以氨熱法形成Ba0.5(NH3)Fe2Se2 的超導性質

全文

(1)國立台灣師範大學. 碩士論文. Superconductivity of Ba0.5(NH3)Fe2Se2 by ammonothermal method. Presented by. Shou-Ting Jian. Department of Physics. National Taiwan Normal University. July 2013.

(2)

(3) 摘要從其他類 122 結構的超導體 KxFe2Se2 (Tc ~ 32 K)，我們發現此類超導體與 Ba1-xKxFe2As2 (I4/mmm)系統是類似的，而此類超導體將 FeSe 的超導溫度從 8.5 K 升高到 32 K 的原因是因為將鹼土或是鹼金屬元素嵌入了 FeSe 的層狀結構中。因此我們試著用液態氨能夠溶解鹼土與鹼金屬融的特性(氨熱法)將 Ba 嵌入至中 FeSe 的層狀結構中。而在四角結構 β-Fe1+xSe 中，我們以 x = 0.008 當做氨熱法的基材，因為在所有我們嘗試的比例中(0.005 到 0.02)，不論從 XRD 還是在低溫時超導的相轉變溫度都有最好的表現。而從中子繞射的實驗中得知 Li0.5(NH3)Fe2Se2 結構與 122 結構的超導相似只是原本鹼土族與鹼金屬的位置(2a)在結構中被 NH3 所佔據，另外 Li 也分別坐落在(2b)以及(4c)位置上，而經過反應後的樣本我們以 XRD 的相對強度模擬出類似的 Ba0.5(NH3)Fe2Se2 結構。此 Ba0.5(NH3)Fe2Se2 擁有 39 K 的超導相轉變溫度，以及經由磁性測量結果推導出此樣本的 μ0Hc1 ~ 21.1 G 與 μ0Hc2 ~ 53.82 T。. 1.

(4) Abstract The recent discovered tetragonal superconductor KxFe2Se2 attracts much attention for its high superconducting transition temperature Tc ~ 32 K. The system can be regarded as inserting alkali metal and alkaline earth atoms between the FeSe layers, which is the iso-structure with Ba1-xKxFe2As2(I4/mmm) and the Tc is enhanced greatly from un-intercalated FeSe (Tc = 8.5 K). Hence, we use the ammonothermal method to intercalate β-Fe1+xSe layer by Ba, trying to enhance the superconducting transition temperature. We used the x = 0.008 tetragonal β-Fe1+xSe as the basic material because of its best performance in the superconducting phase transition. with Tc,onset = 13 K and Tc,zero = 9.5 K between x = 0.005 to 0.02 From neutron powder diffraction pattern of Li0.5(NH3)Fe2Se2, the structure is similar to 122 phase but the the site of the alkali metals and alkaline earths are replaced by NH3(2a) and Li located in (2b) and (4c) site respectively. From XRD pattern of Ba intercalated in tetragonal β-FeSe layer by ammonothermal method, we simulated the structure is Ba0.5(NH3)Fe2Se2.which is Li0.5(NH3)Fe2Se2 type structure. The Ba0.5(NH3)Fe2Se2.is an type II superconductor with high Tc = 39 K. we also derived the μ0Hc1(0 K) ~ 21.1 G and μ0Hc2(0 K) ~ 53.82 T.. 2.

(5) Contents Abstract (Chinese). 1. Abstract (English). 2. Contents. 3. Acknowledgments. 4. List of Figures and Tables. 5. Chapter 1 Introduction. 10. 1.1 Prologue. 10. 1.2 Review of iron-based superconductor. 12. 1.3 Motivation. 18. Chapter 2 Experimental Details. 19. 2.1 Sample preparation. 19. 2.2 X-ray diffraction. 24. 2.3 Magnetism measurement. 26. 2.4 Resistance measurement. 28. Chapter 3 Result and Discussion. 34. 3.1 The selection of basic Fe1+xSe material. 34. 3.2 Analysis of Ba intercalated in β–FeSe by ammonothermal method. 39. 3.3 Superconducting properties of Ba0.5(NH3)Fe2Se2. 45. 3.4 Characteristic of Ba0.5(NH3)Fe2Se2. 53. Chapter 4 Conclusion. 55. References. 57. 3.

(6) 致謝此項研究於國立台灣師範大學物理系徐永源教授的指導下完成。首先我非常感激徐永源老師的指導，以引導的方式能讓我自己摸索並且扶正我處理事情的態度，而讓我不論在學科還是事務上的處理成熟許多。實驗室中的陳弘詔學長、楊名正學長，感謝你們研究上的建議與事務上的協助，並且在實驗上陪我度過許多日子。我還要感謝國立清華大學古煥球老師幫助我在物理概念上的增強，鄭武漢先生在磁性測量上大力的支持，以及國立台灣師範大學駱芳鈺老師在實驗上的協助，使我們的樣本能擁有足夠的時間測量與製作，使我的畢業論文數據能夠如期完成。另外感謝王一智在學業上與樣本 X-ray 測量的協助以及黃俊豪在實驗上的幫忙，使我的樣本能在時間內如期測量。還有淡江大學王玉富學長的教導使我對低溫電阻測量與 LabVIEW 能有所熟悉。最後我要感謝陪我度過這兩年的家人和朋友，謝謝你們的支持與包容。. 4.

(7) List of Figures and Tables Fig. 1-1. Crystal structures of some of iron-based superconductors [11]. 11. Fig. 1-2. LaO1-xFxFeAs phase diagram[15]. 13. Fig. 1-3. Crystal structure of LiFeAs (Cu2Sb-type structure, space group P4/nmm. 14. Fig. 1-4. (a) Crystal structure of BaFe2As2 (ThCr2Si2-type structure, space group I4/mmm). (b) Ba1-xKxFe2As2 phase diagram[20]. 15. Fig. 1-5. (a) Crystal structure of FeSe (b) Crystal structure of FeSe- from top view 15. Fig. 1-6. Crystal structure of KFexSe2[26]. Fig. 1-7. (a) Iron-vacancy ordered corresponding to AFe1.5Se2. The blue solid. 17. circles are iron atoms. The green open circles are vacancies. This type of order is called here the 2 × 4 iron vacancy order. (b) The case of AFe1.6Se2 with its. 5 × 5 iron vacancy distribution. All the iron atoms. have three iron neighbors. (c) State with no iron vacancies, corresponding to AFe2Se2[27]. 18. Fig. 2-1. Binary phase diagram of Fe1+xSe[25]. 21. Fig. 2-2. Binary phase diagram of Fe1+xSe under 480 oC[28]. 21. Fig. 2-3. Ammonal Thermal Method Schematic diagram. 23. Fig. 2-4. Block Diagram of Rigaku rotating anode powder x-ray diffractometer. 25. Fig. 2-5. Block diagram of SQUID detector model SPMS for magnetization and Magnetic susceptibility measurement system of SQUID (QUANTUM DESIGN, MPMS/MPMS-7 systems). Fig. 2-6. Fig. 2-7. 27. MPMS Response to Dipole Point Source of SQUID (QUANTUM DESIGN, MPMS/MPMS-7 systems). 27. Schematic circuit of electrical resistance measurement. 28. 5.

(8) Fig. 2-8. Schematic diagram of closed-cycle refrigerator. Fig. 2-9. Schematic diagram of closed-cycle refrigerator with resistence box and sample holder. Fig. 3-1. 29. 32. The powder X-ray of β–FeSe with different Fe context from 1.008 to 1.02. Each of derived lattice constant is a = 3.7757 Å and c = 5.5306 Å for x = 0.008 a = 3.7743 Å and c = 5.5287 Å for x = 0.008, a = 3.7759 Å and c = 5.5297 Å for x = 0.01, a = 3.7749 Å and c = 5.5275 Å for x = 0.015, and a = 3.77467 Å and c = 5.5261 Å for x = 0.02. Small peaks corresponding to hexagonal δ-FeSe were observed at 2θ = 42o in x = 0.005, 0.008 and 0.02 which is marked by asteroids.. 35. Fig. 3-2. The c/a ratio of β–FeSe with different Fe content from 1.005 to 1.02. 36. Fig. 3-3. Low-T relative resistance R (T)/R(30 K) of β–FeSe with various iron contents. The observed onsets of superconducting transitions are Tc,onset = 12 K for x = 0.005, 13 K for x = 0.008, 12 K for x = 0.01, and 10 K for x = 0.015. The zero resistance temperatures are Tc,0 = 9 K for x = 0.005, Tc,0 = 9.5 K for x = 0.01 and 0.015, and Tc,0 =8 K for x = 0.015. Fig. 3-4. 36. Low field magnetic susceptibility of Fe1.008Se as a function of temperature χg(T) under FC and ZFC modes. Superconducting transition onset was observed at Tc = 9.5 K. Inset: The superconductivity transition observed by low temperature resistance shows an onset of Tc,onset = 13 K and zero resistance at Tc,0 = 9.5 K. Fig. 3-5. The powder X-ray diffraction pattern of Fe1.008Se. The diffraction peaks can be well indexed by tetragonal β–FeSe structure. The derived lattice constants are a = 3.7743 Å and c = 5.5287 Å. Small peaks corresponding to hexagonal δ–FeSe were observed at 2θ = 32o and 42o 6. 37.

