Magnetic and magnetotransport properties of Fe₂P nanocrystallites via a solvothermal route

Abstract

Hexagonal nanocrystalline Fe₂P has been synthesized via a solvothermal route at a relatively low temperature of 180 °C. X-Ray diffraction (XRD), X-ray photoelectron spectra (XPS) and transmission electron microscopy (TEM) were used to investigate the phase, composition, and morphology of the final crystallized product. The particles with a uniform hexagonal structure are stable in air and their size is in the range of 100–120 nm. A magnetoresistance effect is observed at comparatively low temperatures, which originates from the spin polarized tunneling effect. The super-paramagnetic state transformation can be found at the blocking temperature in Fe₂P nanoparticles, further confirmed by electron spin resonance (ESR) experiments. Metamagnetic behavior is also observed in nanocrystalline Fe₂P with variation of field because of the coexistence of ferromagnetic and super-paramagnetic states.

---

Introduction

Transition metals form a large number of metal phosphide compounds comprising over 125 binary and over 80 ternary substances because of the diversity in metal identity and metal/phosphorus stoichiometry.¹ Generally, bulk band gap energies for most of these transition metal phosphides fall into the range of near 0 to over 3 eV similarly to semiconductors, and many metal phosphides display unique physical properties such as phosphorescent, magnetic, electronic and catalytic properties.^2–4 Owing to their important properties and potential applications in many fields, such as semiconductors, luminescent devices, electronic components and hydro-processing catalysts,^2,5–7 transition metal phosphides have attracted considerable interest in the past decades. Nevertheless, the magnetoresistance (MR) effect of transition metal phosphides has not been reported except for the alloy Fe_1.96Ru_0.04P.⁸ Since magnetoresistive materials are the active components of magnetic field sensors and read heads for computer hard drives because of their dependence of electric resistance on magnetic field,^9,10 it is worth investigating the origin of magnetoresistance and exploring new transition metal phosphides. As a excellent magnetic material, the dimetal iron phosphide Fe₂P is a well-known ferromagnet with a first order para- to ferromagnetic transition at 217 K.¹¹ Considering that the d electrons of Fe are partially polarized,¹² when the size of Fe₂P is reduced to a few hundred nanometers, the inter-grain spin polarized tunneling is enhanced,¹³ making the magnetoresistance effect possible. Meanwhile, the magnetic properties of nanocrystalline particles are different compared with bulk Fe₂P, stemming from its reduced dimensions.

Conventional synthesis of metal phosphides usually involves the direct reaction of the appropriate elements for prolonged periods at high temperatures,¹⁴ reaction of phosphine with metals or metal oxides, reduction of metal phosphates by carbon, electrolysis of molten metal phosphate salts,¹⁵ metal organic chemical vapor deposition, and the high-temperature self-propagating route.^16–18 These procedures typically require high reaction temperatures and long annealing periods to produce crystalline materials. However, most of these methods are only suitable for preparing bulk materials, and the overriding synthetic challenge is still the preparation of nanocrystalline transition metal phosphides with precise metal/phosphorus stoichiometry. Recently, Lukehart et al.¹⁹ reported a low-temperature reaction, which affords crystalline nanoclusters of transition metal phosphides in a silica xerogel matrix by the decomposition of molecular precursors. Clark et al.²⁰ also reported a similar route for preparing a hydroprocessing catalyst. Furthermore, solution chemical synthesis techniques have also been developed for synthesizing nanocrystalline Co₂P, Ni₂P and FeP particles under rather mild temperatures.^21–23 Among these techniques, the solvothermal method promises a single step of low-temperature synthesis, superior composition and morphological control.

Though nanocrystalline Fe₂P was first reported by Lukehart et al.²⁴ in 1995, the nanocrystallites were dispersed throughout a silica xerogel matrix and formed a nanocomposite, which cannot afford an opportunity to investigate the unique properties of nanocrystalline Fe₂P itself. In this paper, through a mild solvothermal method, we have synthesized, for the first time, a nanocrystalline dimetal iron phosphide that is stable in air. The magnetic and magnetotransport properties of Fe₂P have also been investigated.

