betaFeedback

WorksPhysics and Astronomy

Deformed spin-12\frac12 square lattice in antiferromagnetic NaZnVOPO4_4(HPO4_4)

Abstract

We report the structural and magnetic properties of a new spin-12\frac12 antiferromagnet NaZnVOPO4_4(HPO4_4) studied via x-ray diffraction, magnetic susceptibility, high-field magnetization, specific heat, and 31^{31}P nuclear magnetic resonance (NMR) measurements, as well as density-functional band-structure calculations. While thermodynamic properties of this compound are well described by the J1−J2J_1-J_2 square-lattice model, ab initio calculations suggest a significant deformation of the spin lattice. From fits to the magnetic susceptibility we determine the averaged nearest-neighbor and second-neighbor exchange couplings of Jˉ1≃−1.3\bar J_1\simeq -1.3 K and Jˉ2≃5.6\bar J_2\simeq 5.6 K, respectively. Experimental saturation field of 15.3,T is consistent with these estimates if 20% spatial anisotropy in J1J_1 is taken into account. Specific heat data signal the onset of a magnetic long-range order at TN≃2.1T_{\rm N} \simeq 2.1 K, which is further supported by a sharp peak in the NMR spin-lattice relaxation rate. The NMR spectra mark the superposition of two P lines due to two noneqivalent P sites where the broad line with the strong hyperfine coupling and short T1T_1 is identified as the P(1) site located within the magnetic planes, while the narrow line with the weak hyperfine coupling and long T1T_1 is designated as the P(2) site located between the planes.

Introduction

Low-dimensional spin systems augmented with strong frustration reveal suppression of a conventional magnetic order and may lead to a quantum disordered ground state like quantum spin liquid (QSL) [1]. Two-dimensional (2D) spin-1/21/2 frustrated square-lattice (FSL) is a well-known example where frustration appears because of the competition between nearest-neighbor (NN, J1J_1) and next-nearest-neighbor (NNN, J2J_2) exchange interactions along the edges and diagonals of a square, respectively (J1−J2J_1-J_2 model). Theoretical studies have determined a global phase diagram with different ground states depending on the sign and relative strength of the exchange couplings (α=J2J1\alpha= \frac{J_2}{J_1}) [2, 3]. Ferromagnetic (FM), Néel antiferromagnetic (NAF), and degenerate collinear antiferromagnetic (CAF) ordered phases are stabilized and extended in the regimes of −0.5<α≤−∞-0.5 < \alpha\leq-\infty, −∞≤α<0.5-\infty\leq\alpha< 0.5, and 0.5<α<−0.50.5 < \alpha< -0.5, respectively. Different quantum disordered phases with novel order parameters are predicted to emerge at the phase boundaries or the quantum critical regimes. For instance, QSL [4, 5, 6], plaquette valence-bond solid (PVBS) [7, 8] or columnar valence-bond solid (CVBS) [9] states are expected around α≃0.5\alpha\simeq0.5 while a spin nematic phase is predicted for α≃−0.5\alpha\simeq-0.5 in the phase diagram [10].

In the recent past, a handful number of Cu2+^{2+} (3d93d^9), V4+^{4+} (3d13d^1), and Mo5+^{5+} (4d14d^1) based spin-1/21/2 FSL magnets have been studied as J1−J2J_1-J_2 model compounds.

The vanadates are the most celebrated ones and include Li2_2VOXO4_4 (XX = Si, Ge) with J1,J2>0J_1,J_2 > 0 and J2≫J1J_2 \gg J_1 [11, 12], along with V4+^{4+} phosphates that typically show J1<0J_1 < 0 and J2>0J_2 > 0, also resulting in the CAF ground state [13, 14, 15, 16, 17, 18]. A few other vanadates, such as Zn2_2VO(PO4_4)2_2 [19], AlVO4_4 [20, 21, 22], and PbVO3_3 [23], feature J1>J2J_1 > J_2 and lie in the NAF region of the phase diagram. A further variability becomes possible with Cu2+^{2+} as well as 4d4d Mo5+^{5+} compounds that span both J2≫J1J_2 \gg J_1 and J2≪J1J_2 \ll J_1 limits [24, 25, 26, 27, 28, 29, 30, 31, 32]. Unfortunately, none of these compounds fall in the quantum critical regimes of the phase diagram around α=12\alpha= \frac{1}{2}. Moreover, some of them show intricate deformations of the magnetic square lattice, because the underlying crystal symmetry is lower than tetragonal [33]. This deformation was verified experimentally in several V4+^{4+} phosphates, including Pb2_2VO(PO4_4)2_2 [34], SrZnVO(PO4_4)2_2 [35], and BaCdVO(PO4_4)2_2 [36]. Each of them showed an interesting pre-saturation phase [35, 37, 38, 39, 40], which is reminiscent of the nematic phase of the J1−J2J_1-J_2 model [10], but deviations from the ideal square lattice put into question the applicability of this theoretical scenario, and alternative interpretations were indeed proposed recently [41, 42].

In this context, finding V4+^{4+} square-lattice compounds with different magnitudes of the deformation is important. Here, we report low-temperature magnetic behavior of the hitherto unexplored NaZnVO4_4(HPO4_4) as a spin-12\frac{1}{2} square-lattice candidate with a different spacer separating the magnetic layers and, potentially, a different regime of exchange couplings compared to the widely studied AA′AA^\primeVO(PO4_4)2_2 phosphates (AA′AA^\prime = Pb2_2, SrZn, BaCd). The monoclinic crystal structure of NaZnVO4_4(HPO4_4) features V4+^{4+} ions in the square-pyramidal coordination. They are joined into layers via P(1)O4_4 tetrahedra (Figure 1), while the interlayer space is filled with the Na+^{+} and Zn2+^{2+} ions, as well as the P(2)O4_4 tetrahedra that represent the HPO4_4 groups.

Figure 1. (a) Crystal structure of NaZnVOPO4_4(HPO4_4) with the hydrogen position and hydrogen bonds (dashed lines) determined in this work. (b) Magnetic [VOPO4_4] layer with the square-lattice-like arrangement of the V4+^{4+} ions. (c) Deformed square lattice with 5 nonequivalent interactions. VESTA software [43] was used for crystal structure visualization.

Methods

Pale-blue coloured powder of the titled compound was synthesized by the conventional hydrothermal method. 0.150 g Na2_2CO3_3 (Aldrich, 99.995%), 0.154 g V2_2O5_5 (Aldrich, 99.99%), and 0.115 g Zn (Aldrich, 99%) powders were mixed with 5 ml of a 1.5 M aqueous solution of H3_3PO4_4, sealed in a 23 ml Teflon lined bomb, and heated at 240°C for 8 days followed by slow cooling (10°C/hour) to room temperature. The obtained blue color product was washed carefully with distilled water and dried in an oven maintained at 100°C for 24 hours. To check the phase purity of the compound, powder x-ray diffraction (XRD) experiment was performed using a PANalytical powder diffractometer with Cu KαK_{\alpha} radiation (λavg≃1.5418\lambda_{\mathrm{avg}} \simeq1.5418 Å) at room temperature. We have also performed temperature-dependent powder XRD measurements on the pure phase powder sample in the temperature range 15 K ≤T≤\leq T \leq 300 K, using a low-temperature attachment (Oxford PheniX) to the x-ray diffractometer. Le-Bail fit of the powder XRD patterns was performed using the FULLPROF software package [44], taking the initial structural parameters from the previous report [45].

Temperature variation of magnetization (MM) was measured in the temperature range 1.8–350 K in different magnetic fields (HH) using a SQUID magnetometer (MPMS3, Quantum Design). The isothermal magnetization was measured at T=1.8T = 1.8 K from 0 to 7 T. The high-field magnetization measurement was performed at T=1.4T = 1.4 K in pulsed magnetic fields up to 30 T using the facility at the Dresden High Magnetic Field Laboratory [46, 47]. The temperature dependent specific heat of this sample was measured on a sintered pellet in a large temperature range (0.5 K ≤T≤\leq T \leq 200 K) using the Physical Property Measurement System (PPMS, Quantum Design) and adopting the thermal relaxation technique. For measurements below 2 K, 3^3He attachment to the PPMS was used.

The Nuclear Magnetic Resonance (NMR) measurements were carried out on the 31^{31}P nuclei (gyromagnetic ratio γN2π=17.237\frac{\gamma_N}{2\pi}=17.237 MHz/T and nuclear spin I=12I=\frac{1}{2}) in the temperature range 1.8 K ≤T≤\leq T \leq 250 K. The NMR spectra at different temperatures were obtained by varying the magnetic field at a constant frequency of 121 MHz. The spin-lattice relaxation rate (1/T11/T_1) was measured by the single saturation pulse method at two frequencies (30.2 MHz and 121 MHz). The temperature-dependent NMR shift, K(T)=[Href/H(T)−1]K(T)=[H_{\mathrm{ref}}/H(T)-1], was calculated from the resonance field of the sample HH with respect to the resonance field of a non-magnetic reference sample (HrefH_{\mathrm{ref}}).

Magnetic couplings were determined by density-functional-theory (DFT) band-structure calculations performed in the FPLO code [48] using experimental structural parameters from Ref. [45], except the hydrogen position that was optimized as further explained in Sec. III E. The Perdew-Burke-Ernzerhof (PBE) approximation for the exchange-correlation potential was employed [49]. We used superexchange theory in the vein of Kugel-Khomskii model [50, 51], as well as the mapping approach [52, 53] based on total energies of collinear spin configurations obtained from DFT+UU calculations with the on-site Coulomb repulsion Ud=4U_d = 4 eV, Hund’s coupling Jd=1J_d = 1 eV, and double-counting correction in the atomic limit [54].

