WorksPhysics and Astronomy
Collinear antiferromagnetic order in spin- triangle lattice antiferromagnet NaFe(PO)
Abstract
We set forth the structural and magnetic properties of the frustrated spin- triangle lattice antiferromagnet NaFe(PO) examined via x-ray diffraction, magnetization, heat capacity, and neutron diffraction measurements on the polycrystalline sample. No structural distortion was detected from the temperature-dependant x-ray diffraction down to 12.5 K, except a systematic lattice contraction. The magnetic susceptibility at high temperatures agrees well with the high-temperature series expansion for a spin- isotropic triangular lattice antiferromagnet with an average exchange coupling of K rather than a one-dimensional spin- chain model. This value of the exchange coupling is consistently reproduced by the saturation field of the pulse field magnetization data. It undergoes a magnetic long-range-order at K. Neutron diffraction experiments elucidate a collinear antiferromagnetic ordering below with the propagation vector . An intermediate value of frustration ratio () reflects moderate frustration in the compound which is corroborated by a reduced ordered magnetic moment of at 1.6 K, compared to its classical value (). Magnetic isotherms exhibit a change of slope envisaging a field induced spin-flop transition at T. The magnetic field vs temperature phase diagram clearly unfold three distinct phase regimes, reminiscent of a frustrated magnet with in-plane (XY-type) anisotropy.
Introduction
Triangular-lattice antiferromagnet (TLA) is a simplest example of geometrically frustrated magnets which has been a motif since the advent of various exotic quantum phases including quantum spin-liquid (QSL) [1, 2]. Ideally, two-dimensional (2D) TLAs with isotropic (Heisenberg) exchange interaction undergo a non-collinear 120° ordering irrespective of their spin value [3, 4, 5, 6, 7]. However, in real materials the inherent exchange anisotropy and inter-layer couplings often have broad implications, leading to more complex and non-trivial ground states [8]. For instance, TLAs with easy-axis anisotropy, the low temperature 120° state is often preceded by a collinear state in zero-field, as in Ba(Mn,Co)NbO and other TLA compounds [9, 10, 11, 12, 13, 14, 15, 16] while for systems with easy-plane anisotropy, the collinear phase is unstable and entail only 120° ordering [17, 18]. The materials with noncollinear magnetic structure give rise to spin chirality, a wellspring of strong magnetoelectric coupling [19]. Moreover, spatially anisotropic spin- TLA model has been proposed to host QSL phase for a critical value of the ratio of exchange couplings [20, 21, 22]. Recently, an extended fluctuating regime with slow dynamics below magnetic ordering is detected in the spin- isotropic TLAs [23]. Overall, diverse phase regimes are anticipated in TLAs while going from a classical (high spin) to quantum (low spin) limit owing to the geometrical frustration in concurrence with magnetic anisotropy, effect of fluctuations
etc [24, 25]. Recently, a series of compounds with general formula ( Ba, K, Rb, and Sr, Na and Ag, Mn, Ni, Co, Cr, Fe, and Cu, and P and V) are reported with varying crystal structures [26, 27, 28, 29, 30, 31, 32, 33, 34]. Interestingly, this series of compounds are projected as low-dimensional spin systems where the superexchange takes place in an elongated path involving VO or PO units. Though majority of the compounds in this series are described in terms of anisotropic triangular lattice model, yet some of them are found to be exceptions. BaAgCu(VO with triclinic () structure was initially proposed to be a spin- anisotropic triangular lattice system, due to its unequal Cu–Cu bond lengths [27]. Subsequently, the band structure calculations ruled out this proposal and established a superposition of double spin-1/2 chains: one is ferromagnetic and another is antiferromagnetic [34]. Similarly, the monoclinic () compound BaNaCu(VO) features antiferromagnetic spin-1/2 crossed chains arranged 62° with respect to each other [26]. Moreover, the signature of QSL is reported in the structurally perfect spin-1/2 TLA BaNaCo(PO) [35] and spin-3/2 TLAs (K,Rb)AgCr(VO) [33, 36], piquing further interest in this series of compounds. Thus, this series of materials serve as a cradle for novel ground states where one can tune the ground state properties by changing the chemical pressure.
In the following, we delineate a comprehensive study of the thermodynamic properties of the frustrated spin-5/2 magnet NaFe(PO) in which the spins are embedded in a triangular lattice. It crystallizes in a monoclinic space group () and belongs to the family [37]. The crystal structure is presented in Figure 1 where the slightly distorted FeO octahedra are corner shared with distorted PO tetrahedra in order to form the triangular Fe (, spin-5/2) layers in the -plane which are stack along the -direction. Isolated Na atoms are found between the layers. In each layer, the Fe triangles are slightly anisotropic with two Fe-Fe distances of 5.182 Å and one of 5.028 Å [Fig. 1(b)]. Further considering the analogy that the shortest Fe-Fe distance leads to a strong exchange interaction and the longer distance leads to a weak interaction, the spin lattice can also be viewed as coupled spin chains. Of course, the magnitude and sign of the interactions may vary depending on the actual path involved. Our magnetization data could be described well using the spin-5/2 isotropic triangle lattice model. It shows the onset of a magnetic ordering at K which is incisively detected to be a collinear AFM type. The phase diagram with a spin-flop (SF) transition confirms magnetic anisotropy in the compound.
Figure 1. (a) Crystal structure of NaFe(PO) showing triangular layers sculpted by the corner-shared FeO octahedra and PO tetrahedra. Location of Na ions is also shown. (b) A section of the layer showing Fe ions linked via PO tetrahedra on an anisotropic triangular lattice.
