Mn-55 NMR and NQR study of the cubic Laves-phase compound UMn2

TLDR: In this paper, NMR and NQR signals of two different crystallographic Mn sites, which are in nearly axial and nonaxial electric field gradients, were found below structural transitions, being consistent with the crystal structure determined from diffraction analysis and confirming the absence of magnetic ordering.

Abstract: We performed ${}^{55}\mathrm{Mn}$ NMR and nuclear quadrupole resonance (NQR) measurements of a cubic Laves-phase compound ${\mathrm{UMn}}_{2},$ which shows structural instability in the temperature range 210--240 K. NMR and NQR signals of two different crystallographic Mn sites, which are in nearly axial and nonaxial electric field gradients, were found below structural transitions, being consistent with the crystal structure determined from diffraction analysis and confirming the absence of magnetic ordering. Analysis of the Knight shift shows that the spin component of the isotropic hyperfine coupling constant does not change at the phase transitions, and the anisotropy of the Knight shift appreciably depends on temperature. The nuclear spin-lattice relaxation is not enhanced except around the structural transitions and is explained as simple conduction-electron relaxation, indicating absence of localized moments at both U and Mn sites. The relaxation behavior also suggests neither critical behavior nor strong electron correlation in the low-temperature range.

Figures

FIG. 5. ~a! Examples of field-swept NMR center-line spectra measured at 76.3 MHz. The bold solid and broken curves represent calculated powder patterns for theaxial andnonaxialsites, respectively. Fine solid curves indicate total spectra after the convolution of Gaussian functions with appropriate widths. Figure~b! shows evolutions of center-line spectra around the structural transitions.

FIG. 6. Temperature dependences of isotropic (K iso , circles! and anisotropic (K1, triangles! components of the Knight shift for the axial ~filled marks! andnonaxial ~open marks! sites. The inset shows the temperature dependence of the susceptibility.

FIG. 9. The inset shows the temperature variation of the spinlattice relaxation rate, 1/T1, estimated from NQR for theaxial ~closed circles! and nonaxial ~open circles! sites, which show the T1T5const behavior~straight lines!. Open squares are averaged values of both the sites, and close squares represent the values estimated by NMR experiments.

FIG. 4. ~a! Examples of field-swept NMR full spectra measured at 76.3 MHz.~b! Calculated powder patterns with quadrupolar parameters estimated from NQR experiments at 4.2 K and appropriates values of the Knight shift.

Frequently asked questions

Q1. what is the gyromagnetic constant of the cubic phase?

The orthorhombic lattice constants, ao , bo and co are equal to ac /A2, ac /A2 and ac , respectively, in the limit of no distortion, where ac means the lattice constant of the cubic phase. 

Q2. What is the symmetry of the atoms in the orthorhombic phase?

In the orthorhombic phase, U atoms occupy the 4e site with mm2 local symmetry and the manganese site separates to two different sites, i.e., 4b and 4d sites. 

Q3. What is the effect of the temperature on the d-spin component?

If the temperature dependence is caused by the Mn 3d-spin component only, as usually in 3d magnets, the positive K iso at low temperatures suggests a comparable or larger positive offset of Van Vleck–type or-bital and/or conduction-electron contributions because the on-site d-spin hyperfine coupling ~via core polarization! is always negative. 

Q4. What is the r(EF) of the UMn2 lattice?

the band Jahn-Teller distortion driven mainly by Mn-3d bands may be one of possible origins of the lattice instability since the authors may expect orbital degeneracies in the axially symmetric MnPRB 61 12 23955Mn NMR AND NQR STUDY OF THE CUBIC LAVES- . . .site in the high-temperature cubic phase. 

Q5. What is the optimum temperature for UMn2?

The observation of 55Mn NQR signals at low temperatures confirm microscopically that there is no magnetic ordering in UMn2 at least down to 1.4 K, being consistent with neutron,11 Mössbauer,19 and m-SR ~Ref. 15! results. 

Q6. What is the resonance field of the constant-frequency experiment?

The resonance field @Hm(u ,f)# of the constant-frequency experiment is obtained by solving Eq. ~3.3! for Hm(u ,f) 5 n0 /g where nm is the operating frequency. 

Q7. What is the reason for the structural instability in UMn2?

For these materials, recently, Chu et al.29 proposed that the structural instability is due to the large density of states at the Fermi level, r(EF), and Fermi surface nesting, which give rise to phonon softening. 

Q8. What is the effect of the relaxation on the UMn2?

the relaxation is considered to be dominated by the conduction electron contribution ~the Korringa mechanism!, indicating clearly the absence of localized moments at both Mn and U sites and also the absence of the enhanced density of states at the Fermi level. 

Q9. How did the authors measure the LS coupling?

The authors measured the 55Mn nuclear spin-lattice relaxation time, T1 by using NQR and NMR signals at low- and high-temperature ranges, respectively. 

Q10. What is the simplest way to estimate the T1 value?

Although the distribution of chemical sites may also cause the nonsingle exponential behavior, the authors try to estimate rather forcibly a unique value of T1 by attributing the behavior to the quadrupole interaction. 

Q11. What are the axial and nonaxial sites?

the authors will call them axial and nonaxial sites, which are considered to correspond to the 4b and 4d Mn sites, respectively, in the orthorhombic phase taking account of the symmetry consideration given in last section. 

Q12. What is the r(EF) of the Mn site?

If only U spin fluctuations are responsible to the large g value, and if the Mn site has no substantial coupling with the U magnetism, this contradiction may be reconciled. 

Q13. Why do the authors not discuss the low temperature anomaly?

since it is known that the low-temperature susceptibility strongly depends on samples, the authors do not discuss further the low-temperature anomaly. 

Q14. What is the average value of 1/T1 for the two sites?

The estimated values of 1/T1 vary linearly with temperature till 100 K which are plotted in the inset of Fig. 9. Average values of 1/T1 for the two sites are shown by open circles in Fig. 9.Above ;100 K, it becomes difficult to estimate reliable T1 values from NQR experiments due to poor signal-to-noise (S/N) ratio.