Experimental thermodynamic properties were modeled by high-temperature series expansion (HTSE) for the J1−J2J_1-J_2 spin-12\frac{1}{2} square lattice [55] as well as full diagonalization (FD) for the 4×44 \times4 finite lattice with periodic boundary conditions. Additionally, quantum Monte Carlo (QMC) simulations were performed for the nonfrustrated spin-12\frac{1}{2} square lattice using directed loop algorithm [56] in the stochastic series expansion (SSE) [57] representation. QMC simulations were performed for 16 × 16 finite lattices with periodic boundary conditions, using 4×1044 \times10^{4} sweeps and 4×1034 \times10^{3} thermalization sweeps. The ALPS package [58] was used for both FD and QMC.

Results

X-ray Diffraction

The powder XRD patterns of NaZnVOPO4_4(HPO4_4) are analyzed by Le Bail fits. Figure 2 presents the data at two end temperatures (T=300T = 300 K and 15 K). The entire XRD pattern down to 15 K could be indexed using the monoclinic crystal structure with the space group P21/cP2_{1}/c. Neither any indication of structural transition nor lattice distortion is found down to 15 K. The refined lattice parameters and unit cell volume (VcellV_{\mathrm{cell}}) are [a=8.5418(4)a = 8.5418(4) Å, b=8.9937(5)b = 8.9937(5) Å, c=9.0765(5)c = 9.0765(5) Å, β=91.186(3)∘\beta= 91.186(3)^\circ, and Vcell≃696.46V_{\mathrm{cell}} \simeq696.46 Å3^{3}] and [a=8.5190(2)a = 8.5190(2) Å, b=8.9281(4)b = 8.9281(4) Å, c=9.1029(4)c = 9.1029(4) Å, β=92.013(5)∘\beta= 92.013(5)^\circ, and Vcell≃691.73V_{\mathrm{cell}} \simeq691.73 Å3^{3}] for T=300T = 300 K and 15 K, respectively. The obtained lat-

Powder XRD patterns at 300 K and 15 K

Figure 2. Powder XRD patterns (open circles) at (a) T=300T = 300 K and (b) T=15T = 15 K. The solid line is the Le-Bail fit, the vertical bars mark the expected Bragg peak positions, and the lower solid line corresponds to the difference between the observed and calculated intensities. The goodness-of-fit is achieved to be χ2∼5.4\chi^{2} \sim5.4 and ∼6.2\sim6.2 for T=300T = 300 K and 15 K, respectively.

tice parameters at room temperature are in close agreement with the values reported earlier [45]. The temperature dependence of the lattice parameters (aa, bb, cc, and β\beta) and VcellV_{\mathrm{cell}} are presented in Figure 3. The lattice constants aa and bb are found to decrease in a systematic way with decreasing temperature while cc and monoclinic angle β\beta increase with decreasing temperature and then reach a plateau. These lead to an overall thermal contraction of the VcellV_{\mathrm{cell}} with temperature.

Figure 3. (a) Lattice constants (aa, bb, and cc) as a function of temperature. (b) Monoclinic angle (β\beta) along with the unit cell volume (VcellV_{\mathrm{cell}}) are plotted as a function of temperature from 15 K to 300 K. The solid line represents the fit of VcellV_{\mathrm{cell}} using Eq. (1).

The variation of VcellV_{\mathrm{cell}} with temperature can be expressed in terms of the internal energy [U(T)][U(T)] of the system [59, 60],

Vcell(T)=γU(T)K0+V0.(1)V_{\mathrm{cell}}(T) = \frac{\gamma U(T)}{K_{0}} + V_{0}. \tag*{(1)}

Here, V0V_{0} is the unit-cell volume at T=0T = 0 K, γ\gamma is the Grüneisen parameter, and K0K_{0} is the bulk modulus of the system. According to the Debye model, U(T)U(T) can be written as,

U(T)=9NkBT(TθD)3∫0θD/Tx3ex−1 dx,(2)U(T) = 9Nk_{\mathrm{B}}T\left(\frac{T}{\theta_{\mathrm{D}}}\right)^{3}\int_{0}^{\theta_{\mathrm{D}}/T}\frac{x^{3}}{e^{x}-1}\,dx, \tag*{(2)}

where NN is the total number of atoms per unit cell, kBk_{\mathrm{B}} is the Boltzmann constant, and θD\theta_{\mathrm{D}} is the Debye temperature [61]. The variable xx inside the integration stands for the quantity ℏωkBT\frac{\hbar\omega}{k_{\mathrm{B}}T} with phonon frequency ω\omega and Planck constant hh. Here, θD=hωDkB\theta_{\mathrm{D}} = \frac{h\omega_{\mathrm{D}}}{k_{\mathrm{B}}} and ωD\omega_{\mathrm{D}} is the upper limit of ω\omega. The best fit of the Vcell(T)V_{\mathrm{cell}}(T) data using Eq. (1) [solid line in Figure 3(b)] yields the parameters: θD≃312\theta_D \simeq312 K, V0≃691.72V_0 \simeq691.72 Å3^3, and γκ0≃8.84×10−12\frac{\gamma}{\kappa_0} \simeq8.84 \times10^{-12} Pa−1^{-1}.

Magnetization

The magnetic susceptibility [χ(T)≡M/H\chi(T) \equiv M/H] of NaZnVOPO4_4(HPO4_4) measured in an applied field of H=0.5H = 0.5 T is shown in Figure 4(a). In the high-temperature region, χ(T)\chi(T) follows a typical Curie-Weiss (CW) behaviour and shows a rounded maximum at Tχ≃4.6T_\chi\simeq4.6 K. Such a maximum represents the short-range AFM order in the low-dimensional spin systems. Below about 2.6 K, a small upturn is likely due to paramagnetic impurities and/or defects present in the powder sample. No trace of magnetic long-range-order (LRO) is detected down to 2 K. We have also measured χ(T)\chi(T) in different applied fields [inset of Figure 4(b)] but no obvious features associated with magnetic LRO is found, except the suppression of a broad maximum towards low temperatures.

Magnetic susceptibility and inverse susceptibility of NaZnVOPO4(HPO4)

Figure 4. (a) χ(T)\chi(T) measured in the magnetic field of H=0.5H = 0.5 T. The lines (solid and dashed) are the HTSE fits using the isotropic spin-12\frac{1}{2} FSL model, with two different solutions (solution-a and solution-b). Inset: The simulated χ(T)\chi(T) using the FD method taking the J1J_1 and J2J_2 values from solutions-a and b, respectively. (b) 1/χ1/\chi vs TT. The red solid line is the CW fit. Inset: χ(T)\chi(T) in the low-TT region measured in different fields.

The inverse susceptibility, 1/χ(T)1/\chi(T), is shown in Figure 4(b). In the paramagnetic regime, 1/χ(T)1/\chi(T) was fitted by the CW law

χ(T)=χ0+CT−θCW,(3)\chi(T) = \chi_0 + \frac{C}{T-\theta_{\mathrm{CW}}}, \tag*{(3)}

where the first term (χ0\chi_0) represents the combination of temperature-independent diamagnetic and Van-Vleck paramagnetic susceptibilities. The second term is the CW law where CC is the Curie constant and θCW\theta_{\mathrm{CW}} is the CW temperature. The fit above 25 K returns the parameters: χ0≃4.26×10−5\chi_0 \simeq4.26 \times10^{-5} cm3^3/mol-V4+^{4+}, C≃0.393C \simeq0.393 cm3^3K/mol-V4+^{4+}, and θCW≃−4.3\theta_{\mathrm{CW}} \simeq-4.3 K. Using the value of CC, the effective moment is calculated to be μeff=[(3kBC/NAμB2)]12\mu_{\mathrm{eff}} = [(3k_{\mathrm{B}}C/N_{\mathrm{A}}\mu_{\mathrm{B}}^2)]^{\frac{1}{2}}, where NAN_{\mathrm{A}} is the Avogadro’s number and μB\mu_{\mathrm{B}} is the Bohr magneton ≃1.77\simeq1.77 μB\mu_{\mathrm{B}}/V4+^{4+}. This value of μeff\mu_{\mathrm{eff}} is close to the actual value 1.73 μB\mu_{\mathrm{B}} for a spin-12\frac{1}{2} transition-metal ion with g=2g = 2. The negative value of θCW\theta_{\mathrm{CW}} indicates the dominant AFM exchange coupling between the V4+^{4+} ions. The core diamagnetic susceptibility (χcore\chi_{\mathrm{core}}) of the compound caused by the core orbital electrons was calculated to be −1.37×10−4-1.37 \times10^{-4} cm3^3/mol by adding the χcore\chi_{\mathrm{core}} of Na+^+, Zn2+^{2+}, V4+^{4+}, P5+^{5+}, and O2−^{2-} ions [62]. The Van-Vleck paramagnetic susceptibility (χvv\chi_{\mathrm{vv}}) was calculated to be ∼1.8×10−4\sim1.8 \times10^{-4} cm3^3/mol by subtracting χcore\chi_{\mathrm{core}} from χ0\chi_0, which is very close to the value reported for other V4+^{4+} based compounds [63].

