Study On Structure,Magnetism And Phase Diagram For FeSe-based Superconductors

Since the discovery of superconductivity (SC) with Tc=26K in LaO1-xFeAs system in2008, the superconducting transition temperature was quickly raised to56K in1111system by means of replacing La by other rare earth elements such as Sm, Nd Pr, Ce, Gd and Y, electron doping by F substitution for O, hole doping by Sr substitution for La, systhesis under pressure and applying physical pressure. This great achievement signified the end of the monopoly of cooper-based superconductors. The research on this new superconductors again attracted great attention. After five years’endeavors, researchers discovered other FeAs-based superconductors with the similar crystal structure and the same Fe2+square lattice as LaO1-xFxFeAs:(Ba,K)Fe2As2(122), LiFeAs (111),’32522’,’42622’,’43822’and Ca10(PtnAs8)(Fe2-xPtxAs2)s and two kinds of FeSe-based superconductors:Fe(TeSe) and (Tl,K,Rb)FexSe2. In the last five years, great effort was contributed to the theory and experiment research in the crystal and electronic structures, magnetic and transport properties and SC, salient progress was made.This dissertation focuses on two topics:1. the intrinsic phase diagram and the flux pinning mechanism of the Fe(Te,Se,S) system;2. S substitution effect in Tl0.5Rb0.3Fe1.72Se2and Tl0.8Fe1.6Se2-xSx system.The Fe1+x(Te,Se) system possesses the simplest structure in the iron-based superconductors, comprises of stacked tetragonal Fe(Te,Se) layers. Shortly after the emergence of SC with Tc～8K in β-FeSe, M. H. Fang firstly discovered the SC with Tc～14K in Fe1+x(Te,Se) system[arXiv:0807.4775]. Later, Wei bao et al. proved that the commensurate or incommensurate bicollinear antiferromagnetic (AFM) order is the ground state of the parent compound FeTe[arXiv:0809.2058], which is different from the collinear AFM order discovered in the parent compound of the FeAs-based superconductors. Besides, there is an AFM order transition in Fe1+xTe accompanied with the structural transition from tetragonal to orthorhombic. Although the long-range AFM order is suppressed in the superconducting Fei+x(Te,Se), strong AFM spin fluctuation still exists at low temperature. Because of the excess iron existing in the FeSe-11system which complicates the magnetic structure, superconductivity and the phase diagram, many groups in the world considered the bicollinear AFM order to be antagonistic to the SC. On the basis of these works, we synthesized a series of Fe1+y(Te1-xSex) and Fe1+y(Tei-xSx) single crystals and annealed them in air or vacuum to partially remove the excess irons or to improve the homogeneity of the crystal. We reached several conclusions via the structure, composition, susceptibility, resistivity and Hall coefficient measurement:1. Annealing in air greatly improves the SC in the Fe1+y(Te1-x(Sex) single crystals by decreasing the excess irons. For example, the Fe1+yTe1-xSex (x=0.1) single crystal as-grown do not show bulk SC. While after annealing in air at300℃for3hours, its volume fraction of SC increases to50%. Annealing in air also improves the SC of the sample with higher Se content, e. g., the Tc of the Fe1+yTe1-xSex (x=0.4) single crystal as-grown is12K, volume fraction of SC is5%. After2hours of annealing at300℃in air, its volume fraction of SC increases to74%, Tc increases to14.7K. It indicates that the excess iron atoms greatly suppress the SC of the Fe1+y(Te1-xSex) and Fei+y(Te1-xSx) system, while annealing in air effectively improve the SC.2. The excess irons in the single crystals as-grown with x<0.3is more than that of the crystals with x>0.3. Annealing in vacuum effectively improves the SC for the crystals with x>0.3and change its normal state from a semiconductor to a metal. However, annealing in vacuum do not improve the SC of the sample with lower Se doping. These results indicate that the excess Fe is the main factor that suppresses the SC of the sample with lower Se doping, while for the sample with higher Se doping, excess Fe and the inhomogeneous distribution of Te and Se are both the adverse factors.3. In order to investigate the effect of annealing on the charge carrier concentration, we measured the Hall coefficients RH of the Fe1+δTe0.6Se0.4single crystal before and after annealing. The concentration of the charge carriers nu increases after annealing in vacuum at400℃for7days, consistent with the metallic behavior of the normal state resistivity. However, the nH decreases after annealing in air at200℃for2hours, consistent with the semiconductor behavior of the normal state resistivity.4. Annealing in air (200-300℃) can also partially remove the excess Fe in the Fe1+yTe single crystals as-grown, resulting