ESR and optical absorption of cupric ion in borate glasses

ESR and optical absorption of cupric ion in borate glasses

Journal of Non-Crystalline Solids 34 (1979) 339-356 © North-Holland Publishing Company ESR AND OPTICAL ABSORPTION OF CUPRIC ION IN BORATE GLASSES H. ...

682KB Sizes 0 Downloads 15 Views

Journal of Non-Crystalline Solids 34 (1979) 339-356 © North-Holland Publishing Company

ESR AND OPTICAL ABSORPTION OF CUPRIC ION IN BORATE GLASSES H. HOSONO, H. KAWAZOE and T. K A N A Z A W A

Department of Industrial Chemistry, Faculty of Technology, Tokyo Metropolitan University, Fukasawa, Setagaya-ku, Tokyo, 158, Japan Received 19 May ?.979 Revised manuscript received 1 August 1979

ESR and optical absorption of Cu 2+ were measured in xNa20(100-x)B203 (1 < x :g 75), x ZnO(100-x)B203 (46 < x < 64) and x PbO(100-x)B203 (20 < x < 75) glasses, where x is expressed in mol. %. Spin hamiltonian parameters and ligand field absorption energy changed abruptly in the regions of 15 < x < 23 and 37 < x <--55 in the soda system, while both parameters were hardly dependent upon the glass composition in zinc and lead systems. The magnitude of micro-environmental fluctuation of Cu2+-complexes in the glasses was estimated qualitatively and correlated with the distribution of the strength of n-bonding between cupric ion and oxygen in the glass. Typical network modifiers and intermediates behaved differently, especially in the composition region of invert glass; the large deformation of the coordination sphere of Cug+ in lead glasses due to the stronger Pb-O bond resulted in the large distribution of g~~.The situation was reverse in the case of soda glasses.

1. Introduction The composition dependence o f structure o f borate glasses has been studied by many authors through the responses o f incorporated transition metal cations such as ESR [ 1 - 4 ] , UV-visible absorption [ 5 - 1 0 ] , fluorescence [11,12] and the M6ssbauer effect [13]. However, the composition region studied was limited to one of the two glass-forming regions, 1 0 - 3 8 tool.% Na20 [14] and the other region, 6 6 . 5 - 7 1 . 5 mol.% Na20 [14] has been disregarded. Recently, Sakka et al. [15] united the two separate glass-forming regions b y adding a small amount o f Al203 to the intermediate non-glass forming region, finding that the ligand field absorption peak o f Cu 2+ shifted abruptly toward higher energies in the range of 4 0 - 5 0 mol.% Na20 content. In preceding papers [ 1 6 - 2 0 ] , we have characterized some oxide glasses from the viewpoint o f the a c i d - b a s e relationship determined b y ESR o f incorporated Cu 2÷ ions. The purpose o f this paper is to show the dependence of ESR and optical absorption spectra of.Cu 2+ on the Na20 concentration in the very wide composition region ( 1 - 7 5 mol.%) o f the N a 2 0 - B 2 0 3 system, to clarify differences 339

340

H. Hosono et al. /ESR and optical absorption

between the effects of typical network modifiers and intermediate oxides on the responses of Cu 2÷, and to correlate the Lewis basicity proposed by Imagawa [1 ] for sodium, lead and zinc borate glasses with the calculated optical basicity (Aca0 proposed by Duffy and Ingram [21].

2.

Experimental

Composition of the glasses used in the present study were xNazO(100-x)B203 (1 < x < 75), x PbO(t00-x)B203 (20 < x < 75) and xZnO(100-x)B2Oa (46 ~: x < 64). The starting materials were reagent grade Na2CO3, ZnO, PbO, CuO and HaBO3. 0.2 mol.% CuO was added to each batch. Most of the glass batches were fused in a platinum crucible in air. Alumina crucible were used for lead borate glasses. Sodium borates with x = 45 and 55, which correspond to the compositions outside the glass-forming regions, were vitrified by adding 3 mol.% Al203. The melts were poured onto a stainless steel plate and pressed with another one into a thin glass plate. No annealing was done. Hygroscopic samples were immediately immersed in liquid paraffin after the quenching to avoid hydration of glass surface. Optical absorption spectra of Cu 2÷ in glasses were recorded on Shimadzu UV200 and -201A spectrophotometers at room temperature. The glasses were crushed into pieces for ESR measurement. ESR spectra were recorded at 20 and -100°C on a JEOL JES PE-3X spectrometer for X-band measurements and an ME-3X instrument for K-band. A magnetic field modulation of 100 kHz was applied. DPPH was used as a standard for the determination of the spin hamiltonian parameters. Optical absorption and ESR spectra of Cu 2+ in [Cu(OH)4] 2- complex were also measured. The complex was prepared in glassy DMSO(dimethyl sulfoxide)-water matrix (1:1 in volume) at lower temperature [22].

