple

as for a substitutional site of C02 +. This thus seems

to corroborate with the hypothesis ^{of an} interstitial model for impurity ^{in RbCI : Co2 + for which} large To is expected.

It would be worthwhile to point ^out that the value of

fo ^has always ^{found to be} quite ^{small in} systems ^with

impurity ion radius much smaller than that of the host cation _e.g. KCI : Be2+ [15] ^{RbCI : Mn2+} [18], ^and

RbCI : Ni2+ [18]. A consideration of the interstitial model of the impurity ^{in such} systems, accordingly,

seems to provide ^a fruitful exercise.

Acknowledgments.

^{We are thankful to Prof.}

C. Ramasastry of Indian Institute of Technology,

Madras for providing experimental facilities and useful

suggestions during ^{the course} of this work.

References

[1] HAVEN, Y., ^{J. Chem.} Phys. ²¹ (1953) ^171.

[2] JAIN, ^{S. C. and} LAL, K., Proc. Phys. ^{Soc. 92} (1967) ^990.

[3] JAIN, ^S. C., LAL, ^{K. and} MITRA, V., ^J. Phys. ^{C 3} (1970) ^2420.

[4] VAROTSOS, P. and MILIOTIS, D., ^J. Phys. Chem. Solids 35

(1974) ^927.

[5] MISHRA, ^{K. D. and} SHARMA, ^M. N., ^J. Phys. ^Soc. Japan ³⁶ (1974) ^154.

[6] WILLIAMS, ^G. P., Jr. and MULLIS, D., Phys. Status Solidi (a) 28 (1975) ^539.

[7] RADHAKRISHNA, S. and SRINIVASA SETTY, ^R. S., Phys. ^Rev.

B 14 (1976) ^969.

[8] WASHIMIYA, S., ^J. Phys. ^Soc. Japan ¹⁸ (1963) ^1719.

[9] MusA, M., Phys. Status Solidi (a) ¹⁶ (1966) ^771.

[10] TRUTIA, ^{A. and} VODA, M., ^{J. Chem.} Phys. ⁶⁴ (1976) ^2715.

[11] HILLS, ^M. E., ^J. Phys. ^Soc. Japan ¹⁹ (1964) ^760.

[12] DREYFUS, ^R. W., Phys. ^{Rev. 121} (1961) ^1675.

[13] PANDEY, ^S. D., Chem. Phys. ^{Lett. 3} (1969) ^646.

[14] KAWANO, K., NAKATA, R., SUMITA, ^{M. and} HIGUCHI, E.,

J. Phys. ^Soc. Japan ⁴¹ (1976) ^72.

[15] BUCCI, C., Phys. ^{Rev. 164} (1967) ^1200.

[16] HARRISON, J. P., PERESSINI, P. P. and POHL, R. O., Phys. ^Rev.

171 (1968) ^15.

[17] LI, ^M. S., ^DE SOUZA, ^{M. and} LUTY, F., Phys. ^{Rev. 7} (1973) 4677.

[18] MURIN, ^D. V., KARNEV, ^{B. F. and} GLUMOV, ^A. V., ^Sov. Phys.

Solid State 18 (1976) ^1963.