(9) with almost invisible Fe2O3.peaks marked by asteroids. Inset: A closer look of diffraction peak around 32o which is composed of hexagonal (101) of δ–FeSe and tetragonal (002) of β–FeSe Fig. 3-6. 38. The powder X-ray diffraction pattern of Ba0.5(NH3)Fe2Se2. The diffraction peaks can be indexed by tetragonal Li0.5(NH3)Fe2Se2–type structure(space group: I4/mmm). The derived lattice constants are a = 3.7868 Å and c = 16.8847 Å. Small peaks corresponding to hexagonal β–FeSe were observed at 2θ = 32o, 42o and 51o with unknown peaks marked by asteroids. The remains located at 2θ= 10.83o (0 0 2)’, 28.86o (1 0 3)’and 40.53o (1 1 4)’ which is marked by symbol v. (a = 3.7550 Å and c = 16.3396 Å). 41. Fig. 3-7. Crystal structure of Ba0.5(NH3)Fe2Se2. 42. Fig. 3-8. The simulation of XRD (Cu Kα) for Ba0.5(NH3)Fe2Se2. 42. Table 3-1. Simulation of structure parameter for Ba0.5(NH3)Fe2Se2. 43. Fig. 3-9. The powder X-ray diffraction pattern of hexagonal δ–FeSe (lower pattern) minor peaks for tetragonal β–FeSe phase were marked by asteroids. The XRD for ammonothermal reacted δ–FeSe with Ba (upper pattern) for 2.5 days. The peaks of δ–FeSe remains strong, which indicates hexagonal FeSe does not participate in ammonothermal reaction. However, ammonia corrosion does happen and results in peaks by asteroids. This peaks are also found in XRD of Ba0.5(NH3)Fe2Se2. Fig. 3-10. 44. Bulk superconducting transition of Ba0.5(NH3)Fe2Se2. Magnetic susceptibility measured with sample cooled under zero field (ZFC) and Ba =10 G measurement field (FC), the inset shows the Tc =39 K and Tirr = 38.5 K. 46 7.

(10) Fig. 3-11. Bulk superconducting transition of Ba0.5(NH3)Fe2Se2. Magnetic susceptibility measured with zero field cooled (ZFC) and Ba = 200 G & 400 G measurement field (FC) with both Tc =39 K and Tirr = 37.5 K & 36.5 K. Fig. 3-12. 46. Bulk superconducting transition of Ba0.5(NH3)Fe2Se2. Magnetic susceptibility measured with zero field cooled (ZFC) and Ba = 600 G measurement field (FC) with both Tc =38.5 K and Tirr = 35.5 K. Fig. 3-13. 47. Bulk superconducting transition of Ba0.5(NH3)Fe2Se2. Magnetic susceptibility measured with zero field cooled (ZFC) and Ba = 800 G measurement field (FC) with both Tc =38.5 K and Tirr = 35 K. Fig 3-14. Temperature dependence of irreversibility line H*(T) of Ba0.5(NH3)Fe2Se2 with applied field Ba up to 1 kG. H*(T) α (1-T/Tc)5/3. Fig. 3-15. 47. 48. Bulk superconducting transition of Ba0.5(NH3)Fe2Se2. Magnetic susceptibility measured with sample cooled under zero field (ZFC) and Ba =10 G & 100 G measurement field (FC) with both Tc =39 K. Fig. 3-16. 49. Bulk superconducting transition of Ba0.5(NH3)Fe2Se2 sample. Magnetic susceptibility measured with sample cooled under zero field (ZFC) and Ba = 1 kG & 1 T measurement field (FC) with Tc = 38.5 K & 38 K respectively. Fig 3-17. 49. Temperature dependence of upper critical field Hc2(T) of Ba0.5(NH3)Fe2Se2 with applied field Ba up to 1 T. Fig 3-18. 50. Field dependence of the virgin magnetization curve at 5 K for zero field cooled Ba0.5(NH3)Fe2Se2. Lower critical field Hc1(5 K) = 21 G. and. Fig 3-19. Hc1(30 K) = 9 G. 51. Field dependence of the virgin magnetization curve at 5 K for zero field. 52. 8.

(11) cooled Ba0.5(NH3)Fe2Se2. Lower critical field Hc1(10 K) = 19 G. and Hc1(20 K) = 16 G Fig. 3-20. Temperature dependence of the lower critical field Hc1(T) for Ba0.5(NH3)Fe2Se2. Hc1(0) ～ 21.1 G was obtained from extrapolation. Fig. 3-21. 52. The temperature dependence of gravimetric magnetic susceptibility χg(T) for Ba0.5(NH3)Fe2Se2 with different ammonothermal reaction time. The optimized reaction time is 3.5~ 4 days. The reaction time for 0.5 and 2.5 days can obviously note a signal which comes from the raw material Fe1.008Se at 9 K. The reaction time up to 12 days is marked by star with a strong paramagnetic signal relatively because the NH3 may corrode Ba0.5(NH3)Fe2Se2 which is reacted under stirring strongly. Fig. 3-22. Fig. 4-1. 54. The decay time of magnetic susceptibility of bulk Ba0.5(NH3)Fe2Se2 sample versus temperature. 54. magnetic phase diagram (< 1T) Ba(T) for Ba0.5(NH3)Fe2Se2. 56. 9.

(12) Chapter 1 Introduction 1.1 Prologue Superconductivity has been found for one hundred years. The unique properties are useful in the electric transfer technology because of superconductivity transferred the electron without any energy loss under phase transition. Even thought the character can solve many energy issues, many problems about the materials, the temperature of phase transition and the complicated mechanism in mirco-scale still need to deal with. Most important of these problems are the interaction mechanism in micro-view. If the interplay details can be controlled well, the first two problems can be solved easily. Unfortunately the interaction must be studied by the materials that we already know. Until now, lots of superconducting material has been discovered that helps us to realize what reason causes the superconducting phase occurred. These materials can divide into many families such as elements, cuprate (high Tc superconductor), and iron-based superconductors. The research turns the point to iron-based superconductors after cuprate because its coexistence of superconductivity and magnetism properties are more interested than cuprate and lots of related compounds (1111-LaO1-xFxFeAs[1], 122-Ba1-xKxFe2As2[2] series, 111- LiFeAs[3][4] and NaFeAs, 11-FeSexTe1-x[5] and intercalate alklia metals to 11 group)[7][8][9] being discovered for study recently. Because of they share the common iron-pnictogen or iron-chalcogen layer so called iron-based superconductors. These compounds with similar iron-based plane will accompany the electronic band structure closely which is multi-band structure and magnetic mechanism in this group [10].. 10.

(13) Fig. 1-1 Crystal structures of some of iron-based superconductors [11].. About this two iron-pnictogen and iron-chalcogen plane, the binary FeAs doesn’t have crystallized into FeAs[10] layer but FeSe and FeTe does. The critical temperature of superconductivity in tetragonal FeSe is at 8 K[12] and intercalates by alkali metals between FeSe layer to layer yield the Tc up to 30 K even more by many ways[8][9][26]. These compounds are called x22 families that are made by intercalation method. This issue will discuss more details in the next section.. 11.

(14) 1.2 Review of iron-based superconductor In 2008, the new superconductor LaO1-xFxFeAs with Tc = 26 K[1] was discovered. This compound starts a new age of superconductor after the high-Tc superconductor “cuprate”. Many related superconducting compounds have been found. They all share the similar Fe-based layer so this family is called iron-based superconductor. The good reason caused iron-based superconductor popular are the coexistence of superconductivity and magnetism and lots of compound which can be studied with their multi-band electronic structure. According the different structure of compounds come from the similar Fe-based, divided the compound into 1111(LaO1-xFxFeAs) ), 111(LiFeAs and NaFeAs), 122(BaFe2As2, 11(FeSexTe1-x) and x22(AxFe2Se2). Here we will discuss this compound briefly. Although the 1111 LaO1-xFxFeAs series has the much high Tc such as NdFeAsO1-y with Tc = 54 K, SmFeAsO1-xF with Tc = 55 K and Gd0.8Th0.2FeAsO with Tc = 56.3 K but this family is hard to study because the size of single crystal which can be produced is too small [13], also the electronic structure is different between bulk and single crystal because of the polar surface state in single crystal boundary. The highly sensitive surface state is hard to study by ARPES [14].. 12.