Methods
Polycrystalline sample of NaFe(PO) was synthesized following the conventional solid-state reaction technique. Prior to synthesis, the precursor FePO was prepared by heating stoichiometric mixture of FeO (Aldrich, 99.99%) and (NH)HPO (Aldrich, 99.995%) at 880 °C for 12 hours in air. In the next step, the raw materials NaCO (Aldrich, 99.995%) and (NH)HPO (Aldrich, 99.995%) were mixed with FePO in stoichiometric ratios, grounded thoroughly, and pressed into pellets. The pellets were then sintered in an alumina crucible at 730 °C for three days in air with several intermediate grindings.
The phase purity of the sample was confirmed from the powder x-ray diffraction (XRD) performed at room temperature. For the powder XRD experiment, a PANalytical powder diffractometer with Cu radiation ( Å) was used. The temperature-dependent powder XRD measurement was performed over the temperature range 12.5 K 300 K using a low-temperature attachment (Oxford Phenix) to the diffractometer. The dc magnetization () was measured on the powder sample using a superconducting quantum interference device (SQUID) (MPMS-3, Quantum Design) magnetometer. The measurements were performed in the temperature range 1.8 K 380 K and in the magnetic field range 0 7 T. High-field magnetization was measured in pulsed magnetic field at the Dresden high magnetic field laboratory. The details of the experimental procedure are described in Ref. [38]. Heat capacity as a function of temperature and magnetic field was measured on a small pellet using a Physical Property Measurement System (PPMS, Quantum Design).
Neutron powder diffraction (NPD) experiments were carried out in the temperature range 1.6 K 15 K using the powder diffractometer (PD-II, Å) at Dhruva reactor, Bhabha Atomic Research Centre (BARC), Mumbai, India. The 1D neutron-depolarization measurements were performed using the polarized neutron spectrometer (PNS) at the Dhruva reactor, BARC, Mumbai ( Å). For this experiment, polarized neutron beam was produced and analysed using the magnetized CuMnAl [reflection from (111) plane] and CoFe [reflection from (200) plane] single crystals, respectively. The two states of the polarization of incident neutron beam were achieved by a flipper placed just before the sample. The flipping ratio () of the neutron beam was determined by measuring the intensities of neutrons in non-spin flip and spin flip channels with the flipper on and off, respectively. Rietveld refinement of the powder XRD and NPD data was performed using FullProf software package [39].
Quantum Monte-Carlo (QMC) simulation and full diagonalization were performed using the loop and fulldiag algorithms [40] of the ALPS package [41].
Results
X-ray Diffraction
In order to check the structural phase transition or lattice distortion, the sample was examined by measuring XRD pattern at different temperatures. Figure 2 displays the powder XRD pattern at two end temperatures along with the Rietveld fits. The obtained lattice parameters from the refinement at room temperature are Å, Å, Å, and and the unit cell volume Å, which are in close agreement with the previous report [37]. Temperature variation of the lattice constants (, , and ) and are shown in Figure 3. All of them are found to decrease systematically upon cooling down to 12.5 K, as expected. No anomaly at the temperature corresponding to the magnetic transition suggests the absence of significant magnetoelastic coupling.
Figure 2. Powder XRD pattern of NaFe(PO) measured at (a) K and (b) K. The open circles are the experimental data and the solid black line is the Rietveld refined fit. The Bragg positions are indicated by green vertical bars and the bottom solid blue line indicates the difference between the experimental and calculated intensities.
Figure 3. The lattice constants (, , and ), and unit cell volume () are plotted as a function of temperature from 12.5 to 300 K. The solid line in (b) represents the fit using Eq. (1).
Such a temperature dependence of is usually described well by [42]
where is the cell volume in the zero temperature limit, is the bulk modulus, and is the Grüneisen parameter. is the internal energy, which can be derived in terms of the Debye approximation as
Here, is the number of atoms in the unit cell and is the Boltzmann constant. As shown in Figure 3(b), above 12.5 K was fitted reasonably by Eq. (1) producing the parameters: Debye temperature K, Pa, and Å.
Magnetization
Temperature-dependent magnetic susceptibility () of the polycrystalline NaFe(PO) sample measured in 0.5 T applied field is shown in Figure 4(a). The most significant feature in is the presence of a broad maximum at K, mimicking the AFM short-range order, typical for low-dimensional spin systems.

Figure 4. (a) as a function of temperature in an applied field T. The downward arrow points to the transition temperature (). The solid line represents the fit using isotropic triangular lattice model [Eq. (4)]. Inset: comparison of experimental data with spin-5/2 uniform chain model and isotropic triangular lattice simulated using QMC and full diagonalization (FD), respectively. (b) Inverse susceptibility vs and the solid line represents the CW fit, as discussed in the text. Inset: measured in different fields, below .
While going further below this broad maxima, a change in slope is observed at K, which marks an alteration towards an AFM long-range-order (LRO). The ordering temperature is very well evident in the vs plot (not shown). To understand the nature of the magnetic ordering, was measured at different applied fields upto 7 T [see the inset of Figure 4(b)]. With increasing , remains unchanged upto 3 T and then moves weakly towards high temperatures above 3 T. For T, below develops into a contort, which emphasizes that there is some field-induced spin canting prevailing in the system.
The zero-field cooled (ZFC) and field-cooled (ZFC) susceptibilities as a function of temperature were measured in a low field of 0.01 T (not shown here). The absence of splitting between ZFC and FC s certainly rules out the possibility of any spin-glass transition or spin freezing at low temperatures.
