Ccmnunications, Great Britain.
OF HIGH-TEMPERATURE K. M. Cirillo
0038-1098/88 $3.00 + .OO Pergamon Press plc
J. C. Wright of Wisconsin,
and J. Seuntjens, Materials (Received
D. C. Larbalestier
Science Program, Department of Metallurgical Engineering. University of Wisconsin, Madison, WI 53706
17, 1987, in revised
11, 1988 by R. H. Silsbee)
starting materials were treated Both YBa2Cu307.x and EuBa2Cu307.x with fluorine gas at 350°C. YBa2Cu307.xFz exhibits a 2-4 K enhancement of the superconducting critical temperature Tc for z = 0.04-0.2. For 0.2 < z -Z 0.4 the T, enhancement of the yttrium compound is about 2 K, and for z > 0.5 the T, and the fraction of superconducting material decrease. For EuBa2Cu307.,F,, T, decreases by about 2 K after fluorine treatment for z = 0.12-0.54. Powder x-ray diffraction patterns for both compounds revea! no structural difference between the fluorine-treated and untreated samples for F2 treatment yielding less than 0.5 mole F-/mole compound. At greater than 0.5 mole F-/mole ccmpound, additional phases and destruction of the perovskite structure are evident. Heat treatment (300-950°C for l-24 h) of F2treated samples in an air or oxygen environment causes the samples to regain their original, untreated superconducting properties. Fluoride ion analysis of these F2-, then 02- or air-treated samples shows that little or no F- remains in the samples.
Workers from our laboratories have, in a parallel study, reported the cl’fcct of low concentration fluorine gas treatment on the semiconducting La2Cu04_y material.1 Fluorine gas treatment yields La2Cu04_gFz, which has a superconducting onset T, of approximately 35 K. The x-ray powder diffraction pattern for this Fs-treated lanthanum cuprate reveals an orthor-hom6ic symunchanged from the unfluorinated, semimetry, conducting starting material. In this communication wc report exoeriments detect that small in the changes onset superconducting critical temperatures of YBaTand EuBa2Cu307.,Fz h%hCu307-xFz temperature superconductors by oxidative fluorine gas treatment. Inductive magnetization measurements were performed to determine onset T,‘s for the untreated, FT-treated, and Fo-, then airlOp-treated YBa2Cu;O7_, and EuBa2Cu307_, samples. These T,‘s with fluoride . were correlated ion content as determined by fluoride ion selective electrode analysis. Powder x-ray diffraction data collected for 28 = 20-70” provided phase information about the samples.
Fluorine gas, a strong oxidizing agent, reacts with oxide high-temperature superconducting materials. In a redox reaction fluorine gas F2 is reduced to fluoride ion upon entering the lattice, and ions in the perovskitc lattice are oxidized. By this method of fluorine introduction, the [Cu-O]+ content of the high-temperature superconductors will be changed. Since FT is a very strong oxidizing agent, it is capable-of extracting electrons from [Cu-Ol” to form ICu-Ol+ and F- species. Interestingly,-Fis abou; the same size as 02- and may, to some extent, replace 02- in the lattice yielding [Cu-F]+. X-ray fluorescence scanning electron microscopy has indicated that fluoride anions in anions replace some oxygen YBa2Cu307.x and Ls~CUO~_~ lattices after oxidative fluorine gas treatment.1 This oxidative fluorine gas treatment is very different from the direct incorporation of fluoride ion, through the mixing of barium fluoride with the oxide and carbonate powders or through the bombardment of the superconductors with F- ions.2-4 In the direct incorporation fluorine enters the lattice as an anion in the (-I) oxidation state and has no effect on the redox state of the superconducting material. Because of our prior cxpericnce with oxidative fluorine gas processing of solid materials5, WC were intcrestcd in dctcrmining whether F2 oxidation would affect the properties of high-T, materials. Studies of these two types of fluorinated yield high-temperature superconductors may added information on the nature of the superconductivity in these new materials.