As evident from the structural data (Figure 1), the system deviates from the isotropic square lattice. This spatial anisotropy is mainly due to five non-equivalent exchange couplings, three between nearest neighbors (J1J_1, J1′J_1^{\prime}, and J1′′J_1^{\prime\prime}) and two between next-nearest neighbors (J2J_2 and J2′J_2^{\prime}). Previous studies suggest that this anisotropy has only a minor effect on thermodynamic properties at higher temperatures [33]. Therefore, fits with the isotropic J1−J2J_1-J_2 model return averaged values of the NN and NNN couplings, Jˉ1=(2J1+J1′+J1′′)/4\bar{J}_1 = (2J_1 + J_1^{\prime} + J_1^{\prime\prime})/4 and Jˉ2=(J2+J2′)/2\bar{J}_2 = (J_2 + J_2^{\prime})/2, respectively. We fitted the data with

χ(T)=χ0+χspin(T)(4)\chi(T) = \chi_0 + \chi_{\mathrm{spin}}(T) \tag*{(4)}

using the temperature-independent term (χ0\chi_0) and the 9th order HTSE [χspin(T)\chi_{\mathrm{spin}}(T)],

χspin(T)=NAg2μB2kBT∑n(Jˉ1kBT)n∑mcmn(Jˉ2Jˉ1)m.(5)\chi_{\mathrm{spin}}(T) = \frac{N_{\mathrm{A}}g^2\mu_{\mathrm{B}}^2}{k_{\mathrm{B}}T} \sum_n \left(\frac{\bar{J}_1}{k_{\mathrm{B}}T}\right)^n \sum_m c_{mn}\left(\frac{\bar{J}_2}{\bar{J}_1}\right)^m. \tag*{(5)}

The values of the coefficients cmnc_{mn} are taken from Ref. [55]. Our fit for T≥7T \geq7 K yields two solutions: solution-a with χ0≃6×10−5\chi_0 \simeq6 \times10^{-5} cm3^3/mol-V4+^{4+}, g≃2.04g \simeq2.04, Jˉ1≃5.18\bar{J}_1 \simeq5.18 K, Jˉ2≃−0.66\bar{J}_2 \simeq-0.66 K and solution-b with χ0≃6.6×10−5\chi_0 \simeq6.6 \times10^{-5} cm3^3/mol-V4+^{4+}, g≃2.04g \simeq2.04, Jˉ1≃−1.28\bar{J}_1 \simeq-1.28 K, Jˉ2≃5.59\bar{J}_2 \simeq5.59 K.

The main difficulty of the HTSE fit is choosing the appropriate TT-range. The convergence of the HTSE depends on the Jˉ2/Jˉ1\bar{J}_2/\bar{J}_1 ratio, hence the lower limit of the fitting range (Tmin⁡T_{\min}) should be chosen with caution depending on the results of the fit [15]. For a precise estimation of the JJ values, we decided the lower limit of the fitting TT-range (Tmin⁡T_{\min}) between 5 and 9 K and estimated the exchange couplings for both the solutions. Figures 5(a) and (b) present the variation of Jˉ1\bar{J}_1 and Jˉ2\bar{J}_2 with Tmin⁡T_{\min} for the solutions-a and b, respectively. To check the convergence of the HTSE, we calculated the ratio of the ninth-order term to the total susceptibility [χ9(T)/χtotal(T)][\chi_9(T)/\chi_{\mathrm{total}}(T)] using the parameters obtained from HTSE fits with a different Tmin⁡T_{\min}. Both solutions are stable above 7 K. We also performed the FD simulations of χ(T)\chi(T) using the values of χ0\chi_0, gg, Jˉ1\bar{J}_1, and Jˉ2\bar{J}_2 and found a good agreement with the experimental data [inset of Figure 4(a)].

Discriminating between the solutions-a and b may be possible using measurements of the saturation field [47]. The MM vs HH data measured up to 25 T in pulsed magnetic fields at T=1.4T = 1.4 K are shown in Figure 6. We have scaled the high-field data with respect to the magnetic isotherm at T=1.8T = 1.8 K measured up to 7 T using a SQUID magnetometer. MM increases linearly with HH in the low-field region, shows a positive curvature in the intermediate fields, and then saturates at around HS≃15.3H_{\mathrm{S}} \simeq15.3 T. The positive curvature is typical for low-dimensional and frustrated spin systems [64].

Magnetization versus field at 1.4 K

Figure 6. Magnetization (MM) vs field (HH) at T=1.4T = 1.4 K measured using pulse magnetic field and scaled with respect to the SQUID data. The solid line represents the QMC simulation, assuming the spatially isotropic nonfrustrated spin-1/2 square-lattice model with Jeff=5.1J_{\mathrm{eff}} = 5.1 K. The arrow indicates the position of the saturation field (HSH_{\mathrm{S}}).

Saturation field of an FSL magnet depends on the type of magnetic order. In the NAF case (solution-a), HS=4Jˉ1kB/(gμB)≃15.1H_{\mathrm{S}} = 4\bar{J}_1 k_{\mathrm{B}}/(g\mu_{\mathrm{B}}) \simeq15.1 T in good agreement with the experiment. On the other hand, in the CAF case HS=(J1+2Jˉ2)2kB/(gμB)H_{\mathrm{S}} = (J_1 + 2\bar{J}_2)2k_{\mathrm{B}}/(g\mu_{\mathrm{B}}) where J1J_1 is the weaker of the couplings J1J_1 and (J1′+J1′′)/2(J'_1 + J''_1)/2. In the CAF state, spins align antiferromagnetically along the direction of this coupling and ferromagnetically along the orthogonal direction. Therefore, only the weaker coupling enters the saturation field. Using J1=Jˉ1J_1 = \bar{J}_1 leads to HSCAF≃14.45H_{\mathrm{S}}^{\mathrm{CAF}} \simeq14.45 T for the solution-b, lower than in the experiment. On the other hand, the actual value of J1J_1 is reduced owing to the deformation of the square lattice. Assuming 20% spatial anisotropy in the NN couplings (J1=0.8Jˉ1J_1 = 0.8\bar{J}_1) according to the DFT results (Sec. III E), we arrive at HSCAF≃15.15H_{\mathrm{S}}^{\mathrm{CAF}} \simeq15.15 T, which is on par with the result for the solution-a and also matches the experimental value.

Interestingly, the magnetization curve can be also well described by the simple NN square-lattice model with an effective coupling Jeff=5.1J_{\mathrm{eff}} = 5.1 K (Figure 6), which is in good agreement with the leading exchange couplings extracted from the HTSE fits.

Specific Heat

In Figure 7(a), we have plotted the temperature-dependent specific heat [Cp(T)][C_{\mathrm{p}}(T)] measured from 0.5 to 140 K in zero applied field. It decreases systematically with temperature and passes through a broad maximum at TCmax⁡≃3.7T_{\mathrm{C}}^{\max} \simeq3.7 K, typical for low-dimensional oxides. An anomalous behavior was observed in a narrow temperature range TN=2−2.2T_{\mathrm{N}} = 2 - 2.2 K where we couldn’t stabilize the temperature. This is a possible indication of the onset of a magnetic LRO.

Figure 7. (a) Variation of CpC_{\mathrm{p}} with temperature in the absence of magnetic field. The red solid line represents the simulated phonon contribution (CphC_{\mathrm{ph}}) taking into account the Debye (dash-dotted line) and Einstein (dashed line) terms. Inset: Cp/TC_{\mathrm{p}}/T vs TT at low temperatures and in different applied fields. The blue solid line is the magnetic contribution (CmagC_{\mathrm{mag}}). (b) Cmag/RC_{\mathrm{mag}}/R vs TT (left yy-axis). The solid line represents the change in the magnetic entropy, SmagS_{\mathrm{mag}} vs TT (right yy-axis). The red (dash-dotted) and blue (solid) lines are the HTSE fits corresponding to the two solutions, respectively. The green (dashed) line is the QMC simulation for an isotropic nonfrustrated spin-1/2 square lattice model with Jeff=6.3J_{\mathrm{eff}} = 6.3 K.

In a magnetic insulator, the total specific heat Cp(T)C_{\mathrm{p}}(T) is the sum of two main contributions: one is the phonon/lattice contribution [Cph(T)][C_{\mathrm{ph}}(T)], which dominates

Results of the $\chi(T)$ fit using HTSE for isotropic spin-$\frac{1}{2}$ FSL model

Figure 5. Results of the χ(T)\chi(T) fit using HTSE for isotropic spin-12\frac{1}{2} FSL model by varying the minimum temperature of the fitting range (Tmin⁡T_{\min}). (a) and (b) Averaged exchange interactions (Jˉ1\bar{J}_1 and Jˉ2\bar{J}_2) vs Tmin⁡T_{\min} for the solutions a and b, respectively. (c) Convergence test of the HTSE fit for both solutions, χ9(T)/χtotal(T)\chi_9(T)/\chi_{\mathrm{total}}(T) vs Tmin⁡T_{\min}. (d) Saturation fields for the solutions a and b vs Tmin⁡T_{\min}.

in the high-temperature region, and another one is the magnetic contribution [Cmag(T)][C_{\mathrm{mag}}(T)], which dominates in the low-temperature region depending upon the strength of the exchange coupling. In order to bring out the magnetic part of the specific heat, we first quantified the lattice contribution and then subtracted it from the total specific heat. We simulated the high-temperature

Cph(T)C_{\mathrm{ph}}(T) data taking into account the sum of one Debye [CD(T)C_{\mathrm{D}}(T)] and three Einstein [CE(T)C_{\mathrm{E}}(T)] terms, i.e., Cph(T)=CD(T)+∑i=13CEi(T)C_{\mathrm{ph}}(T) = C_{\mathrm{D}}(T) + \sum_{i=1}^{3} C_{\mathrm{E}i}(T). The Debye and Einstein terms are expressed as

CD(T)=9nDR(TθD)3∫0θD/Tx4ex(ex−1)2 dx(6)C_{\mathrm{D}}(T) = 9n_{\mathrm{D}}R\left(\frac{T}{\theta_{\mathrm{D}}}\right)^3 \int_{0}^{\theta_{\mathrm{D}}/T} \frac{x^4 e^x}{(e^x-1)^2}\,dx \tag*{(6)}

and

CEi(T)=3nEiR(θEiT)2e(θEi/T)[e(θEi/T)−1]2.(7)C_{\mathrm{E}i}(T) = 3n_{\mathrm{E}i}R\left(\frac{\theta_{\mathrm{E}i}}{T}\right)^2 \frac{e^{\left(\theta_{\mathrm{E}i}/T\right)}}{\left[e^{\left(\theta_{\mathrm{E}i}/T\right)}-1\right]^2}. \tag*{(7)}