For the purpose of analysis, data in the high temperature regime were fitted using the Curie-Weiss (CW)
law,
where is the temperature-independent susceptibility, is the Curie constant, and is the characteristic CW temperature. The fit shown in Figure 4(b) for K yields the parameters: , , and K. A positive value of indicates the presence of dominate AFM interactions between the Fe ions. From the and values, we can quantify frustration by the frustration ratio, (= ) , which apprises us that the system is moderately frustrated. The core diamagnetic susceptibility of NaFe(PO) was
calculated to be by summing the core diamagnetic susceptibilities of individual ions Na, Fe, P, and O [43]. The Van-Vleck paramagnetic susceptibility (), which arises from the second-order contribution to free energy in the presence of magnetic field was obtained by subtracting from to be . Further, from the value of Curie constant , the effective moment calculated using the relation to be , where is the Boltzmann constant, is the Bohr magneton, and is the Avogadro’s number. For a spin-5/2 system, the spin-only effective moment is expected to be , assuming Landé -factor . However, our experimental value of is slightly higher than the spin-only value and corresponds to a -factor of , which is consistent with our spin model fit discussed later. In the mean-field approximation, is a measure of energy scale of the total exchange interactions which is related to the exchange coupling as , where is the number of nearest-neighbours of Fe ions and is the average intra-layer exchange coupling [44]. In this way, we estimated the average exchange coupling within the triangular planes to be K.
Further, for a direct estimation of exchange coupling between the Fe ions and to understand the spin-lattice, we decomposed into two components,
Here, typifies the intrinsic spin susceptibility which can be chosen according to the intrinsic magnetic model. As the Fe ions in the crystal structure are arranged in triangles, we took the expression of high temperature series expansion (HTSE) of for a spin-5/2 Heisenberg isotropic TLA model which has the form [45]
with . This expression is valid for [46]. The solid line in Figure 4(a) represents the best fit to the data above 20 K by Eq. (4) resulting , , and the average AFM exchange coupling K. Indeed, this value of is in good agreement with the value estimated from . We also tried to fit the data using HTSE for of the spin-5/2 Heisenberg 1D chain model in Eq. (4) [47]. The fit however did not converge giving unphysical parameters. Thus, the agreement of our experimental data with the isotropic TLA model suggests frustrated triangular spin-lattice in the system.
To look for the field-induced transition, we measured the magnetization isotherms ( vs ) at different temperatures up to 7 T. As shown in Figure 5(a), for K a bend is observed in the intermediate magnetic fields which is the signature of a metastable field-induced transition. It is more pronounced in the vs plots shown in Figure 5(b) where this feature is manifested in a well defined peak. As the temperature rises, one can observe that the peak is moving slightly towards higher fields and then disappears completely for K.

Figure 5. (a) Magnetization () vs measured at various temperatures from 1.8 K to 50 K. The encircled portion depicts the field induced transition regime. (b) The derivative vs is plotted at various temperatures from 1.8 K to 9 K in-order to highlight the field induced transition.
Further details about field induced transition and saturation magnetization can be obtained from the high field magnetization measurements. The vs data measured in pulsed magnetic field up to 50 T at K are plotted in Figure 6. The pulse field magnetization curve is quantified by scaling it with respect to the SQUID data measured up to 7 T at K. It exhibits a kink at T, similar to that observed in Figure 5, reminiscent of a spin-flop (SF) transition. Above , increases almost linearly with and develops a sharp bend towards saturation above T. These two critical fields are very well visualized as sharp peaks in the vs plot. No obvious features associated with the 1/3 magnetization plateau is observed in the intermediate field range. In the polycrystalline sample,
Figure 6. Magnetization () and its field derivative vs measured at K using pulsed magnetic field in the left and right y-axes, respectively. The pulse field data are scaled with respect to the SQUID data measured up to 7 T. The dash-dotted line represents the saturation magnetization per Fe ion. The sharp peaks in vs plot depict the spin-flop and saturation fields T and T, respectively.
the random orientation of the grains with respect to the applied field direction might obscure this plateau. The linear interpolation of a straight line fit to the magnetization above intercepts y-axis at . This saturation magnetization corresponds to the value expected for a spin-5/2 (Fe) ion with . Further, in a spin AFM defines the energy required to overcome the AFM interactions and to polarize the spins in the direction of applied field. In particular, in a Heisenberg TLA, can be written in terms of the intra-layer exchange coupling as [48]. Our experimental value of T yields an average exchange coupling of K which is indeed close to the value obtained from the analysis of and . A small difference can be attributed to the magnetic anisotropy present in the compound.
Heat Capacity
The temperature-dependent heat capacity of the polycrystalline sample measured in zero field is shown in Figure 7(a). As we go lower in temperature, also goes down systematically and then shows a pronounced -type anomaly at K, illustrating the crossover to a magnetically ordered state. Typically, in high temperatures, in magnetic insulators has a dominant contribution from phonon excitations (), whereas in low-temperatures, the magnetic part of the heat capacity () predominates over . Therefore, in magnetic systems with low energy scale of the exchange coupling, one can separate the magnetic part from the phononic part by analyzing in the high temperature regime.
Figure 7. (a) Heat capacity () as a function of temperature in zero applied field. The solid line represents the phonon contribution () Eq. (6) while the dashed line indicates the magnetic contribution (). Inset: Field dependence of around . (b) and vs in the left and right -axes, respectively. Inset: vs and the solid line is the linear fit.
Inorder to demarcate from the total heat capacity , the phonon contribution was first estimated by fitting the high- data by a linear combination of one Debye and three Einstein terms [Debye-Einstein (DE) model] as [26, 49]
The first term in Eq. (6) represents the Debye model,
where, , is the frequency of oscillation, is the universal gas constant, and is the characteristic Debye temperature. The flat optical modes in the phonon spectra is accounted by the second term in Eq. (6), known
as the Einstein term
where, is the characteristic Einstein temperature. In Eq. (6) one can tentatively assign to the low-energy vibrations of the heavy atom (Fe) and the remaining three lighter atoms (Na, P, O) can be accounted separately by the three Einstein terms, respectively. The coefficients , , , and are the weight factors, which take into account the number of atoms per formula unit () and are chosen in such a way that at high temperatures the Dulong-Petit value () is satisfied [50].
