betaFeedback

WorksPhysics and Astronomy

Rhenium oxyhalides: a showcase for anisotropic-triangular-lattice quantum antiferromagnets

Abstract

The spin-1/2 Heisenberg antiferromagnet on an anisotropic triangular lattice (ATL) is an archetypal spin system hosting exotic quantum magnetism and dimensional crossover. However, the progress in experimental research on this field has been limited due to the scarcity of ideal model materials. Here, we show that rhenium oxyhalides A3A_{3}ReO5X2_{5}X_{2}, where spin-1/2 Re6+^{6+} ions form a layered structure of ATLs, allow for flexible chemical substitution in both cation A2+A^{2+} (AA = Ca, Sr, Ba, Pb) and anion X−X^{-} (XX = Cl, Br) sites, leading to seven synthesizable compounds. By combining magnetic susceptibility and high-field magnetization measurements with theoretical calculations using the orthogonalized finite-temperature Lanczos method, we find that the anisotropy J′/JJ'/J ranges from 0.25 to 0.45 depending on the chemical composition. Our findings demonstrate that A3A_{3}ReO5X2_{5}X_{2} is an excellent platform for realizing diverse effective spin Hamiltonians that differ in the strength of the anisotropy J′/JJ'/J as well as the relevance of perturbation terms such as the Dzyaloshinskii-Moriya interaction and interlayer exchange coupling.

Frustrated quantum magnets have long drawn considerable attention due to the possibility of realizing exotic magnetic states [1] and fractional quasi-particle excitations [2]. Although recent advances in analytical and numerical approaches to describe many-body quantum physics have promoted our understanding of various magnetic phenomena, many issues remain unresolved [3], highlighting the need for further experimental insights. To this end, it is desirable to develop model compounds with a simple effective spin Hamiltonian, an accessible energy scale (i.e., temperature and magnetic field), and a broad tunability of perturbative terms as well as exchange couplings through chemical substitution.

The spin-1/2 Heisenberg antiferromagnet on an anisotropic triangular lattice (ATL) is an archetypal frustrated spin system [4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20]. The minimal model consists of two kinds of antiferromagnetic (AFM) exchange couplings, JJ and J′J', which dominate between the nearest-neighbor sites along the intrachain and interchain directions, respectively. By varying the ratio J′/JJ'/J, this model provides an interpolation between a regular triangular lattice (J′/J=1J'/J = 1) and a decoupled one-dimensional (1D) chain (J′/J=0J'/J = 0), with known ground states of the 120° long-range order (LRO) [21, 22] and the Tomonaga-Luttinger liquid (TLL) [23, 24], respectively. Theoretically, it is well established that the TLL-like gapless quantum spin liquid (QSL) is robust for a wide parameter range of 0<J′/J≲0.60 < J'/J \lesssim0.6 [4, 5, 6]. This is a manifestation of the so-called dimensional reduction caused by geometrical frustration [7]. On the other hand, the ground state for 0.6≲J′/J≲0.80.6 \lesssim J'/J \lesssim0.8 is controversial; some theories suggest a direct transition from the TLL-like QSL to the spiral LRO [4, 8, 9, 10, 11], while others suggest the appearance of another phase, such as a dimer-ordered state [12] and a gapped QSL [5, 6].

Experimentally, two types of compounds have been intensively studied as candidates for the ATL quantum antiferromagnet: Cs2CuX4\mathrm{Cs}_{2}\mathrm{Cu}X_{4} (X=Cl,BrX = \mathrm{Cl}, \mathrm{Br}) [25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39] and organic Mott insulators represented by κ\kappa-(BEDT-TTF)2Cu2(CN)3_{2}\mathrm{Cu}_{2}(\mathrm{CN})_{3} [40, 41, 42, 43, 44, 45, 46, 47]. The magnetic properties of Cs2CuX4\mathrm{Cs}_{2}\mathrm{Cu}X_{4} have been unveiled in detail thanks to relatively weak AFM exchange couplings: J/kB=4.3J/k_{\mathrm{B}} = 4.3 K and J′/J=0.30J'/J = 0.30 (J/kB=14.9J/k_{\mathrm{B}} = 14.9 K and J′/J=0.41J'/J = 0.41) for X=ClX = \mathrm{Cl} (Br\mathrm{Br}) [33], where kBk_{\mathrm{B}} is the Boltzmann constant. However, Cs2CuX4\mathrm{Cs}_{2}\mathrm{Cu}X_{4} undergoes a magnetic transition into an incommensurate spiral LRO at low temperatures [25, 36] due to the presence of the Dzyaloshinskii-Moriya (DM) interaction [16], which interfere with the observation of essential properties expected for the ideal ATL quantum antiferromagnet. For the organic Mott insulators, the exchange couplings are much stronger, on the order of several hundred kelvin [19, 40, 41, 42, 43], making the accurate estimation of J′/JJ'/J elusive. The contribution of multi-spin exchange and spin-lattice coupling can be important in the organic salts [19], as suggested from a pressure-induced insulator-metal transition [44, 45] and a magnetic-field-induced AFM transition reminiscent of a spin-Peierls phase [46], respectively.

Compared to the aforementioned compounds, recently-discovered rhenium oxychlorides A3ReO5Cl2A_{3}\mathrm{ReO}_{5}\mathrm{Cl}_{2} (A=Ca,Sr,BaA = \mathrm{Ca}, \mathrm{Sr}, \mathrm{Ba}) [48, 49, 50, 51, 52, 53] potentially hold an advantage in the search for low-temperature magnetism of the ATL quantum antiferromagnet. The Re6+\mathrm{Re}^{6+} ion has a 5d15d^{1} electronic configuration, and the unique crystal field resulting from the surrounding mixed anions causes the dxyd_{xy} orbital to become the lowest-energy level without degeneracy. As a consequence, A3ReO5Cl2A_{3}\mathrm{ReO}_{5}\mathrm{Cl}_{2} can be regarded as a spin-1/2 Heisenberg system with quenched orbital angular momentum. J′/JJ'/J is estimated to be 0.32, 0.43, and 0.47 for A=CaA = \mathrm{Ca}, Sr\mathrm{Sr}, and Ba\mathrm{Ba}, respectively, based on theoretical fits to the temperature dependence of magnetic susceptibility using high-temperature series expansion [49, 50]. A3ReO5Cl2A_{3}\mathrm{ReO}_{5}\mathrm{Cl}_{2} exhibits TLL-like characteristics indicative of one-dimensionality at low temperatures, as evidenced by several experimental observations such as the Bonner–Fisher-like magnetic susceptibility [54], large TT-linear magnetic heat capacity [49, 50], and spinon and triplon excitations in inelastic neutron scattering [51] and Raman spectra [53]. The frustration factor, defined as f≡∣ΘW∣/TNf \equiv|\Theta_\mathrm{W}|/T_\mathrm{N}, where ΘW\Theta_\mathrm{W} and TNT_\mathrm{N} represent a Weiss temperature and an ordering temperature, respectively, is larger in A3ReO5Cl2A_3ReO_5Cl_2 than in Cs2CuX4Cs_2CuX_4 [49, 50]. This suggests that additional interactions, aside from JJ and J′J', play a less significant role in A3ReO5Cl2A_3ReO_5Cl_2 than in Cs2CuX4Cs_2CuX_4. Further development of chemical substitutes for A3ReO5Cl2A_3ReO_5Cl_2, along with the evaluation of their effective spin models, should open up the route for a systematic understanding of the fundamental properties of the ATL quantum antiferromagnet.

Here, we report physical properties of four novel Re-based quantum ATL antiferromagnets, A3ReO5Br2A_3ReO_5Br_2 (A=CaA = Ca, SrSr, BaBa) and Pb3ReO5Cl2Pb_3ReO_5Cl_2. These compounds are isostructural with A3ReO5Cl2A_3ReO_5Cl_2 (Figure 1a–Figure 1c) and exhibit one-dimensional character at low temperatures as well [48, 49, 50]. We also investigate magnetization processes of all the seven A3ReO5X2A_3ReO_5X_2 families in pulsed high magnetic fields of up to 130 T [55]. On the basis of theoretical calculations using the orthogonalized finite-temperature Lanczos method (OFTLM) [18, 56], J′/JJ'/J is found to vary between 0.25 and 0.45 from compound to compound. We propose that, considering the difference in their crystal structures and spatial distributions of electronic density, A3ReO5X2A_3ReO_5X_2 can be categorized into three groups in terms of an effective spin Hamiltonian (Figure 1d): (1) A=CaA = Ca with the DM interaction, (2) A=PbA = Pb with the relatively strong interlayer exchange coupling, and (3) A=SrA = Sr and BaBa with no dominant perturbations.

Structural and magnetic properties of A3ReO5X2

Figure 1. Structural and magnetic properties of A3ReO5X2A_3\mathrm{ReO}_5X_2. a Two kinds of crystal structures. Ca3ReO5X2Ca_3ReO_5X_2 crystallizes in Type I with the space group PnmaPnma (left), whereas the system with A=SrA = Sr, BaBa, and PbPb crystallizes in Type II with the space group CmcmCmcm (right). The illustrations are drawn with VESTA software [57]. b Schematic of an anisotropic triangular lattice (ATL), where upward and downward ReO5ReO_5 square pyramids running along the bb (aa) direction are alternately arranged along the cc direction in the Type I (II) structure. Exchange couplings between nearest-neighbor and next-nearest-neighbor Re ions with the bond length dd and d′d' are defined as JJ and J′J', respectively. c Plots of dd and d′d' for the seven compounds. Elements of AA and XX are written next to each marker. The data of A3ReO5Cl2A_3ReO_5Cl_2 (A=CaA = Ca, SrSr, BaBa) are taken from Refs. [48, 50]. d Summary of effective spin Hamiltonians with the ATL quantum Heisenberg model. The anisotropy J′/JJ'/J for each compound is revealed in the present work and displayed in the left axis. The DM interaction and interlayer exchange coupling dominate the low-temperature magnetism for A=CaA = Ca and PbPb, respectively, whereas there are no significant effects of such perturbations for A=SrA = Sr and BaBa.

Results

Crystal structures

We first present the crystal structures of newly synthesized compounds: Ca3ReO5Br2Ca_3ReO_5Br_2 (CROB), Sr3ReO5Br2Sr_3ReO_5Br_2 (SROB), Ba3ReO5Br2Ba_3ReO_5Br_2 (BROB), and Pb3ReO5Cl2Pb_3ReO_5Cl_2 (PROC). Regarding the previously reported oxychlorides, Ca3ReO5Cl2Ca_3ReO_5Cl_2 (CROC) crystallizes in an orthorhombic structure with the space group PnmaPnma (No. 62) [48], whereas Sr3ReO5Cl2Sr_3ReO_5Cl_2 (SROC) and Ba3ReO5Cl2Ba_3ReO_5Cl_2 (BROC) with the space group CmcmCmcm (No. 63) [50], as depicted in Figure 1a. Hereafter, we refer to these two structures as Type I and Type II, respectively. Both structures consist of alternating layers of A3ReO5A_3ReO_5 and Cl slabs. The coordination numbers of the A2+A^{2+} ions are 5 or 6 in the Type I structure, whereas they are 8 or 9 in the Type II structure. This difference arises due to the smaller ionic radius of Ca2+Ca^{2+} (1.00 Å in the 6-fold coordination) compared to Sr2+Sr^{2+} and Ba2+Ba^{2+} (1.26 and 1.42 Å, respectively, in the 8-fold coordination). We find that the substitution of Cl−Cl^- (ionic radius 1.81 Å) with Br−Br^- (ionic radius 1.96 Å) does not impact the original crystal structure; CROB crystallizes in Type-I, while SROB and BROB crystallize in Type II. PROC also crystallizes in TypeII (the ionic radius of Pb2+^{2+} is 1.29 Å in the 8-fold coordination). Further details on the crystal structures are provided in Supplementary Note 2.