3. Resets 3.1. ESR 3.1.1. Sodium borate glasses

Three distinct spectra I, II and III were obtained for low-, high- and ultra-highalkali glasses, as shown in fig. 1. Characteristics of the spectra are summarized in table 1. Spectra I and II were already reported [1,17,23] and spectrum III has been newly observed. Two features can be read from the line shape of the spectra. First, linewidth (AH) of the parallel hyperfine structure (hfs) increases with increasing m (nuclear magnetic quantum number of the copper nucleus), although the dependence of AH on m is different for I, II and III. The magnitude of AH is greater in the order

H. Hosono et al. / ESR and optical absorption

341

t~

"0 L

o\ t~

t~

t~

\ I

© Z

8

342

H. Hosono et al. / E S R and optical absorption

Table 1 Characteristics of three representative ESR spectra (I, II and III) of Cu2+ observed in the xNa 20 (100-x) B203 glass system. g//

IA// L(10-4 cm-1)

I

ca. 2.37

ca. 138

II

ca. 2.33

III

ca. 2.26

Na20 content (mol.%)

Splitting of perpendicular hfs

Linewidth of parallel hfs peaks

5 ~
Ambiguous

Slight increase in the order of m

ca. 153

20 ~
Resolved

Considerable increase in the order of m

ca. 175

55 <~x ~<75

Well-resolved

Narrow and independent of m

II > I > > III. Second, the perpendicular splitting becomes clearer in the order I < II < < III. Extra divergence is observed in spectrum III [ 2 4 - 2 6 ] . Spectroscopic splitting (g) and hyperfine (A) tensors with axial symmetry have been assumed in the analysis of the spectra of oxide glasses [1,2,17,19,23]. The axial symmetry has been confirmed by measuring K-band spectra which remove the overlap between parallel and perpendicular hfs. Figure 2 showing the perpendicular hfs for ultra-

r--r--r-q ra

=

-~

'2-1/2+1/2+3/2 f

x=55

/ I

I

I

8400

I 8500

I

I 8600

I 8700

MAGNETICFIELD (GAUSS)

Fig. 2. K-band ESR spectra of Cu2+ in ultra-high-soda composition (perpendicular hyperfine structure), x is the Na20 content in mol.%.

H. Hosono et aL / ESR and optical absorption

343

high-soda glasses indicates that axial symmetry still holds tbr these glasses. It is natural, therefore, to employ the following axial spin hamiltonian in the analysis of the spectra, ~C = g//~HzS z + ga[3(HxSx + H y S y ) + A //IzSz + A ±(IxSx + I y g y ) .

(1)

The solutions of the spin hamiltonian (1) are given in eq. (2a) and (2b) for the parallel and perpendicular hf peaks, respectively [24]: hv = g//f3H + mAll + (~- - m2)(A~/2gfl3H) ,

(2a)

hu = g±f3H + m A ± + ( ~ - m~)(A~/ + A~)/4gll3H .

(2b)