The zero field data above K are fitted by Eq. (6) [see, solid curve in Figure 7(a)] and the obtained parameters are , , , , K, K, K, and K. One may notice that the sum of , , , and is close to one, as expected. Finally, the high- fit was extrapolated down to 2 K and subtracted from to get [see, Figure 7(a)]. Figure 7(b) presents and the corresponding magnetic entropy [] vs . The obtained magnetic entropy which saturates above 50 K approaches a value J/mol-K which is close to the expected theoretical value J/mol K for a system.
measured in different applied fields in the low temperature regime is shown in the inset of Figure 7(a). No noticeable shift in is apparent for a field change from 0 to 9 T. Further, well below , follows a behaviour [inset of Figure 7(b)], depicting the dominance of three-dimensional (3D) magnon excitations [23, 51].
Neutron diffraction
The temperature evolution of the NPD pattern at different temperatures (1.6 to 15 K) are shown in Figure 8. New peaks were found to appear below at , 0.8180, 1.1227, 1.2503, and 1.5455 Å. As we reduce the temperature, the intensity of these peaks increases gradually, suggesting the emergence of magnetic reflections. The appearance of new magnetic reflections suggests the AFM nature of the transition [52]. In Figure 8, the -axis is shown in log-scale in order to highlight these magnetic reflections in the low- (momentum transfer vector) regime. The magnetic peaks developed below can be indexed with an ordering vector , which insinuates a commensurate magnetic ordering. The Rietveld refinement of the nuclear pattern at K was performed considering monoclinic crystal structure (space group: ). The best fit of the data [see, Figure 9(a)] yields Å, Å, Å, , and Å. These obtained lattice parameters are in close agreement with the refined values from the powder XRD data at room temperature [37].

Figure 8. (a) Temperature evolution of the neutron powder diffraction (NPD) patterns of Na₃Fe(PO₄)₂ in the low- regime and for temperatures below 15 K. An emergence of magnetic reflections with is evident below .
Figure 9. Rietveld refinement of the neutron powder diffraction data at (a) K and (b) K, considering only nuclear reflections and nuclear + magnetic reflections, respectively. The experimental data are depicted as open black circles. The red solid line represents the calculated pattern. The difference in intensities of experimental and calculated data is shown as blue solid line at the bottom. The allowed Bragg peaks corresponding to the nuclear (green) and magnetic (pink) reflections are shown as vertical bars. Inset of (a) represents the temperature variation of ordered magnetic moment, below . Inset of (b) represents the variation of flipping ratio () in the -range .
In the crystal structure of Na₃Fe(PO₄)₂, magnetic Fe-atom occupies crystallographic site. For determining the magnetic structure the standard representational theory was used [53]. For the propagation vector , the little group can be generated using two symmetry elements, and . Our symmetry analysis shows that the magnetic representation for site can be written in terms of the irreducible representations (IR) as . Here, both IRs ( and ) are one-dimensional and are composed of basis vectors and , respectively (see, Table I).
| IR | Basis Vectors | Fe1 | Fe2 |
Table I. Irreducible representations and basis vectors for the space group with propagation vector , obtained using the BasIreps software. The atoms are defined according to Fe1: and Fe2: .
The observed NPD patterns cannot be fitted with the magnetic structure corresponding to the representations . Whereas, the magnetic structure corresponding to the representation is found to be appropriate to describe the observed NPD patterns. The magnetic structure belonging to the representation corresponds to an antiferromagnetic structure, where moments are aligned in the -plane with dominant component along the crystallographic -axis (Figure 10). Further details of the spin structure is discussed in the next section. From the magnetic structure refinement, we were able to quantify the magnetic moment values along different crystallographic directions. At the lowest temperature ( K), these values are and along the - and -directions, respectively. The temperature variation of total moment () is shown in the inset of Figure 9(a). The observed reduced value of ordered moment with respect to the theoretically expected classical value of for spin-5/2 could be attributed to the effect of low-dimensionality and magnetic frustration and/or presence of covalence effect [32].
Figure 10. (a) Spin structure of NaFe(PO) derived from the neutron powder diffraction at K. (b) A section of the triangular layer (-plane) showing the spin orientation and exchange interactions.
Further, to check the presence of any ferromagnetic-type correlations, we have also carried out the neutron depolarization measurements.
The absence of depolarization is apparent from the flipping ratio () vs plot, where is found to be constant over the entire temperature range () (see, inset of Figure 9(b)). This rules out the possibility of any ferromagnetic-type correlations or spin canting in the system.
Discussion and Summary
Theoretically, for a spin-5/2 chain system, the broad maxima in is expected at [54]. The broad maximum in our experimental data appears at K which corresponds to the intra-chain coupling of K. In order to reproduce the overall shape of the plot, we simulated the data using QMC technique, assuming a spin-chain model with and K. As depicted in the inset of Figure 4(a), the simulated data show a significant deviation from the experimental data, discarding the possibility of spin-chain physics. On the other hand, the fit assuming isotropic triangular lattice [Eq. (4)] reproduces the experimental data very well in the high temperature regime, despite a small spatial anisotropy. As a further test for the isotropic TLA model, we performed the full diagonalization for a spin-5/2 isotropic TLA with K [obtained from the analysis]. As evident from the inset of Figure 4(a), the simulated data using full diagonalization superimpose the experimental data, above . This confirms that the spin-lattice in NaFe(PO) behaves more of a 2D triangular lattice, rather than a 1D spin-chain.