Experimental

In the synthesis of Fe₂P nanocrystallites, freshly obtained FeCl₂·6H₂O microcrystals and analytically pure yellow phosphorus were used as the starting materials in a molar ratio of 1∶5. FeCl₂·6H₂O crystals were prepared by dissolving appropriate amounts of FeCl₃ and Fe in aqueous HCl, subsequently filtering and then cooling in a refrigerator. The typical synthesis procedure is as follows. 1.24 g (10 mmol) of yellow phosphorus, washed free of water by using absolute ethanol and toluene successively, and 0.470 g (2 mmol) FeCl₂·6H₂O were put into a 100 ml Teflon lined autoclave, which was then filled with ethylenediamine up to 85% of its capacity. The autoclave was maintained at 150–200 °C for 48 h and then cooled to room temperature naturally. The black precipitate was collected and washed with benzene, distilled water, and absolute ethanol several times in sequence. The final product weighed 0.120 g after drying in vacuum at 80 °C for 12 h. The yield was about 84%.

The obtained products were characterized by D/max-2000 rotating powder diffractometer (Rigaku, Japan) using Cu-Kα radiation (λ = 1.5418 Å). X-Ray diffraction (XRD) patterns were obtained at a scanning rate of 8° min⁻¹ over the 2θ range 20–80°. The purity and composition of the sample were determined by X-ray photoelectron spectra (XPS), which were carried out in an ion-pumped chamber (evacuated to 2 × 10⁻⁹ Torr) of an Escalab5 (UK) spectrometer, employing Mg-Kα radiation (binding energy (BE) = 1253.6 eV). Moreover, the morphology and size were both examined by transmission electron microscopy (TEM, H-800, Hitachi, Japan). Resistance was measured in the temperature range of 300–5 K on a MagLab System 2000 (Oxford, UK) using a standard dc four-probe technique with the target products pressed into pellets with diameter of 6 mm. Magnetization measurements were performed on a commercial superconducting quantum interference device (SQUID). Electron spin resonance (ESR) experiments were conducted on an ER-200D spectrometer (Bruker, Germany) at a microwave frequency of 9.5 GHz in the temperature range 110–300 K.

Results and discussion

Fe₂P crystallizes in the hexagonal phase with space group P6̄2m. This structure presented in Fig. 1 shows two different metal sites: tetrahedral Fe (3f) in position (0.2568, 0, 0) with four P atoms as nearest neighbors (NN) and pyramidal Fe (3g) in position (0.5962, 0, ½) with five P atoms as NN.²⁵ In the XRD patterns of Fig. 2, all peaks of Fe₂P obtained at 180 °C can be well indexed to a single hexagonal cell with a = 5.88 Å, c = 3.43 Å. These cell parameters are in good agreement with the JCPDS card file 33-0670.²⁶ Moreover, the relative intensity of (111) to (002) obtained for the as-prepared nanocrystals is 6.0, which is higher than that of the bulk material, indicating the existence of somewhat preferential orientation in the Fe₂P nanocrystals. In the solvothermal process, we have done some related experiments to explore the possible mechanism in the formation of iron phosphide. During the whole reaction process, ethylenediamine not only acted as an N-chelating ligand to chelate with the iron ions and form stable complexes, but also as a basic solvent, which activates the reductive ability of yellow phosphorus. Thus iron ions are easily reduced to active atoms. Meanwhile, yellow phosphorus also disperses homogeneously in ethylenediamine and is rather active under the solvothermal conditions. Once generated, the metallic atoms will immediately combine with elemental phosphorus to form the corresponding metal phosphide. In this solvothermal process, the possible reactions in the formation of iron phosphide can be described as follows:²¹

P₄ + 14 en + 16 H₂O → 14 H₂en²⁺ + 4 HPO₄²⁻ + 20 e (1)

Fe²⁺ + 3 en → Fe(en)₃²⁺ (2)

Fe(en)₃²⁺ + 2 e → Fe + 3 en (3)

8 Fe + P₄ → 4 Fe₂P (4)

Fig. 1 Crystal structure of hexagonal Fe₂P. Two inequivalent iron positions, tetrahedral and pyramidal, are distinguished and their nearest atomic neighbors are shown.

Fig. 2 The XRD patterns of Fe₂P nanoparticles prepared under different temperatures.

During the reaction process, an appropriate amount of water is necessary. It is concluded that the appropriate reaction temperature for synthesizing nanocrystalline Fe₂P is about 180 °C, above which the FeP appears, consistent with the XRD peaks below 40°, as observed in Fig. 2.²⁷ If the reaction temperature is below 180 °C, the final products are only amorphous iron and phosphorus. Since the yellow phosphorus is in excess, Fe₃P does not form in the reaction process.²⁸

