174
NOTES
CO concentration a straight line is obtained having exactly the slope predicted by the Nernst equation for the cell involved. Thus it is seen that the cell containing the glass electrode is behaving in the proper fashion once the effect of drift is removed. I n order to be a useful reference, however, the electrode must be stable. It was observed that the drift decreased slowly with time, suggesting thqt an “aging” process was occurring. I n order to study this effect the following experiment was set up. Three bulbs, selected at random, were prepared as above and placed in the molten solvent. One of these was arbitrarily selected as reference and the other two measured against it over a period of several days. Voltage readings were taken periodically over each eighthour working day. For the first two days the voltages changed a t rates greater than 50 mv./hr. On the third day one bulb showed a change of 2.5 and the other 6.5 mv./hr. On the fourth and fifth days the rates for each had fallen to less than 0.5 mv./hr. Beyond this time period the e.m.f. showed no further change. Thus five days aging yields a reference electrode that is stable for direct potentiometric measurements. Removal from the melt, washing and drying had no effect on this stability. Several electrodes were selected a t random and aged for one week in Li-K nitrate eutectic. Each then was placed in a cell of the type described earlier and the e.m.f. measured as a function of CoClz concentration. Each cell yielded a linear Nernst plot having exactly the predicted slope. E o drift was observed in any of these measurements. On the basis of the information presented above it is not feasible io attempt a complete explantion for the mechanism occurring a t this electrode. Certain observations may be made, however. The fact that the material inside the bulb made little difference suggests that this material serves only to establish electrical contact. The glass does not act as a membrane separating electrode compartments. Equilibration seems to follow an exponential rate suggesting zi diffusion -controlled process. Once equilibration is established, however, it is apparently quite stabie since random agitation (Le,, K2 bubbling) or complete removal appears to have no effect. Additional investigations are being conducted to answer these questions more fully. It should be pointed out, however, that this investigation, while establishing a useful reference electrode, also has shown the ideal behavior of Co(I1) in the solvents studied. The maximum concentration used was 0.1 mole % and to this concentration the solutions behaved ideally. The presence of chloride ion added as CoClz apparently has no effect at these concentrations. H. T. Tien wishes to express his thanks to Dr. D. 0. Rudkin, Eastern Psychiatric Institute, for his kind permission to use some of the facilities of the Institute.
Vol. 66 NUCLEAR MAGNETIC RESONANCE STUDY OF BORON COoRDIh’ATION I N POTASSIUM BORATE GLASSES’ BY %-E. SVAITSON, E. FORSLIND AND
Research Group for N M R , Divzsion of Physzcal Chemistry,, The Raga1 Institute of Technology, Stocloholm, Sweden
J. KROGH-MOE Suedash Institute of Siltcute Research. Gothenburg, Sweden Received June 6, 1961
The coordination of boron in the alkali borate glasses has been studied earlier with n.ni.r. by Silver and bra^.^,^ The B1l-resonance indicates that the boron nuclei in these glasses have two different local environments. Due to an anisotropic distribution of electronic charge around the boron nucleus, the resonance line in vitreous boron oxide shows a second-order broadening effect of nuclear quadrupole interaction, and this broad line is assigned to boron in threefold oxygen coordination. By addition of alkali oxide a sharp resonance line is produced which is attributed to the appearance of boron in fourfold coordination belonging to BO4 tetrahedra with low quadrupole interaction. It would in principle be possible to calculate the amounts of threefold and fourfold coordinated boron in the sample after separation of the overlapping resonance lines and evaluation of the areas under each line. The results of Silver and Bray, however, were evaluated in a very simplified manner from the observed signal amplitudes. As pointed out by the authors themselves their method is likely to overestimate the fraction of fourfold coordinated boron at high alkali content, since the width of the broad line increases with increasing alkali content and, indeed, their measurements (filled circles in Fig. 1) lead to a fraction of fourfold coordinated boron definitely higher than that assumed by Warren4 and the one observed in some crystalline borates investigated by Krogh-R/Ioe.6t6 Warren assumed that boron changes from threefold to fourfold coordination as alkali oxide is added to the glass, supposing each oxygen bonded to two b o r a s at low alkali content, all oxygens being engaged in boron-oxygen bonds. Krogh-Moe6 has shown that the boron-oxygen networks of crystalline potassium pentaborate,6 cesium triborate6 and lithium diborate’ are built in agreement with the following rule regarding the coordination of boron and oxygen: each “molecule” of alkali oxide added to boron oxide converts two boron atoms from threefold to fourfoId coordination. This corresponds to a fraction of boron in fourfold coordination Nd = X / ( l - X) (shown by the solid line in Fig. l), where S is t8hemolar alkali concentration. The above rule and expression for N4 follows from (1) The Swedish Xatuial Science Research Council and the S t a t e Council of Technical Research have provided financial support and the n.m.r. apparatus cost has been defrayed by grants from the Knut and Alice Wallenberg Foundation. (2) A. H. Silver and P. J. Bray, J . Chem. Phys., 29, 984 (1968). (3) J. D. Mackennie, “Modern Aspects of the Vitreous S t a t e ” Vol. 1, Butterworths, London, 1960. (4) E. E. Warren, J Am. Ceram. Soc., 24, 256 (1841). ( 5 ) J . Krogh-Moe, Arkw Kemi, 14, 489 (1959). ( 6 ) J. Krogh-Moe, Acta Cwst.. 18, 889 (1960). (7) J. Kiogh-Moe, to be published.