(15) FIG. 1-2 LaO1-xFxFeAs phase diagram [15]. The 111 family such as LiFeAs(Tc = 18 K) and NaFeAs(Tc = 23 K) doesn’t have the polar surface from the well grown single crystal so they can be studied by ARPES[3][4] which can measure the superconducting gap and compares the difference with bulk. The experiment shows the LiFeAs electronic bands are very independent from one to another because of no magnetic phase in LiFeAs compound, that’s why the band structure can be analyzed easily. Another one NaFeAs has the magnetic phase transition at T = 41 K and structural phase phase transition at T = 52 K [16][17]. Also other research shows that replacing Fe by Co or Ni can suppress the magnetism and enhance the superconductivity [19].. 13.

(16) Fig. 1-3 Crystal structure of LiFeAs (Cu2Sb-type structure, space group P4/nmm). Lots of iron-based superconducting compound in the 122 family can be researched by hole doping or electron doping even part of isovalent replaced [20]. The series of superconducting transition can be drew a rich phase diagram. The common feature is its coexistence of superconductivity and magnetism in the three different phase diagram. The hole doped Ba1-xKxFe2As2 are studied widely with Tc = 38 K[21] after optimized the doping concentration and the electron doped Ba(Fe1-xCox)2As2 with the highest Tc = 22 K[22][23], both they all share the same basis BaFe2As2. In other view, this two doping ways are the same method to increase the carrier density with different type of carriers. The magnetic order transition below 140 K occurred in the non-superconducting BaFe2As2 sample without any doped, and the over doped KFe2Se2 is a nonmagnetic material with Tc = 3 K. Another similar compound is isovalent doped BaFe2 (As1-xPx)2 with Tc = 30 K. This 122 family still has many similar compound with plenty elements replaced in the same basis structure brings us to study their physical properties in lots way. 14.

(17) (a). (b). Fig. 1-4 (a) Crystal structure of BaFe2As2 (ThCr2Si2-type structure, space group I4/mmm) (b) Ba1-xKxFe2As2 phase diagram[7]. In iron-based superconductor families, the most structures all share the same basis-FeAs layer but the binary FeAs phase diagram shows no crystalline in the FeAs layer. For this reason, the binary FeSe superconductor with Tc = 8 K[12] which was found is surprising for everyone. After FeSe superconductor had been studied, the Tc of FeSe can be enhanced to 37 K under pressure. And FeSe combined FeTe to the ternary FeSexTe1-x compound with Tc = 14 K for x = 0.5.. (a). (b). Fig. 1-5 (a) Crystal structure of FeSe (b) Crystal structure of FeSe- from top view 15.

(18) Here is a thing about FeSe. All evidence of research about the ratio between Fe and Se shows that excess Fe plays an important role in superconducting phase[12][24][25]. But the ratio of superconducting phase is different from recent literatures about the composition between Fe and Se. The composition depends on the original form of elements (such as powder or shot) and the reacting route which you choose. On the other hand, FeSe superconductor is a sensitive compound about its ratio. The ternary compound FeSexTe1-x also has the problem with the Fe content but lack. What kind of rule about the Fe concentration is still needed to understand but hard due to the little different in Fe’s concentration. The original idea of x22 family comes from the similar FeAs layer in Ba1-xKxFe2As2 [20]. The experiment discovered a new superconducting family AxFe2Se2 by intercalated “A” (K, Rb, Cs, Tl) element to enhance the c-axis between FeSe layer to layer[8]. The result shows Tc up to 30 K and high Neel temperature (> 500 K). The research shows that x22 structure may have three phases in the same condition (the level of Fe-vacancy order/disorder) by STM and other measurement method (AEPES, neutron scattering, and others). For this reason, this chemical formula also is called 245 from the chemical view. The three phases, according their space group, can divide into AFe1.5Se2, AFe1.6Se2, and AxFe2Se2. By the observation of these experiments, the occurrence of superconducting phase is AxFe2Se2. That’s why this compound is called x22 family. Trying to find the new parent compounds, changing the electronic concentration is a good way because this way is the same concept from the 122 family. Unfortunately, replacement methods of other alkali metals and alkaline earths have failed by the reaction of high-temperature route. For example, taking Sr/Ba and FeSe together in the sealed quartz tube before reactants are put in the furnace. The finished samples turn out to become the structure of NaCl like binary (Sr/Ba)Se which isn’t the superconducting tetragonal structure and part of FeSe. 16.

(19) Another problem is that x22 compounds are unstable in the air.. Fig 1-6 Crystal structure of KFexSe2[26]. (a). (b). (c). Fig.1-7 (a) Iron-vacancy ordered corresponding to AFe1.5Se2. The solid circles are iron atoms. The open circles are vacancies. This type of order is called here the. 2 × 4 iron vacancy order. (b) The case of AFe1.6Se2 with its. 5 × 5 iron vacancy distribution. All the iron atoms have three iron neighbors. (c) State with no iron. vacancies, corresponding to AFe2Se2 [29]. After the x22 compounds, researchers try to intercalate alkali metals and alkaline earths in FeSe layer by other routes which avoid the high-temperature method. Things always can find ways out. The intercalated way via liquid ammonia that can dissolve alkali metals and alkaline earths construct another x22 with NH3 successfully. The c-axis of this structure with extremely long c-axis than the normal x22 family is a feature. Recently, researches show the LA method also took NH3 [8][9][27] in the 17.

(20) structure by neutron scattering diffraction. This family is still a totally new compound and lots of properties need to be studied.. 1.3 Motivation The KFe2Se2 system has been studies widely by sintering method after AFe2As2 system, but the research discovered the system is not pure about 122-phase by structural analysis. This system is called 245-family with Fe-vacancy order in structure. In order to study more detail about the new family, the basic idea is trying to find the same structural compounds and figure out the electric and magnetic mechanism. But the similar compounds which are changed the site of potassium to other alkali metals are produced hardly because the 122 phase of structure will turn to another structure after the raw material which combines two binary compounds such as BaSe, SrSe and other FeSe phase after high-temperature reaction. That’s why the route of sintering method needs to be change by others. Liquid ammonia can dissolve the alkali metals and alkaline earths that is already known and FeSe is bound by Van der Waal force between layer by layer, so that x22 structure may intercalate by liquid ammonia dissolved the alkali metals and alkaline earths in auto-cleaver. Other researches show that the NH3 also intercalates in the structure with alkali metals and alkaline earths[8][9][27] actually, but more details about structure analysis hasn’t discover very well by experiment, we trying to produce the sample by intercalated Strontium and Barium in FeSe layer and studies these sample by XRD, magnetic measure, resistance.. 18.

(21) Chapter 2Experimental Details 2.1 Sample preparation Ba0.5(NH3)xFe2Se2 samples are produced by two main procedures: The first step is that well quality of Fe1.0XSe (the superconducting volume is higher than 20%) needs to be produced and the second is that Ba and Fe1.0XSe react in the autoclave with proper ratio by ammonal thermal method. Here will describe the detail as this below:. Step1: production process of β-Fe1.0XSe samples Iron granules (99.98%, Alfa Aesar) and Selenium shots (99.999%, Alfa Aesar) with a proper ratio were loaded in the cylindrical crucible (9.9 mm × 11.9 mm × 16 mm) and sealed in the quartz tube. The sample was placed in the furnace. The temperature was gradually increased to 750 OC (180 OC/hour) and kept 750 OC for 1-3 days (This step evaporated the selenium were evaporated totally and then reacted with iron at 750 OC.) then heated to 1075 OC (100 OC/hour). This temperature was held for 1-3 days in order to melt and mix the FeSe compound well then fast decreased to 420 O. C (80 OC/hour). Due to the environment in 420 OC is a condition in β-Fe1+xSe phase. diagram, the temperature were held 420 OC for 1-4 days then quenched in liquid nitrogen (-77 OC) .. 19.

(22) Prepare the iron granules and Selenium shots with proper ratio.. ↓ sealed these in quartz tube with crucible Heated to 750 OC. ↓ held for 1-3 days Heated to 1075 OC. ↓ held for 1-3 days Fast decreased to 420 OC. ↓ held for 1-4 days Quenched in liquid nitrogen. 20.