This compound orders antiferromagnetically below K which is clear from the appearance of magnetic reflections in the NPD pattern, -type anomaly in , and slope change in . The magnetic structure refined from the NPD data is shown in Figure 10 which demonstrates a collinear AFM ordering. The spins are coupled ferromagnetically () along the -axis or in the [100] plane which is the direction of the shortest Fe-Fe distance. The AFM interchain coupling () ties the chains into a 2D anisotropic triangular layer in the -plane. Alternatively, one can visualize it as AFM cross chains running along the [110] and [-110] directions which are coupled ferromagnetically in the -plane. Though a non-collinear 120° structure is understood as a hallmark of TLA, some compounds tend to deviate from this spin structure either due to anisotropy or inter-layer coupling [55].
The crystal symmetry plays a key role in deciding the ground states of the compounds in the family. The correspondence between crystal symmetry and exchange couplings is well demonstrated for the series ( Ag, K, Rb) [36]. Here, replacement of A site induces a symmetry change in the CrO octahedra. Therefore, AgCr(VO) (space group: ) is a distorted/anisotropic TLA with a collinear AFM ordering below K, whereas the other compounds ( K, Rb) with higher symmetry (space group: ) are undistorted/isotropic TLAs and are reported show signature of QSL [36]. Several other compounds with higher symmetry in this series also feature isotropic triangular lattices [28, 32, 33, 36, 56]. As pointed out in Figure 1(b), for NaFe(PO), the PO tetrahedra which acts as a medium of interaction among the Fe ions is distorted where the P-O bond distance varies from 1.50 Åto 1.56 Åand the angles are also different. These give rise to the non-equivalent exchange interactions along the edges of the triangle. The anisotropic interactions can also be assessed from the deviation of O-Fe-O bond angles from 90°, typically expected for an isotropic triangular lattice formed by undistorted FeO octahedra. It is also to be noted that despite the shortest Fe-Fe distance, the interaction along the [100] direction is FM. A close inspection of the bond angles reveals that along the [100] direction has the smallest value of 97.13° compared to the other directions which possibly favours parallel alignment. The collinear order in NaFe(PO) may thus be attributed to the distortion in FeO octahedra and inter-layer coupling along the stacking -direction, similar to AgCr(VO).
The obtained from and data and the critical field () corresponding to the metastable transition obtained from the magnetic isotherms are summarized in Figure 11 as a phase diagram. It displays three well defined regions represented as paramagnetic (PM), antiferromagnetic (AFM), and spin-flop (SF) phases. Majority of TLAs feature a magnetization plateau at 1/3 of the saturation magnetization in the magnetic isotherms [57, 58, 59, 60], where the system evolves from a 120° AFM order, which is the true ground state of a Heisenberg TLA [5] to a up-up-down () phase. The phase is stabilized by the magnetic field that introduces axial (Ising) anisotropy and disfavors the noncollinear 120° order. Though our low temperature isotherms illustrate a plateau but it is well below the 1/3 of the saturation magnetization [61]. Further, the magnetic ordering in zero field is a collinear AFM order rather than a noncollinear 120° order. Hence, the 1/3 magnetization plateau is overruled. Moreover, the presence of a easy-plane (XY-type) anisotropy often triggers field emergent phenomenon like SF transition [62] above a critical field , where the moments flip from parallel to perpendicular direction with respect to the applied field. Thus, the metastable/SF transition in Na₃Fe(PO₄)₂ can be attributed to the easy-plane anisotropy in the compound. Na₂BaMnV₂O₈, which belongs to the same series also shows a SF transition at low temperatures [30]. Similar phase diagram is also reported previously in the frustrated spin chain compounds SrCuTe₂O₆ and α-Cu₂As₂O₇ [63].

Figure 11. phase diagram of Na₃Fe(PO₄)₂ drawn from the magnetic isotherm, , and measurements. represents the critical field at which the SF transition takes place. The point denotes the bi-critical point, where the SF transition ends.
In summary, we have demonstrated the ground state properties of the frustrated magnet Na₃Fe(PO₄)₂. Based on the structural data one may anticipate a small anisotropy in the exchange couplings in the triangular unit. However, the analysis of our experimental data emphasizes that the system behaves more like a spin-5/2 isotropic triangular lattice. It stabilizes in a collinear antiferromagnetic ordering below likely due to moderate inter-layer coupling. The occurrence of a field induced SF transition in the low temperature magnetic isotherms insinuates the presence of XY-type anisotropy. A comparison with other members of the family reveals that the bond distances between metal ions and the corresponding angles become more anisotropic with reduced symmetry, leading to a potential structural complexity and exotic ground states.
Acknowledgments
SJS and RN would like to acknowledge SERB, India for financial support bearing sanction Grant No. CRG/2019/000960. SJS 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).
References
- [1]M. F. Collins and O. A. Petrenko, Review/Synthèse: Triangular antiferromagnets, Can. J. Phys 75, 605 (1997); O. A. Starykh, Unusual ordered phases of highly frustrated magnets: a review, Rep. Prog. Phys. 78, 052502 (2015).
- [2]L. Balents, Spin liquids in frustrated magnets, Nature 464, 199 (2010).
- [3]T. Jolicoeur and J. C. Le Guillou, Spin-wave results for the triangular Heisenberg antiferromagnet, Phys. Rev. B 40, 2727 (1989).
- [4]A. V. Chubukov, S. Sachdev, and T. Senthil, Large-S expansion for quantum antiferromagnets on a triangular lattice, J. Phys. Condens. Matter 6, 881 (1994).DOI
- [5]L. Capriotti, A. E. Trumper, and S. Sorella, Long-Range Néel Order in the Triangular Heisenberg Model, Phys. Rev. Lett. 82, 3899 (1999).DOI
- [6]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).