An ATL composed of magnetic Re6+^{6+} ions is depicted in Figure 1b. The ReO5_5 square pyramids, which point upward and downward alternately along the interchain direction, do not share any oxygen ions. Consequently, each Re6+^{6+} ion is well separated from one another by more than 5 Å. Figure 1c compares the lengths of the nearest-neighbor and next-nearest-neighbor Re–Re bonds, dd and d′d', respectively, among all the seven A3_3ReO5_5X2_2 compounds. In common with A=A= Ca, Sr, and Ba, the substitution of Cl with Br leads to the elongation of both dd and d′d'; the increase rates are 1.4% and 0.7% for A=A= Ca, 1.2% and 1.7% for A=A= Sr, and 1.4% and 1.5% for A=A= Ba, respectively.

One dimensionality of the low-temperature magnetism

Next, we show the basic physical properties of A3_3ReO5_5Br2_2 (A=A= Ca, Sr, Ba) and PROC. Figure 2a–Figure 2d shows the temperature dependence of magnetization divided by a magnetic field M/HM/H measured at 7 T. All the M/HM/H data above 150 K are well fitted by the Curie–Weiss law M/H=C0/(T−ΘW)+χ0M/H=C_0/(T-\Theta_{\mathrm{W}})+\chi_0, where C0C_0 is the Curie constant, and χ0\chi_0 is the temperature-independent term, as shown by black lines in Figure 2a–Figure 2d. The fitting parameters as well as the effective magnetic moment μeff\mu_{\mathrm{eff}} calculated from C0C_0 are listed in Table I. The obtained χ0\chi_0 is comparable to the diamagnetic contribution of core electrons χdia\chi_{\mathrm{dia}} for each compound: χdia=−1.76×10−4\chi_{\mathrm{dia}}=-1.76\times10^{-4}, −2.02×10−4-2.02\times10^{-4}, −2.24×10−4-2.24\times10^{-4} and −2.18×10−4 cm3 mol−1-2.18\times10^{-4}\,\mathrm{cm}^3\,\mathrm{mol}^{-1} for CROB, SROB, BROB, and PROC, respectively. The reduction of μeff\mu_{\mathrm{eff}} from 1.73 μB\mu_{\mathrm{B}} expected for the pure S=1/2S=1/2 case would be attributed to the effect of the spin-orbit interaction, which cancels the spin and orbital angular momentum with each other in the 5d15d^1 electronic configuration.

Figure 2. Basic physical properties of four novel compounds. a–d Temperature dependence of the magnetic susceptibility M/HM/H measured at 7 T (left axis) and the inverse susceptibility (M/H−χ0)−1(M/H-\chi_0)^{-1} (right axis), where a TT-independent component χ0\chi_0 is subtracted, for CROB (a), SROB (b), BROB (c), and PROC (d). The black lines represent the Curie–Weiss law, i.e., M/H=C/(T−ΘW)+χ0M/H=C/(T-\Theta_{\mathrm{W}})+\chi_0, fitted to the experimental data between 150 and 400 K. e–h Heat capacity divided by temperature C/TC/T as a function of TT below 100 K (bottom axis) and as a function of T2T^2 below 10 K (top axis) for CROB (e), SROB (f), BROB (g), and PROC (h). The black lines represent fittings C/T=α+βT2C/T=\alpha+\beta T^2 to the experimental data between 2 and 10 K for CROB and SROB, between 2 and 4 K for BROB, and between 7 and 10 K for PROC.

CROCSROCBROCPROC
μeff\mu_{\mathrm{eff}} (μB\mu_{\mathrm{B}})1.538(2)1.648(1)1.480(3)1.493(1)
ΘW\Theta_{\mathrm{W}} (K)−37.8(1)-37.8(1)−49.5(1)-49.5(1)−21.6(1)-21.6(1)−47.8(2)-47.8(2)
χ0\chi_{0} (10−4 cm3 mol−110^{-4}\ \mathrm{cm}^{3}\ \mathrm{mol}^{-1})−1.361(5)-1.361(5)−2.219(5)-2.219(5)−1.931(7)-1.931(7)−2.164(7)-2.164(7)
α\alpha (mJ K−2 mol−1\mathrm{mJ}\ \mathrm{K}^{-2}\ \mathrm{mol}^{-1})114.7(4)93.3(5)246.7(4)109(2)
β\beta (mJ K−4 mol−1\mathrm{mJ}\ \mathrm{K}^{-4}\ \mathrm{mol}^{-1})1.64(1)2.21(2)6.11(3)4.93(3)
CROBSROBBROB
μeff\mu_{\mathrm{eff}} (μB\mu_{\mathrm{B}})1.609(2)1.524(2)1.471(1)
ΘW\Theta_{\mathrm{W}} (K)−36.5(2)-36.5(2)−37.4(4)-37.4(4)−17.2(2)-17.2(2)
χ0\chi_{0} (10−4 cm3 mol−110^{-4}\ \mathrm{cm}^{3}\ \mathrm{mol}^{-1})−1.740(8)-1.740(8)−1.936(16)-1.936(16)−2.188(6)-2.188(6)
α\alpha (mJ K−2 mol−1\mathrm{mJ}\ \mathrm{K}^{-2}\ \mathrm{mol}^{-1})126.2(7)99.0(7)373(9)
β\beta (mJ K−4 mol−1\mathrm{mJ}\ \mathrm{K}^{-4}\ \mathrm{mol}^{-1})1.92(2)2.89(2)8.65(9)

Table I. Magnetic and thermal parameters for A3_3ReO5_5X2_2 (A = Ca, Sr, Ba, Pb; X = Cl, Br). Data for CROC are taken from Ref. [49], and those for SROC and BROC are taken from Ref. [50]. See text for details of the definition of each parameter.

The large negative Weiss temperatures indicate the dominance of AFM exchange interactions in these compounds. The smaller value of ∣ΘW∣|\Theta_{\mathrm{W}}| for BROB (ΘW=−17.2(2)\Theta_{\mathrm{W}}=-17.2(2) K) compared to CROB (−36.5(2)-36.5(2) K) and SROB (−37.4(2)-37.4(2) K) can be attributed to the relatively long Re–Re bond lengths, dd and d′d', in BROB (Figure 1c). This trend is common with the A3_3ReO5_5Cl2_2 case: ΘW=−37.8(1)\Theta_{\mathrm{W}}=-37.8(1), −49.5(1)-49.5(1) and −21.6(1)-21.6(1) K for CROC, SROC, and BROC, respectively [49, 50]. Furthermore, the substitution of Cl with Br leads to a decrease in ∣ΘW∣|\Theta_{\mathrm{W}}|; the decrease rates are 3.4%, 24.4%, and 20.4% for A=A= Ca, Sr, and Ba, respectively. The much smaller decrease rate for A=A= Ca compared to A=A= Sr and Ba would be due to the difference in their crystal structures, as mentioned above (Figure 1a).

On further cooling below 100 K, all the M/HM/H–TT curves gradually deviate from the Curie–Weiss behavior and exhibit a broad peak around 10–20 K (Figure 2a–Figure 2d), signaling the development of AFM short-range correlations. In the ATL Heisenberg antiferromagnet with moderate anisotropy J′/JJ'/J, a dimensional crossover from 2D to 1D, i.e., the so-called one-dimensionalization, is expected at low temperatures, where the interchain zigzag couplings J′J' are effectively cancelled out due to geometrical frustration, and consequently a TLL-like disordered state would be realized [49, 50, 51, 52, 53].

Figure 2e–Figure 2h shows the temperature dependence of heat capacity divided by temperature C/TC/T for each compound. We also display C/TC/T versus T2T^{2} plots in the top axes. For A3_3ReO5_5Br2_2, C/TC/T exhibits a linear T2T^{2} dependence at low temperatures with a finite intercept, the origin of which would be gapless-spin excitation expected in the TLL-like disordered state. Note that a hump structure is observed below 10 K in the C/TC/T–TT curve for BROB (Figure 2g). This would reflect the entropy release caused by the development of magnetic short-range correlations rather than a magnetic LRO, as suggested from a broad peak around 8 K in the magnetic susceptibility data (Figure 2c). By fitting the C/TC/T–T2T^{2} curve to the equation

C/T=α+βT2,C/T = \alpha+ \beta T^{2},

where the first and second terms originate from magnetic and phonon contributions, respectively, in a temperature range between 2 and 10 K for CROB and SROB and between 2 and 4 K for BROB, we obtain the values α\alpha and β\beta for each compound as summarized in Table I (Note that the estimation of β\beta for BROC has an uncertainty of approximately ±30%\pm30\% depending on the fitting range). In the 1D chain limit (J′=0J^{\prime}=0), the TT-linear contribution to the heat capacity is given by 2RT/3J2RT/3J [54], where RR is the gas constant, so that α\alpha is inversely related to JJ. Indeed, α\alpha in BROB is much larger than those in CROB and SROB. This is compatible with the much smaller value of ∣ΘW∣|\Theta_{\mathrm{W}}| in BROB. Furthermore, we find that α\alpha in A3_3ReO5_5Br2_2 is a bit larger than that in the oxychloride counterpart A3_3ReO5_5Cl2_2 [49, 50] (Table I). This tendency also agrees with the decrease of ∣ΘW∣|\Theta_{\mathrm{W}}| by the substitution of Cl with Br. In contrast, C/TC/T approaches zero toward the low-temperature limit for PROC. This is owing to the occurrence of a magnetic LRO at TN=6.6T_{\mathrm{N}} = 6.6 K, as discussed below. Accordingly, we perform a linear fit for the C/TC/T–T2T^{2} curve of PROC between 7 and 10 K (Table I). The magnitudes of α\alpha and ∣ΘW∣|\Theta_{\mathrm{W}}| in PROC are both intermediate between those in CROC and SROC.

Magnetic phase transition in Ca3_3ReO5_5Br2_2 and Pb3_3ReO5_5Cl2_2

In real spin systems, deviations from an ideal spin model more or less exist. The ground state of the ATL quantum antiferromagnet is highly sensitive to weak perturbations such as the DM interaction and interlayer exchange couplings [16].