(23) Fig. 2-1 Binary phase diagram of Fe1+xSe[25]. Fig. 2-2 Binary phase diagram of Fe1+xSe under 480 OC[28]. 21.

(24) Step2: production process of Ba0.5(NH3)Fe2Se2 by ammonal thermal method The alkaline earths (Barium and strontium) and as-prepared Fe1.0XSe (powder like) were loaded in autoclave before the autoclave is evacuated air out and checked the level of vacuum by gauge. The autoclave was placed in liquid nitrogen until it had been cooled down to the condensation point of NH3, then flow in the gas form of NH3 (99.9%). The flow was detected by flow meter.. Load the sample of Fe1+xSe powder and alkaline earths (Br/Sr) with proper ratio and the magnetic stirrer in the autoclave. ↓ Evacuate air of the autoclave by switching the 3-way valve to the vacuum pump side. ↓ Switch the 3-way vale to ammonia side, open the needle valve and pre-load 1-atm-ammonia into the autoclave then close the needle valve again.. ↓ Place the autoclave in liquid nitrogen and wait till the pre-loaded ammonia condenses, which can be observed by the decreasing pressure.. ↓ When the pressure remains stable, flow the ammonia into the autoclave and control the flux speed by the needle valve till the purpose volume is loaded.. ↓ Sealed the autoclave, and stir the sample during the reaction.. 22.

(25) Manometer Needle Valve 2. 3-Way Valve Autoclave. Flow meter. Needle Valve 1. Liquid Nitrogen. Ammonia Tank. Vacuum Pump. Fig. 2-3 Ammonothermal Method Schematic diagram. 23.

(26) 2.2 X-ray diffraction Parts short length electromagnetic wave is called X-ray which is produced by a high energy electron after collision with the Cu target. The electron was free after it interacted with the high energy one in K layer (1s). The hole-like place was filled by the other electron in L layer (2p) after the free electron was hit, then the two type of X-ray which were Kα1 (λ = 1.5418 Å) and Kα2 (λ = 1.5448 Å) were injected. X-ray will diffraction after it interacted with long range ordering materials which has specific plane (hkl) and the type of constructive and destructive diffraction detected by detector. The specific structure has one unique pattern which could kwon by the intensity and diffraction angle (2θ). When atoms or molecular accumulate in space periodically, which are called lattice. The lattice also regarded as layer by layer. The distance between layers is called d, according to Bragg’s law 2dsinθ = nλ, the structure which is researched can be realized. The powder X-ray diffractometer is made by Rigaku in Japan with Cu Kα (λ = 1.54187 Å) anode. The scanning step is 0.02o per step ( 15 seconds cunting time per step in 2θ range of 5o to 55o. the instruments are followed as schematic diagram in this below. 24.

(27) Fig. 2-4. Block Diagram of Rigaku rotating anode powder x-ray diffractometer. 25.

(28) 2.3 Magnetism measurement The magnetization and magnetic susceptibility measurements were carried out by a QUANTUM DESIGN MPMS SQUID magnetometer (SQUID) with a temperature control module that provides an active-regulated, precision thermal environment over range from 2 to 400 K. The superconducting magnet system provides a reversible field operation ±7 Tesla using a non-overshoot technique for irreversibility of magnetic susceptibility measurement. The SQUID detector system that includes model 2000 SQUID amplified control electronics, sensing pick-up and special filter with full computer control via the HP-150 interfaced computer. The block diagram of SQUID detector system is shown in Fig. 2.3 and Fig. 2.4. The sample was attached on a sample holder and put into the sample chamber then calibrated sample position to make it site at the center of SQUID detecting coil array. The data were measured and stored automatically by MPMSR2 software. For zero-field cooled measurements in the MPMS SQUID, the “Magnetic reset” option was used to quench the superconducting magnet and reduce the residual or remnant field to less than 2 G. When the temperature was raised from lower than 4.4 K to higher than 4.6 K the sample was heated to a temperature higher than its Tc and field was set to zero before cooling down to the measurement temperature. These additional operations were added under the consideration of the limitation of the SQUID temperature control processes.. 26.

(29) Fig. 2-5 Block diagram of SQUID detector model SPMS for magnetization and Magnetic susceptibility measurement system of SQUID (QUANTUM DESIGN, MPMS/MPMS-7 systems). MPMS response to Dipole Point Source. Fig. 2-6 MPMS Response to Dipole Point Source of SQUID (QUANTUM DESIGN, MPMS/MPMS-7 systems). 27.

(30) 2.4 Resistance measurement The system will separate three parts to description in this section: a. Measurement system Resistance of sample is measured by four-probe technique. The schematic diagram show as this below: The power of circuit comes from lock-in amplifier then connects to the resistance switching box. On the other hand, the Rreg not only can prevent the total current overload but control the total current during measure. The different temperature will caused the total resistance changed dramatically only if this Rreg is 10 to 100 times than sample which is measured. The work of Rref helped us to calculate the total current in this circuit. Two volts of Rref and Rsample are measured by two lock-in amplifier which are connected to computer and controlled by LabVIEW.. Rreg. Rreg: prevent the total current overload Rref: convert this volt for current Resistance switching box. Lock-in generator Rref Measured by Lock-in amplifyer Sample. Sample space I+ V+. V- I-. Rreg: total six resistance can be switched (1, 10, 100, 1K, 10K, 100K Ω) Rref: total six resistance can be switched (0.1, 1, 10, 100, 1K, 10K Ω). Fig. 2-7 Schematic circuit of electrical resistance measurement. 28.

(31) b. Closed-cycle refrigerator system. Basic principle: The principle of closed-cycle refrigerator is called Gifford-McMahon (GM) refrigerator. Here have 4 steps to cool the metal which is chose. PH. TL. TH. Switching valve. Displacer. PL. QH. Heat exchanger. Cold finger. Shield QL. Fig. 2-8 Schematic diagram of closed-cycle refrigerator. Step1 (a) to (b): while high pressure enters the space, the displacer is moved to the left. High pressure gas is set at the right. Step2 (b) to (c): the valve is switched from high to low pressure. The high pressure gas expends then absorbs heat from TL state. Step3 (c) to (d): the displacer is reset to the right and let the displacer pass through low temperature gas. Step3 (d) to (a): through the high pressure in the left space and repeat Step1 again.. 29.

(32) PH. TH. TL. PH. Displacer. TH. TL. Displacer. PL. PL. (a) PH. TH. (b). TL. PH. TH. TL. Displacer. Displacer. PL. PL. (c). (d). The 4 Kelvin Closed Cycle Refrigerator Systems from Janis and SHI Sumitomo (SHI) 4 K Refrigerator - RDK-408D2 have two GM refrigerator stages and the working gas is helium. The system can cool down the cold finger (sample holder) to 4.2 Kelvin at second stage. The connection between lock-in, resistance switching box, heater and closed cycle system will show in next paragraph.. 30.

(33) Instrument setup: The R-T measurement system is connected by total 30 (with back up) enameled insulated wires from sample holder space to thermometer and resistance switching box. From the fig, the sample space is desired to place 3 samples which are measured in the same while (12 enameled wire need to be used) for convenience and advanced hardware in the future. The two sensors are placed in different position. The sensor I is contacted well to the cold finger in order to detect the temperature right at the place and the sensor II is placed to the sample holder because the distance between sensor I and sample holder which is removable provides the temperature which have a gradient, also, the sensor II can avoid the different temperature from sensor II to sample when they are closed (8 enameled wire need to be used). Last one thing is heater which is placed at near the cold finger and is used to control temperature at the second between the measurements from this moment to next one. This heating controller usually needs couple, one of them offers a main power to resist the power of cold head and another one maintain the temperature which is set. But now this system has one heater to control for desire. Because the well two layer shielding can resist the radiation and offer a vacuum space to defend the thermal convection from gas, the system can be consider a thermal equilibrium state during measurement from 4 K to 280 K (4 enameled wire need to be used).. 31.

(34) Lock-in amplifier Thermometer. Helium Refrigerator. Lock-in amplifier. Resistance Cold head. switching box. Connect to sample and circuit. I. Thermometer sensor I. II. III. Thermometer sensor II Sample space. sensor II. Sample holder Fig. 2-9 Schematic diagram of closed-cycle refrigerator with resistence box and sample holder. 32.

