We report the bulk magnetic properties of a yet unexplored vanadium-based multivalued spinel system, Zn3V3O8. A Curie-Weiss fit of our dc magnetic susceptibility data in the temperature region from 140 to 300 K yields a Curie constant C = 0.75cm3K/mo
le V, theta CW = -370 K. We have observed a splitting between the zero field cooled ZFC and field cooled FC susceptibility curves below a temperature Tirr of about 6.3 K. The value of the frustration parameter nearly equals to 100 suggests that the system is strongly frustrated. From the ac susceptibility measurements we find a logarithmic variation of freezing temperature (Tf ) with frequency attesting to the formation of a spin glass below Tf . However, the value of the characteristic frequency obtained from the Vogel-Fulcher fit suggests that the ground state is closer to a cluster glass rather than a conventional spin glass. We explored further consequences of the spin glass behavior and observed aging phenomena and memory effect (both in ZFC and FC). We found that a positive temperature cycle erases the memory, as predicted by the hierarchical model. From the heat capacity CP data, a hump-like anomaly was observed in CP /T at about 3.75 K. Below this temperature the magnetic heat capacity shows a nearly linear dependence with T which is consistent with the formation of a spin glass state below Tf in Zn3V3O8.
We report the synthesis and characterization of Li2ZnV3O8, which is a new Zn-doped LiV2O4 system containing only tetravalent vanadium. A Curie-Weiss susceptibility with a Curie-Weiss temperature of <theta>CW ~214 K suggests the presence of strong ant
iferromagnetic correlations in this system. We have observed a splitting between the zero-field cooled ZFC and field cooled FC susceptibility curves below 6 K. A peak is present in the ZFC curve around 3.5 K suggestive of spin-freezing . Similarly, a broad hump is also seen in the inferred magnetic heat capacity around 9 K. The consequent entropy change is only about 8% of the value expected for an ordered S = 1=2 system. This reduction indicates continued presence of large disorder in the system in spite of the large <theta>CW, which might result from strong geometric frustration in the system. We did not find any temperature T dependence in our 7Li nuclear magnetic resonance NMR shift down to 6 K (an abrupt change in the shift takes place below 6 K) though considerable T-dependence has been found in literature for LiV2O4- undoped or with other Zn/Ti contents. Consistent with the above observation, the 7Li nuclear spin-lattice relaxation rate 1/T1 is relatively small and nearly T-independent except a small increase close to the freezing temperature, once again, small compared to undoped or 10% Zn or 20% Ti-doped LiV2O4.
BaV3O8 contains both magnetic V4+(S=1/2) ions and non-magnetic V5+(S=0) ions. The V4+ ions are arranged in a coupled Majumdar-Ghosh chain like network. Our magnetic susceptibility chi(T) data fit well with the Curie-Weiss formula in the temperature r
ange of 80-300K and it yields a Curie constant C=0.39cm3K/mole-V4+ and an antiferromagnetic Weiss temperature theta=-26K. The chi(T) curve shows a broad maximum at T~25K indicative of short-range order (SRO) and an anomaly corresponding to long-range order (LRO) at TN~6K. The value of the frustration index (f=mod[theta/TN]~5) suggests that the system is moderately frustrated. Above the LRO temperature the experimental magnetic susceptibility data match well with the coupled Majumdar-Ghosh chain model with the ratio of the nnn (next-nearest neighbor) to nn (nearest neighbor) magnetic coupling alpha=2 and Jnnn/kB=40K. In a mean-field approach when considering the inter-chain interactions, we obtain the total inter-chain coupling to be about 16K. The LRO anomaly at TN is also observe in the specific heat Cp(T) data and is not sensitive to an applied magnetic field up to 90kOe. A 51V NMR signal corresponding to the non-magnetic vanadium was observed. Anomalies at 6K were observed in the variation with temperature of the 51V NMR linewidth and in the spin-lattice relaxation rate 1/T1, indicating that they are sensitive to the LRO onset and fluctuations at the magnetic V sites. The existence of two components (one short and another long) is observed in the spin-spin relaxation rate 1/T2 data in the vicinity of TN. The shorter component seems to be intimately connected with the magnetically ordered state. We suggest that both magnetically ordered and non-long range ordered (non-LRO) regions coexist in this compound below the long range ordering temperature.