Fig. 3 displays the XPS results of as-prepared Fe₂P. The binding energy (BE) of Fe 2p is about 707.4 eV, close to the BE of elemental Fe, which is nearly 707.0 eV. This indicates that the active valence of Fe is comparatively low and almost zero, consistent with the structure of Fe₂P. Moreover, the BE peaks of 2s, 2p for phosphorus and 3s, 3p for iron were all observed in Fig. 3 and accordant with the normal values. The BE value of P 2p is only 130.0 eV, close to the BE value of elemental phosphorus (129.1 eV), indicating that the nanocrystalline Fe₂P is stable in air and is not oxidized. The BE peaks indexed as O KLL, O 1s, N 1s and C 1s in Fig. 3 originate from a small amount of CO₂ and N₂ molecules in the atmosphere. From the integral peak areas of Fe and P in the XPS spectra, the product is rather pure with molar ratio Fe∶P = 1.985∶1, which is close to the chemical formula of Fe₂P. No obvious impurities such as chloride or phosphate could be detected in the sample, indicating that the level of impurities is lower than the resolution limit of XPS (1–5 atom%). From the transmission electron microscopy (TEM) images shown in Fig. 4, we can see that the size of Fe₂P nanoparticles ranges from about 100 to 120 nm, with a uniform hexagonal shape, coincident with the trend of preferential orientation observed in Fig. 2. The electron diffraction (ED) pattern in the inset reveals that the Fe₂P nanoparticles are polycrystalline and externally well crystallized.

Fig. 3 XPS analysis of as-prepared Fe₂P product.

Fig. 4 TEM micrograph for Fe₂P nanoparticles.

Fig. 5 displays the temperature dependence of resistance and MR ratio (defined as (R_0 T − R_5 T)/R_5 T) for nanocrystalline Fe₂P particles. It is clear that the resistance values dramatically increase with the temperature, showing a metallic behavior. Compared with the bulk material, the influence of the scattering of carriers at grain boundaries on the temperature dependence of electric conductivity ought to be considered, which may lead to the enhancement of resistivity for nanocrystals. The MR% values are monotonously decreased along with increasing temperature, and then slightly increased above 245 K, which is very close to the ferro–paramagnetic transition temperature confirmed from Fig. 8. As shown in Fig. 6 (MR ratio is redefined as (R_0 T − R_B T)/R_B T, where B is the external magnetic field), we measured the evolution of MR% with external field H at a temperature of 5 K. The MR ratio is appropriately linearly increased dependent on the external field and reaches about 1.2% in a field of 5 T. This MR effect is mainly caused by the spin polarized tunneling of polarized free electron on the d orbitals of Fe. Moreover, the enhanced inter-grain spin polarized tunneling in nanocrystalline Fe₂P also contributes to the MR enhancement.¹³

Fig. 5 Temperature dependence of resistance and MR ratio obtained at 5 T for nanocrystalline Fe₂P obtained at 180 °C.

Fig. 6 Field dependence of MR values for Fe₂P at 5 K.