- [7]A. L. Chernyshev and M. E. Zhitomirsky, Spin waves in a triangular lattice antiferromagnet: Decays, spectrum renormalization, and singularities, Phys. Rev. B 79, 144416 (2009).
- [8]H. Kawamura and S. Miyashita, Phase Transition of the Two-Dimensional Heisenberg Antiferromagnet on the Triangular Lattice, J. Phys. Soc. Jpn 53, 4138 (1984).
- [9]S. Miyashita and H. Kawamura, Phase Transitions of Anisotropic Heisenberg Antiferromagnets on the Triangular Lattice, J. Phys. Soc. Jpn. 54, 3385 (1985).DOI
- [10]P.-E. Melchy and M. E. Zhitomirsky, Interplay of anisotropy and frustration: Triple transitions in a triangular-lattice antiferromagnet, Phys. Rev. B 80, 064411 (2009).DOI
- [11]R. H. Clark and W. G. Moulton, Cs¹³³ and Cl³⁵ NMR in Antiferromagnetic CsNiCl₃, Phys. Rev. B 5, 788 (1972).
- [12]H. Kadowaki, K. Ubukoshi, and K. Hirakawa, Neutron Scattering Study of Successive Phase Transitions in Triangular Lattice Antiferromagnet CsNiCl₃, J. Phys. Soc. Jpn. 56, 751 (1987).
- [13]Y. Ajiro, T. Inami, and H. Kadowaki, Magnetic Phase Transition of the Triangular Lattice Antiferromagnet CsMnI₃, J. Phys. Soc. Jpn. 59, 4142 (1990).
- [14]A. Harrison, M. F. Collins, J. Abu-Dayyeh, and C. V. Stager, Magnetic structures and excitations of CsMnI₃: A one-dimensional Heisenberg antiferromagnet with easy-axis anisotropy, Phys. Rev. B 43, 679 (1991).pubmed.ncbi.nlm.nih.gov/9996259
- [15]K. M. Ranjith, K. Brinda, U. Arjun, N. G. Hegde, and R. Nath, Double phase transition in the triangular antiferromagnet Ba₃CoTa₂O₉, J. Phys.: Cond. Mat. 29, 115804 (2017).
- [16]M. Lee, E. S. Choi, X. Huang, J. Ma, C. R. Dela Cruz, M. Matsuda, W. Tian, Z. L. Dun, S. Dong, and H. D. Zhou, Magnetic phase diagram and multiferroicity of Ba₃MnNb₂O₉: A spin-½ triangular lattice antiferromag- net with weak easy-axis anisotropy, Phys. Rev. B 90, 224402 (2014); M. Lee, J. Hwang, E. S. Choi, J. Ma, C. R. Dela Cruz, M. Zhu, X. Ke, Z. L. Dun, and H. D. Zhou, Series of phase transitions and multiferroicity in the quasi-two-dimensional spin-1/2 triangular-lattice antiferromagnet Ba3CoNb2O9, Phys. Rev. B 89, 104420 (2014).
- [17]S. Miyashita, Magnetic Properties of Ising-Like Heisenberg Antiferromagnets on the Triangular Lattice, J. Phys. Soc. Jpn. 55, 3605 (1986).DOI
- [18]R. Rawl, L. Ge, Z. Lu, X. Evenson, C. R. Dela Cruz, Q. Huang, M. Lee, E. S. Choi, M. Mourigal, H. D. Zhou, and J. Ma, Ba8MnNb6O24: A model two-dimensional spin-1/2 triangular-lattice antiferromagnet, Phys. Rev. Materials 3, 054412 (2019).
- [19]S.-W. Cheong and M. Mostovoy, Multiferroics: a magnetoelectric twist for ferroelectricity, Nat. Mater. 6, 13 (2007).pubmed.ncbi.nlm.nih.gov/17199121
- [20]Z. Zhu, P. A. Maksimov, S. R. White, and A. L. Chernyshev, Disorder-Induced Mimicry of a Spin Liquid in YbMgGaO4, Phys. Rev. Lett. 119, 157201 (2017).DOI
- [21]Z. Zhu, P. A. Maksimov, S. R. White, and A. L. Chernyshev, Topography of spin liquids on a triangular lattice, Phys. Rev. Lett. 120, 207203 (2018).DOI
- [22]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).
- [23]K. Somesh, Y. Furukawa, G. Simutis, F. Bert, M. Prinz-Zwick, N. Büttgen, A. Zorko, A. A. Tsirlin, P. Mendels, and R. Nath, Universal fluctuating regime in triangular chromate antiferromagnets, Phys. Rev. B 104, 104422 (2021).DOI
- [24]D. Yamamoto, T. Sakurai, R. Okuto, S. Okubo, H. Ohta, H. Tanaka, and Y. Uwatoko, Continuous control of classical-quantum crossover by external high pressure in the coupled chain compound CsCuCl3, Nat. Commun. 12, 4263 (2021).
- [25]K. M. Ranjith, R. Nath, M. Majumder, D. Kasinathan, M. Baenitz, L. Keller, Y. Skourski, M. Baenitz, and A. A. Tsirlin, Commensurate and incommensurate magnetic order in spin-1 chains stacked on the triangular lattice in Li2NiW2O8, Phys. Rev. B 94, 014415 (2016).
- [26]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-1/2 antiferromagnet BaNa2Cu(VO4)2, Phys. Rev. B 103, 064413 (2021).