For instance, Cs2_2CuCl4_4 undergoes a magnetic transition into a spiral LRO phase with a magnetic propagation vector of q=(0,0.475,0)\boldsymbol{q} = (0, 0.475, 0) [25], which is attributed to a combined effect of magnetic frustration and the DM interaction [31]. As for A3_3ReO5_5X2_2, the DM interaction is prohibited for A = Sr, Ba, and Pb with the Type II structure (Supplementary Note 3), whereas this is not the case for A = Ca with the Type I structure. Indeed, a neutron scattering experimentiferromagnet is highly sensitive to weak perturbations such as the DM interaction and interlayer exchange couplings [16]. For instance, Cs2_2CuCl4_4 undergoes a magnetic transition into a spiral LRO phase with a magnetic propagation vector of q=(0,0.475,0)\boldsymbol{q} = (0, 0.475, 0) [25], which is attributed to a combined effect of magnetic frustration and the DM interaction [31]. As for A3_3ReO5_5X2_2, the DM interaction is prohibited for A = Sr, Ba, and Pb with the Type II structure (Supplementary Note 3), whereas this is not the case for A = Ca with the Type I structure. Indeed, a neutron scattering experiment on CROC confirmed the appearance of a spiral LRO with q=(0,0.465,0)q = (0, 0.465, 0) below TN=1.13T_{\mathrm{N}} = 1.13 K [49, 51]. Therefore, a similar magnetic transition is expected for CROB at lower temperatures, though no anomaly is observed in the magnetic susceptibility and heat capacity data above 2 K for CROB as well as SROB and BROB (Figure 2). Figure 3a shows the temperature dependence of C/TC/T measured down to 0.24 K for CROB. A clear lambda-shaped peak is observed at TN=1.15T_{\mathrm{N}} = 1.15 K, indicating a second-order transition. Figure 3b illustrates the DM vectors in one ATL layer allowed for the Type I structure (space group PnmaPnma) [16]. As discussed in Ref. [16], the cc component of the DM vector on the intrachain bond DcD_c and the aa component of the DM vector on the diagonal bond Da′D'_a are most relevant in the ground state selection. For Cs2_2CuCl4_4, it is proposed that the contribution of Da′D'_a is significant [27], resulting in the spiral moment lying almost within the bcbc plane [25]. On the other hand, a recent electron-spin-resonance (ESR) study suggested that the intrachain DM interaction DcD_c is the primary source for stabilizing the spiral LRO in CROC, as indicated by the observed shift of the soft mode in the disordered phase [52]. Despite the slightly weaker exchange interactions in CROB than in CROC, they exhibit nearly the same transition temperature, suggesting that the intrachain DM interaction is slightly stronger in CROB.

Magnetic transition and its microscopic origin in CROB and PROC

Figure 3. Magnetic transition and its microscopic origin in CROB and PROC. a Temperature dependence of the heat capacity divided by temperature C/TC/T measured at 0 T using 3^{3}He for CROB. b Schematic of symmetrically-allowed DM vectors in one ATL layer in CROB [16]. DiD_i (Di′D_i^{\prime}) (i=a,b,ci = a, b, c) indicates the ii component of the DM vector on the JJ (J′J^{\prime}) bond. In the neighboring ATL layer, all the DM components except for DcD_c orient in the opposite direction. c, d Temperature dependence of the magnetic susceptibility M/HM/H measured at various magnetic fields (c) and the heat capacity divided by temperature C/TC/T at 0 and 7 T (d) for PROC. The inset of panel c shows a magnetization curve at 2 K measured in a static magnetic field. The gray dashed line is a guide for the eye to clarify the slope change of the magnetization curve. e Schematic of the Wannier orbitals around the Pb/Sr and Re atoms for PROC (top) and BROC (bottom) revealed by the first-principles calculations. The paths of intralayer exchange J′J^{\prime} and interlayer exchange J′′J^{\prime\prime} and J′′′J^{\prime\prime\prime} couplings are depicted by red arrows.

In addition, we observe a magnetic transition at TN=6.6T_{\mathrm{N}} = 6.6 K for PROC. As shown in Figure 3c, the M/HM/H–TT curve at 0.1 T exhibits a sudden drop below TNT_{\mathrm{N}}, indicating an AFM phase transition. M/HM/H below TNT_{\mathrm{N}} gradually rises with increasing an applied magnetic field up to 7 T, suggesting a spin-flop transition (see also the MM–HH curve at 2 K shown in the inset of Figure 3c). The C/TC/T–TT curve exhibits a broad lambda-shaped anomaly indicative of a second-order transition, as shown in Figure 3d. This anomaly persists at almost the same temperature until 7 T, indicating that the AFM or field-induced spin-canted phases are robust against a higher magnetic field. The magnetic contribution of CC is found to be proportional to T3T^3 below TNT_{\mathrm{N}} (Supplementary Note 4), suggesting the presence of 3D magnetic correlations, i.e., interlayer exchange couplings. This is confirmed by the first-principles calculations: the strengths of interlayer exchange couplings, J′′J'' and J′′′J''', defined as shown in Figure 3e, are estimated to J′′/J=0.006J''/J = 0.006 and J′′′/J=0.190J'''/J = 0.190 for PROC, whereas both J′′/JJ''/J and J′′′/JJ'''/J are much smaller than 0.01 for other compounds (for details, see Supplementary Note 5). The relatively strong J′′′J''' coupling can be attributed to the electron hopping via Pb 6p6p orbital.

High-field magnetization processes

Here, we move on to the in-field properties of all the seven A3_3ReO5_5X2_2 families, which are useful for estimating the strengths of JJ and J′J'. Figure 4 summarizes the magnetization data of BROC and BROB. The magnetic susceptibility data for BROC (Figure 4a) are taken from Ref. [50], and those for BROB (Figure 4b) are identical to Figure 2c. The MM–HH curves in Figure 4c–Figure 4f are quasi-isothermal processes obtained using a non-destructive pulsed magnet. The saturation field at 1.3 K is μ0Hsat=55(1)\mu_0H_{\mathrm{sat}} = 55(1) T for BROC and 42(1) T for BROB. The MM–HH curves exhibit a concave shape followed by an inflection point immediately below HsatH_{\mathrm{sat}}, which is a typical behavior of low-dimensional quantum magnets.

Figure 4. Comparison of magnetization data of Ba3_3ReO5X2_5X_2 with theoretical calculations based on the OFTLM. Panels a and b show the temperature dependence of the magnetic susceptibility measured at 7 T [50], where a TT-independent component χ0\chi_0 is subtracted. Panels c–f show quasi-isothermal magnetization curves in the field-increasing process measured at 4.2 K (c, d) and 1.3 K (e, f) using a non-destructive pulsed magnet. Here, the field derivative of the magnetization dM/d(μ0H)dM/d(\mu_0H) is displayed in the right axis. Experimental curves are shown by cyan lines, and theoretical curves fitted to each experimental data are shown by black lines. Theoretical MM–HH curves represent isothermal curves at temperatures corresponding to the experimental ones. The fitting parameters for each compound are summarized in Table II.

The magnetic susceptibility data for CROC (Figure 5a) and SROC (Figure 5c) are taken from Refs. [49] and [50], respectively. Those for CROB (Figure 5b), SROB (Figure 5d), and PROC (Figure 5e) are identical to Figure 2a, Figure 2b, and Figure 2d, respectively. The MM–HH curves in Figure 5f–Figure 5j are quasi-adiabatic processes with the initial temperature of Tini=5T_{\mathrm{ini}} = 5 K obtained using a single-turn coil system, i.e., a semi-destructive pulsed magnet. The absolute values of MM are calibrated using the MM–HH curves at 4.2 K obtained in a non-destructive pulsed magnet (Supplementary Note 6). For all the MM–HH curves, an inflection point indicative of the nearly saturation is observed at μ0Hsat=83(3)\mu_0H_{\mathrm{sat}} = 83(3) T for CROC, 78(3) T for CROB, 95(5) T for SROC, 80(5) T for SROB, and 115(5) T for PROC.

Figure 5. Comparison of magnetization data of A3ReO5X2A_{3}\mathrm{ReO}_{5}X_{2} (AA = Ca, Sr, Pb) with theoretical calculations based on the OFTLM. Panels a–e show the temperature dependence of the magnetic susceptibility measured at 7 T [49, 50], where a TT-independent component χ0\chi_{0} is subtracted. Panels f–j show quasi-isentropic magnetization curves in the field-increasing (solid lines) and field-decreasing (dashed lines) processes measured at the initial temperature of Tini=5T_{\mathrm{ini}} = 5 K using a single-turn coil megagauss generator. The field derivative of the magnetization dM/d(μ0H)\mathrm{d}M/\mathrm{d}(\mu_{0}H) in the field-decreasing process is displayed in the right axis. Experimental curves are shown by colored lines, and theoretical curves fitted to each experimental data are shown by black lines. Theoretical MM–HH curves represent isentropic curves for Tini=5T_{\mathrm{ini}} = 5 K. The fitting parameters for each compound are summarized in Table II.

Comparison with theoretical calculations based on the OFTLM

The accurate estimation of JJ and J′J^{\prime} in the quantum ATL antiferromagnet is generally challenging. One might consider achieving this by relating the Weiss temperature and the saturation field to the exchange interactions using the equations

ΘW=−(J/2+J′)andgμBHsat=2J(1+J′/2J)2,\Theta_{\mathrm{W}} = -(J/2 + J^{\prime}) \quad\text{and} \quad g\mu_{\mathrm{B}}H_{\mathrm{sat}} = 2J(1 + J^{\prime}/2J)^2,

respectively [19, 33]. However, the experimentally-determined ΘW\Theta_{\mathrm{W}} tends to underestimate the actual value, especially for small J′/JJ^{\prime}/J, even when performing the Curie-Weiss fit to the magnetic susceptibility data in a high-temperature range [19]. Besides, HsatH_{\mathrm{sat}} is also highly temperature-dependent.

Recently, the OFTLM was developed [56] and applied to the calculation of the magnetic susceptibility and magnetization process in the quantum ATL antiferromagnet [18]. This technique has an advantage over the standard FTLM [58] in accurately evaluating the magnetic entropy at low temperatures, which is crucial for the calculation of the adiabatic magnetization process [18]. Here, we adopt the OFTLM to the Hamiltonian of the quantum ATL antiferromagnet in a magnetic field,

H=J∑⟨i,j⟩Si⋅Sj+J′∑⟨i,k⟩Si⋅Sk−h∑iSiz,(1)\mathcal{H} = J \sum_{\langle i,j\rangle} \mathbf{S}_{i}\cdot\mathbf{S}_{j} + J^{\prime} \sum_{\langle i,k\rangle} \mathbf{S}_{i}\cdot\mathbf{S}_{k} - h \sum_i S_i^z, \tag*{(1)}

where Si\mathbf{S}_{i} denotes a spin-1/2 operator localized on the ATL, ⟨i,j⟩\langle i,j\rangle and ⟨i,k⟩\langle i,k\rangle indicate pairs of neighboring sites along the intrachain and interchain directions, respectively, and hh is a magnetic field applied along the zz axis. Further details of the calculation are found in the Method section. Importantly, the peak temperature of the magnetic susceptibility TpT_{\mathrm{p}} normalized to JJ shifts to a higher temperature side as J′/JJ^{\prime}/J decreases [18, 19]. Thanks to this feature, the strengths of JJ and J′J^{\prime} can be uniquely determined once we reveal experimental values of the Landé gg-factor, TpT_{\mathrm{p}}, and HsatH_{\mathrm{sat}}, which are listed in Table II. Here, we set the gg-value using a relation μeff=gS(S+1)\mu_{\mathrm{eff}} = g\sqrt{S(S+1)} with S=1/2S = 1/2, where μeff\mu_{\mathrm{eff}} is estimated from the Curie-Weiss fit (Table I).