The magnetic characterization was carried out by several experiments. For single crystal Fe₂P, the measured atomic magnetic moments of tetrahedral Fe (3f) and pyramidal Fe (3g) sites are 0.59 and 2.22 μ_B respectively, resulting in a net moment of 2.81 μ_B.²⁹ From the hysteresis loop at 5 K of Fig. 7, we find that the magnetization of Fe₂P is still not saturated even under an external field of 50 kG, and the unsaturated magnetization moment is 32.1 emu g⁻¹ (0.82 μ_B/f.u.), which is much lower than the saturated moment of bulk Fe₂P.²⁹ The rather unusual shape of this hysteresis loop possibly indicates this material is comprised of a combination of paramagnetic and ferromagnetic constituents. Resolution of this complexity will result from more detailed measurements of the temperature dependence of the sample magnetization. Fig. 8 shows the temperature dependence of magnetization in the zero-field-cooled (ZFC) and field-cooled (FC) processes with an applied low field of 100 and 500 Oe. In the ZFC mode, the magnetization increases at first, and then decreases with an increase of temperature. The temperature, at which the maximum ZFC magnetization occurs, is characterized as the blocking temperature (T_B ≈ 248 K) for the super-paramagnetic transformation.³⁰ Meanwhile, the magnetization data as a function of temperature taken at 100 and 500 Oe in FC mode were used to determine the Curie temperature (T_C). From the maximum value in the dM/dT vs. T curve, the obtained T_C is about 250 K,³¹ different from the value of the bulk material.¹¹ Usually, the blocking temperature is sensitive to the volume of single particles and magnetic anisotropy constant. The magnetic anisotropy constant (K), was deduced from the blocking temperature using the equation K = 25k_bT_B/V, where k_b is the Boltzman constant and V is the volume of the nanoparticle.³² When the size of nanoparticles increases, T_B will approach T_C and disappear with further increase of particle size. As for the Curie temperature, it is related to the size and crystallization of particles as well as the magnitude of the applied external field. From Fig. 9, we can judge that nanocrystalline Fe₂P is a metamagnetic compound. The value of dM/dH first increases with H, reaches a maximum, and then decreases with a further increase of H. The given field, at which the maximum value is reached, is the critical field related to the magnetic transition. This magnetic transition may be attributed to the transformation from the super-paramagnetic to ferromagnetic state. In order to further understand the complicated magnetic behavior of this compound, we investigated ESR signals of the Fe₂P nanoparticles. ESR detects the power P absorbed by the sample from the transverse magnetic microwave field as a function of the static magnetic field H. As shown in Fig. 10(a), the sample shows gradual X-band paramagnetic (PM) line transformations at above 260 K, which is consistent with the T_B value in Fig. 8. Within the whole paramagnetic regime above 285 K, the spectrum consists of a broad, exchange narrowed resonance line, which is well fitted by a Dysonian line shape with a g factor of 1.995.³³ As observed in Fig. 10(b), a ferromagnetic (FM) signal appears at 110 K and gradually disappears when the temperature approaches 230 K, consistent with the value of T_C obtained from the magnetization (Fig. 8). When the temperature is lowered to 230 K, an additional small peak occurs around H = 1000 G. This additional peak is related to the ferromagnetism.^34,35 When the temperature is further lowered, the intensity of the additional peak gradually increases, accompanied by a gradual decrease of the PM peak. When the FM state occurs, the remaining PM regime is actually surrounded by an internal magnetic field induced by the FM phase. Therefore, at a lower applied magnetic field, the sum of the internal and applied fields can satisfy the required magnitude of the resonance field. Thus, the location of the PM peak is shifted to a lower field with decreasing temperature and an additional peak appears at lower field. On the other hand, temperature-dependent ESR signals of PM lines with g = 1.996 also occur below 285 K, different from the PM lines above 285 K. Since the super-paramagnetic state coexists with the ferromagnetic state below T_B (as deduced from Fig. 8), this newly PM state may be due to the contribution of the super-paramagnetism of nanocrystalline Fe₂P. The two magnetic phases, FM and PM, coexist from 110 to 230 K, according to the ESR spectra, providing a direct evidence for the coexistence of FM and super-paramagnetic states in nanocrystalline Fe₂P.

Fig. 7 Hysteresis loop of nanocrystalline Fe₂P at 5 K.

Fig. 8 ZFC and FC magnetization as a function of temperature in applied fields of 100 and 500 Oe.

Fig. 9 Typical M–H and dM/dH–H curves of Fe₂P at 5 K.

Fig. 10 ESR signals at various temperatures for nanocrystalline Fe₂P: (a) from 300 to 255 K (5 K intervals), (b) from 255 to 110 K (5 K intervals in the range 255–200 K, 10 K intervals in the range 200–110 K).

Conclusions

We have synthesized nanosized Fe₂P particles through a mild solvothermal route based on the direct reaction of the metal halide with yellow phosphorus. The magnetoresistance effect is observed and reaches 1.2% at 5 K in a field of 50 kG, originating from the spin polarized tunneling effect. The super-paramagnetic state confirmed by magnetic measurements affects the metamagnetic behavior of nanocrystalline Fe₂P, and the coexistence of ferromagnetic and super-paramagnetic states is also observed from 230 to 110 K according to ESR spectroscopy.

Acknowledgements

The authors would like to acknowledge the support of MOST of China (G19980613), NSFC (Nos. 50272006, 20221101 & 20023005), Training Project for Doctoral Student of MOE, and the Founder Foundation of PKU.