We employ the first order perturbation theory, considering that the second order perturbation term in negligibly small. The g//and IA//I, determined from the positions of sharper peaks (those of m = -3•2 and - 1 / 2 ) using eq. (2a), are shown in fig. 3(a). Stepwise changes o f g / / a n d IA//I are observed in the regions around x = 17 and around x = 45. The former transition region was noticed by several authors [1,2,17,23], whereas the latter has been found in this work. Magnitude of the change in g// (or IA//I) in the latter region is greater than that in the former. The spectra of the glasses in the former transition region (x = 15 and 17) are recognized as the superposition of spectra I and II, which is most clearly seen in the parallel hf peak with m = - 1 / 2 as shown in fig. 4. Similar phenomena were observed in other alkali borate glasses [17,23]. It appears that these two regions correspond to transitions from spectra I to II and II to III. Low soda glasses (x ~ 5) also exhibit two sets of hfs (fig. 5). One is regarded as spectrum I. The other is characterized by the same values of ESR parameters as spectrum II. However, it differs from II in the dependence of AH on m and thermal stability. This spectrum is abbreviated as II'. Characteristics of spectrum II and II' are given in table 2. The coordination structure of the CuZ+-complex which produces spectrum II' and its thermal relaxation will be discussed elsewhere. It may be noted that in the spectra measured at the K-band (fig. 6), splitting of parallel hf shoulders is faint for spectrum I, unrecognizable for II and well-resolved for III. Extra broadening caused by increasing microwave frequency will be discussed in terms of micro-environmental fluctuation in section 4.2. 3.1.2. Zinc borate and lead borate glasses

Representative X-band spectra are shown in fig. 7. Line profile and spin hamiltonian parameters for these glasses are found to be almost invariant in the whole range of glass-formation (fig. 3b, 3c). These spectra correspond to spectrum II observed in high Na20 glasses. The composition dependence of the characteristics of the spectra is not so marked as that of Na20-B203 glasses (fig. 3a). It is seen that the line width of parallel hfs for an ultra-high-lead glass (x = 75) increases considerably in order of m (fig. 7) and with increasing microwave frequency (fig. 6), which is not found for ultra-high-soda glasses.

2.25

2.30

I

I

I

I

1

I

I

130

NA20 (MOL%) (a)

i0 20 30 40 50 60 70 80

-

"T'

150 ~

170

190

2.25

2,30

2.35

2.40

L

/x

O

l

l

1

l

I

o /x

l

I

l

I

o A

l

1

0 A

i

I

l

I

170

--

130

I

i

t

,

t

i

I

I

i

I

l

I

I

I|

ZN0 C~TENT ( MOLI ) (c)

~:

%

i--170

i0 20 30 40 50 60 70 80 90 i00

r

,

(b)

PB0 CONTENT ( MOL%

'7, -- 150 %

-

i0 20 30 40 50 60 70 80 90 i00

1

l

Fig. 3. Variation o f ESR parameters o f Cu 2÷ with Na20 (a), PbO (b) and ZnO (c) content. Open circles and triangles indicate g//and IA//I, respectively.

{2)

=

2.35

2,40

2.25

2,30

2,35

2,40

O"

.8

H. Hosono et al. / ESR and optical absorption

345

/ ,0 s. " ' ' ° ' " /

I

s st"

/

!

/ 2600

,

,

I

2800

3000

MAGNETIC

I

3400

FIELD ( GAUSS )

Fig. 4. ESR spectra o f Cu 2+ in 1 5 N a 2 0 - 8 5 B 2 0 a glass: s p e c t r u m 1; . . . . ; s p e c t r u m II. (See table 1.)

"_+ ~L+ ~'_-÷4 '

2600

I

3200

2800 MAGNETIC

.z

; 15Na20-85B203

xd

3000 3200 FIELD (6AOSS)

3400

Fig. 5. E S R s p e c t r a of Cu 2+ i n x N a 2 0 ( l O O - x ) B 2 0 3 glasses (x < 10).

; ......

;

346

H. Hosono et al. / ESR and optical absorption

Table 2 Characteristics of spectrum II and II'. g//

IA//I (10 --4 cm -1)

Na20 content (mol.%)

Linewidth of parallel hfs peaks

Stability upon annealing

II

ca. 2.33

ca. 154

20 < x <~ 37

Considerable increase in the order of m

Stable

II'

ca. 2.33

ca. 154

x<5

Slight increase in the order of m

Unstable

3.2. Optical absorption The absorption band located in the region o f 12 500 cm -1 can be attributed to elongated octahedral coordination o f cupric ions [27,28]. Several workers measured the composition dependence of the peak energy (AE) of this band for N a 2 0 B203 glasses [ 7 - 1 0 , 1 5 ] . However, as far as the authors are aware, no data have been reported for zinc borate glasses, lead borate glasses and sodium borate glasses with low soda composition (x < 5). Figure 8 shows the variation o f the peak position for the three systems with B203 content. Dependence o f the peak energy on B203 content is marked for Na20 glasses, whereas it is vague for ZnO and PbO glasses. In the sodium system three transition regions are recognized as 5 ~ x (first), 15 ~ x ~ 23 (second) and 37 ~ x ~ 55 (third). The first transition region of AE wa~ characterized by its marked blue shift, which is compatible with the appearance of spectrum II' in ESR of low Na20 glasses. Table 3 shows the variation o f ESR spectral patterns and optical absorption energy with the Na20 content. Three invariant and transition regions appear alternatively in the whole range o f glass formation.