Here, the Einstein temperatures θEi=hωEikB\theta_{\mathrm{E}i} = \frac{h\omega_{\mathrm{E}i}}{k_{\mathrm{B}}}, ωEi\omega_{\mathrm{E}i} are the respective Einstein frequencies, and RR denotes the universal gas constant. The values of nDn_{\mathrm{D}} and nEin_{\mathrm{E}i} are chosen in such a way that the sum nD+∑i=13nEin_{\mathrm{D}} + \sum_{i=1}^{3} n_{\mathrm{E}i} matches with the total number of atoms per formula unit. The best fit of the CpC_{\mathrm{p}} data in the high-temperature regime using one Debye and three Einstein branches yields the characteristic temperatures: θD≃122\theta_{\mathrm{D}} \simeq122 K, θE1≃307\theta_{\mathrm{E}1} \simeq307 K, θE2≃850\theta_{\mathrm{E}2} \simeq850 K, and θE3≃161\theta_{\mathrm{E}3} \simeq161 K with nD=1n_{\mathrm{D}} = 1, nE1=5n_{\mathrm{E}1} = 5, nE2=7n_{\mathrm{E}2} = 7, and nE3=2n_{\mathrm{E}3} = 2, respectively [65]. The red solid line in Figure 7(a) is the total phononic contribution to the specific heat (CphC_{\mathrm{ph}}) extrapolated down to low temperature, while the dashed and dash-dotted lines are the Einstein and Debye contributions, respectively. The value of θD\theta_{\mathrm{D}} and the average value of θE\theta_{\mathrm{E}} match well with θD\theta_{\mathrm{D}} estimated from the VcellV_{\mathrm{cell}} vs TT analysis.

CmagC_{\mathrm{mag}} estimated after subtracting CphC_{\mathrm{ph}} from CpC_{\mathrm{p}} is plotted in Figure 7(b) as a function of temperature. The

pronounced broad maximum at TCmax≃3.7T_{\mathrm{C}}^{\mathrm{max}} \simeq3.7 K mimics short-range antiferromagnetic correlations. The change in magnetic entropy [Smag(T)S_{\mathrm{mag}}(T)] is obtained by integrating Cmag/TC_{\mathrm{mag}}/T over temperature. It saturates to a value of ∼5.6\sim5.6 J.mol−1^{-1}.K−1^{-1} at around T≃25T \simeq25 K, which is close to the expected value Rln⁡2=5.76R\ln2 = 5.76 J.mol−1^{-1}.K−1^{-1} for a two-level (spin-12\frac{1}{2}) system, thus justifying our subtraction procedure and the evaluation of CmagC_{\mathrm{mag}}.

The magnetic LRO is highlighted by plotting Cp/TC_{\mathrm{p}}/T vs TT in the inset of Figure 7(a) in different applied fields. While the anomaly associated with the magnetic LRO is not pronounced in the zero-field data, the peak at TN≃2.1T_{\mathrm{N}} \simeq2.1 K becomes more pronounced with increasing the field. This is due to the transfer of entropy from the broad maximum to the transition anomaly. Surprisingly, no visible shift in TNT_{\mathrm{N}} is perceived even with the field change of 9 T, indication of a robust AFM transition.

One can estimate the exchange couplings by analyzing Cmag(T)C_{\mathrm{mag}}(T) using the HTSE of a spin-12\frac{1}{2} FSL model [55],

Cmag(T)R=J‾1kBT∑n(−n)(J‾1kBT)n∑memn(J‾2J‾1)m(8)\frac{C_{\mathrm{mag}}(T)}{R}=\frac{\overline{J}_1}{k_{\mathrm{B}}T}\sum_n(-n)\left(\frac{\overline{J}_1}{k_{\mathrm{B}}T}\right)^n\sum_m e_{mn}\left(\frac{\overline{J}_2}{\overline{J}_1}\right)^m \tag*{(8)}

We have fitted the Cmag(T)/RC_{\mathrm{mag}}(T)/R data at T>6T>6 K [see Figure 7(b)] and arrived at two solutions that strongly resemble the two solutions from the susceptibility fit: solution-a (J‾1/kB≃5.9\overline{J}_1/k_{\mathrm{B}}\simeq5.9 K, J‾2/kB≃−0.2\overline{J}_2/k_{\mathrm{B}}\simeq-0.2 K) and solution-b (J‾1/kB≃−1.6\overline{J}_1/k_{\mathrm{B}}\simeq-1.6 K, J‾2/kB≃6.1\overline{J}_2/k_{\mathrm{B}}\simeq6.1 K). Alternatively, we can compare our experimental data with the QMC simulation for the nonfrustrated square-lattice model [see Figure 7(b)]. The position of the maximum in CmagC_{\mathrm{mag}} is reproduced with Jeff=6.3J_{\mathrm{eff}}=6.3 K, which is notably higher than in the M(H)M(H) fit (Figure 6). Moreover, the maximum value of Cmag(T)/RC_{\mathrm{mag}}(T)/R is higher than in the experiment. The simple NN square-lattice model is thus insufficient to describe the magnetic behavior of NaZnVOPO4_4(HPO4_4). Both frustration and deformation of the square lattice affect thermodynamic properties of this compound.

31^{31}P NMR

NMR is a convenient local probe of both static and dynamic properties. In NaZnVOPO4_4(HPO4_4), the P(1) site is strongly coupled to the V4+^{4+} ions within the VOP(1)O4_4 layer, while the P(2) site which is located between the adjacent layers is weakly coupled to the V4+^{4+} ions [see Figure 1(a)]. This difference allows a useful comparison and highlights the magnetic behavior of the square-lattice planes.

31^{31}P NMR Spectra

The field-sweep 31^{31}P NMR spectra above TNT_{\mathrm{N}} measured in a radio frequency of 121 MHz are shown in Figure 8. Each spectrum is normalized by its maximum amplitude and offset vertically by adding a constant. At high temperatures, the line is found to be narrow but asymmetric and the central peak appears at the zero-shift position. As the temperature is lowered, the line width increases drastically and becomes more anisotropic with two shoulders on either side of the central peak. This abnormal spectral shape can be attributed to two nonequivalent P sites in the crystal structure [45]. Remarkably, the complete shape of the spectra could be reproduced considering the superposition of two spectral lines. The inset of Figure 8 portrays the spectral fit at T=12T=12 K. We have a narrow central line with weak anisotropy and a broad asymmetric background with two distinct shoulders. The narrow central peak shifts weakly, whereas the shoulders move significantly with decreasing temperature. Thus, the narrow central line with a weak shift

Temperature evolution of the $^{31}$P NMR spectra

Figure 8. Temperature evolution of the 31^{31}P NMR spectra of NaZnVOPO4_4(HPO4_4) measured at 121 MHz. The dashed line indicates the reference field position. Inset: 31^{31}P NMR spectrum at T=12T=12 K with the dashed and dotted lines are the fits of the P(1) and P(2) sites, respectively, and the solid line (final fit) is the superposition of the P(1) and P(2) fits. The NMR shift values along the xx, yy, and zz directions, obtained from the fitting are Kx(1)≃0.48%K_x^{(1)}\simeq0.48\%, Ky(1)≃−0.68%K_y^{(1)}\simeq-0.68\%, and Kz(1)≃0.24%K_z^{(1)}\simeq0.24\% for P(1) site and Kx(2)≃0.13%K_x^{(2)}\simeq0.13\%, Ky(2)≃−0.1%K_y^{(2)}\simeq-0.1\%, and Kz(2)≃−0.09%K_z^{(2)}\simeq-0.09\% for P(2) site, respectively.

can be assigned to the P(2) site, which is weakly coupled while the broad line with the strong temperature-dependent behavior corresponds to the in-plane P(1) site, which is strongly coupled to the V4+^{4+} spins. The asymmetric shape of both P(1) and P(2) sites is likely due to the anisotropy in χ(T)\chi(T) or asymmetry in the hyperfine coupling constant between the P nuclear spins and the V4+^{4+} electronic spins. The overall spectral shape matches exactly with the spectral shape reported for other V4+^{4+}-based FSL compounds, A′'A′′^{\prime\prime}VO(PO4_4)2_2 (A′'A′′^{\prime\prime} = Pb2_2, SrZn), on the polycrystalline sample [13, 16].

The nature of the magnetic ordering can also be gleaned from the analysis of NMR spectra below TNT_{\mathrm{N}}. Figure 9 shows the normalized 31^{31}P NMR spectra measured at 121 MHz around TNT_{\mathrm{N}}. Neither significant change in the line shape nor any visible line broadening are observed below TNT_{\mathrm{N}}. This indicates that the 31^{31}P site experiences only a weak static field in the ordered state. This observation is quite opposite to that reported for Pb2_2VO(PO4_4)2_2 [13] but similar to the 29^{29}Si NMR results on Li2_2VOSiO4_4 where local field on the Si site cancels out [66]. All these compounds show a similar type of magnetic layers and a similar mutual arrangement of the V4+^{4+} ions and PO4_4/SiO4_4 tetrahedra. However, only in Pb2_2VO(PO4_4)2_2 the magnetic [VOPO4_4] layers are strongly buckled [33]. This may explain why the hyperfine couplings to V4+^{4+} spins with the opposite alignment do not lead to the cancellation of the local field. On the other hand, in NaZnVOPO4_4(HPO4_4) the layers are almost flat (Figure 1), and the absence of the 31^{31}P NMR line broadening below TNT_{\mathrm{N}} can be still ascribed to the filtering of the hyperfine fields due to the AFM spin alignments in the ground state.