CROCSROCBROCPROC
TpT_{\mathrm{p}} (K)27291229
μ0Hsat\mu_{0}H_{\mathrm{sat}} (T)83(3)95(5)55(1)115(5)
gg1.7761.9031.7091.724
J/kBJ/k_{\mathrm{B}} (K)42472046.5
J′/JJ^{\prime}/J0.250.350.400.35
CROBSROBBROB
TpT_{\mathrm{p}} (K)24258.0
μ0Hsat\mu_{0}H_{\mathrm{sat}} (T)78(3)80(5)42(1)
gg1.8581.7601.699
J/kBJ/k_{\mathrm{B}} (K)39.54014
J′/JJ^{\prime}/J0.250.250.45

Table II. Experimental values of the peak temperature of the magnetic susceptibility TpT_{\mathrm{p}}, the saturation magnetic field HsatH_{\mathrm{sat}}, and the Landé gg factor as well as the exchange couplings JJ and J′J^{\prime} estimated based on theoretical calculations using the orthogonalized finite-temperature Lanczos method in the seven A3_{3}ReO5_{5}X2_{2} compounds.

For BROC and BROB, all the magnetization data can be simultaneously reproduced by theoretical curves with a single parameter set, as shown in Figure 4. J/kB=20J/k_{\mathrm{B}} = 20 K and J′/J=0.40J^{\prime}/J = 0.40 for BROC, and J/kB=14J/k_{\mathrm{B}} = 14 K and J′/J=0.45J^{\prime}/J = 0.45 for BROB. J′/JJ^{\prime}/J for BROB is almost the same with that for BROC. This is reasonable given that the increase rates of dd and d′d^{\prime} by the substitution of Cl with Br are almost the same. No signatures of a 1/3-magnetization plateau are observed for both compounds in contrast to the theoretical MM–HH curves at 1.3 K (Note that fluctuations in the theoretical curves on the low-field side below the 1/3-magnetization plateau are caused by finite-size effects). We infer that the appearance of the 1/3-magnetization plateau is blurred by averaging over the MM–HH curves on randomly-oriented polycrystals in the present experiments using polycrystalline powder samples. Indeed, a previous ESR experiment on a BROC single crystal revealed a weak anisotropy: g=1.88g = 1.88, 1.85 and 1.92 for H∥aH \parallel a, bb, and cc, respectively [51].

Similarly, we calculate magnetic susceptibility and adiabatic magnetization curves for the rest five compounds, as shown in Figure 5. The exchange parameter set which can reproduce the M/HM/H–TT curve as well as the saturation field HsatH_{\mathrm{sat}} is obtained for each compound as follows: J/kB=42J/k_{\mathrm{B}} = 42 K and J′/J=0.25J^{\prime}/J = 0.25 for CROC, J/kB=39.5J/k_{\mathrm{B}} = 39.5 K and J′/J=0.25J^{\prime}/J = 0.25 for CROB, J/kBJ/k_{\mathrm{B}} = 47 K and J′/JJ^{\prime}/J = 0.35 for SROC, J/kBJ/k_{\mathrm{B}} = 40 K and J′/JJ^{\prime}/J = 0.25 for SROB, and J/kBJ/k_{\mathrm{B}} = 46.5 K and J′/JJ^{\prime}/J = 0.35 for PROC. The deviation between the experimental MM–HH curves and theoretical ones in the high-field region would be mainly due to the imperfect background subtraction in the experimental data (Note that the detection sensitivity of the magnetization pickup coil is significantly reduced near the maximum field). Given that the error in the experimental HsatH_{\mathrm{sat}} is approximately ±5\pm5 T, the error in the estimated value of J′/JJ^{\prime}/J is at most ±0.05\pm0.05.

Discussion

We demonstrate that all three previously reported compounds A3_{3}ReO5_{5}Cl2_{2} (A = Ca, Sr, Ba) [48, 50] can accommodate the substitution of Cl with Br. The halide substitution results in the elongation of both dd and d′d^{\prime} in the ATL composed of Re6+^{6+} ions (Figure 1c). These elongations lead to a weakening of the exchange couplings JJ and J′J^{\prime}, as indicated by the decrease in ∣ΘW∣|\Theta_{\mathrm{W}}| as well as the increase in α\alpha (∝J−1\propto J^{-1}) (Table I). Notably, this tendency is in contrast with another model compound of an ATL quantum antiferromagnet Cs2_{2}CuCl4_{4}, where the substitution of Cl with Br leads to the enhancement of JJ and J′J^{\prime} [33]. We also reveal that the anisotropy J′/JJ^{\prime}/J in A3_{3}ReO5_{5}X2_{2} ranges from 0.25 to 0.45, although the low-temperature magnetism is commonly characterized by one-dimensionalization irrespective of the types of the A cation and X anion. The halide substitution does not significantly modify J′/JJ^{\prime}/J for A = Ca and Ba, while a substantial decrease in J′/JJ^{\prime}/J (from 0.35 to 0.25) is found for A = Sr (Table II). This can be attributed to a relatively larger increase in d′d^{\prime} (∼1.7%\sim1.7\%) compared to dd (∼1.2%\sim1.2\%) for A = Sr (Figure 1c).

The most fascinating aspect of A3_{3}ReO5_{5}X2_{2} is the controllability of perturbative interactions, which is responsible for selecting the ground state. CROC and CROB undergo a magnetic transition at TNT_{\mathrm{N}} = 1.13 [49] and 1.15 K (Figure 3a), respectively, due to the uniform DM interaction allowed by the space group of the crystal lattice (PnmaP n m a). The identification of a magnetic structure in CROB, as well as in CROC, is an important issue to be addressed in future works. Theoretically, the application of a magnetic field on the ATL quantum antiferromagnet with the DM interaction can induce successive phase transitions associated with the emergence of versatile LRO phases, such as commensurate AFM and incommensurate cone phases, aside from the low-field spiral and forced ferromagnetic phases [9, 16]. Indeed, Cs2_{2}CuCl4_{4} exhibits multiple phase transitions for H∥bH \parallel b and cc [30], and interestingly, the phase diagram becomes more complicated under pressure [34]. However, the strength of the interlayer exchange (J′′/JJ^{\prime\prime}/J = 0.045) is comparable to that of the DM interaction (D/JD/J = 0.05) in Cs2_{2}CuCl4_{4} [27], making assignments of several in-field phases somehow puzzling. For CROC, the interlayer exchange couplings (J′′/J≈0.0007J^{\prime\prime}/J \approx0.0007) [49] is much weaker than the DM interaction (D/J≈0.23D/J \approx0.23) [52], so that the role of J′′J^{\prime\prime} could be safely neglected. Detailed magnetization measurements on CROC and CROB single crystals below TNT_{\mathrm{N}} will explicitly elucidate the effect of the DM interaction on the field-induced phases of the ATL quantum antiferromagnet. Furthermore, PROC undergoes a magnetic transition at TNT_{\mathrm{N}} = 6.6 K due to the interlayer exchange couplings via the Re–O–Pb–O–Re superexchange path, which arises from the hybridization of the O 2p2p and Pb 6p6p orbitals. The exchange paths of the interlayer J′′J^{\prime\prime} and J′′′J^{\prime\prime\prime} couplings in PROC (Figure 3e) are more complex than those considered in the previous theoretical work [16]. It is necessary to construct a new theoretical model in line with the spin Hamiltonian of PROC. The rest four A3_{3}ReO5_{5}X2_{2} compounds with A = Sr or Ba are close to the ideal ATL quantum antiferromagnet free from the interlayer exchange and DM interactions. Since J′/JJ^{\prime}/J for these compounds is smaller than 0.6, the ground states are expected to be a TLL-like gapless QSL, in accordance with previous theories [4, 5, 6]. The diversity in J′/JJ^{\prime}/J values (0.25–0.45) offers an excellent opportunity to meticulously investigate the impact of magnetic frustration on the TLL state. Investigating the magnetic excitation as well as the dynamical magnetic property at extremely low temperatures would also be promising for understanding the essential nature of the ground state of the ATL quantum antiferromagnet.

Method

Sample preparation

Polycrystalline samples of A3_{3}ReO5_{5}X2_{2} (A = Ca, Sr, Ba, Pb; X = Cl, Br) and an isostructural nonmagnetic compound Pb3_{3}WO5_{5}Cl2_{2} were synthesized by conventional solid-state reaction. AO, AX2_{2}, and ReO3_{3} (WO3_{3}) were mixed in a molar ratio of 2:1:1 in an argon-filled glovebox. The mixture was pressed into a pellet and wrapped in gold foil, and then sealed in an evacuated quartz tube. All the samples were sintered twice with an intermediate grinding step. The sintering conditions for each compound are listed in Supplementary Table S1. Single crystals of A3_{3}ReO5_{5}Br2_{2} (A = Ca, Sr, Ba) with a size of approximately 50 μm\mu\mathrm{m} were grown by partially melting a stoichiometric mixture in an evacuated quartz tube at 1050, 1030, 900°C for 20h for A=CaA = \mathrm{Ca}, Sr\mathrm{Sr}, and Ba\mathrm{Ba}, respectively.

CompoundInitial sintering conditionSecond sintering condition
Ca3_3ReO5_5Cl2_2800°C for 24h900°C for 48h
Ca3_3ReO5_5Br2_2700°C for 24h750°C for 24h
Sr3_3ReO5_5Cl2_2, Sr3_3ReO5_5Br2_2750°C for 24h800°C for 48h
Ba3_3ReO5_5Cl2_2, Ba3_3ReO5_5Br2_2750°C for 24h800°C for 48h
Pb3_3ReO5_5Cl2_2, Pb3_3WO5_5Cl2_2550°C for 24h570°C for 24h

Table S1. Synthesis conditions for the polycrystalline samples used in this work.

X-ray diffraction measurement and structural analysis

The synthesized polycrystalline samples were characterized by X-ray diffraction (XRD) measurements with Cu KαK\alpha radiation using a diffractometer (RINT-2500, Rigaku). All the polycrystalline samples used for physical property measurements were confirmed to be single phase without any noticeable impurities. The crystal structures of A3ReO5Br2A_3\mathrm{ReO}_5\mathrm{Br}_2 (A=CaA = \mathrm{Ca}, Sr\mathrm{Sr}, Ba\mathrm{Ba}) were determined by single-crystal XRD measurements at room temperature using an R-AXIS RAPID IP diffractometer (Rigaku) with monochromated Mo KαK\alpha radiation. The structures were solved by direct methods and refined by full-matrix least-squares methods on ∣F∣2|F|^2 by using the SHELXL2013 software. The structural analysis of Pb3ReO5Cl2\mathrm{Pb}_3\mathrm{ReO}_5\mathrm{Cl}_2 was performed by the Rietveld method with RIETAN-FP program [67] on the basis of the Ba3ReO5Cl2\mathrm{Ba}_3\mathrm{ReO}_5\mathrm{Cl}_2 structure [50]. The powder XRD data used for the structure analysis were collected at room temperature using a SmartLab diffractometer (Rigaku) with Cu Kα1K\alpha1 radiation monochromated by a Ge(111)-Johansson-type monochromator.