(35) c. Connection by LabVIEW This program is desired to calculate the resistance of sample when the specific temperature is stood still then plot and save data. One of two functions is data selection, another one is desired as a controller for heater in this program.. Data selection: The basic idea is that we want to clean up the data which isn’t need, so the data is selected in a specific range. For temperature, the condition of the data which is decided to save is how different from the last point that have been saved to newest one. Constrain of decision is not the same for every range (for instance, form 273 K to 200K, per 1 K saved one point). In order to save the phase transition in R-T diagram precisely, for the resistance part, the condition is that the newest one value of resistance is different from the last one about 5 %. If the condition is agreed, the data will be saved and become the last one data.. Heat Controller: This heat controller relies on the temperature which is detected by Sensor I to optimize the value of PID and power limit. The first point about temperature controller is that PID of thermometer is needed to be found for each range of specific temperature and every PID accompany different power limit must to fix one. Due to the data will out of control during the sensitive of PID and power limit switching, the program has to by pass the unreal data about combination of data selection and heat controller by case selection.. 33.

(36) Chapter 3 Result and Discussion 3.1 The selection of basic Fe1+xSe material The most important thing about Ba intercalated into β-FeSe layer with NH3 is needed to realize how to create the basic material β-FeSe purely and what kinds of impurity may compete with Ba0.5(NH3)Fe2Se2 during formation reaction. First step is trying to find the condition of β-FeSe about its sintering temperature and Fe concentration. The sintering condition quote Cave’s paper which refers to the phase diagram [28] and the Fe concentration that we need to find because the condition varies from the raw material which is powder or shot and even uses the same type of material from other researches [12] (the tetragonal Fe1+xSe range of x from -0.12 to 0.3). The Fig. 3-1 of x-ray reveals that Fe content is changed from x = 0.005, 0.008, 0.01, 0.015, and 0.02 respectively. The difference can be observed that hexagonal δ–FeSe phase which is marked by asteroids occurs at 2θ = 42o for x = 0.005, 0.008, 1.02 very clearly but δ–FeSe is invisible for x = 1.01 & 1.015 patterns. The range x of perfect tetragonal β-Fe1+xSe is around 0.01. From the ratio of c/a (Fig. 3-2), the value will locate in the range of 1.4645 to 1.4650, when the guide of well superconducting transition in R-T graph. Also the fitting curve of c/a ratio versus Fe concentration x is still different with other report [25] which is shifted from 0.01 to 0.02. But the best superconducting performance is x = 0.008 in the Fig. 3-3 Low-T relative resistance that is noted the Tc,onset = 13 K and Tc, zero = 9.5 K and from low field gravimetric magnetic susceptibility of Fe1.008Se as a function of temperature χg(T) under FC and ZFC modes (Fig. 3-4), we can show the well quality of superconducting volume under superconducting phase is 68%. This is reason why we want to choose x = 0.008. The basic material of ammonothermal method is had to analyze what kinds of background in β-FeSe before reaction. This step is helpful to find the impurities phase 34.

(37) after ammonothermal method. All peaks can well define by tetragonal β-FeSe except two small peaks for hexagonal δ–FeSe and almost invisible Fe2O3 phase which is. a = 3.7757 Å c = 5.5306 Å *. a = 3.7743 Å c = 5.5287 Å. Fe1.008Se. Intensity (arb. unit). (1 1 2) (2 0 0) (0 0 3). (1 1 1). Fe1.005Se. (0 0 2). (1 0 1). (0 0 1). marked by asteroids by the powder X-ray diffraction pattern of Fe1.008Se (Fig. 3-5).. *. Fe1.01Se. a = 3.7759 Å c = 5.5297 Å. Fe1.015Se. a = 3.7749 Å c = 5.5275 Å. Fe1.02Se. a = 3.7767 Å c = 5.5261 Å *. 10. 15. 20. 25. 30. 2θ degree. 35. 40. 45. 50. Fig. 3-1: The powder X-ray of β–FeSe with different Fe context from 1.008 to 1.02. Each of derived lattice constant is a = 3.7757 Å and c = 5.5306 Å for x = 0.008 a = 3.7743 Å and c = 5.5287 Å for x = 0.008, a = 3.7759 Å and c = 5.5297 Å for x = 0.01, a = 3.7749 Å and c = 5.5275 Å for x = 0.015, and a = 3.77467 Å and c = 5.5261 Å for x = 0.02. Small peaks corresponding to hexagonal δ–FeSe were observed at 2θ = 42o in x = 0.005, 0.008 and 0.02 which is marked by asteroids.. 35.

(38) 1.4650. Fe1+xSe. c/a ratio. 1.4645. 1.4640. 1.4635. Fitting curve. 1.4630 0.000. 0.005. 0.010. 0.015. 0.020. Fe content. 0.025. Fig. 3-2: The c/a ratio of β–FeSe with different Fe content from 1.005 to 1.02.. 1.00. Fe1+xSe. onset. Tc. = 12 K. 0.75. R (Ω)/R (30 K). onset. Tc. = 10 K. 0.50 onset. Tc. 0.25. x = 1.005 x = 1.008 x = 1.01 x = 1.015. = 13 K. Tc,0 = 8 K 0.00. 0. 5. 10. 15. 20. Temperature (K). 25. 30. Fig. 3-3 Low-T relative resistance R (T)/R(30 K) of β–FeSe with various iron contents. The observed onsets of superconducting transitions are Tc,onset = 12 K for x = 0.005, 13 K for x = 0.008, 12 K for x = 0.01, and 10 K for x = 0.015. The zero resistance temperatures are Tc,0 = 9 K for x = 0.005, Tc,0 = 9.5 K for x = 0.01 and 0.015, and Tc,0 =8 K for x = 0.015. 36.

(39) 2. Fe1.008Se. 0. FC 1.0. -4 0.8. -6 0.6. R (mΩ). -3. χg (10 emu/G g). -2. Tc = 9.5 K. -8 -10. 0.0. Ba = 10 G 4. 5. 6. 7. 8. = 13 K. 0.4. 0.2. ZFC. -12 -14. onset. Tc. Tc,0 = 9.5 K 0. 5. 9. 10. 10. 15. 20. T (K). 11. 12. 25. 13. 30. 14. 15. Temperature Fig. 3-4: Low field magnetic susceptibility of Fe1.008Se as a function of temperature χg(T) under FC and ZFC modes. Superconducting transition onset was observed at Tc = 9.5 K. Inset: The superconductivity transition observed by low temperature resistance shows an onset of Tc,onset = 13 K and zero resistance at Tc,0 = 9.5 K.. 37.

(40) β - FeSe. a = 3.7743 Å c = 5.5287 Å. 1000 800 31.5. *. 15. 20. 25. 33.0. H(1 0 2). - FeSe. v. 0 10. 32.5. (1 1 1). (0 0 2). Fe2O3. v δ. 32.0. 2θ degree. 4000 2000. v. 1200. 6000. *. H(1 0 1). 1400. (1 1 2) (2 0 0) (0 0 3). 8000. 1600. (0 0 2). 1800. (1 0 1). Intensity (arb. units). 10000. Fe1.008Se. Intensity (arb. units). (0 0 1). 12000. 30. 2θ degree. *. 35. 40. v. 45. 50. Fig. 3-5: The powder X-ray diffraction pattern of Fe1.008Se. The diffraction peaks can be well indexed by tetragonal β–FeSe structure. The derived lattice constants are a = 3.7743 Å and c = 5.5287 Å. Small peaks corresponding to hexagonal δ–FeSe were observed at 2θ = 32o and 42o with almost invisible Fe2O3.peaks marked by asteroids. Inset: A closer look of diffraction peak around 32o which is composed of hexagonal (101) of δ–FeSe and tetragonal (002) of β–FeSe.. 38.