Temperature-dependent $^{31}$P NMR spectra

Figure 9. Temperature-dependent 31^{31}P NMR spectra of NaZnVOPO4_4(HPO4_4) measured at 121 MHz around TNT_{\mathrm{N}}. The dashed line indicates the reference field position.

31^{31}P NMR Shift

The spectrum at each temperature was fitted following the same procedure as described above for T=12T = 12 K. The estimated temperature-dependent NMR shift [K(T)][K(T)] along different orientations (KxK_x, KyK_y, and KzK_z) for both P(1) and P(2) sites are plotted in Figure 10(a) and (b), respectively. All the components of K(T)K(T) exhibit a broad maximum/minimum at around 4.3 K, an indication of the 2D AFM short-range-ordering. Further, the magnitude of K(T)K(T) for the P(2) site is weaker than the P(1) site, as expected. The isotropic NMR shift was calculated as Kiso=(Kx+Ky+Kz)/3K_{\mathrm{iso}} = (K_x + K_y + K_z)/3, which is found to be almost temperature-independent for both the 31^{31}P sites. This also suggests that the isotropic part of hyperfine coupling at the P site from the four neighboring V4+^{4+} spins is nearly averaged out.

Anisotropic components of $K$ and $K$ versus $\chi$ plots

Figure 10. Anisotropic components of KK along the xx, yy, and zz-directions as a function of temperature for (a) P(1) and (b) P(2) sites, respectively. KK vs χ\chi measured at 7 T are plotted for all the three orientations for (c) P(1) and (d) P(2) sites, respectively. The solid lines are the straight line fits.

As K(T)K(T) is an intrinsic measure of the spin susceptibility χspin(T)\chi_{\mathrm{spin}}(T) and is free from extrinsic contributions,

one can write

K(T)=K0+AhfNAχspin(T).(9)K(T) = K_0 + \frac{A_{\mathrm{hf}}}{N_{\mathrm{A}}}\chi_{\mathrm{spin}}(T). \tag*{(9)}

Here, K0K_0 is the temperature-independent chemical (orbital) shift and AhfA_{\mathrm{hf}} is the hyperfine coupling constant between the 31^{31}P nucleus and V4+^{4+} spins. In order to estimate AhfA_{\mathrm{hf}}, we have plotted KK vs χ\chi, assuming temperature as an implied variable in Figure 10(c) and (d) for P(1) and P(2) sites, respectively. In every case, KK vs χ\chi plot is linear in the whole temperature range. A straight line fit results Ahf(1)=(1134±19)A_{\mathrm{hf}}^{(1)} = (1134 \pm19) Oe/μB\mu_{\mathrm{B}}, (−1716±20)(-1716 \pm20) Oe/μB\mu_{\mathrm{B}}, and (582.59±11)(582.59 \pm11) Oe/μB\mu_{\mathrm{B}} along the xx, yy and zz-directions for the P(1) site and Ahf(2)=(343±13)A_{\mathrm{hf}}^{(2)} = (343 \pm13) Oe/μB\mu_{\mathrm{B}}, (−248±4)(-248 \pm4) Oe/μB\mu_{\mathrm{B}}, and (−192±3)(-192 \pm3) Oe/μB\mu_{\mathrm{B}} along the xx, yy and zz-directions for the P(2) site, respectively. Clearly, the magnitude of Ahf(1)A_{\mathrm{hf}}^{(1)} is almost one order of magnitude larger than Ahf(2)A_{\mathrm{hf}}^{(2)} in all the three directions proving the stronger coupling for P(1) than P(2).

For the P1 site, the transferred hyperfine coupling mainly arises from the interactions with the four nearest-neighbor V4+^{4+} spins in the plane. The isotropic and anisotropic transferred hyperfine couplings originate from P(3s)-O(2p)-V(3d) and P(3p)-O(2p)-V(3d) covalent bonds, respectively. Since P1 is surrounded by four V4+^{4+} ions forming a nearly square lattice in the plane, the experimentally observed asymmetry in hyperfine field indicates inequivalent P(3p)-O(2p)-V(3d) bonds for the four NN V4+^{4+} ions and hence a distortion in the square lattice, consistent with the low symmetry of the crystal structure as pointed out earlier.

Spin-lattice relaxation rate 311/T1^{31}1/T_1

We have performed 31^{31}P spin-lattice relaxation rate (1/T11/T_1) measurements as a function of temperature down to 1.7 K in two different frequencies, 30.2 MHz and 121 MHz. The recovery of the longitudinal nuclear magnetization after a saturation pulse could be fitted by a double exponential function

1−M(t)M(∞)=Ae−t/T11+Be−t/T12,(10)1-\frac{M(t)}{M(\infty)}=Ae^{-t/T_{11}}+Be^{-t/T_{12}}, \tag*{(10)}

where M(t)M(t) is the nuclear magnetization at a time tt after the saturation pulse and M(∞)M(\infty) is the equilibrium nuclear magnetization. As each NMR spectrum is a superposition of two P-sites, which are inseparable, a double exponential function is used to fit the recovery curves. In Eq. (10), 1/T111/T_{11} and 1/T121/T_{12} are the spin-lattice relaxation rates for P(1) and P(2) sites, respectively, and AA and BB account for their respective weight factors. At 121 MHz, the total spectral width was large and we were not able to saturate the whole spectrum using a single saturation pulse. Therefore, 1/T11/T_1 is also measured at a

dV−Vd_{\mathrm{V-V}}JiAFMJ_i^{\mathrm{AFM}}JiFMJ_i^{\mathrm{FM}}JiJ_i
J1J_14.6161.4−5.6−5.1
J1′J'_14.5421.4−7.4−7.9
J1′′J''_14.7852.6−3.1−1.0
J2J_26.38018.7−0.48.1
J2′J'_26.3959.1−0.44.6

Table 1. Interatomic V–V distances (in Å) and exchange couplings (in K). The FM and AFM contributions are obtained from the superexchange model, Eq. (11), whereas total exchange couplings JiJ_i are calculated via the DFT+UU mapping procedure and thus deviate from JiFM+JiAFMJ_i^{\mathrm{FM}}+J_i^{\mathrm{AFM}}.

lower frequency of 30.2 MHz where the spectral width is reduced significantly and we were able to saturate the whole spectrum above TNT_N.

The extracted 1/T111/T_{11} and 1/T121/T_{12} as a function of temperature at 30.2 MHz and 121 MHz and for both the P-sites are presented in Figure 11(b). In the high-temperature (T≥4.3T \geq4.3 K) region, 1/T11/T_1 for both the P-sites are almost constant, typically expected in the paramagnetic regime [67]. At low temperatures, 1/T111/T_{11} and 1/T121/T_{12} show sharp peaks at around TN≃2.04T_N \simeq2.04 K and 2.36 K for 30.2 MHz and 121 MHz, respectively, indicating the slowing down of fluctuating moments as we approach the magnetic LRO. Below TNT_N, both 1/T111/T_{11} and 1/T121/T_{12} decrease toward zero due to the scattering of magnons by the nuclear spins [68].

Longitudinal recovery curves and spin-lattice relaxation rates

Figure 11. (a) Longitudinal recovery curves at three selective temperatures and the solid lines are fits using Eq. (10). (b) 31^{31}P spin-lattice relaxation rates for P(1) (1/T111/T_{11}) and P(2) (1/T121/T_{12}) sites as a function of temperature measured in 30.2 MHz and 121 MHz. The data are shown in log-log scale in order to highlight the peak at TNT_N. The vertical dash-dotted lines indicate the magnetic LRO at TN≃2.04T_N \simeq2.04 K and TN≃2.2T_N \simeq2.2 K for 30.2 MHz and 121 MHz data, respectively.

Microscopic magnetic model

Before discussing the magnetic model, we determine the position of hydrogen, which may be crucial for the correct evaluation of exchange couplings [69]. In hydrophosphates, one expects a deformation of PO4_4 tetrahedra because one of the oxygens is linked to hydrogen and should thus weaken its bond to phosphorus in order to keep the overall bond valence unchanged. Such a deformation is observed in the P(2)O4_4 tetrahedra, with the P–O bond distances of 1.490 Å [O(1)], 1.530 [O(7)], 1.537 Å [O(2)], and 1.585 Å [O(9)] [45]. We thus considered different hydrogen positions in the vicinity of O(9) and found the lowest energy for hydrogen located at (0.7658, 0.2823, 0.7588) and separated from oxygen by 0.988 Å. The O–H bond is directed toward the [VOPO4_4] layer, with the H atom forming three hydrogen bonds of 2.2–2.4 Å to oxygen atoms of the VO5_5 pyramids and P(1)O4_4 tetrahedra (Figure 1).