in a little increase of TN and the transition from a semiconductor to a metal below TN, demonstrating a carrier localization due to the excess Fe. 5. We also found that the solid solution limit of S in FeTe is about0.12, it is hard to further increase the content of S. Partially substitution of Te by S leads to the SC with Tc～10K. Similar to the Fei+y(Te1-xSex) system, there is also excess Fe existing in the Fei+yCTe1-xSx) compound. Its SC can be distinctly improved by annealing in air as well.6. Based on the result of the resistivity and susceptibility measurement, we obtained the phase diagrams of the Fe1+y(Te1-xSex) and Fe1+y(Te1-xSx) system. In contrast to the results of other groups, we found that the bicollinear AFM order coexists with the bulk SC in the low doping region. Our phase diagram is similar to that of the (Ba,K.)Fe2As2and SmFeAsO1-xFx system, indicating that the coexistence of the long-range AFM order and the bulk SC is a commonality of iron-based superconductors.7. We measured the magnetization hysteresis loop of the optimal doped FeTe0.6Se0.4single crystals after annealing. We found that the critical current density at2K under zero field reaches3.8×105A/cm2. The flux pinning mechanism is determined as core normal point-like pinning.At the end of2010, our group firstly discovered the SC with Tc～30K. and the iron vacancy order in (Tl,K)FexSe2compound[arXiv:1012.5236], and consider the SC to be close to the AFM insulator. W. Bao et al. confirmed the existence of the (?)×(?) Fe vacancy order proposed by our group and the block checkerboard AFM order (BCAF-c) in KFe2Se2,(Tl,K)FexSe2and (Tl,Rb)Fe*Se2compounds, which is also proved by the theory calculation result of Cao C. et al. It is puzzling that the A2Fe4Se5(A=Tl,K.,Rb,Cs) phase with the iron vacancy and the BCAF-c order is observed in the insulating as well as the superconducting compounds. Several questions are naturally posed:what is the correlation between the AFM order phase and the superconducting phase? Is the former the parent compound of the later or they are just phase separated? These questions became the hot issue in the field of SC in the last two years. Based on these results, we studied the S doping effect in the superconducting Tl0.5Rb0.3Fe1.72Se2system and the insulating Tl0.8Fe1.6Se2system. Great details of the structure, composition, resistivity and susceptibility data are discussed in Chapter4. Here, we present several conclusions:1. The undoped Tl0.5Rb0.3Fe1.72Se2is a superconductor with Tc～32K. In the Tl0.5Rb0.3Fe1.72Se2-xSx system, the Tc increased to33.1-33.6K with0≤x≤0.4. However, further increase of the S content results in the lowering Tc. The SC is fully suppressed with x>1.0. Phase separation of the insulating (Tl,Rb)2Fe4Ses phase with the BCAF-c order and the superconducting (Tl,Rb)0.53Fe2Se2phase with no magnetic order was determined by neutron diffraction. This phase separation makes it hard to analysis the intrinsic S doping effect in the (Tl,Rb)0.53Fe2Se2system. As a result, we determined to study the S doping effect in the Tl0.8Fe1.6Se2system with the (?)×(?) superlattice and the BCAF-c order.2. All the Tl0.8Fe1.6Se2-xSx (0<x<2.0) samples remain a single phase. We found that the lattice constants and volume decrease with increasing S content, the Se(S)-Fe-Se(S) angle varies systematically with S doping and reached109.47°when x=1.7, which is exactly the Se(S)-Fe-Se(S) angle of the regular Fe(Se/S)4tetrahedron, but the sample did not show any sign of SC.3. We concluded from the resistivity and susceptibility data that when S content increases, the AFM transition temperature TN and the resistivity decrease, the system evolves from a insulator into a metal. When x>].6, the system no longer exhibits BCAF-c order, but enters into a spin glass state with spin lying in the ab plane.4. Based on these results, we firstly obtained the phase diagram of the Tl0.8Fe1.6Se2-xSx system. S substitution for Se suppresses the long-range BCAF-c order, leading to a spin glass state at low temperature rather than a superconducting state. In view of the similarities between our results and that of the (TI,Rb)0.8Fe1.56Se2insulator under pressure, we consider the variation of the physical properties of the Tl0.8Fe1.6Se2-xSx system to be the consequence of chemical pressure introduced by S doping. It is noteworthy that the S substitution for Se may lead to the redistribution of the iron vacancies, but details need to be confirmed by neutron diffraction.The results concluded above clarify some opinions about the phase diagrams of the Fe(Te,Se,S) and (Tl,K,Rb)FexSe2system. It is of great significance and making a better knowledge of the iron-based superconductors.