(41) 3.2 Analysis of Ba intercalated in β–FeSe by ammonothermal method The powder X-ray diffraction pattern of Ba0.5(NH3)Fe2Se2 Fig. 3-6 shows that diffraction peaks can be indexed by tetragonal Li0.5(NH3)Fe2Se2–type structure(space group: I4/mmm). The derived lattice constants are a = 3.7868 Å and c = 16.8847 Å. The a-axis is similer to β–FeSe (a ≒ 3.775 Å) but c-axis is larger than KxFe2Se2 (c ≒ 14.91 Å). The reason which caused the high c-axis in iron-based superconducting family is the NH3 with Ba embedded in FeSe layer (Li0.5(NH3)Fe2Se2 (c ≒ 16.5266 Å)) and the site of Sr in SrFe2As2 are replaced by N with H, on the other hand, the NH3 occupies the 8-coordinate site in SrFe2As2 structure type, also from NPD of Li0.5(NH3)Fe2Se2[9] shows the Li or H (because of its atomic size) still have contribution after they predicted NFe2Se2. This report refined the site occupation factor 26% for Li1, 19% for Li2 (same position with Ba1 and Ba2), 23% for D1 and 25% for D2(replaced the H by D for NPD measurements). The simulation (Fig. 3-7) by diamond 3 shows the relative intensity is match with the XRD of Fig. 3-6 by adjustment for the site occupation factor (table. 3-1) which are 10% for Ba1 (2b-site), 7.5% for Ba2 (4c-site), as well as 25% for H1 and H2, then the formulation of Ba intercalated in β–FeSe by ammonothermal method should be rewrote as Ba0.25(NH3)Fe2Se2. But we still need more evidence to confirm this factor. Adjustment of Ba1 and Ba2 site occupation factor is based on the ratio of intensity for (0 0 2) and (1 0 3) which is 0.5 approximately in Ba0.5(NH3)Fe2Se2 XRD pattern. The structure of Ba0.5(NH3)Fe2Se2 reveals in Fig. 3-8. X-ray powder diffraction shows no evidence in Fig. 3-6 for the starting material but still no match phase can index on the peak which is marked by symbol v, however other report [8] shows the ammonothermal method yield two tetragonal cell on K, Sr, Eu and Yb but Ba, contradiction with Ba0.5(NH3)Fe2Se2 in our experiment. But another report says Kx(NH3)yFe2Se2 structure separated to two tetragonal cell during 39.

(42) the concentration of K is right on x = 0.4 and 0.5[27]. The contribution of Ba0.5(NH3)Fe2Se2 XRD at 10.83o is reasonable to attribute to the second cell because the excess of Ba (mole ratio = 0.8)make sure the enough Ba concentration during reaction. In order to realize the unknown peaks further in the Ba0.5(NH3)Fe2Se2 XRD, the intercalated basis of β–FeSe replaced by δ–FeSe with a small count of β–FeSe (Fig. 3-9 lower pattern) and repeated the ammonothermal method and the unknown peaks at 2θ = 24.12o 34.38o, 44.81o, and 46.99o shows up (Fig. 3-9 upper pattern) after ammonothermal reacted δ–FeSe with Ba for 2.5 days. The peaks of δ–FeSe remains strong, which indicates hexagonal FeSe does not participate in ammonothermal reaction. These unknown peaks may be also found in XRD of Ba0.5(NH3)Fe2Se2 during the reaction with impurities β–FeSe. Beside, the small count of β–FeSe had disappeared in the XRD pattern that tells ammonia corrosion may cause new impurities which are the source of catalysts [27].. 40.

(43) a = 3.7868 Å c = 16.8847 Å. 1400. Ba0.5(NH3)Fe2Se2. 400. (2 1 3). (2 1 1). H(1 1 3). (1 0 7) (1 1 6) (2 0 0) (2 0 2). (1 1 4). H(1 0 2). v. (1 1 0). 600. Fe O 2 3. 800. H(1 0 1). v. (1 0 1). 1000. (1 0 3). 1200. (0 0 4). Intensity (arb. unit). 1600. (0 0 2). 1800. v. 200 10. 15. 20. 25. 30. 35. 40. Temperature (K). 45. 50. 55. Fig. 3-6: The powder X-ray diffraction pattern of Ba0.5(NH3)Fe2Se2. The diffraction peaks can be indexed by tetragonal Li0.5(NH3)Fe2Se2–type structure(space group: I4/mmm). The derived lattice constants are a = 3.7868 Å and c = 16.8847 Å. Small peaks corresponding to hexagonal δ–FeSe were observed at 2θ = 32o, 42o and 51o with unknown peaks marked by asteroids. The remains located at 2θ = 10.83o (0 0 2)’, 28.86o (1 0 3)’and 40.53o (1 1 4)’ which is marked by symbol v. (a = 3.7550 Å and c = 16.3396 Å). 41. 60.

(44) Fig. 3-7: Crystal structure of Ba0.5(NH3)Fe2Se2. Fig. 3-8: The simulation of XRD (Cu Kα) for Ba0.5(NH3)Fe2Se2. 42.

(45) Table 3-1: Simulation of structure parameter for Ba0.5(NH3)Fe2Se2. 43.

(46) Hexagonal *. *. *. (1 0 1). Intensity (arb. unit). *. Fe2O3. δ-FeSe and Ba after reaction with NH3. (1 0 2). FeSe. 10. 15. 20. 25. 30. 2θ degree. Fe2O3. T. T. (0 0 2). Hexagonal T - Tetragonal. 35. T. T. 40. 45. T. 50. Fig. 3-9: The powder X-ray diffraction pattern of hexagonal δ–FeSe (lower pattern) minor peaks for tetragonal β–FeSe phase were marked by asteroids. The XRD for ammonothermal reacted δ–FeSe with Ba (upper pattern) for 2.5 days. The peaks of δ–FeSe remains strong, which indicates hexagonal FeSe does not participate in ammonothermal reaction. However, ammonia corrosion does happen and results in peaks by asteroids. This peaks may be found in XRD of Ba0.5(NH3)Fe2Se2.. 44.

(47) 3.3. Superconducting properties of Ba0.5(NH3)Fe2Se2 Low field (10 G & 100 G) magnetization measurements for both. zero-field-cooled (field exclusion) and field-cooled (field expulsion) data are shown in Fig. 3-10. The phase transition of Tc = 39 K was observed in both applied field. The irreversibility temperature which is shown in inset Tirr = 38.5 K is slightly small than critical temperature. From the low field (< 1000 G) magnetization data (Fig. 3-11 to Fig. 3-13), the irreversible line H*(T) which separating the vortex lattice state from the vortex glass state was found to vary as (1-T/Tc)5/3 (Fig. 3-14). Based on the data in Fig. 3-15 and Fig. 3-16, the Tc is slightly down to 38.5 K at Ba = 1 kG and 37 K at Ba = 1 T. An initial slope dHc2/dT|Tc ≒ -2 T/K is obtained. By using the limit Werthamer-Helfand-Hohenberg formula, Hc2(0) = -0.69Tc × dHc2/dT [] (shows in Fig. 3-17), which givens an μ0Hc2(0) ≒ 53.82 T. Field dependence of the virgin magnetization curve at 5 K, 10 K, 20 K as well as 30 K for zero field cooled Ba0.5(NH3)Fe2Se2 are shows in Fig. 3-18 and Fig. 3-19. Lower critical field μ0Hc1(5 K) = 21 G, μ0Hc1(10 K) = 19, μ0Hc1(20 K) = 16 G G and μ0Hc1(30 K) = 9 G can be marked by the two linear cross in one point. Lower critical field at 0 K μ0Hc1(0) ≒ 21.1 G (shows in Fig. 3-20) can be predicted by parabola exploration. With small μ0Hc1(0) and large μ0Hc2(0), the Ba0.5(NH3)Fe2Se2 is a typical type II superconductors and accompanies the long penetration depth λ and short coherence length ξ. Ginzburg-Landau parameter κ and thermodynamic critical field μ0Hc(0) are estimated to be λ(0) ≒ 5581 Å, ξ(0) ≒ 34.97 Å, κ ≒ 159 and μ0Hc(0) ≒ 3.37 kG. Because the data all come from polycrystalline sample as well as powder form of sample which is random orientation, the properties of anisotropic effect are averaged. 45.

(48) 0.5 0.0 Tc = 39 K. Ba0.5(NH3)Fe2Se2. -1.0. Ba = 10 G. -1.5 FC. -2.0. 0.40. χg (10 emu/gG). -2.5. -3. -3. χg (10 emu/gG). -0.5. -3.0 ZFC. -3.5 -4.0. 0.35. Tc = 39 K. 0.30 0.25 0.20 0.15. Tirr = 38.5 K. 0.10 0.05 0.00 37. 5. 10. 15. 20. 25. 38. 30. 39. 40. Temperature (K). 35. 40. Temperature (K). 41. 42. 45. 50. Fig. 3-10 Bulk superconducting transition of Ba0.5(NH3)Fe2Se2. Magnetic susceptibility measured with sample cooled under zero field (ZFC) and Ba =10 G measurement field (FC), the inset shows the Tc =39 K and Tirr = 38.5 K. 0.24. Tirr = 37.5 K. Tc = 39 K. 0.20. Tc = 39 K Tirr = 36.5 K. 0.18 200 G, FC. -3. χg (10 emu/gG). 0.22. 200 G, ZFC. 0.16. 400 G, FC. Ba0.5(NH3)Fe2Se2. 0.14 400 G, ZFC 34. 35. 36. 37. 38. 39. 40. 41. 42. Temperature (K) Fig. 3-11 Bulk superconducting transition of Ba0.5(NH3)Fe2Se2. Magnetic susceptibility measured with zero field cooled (ZFC) and Ba = 200 G & 400 G measurement field (FC) with both Tc =39 K and Tirr = 37.5 K & 36.5 K. 46.