We now adopt this hydrogen position in DFT calculations and evaluate exchange couplings using two complementary methods. The first one is the superexchange model based on electron hoppings extracted by Wannier fits to the PBE band structure, resulting in [50]

Ji=4txy→xy2Ueff−∑α4txy→α2Jeff(Ueff+Δα)(Ueff+Δα−Jeff)(11)J_i = \frac{4t_{xy\to xy}^{2}}{U_{\mathrm{eff}}} - \sum_{\alpha} \frac{4t_{xy\to\alpha}^{2}J_{\mathrm{eff}}}{(U_{\mathrm{eff}}+\Delta_{\alpha})(U_{\mathrm{eff}}+\Delta_{\alpha}-J_{\mathrm{eff}})} \tag*{(11)}

where dxyd_{xy} is the half-filled orbital of V4+\mathrm{V}^{4+}, α\alpha labels unoccupied dd-orbitals, and Δα\Delta_{\alpha} is the crystal-field splitting. The first term stands for AFM superexchange arising from electron hoppings between the half-filled orbitals, whereas the second term is FM superexchange due to electron hoppings between the half-filled and empty orbitals. Using the effective Coulomb repulsion of Ueff=4 eVU_{\mathrm{eff}}=4\,\mathrm{eV} and Hund’s coupling of Jeff=1 eVJ_{\mathrm{eff}}=1\,\mathrm{eV} [70] from Refs. [51, 53], we arrive at the exchange couplings JiAFMJ_i^{\mathrm{AFM}} and JiFMJ_i^{\mathrm{FM}} listed in Table I. Additionally, we compute total exchange couplings JiJ_i via the DFT+UU mapping approach [52, 53].

Both methods arrive at qualitatively similar results. The NN couplings are FM, whereas the NNN couplings are AFM in nature, thus favoring solution-b. The largest anisotropy is observed between J1′J_1' and J1′′J_1''. However, on average these couplings – both running along cc – differ from J1J_1 (along bb) by 20% only. The spatial anisotropy of the NNN couplings is more significant with J2/J2′≃1.8J_2/J_2' \simeq1.8. The overall energy scale of the NNN couplings, Jˉ2≃5.6 K\bar{J}_2 \simeq5.6\,\mathrm{K}, is similar to Li2_2VOSiO4_4 (J2≃5.9 KJ_2 \simeq5.9\,\mathrm{K} [55]) and Na1.5_{1.5}VOPO4_4F0.5_{0.5} (Jˉ2≃6.6 K\bar{J}_2 \simeq6.6\,\mathrm{K}) and much lower than in Pb2_2VO(PO4_4)2_2 with Jˉ2≃9.3 K\bar{J}_2 \simeq9.3\,\mathrm{K} [13] or SrZnVO(PO4_4)2_2 with Jˉ2=8.6 K\bar{J}_2=8.6\,\mathrm{K} [16]. This difference can be traced back to the buckling of the magnetic layers in AA′VO(PO4_4)2_2 (AA′ = Pb2_2, SrZn), while the layers are almost or even perfectly flat in NaZnVOPO4_4(HPO4_4) (Figure 1), Li2_2VOSiO4_4, and Na1.5_{1.5}VOPO4_4F0.5_{0.5}.

Discussion

Our data suggest that NaZnVOPO4_4(HPO4_4) is well described by the spin-12\frac{1}{2} FSL model if deformation of the square lattice is taken into account. Although individual data sets such as field-dependent magnetization may be consistent even with a nonfrustrated square lattice, the reduced size of the specific heat maximum, Cmax⁡/R≃0.41C_{\max}/R \simeq0.41, indicates the presence of frustration. Indeed, the value of the frustration parameter f=∣θCW∣TN≃2.1f=\frac{|\theta_{\mathrm{CW}}|}{T_{\mathrm{N}}}\simeq2.1 reflects a moderate frustration in the compound.

In a spin system, the spin-lattice relaxation rate carries information on the low-lying excitations or spin dynamics in the momentum space. Typically, 1T1T\frac{1}{T_1T} can be expressed in terms of the dynamic susceptibility χM(q⃗,ω0)\chi_M(\vec{q},\omega_0) as [71]

1T1T=2γN2kBNA2∑q⃗∣A(q⃗)∣2χM′′(q⃗,ω0)ω0(12)\frac{1}{T_1T}=\frac{2\gamma_N^2 k_B}{N_A^2}\sum_{\vec{q}} |A(\vec{q})|^2\frac{\chi_M''(\vec{q},\omega_0)}{\omega_0} \tag*{(12)}

where the sum is over the wave vector q⃗\vec{q} within the first Brillouin zone, A(q⃗)A(\vec{q}) is the form-factor of the hyperfine interaction, and χM′′(q⃗,ω0)\chi_M''(\vec{q},\omega_0) is the imaginary part of the dynamic susceptibility at the nuclear Larmor frequency ω0\omega_0. The dominance of different qq-components [q⃗=0\vec{q}=0 and q⃗=(±π/a,±π/b)\vec{q}=(\pm\pi/a,\pm\pi/b)] is often visible in the 1/T11/T_1 data when plotted against temperature, especially for the low-dimensional spin systems with strong exchange coupling. At very high temperatures (T>J/kBT>J/k_B), 1/T11/T_1 is almost temperature-independent due to uncorrelated moments and can be expressed as [13]

(1T1)T→∞=(γNgμB)22πz′S(S+1)3ωex×(Axz′)2+(Ayz′)2+(Azz′)23(13)\left(\frac{1}{T_1}\right)_{T\to\infty}=\frac{(\gamma_N g\mu_B)^2\sqrt{2\pi}z'S(S+1)}{3\omega_{\mathrm{ex}}}\times\frac{\left(\frac{A_x}{z'}\right)^2+\left(\frac{A_y}{z'}\right)^2+\left(\frac{A_z}{z'}\right)^2}{3} \tag*{(13)}

where ωex=[max⁡(∣J1∣,∣J2∣)kB/ℏ]2zS(S+1)3\omega_{\mathrm{ex}}=[\max(|J_1|,|J_2|)k_B/\hbar]\sqrt{\frac{2zS(S+1)}{3}} is the Heisenberg exchange frequency, zz is the number of NN spins of each V4+^{4+} ion, and z′z' is the number of NN V4+^{4+} spins of the P(1) site. In the above expression, the hyperfine couplings along different directions are divided by z′z' in order to account for the coupling of P site with the individual V4+^{4+} ion. As the measurements are carried out on the powder sample, we have taken the rms average of the couplings along three directions.

For a tentative estimation of the in-plane exchange coupling between the V4+^{4+} ions we took the high temperature value of 1/T111/T_{11} for the strongly coupled P(1) site. Using the experimental parameters obtained for this site (Ax1≃1134 Oe/μBA_x^1\simeq1134\,\mathrm{Oe}/\mu_B, Ay1≃−1716 Oe/μBA_y^1\simeq-1716\,\mathrm{Oe}/\mu_B, Az1≃582 Oe/μBA_z^1\simeq582\,\mathrm{Oe}/\mu_B, γN=108.303×102 rad.sec−1/Oe\gamma_N=108.303\times10^2\,\mathrm{rad.sec}^{-1}/\mathrm{Oe}, z′=4z'=4, z=4z=4, g=2.04g=2.04, S=1/2S=1/2, and 1/T11≃150 sec−11/T_{11}\simeq150\,\mathrm{sec}^{-1}), the magnitude of the maximum exchange coupling strength between V4+^{4+} ions is estimated to be max⁡(∣J1∣,∣J2∣)≃4.2 K\max(|J_1|,|J_2|)\simeq4.2\,\mathrm{K}. This value is indeed very close to the dominant in-plane AFM exchange coupling for both the solutions (a and b), obtained from the χ(T)\chi(T) analysis.

It is also instructive to compare the transition temperatures of V4+^{4+}-based FSL magnets. NaZnVOPO4_4(HPO4_4) and Li2_2VOSiO4_4 feature almost the same J2≃6 KJ_2\simeq6\,\mathrm{K} but different signs of J1J_1 (FM and AFM, respectively), while the former compound has a much lower TN≃2.1 KT_N\simeq2.1\,\mathrm{K} than the latter (TN≃2.8 KT_N\simeq2.8\,\mathrm{K} [11]). This indicates that FM couplings J1J_1 together with the deformation of the square lattice lead to a visible reduction in TNT_N in NaZnVOPO4_4(HPO4_4). On the other hand, Pb2_2VO(PO4_4)2_2 shows an even higher TN≃3.65 KT_N\simeq3.65\,\mathrm{K} [13] because of the 30% increase in the magnitude of J2J_2. This comparison illustrates that buckling of the [VOPO4_4] layers caused by variable spacers between the magnetic layers effectively tunes magnetic interactions in FSL-like compounds.

Summary

We have studied the magnetism of NaZnVOPO4_4(HPO4_4) using wide variety of experimental techniques and complementary ab initioab\ initio calculations. With almost flat magnetic layers, this compound shows weaker next-nearest-neighbor couplings (Jˉ2\bar{J}_2) than other V4+\mathrm{V}^{4+} phosphates. This reduction in Jˉ2\bar{J}_2 along with the deformation of the spin-12\frac{1}{2} square lattice lead to a lower magnetic transition temperature than in most of the other FSL candidates.

Acknowledgments

SG and RN would like to acknowledge SERB, India for financial support bearing sanction Grant No. CRG/2019/000960. SG is supported by the Prime Minister’s Research Fellowship (PMRF) scheme, Government of India. We also acknowledge the support of the HLD at HZDR, member of European Magnetic Field Laboratory (EMFL). Work at the Ames Laboratory was supported by the U.S. Department of Energy, Office of Science, Basic Energy Sciences, Materials Sciences and Engineering Division. The Ames Laboratory is operated for the U.S. Department of Energy by Iowa State University under Contract No. DEAC02-07CH11358.