Physical property measurements

Magnetization measurements up to 7 T were performed using a commercial magnetometer (MPMS-3, Quantum Design). Heat capacity measurements were performed in zero field by a thermal relaxation method using a commercial cryostat equipped with a superconducting magnet (PPMS, Quantum Design) down to 2 K and by a quasi-adiabatic heat-pulse method in a 3^3He cryostat down to 0.24 K. Magnetization measurements up to ∼\sim60 T and ∼\sim130 T were performed by the induction method using a non-destructive pulsed magnet (∼\sim4 ms duration) and a horizontal single-turn coil system (∼\sim8 μ\mus duration) [55], respectively. Polycrystalline samples were used in all the magnetization measurements. A3ReO5Br2A_3\mathrm{ReO}_5\mathrm{Br}_2 (A=CaA = \mathrm{Ca}, Sr\mathrm{Sr}, Ba\mathrm{Ba}) single crystals and Pb3ReO5Cl2\mathrm{Pb}_3\mathrm{ReO}_5\mathrm{Cl}_2 polycrystalline samples were used in the heat capacity measurements.

First-principles calculation

First-principles calculations are performed based on density functional theory (DFT) using the program package Quantum ESPRESSO [59], which employs plane waves and pseudopotentials to describe the Kohn–Sham orbitals and the crystalline potential, respectively. We set the plane-wave cutoff for a wave function to 120 Ry, suitable for the optimized norm-conserving pseudopotentials [60] provided in the SG15 pseudopotential library [61]. The exchange and correlation effect is described with the Perdew–Burke–Ernzerhof functional [62] based on the generalized gradient correction. We perform the Brillouin-zone integral on 5×10×55 \times10 \times5 (10×10×510 \times10 \times5) kk-point grids for the charge-density calculation of the Type I (Type II) structure by using the optimized tetrahedron method [63]. The fluctuated position of halogen atoms is fixed to the middle of the fluctuation to make the structure suitable for the DFT calculation. We computed the maximally localized Wannier function (MLWF) [64] associated with the 5dxy5d_{xy} orbitals of the Re atoms. The onsite screened Coulomb integral (UU) is computed with the constrained random phase approximation (cRPA) [65]. For the calculations of MLWF and cRPA, we use the RESPACK [66] program package. The susceptibility for the screened Coulomb interaction in Type I (II) materials is computed with 878 (439) empty bands whose upper limit is 35 eV from the Fermi level. The plane-wave cutoff for the susceptibility is set to 12 eV.

Orthogonalized finite-temperature Lanczos method

The finite-temperature Lanczos method (FTLM) is useful for the analysis of frustrated quantum spin systems, though [58]. The orthogonalized finite-temperature Lanczos method (OFTLM) is a more accurate method than the standard FTLM, particularly at low temperatures [18, 56]. In the OFTLM, we first calculate several low-lying exact eigenvectors ∣Ψi,m⟩|\Psi_{i,m}\rangle with NVN_V levels, where we define the order of the corresponding eigenvalues {Ei,m}\{E_{i,m}\} as E0,m≤E1,m≤⋯≤ENV−1,mE_{0,m} \le E_{1,m} \le\cdots\le E_{N_V-1,m}. By using a normalized random initial vector with Stotz=mS^z_{\mathrm{tot}} = m, ∣Vr,m⟩|V_{r,m}\rangle, we calculate the modulated random vector

∣Vr,m′⟩=[I−∑i=0NV−1∣Ψi,m⟩⟨Ψi,m∣]∣Vr,m⟩,(2)|V'_{r,m}\rangle= \left[I - \sum_{i=0}^{N_V-1} |\Psi_{i,m}\rangle\langle\Psi_{i,m}|\right]|V_{r,m}\rangle, \tag*{(2)}

with normalization

∣Vr,m′⟩⇒∣Vr,m′⟩⟨Vr,m′∣Vr,m′⟩.(3)|V'_{r,m}\rangle\Rightarrow\frac{|V'_{r,m}\rangle}{\sqrt{\langle V'_{r,m}|V'_{r,m}\rangle}}. \tag*{(3)}

The partition function of the OFTLM is obtained as follows:

Z(T,h)OFTL=∑m=−MsatMsat[Nst(m)−NVR∑r=1R∑j=0ML−1e−βEj(r)(h)∣⟨Vr,m′∣ψj(r)⟩∣2+∑i=0NV−1e−βEi,m(h)],(4)Z(T,h)_{\mathrm{OFTL}} = \sum_{m=-M_{\mathrm{sat}}}^{M_{\mathrm{sat}}} \left[\frac{N_{\mathrm{st}}^{(m)}-N_V}{R}\sum_{r=1}^{R}\sum_{j=0}^{M_L-1} e^{-\beta E_j^{(r)}(h)}|\langle V'_{r,m}|\psi_j^{(r)}\rangle|^2 + \sum_{i=0}^{N_V-1}e^{-\beta E_{i,m}(h)}\right], \tag*{(4)}

where MsatM_{\mathrm{sat}} is the saturation magnetization, Nst(m)N_{\mathrm{st}}^{(m)} is the dimension of the Hilbert subspace with Stotz=mS^z_{\mathrm{tot}} = m, RR denotes the number of random samplings of the OFTLM, MLM_L denotes the dimension of the Krylov subspace, β\beta is the inverse temperature 1/kBT1/k_{\mathrm{B}}T, and ∣ψj(r)⟩|\psi_j^{(r)}\rangle [Ej(r)(h)E_j^{(r)}(h)] are the eigenvectors (eigenvalues) in the MLM_L-dimensional Krylov subspace with Stotz=mS^z_{\mathrm{tot}} = m. The energy E(T,h)OFTLE(T,h)_{\mathrm{OFTL}}, magnetization M(T,h)OFTLM(T,h)_{\mathrm{OFTL}}, magnetic susceptibility χ(T)OFTL\chi(T)_{\mathrm{OFTL}}, and magnetic entropy Smag(T,h)OFTLS_{\mathrm{mag}}(T,h)_{\mathrm{OFTL}} are obtained as follows:

E(T,h)OFTL=1Z(T,h)OFTL∑m=−MsatMsat[Nst(m)−NVR∑r=1R∑j=0ML−1εj,m(r)(h)e−βεj,m(r)(h)∣⟨Vr,m∣ψj,m⟩∣2+∑i=0NV−1Ei,m(h)e−βEi,m(h)](5)E(T,h)_{\mathrm{OFTL}}=\frac{1}{Z(T,h)_{\mathrm{OFTL}}}\sum_{m=-M_{\mathrm{sat}}}^{M_{\mathrm{sat}}}\left[\frac{N_{\mathrm{st}}^{(m)}-N_V}{R}\sum_{r=1}^{R}\sum_{j=0}^{M_L-1}\varepsilon_{j,m}^{(r)}(h)e^{-\beta\varepsilon_{j,m}^{(r)}(h)}\lvert\langle V_{r,m}\lvert\psi_{j,m}\rangle\rvert^2+\sum_{i=0}^{N_V-1}E_{i,m}(h)e^{-\beta E_{i,m}(h)}\right] \tag*{(5)}
M(T,h)OFTL=1Z(T,h)OFTL∑m=−MsatMsat[Nst(m)−NVR∑r=1R∑j=0ML−1me−βεj,m(r)(h)∣⟨Vr,m∣ψj,m⟩∣2+∑i=0NV−1me−βEi,m(h)](6)M(T,h)_{\mathrm{OFTL}}=\frac{1}{Z(T,h)_{\mathrm{OFTL}}}\sum_{m=-M_{\mathrm{sat}}}^{M_{\mathrm{sat}}}\left[\frac{N_{\mathrm{st}}^{(m)}-N_V}{R}\sum_{r=1}^{R}\sum_{j=0}^{M_L-1}m e^{-\beta\varepsilon_{j,m}^{(r)}(h)}\lvert\langle V_{r,m}\lvert\psi_{j,m}\rangle\rvert^2+\sum_{i=0}^{N_V-1}m e^{-\beta E_{i,m}(h)}\right] \tag*{(6)}
χ(T)OFTL=1kBTZ(T,0)OFTL∑m=−MsatMsat[Nst(m)−NVR∑r=1R∑j=0ML−1m2e−βεj,m(r)∣⟨Vr,m∣ψj,m⟩∣2+∑i=0NV−1m2e−βEi,m](7)\chi(T)_{\mathrm{OFTL}}=\frac{1}{k_{\mathrm{B}}TZ(T,0)_{\mathrm{OFTL}}}\sum_{m=-M_{\mathrm{sat}}}^{M_{\mathrm{sat}}}\left[\frac{N_{\mathrm{st}}^{(m)}-N_V}{R}\sum_{r=1}^{R}\sum_{j=0}^{M_L-1}m^2e^{-\beta\varepsilon_{j,m}^{(r)}}\lvert\langle V_{r,m}\lvert\psi_{j,m}\rangle\rvert^2+\sum_{i=0}^{N_V-1}m^2e^{-\beta E_{i,m}}\right] \tag*{(7)}
Smag(T,h)OFTL=E(T,h)OFTLT−kBln⁡Z(T,h)OFTL.(8)S_{\mathrm{mag}}(T,h)_{\mathrm{OFTL}}=\frac{E(T,h)_{\mathrm{OFTL}}}{T}-k_{\mathrm{B}}\ln Z(T,h)_{\mathrm{OFTL}}. \tag*{(8)}

Since the last terms in Eqs. (4), (5), (6), and (7) are exact values, they are more accurate than those obtained using the standard FTLM, particularly at low temperatures.

In the adiabatic processes, the entropy remains constant while the temperature changes. The temperature under the adiabatic processes can be determined based on the following relationship with hh:

Smag(T,h)OFTL−Smag(Tini,0)OFTL=0,(9)S_{\mathrm{mag}}(T,h)_{\mathrm{OFTL}}-S_{\mathrm{mag}}(T_{\mathrm{ini}},0)_{\mathrm{OFTL}}=0, \tag*{(9)}

where Smag(Tini,0)OFTLS_{\mathrm{mag}}(T_{\mathrm{ini}},0)_{\mathrm{OFTL}} corresponds to the initial entropy, and TiniT_{\mathrm{ini}} represents the initial temperature. We solved Eq. (9) for TT using the Newton-Raphson method, and then substituted the obtained TT into Eq. (6) to calculate the magnetization under the adiabatic processes.

We performed OFTLM calculations for a cluster of 36 sites under periodic boundary conditions with R=10R=10, NV=4N_V=4, and ML=100M_L=100. Our calculations have revealed that there are almost no finite-size effects in the magnetization for kBT/J>0.1k_{\mathrm{B}}T/J>0.1 and in the magnetic susceptibility for kBT/J>0.2k_{\mathrm{B}}T/J>0.2 [18]. Therefore, the analysis of the magnetic susceptibility and magnetization curves in this study is sufficiently accurate.