Table 3 Trend of changes in ESR and optical absorption of Cu 2÷ in xNa20(100-x) B203 glasses. Na20 content (mol.%)

ESR spectrum

Optical absorption energy (cm -1 )

Region

x~ 5 5~x 15 ~ x 23 <~x 37 ~
I + II I I + II II

12 800-12 500 12 500 12 500-12 800 12 800 12 800-15 200 15 200

Transition Invariant Transition Invariant Transition Invariant

~< 13 ~ 23 ~< 37 ~<55 <~ 75

III

8000

8400

MAGNETIC FIELD (GAuss)

7600

/

8800

7200

8400

FIELD(GAuss)

8000

MAGNETIC

7600

/

f

I

I v

I

I/ i

I

8800

'

i "~". . . .

Fig. 6. K-band ESR spectra ot Cu 2+ in xNa20(100-x)B203 and 75PbO 25B203 glasses. The intensity of perpendicular hyperfine peaks is ten times larger than that of parallel ones.

7200

I (X=lO)

,/

i i

g~

r~

H. Hosono et al. / ESR and optical absorption

348

I

I

I

2600

]

I

2800

I

i

3000

3200

MAGNETICFIELD (GAuss) Fig. 7. E S R s p e c t r a of Cu 2+ in xPbO(lOO-x)B20 3 glasses.

I

I

I

I

I

I

I

_

1.2

800 1.3

1.4 ~o

1.5

,=,

1,6 600

V0

I 10

I 20

1 30

I 40

I 50

I 60

I 70

80

RO(NA20) C~TENT(MoL%) Fig. 8. Variation of peak wavelengths of Cu 2+ in borate glasses with B 2 0 3 content.

H. Hosono et aL / ESR and optical absorption

349

4. Discussion 4.1. Analysis by L CA O - M O The spin hamiltonian parameters and ligand field absorption energy were correlated to the bonding parameters between Cu 2÷ and ligands, I'o and Prr, by using Maki and McGarvey's LCAO MO analysis [29,30]. Calculation of the bonding parameters was conducted using the same procedures as in the previous works [17,19]. The covalencies, I'o and I'Tr, for soda glasses are plotted against the Na20 content in fig. 9 and those for zinc and lead systems are in fig. 10, where l-'o and PTr mean the covalency of the Cu2+-O o-bonding and the in-plane pTr-dlr n-bonding, respectively. The results are summarized as follows: (1) Po seems to be independent of the glass composition and has a value o f about 40%. (2) rTr increases nearly stepwisely around x = 17 and x = 40 in the case o f Na20 glasses. The increase at x = 17 is similar to the results of Imagawa [ 1 ]. This pheno-

70 60

I

I

I

'

I

J

l

I

I

I

l

50 00000

40 30

~

~OoO

2o

~ lO o

rl

nlJ

10 20 30 40 50 60 70 ~ 70 60 50

[

I

I

I

I

,

r

ZNO CONTENT( ~ ) l l l l l l

t

60

I

50

40

40

0

0

30

30

@



~ 20 0

l

i

90 100

..J I

1

L

I

]

]

i

0 i0 20 30 40 50 60 70 80 NA20CONTENT(MOL%)

0

0

2o

o

i , , t I I I t t 10 20 30 40 50 60 70 80 90 100 PBO CONTENT(MOL%)

Fig. 9. Bonding parameters between cupric ion and oxygen for Na20-B203 glasses: o, Fo; O,

rff.

Fig. 10. Bonding parameters between cupric ion and oxygen for zinc and lead borate glasses: o, Fo; o, Fir.