References

  1. [1]L. Savary and L. Balents, Quantum spin liquids: a review, Rep. Prog. Phys. 80, 016502 (2016); L. Balents, Spin liquids in frustrated magnets, Nature 464, 199 (2010).DOI
  2. [2]N. Shannon, B. Schmidt, K. Penc, and P. Thalmeier, Finite temperature properties and frustrated ferromagnetism in a square lattice Heisenberg model, Eur. Phys. J. B 38, 599 (2004).DOI
  3. [3]B. Schmidt and P. Thalmeier, Néel temperature and reentrant H−T phase diagram of quasi-two-dimensional frustrated magnets, Phys. Rev. B 96, 214443 (2017).DOI
  4. [4]G.-M. Zhang, H. Hu, and L. Yu, Valence-Bond Spin-Liquid State in Two-Dimensional Frustrated Spin-1/2 Heisenberg Antiferromagnets, Phys. Rev. Lett. 91, 067201 (2003).DOI
  5. [5]W.-J. Hu, F. Becca, A. Parola, and S. Sorella, Direct evidence for a gapless Z2 spin liquid by frustrating Néel antiferromagnetism, Phys. Rev. B 88, 060402(R) (2013).DOI
  6. [6]L. Wang and A. W. Sandvik, Critical Level Crossings and Gapless Spin Liquid in the Square-Lattice Spin-1/2 J1 − J2 Heisenberg Antiferromagnet, Phys. Rev. Lett. 121, 107202 (2018).DOI
  7. [7]S.-S. Gong, W. Zhu, D. N. Sheng, O. I. Motrunich, and M. P. A. Fisher, Plaquette Ordered Phase and Quantum Phase Diagram in the Spin-1/2 J1−J2 Square Heisenberg Model, Phys. Rev. Lett. 113, 027201 (2014).DOI
  8. [8]R. L. Doretto, Plaquette valence-bond solid in the square-lattice J1-J2 antiferromagnet Heisenberg model: A bond operator approach, Phys. Rev. B 89, 104415 (2014).DOI
  9. [9]R. Haghshemash and D. N. Sheng, U(1)-symmetric infinite projected entangled-pair states study of the spin-1/2 square J1−J2 Heisenberg model, Phys. Rev. B 97, 174408 (2018).
  10. [10]N. Shannon, T. Momoi, and P. Sindzингre, Nematic Order in Square Lattice Frustrated Ferromagnets, Phys. Rev. Lett. 96, 027213 (2006).
  11. [11]R. Melzi, P. Carretta, A. Lascialfari, M. Mambrini, M. Troyer, P. Millet, and F. Mila, Li2VO(Si,Ge)O4, a Prototype of a Two-Dimensional Frustrated Quantum Heisenberg Antiferromagnet, Phys. Rev. Lett. 85, 1318 (2000).pubmed.ncbi.nlm.nih.gov/10991541
  12. [12]H. Rosner, R. R. P. Singh, W. H. Zheng, J. Oitmaa, S.-L. Drechsler, and W. E. Pickett, Realization of a large J2 quasi-2D spin-half Heisenberg system: Li2VOSiO4, Phys. Rev. Lett. 88, 186405 (2002).pubmed.ncbi.nlm.nih.gov/12005707
  13. [13]R. Nath, Y. Furukawa, F. Borsa, E. E. Kaul, M. Baenitz, C. Geibel, and D. C. Johnston, Single-crystal 31P NMR studies of the frustrated square-lattice compound Pb2VO(PO4)2, Phys. Rev. B 80, 214430 (2009).
  14. [14]B. Roy, Y. Furukawa, R. Nath, and D. C. Johnston, Low-temperature 31P NMR study of the two-dimensional frustrated square lattice compound BaCdVO(PO4)2, J. Phys. Conf. Ser 320, 021048 (2011).
  15. [15]A. A. Tsirlin, R. Nath, A. M. Abakumov, R. V. Shpanchenko, C. Geibel, and H. Rosner, Frustrated square lattice with spatial anisotropy: Crystal structure and magnetic properties of PbZnVO(PO4)2, Phys. Rev. B 81, 174424 (2010).DOI
  16. [16]L. Bossoni, P. Carretta, R. Nath, M. Moscardini, M. Baenitz, and C. Geibel, NMR and μSR study of spin correlations in SrZnVO(PO4)2: An S = 1/2 frustrated magnet on a square lattice, Phys. Rev. B 83, 014412 (2011).arxiv.org/abs/1010.3495
  17. [17]A. A. Tsirlin, R. Nath, A. M. Abakumov, Y. Furukawa, D. C. Johnston, M. Hemmida, H.-A. Krug von Nidda, A. Loidl, C. Geibel, and H. Rosner, Phase separation and frustrated square lattice magnetism of Na1.5VOPO4F0.5, Phys. Rev. B 84, 014429 (2011).DOI
  18. [18]R. Nath, A. A. Tsirlin, H. Rosner, and C. Geibel, Magnetic properties of BaCdVO(PO4)2: A strongly frustrated spin-1/2 square lattice close to the quantum critical regime, Phys. Rev. B 78, 064422 (2008).DOI
  19. [19]A. Yogi, N. Ahmed, R. Nath, A. A. Tsirlin, S. Kundu, A. V. Mahajan, J. Sichelschmidt, B. Roy, and Y. Furukawa, Antiferromagnetism of Zn2VO(PO4)2 and the dilution with Ti4+, Phys. Rev. B 91, 024413 (2015).DOI
  20. [20]A. Kiani and E. Pavarini, Electronic correlation and magnetic frustration in Li2VOSiO4 and VOMoO4, Phys. Rev. B 94, 075112 (2016).DOI
  21. [21]A. Bombardi, L. C. Chapon, I. Margiolaki, C. Mazzoli, S. Gonthier, F. Duc, and P. G. Radaelli, Magnetic order and lattice anomalies in the J1 − J2 model system VOMoO4, Phys. Rev. B 71, 220406 (2005).DOI
  22. [22]P. Carretta, N. Papinutto, C. B. Azzoni, M. C. Mozzati, E. Pavarini, S. Gonthier, and P. Millet, Frustration-driven structural distortion in VOMoO4, Phys. Rev. B 66, 094420 (2002).DOI
  23. [23]A. A. Tsirlin, A. A. Belik, R. V. Shpanchenko, E. V. Antipov, E. Takayama-Muromachi, and H. Rosner, Frustrated spin-1/2 square lattice in the layered perovskite PbVO3, Phys. Rev. B 77, 094202 (2008).DOI
  24. [24]H. Ishikawa, N. Nakamura, M. Yoshida, M. Takigawa, P. Babkevich, N. Qureshi, H. M. Rønnow, T. Yajima, and Z. Hiroi, J₁ − J₂ square-lattice Heisenberg antiferromagnets with 4d¹ spins: AMoPO₄Cl (A = K,Rb), Phys. Rev. B 95, 064408 (2017).
  25. [25]H. Takeda, T. Yamauchi, M. Takigawa, H. Ishikawa, and Z. Hiroi, Pressure-induced phase transition in the J₁−J₂ square lattice antiferromagnet RbMoPO₄Cl, Phys. Rev. B 103, 104406 (2021).
  26. [26]Y. Xu, S. Liu, N. Qu, Y. Cui, Q. Gao, R. Chen, J. Wang, F. Gao, and X. Hao, Comparative description of magnetic interactions in Sr₂CuTeO₆ and Sr₂CuWO₆, J. Phys. Condens. Matter 29, 105801 (2017).DOI
  27. [27]S. Vasala, M. Aydemir, S. Danilkin, O. Chmaissem, and M. Karppinen, Magnetic structure of Sr₂CuWO₆, J. Phys. Condens. Matter 26, 496001 (2014).DOI
  28. [28]T. Koga, N. Kurita, M. Avdeev, S. Danilkin, T. J. Sato, and H. Tanaka, Magnetic structure of the S = 1/2 quasi-two-dimensional square-lattice Heisenberg antiferromagnet Sr₂CuTeO₆, Phys. Rev. B 93, 054426 (2016).
  29. [29]P. Babkevich, V. M. Katukuri, B. Fåk, S. Rols, T. Fennell, D. Pajić, H. Tanaka, T. Pardini, R. R. P. Singh, A. Mitrushchenkov, O. V. Yazyev, and H. M. Rønnow, Magnetic Excitations and Electronic Interactions in Sr₂CuTeO₆: A Spin-1/2 Square Lattice Heisenberg Antiferromagnet, Phys. Rev. Lett. 117, 237203 (2016).
  30. [30]H. C. Walker, O. Mustonen, S. Vasala, D. J. Voneshen, M. D. Le, D. T. Adroja, and M. Karppinen, Spin wave excitations in the tetragonal double perovskite Sr₂CuWO₆, Phys. Rev. B 94, 064411 (2016).
  31. [31]M. Watanabe, N. Kurita, H. Tanaka, W. Ueno, K. Matsui, T. Goto, and M. Hagihara, Contrasting magnetic structures in SrLaCuSbO₆ and SrLaCuNbO₆: Spin-1/2 quasi-square-lattice J₁−J₂ Heisenberg antiferromagnets, Phys. Rev. B 105, 054414 (2022).
  32. [32]S. Guchhait, S. Baby, M. Padmanabhan, A. Medhi, and R. Nath, Quasi-two-dimensional magnetism in spin-1/2 square lattice compound Cu[C₆H₂(COO)₄][H₃N-(CH₂)₂-NH₃]·3H₂O, EPL 133, 57006 (2021).
  33. [33]A. A. Tsirlin and H. Rosner, Extension of the spin-1/2 frustrated square lattice model: The case of layered vanadium phosphates, Phys. Rev. B 79, 214417 (2009).
  34. [34]S. Bettler, F. Landolt, O. M. Aksoy, Z. Yan, S. Gvasaliya, Y. Qiu, E. Ressouche, K. Beauvois, S. Raymond, A. N. Ponomaryov, S. A. Zvyagin, and A. Zheludev, Magnetic structure and spin waves in the frustrated ferro-antiferromagnet Pb₂VO(PO₄)₂, Phys. Rev. B 99, 184437 (2019).
  35. [35]F. Landolt, Z. Yan, S. Gvasaliya, K. Beauvois, E. Ressouche, J. Xu, and A. Zheludev, Phase diagram and spin waves in the frustrated ferro-antiferromagnet SrZnVO(PO₄)₂, Phys. Rev. B 104, 224435 (2021).