- [27]N. E. Amuneke, D. E. Gheorghe, B. Lorenz, and A. Möller, Synthesis, Crystal Structure, and Physical Properties of BaAg2Cu(VO4)2: A New Member of the S = 1/2 Triangular Lattice, Inorg. Chem. 50, 2207 (2011).DOI
- [28]N. E. Amuneke, J. Tapp, C. R. de la Cruz, and A. Möller, Experimental realization of a unique class of compounds: XY-antiferromagnetic triangular lattices, KAg2Fe[VO4]2 and RbAg2Fe[VO4]2, with ferroelectric ground states, Chem. Mater. 26, 5930 (2014).DOI
- [29]A. Möller, N. E. Amuneke, P. Daniel, B. Lorenz, C. R. de la Cruz, M. Gooch, and P. C. W. Chu, AAg2M[VO4]2 (A = Ba, Sr; M = Co, Ni): A series of ferromagnetic insulators, Phys. Rev. B 85, 214422 (2012).
- [30]G. Nakayama, S. Hara, H. Sato, Y. Narumi, and H. Nojiri, Synthesis and magnetic properties of a new series of triangular-lattice magnets, Na2BaMV2O8(M = Ni, Co, and Mn), J. Phys.: Condens. Matter 25, 116003 (2013).
- [31]A. Reuß, V. Ksenofontov, J. Tapp, D. Wulferding, P. Lemmens, M. Panthöfer, and A. Möller, Screw-Type Motion and Its Impact on Cooperativity in BaNa2Fe[VO4]2, Inorg. Chem. 57, 6300 (2018).
- [32]L. D. Sanjeewa, V. O. Garlea, M. A. McGuire, C. D. McMillen, and J. W. Kolis, Magnetic Ground State Crossover in a Series of Glaserite Systems with Triangular Magnetic Lattices, Inorg. Chem. 58, 2813 (2019).
- [33]S. Lee, R. Klauer, J. Menten, W. Lee, S. Yoon, H. Luetkens, P. Lemmens, A. Möller, and K.-Y. Choi, Unconventional spin excitations in the S = 3/2 triangular antiferromagnet RbAg2Cr[VO4]2, Phys. Rev. B 101, 224420 (2020).
- [34]A. A. Tsirlin, A. Möller, B. Lorenz, Y. Skourski, and H. Rosner, Superposition of ferromagnetic and antiferromagnetic spin chains in the quantum magnet BaAg2Cu[VO4]2, Phys. Rev. B 85, 014401 (2012).DOI
- [35]S. Lee, C. H. Lee, A. Berlie, A. D. Hillier, D. T. Adroja, R. Zhong, R. J. Cava, Z. H. Jang, and K.-Y. Choi, Temporal and field evolution of spin excitations in the disorder-free triangular antiferromagnet Na2BaCo(PO4)2, Phys. Rev. B 103, 024413 (2021).DOI
- [36]J. Tapp, C. R. dela Cruz, M. Bratsch, N. E. Amuneke, L. Postulka, B. Wolf, M. Lang, H. O. Jeschke, R. Valentí, P. Lemmens, and A. Möller, From magnetic order to spin-liquid ground states on the S = 3/2 triangular lattice, Phys. Rev. B 96, 064404 (2017).
- [37]V. Morozov, B. Lazoryak, A. Malakho, K. Pokholok, S. Polyakov, and T. Terekhinа, The Glaserite-like Structure of Double Sodium and Iron Phosphate Na3Fe(PO4)2, J. Solid State Chem. 160, 377 (2001); M. Belkhiria, S. Laaribi, A. Ben Hadj Amara, and M. Ben Amara, Structure of Na3Fe(PO4)2 from powder X-ray data, Ann. Chim. Sci. Mater. 23, 117 (1998).DOI
- [38]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); 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).
- [39]J. Rodríguez-Carvajal, Recent advances in magnetic structure determination by neutron powder diffraction, Physica B: Condensed Matter 192, 55 (1993).
- [40]S. Todo and K. Kato, Cluster algorithms for general-S quantum spin systems, Phys. Rev. Lett. 87, 047203 (2001).
- [41]A. Albuquerque, F. Alet, P. Corboz, P. Dayal, A. Feiguin, S. Fuchs, L. Gamper, E. Gull, S. Gürterl, A. Honecker, R. Igarashi, M. Körner, A. Kozhevnikov, A. Läuchli, S. Manmana, M. Matsumoto, I. McCulloch, F. Michel, R. Noack, G. Pawłowski, L. Pollet, T. Pruschke, U. Schollwöck, S. Todo, S. Trebst, M. Troyer, P. Werner, and S. Wessel, The alps project release 1.3: Open-source software for strongly correlated systems, J. Magn. Magn. Mater 310, 1187 (2007).DOI
- [42]S. S. Islam, K. M. Ranjith, M. Baenitz, Y. Skourski, A. A. Tsirlin, and R. Nath, Frustration of square cupola in Sr(TiO)Cu(PO4)2, Phys. Rev. B 97, 174432 (2018).
- [43]P. W. Selwood, Magnetochemistry (Read Books Ltd, 2013); Mendelsohn, Biggs, and Mann, Hartree-fock diamagnetic susceptibilities, Phys. Rev. A 2, 1130 (1970).
- [44]C. Domb and A. Miedema, Chapter VI Magnetic Transitions, J. Low Temp. Phys. 4, 296 (1964).DOI
- [45]C. Delmas, G. Le Flem, and C. Foussier, and P. Hagenmuller, Etude comparative des proprietes magnetiques des oxydes lamellaires ACrO₂ (A = Li, Na, K)—II: Calcul des integrales d’echange, J. Phys. Chem. Solids 39, 55 (1978).