350

H. Hosono et al. /ESR and optical absorption

menon may be caused by the change of the ligand borate groups from penta- or tri-borate group to the tetra- or di-borate group [31,32]. The molecular orbital calculation made assuming that no non-bridging oxygens are formed in this composition region predicted that the Lewis basicity of the ligand system would increase upon the change of borate structures [16]. The increase of Frr in the vicinity of x = 40 may be attributed to the occurrence of non-bridging oxygens. Such an explanation is consistent with the results obtained by refined NMR measurements [32,33]. (3) Fzr slightly decreases with decreasing B203 content in PbO glasses, whereas it remains constant in ZnO glasses. Here, we will confine ourselves to discussing the correlation between PTr and the calculated optical basicity (Acal) which is expected to express the Lewis Basicity of bulk glasses. Variation of mcaI with the glass composition are shown in fig. 11 for the Na20 and ZnO systems, mcaI increases remarkably with decreasing B203 content in Na20 glasses, while it remains unchanged in ZnO glasses. Roughly speaking, such composition dependences of AcaI resemble those of I~Trin both systems. As Acal is defined to be a monotonous function of the composition, it cannot predict the abrupt change of Fzr in Na20 glasses. AcaI of the lead system could not be calculated, because Pauling's electronegativity of divalent lead ion prerequisite for the calculation has not been reported yet. 4.2. Micro-environmental fluctuation o f Cu2+-complexes in borate glasses

In general, perfect shape fitting of magnetic resonance spectra for glassy matedais is impossible if one considers only an ensemble of spins having mutual dipolar interactions and random orientation to the applied magnetic fields. There remains an additional broadening effect, which is attributed to the micro-environmental fluctuation around a spin [24]. Here, we discuss structural distribution in borate glasses detected through the ESR of Cu2*. First, we briefly review the semiquantative treatment of Imagawa [ 1 ].

1,2

,

f

|

i

t

a

l

i

1,0 0,8 0.6 0,4 0,2 0,0

20 qO 60 80 100 NA20 (ZNO) CONTENT

(MOL%)

Fig. 11. Calculated optical basicity (Aca1) for sodium and zinc borate systems.

H. Hosono et aL / ESR and optical absorption

351

He assumes that as the value of Po's of Cu2+-complexes does not depend on the chemical composition of the glasses, the distribution for an ensemble of Cu 2+ in the glass under consideration may be negligible. This assumption leads to the expression; g//- A de = K

(3)

where P and K are constants. In the present experiments, the average value of K is 2.752 and the standard deviation, OK, is 0.006, indicating that the above assumption is satisfactory. On the other hand Pn is very sensitive to glass composition (fig. 9). We may expect that dg//and dA//, fluctuations of g# and A,, respectively, cause dH, fluctuation of H according to eq. (4) which is derived from (2a) 1 [hv--dg// g,l~ gll

dH

|-Tld'4~ dg//l"]

(4)

Then, the extra peak-width z3J/(m) due to the fluctuation is related to the variation of g#, ~g// AII (m) = 6g//

hv + rn(g//P - A/l) g2/3

= 8g//(Cv + m × CF) ( G ) ,

(5)

where Cv and CF are expressed as c.

=

CF

=

(g//e

-

A//)/g~/3 ,

(6)

respectively. Second, we examine the validity of the above treatment qualitatively in a simple experiment. In fig. 12 X-band and K-band ESR spectra of Cu 2+ in zinc metaphosphate glass and its polycrystalline form are shown in fig. 12. The polycrystalline form was obtained by devitrifying the glass at 600°C for 24 h and the crystalline phase was identified as H-zinc metaphosphate. It may be noted in the figure that z2d-/(m) for the crystalline material is nearly independent of m and v, whereas it increases with increasing m and v for the glassy sample. These results are in good agreement with the prediction given by eq. (5). Third, the treatment was applied to the ESR of Cu 2+ in the borate glasses. Values of the proportional constants, Cv and Cv [eq. (5)] for the spectra I, II and III were calculated from the parameters given in fig. 3 and table 4. Three points are easily recognized from the table: (1) in spite of the greater difference in glass composition, CF which depends only on g// and All concentrates around 400; (2) another constant C~, which depends on both g# and v, has nearly the same value for all the specimens, when compared at the same v. The deviation of the constant is about 10% at most; (3) the constant Cv for the K-band measurement is approximately three times as large as that for the X-band measurement. This is naturally

352

H. Hosono et al. / ESR and optical absorption

J

2600

270O

280O

'2980

~0

3000

~ETIC FIELD(G) f,a)