  36. [36]V. K. Bhartiya, S. Hayashida, K. Y. Povarov, Z. Yan, Y. Qiu, S. Raymond, and A. Zheludev, Inelastic neutron scattering determination of the spin hamiltonian for BaCdVO(PO₄)₂, Phys. Rev. B 103, 144402 (2021).
  37. [37]K. Y. Povarov, V. K. Bhartiya, Z. Yan, and A. Zheludev, Thermodynamics of a frustrated quantum magnet on a square lattice, Phys. Rev. B 99, 024413 (2019).
  38. [38]V. K. Bhartiya, K. Y. Povarov, D. Blosser, S. Bettler, Z. Yan, S. Gvasaliya, S. Raymond, E. Ressouche, K. Beauvois, J. Xu, F. Yokaïchya, and A. Zheludev, Presaturation phase with no dipolar order in a quantum ferro-antiferromagnet, Phys. Rev. Research 1, 033078
  39. [39]M. Skoulatos, F. Rucker, G. J. Nilsen, A. Bertin, E. Pomjakushina, J. Ollivier, A. Schneidewind, R. Georgii, O. Zaharko, L. Keller, C. Rüegg, C. Pfleiderer, B. Schmidt, N. Shannon, A. Kriele, A. Senyshyn, and A. Smerald, Putative spin-nematic phase in BaCdVO(PO₄)₂, Phys. Rev. B 100, 014405 (2019).
  40. [40]F. Landolt, S. Bettler, Z. Yan, S. Gvasaliya, A. Zheludev, S. Mishra, I. Sheikin, S. Krämer, M. Horvatić, A. Gazizulina, and O. Prokinenko, Presaturation phase in the frustrated ferro-antiferromagnet Pb₂VO(PO₄)₂, Phys. Rev. B 102, 094414 (2020).
  41. [41]K. M. Ranjith, F. Landolt, S. Raymond, A. Zheludev, and M. Horvatić, NMR evidence against a spin-nematic nature of the presaturation phase in the frustrated magnet SrZnVO(PO₄)₂, Phys. Rev. B 105, 134429 (2022).
  42. [42]F. Landolt, K. Povarov, Z. Yan, S. Gvasaliya, E. Ressouche, S. Raymond, V. O. Garlea, and A. Zheludev, Spin correlations in the frustrated ferro-antiferromagnet SrZnVO(PO₄)₂ near saturation, arXiv preprint arXiv:2204.08778 (2022).
  43. [43]K. Momma and F. Izumi, VESTA 3 for three-dimensional visualization of crystal, volumetric and morphology data, J. Appl. Crystallogr. 44, 1272 (2011).
  44. [44]J. Rodríguez-Carvajal, Recent advances in magnetic structure determination by neutron powder diffraction, Physica B Condens. Matter 192, 55 (1993).
  45. [45]E. Le Fur, Y. Moreno, and J. Pivan, On the new oxovanadium phosphate NaZnVOPO₄(HPO₄) obtained at 510 K from hydrothermal treatment, J. Mater. Chem. 11, 1735 (2001).
  46. [46]Y. Skourski, M. D. Kuz’min, K. P. Skokov, A. V. Andreev, and J. Wosnitza, High-field magnetization of Ho₂Fe₁₇, Phys. Rev. B 83, 214420 (2011).
  47. [47]A. A. Tsirlin, B. Schmidt, Y. Skourski, R. Nath, C. Geibel, and H. Rosner, Exploring the spin-1/2 frustrated square lattice model with high-field magnetization studies, Phys. Rev. B 80, 132407 (2009).
  48. [48]K. Kopérnik and H. Eschrig, Full-potential nonorthogonal local-orbital minimum-basis band-structure scheme, Phys. Rev. B 59, 1743 (1999).
  49. [49]J. P. Perdew, K. Burke, and M. Ernzerhof, Generalized gradient approximation made simple, Phys. Rev. Lett. 77, 3865 (1996).
  50. [50]V. V. Mazurenko, F. Mila, and V. I. Anisimov, Electronic structure and exchange interactions of Na₂V₃O₇, Phys. Rev. B 73, 014418 (2006).
  51. [51]A. A. Tsirlin, O. Janson, and H. Rosner, Unusual ferromagnetic superexchange in CdVO₃: The role of Cd, Phys. Rev. B 84, 144429 (2011).
  52. [52]H. J. Xiang, E. J. Kan, S.-H. Wei, M.-H. Whangbo, and X. G. Gong, Predicting the spin-lattice order of frustrated systems from first principles, Phys. Rev. B 84, 224429 (2011).
  53. [53]A. A. Tsirlin, Spin-chain magnetism and uniform Dzyaloshinsky-Moriya anisotropy in BaV₃O₈, Phys. Rev. B 89, 014405 (2014).
  54. [54]F. Weickert, N. Harrison, B. L. Scott, M. Jaime, A. Leitmaier, I. Heinmaa, R. Stern, O. Janson, H. Berger, H. Rosner, and A. A. Tsirlin, Magnetic anisotropy in the frustrated spin-chain compound β-TeVO₄, Phys. Rev. B 94, 064403 (2016).
  55. [55]H. Rosner, R. R. P. Singh, W. H. Zheng, J. Oitmaa, and W. E. Pickett, High-temperature expansions for the J₁–J₂ Heisenberg models: Applications to ab initio calculated models for Li₂VOSiO₄ and Li₂VOGeO₄, Phys. Rev. B 67, 014416 (2003).
  56. [56]F. Alet, S. Wessel, and M. Troyer, Generalized directed loop method for quantum Monte Carlo simulations, Phys. Rev. E 71, 036706 (2005).pubmed.ncbi.nlm.nih.gov/15903632
  57. [57]A. W. Sandvik, Stochastic series expansion method with operator-loop update, Phys. Rev. B 59, R14157 (1999).DOI
  58. [58]B. Bauer, L. D. Carr, H. G. Evertz, A. Feiguin, J. Freire, S. Fuchs, L. Gamper, J. Gukelberger, E. Gull, S. Guertler, et al., The ALPS project release 2.0: open source software for strongly correlated systems, J. Stat. Mech. Theory Exp. 2011, P05001 (2011).
  59. [59]D. C. Wallace, Thermodynamics of Crystals (Dover Publications, 1998).
  60. [60]L. E. Budd, R. M. Ibberson, W. G. Marshall, and S. Parsons, The effect of temperature and pressure on the crystal structure of piperidine, Chem. Cent. J. 9, 18 (2015); S. Guchhait, Q.-P. Ding, M. Sahoo, A. Giri, S. Maji, Y. Furukawa, and R. Nath, Quasi-one-dimensional uniform spin-½ Heisenberg antiferromagnet KNaCu₂O₇ probed by ³¹P and ²³Na NMR, Phys. Rev. B 103, 224415 (2021).
  61. [61]C. Kittel, Introduction to Solid State Physics (Wiley, 2004).
  62. [62]P. W. Selwood, Magnetochemistry (Read Books Ltd, 2013).
  63. [63]P. K. Mukharjee, K. M. Ranjith, M. Baenitz, Y. Skourski, A. A. Tsirlin, and R. Nath, Two types of alternating spin-½ chains and their field-induced transitions in ε−LiVOP₄, Phys. Rev. B 101, 224403 (2020).
  64. [64]P. Thalmeier, M. E. Zhitomirsky, B. Schmidt, and N. Shannon, Quantum effects in magnetization of J₁–J₂ square lattice antiferromagnet, Phys. Rev. B 77, 104411 (2008).DOI
  65. [65]S. J. Sebastian, K. Somesh, M. Nandi, N. Ahmed, P. Bag, M. Baenitz, B. Koo, J. Sichelschmidt, A. A. Tsirlin, Y. Furukawa, and R. Nath, Quasi-one-dimensional magnetism in the spin-½ antiferromagnet BaNa₂Cu(VO₄)₂, Phys. Rev. B 103, 064413 (2021).
  66. [66]R. Melzi, S. Aldrovandi, F. Tedoldi, P. Carretta, P. Millet, and F. Mila, Magnetic and thermodynamic properties of Li₂VOSiO₄: A two-dimensional S = 1/2 frustrated antiferromagnet on a square lattice, Phys. Rev. B 64, 024409 (2001), cond-mat/0101066.
  67. [67]T. Moriya, Nuclear magnetic relaxation in antiferromagnetics, Prog. Theor. Phys. 16, 23 (1956).
  68. [68]S. S. Islam, K. M. Ranjith, M. Baenitz, Y. Skourski, A. A. Tsirlin, and R. Nath, Frustration of square cupola in Sr(TiO)Cu₄(PO₄)₄, Phys. Rev. B 97, 174432 (2018).DOI
  69. [69]S. Lebernegg, A. A. Tsirlin, O. Janson, and H. Rosner, Two energy scales of spin dimers in clinoclase Cu₃(AsO₄)(OH)₃, Phys. Rev. B 87, 235117 (2013).DOI
  70. [70]Note that Ueff and Jeff apply to V-based Wannier functions that include both V 3d and O 2p orbitals, while the DFT+U parameters Ud and Jd mentioned in Sec. II describe, respectively, the Coulomb repulsion and Hund’s coupling in V 3d orbitals only.
  71. [71]T. Moriya, The effect of electron-electron interaction on the nuclear spin relaxation in metals, J. Phys. Soc. Jpn. 18, 516 (1963).DOI

Paper details

vlnc:W_bayy2xb5
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 0

downloads · 0

A view counts once per browser per day, a download each time the PDF is requested.

  • Physics and Astronomy

Not recorded for this paper

View statement
Looking for related papers…
Contents