DATA AVAILABILITY

The datasets generated during and/or analyzed during the current study are available from the corresponding author upon reasonable request. The data from the first-principles calculations are available in the ARIM-mdx data repository [68, 69].

ACKNOWLEDGEMENTS

We thank the Supercomputer Center, the Institute for Solid State Physics, the University of Tokyo for providing us with the CPU time. This work was partly supported by the Japan Society for the Promotion of Science (JSPS) KAKENHI Grants-In-Aid for Scientific Research (No. 19H05821, No. 20H01858, No. 20J10988, No. 22H01167, No. 23H04860). M.G. was a postdoctoral research fellow of the JSPS.

AUTHOR CONTRIBUTIONS

M.G. and D.H. conceived and organized the project. S.K. and D.H. synthesized samples under supervision of Z.H.; T.Y. performed the XRD measurement and structural analysis. D.H. and K.D. performed the magnetic susceptibility and heat capacity measurements. A.M. performed the magnetization measurement in the non-destructive pulsed magnet under supervision of M.G. and K.K.; M.G. and N.M. performed the magnetization measurement in the single-turn coil system under supervision of Y.K.; K.M. performed the OFTLM calculations. M.K. performed the first-principles calculations. M.G. wrote the manuscript with input comments from all co-authors.

COMPETING INTERESTS

The authors declare no competing financial or non-financial interests.

ADDITIONAL INFORMATION

Supplementary Information accompanies this article.

Rhenium oxyhalides: a showcase for anisotropic-triangular lattice quantum antiferromagnets

M. Gen; D. Hirai; K. Morita; S. Kogane; N. Matsuyama; T. Yajima; M. Kawamura; K. Deguchi; A. Matsuo; K. Kindo; Y. Kohama; Z. Hiroi

This Supplementary Information includes contents as listed below:

Note 1. Synthesis conditions for polycrystalline samples

Note 2. Structural analysis for four novel compounds

Note 3. Absence of Dzyaloshinskii-Moriya interaction for the Type-II structure

Note 4. Magnetic heat capacity below TNT_{\mathrm{N}} for Pb3_3ReO5_5Cl2_2

Note 5. Details of first-principles calculations results

Note 6. Magnetization curves measured using a nondestructive pulsed magnet

Synthesis conditions for polycrystalline samples

Table S1 summarizes the synthesis conditions for polycrystalline samples of A3_3ReO5_5X2_2 (A = Ca, Sr, Ba, Pb; X = Cl, Br) and Pb3_3WO5_5Cl2_2 in the solid state reactions. The heating and cooling rates were set at 200°C/h for all the syntheses.

Structural analysis for four novel compounds

Table S2 summarizes the crystallographic parameters of the four novel compounds, A3ReO5Br2\mathrm{A}_3\mathrm{ReO}_5\mathrm{Br}_2 (A=Ca,Sr,Ba\mathrm{A} = \mathrm{Ca}, \mathrm{Sr}, \mathrm{Ba}) and Pb3ReO5Cl2\mathrm{Pb}_3\mathrm{ReO}_5\mathrm{Cl}_2. The structural analysis was performed using the single-crystal XRD data for the three bromides, and the Rietveld analysis was performed using the powder XRD data for Pb3ReO5Cl2\mathrm{Pb}_3\mathrm{ReO}_5\mathrm{Cl}_2.

Absence of Dzyaloshinskii-Moriya interaction for the Type-II structure

The DM interaction is absent for A=Sr\mathrm{A} = \mathrm{Sr}, Ba\mathrm{Ba}, and Pb\mathrm{Pb} with the Type II structure due to the following reasons:

  1. An inversion symmetry exists at the center of the Re–Re diagonal J′J' bonds.

  2. There are mirror planes perpendicular to the aa and cc axes.

  3. Two-fold rotation axis along the bb axis exists at the center of the Re–Re intrachain JJ bonds.

Magnetic heat capacity below TNT_{\mathrm{N}} for Pb3ReO5Cl2\mathrm{Pb}_{3}\mathrm{ReO}_{5}\mathrm{Cl}_{2}

To estimate the lattice heat capacity of Pb3ReO5Cl2\mathrm{Pb}_{3}\mathrm{ReO}_{5}\mathrm{Cl}_{2}, we measured the heat capacity of isostructural nonmagnetic compound Pb3WO5Cl2\mathrm{Pb}_{3}\mathrm{WO}_{5}\mathrm{Cl}_{2} in zero field. Figure S1 shows C/TC/T–TT curves of Pb3ReO5Cl2\mathrm{Pb}_{3}\mathrm{ReO}_{5}\mathrm{Cl}_{2} and Pb3WO5Cl2\mathrm{Pb}_{3}\mathrm{WO}_{5}\mathrm{Cl}_{2} at low temperatures. By subtracting the experimental C/TC/T values of Pb3WO5Cl2\mathrm{Pb}_{3}\mathrm{WO}_{5}\mathrm{Cl}_{2} from those of Pb3ReO5Cl2\mathrm{Pb}_{3}\mathrm{ReO}_{5}\mathrm{Cl}_{2}, the temperature dependence of the magnetic heat capacity, CmagC_{\mathrm{mag}}, for Pb3ReO5Cl2\mathrm{Pb}_{3}\mathrm{ReO}_{5}\mathrm{Cl}_{2} was obtained, as shown in Figure S1. Below TNT_{\mathrm{N}}, CmagC_{\mathrm{mag}} is proportional to T3T^{3}, as indicated by a red fitting curve. This suggests the presence of 3D magnetic correlations in the magnetically ordered phase of Pb3ReO5Cl2\mathrm{Pb}_{3}\mathrm{ReO}_{5}\mathrm{Cl}_{2}.

Heat capacity curves and estimated magnetic heat capacity

Figure S1. (a) Heat capacity divided by temperature C/TC/T at 0 T for Pb3ReO5Cl2\mathrm{Pb}_{3}\mathrm{ReO}_{5}\mathrm{Cl}_{2} (diamonds) and Pb3WO5Cl2\mathrm{Pb}_{3}\mathrm{WO}_{5}\mathrm{Cl}_{2} (circles). (b) Estimated magnetic heat capacity divided by temperature Cmag/TC_{\mathrm{mag}}/T for Pb3ReO5Cl2\mathrm{Pb}_{3}\mathrm{ReO}_{5}\mathrm{Cl}_{2}. The red line represents a fit of Cmag=kT3C_{\mathrm{mag}} = kT^{3} to the experimental data between 2 and 5.5 K.

Details of first-principles calculations results

In order to evaluate the strengths of the interlayer exchange couplings in A3ReO5X\mathrm{A}_{3}\mathrm{ReO}_{5}\mathrm{X}, we performed the first-principles calculations based on density functional theory. Here, we used the structural parameters shown in Table S2. Note that, for Sr3ReO5Br2\mathrm{Sr}_{3}\mathrm{ReO}_{5}\mathrm{Br}_{2} and Ba3ReO5Br2\mathrm{Ba}_{3}\mathrm{ReO}_{5}\mathrm{Br}_{2}, we fixed the atomic position of Br in the average position of two different Br sites with half occupancy. Table S3 summarizes the calculation results of the hopping integrals between two Re6+\mathrm{Re}^{6+} ions along the intrachain and interchain bonds, tt and t′t', respectively, as well as those between the interlayer bonds, t′′t'', and t′′′t''', defined for the Type-II structure as shown in Figure 3e in the main text. The results for A3ReO5Cl2\mathrm{A}_{3}\mathrm{ReO}_{5}\mathrm{Cl}_{2} (A=Ca\mathrm{A} = \mathrm{Ca}, Sr\mathrm{Sr}, Ba\mathrm{Ba}) were already published in Refs. [1, 2]. Considering that the magnetic exchange energy can be related as J=4t2/UJ = 4t^{2}/U, where UU is the Coulomb integral, it is suggested that the strengths of the interlayer exchange couplings J′′J'' and J′′′J''' in PROC are one to two orders of magnitude stronger than in other compounds.

xxyyzzOccupancy100Uiso100U_{\mathrm{iso}} (Å2^2)
Ca3ReO5Br2\mathrm{Ca}_3\mathrm{ReO}_5\mathrm{Br}_2 (R=4.58 %R = 4.58\ \%, wR=10.24 %wR = 10.24\ \%)
Space group PnmaPnma (No. 62), a=12.1219(14)a = 12.1219(14) Å, b=5.6447(7)b = 5.6447(7) Å, c=11.2205(10)c = 11.2205(10) Å
Ca14c4c0.35385(15)1/41/40.84889(14)11.14(4)
Ca24c4c0.33714(17)3/43/40.59350(11)11.04(4)
Ca34c4c0.33380(15)1/41/40.34145(16)11.09(4)
Re4c4c0.19045(3)1/41/40.59694(2)10.78(2)
O18d8d0.2431(4)0.0269(6)0.7113(3)11.07(8)
O28d8d0.2392(3)0.4784(7)0.4817(3)11.08(8)
O34c4c0.0495(6)1/41/40.5984(4)11.33(13)
Br14c4c0.52924(7)1/41/40.19156(6)11.39(3)
Br24c4c0.50835(11)3/43/40.40308(7)12.29(3)
Sr3ReO5Br2\mathrm{Sr}_3\mathrm{ReO}_5\mathrm{Br}_2 (R=2.77 %R = 2.77\ \%, wR=5.13 %wR = 5.13\ \%)
Space group CmcmCmcm (No. 63), a=5.7145(3)a = 5.7145(3) Å, b=13.3445(7)b = 13.3445(7) Å, c=11.3044(6)c = 11.3044(6) Å
Sr14c4c00.38997(8)1/41/411.29(2)
Sr28f8f00.15981(6)0.46553(6)11.563(19)
Re4c4c1/21/20.23305(3)1/41/411.088(14)
O116h16h0.7273(6)0.2745(3)0.3639(3)11.47(8)
O24c4c1/21/20.1059(6)1/41/412.44(19)
Br18f8f1/21/20.49837(11)0.27498(15)0.52.29(6)
Br28e8e0.0573(3)1/21/21/21/20.52.98(6)
Ba3ReO5Br2\mathrm{Ba}_3\mathrm{ReO}_5\mathrm{Br}_2 (R=3.38 %R = 3.38\ \%, wR=4.82 %wR = 4.82\ \%)
Space group CmcmCmcm (No. 63), a=5.87477(4)a = 5.87477(4) Å, b=14.19916(11)b = 14.19916(11) Å, c=11.60519(9)c = 11.60519(9) Å
Ba18f8f00.15430(5)0.46319(9)10.058(12)
Ba24c4c1/21/20.10984(8)3/43/410.058(12)
Re4c4c00.26585(6)3/43/410.1
O116h16h0.2366(10)0.2314(4)0.6396(5)12.2(2)
O24c4c00.3914(8)3/43/412.2(2)
Br18f8f00.00700(14)0.7624(7)0.50.54(8)
Br28e8e0.5220(17)01/21/20.51.81(11)
Pb3ReO5Cl2\mathrm{Pb}_3\mathrm{ReO}_5\mathrm{Cl}_2 (R=3.77 %R = 3.77\ \%, wR=3.85 %wR = 3.85\ \%)
Space group CmcmCmcm (No. 63), a=5.59462(4)a = 5.59462(4) Å, b=13.26905(8)b = 13.26905(8) Å, c=10.97535(7)c = 10.97535(7) Å
Pb18f8f00.35012(3)0.52789(4)11.12(10)
Pb24c4c00.89107(4)1/41/411.125(15)
Re4c4c00.22927(5)1/41/410.425(14)
O116h16h0.2346(6)0.2682(3)0.1300(4)11.17(11)
O24c4c00.1074(6)1/41/412.1(3)
Cl14a4a00014.81(14)
Cl24c4c00.4752(3)1/41/413.38(11)

Table S2. Crystallographic parameters for Ca3ReO5Br2\mathrm{Ca}_3\mathrm{ReO}_5\mathrm{Br}_2, Sr3ReO5Br2\mathrm{Sr}_3\mathrm{ReO}_5\mathrm{Br}_2, Ba3ReO5Br2\mathrm{Ba}_3\mathrm{ReO}_5\mathrm{Br}_2, and Pb3ReO5Cl2\mathrm{Pb}_3\mathrm{ReO}_5\mathrm{Cl}_2.