I

I

I

7000 72O0 74O0 t~lf'flC FIELD(G) (b)

I

7f~O

Fig. 12. ESR spectra of Cu2+ in zinc metaphosphate. Parallel hyperfine structures are shown. The upper traces are the spectra for polycrystal and the lower curves for glassy ZnP206. (a) X-band (~9.52 GHz). (b) K-band (~26 GHz).

predicted from the definition of C v [eq. (6)]. In spite of the small deviation among CF and C~ for each band, an additional broadening can be seen for each glass sample (figs. 1 and 6). Consequently, the observed AH(m) is substantially attributed to the fluctuation of g// [eq. (5)]. As noted in section 3.1, 5g// for the representative spectra decreases in the order II > I > > I I I . Last, we try to interpret the behavior of 6g// from the chemical point of view. A boron atom and a cupric ion having an available vacant p-orbital and 3d hole orbital are Lewis acids. An oxygen atom is a Lewis base and coordinates to a boron or borons and a Cu 2÷ simultaneously. Therefore, the boron atom and cupric ion are in mutual competition for attracting lone pairs of the oxygen. Double bond character of B - O and Cu2+-O bonds is controlled by donation of

Table 4 Values of proportional constants for accounting for additional broadening observed in the ESR spectra of Cu2+. Spectrum

Cv a (gauss)

Cv b (gauss)

CF (gauss)

I (x c = 10) II (x = 30) III (x = 70) 75 PbO-25 B203

1220 1260 1340 1270

3070 3190 3370 3190

380 390 420 390

a At X-band frequency (v = 9.52 GHz). b At K-band frequency (v = 24.0 GHz). c x indicates the N a 2 0 content in mol.% for N a 2 0 - B 2 0 a glasses.

H. Hosono et al. / ESR and optical absorption

353

lone pairs of the oxygen. This means that correlation between strength of B - O and Cu2+-O bonds is negative. In the composition range where spectrum I appears (5 <- x < 13), group structures with less basic oxygens (boroxol, pentaborate or triborate) are dominant according to the results of MO calculation [16], namely double bond character of B - O bond is strong. Consequently, the strnegth of Cu2+-O bond in the competition with the stronger B - O bond is suppressed to a low strength level. Then, the spectrum I with small ~gn and PTr would be expected. In the composition range 55 < x < 75 appeared spectrum III with the narrowest grdistribution. Here, discrete pyro- and ortho-borate ions are expected to form from the chemical composition. The nominal vacant p-orbital of three coordinated borons in these borate ions is saturated with adjoining non-bridging oxygens carrying negative charges. An excess charge on the oxygen results in strengthening of competitive Cu2+-O bondings. This would lead to the formation of Cu 2+ coordinated by ortho- or pyro-borate ions. The coordination structure of these complexes are considered to be framed up most regularly because the structural distortion required for glassy structure concentrates on the coordination sphere of the alkali ions which are more deformable. Thus, in this composition region, spectrum III characterized by the strongest Cu2+-O bonds and the narrowest distribution of g// is expected to appear. The complex [Cu(OH)4] 2- in DMSO-water glass serves as an extreme example. The bond strength of glass network, being constructed by hydrogen bonding, is very weak, whereas that of Cu2+-OH - bond is extraordinarily strong, because OH- is highly basic. The 5g~ is surprisingly so small that even the splitting between parallel hfs produced by 63Cu and 6SCu can be observed in the hfs shoulder for m --- - 3 / 2 (fig. 13). The structural distortion tends to concentrate on most deformable, weak points in glassy materials. In the composition region 20 <--x <-- 37, several group structures coexist [31,32] and the oxygens expected to be a ligand for Cu 2÷ in the respective groups have distinct electron densities whose values distribute in the region of the critical basicity for the coordination structure of Cu2+-complex;the resulting bond strength of Cu2÷-O bonds has a large distribution. Then, 8gg of spectrum II is expected to be the largest. The results of above discussions are schematically shown in fig. 14. It is worth noting that in a high lead glass of 75PbO-25B203, discrete borate ions are expected to be predominant anion species, but the magnitude of its 8g//is as large as that of spectrum II observed in soda glass. It has been considered that Pb 2+ can behave as a network former in the low borate glasses [34-36] owing to its large polarizability [37,38] originated from the soft-character of its outer electron pair (6S2), i.e., Pb 2+ makes a part of the network. This means that most of orthoborate present in the glass cannot be regarded as "free" ion (fig. 15). In this case, Cu2÷-O and Pb2+-O bonding are in competition with each other in constructing each favorite coordination structure. Distinction between a typical network modifier (Na ÷) and an intermediate (Pb 2÷) becomes pronounced when viewed through the magnitude of spin hamiltonian parameters and their distributions especially for the invert glass [39,40] region.