References

1. H. G. von Schnering and W. Honle, Encyclopedia of Inorganic Chemistry, John Wiley & Sons, Chichester, UK, 1994, vol. 6, p. 3106.
2. B. Arinsson, T. Landstron and S. Rundquist, Borides, Silicides and Phosphides, John Wiley & Sons, New York, 1965.
3. R. Wappling, L. Haggstrom, T. Ericsson, S. Devanarayan, E. Karlsson, B. Carlsson and S. Rundquist, J. Solid State Chem., 1975, 13, 258.
4. C. Stinner, R. Prins and Th. Weber, J. Catal., 2001, 202, 187.
5. V. B. Chernogorenko, S. V. Muchnik, K. A. Lynchak and Z. A. Klimak, Mater. Res. Bull., 1981, 16, 1.
6. A. Catalano and R. B. Hall, J. Phys. Chem. Solids, 1980, 41, 635.
7. M. Jian, J. L. Rico Cerda and R. Prins, Catal. Lett., 1995, 35, 193.
8. T. Oki, T. Sato, Y. Uwatoko, H. Fujii and A. V. Andreev, J. Alloys Compd., 2000, 306, 1.
9. J. Brig, T. C. Anthony and J. Nickel, Mater. Res. Soc. Bull., 1996, 21, 23.
10. J. M. Daughton, J. Magn. Magn. Mater., 1999, 192, 334.
11. H. Fujii, Y. Uwatoko, K. Motoya, Y. Ito and T. Okamoto, J. Phys. Soc. Jpn., 1988, 57, 2143.
12. M. B. Stearn, Physica, 1977, 91(B + C), 37.
13. H. Y. Hwang, S. W. Cheong, N. P. Ong and B. Batlogg, Phys. Rev. Lett., 1996, 77, 2041.
14. Comprehensive Inorganic Chemistry, ed. J. C. Bailer, H. J. Emelius, R. Nyholm and A. F. Trotman-Dickenson, Pergamon, Oxford, 1973, vol. 2.
15. J. R. van Wazer, Phosphorus and Its Compounds, Interscience, New York, 1958, vol. 1.
16. P. R. Bonneau, R. F. Jarris and R. B. Kaner, Nature, 1991, 349, 510.
17. I. P. Parkin, Chem. Soc. Rev., 1996, 199.
18. J. Gopalakrishnan, S. Pandey and K. K. Rangan, Chem. Mater., 1997, 9, 2113.
19. C. M. Lukehart, B. Milne Stephen and S. R. Stock, Chem. Mater., 1998, 10, 903.
20. Xian-Qin Wang, P. Clark and S. Ted Oyama, J. Catal., 2002, 208, 321.
21. Y. Xie, H. L. Su, X. F. Qian, X. M. Liu and Y. T. Qian, J. Solid State Chem., 2000, 149, 88.
22. Shu-Hong Yu, Jian Yang, Yong-Sheng Wu, Zhao-Hui Han, Lei Shu, Yi Xie and Yi-Tai Qian, J. Mater. Res., 1998, 13, 3365.
23. Yun-Le Guo, Fan Guo, Yi-Tai Qian, Hua-Gui Zheng and Zi-Ping Yang, Mater. Res. Bull., 2002, 37, 1101.
24. C. M. Lukehart, B. Milne Stephen, S. R. Stock, R. D. Shull and E. Wittig James, Mater. Sci. Eng. A, 1995, 204, 176.
25. H. Fujii, S. Komura, T. Takeda, T. Okamoto, Y. Ito and J. Akimistsu, J. Phys. Soc. Jpn., 1979, 51, 1616.
26. B. Carlson, J. Solid State Chem., 1973, 8, 57.
27. S. Rundqvist and P. C. Nawapong, Acta Chem. Scand., 1965, 19, 1006.
28. Fasiska and L. Zwell, Trans. Am. Inst. Min. Metall. Pet. Eng., 1967, 239, 924.
29. A. Koumina, M. Bacmann, D. Fruchart, J. L. Soubeyroux, P. Wolfers, J. Tobola, S. Kaprzyk, S. Niziol, M. Mesnaoui and R. Zach, Ann. Chim. Sci. Mater., 1998, 23, 177.
30. R. W. Li, H. Xiong, J. R. Sun, Q. A. Li, Z. H. Wang, J. Zhang and B. G. Shen, J. Phys.: Condens. Mater., 2001, 13, 141.
31. A. J. P. Meyer and M. C. Cadeville, J. Phys. Soc. Jpn. Suppl. (Proc. Int. Conf. Mag. Crys. Vol. 1), 1962, 17, 223.
32. S. J. Park, S. Kim, S. Lee, Z. G. Khim, K. Char and T. Hyeon, J. Am. Chem. Soc., 2000, 122, 8581.
33. G. Feher and A. F. Kip, Phys. Rev., 1955, 98, 337.
34. S. M. Zhou, M. Yoshizawa, P. Lou, S. Z. Jin and Y. H. Zhang, J. Appl. Phys., 2001, 89, 3377.
35. N. O. Moreno, P. G. Pagliuso, C. Rettori, J. S. Gardner, J. L. Sarrao, J. D. Thompson, D. L. Huber, J. F. Mitchell, J. J. Martinez and S. B. Oseroff, Phys. Rev. B, 2001, 63, 174413.

---