Magnetic-field effects in non-magnetic glasses

Abstract

Recently, it was found that the multi-component glass a-BaO-Al2O3-SiO2 exhibits unusual magnetic properties at very low temperatures. Thus the question arises whether this is a specialty of that particular glass or a more general phenomenon. We report here on our studies of the magnetic-field dependence of the dielectric properties of the borosilicate glass BK7 which contains only a negligible amount of magnetic impurities. Since this glass also responds sensitively to magnetic fields, our investigations demonstrate that the reaction of glasses to magnetic fields is not caused by magnetic impurities but reflects a more general phenomenon. In addition, we have observed that the variation of the dielectric constant and the loss angle with magnetic field depend on the amplitude of the electric field that is used to measure the glass capacitance. We present the data and discuss possible origins of the magnetic-field phenomena in non-magnetic glasses.

Multiple-well tunneling model for the magnetic-field effect in ultracold glasses

“…A best fit to the 118 mK experimental data yields the parameters reported in Table II. The T -dependence of the calculated magnetocapacitance is also in qualitative agreement with the experimental data [3,4]. In a comparison with data for the BK7 and Duran glasses [4] (at the much lower only available temperature of 15.4 mK, however) the present theory reproduces the main features correctly.…”

“…(7) gives rise to a peak in δǫ/ǫ at very low B, while the first term (present only if n w > 2) gives rise to a negative contribution to δǫ/ǫ at larger B which can win over the enhancement term for all values of B if D 0max ≫ D 0min (D 0min , D 0max corresponding to cutoffs in the distribution of ATS energy barriers U B ). The observations in Duran and BK7 indeed show a significant depression of ǫ(B) for weak fields [4], thus giving direct evidence for the existence of ATS with n w > 2 in the multisilicate glasses. Carrying out the averaging [.…”

“…A typical glass showing a strong response has some 100 ppm Fe 3+ and the composition Al 2 O 3 -BaO-SiO 2 (in short AlBaSi-O, in this paper). Measurements made on a thick sol-gel fabricated AlBaSi-O film showed changes in δǫ/ǫ = [ǫ(B) − ǫ(0)]/ǫ(0) of order 10 −4 and characterised by an enhancement peaking around 0.03 T for 10 mK < T < 200 mK then followed by a reduction of ǫ for B > 0.1 T. A further enhancement was also observed at much higher fields (B > 10 T).Another, cleaner, multi-component silicate glass (borosilicate BK7, with 6 ppm Fe 3+ ) and a dirtier one (borosilicate Duran, with 120 ppm Fe 3+ ) have shown similar -but weaker -magnetic anomalies, |δǫ/ǫ| ∼ 10 −5 , seemingly excluding the paramagnetic impurities as their source [4]. Yet another multi-silicate glass, aSiO 2+x C y H z was investigated [5], confirming the unusual findings in AlBaSi-O.…”

“…AlBaSi-O [11], [3], [7] YES YES ? YES Duran [11], [4], [7] YES weak ? YES BK7 [11], [4], [13], [7] weak weak NO YES AF45 [13] ?…”

“…Fig. 3 shows the change due to a magnetic field in the dielectric constant ǫ(T, B) = ǫ 0 + α(T, B) for the AlBaSi-O glass within the present oversimplified theory, which is used to fit the data at 118 mK [4]. It is seen that using the parameters extracted from fitting this theory [12] to the C p (T, B) data [11] and a value A = 20.0 × 10 −5 for the dimensionless prefactor A ≡ p 2 i P * /(ǫ 0 ǫ r D min ) one can reproduce all the qualitative features of the puzzling B-dependence of the magnetocapacitance for this glass.…”