Compoundtt (eV)t′t' (eV)t′′t'' (eV)t′′′t''' (eV)UU (eV)
Ca3ReO5Cl2\mathrm{Ca}_{3}\mathrm{ReO}_{5}\mathrm{Cl}_{2}−0.050765−0.027532N/AN/A3.391358
Ca3ReO5Br2\mathrm{Ca}_{3}\mathrm{ReO}_{5}\mathrm{Br}_{2}−0.048177−0.026705N/AN/A3.277585
Sr3ReO5Cl2\mathrm{Sr}_{3}\mathrm{ReO}_{5}\mathrm{Cl}_{2}−0.058905−0.0202740.000002−0.0023593.373051
Sr3ReO5Br2\mathrm{Sr}_{3}\mathrm{ReO}_{5}\mathrm{Br}_{2}−0.055679−0.019031−0.000070−0.0031753.302044
Ba3ReO5Cl2\mathrm{Ba}_{3}\mathrm{ReO}_{5}\mathrm{Cl}_{2}−0.040984−0.022597−0.000587−0.0027153.159830
Ba3ReO5Br2\mathrm{Ba}_{3}\mathrm{ReO}_{5}\mathrm{Br}_{2}−0.051275−0.0202290.001340−0.0038112.912788
Pb3ReO5Cl2\mathrm{Pb}_{3}\mathrm{ReO}_{5}\mathrm{Cl}_{2}−0.064862−0.0201210.005146−0.0282632.112172

Table S3. Hopping integrals in A3ReO5X\mathrm{A}_{3}\mathrm{ReO}_{5}\mathrm{X} obtained by the first-principles calculations.

Magnetization curves measured using a nondestructive pulsed magnet

Figure S2 shows the MM–HH curves of five compounds, A3ReO5X2\mathrm{A}_3\mathrm{ReO}_5\mathrm{X}_2 (A=Ca,Sr,Pb\mathrm{A} = \mathrm{Ca}, \mathrm{Sr}, \mathrm{Pb}; X=Cl,Br\mathrm{X} = \mathrm{Cl}, \mathrm{Br}), measured at 4.2 K using a non-destructive pulsed magnet. These data are used for calibrating the absolute value of MM for the MM–HH curves measured using a single-turn coil system (shown in Figure 5f–Figure 5j in the main text).

Quasi-isothermal magnetization curves for five A3ReO5X2 compounds

Figure S2. Quasi-isothermal magnetization curves for five A3ReO5X2\mathrm{A}_3\mathrm{ReO}_5\mathrm{X}_2 compounds measured at 4.2 K using a non-destructive pulsed magnet.

[1] D. Hirai, K. Nawa, M. Kawamura, T. Misawa, and Z. Hiroi, One-dimensionalization by Geometrical Frustration in the Anisotropic Triangular Lattice of the 5d5d Quantum Antiferromagnet Ca3ReO5Cl2\mathrm{Ca}_3\mathrm{ReO}_5\mathrm{Cl}_2, J. Phys. Soc. Jpn. 88, 044708 (2019).

[2] D. Hirai, T. Yajima, K. Nawa, M. Kawamura, and Z. Hiroi, Anisotropic Triangular Lattice Realized in Rhenium Oxychlorides A3ReO5Cl2\mathrm{A}_3\mathrm{ReO}_5\mathrm{Cl}_2 (A=Ba,Sr\mathrm{A} = \mathrm{Ba}, \mathrm{Sr}), Inorg. Chem. 59, 10025 (2020).