(49) -3. χg (10 emu/gG). 0.16. Tc = 38.5 K. 0.15. Ba0.5(NH3)Fe2Se2. Tirr = 35.5 K 0.14. Ba = 600 G 0.13 34. 35. 36. 37. 38. 39. Temperature (K). 40. 41. 42. Fig. 3-12 Bulk superconducting transition of Ba0.5(NH3)Fe2Se2. Magnetic susceptibility measured with zero field cooled (ZFC) and Ba = 600 G measurement field (FC) with both Tc =38.5 K and Tirr = 35.5 K. 0.14. Tirr = 35 K. Tc = 38.5 K. Ba0.5(NH3)Fe2Se2. -3. χg (10 emu/gG). 0.15. 0.13 Ba = 800 G. 0.12 33. 34. 35. 36. 37. 38. 39. Temperature (K). 40. 41. 42. Fig. 3-13 Bulk superconducting transition of Ba0.5(NH3)Fe2Se2. Magnetic susceptibility measured with zero field cooled (ZFC) and Ba = 800 G measurement field (FC) with both Tc =38.5 K and Tirr = 35 K. 47.

(50) 1000. 800. *. μ0H (T) (G). *. 5/3. H α (1-T/Tc) 600. 400. Ba0.5(NH3)Fe2Se2 200. 0 30. 31. 32. 33. 34. 35. 36. Temperature T (K). 37. 38. 39. 40. Fig 3-14 Temperature dependence of irreversibility line μ0H*(T) of Ba0.5(NH3)Fe2Se2 with applied field Ba up to 1 kG., H*(T) α (1-T/Tc)5/3.. 48.

(51) 1 Tc = 39 K. Ba0.5(NH3)Fe2Se2 0. -3. χg (10 emu/gG). 100 G, FC. -1 10 G, FC. -2. 100 G, ZFC. -3. Ba = 10 G ZFC. -4. 0. 5. 10. 15. 20. 25. 30. Temperature (K). 35. 40. 45. Fig. 3-15 Bulk superconducting transition of Ba0.5(NH3)Fe2Se2. Magnetic susceptibility measured with sample cooled under zero field (ZFC) and Ba =10 G & 100 G measurement field (FC) with both Tc =39 K. 0.4 Tc (1 kG) = 38.5 K 0.2 1 kG, FC 0.0. Tc (1 T) = 37 K. 1 T, ZFC. -0.2. -3. χg (10 emu/gG). 1 T, FC. -0.4. Ba0.5(NH3)Fe2Se2. -0.6 -0.8. ZFC, 1 kG. 0. 5. 10. 15. 20. 25. 30. Temperature (K). 35. 40. 45. Fig. 3-16 Bulk superconducting transition of Ba0.5(NH3)Fe2Se2 sample. Magnetic susceptibility measured with sample cooled under zero field (ZFC) and Ba = 1 kG & 1 T measurement field (FC) with Tc = 38.5 K & 38 K respectively. 49.

(52) 10. Ba0.5(NH3)Fe2Se2. μ0Hc2 (kG). 8. 6 dHc2/dT ≅ -2 T/K. 4. 2. 0. 0. 5. 10. 15. 20. 25. 30. 35. Temperature T (K). 40. 45. 50. Fig 3-17 Temperature dependence of upper critical field μ0Hc2(T) of Ba0.5(NH3)Fe2Se2 with applied field Ba up to 1 T.. 50.

(53) 0.1. Moment (emu/g). Ba0.5(NH3)Fe2Se2 μ0Hc1 (30 K) = 9 G. 0.0. -0.1. μ0Hc1 (5 K) = 21 G. -0.2. 0. 10. 20. 30. 40. 50. 60. 70. Applied field Ba (G). 80. 90. 100. Fig 3-18 Field dependence of the virgin magnetization curve at 5 K for zero field cooled Ba0.5(NH3)Fe2Se2. Lower critical field μ0Hc1(5 K) = 21 G. and μ0Hc1(30 K) = 9 G.. Moment (emu/g). 0.00. μ0Hc1(20 K) = 16 G. -0.01 μ0Hc1(10 K) = 19 G. -0.02. -0.03. -0.04. 0. 10. 20. 30. 40. 50. 60. 70. 80. 90. 100. Applied field Ba (G) Fig 3-19 Field dependence of the virgin magnetization curve at 5 K for zero field cooled Ba0.5(NH3)Fe2Se2. Lower critical field μ0Hc1(10 K) = 19 G. and μ0Hc1(20 K) = 51.

(54) 16 G. 25 μ0Hc1(0 K) ~ 21.1 G. μ0Hc1(T) (G). 20. Ba0.5(NH3)Fe2Se2. 15. 10. 5. 0. 0. 5. 10. 15. 20. 25. 30. 35. 40. Temperature T (K) Fig. 3-20 Temperature dependence of the lower critical field Hc1(T) for Ba0.5(NH3)Fe2Se2. μ0Hc1(0) ～ 21.1 G was obtained from parabola extrapolation.. 52. 45.

(55) 3.4 Characteristic of Ba0.5(NH3)Fe2Se2 The temperature dependence of gravimetric magnetic susceptibility χg(T) for Ba0.5(NH3)Fe2Se2 with different ammonothermal reaction time shows (Fig. 3-21) that data accompany a random paramagnetic background in each time after the sample reacted. This background signals origin the impurities after ammonothermal method erode part of Ba0.5(NH3)Fe2Se2 which is proved by reaction time up to 12 days with a strong paramagnetic signal relatively and the phase transition occurred gradually because the NH3 may corrode Ba0.5(NH3)Fe2Se2[27]under stirring strongly after Ba0.5(NH3)Fe2Se2 is formed and other small amount of impurities from the material (Fig. 3-9). Hence, from the Fig. 3-21, the optimized reaction time is 3.5~ 4 days. The reaction time for 0.5 and 2.5 days can obviously note a signal which comes from the raw material Fe1.008Se at 9 K. One last thing is that we discussed about the Ba0.5(NH3)Fe2Se2 which is extremely air-sensitive. In usual case, because the grain size is very small after the ammonothermal method, the period of superconducting life time is 1 to 2 days approximately. But the way that we keep by pressed in to bulk with 250 to 350 kg-cm-2, the period can extent to one week or more as shown in Fig. 3-22. 53.

(56) 1.5 1.0. 12 days. Ba0.5(NH3)Fe2Se2. 0.5. -3. χg (10 emu/gG). 0.0 Fe1.008Se. -0.5. 0.5 days. -1.0 -1.5. 2.5 days. -2.0 -2.5 Reaction time: 4 days. -3.0. ZFC. -3.5. Ba = 10 G. -4.0 0. 5. 10. 15. 20. 25. 30. 35. Temperature (K). 40. 45. Fig. 3-21 The temperature dependence of gravimetric magnetic susceptibility χg(T) for Ba0.5(NH3)Fe2Se2 with different ammonothermal reaction time. The optimized reaction time is 3.5~ 4 days. The reaction time for 0.5 and 2.5 days can obviously note a signal which comes from the raw material Fe1.008Se at 9 K. The reaction time up to 12 days is marked by star with a strong paramagnetic signal relatively because the NH3 may corrode Ba0.5(NH3)Fe2Se2 which is reacted under stirring strongly. Tc = 38 K. Ba0.5(NH3)Fe2Se2. Tc = 39 K. Ba = 10 G. -0.5 After 9 days. ZFC. -3. χg (10 emu/g G). 0.0. -1.0. Sealed in Ar After 2 days. -1.5 10. 15. 20. 25. 30. Temperature (K). 35. 40. 45. Fig. 3-22 The decay time of magnetic susceptibility of bulk Ba0.5(NH3)Fe2Se2 sample versus temperature. 54.