- [46]H. J. Schmidt, A. Lohmann, and J. Richter, Eighth-order high-temperature expansion for general heisenberg hamiltonians, Phys. Rev. B 84, 104443 (2011).DOI
- [47]R. Dingle, M. E. Lines, and S. L. Holt, Linear-Chain Antiferromagnetism in [(CH₃)₄N] [MnCl₃], Phys. Rev. 187, 643 (1969).
- [48]H. Kawamura and S. Miyashita, Phase Transition of the Heisenberg Antiferromagnet on the Triangular Lattice in a Magnetic Field, J. Phys. Soc. Jpn. 54, 4530 (1985).DOI
- [49]E. S. R. Gopal, Specific heats at low temperatures (Springer Science & Business Media, 2012).
- [50]R. K. Fitzgerald and F. H. Verhoek, The low-temperature law of Dulong and Petit, J. Chem. Educ. 37, 545 (1960).
- [51]P. Bag, N. Ahmed, V. Singh, M. Sahoo, A. A. Tsirlin, and R. Nath, Low-dimensional magnetism of BaCuTe₂O₆, Phys. Rev. B 103, 134416 (2021).DOI
- [52]V. K. Anand, L. Opherden, J. Xu, D. T. Adroja, A. D. Hillier, P. K. Biswas, T. Herrmannsdörfer, M. Uhlarz, J. Hornung, J. Wosnitza, E. Canévet, and B. Lake, Evidence for a dynamical ground state in the frustrated pyrochlore Tb₂Hf₂O₇, Phys. Rev. B 97, 094402 (2018); S. S. Islam, V. Singh, K. Somesh, P. K. Mukharjee, A. Jain, S. M. Yusuf, and R. Nath, Unconventional superparamagnetic behavior in the modified cubic spinel compound LiNi₀.₅Mn₁.₅O₄, Phys. Rev. B 102, 134433 (2020).DOI
- [53]E. Bertaut, Spin Configurations of Ionic Structures: Theory and Practice, Magnetism (GT Rado and H. Shul, eds.) 3, 149 (1963).DOI
- [54]L. J. D. Jongh and A. R. Miedema, Experiments on simple magnetic model systems, Adv. Phys. 50, 947 (2001).
- [55]K. M. Ranjith, R. Nath, M. Skoulatos, L. Keller, D. Kashnathan, Y. Skourski, and A. A. Tsirlin, Collinear order in the frustrated three-dimensional spin−1/2 antiferromagnet Li₂CuWO₈, Phys. Rev. B 92, 094426 (2015); K. M. Ranjith, M. Majumder, M. Baenitz, A. A. Tsirlin, and R. Nath, Frustrated three-dimensional antiferromagnet Li₂CuWO₈: ⁷Li NMR and the effect of nonmagnetic dilution, Phys. Rev. B 92, 024422 (2015).
- [56]C. Wellm, W. Rosner, J. Zeisner, A. Alfonsov, R. Zhong, R. J. Cava, A. Savoyant, R. Hayn, J. van den Brink, B. Büchner, O. Janson, and V. Kataev, Frustration enhanced by Kitaev exchange in a Jₑff=1/2 triangular antiferromagnet, Phys. Rev. B 104, L100420 (2021).
- [57]A. V. Chubukov and D. I. Golosov, Quantum theory of an antiferromagnet on a triangular lattice in a magnetic field, J. Phys. Condens. Matter. 3, 69 (1991).DOI
- [58]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); T. Susuki, N. Kurita, T. Tanaka, H. Nojiri, A. Matsuo, K. Kindo, and H. Tanaka, Magnetization Process and Collective Excitations in the S=1/2 Triangular-Lattice Heisenberg Antiferromagnet Ba₃CoSb₂O₉, Phys. Rev. Lett. 110, 267201 (2013).
- [59]A. I. Smirnov, H. Yashiro, S. Kimura, M. Hagiwara, Y. Narumi, K. Kindo, A. Kikkawa, K. Katsumata, A. Y. Shapiro, and L. N. Demianets, Triangular lattice antiferromagnet RbFe(MoO₄)₂ in high magnetic fields, Phys. Rev. B 75, 134412 (2007).
- [60]J. Hwang, E. S. Choi, F. Ye, C. R. Dela Cruz, Y. Xin, H. D. Zhou, and P. Schlottmann, Successive Magnetic Phase Transitions and Multiferroicity in the Spin-One Triangular-Lattice Antiferromagnet Ba₃NiNb₂O₉, Phys. Rev. Lett. 109, 257205 (2012).pubmed.ncbi.nlm.nih.gov/23368494
- [61]L. Seabra, T. Momoi, P. Sindzingre, and N. Shannon, Phase diagram of the classical Heisenberg antiferromagnet on a triangular lattice in an applied magnetic field, Phys. Rev. B 84, 214418 (2011).DOI
- [62]O. A. Petrenko, G. Balakrishnan, M. R. Lees, D. McK. Paul, and A. Hoser, High-magnetic-field behavior of the triangular-lattice antiferromagnet CuFeO₂, Phys. Rev. B 62, 8983 (2000).DOI
- [63]N. Ahmed, A. A. Tsirlin, and R. Nath, Multiple magnetic transitions in the spin−1/2 chain antiferromagnet SrCuTe₂O₆, Phys. Rev. B 91, 214413 (2015); Y. C. Arango, E. Vavilova, M. Abdel-Hafiez, O. Janson, A. A. Tsirlin, H. Rosner, S.-L. Drechsler, M. Weil, G. Nénert, R. Klinger, O. Volkova, A. Vasiliev, V. Kataev, and B. Büchner, Magnetic properties of the low-dimensional spin−1/2 magnet α-Cu₂As₂O₇, Phys. Rev. B 84, 134430 (2011).