References

  1. [1]L. Balents, Spin liquids in frustrated magnets, Nature 464, 199 (2010).
  2. [2]M. Kohno, O. A. Starykh, and L. Balents, Spinons and triplons in spatially anisotropic frustrated antiferromagnets, Nat. Phys. 3, 790 (2007).
  3. [3]M. R. Norman, Colloquium: Herbertsmithite and the search for the quantum spin liquid, Rev. Mod. Phys. 88, 041002 (2016).DOI
  4. [4]E. Ghorbani, L. F. Tocchio, and F. Becca, Variational wave functions for the S = 1/2 Heisenberg model on the anisotropic triangular lattice: Spin liquids and spiral orders, Phys. Rev. B 93, 085111 (2016).
  5. [5]S. Yunoki and S. Sorella, Two spin liquid phases in the spatially anisotropic triangular Heisenberg model, Phys. Rev. B 74, 014408 (2006).DOI
  6. [6]D. Heidarian, S. Sorella, and F. Becca, Spin-1/2 Heisenberg model on the anisotropic triangular lattice: From magnetism to a one-dimensional spin liquid, Phys. Rev. B 80, 012404 (2009).
  7. [7]Y. Hayashi and M. Ogata, Possibility of Gapless Spin Liquid State by One-Dimensionalization, J. Phys. Soc. Jpn. 76, 053705 (2007).DOI
  8. [8]M. Q. Weng, D. N. Sheng, Z. Y. Weng, and R. J. Bursill, Spin-liquid phase in an anisotropic triangular-lattice Heisenberg model: Exact diagonalization and density-matrix renormalization group calculations, Phys. Rev. B 74, 012407 (2006).DOI
  9. [9]O. A. Starykh and L. Balents, Ordering in Spatially Anisotropic Triangular Antiferromagnets, Phys. Rev. Lett. 98, 077205 (2007).pubmed.ncbi.nlm.nih.gov/17359057
  10. [10]J. Reuther and R. Thomale, Functional renormalization group for the anisotropic triangular antiferromagnet, Phys. Rev. B 83, 024402 (2011).DOI
  11. [11]K. Harada, Numerical study of incommensurability of the spiral state on spin-1/2 spatially anisotropic triangular antiferromagnets using entanglement renormalization, Phys. Rev. B 86, 184421 (2012).DOI
  12. [12]Z. Weihong, R. H. McKenzie, and R. R. P. Singh, Phase diagram for a class of spin-1/2 Heisenberg models interpolating between the square-lattice, the triangular-lattice, and the linear-chain limits, Phys. Rev. B 59, 14367 (1999).
  13. [13]C.-H. Chung, K. Voelker, and Y. B. Kim, Statistics of spinons in the spin-liquid phase of Cs2CuCl4, Phys. Rev. B 68, 094412 (2003).DOI
  14. [14]S. Yunoki and S. Sorella, Resonating Valence Bond Wave Function for the Two-Dimensional Fractional Spin Liquid, Phys. Rev. Lett. 92, 157003 (2004).pubmed.ncbi.nlm.nih.gov/15169309
  15. [15]W. Zheng, J. O. Fjærestad, R. R. P. Singh, R. H. McKenzie, and R. Coldea, Anomalous Excitation Spectra of Frustrated Quantum Antiferromagnets, Phys. Rev. Lett. 96, 057201 (2006).pubmed.ncbi.nlm.nih.gov/16486972
  16. [16]O. A. Starykh, H. Katsura, and L. Balents, Extreme sensitivity of a frustrated quantum magnet: Cs2CuCl4, Phys. Rev. B 82, 014421 (2010).DOI
  17. [17]O. A. Starykh, W. Jin, and A. V. Chubukov, Phases of a Triangular-Lattice Antiferromagnet Near Saturation, Phys. Rev. Lett. 113, 087204 (2014).pubmed.ncbi.nlm.nih.gov/25192123
  18. [18]K. Morita, Isothermal and adiabatic magnetization processes of the spin-1/2 Heisenberg model on an anisotropic triangular lattice, Phys. Rev. B 105, 064428 (2022).DOI
  19. [19]W. Zheng, R. R. P. Singh, R. H. McKenzie, and R. Coldea, Temperature dependence of the magnetic susceptibility for triangular-lattice antiferromagnets with spatially anisotropic exchange constants, Phys. Rev. B 71, 134422 (2005).DOI
  20. [20]Y. Yu, S. Li, S. Isk, and E. Gul, Magnetic phases of the anisotropic triangular lattice Hubbard model, Phys. Rev. B 107, 075106 (2023).DOI
  21. [21]B. Bernu, P. Lecheminant, C. Lhuillier, and L. Pierre, Exact spectra, spin susceptibilities, and order parameter of the quantum Heisenberg antiferromagnet on the triangular lattice, Phys. Rev. B 50, 10048 (1994).DOI
  22. [22]S. R. White and A. L. Chernyshev, Néel Order in Square and Triangular Lattice Heisenberg Models, Phys. Rev. Lett. 99, 127004 (2007).DOI
  23. [23]Y. Endoh, G. Shirane, R. J. Birgeneau, Peter M. Richards, and S. L. Holt, Dynamics of an S = 1/2, One-Dimensional Heisenberg Antiferromagnet, Phys. Rev. Lett. 32, 170 (1974).
  24. [24]D. A. Tennant, T. G. Perring, R. A. Cowley, and S. E. Nagler, Unbound spinons in the S = 1/2 antiferromagnetic chain KCuF3, Phys. Rev. Lett. 70, 4003 (1993).pubmed.ncbi.nlm.nih.gov/10054020
  25. [25]R. Coldea, D. A. Tennant, R. A. Cowley, D. F. McMorrow, B. Dorner, and Z. Tylczynski, Neutron scattering study of the magnetic structure of Cs2CuCl4, J. Phys.: Condens. Matter 8, 7473 (1996).
  26. [26]R. Coldea, D. A. Tennant, A. M. Tsvelik, and Z. Tylczynski, Experimental Realization of a 2D Fractional Quantum Spin Liquid, Phys. Rev. Lett. 86, 1335 (2001).pubmed.ncbi.nlm.nih.gov/11178077
  27. [27]R. Coldea, D. A. Tennant, K. Habicht, P. Smeibidl, C. Wolters, and Z. Tylczynski, Direct Measurement of the Spin Hamiltonian and Observation of Condensation of Magnons in the 2D Frustrated Quantum Magnet Cs2CuCl4, Phys. Rev. Lett. 88, 137203 (2002).pubmed.ncbi.nlm.nih.gov/11955122
  28. [28]R. Coldea, D. A. Tennant, and Z. Tylczynski, Extended scattering continua characteristic of spin fractionalization in the two-dimensional frustrated quantum magnet Cs2CuCl4 observed by neutron scattering, Phys. Rev. B 68, 134424 (2003).DOI
  29. [29]T. Radu, H. Wilhelm, V. Yushankhai, D. Kovrizhin, R. Coldea, Z. Tylczynski, T. Lühmann, and F. Steglich, Bose-Einstein Condensation of Magnons in Cs2CuCl4, Phys. Rev. Lett. 95, 127202 (2005).pubmed.ncbi.nlm.nih.gov/16197102
  30. [30]Y. Tokiwa, T. Radu, R. Coldea, H. Wilhelm, Z. Tylczynski, and F. Steglich, Phys. Rev. B 73, 134414 (2006).
  31. [31]K. Yu, Povarov, A. I. Smirnov, O. A. Starykh, S. V. Petrov, and A. Ya. Shapiro, Modes of Magnetic Resonance in the Spin-Liquid Phase of Cs2CuCl4, Phys. Rev. Lett. 107, 037204 (2011).
  32. [32]A. I. Smirnov, K. Yu. Povarov, S. V. Petrov, and A. Ya. Shapiro, Magnetic resonance in the ordered phases of the two-dimensional frustrated quantum magnet Cs2CuCl4, Phys. Rev. B 85, 184423 (2012).DOI
  33. [33]A. A. Zvyagin, D. Kamensky, M. Ozerov, J. Wosnitza, M. Ikeda, T. Fujita, M. Hagiwara, A. I. Smirnov, T. A. Soldatov, A. Ya. Shapiro, J. Krzystek, R. Hu, H. Ryu, C. Petrovic, and M. E. Zhitomirsky, Phys. Rev. Lett. 112, 077206 (2014).
  34. [34]S. A. Zvyagin, D. Graf, T. Sakurai, S. Kimura, H. Nojiri, J. Wosnitza, H. Ohta, T. Ono, and H. Tanaka, Pressure-tuning the quantum spin Hamiltonian of the triangular lattice antiferromagnet Cs2CuCl4, Nat. Commun. 10, 1064 (2019).DOI
  35. [35]E. Schulze, S. Arsenijevic, L. Opferden, A. N. Ponomaryov, J. Wosnitza, T. Ono, H. Tanaka, and S. A. Zvyagin, Evidence of one-dimensional magnetic heat transport in the triangular-lattice antiferromagnet Cs2CuCl4, Phys. Rev. Research 1, 032022(R) (2019).
  36. [36]T. Ono, H. Tanaka, H. Aruga Katori, F. Ishikawa, H. Mitamura, and T. Goto, Magnetization plateau in the frustrated quantum spin system Cs2CuBr4, Phys. Rev. B 67, 104431 (2003).
  37. [37]T. Ono, H. Tanaka, O. Kolomiyets, H. Mitamura, T. Goto, K. Nakajima, A. Oosawa, Y. Koike, K. Kakurai, J. Klenke, P. Smeibidle, and M. Meißner, Magnetization plateaux of the S = 1/2 two-dimensional frustrated antiferromagnet Cs2CuBr4, J. Phys.: Condens. Matter 16, S773 (2004).DOI
  38. [38]H. Tsuji, C. R. Rotundu, T. Ono, H. Tanaka, B. Andraka, K. Ingersent, and Y. Takano, Thermodynamics of the up-up-down phase of the S = 1/2 triangular-lattice antiferromagnet Cs2CuBr4, Phys. Rev. B 76, 060406(R) (2007).
  39. [39]N. A. Fortune, S. T. Hannahs, Y. Yoshida, T. E. Sherline, T. Ono, H. Tanaka, and Y. Takano, Cascade of Magnetic-Field-Induced Quantum Phase Transitions in a Spin-1/2 Triangular-Lattice Antiferromagnet, Phys. Rev. Lett. 102, 257201 (2009).pubmed.ncbi.nlm.nih.gov/19659114
  40. [40]Y. Shimizu, K. Miyagawa, K. Kanoda, M. Maesato, and G. Saito, Spin Liquid State in an Organic Mott Insulator with a Triangular Lattice, Phys. Rev. Lett. 91, 107001 (2003).pubmed.ncbi.nlm.nih.gov/14525498
  41. [41]H. C. Kandpal, I. Opahle, Y.-Z. Zhang, H. O. Jeschke, and R. Valentí, Revision of Model Parameters for κ-Type Charge Transfer Salts: An Ab Initio Study, Phys. Rev. Lett. 103, 067004 (2009).DOI
  42. [42]T. Koretsune and C. Hotta, Evaluating model parameters of the κ- and β′-type Mott insulating organic solids, Phys. Rev. B 89, 045102 (2014).DOI
  43. [43]Y. Yoshida, H. Ito, M. Maesato, Y. Shimizu, H. Hayama, T. Hiramatsu, Y. Nakamura, H. Kishida, T. Koretsune, C. Hotta, and G. Saito, Spin-disordered quantum phases in a quasi-one-dimensional triangular lattice, Nat. Phys. 11, 679 (2015).DOI
  44. [44]T. Komatsu, N. Matsukawa, T. Inoue, and G. Saito, Realization of Superconductivity at Ambient Pressure by Band-Filling Control in κ-(BEDT-TTF)2Cu2(CN)3, J. Phys. Soc. Jpn. 65, 1340 (1996).DOI
  45. [45]R. Kato, Y. Kashimura, S. Aonuma, N. Hanashaki, and H. Tajima, A new molecular superconductor β′-Et2Me2P[Pd(dmit)2]2 (dmit = 2-thioxo-1,3-dithiole-4,5-dithiolate), Solid State Commun. 105, 561 (1998).
  46. [46]Y. Shimizu, M. Maesato, M. Yoshida, M. Takigawa, M. Itoh, A. Otsuka, H. Yamochi, Y. Yoshida, G. Kawaguchi, D. Graf, and G. Saito, Magnetic field driven transition between valence bond solid and antiferromagnetic order in a distorted triangular lattice, Phys. Rev. Research 3, 023145 (2021).DOI
  47. [47]B. J. Powell, E. P. Kenny, and J. Merino, Dynamical Reduction of the Dimensionality of Exchange Interactions and the “Spin-Liquid” Phase of κ-(BEDT-TTF)2X, Phys. Rev. Lett. 119, 087204 (2017).DOI
  48. [48]D. Hirai, T. Yajima, D. Nishio-Hamane, C. Kim, H. Akiyama, M. Kawamura, T. Misawa, N. Abe, T. Arima, and Z. Hiroi, “Visible” 5d Orbital States in a Pleochroic Oxychloride, J. Am. Chem. Soc. 139, 10784 (2017).DOI
  49. [49]D. Hirai, K. Nawa, M. Kawamura, T. Misawa, and Z. Hiroi, One-dimensionalization by Geometrical Frustration in the Anisotropic Triangular Lattice of the 5d Quantum Antiferromagnet Ca3ReO5Cl2, J. Phys. Soc. Jpn. 88, 044708 (2019).DOI
  50. [50]D. Hirai, T. Yajima, K. Nawa, M. Kawamura, and Z. Hiroi, Anisotropic Triangular Lattice Realized in Rhenium Oxychlorides A3ReO5Cl2 (A = Ba, Sr), Inorg. Chem. 59, 10025 (2020).DOI
  51. [51]K. Nawa, D. Hirai, M. Kofu, K. Nakajima, R. Murasaki, S. Kogane, M. Kimata, H. Nojiri, Z. Hiroi, and T. J. Sato, Bound spinon excitations in the spin-1/2 anisotropic triangular antiferromagnet Ca3ReO5Cl2, Phys. Rev. Research 2, 043121 (2020).
  52. [52]S. A. Zvyagin, A. N. Ponomaryov, J. Wosnitza, D. Hirai, Z. Hiroi, M. Gen, Y. Kohama, A. Matsuo, Y. H. Matsuda, and K. Kindo, Dimensional reduction and incommensurate dynamic correlations in the S = 1/2 triangular-lattice antiferromagnet Ca3ReO5Cl2, Nat. Commun. 13, 6310 (2022).DOI
  53. [53]Y. Choi, S. Lee, J.-H. Lee, S. Lee, M.-J. Seong, and K.-Y. Choi, Bosonic spinons in anisotropic triangular antiferromagnets, Nat. Commun. 12, 6453 (2021).DOI
  54. [54]J. C. Bonner and M. E. Fisher, Linear Magnetic Chains with Anisotropic Coupling, Phys. Rev. 135, A640 (1964).DOI
  55. [55]M. Gen, Y. Okamoto, M. Mori, K. Takenaka, and Y. Kohama, Magnetization process of the breathing pyrochlore magnet CuInCr4S8 in ultrahigh magnetic fields up to 150 T, Phys. Rev. B 101, 054434 (2020).DOI
  56. [56]K. Morita and T. Tohyama, Finite-temperature properties of the Kitaev-Heisenberg models on kagome and triangular lattices studied by improved finite-temperature Lanczos methods, Phys. Rev. Research 2, 013205 (2020).DOI
  57. [57]K. Momma and F. Izumi, VESTA 3 for three-dimensional visualization of crystal, volumetric and morphology data, J. Appl. Crystallogr. 44, 1272 (2011).DOI
  58. [58]J. Jaklič and P. Prelovšek, Lanczos method for the calculation of finite-temperature quantities in correlated systems, Phys. Rev. B 49, 5065(R) (1994).pubmed.ncbi.nlm.nih.gov/10011449
  59. [59]P. Giannozzi, O. Andreussi, T. Brumme et al., J. Phys.: Condens. Matter 29, 465901 (2017).
  60. [60]D. R. Hamann, Phys. Rev. B 88, 085117 (2013).
  61. [61]M. Schlipf and F. Gygi, Comp. Phys. Commun. 196, 36 (2015).
  62. [62]J. P. Perdew, K. Burke, and M. Ernzerhof, Phys. Rev. Lett. 77, 3865 (1996).
  63. [63]M. Kawamura, Y. Gohda, and S. Tsuneyuki, Phys. Rev. B 89, 094515 (2014).
  64. [64]N. Marzari and D. Vanderbilt, Phys. Rev. B 56, 12847 (1997).
  65. [65]T. Miyake, F. Aryasetiawan, and M. Imada Phys. Rev. B 80, 155134 (2009).
  66. [66]K. Nakamura, Y. Yoshimoto, Y. Nomura, T. Tadano, M. Kawamura, T. Kosugi, K. Yoshimi, T. Misawa, and Y. Motoyama, Comp. Phys. Commun. 261, 107781 (2021).
  67. [67]F. Izumi and K. Momma, Three-Dimensional Visualization in Powder Diffraction, Solid State Phenom. 130, 15 (2007).
  68. [68]https://arim.mdx.jp/nextcloud/index.php/s/Q4TRDdnSYArf038
  69. [69]T. Tsuzawa, A. Sugiki, H. Takizawa et al., Proc. IEEE Int. Conf. Cloud Big Data Comput., pp. 1-7 (2022).

Paper details

vlnc:W_yfn8wgmp
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