(57) Chapter 4 conclusion From section 3.1, tetragonal β–FeSe can be well produced by the sintering solution which is mixed the Fe and Se well by melting in 1075 oC for 1.5 days before the gas-solid reaction with Fe and Se at 750 oC and keep the sample in 420 oC for 1 to 2 days and than quenched in liquid nitrogen. The result can be proved by XRD and R-T graph. About the Fe mole concentration in β–Fe1+xSe, the best performance of superconductivity in R-T is x = 0.008 with a small amount hexagonal phase which shows in XRD pattern. The basic material of Ba0.5(NH3)Fe2Se2 is also based on the best superconducting quality of tetragonal β–Fe1.008Se. According to the ratio of intensities between (0 0 2) and (1 0 3), the Ba0.5(NH3)Fe2Se2 sample with a long c-axis can be derived by Li0.5(NH3)Fe2Se2-type structure(I4/mmm) which is a similar structure of 122 family but the site of alkali metals and alkaline earths is replaced by NH3 with the site occupation factor in H1(8i) & H2(16m)～ 0.25. as well as the Ba is located at Ba1(2b) (0 0 1/2) ～ 0.1 & Ba2(4c) (0 1/2 0) ～ 0.075. However, the intensities between (1 1 6) and (2 0 0) still conflict with XRD pattern after derived, also the phase separation occurred in (0 0 2), (1 0 3) and (1 1 4) with a = 3.7550 Å and c = 16.3396 Å which are derived by the same space group (I4/mmm). The the superconducting phase transition of Ba0.5(NH3)Fe2Se2 can be observed at Tc = 39 K in applied field Ba = 10 G and Tirr = 38.5 K. Tc dropped slightly from 39 to 37 K in M-T graph with changing the applied field from 10 G to 1 T. Beside, the changed H* is proportional to (1-T/Tc)5/3 under 1 kG but the Tirr still doesn’t change at Ba = 1T which is the same with Ba = 1 kG. From the M-H graph, the Lower critical field μ0Hc1(5 K) = 21 G, μ0Hc1(10 K) = 19, μ0Hc1(20 K) = 16 G G and μ0Hc1(30 K) = 9 55.

(58) G can be observed. The long penetration depth λ and short coherence length ξ. Ginzburg-Landau parameter κ and thermodynamic critical field μ0Hc(0) are estimated to be λ(0) ≒ 5581 Å, ξ(0) ≒ 34.97 Å, κ ≒ 159 and μ0Hc(0) ≒ 3.37 kG. 10. Vortex glass. Hc2. Vortex lattice. 8. *. Ba (kG). H. Normal. 6. Ba0.5(NH3)Fe2Se2 4. 2 Meissner Hc1 0. 5. 10. 15. 20. 25. 30. Temperature (K). 35. 40. 45. Fig. 4-1 magnetic phase diagram (< 1T) Ba(T) for Ba0.5(NH3)Fe2Se2. About the life time of Ba0.5(NH3)Fe2Se2 sample, it can be kept for one week long in the Argon after bulk. The most of impurities are observed non structure in XRD pattern and small amount is contributed by hexagonal δ–FeSe and Fe2O3.. 56.

(59) Reference [1] Y. Kamihara, T. Watanabe, M. Hirano, and H. Hosono, J. Am. Chem. Soc. 130, 3296 (2008). [2] H.-H. Wen and S. Li, Ann. Rev. Condens. Matter Phys. 2, 121 (2011) [3] X. Wang, Q. Q. Liu, Y. X. Lv, W. B. Gao, L. X. Yang, R. C. Yu, F. Y. Li, and C. Q. Jin, Solid State Commun. 148, 538 (2008). [4]J. H. Tapp, Zh. Tang, B. Lv, K. Sasmal, B. Lorenz, P. C. W. Chu, and A. M. Guloy, Phys. Rev. B 78, 060505 (2008). [5] B. C. Sales, A. S. Sefat, M. A. McGuire, R. Y. Jin, D. Mandrus and Y. Mozharivskyj, Phys. Rev. B 79, 094521 (2009). [6] N. Katayama, N. Katayama, S. Ji, D. Louca, S. Lee, M. Fujita, T. J. Sato, J. Wen, Zh. Xu, G. Gu, G. Xu, Z. Lin, M. Enoki, S. Chang, K. Yamada, and J. M. Tranquada, J. Phys. Soc. Jpn. 79, 113702 (2010). [7] Marianne Rotter, Michael Pangerl,Marcus Tegel and DirkJohrendt, Angew. Chem., Int Ed 47,7949 (2008) [8] T. P. Ying, X. L. Chen, G. Wang, S. F. Jin, T. T. Zhou, X. F. Lai, H. Zhang & W. Y. Wang Sci. Rep.2,426 (2012) [9] Matthew Burrard-Lucas1, David G. Free1, Stefan J. Sedlmaier1, Jack D.Wright2, Simon J. Cassidy1, Yoshiaki Hara1,3, Alex J. Corkett1, Tom Lancaster4, Peter J. Baker5, Stephen J. Blundell2 and Simon J. Clarke1. Nat.Mater. 2013, 12, 15 [10] A. A. Kordyuk. Low Temp. Phys. 38, 888 (2012) [11] P. J. Hirschfeld, M. M. Korshunov, and I. I. Mazin, Rep. Prog. Phys. 74, 124508 (2011). [12] F. C. Hsu, J. Y. Luo, K. W. Yeh, T. K. Chen, T. W. Huang, P. M. Wu, Y. C. Lee, Y. L. Huang, Y. Y. Chu, D. C. Yan, and M. K. Wu, Proc. Natl. Acad. Sci. U.S.A. 105, 14262 (2008). [13] P. J. W. Moll, R. Puzniak, F. Balakirev, K. Rogacki, J. Karpinski, N. D. Zhigadlo, and B. Batlogg, Nat. Mater. 9, 628 (2010). [14] C. Liu, Y. Lee, A. D. Palczewski, J.-Q. Yan, T. Kondo, B. N. Harmon, R. W. McCallum, T. A. Lograsso, and A. Kaminski, Phys. Rev. B 82, 075135 (2010). [15] H. Luetkens, H.-H. Klauss, M. Kraken, F. J. Litterst, T. Dellmann, R. Klingeler, C. Hess, R. Khasanov, A. Amato, C. Baines, M. Kosmala, O. J. Schumann, M. Braden, J. Hamann-Borrero, N. Leps, A. Kondrat, G. Behr, J. Werner, and B. B€uchner, Nature Mater. 8, 305 (2009). [16] G. F. Chen, W. Z. Hu, J. L. Luo, and N. L. Wang, Phys. Rev. Lett. 102, 227004 (2009). [17] S. Li, C. de la Cruz, Q. Huang, G. F. Chen, T.-L. Xia, J. L. Luo, N. L. Wang, and 57.

(60) P. Dai, Phys. Rev. B 80, 020504 (2009). [18] X. Wang, Q. Q. Liu, Y. X. Lv, W. B. Gao, L. X. Yang, R. C. Yu, F. Y. Li, and C. Q. Jin, Solid State Commun. 148, 538 (2008). [19] D. R. Parker, M. J. P. Smith, T. Lancaster, A. J. Steele, I. Franke, P. J. Baker, F. L. Pratt, Michael J. Pitcher, S. J. Blundell, and Simon J. Clarke, Phys. Rev. Lett. 104, 057007 (2010). [20] H.-H. Wen and S. Li, Ann. Rev. Condens. Matter Phys. 2, 121 (2011). [21] M. Rotter, M. Tegel, and D. Johrendt, Phys. Rev. Lett. 101, 107006 (2008). [22] A. S. Sefat, R. Jin, M. A. McGuire, B. C. Sales, D. J. Singh, and D. Mandrus, Phys. Rev. Lett. 101, 117004 (2008). [23] N. Ni, M. E. Tillman, J.-Q. Yan, A. Kracher, S. T. Hannahs, S. L. Bud’ko, and P. C. Canfield, Phys. Rev. B 78, 214515 (2008). [24] Junyi Ge, Shixun Cao, Shujuan Yuan, Baojuan Kang, and Jincang Zhang, J. Appl. Phys. 108, 053903 (2010); [25] T. M. McQueen, Q. Huang, V. Ksenofontov, C. Felser, Q. Xu, H. Zandbergen, Y. S. Hor, J. Allred, A. J. Williams, D. Qu, J. Checkelsky, N. P. Ong, and R. J. Cava [26] Wei Bao, G. N. Li, Q. Huang,2 G. F. Chen, J. B. He, M.A. Green, Y. Qiu, D. M. Wang, J. L. Luo, and M. M. Wu arXiv:1102.3674v1 [27] Tianping Ying , Xiaolong Chen, Gang Wang , Shifeng Jin , Xiaofang Lai , Tingting Zhou , Han Zhang , Shijie Shen , and Wanyan Wang J. Am. Chem. Soc, 2013, 135 [28] W. Schuster, H. Mikler, and K. L. Komarek, Monatsch. Chem. 110, 1153 (1979). [29] Elbio Dagotto arXiv:1210.6501v1. 58.

(61)