Jan., 1962 the Warren scheme for the region of low alkali oxide content (< 15 mole %). At higher alkali contcnt, however, Warren assumes that some oxygens are bonded only to one boron, in which case the frartion of fourfold coordinated boron should be lcss than that given by the ratio S i (1- S). The dashed line in Fig. 1 indicates the values expected by Warren. I n order to investigate further the discrepancy between experimental and theoretical data we have repeated aiid extended some of Silver and Bray’s measurements. During this work we made the observation that there is a close agreement between the spectra of crystalline boron oxide and vitreous boron oxide, from which we infer that the similarity of the spectra is based on a corresponding similarity of the two states with regard tx, the boron coordination. These findings, on the other hand, suggest the existence of corresponding relationships between other crystalline and vitreous boron compounds. Sincc the crystalline compounds so far investigated all lie on the theoretical N4(S) curvc, the corresponding vitreous states are expected to yield Ne-values in the close vicinity of the theoretical line, an assumption which actually is borne out by experiment. I n particular, we may refer to a study of potassium borate glass a t room temperature. The samples were prepared by melting boric acid and potassium carbonate in a platinum crucible. After solidification, the samples were coated with a plastic to be protected against moisture. The cylinders prepared were 16 mm. in diameter and of 70 mm. length. The compositions were determined by analyses within 0.1%. Spectra were recorded with a Varian-4250 wide line spectrometer a t 13.0 Mc./ see. and three different settings of the radiofrequency field. Our recorded spectra were in agreement with those reported by Silver and Bray.2 A rigorous calculation of the fraction of fourfold coordinated boron based on a detailed analysis of the two lines would, however, require considerable work, while still yielding values of limited accuracy because of the poor signal-to-noise ratio of the recorded broad line spectrum. As the signal-tonoise ratio of the sharp line much exceeds that, of the broad one, we have therefore, to a first approximation, chosen to calculate only the area under the sharp absorption line as a measure of the changes in the amount of four-co6rdinated boron. The area is estimated to be of the magnitude y‘,.2’, where ylrll is the ordinate and zm the abscissa of the milximuni of the experimental sharp line derivative, taking the center of the line as origin. For absorp tiori curves of Gaussian or Lorentz line shapes, the area is proportional to this quantity. As y,’ and c,. refer to the criitral part of the sharp line, thp influence on these quantities from the superposition of the broad line could be neglected. The calculated absorption line area is taken to be proportional to the fraction of fourfold coordinated boron and l o the volume density of boron atoms in the sample. In this way it was possible to calculate the ratio between Nrvalues for samples of different, K20contcnt. The method, of course, does not allow an absolutc deterniination of N 4 . Our values, shown
NOTES
*
175
............ o
-e---
SILVER 8 BRAY PRESENT AUTHORS WARREN
...r 0
30 S % 0 10 20 Fig. I.-The fraction of fourfold cotirdinated boron atoms, Ai,, aa a function of the molar KzO coricentrstion S. The solid line represents the theoretical curve N , = S/(1 - S ) .
in Fig. 1 by open circles, therefore have been fitted to the theoretical line N4 = S/ (1 - 5)by the method of least squares, yielding a standard deviation from the theoretical value of 3y0. A significance test of the deviation indicates that the boron-oxygcn coordination in the potassium borate glass is in agreement with the Si(1 S)-rule, in cont,radict,ionto the results obtained by Silver and bra^.^.^
-
T H E KINETICS OF T H E HYDROLYSIS OF POLYETHYLENE TEREPHTHALATE FILM BY T. DAVIES,P. L. GOLDSMITH, D. A. S. RAVENSA N D I. M. WARD Reaaarch Deparfment, Imperial Chemical Industries Limited, Fibres Didnun, Hookstone Road, Harrogate, Yorkshire, England Received June 2031981
In a recent papcr,l Golike and Lasoski have reported results for the hydrolysis of polyethylene terephthalatc film. The problem of polyethylene terephthalate degradation has received considerablo attention in these Laboratories and a recent account As our conclusions of this work has been differ materially from those drawn by Golike and Lasoski, we feel that it, is worth considering the cxperimental evidence of both investigations. We shall show that although the experimental results of Golike and Lasoski are in good agreement with our own, a very different interpretation of these results can be obtained. The principal conclusion of Golike and Lasoski is that the reaction rate is controlled by diffusion of water into the sample. We (1) R. C.Golike and S. W . Lasoeki, J . Phya. Chem., 64, 895 (1960). (2) D. A. S. Ravens stid I. M. Ward, Trans. Furaday Soc., 67, 150 (1961).
