Acta Physica Academiae Scientiarum Hungaricae 2. (1952)

1952 / 1. szám - Brief Reports - E. Nagy: Temperature-Dependence of Willemite Luminescence

TEMPERATURE-DEPENDENCE OF WILLEMITE LUMINESCENCE (RECEIVED : 15. XII. 1951) In previous communications it was reported [1 ], [2 ], [3 ] that the tempera­ture dependence of willemite luminescence may be de:cribed by a formula: (1)V I + Ce lt was also established that by varying the activator concentration log C and E change in a similar manner. The present paper deals with a generalization of the above assertion. 24 willemites, containing various amounts of manganese activator (0,001, 0,01, 0,1, 0,7, 2 and 5%) and iron killer (0,001, 0,01, 0,1, 1%) were investigated. It was found that the temperature dependence of the lumines­cence for all the specimens could be described by a modification of formula (1), namely : The values of the exponentials in the denominator group themselves into two distinct classes. For the first exponential, C1 is appr. 102—TO3 and El 0,2 — 0,4 e. v., for the second, C2 appr. 109, E2 1 — 1,4 e. v. This means, that the second exponential cannot be observed below appr. 600° K, and therefore, in materials containing substantial amounts of manganese or iron, it must have escaped attention. The effect of iron can be studied thus only on the first exponen­tial with the result that increasing amounts of iron monotonously lower both Cj and Ег. The same holds for manganese above the optimum concentration. The constants of the second exponential, C2 and E„, may be determined, but less accurately, though the magnitude mentioned is undoubtedly correct. There seems to exist a definite correlation between the corresponding activation energies (E) and constants C. Plotting log C as a function of E, (log Cj vs. Fj and log C2 vs. E2), we get quite an accurate straight line (Fig. 1.). This relationship can therefore be expressed thus : 10log C — a -f- bE. The existence of this relationship is not destroved by possible errors in the determination of the corresponding C and E, for the eventual variations in C and E merely cause a shift of the locus almost exactly parallel to the above line. E ti­ A V = (2) 1 + Cje Jh kT C„e 3 kT

Next