354

H. Hosono et al. / ESR and optical absorption

63Cu2+ 65Cu2~ .)

I 2600

I

Ij

I I 2800 3000 MAGNETC IFIELD( GAUSS)

I 3200

I 3400

I

Fig. 13. ESR spectra of [Cu(OH)4 ]2- complex in DMSO-water (1:1 in volume) matrix (pH = 12.7) at -100°C. Note that the splitting due to the two isotopes of the copper nucleus is found in the peak rn = - 3 / 2 of parallel hyperfine structure.

III I

,o-

-NA+O NA+-Of"O-NA+ Cu2+-0 BONDSTRENGTH

( a

o-g~°

-_ . , o - P B -

)

-O.. pB2+O__B. -0,~-0( b

)

Fig. 14. Distribution of Cu2+-O bond strength (schematic): I, 5 < x < 13; II, 20 < x <-- 38; III, 55 < x K 75 (x is the Na20 content in mol.%). Fig. 15. Schematic representation of the structure for invert glass: (a) 70Na20-30B203 (N4 = 0 [41]). (b) 75PbO-25B203 (N4 ~- 0.2 [35,36]).

5. C o n c l u s i o n E S R and optical absorption spectra o f Cu 2+ were measured for sodium, zinc and lead-borate glasses whose c o m p o s i t i o n s covered the whole range o f glass-formation.

H. Hosono et al. / ESR and optical absorption

355

Results obtained are summarized below: (1) an elongated octahedral coordination of Cu2*-complex is held in all the glasses measured. (2) Spin hamiltonian parameters (g//and IA//I) and ligand field absorption energy of Cu 2÷ vary stepwisely in the vicity o f x = 15 and x = 45 in x N a 2 0 (100-x)B203 glasses. The magnitude of the variation in the latter composition range is larger than that in the former. The former is attributed to the basicity change due to the change among group structures without non-bridging oxygen and the latter due to the occurrence of non-bridging oxygens, respectively. In contrast with the sodium system, marked dependence of the parameters on the B203 content is not observed for the lead and zinc systems. (3) Imagawa basicity (FTr) and calculated optical basicity (Acal) are in a rough agreement for those glasses. (4) The magnitude of g//-distribution varies strikingly with the soda content. It decreased for the three types of spectra in the order, II (high soda, 20 < x ~ 37) > I (low soda, 5 < x < 13) > > III (ultra-high-soda, 55 <--x <-- 75). Distribution of g// for the spectrum III is negligibly small. This was interpreted on the assumption that the formation of Cu2÷-discrete ion complexes such as Cu2+-orthoborate and Cu 2÷pyroborate is expected from the chemical composition of the glass and the strong Cu2÷-O bond in the complexes. (5) On the other hand, the magnitude of gg-distribution (Sg//) of Cu 2÷ in lead orthobor~ite glasses is considerably larger than that in ultra-high-soda composition. The difference between/Sg//for ultra-high-soda and lead glasses suggests that "free" orthoborate ions are not present in the lead glass as major anions because lead ions make a part of glass network with their large polarization.

Acknowledgements The authors wish to thank Dr. H. Imagawa of Tokyo Shibaura Electric Co. Ltd., for helpful suggestions and discussions and Prof. M. Sukigara of University of Tokyo for facilities for performing K-band ESR measurements. We appreciate Prof. S. Sakka of Mie University for reading the manuscript. This work was partly supported by the Asahi Glass Foundation for Industrial Technology.

References [1] [2] [3] [4] [5] [6]

H. Imagawa, Phys. Stat. Solidi 30 (1968) 469. H.G. Hecht, Phys. Chem. Glasses9 (1968) 178. H.G. Hecht and T.S. Johnston, J. Chem. Phys. 46 (1967) 23. H. Toyuki and S. Akagi, Phys. Chem. Glasses 13 (1972) 15. N.A. Ghoneim and H.A. Elbatal, J. Non-Crystalline Solids 12 (1973) 189. A. Paul and R. Douglas, Phys. Chem. Glasses 9 (1968) 21.

356

H. Hosono et aL /ESR and optical absorption

[7] R. Juza, H. Seidel and J. Tiedemann, Angew. Chem. 78 (1966) 41. [8] A.A. Ahmed and G.N. Ashour, Recent Advances in Science and Technology of Materials, ed., A. Bishay, Vol. 2 (Plenum Press, New York, 1974) p. 83. [9] C.R. Bamford, Phys. Chem. Glasses 3 (1962) 189. [10] J.A. Duffy and M.D. Grant, ibid. 16 (1975) 124. [11] S. Park and R.S. Webb, Phys. Chem. Glasses 13 (1972) 157. [12] M.I. Weber, J. Hegarty, D.H. Blackburn, Borate Glasses, eds., L.D. Pye, V.D. Frechette and N.J. Kreidl (Plenum Press, New York, 1978) p. 215. [13] T. Raman and G.N. Rao, J. Non-Crystalline Solids 29 (1978) 85. [14] M. Imaoka, Advances in Glass Technology, Part I (Plenum Press, New York, 1962) 149. [15] S. Sakka, K. Kamiya and H. Yoshikawa, J. Non-Crystalline Solids 27 (1978) 289. [16] H. Kawazoe, H. Hosono and T. Kanazawa, J. Non-Crystalline Solids 29 (1978) 159. [17] H. Kawazoe, H. Hosono and T. Kanazawa, ibid. 29 (1978) 173. [18] H. Kawazoe, H. Hosono and T. Kanazawa, Yogyo-Kyokai-Shi, 87 (1979) 1 (in Japanese). [19] H. Hosono, H. Kawazoe and T. Kanazawa, J. Non-Crystalline Solids 33 (1979) 103. [20] H. Hosono, H. Kawazoe and T. Kanazawa, ibid. 33 (1979) 125. [21] J.A. Duffy and M.D. Ingrain, J. Inorg. Nucl. Chem. 37 (1975) 1203. [22] Y.H. Chao and D.R. Kearns, J. Phys. Chem. 81 (1977) 666. [23] L.D. Bogomolova, V.A. Zhachkin, V.N. Lazukin, N.F. Shapovalova and V.A. Shmukler, Fiz. Tverd. Tela 12 (1970) 3370. [24] P.C. Taylor, J.F. Baugher and H.M. Kriz, Chem. Rev. 75 (1975) 203. [25] R. Neiman and D. Kivelson, J. Chem. Phys. 35 (1961) 156. [26] H.R. Gersman and J.D. Swalen, ibid. 41 (1964) 379. [27] C.J. Ballhausen, Introduction to Ligand Field Theory (McGraw-Hill, New York, 1962). [28] T. Bates, Modern Aspects of the Vitreous State, Vol. 2, (Butterworths, London, 1962) p. 195. [29] A.H. Maki and B.R. McGarvey, J. Chem. Phys. 29 (1958) 31. [30] D. Kivelson and R. Neiman, ibid. 35 (1961) 149. [31] W.L. Konijnendijk and J.M. Stevels, J. Non-Crystalline Solids 18 (1975) 307. [32] G.E. JeUison and P.J. Bray, ibid. 29 (1978) 187. [33] G.E. JeUison, S.A. Feller and P.J. Bray, Phys. Chem. Glasses 19 (1978) 52. [34] P. Tarte and M.J. Pottier, The structure of Non-Crystalline Materials, ed., P.H. Gaskell (1978) p. 227. [35] P.J. Bray, M. Leventhal and H.O. Hooper, Phys. Chem. Glasses 4 (1963) 47. [36] M. Leventhal and P.J. Bray, ibid. 6 (1963) 113. [37] K. Fajans and N.J. Kreidl, J. Am. Ceram. Soc. 31 (1948) 105. [38] J.E. Stanworth, J. Soc. Glass Technol. 32 (1948) 154. [39] J.L. Trap and J.M. Stevels, Phys. Chem. Glasses 1 (1960) 107. [40] J.L. Trap and J.M. Stevels, ibid. 1 (1960) 181. [41] P.J. Bray and J.G. O'Keefe, ibid. 4 (1963) 37.