Experimental of Starting
Polycrystalline samples of YBa2Cu307_x prepared by grinding together appropriate ounts of high-purity Y2O3 (Aldrich, 99.99%), CO3 (Aldrich, 99.999Yo), and CuO (Spex, Hi The samples were heated in air for 24 ity).(j 870°C, cooled slowly, reground, and heated 1237
were amBapurh at at
FLUORINE OXIDATION OF HIGH-TEMPERATURESUPERCONDUCTORS
1238 950°C for an samples were or air for 12 and annealed Slowly cooling Ba2Cu307_x using Eu2O3. 2.2
additional 24 h. After cooling, the pressed into pellets, sintered in 02 h at 700°C, cooled slowly to 450°C, at that temperature for 4 h before to room tcmpcrature.‘l The Eusamples were prepared similarly
Pellets of the superconducting starting materials were placed on dec tubes (Coors Porcelain Co., 99.8% alumina) and loaded into an alumina tube (Coors Porcelain Co., 99.8%, 32 inch length) that serves as the reaction region of the F2 oxidation line.5 The alumina tube, which is surrounded by a tube furnace (Lindbcrg, Model 54032A), conThe fluorine line afnects to the fluorine line. fords contained gas delivery, reaction at temperatures from ambient to 1300°C, and scrubbing of unreacted fluorine gas. After sample introduction, the line is purged with dry nitrogen for several hours, The tube furnace temperature is set at 350°C and varying amounts of F2 arc introduced into the system from a 9.9% F2/90.1% N2 cylinder (Mnthcson Gas Products, size 3) to create samples of varying Fion content. The fluorine is allowed to react with the oxide superconductors for 0.5-2 h before the furnace temperature is lowcrcd and the flow of pure N2 resumed. 2.3
Powder x-ray diffraction patterns were obtained with a Nicolct 12/V powder diffractometer having a CuKa radiation source. T, measurements were performed inductively on 2 mm x 2 mm x 4 mm bars cut from F2-trcatcd and untreated pellets. The inductive measurement apparatus consists of a pair of balanced pick-up coils, one empty and one surrounding the sample, inside a primary coil that produces a 17 Hz sinusoidal ac magnetic field (typically 58 PT rms). The zero level at this low frcqucncy is flat with changes in temperature and is very reproducible. The differential signal from the two pick-up coils is measured as a function of temperature. The earth’s magnetic field was not shielded, and demagnetization factor and void fraction effects were neglected. The inductive measurements have been described previously.138 Cutting the pellets into bars after treatment should result in inductive T, measurements that are not influenced by surface shielding effects due to depth gradients of fluorine oxidation9 Great cart was taken in the temperature measurcmcnts because the observed changes from F2 treatment are small. Multiple measurements wcrc made for each material and frequent calibrations were performed using a reference pellet of untreated YBa2Cu307_x, The samples were analyzed for F- content using fluoride ion selective elcctrodc analysis (Orion, Model 94-09-00 with KCI reference electrode). The samples were ground and dissolved in small quantities of acid (HCl or HN03). Standards(lO-3, 10m4, toe5 M) were prcparcd by dissolving appropriate amounts of NaF (Aldrich, 99.999%) and untreated starting material in acid. The addition of untreated YBa2Cu307.x or EuBs2Cu307_, to the standards compcnsntcs for any possible cffccts of the superconductor matrix on the fluoride ion de-
Vol. 66, No. 12
dissolution, TISAB IV11 termination.lO Ionic Strength (Orion preparation), a Total Adjusting Buffer with added cation complcxing agent, and dcionizcd-distilled water wcrc added to the standard and sample solutions. The solutions were adjusted to a pH = 5.5 with NaOH powder (CCL, 99.9%). Measurements of the potential difference between the F- electrode and the rcference electrode wcrc made with a Pope pH/Ion Meter (Model 1501, rclativc mV scale). A plot of the differences in potential versus the log of the concentrations of the standards gives a straight line, and the F‘ concentrations of the unknown samples were intcrpolatcd from that plot. 3.
Table 1 lists the cxtrapolatcd superconducting critical onset tcmpcraturcs and fluoride ion content of the YBa2Cu307_xFz and EuBa2Cu307_,Fz materials employed in this study. The uncertainty in the tcmpcraturc mcasurcment is f 0.5 K. The uncertainty in the F- concentration is either 1% for samples annlyzcd by fluoride ion selective electrode analysis or 20% for concentration cstimations based on amuunt of fluorine gas used in sample trcatmcnt. The Y Ba2Cu307_x starting materials exhibited Tc onset’s in the range of 91.592.5 K, while the EuBa2Cu307., starting material (all samples wcrc mndc from one preparation of starting material) had a T, onset of 95.0 K. Thcsc values are in agrccmcnt with rcccnt mcasurcmcnts Oxidativc of T, in the Y and Eu mntcrials.t2 fluorine treatment of Y’B~~CU~O~_~ leads to T, enhancement for low concentrations of fluorine and mild reaction conditions. The YBa2Cu307_x Fz data arc plotted in Figure I. The highest valucs of T, result for concentrations in the region of 0.04-0.2 molt F-/molt compound. Samples containing ca. I molt F-:molc compound exhibit a decrease in T, (Figure I). The fluorine-treated EuBa2Cu307_x samples show a dccrcase in T, for the range of fluoride ion concentrations studied (0.12-0.54 mole F-/ mole compound, see Tnblc 1). YBa2Cu307_x and EuBn2Cu307_, samples containing less than 0.5 molt F-,‘molc compound yield powder x-ray diffraction pnttcrns that arc virtually identical to the patterns of their rcspcctivc untrcatcd starting materials (Figure 2).
1. Plot of onset T, (K) versus in YBa2Cu307_xFz.
Vol. 66, No. 12
FLUORINE OXIDATION OF HIGH-TEMPERATURESUPERCONDUCTORS Table 1. Table of sample designation, sample type (YBa2CU307., Or EuBa2Cu307_& fluoride ion content as mole F-/mole compound, compound and onset Tc. The uuccrtainty in T, is 0.5 K.
Table 1. Z (mole F-/ molt
T, (K) onsets
a 1% uncertainty.
b 20% uncertainty.
Substantial disruption of the orthorhombic structure and multiphase character occur for both yttrium and europium sumplcs containing on the order of 1 molt F-/molt compound (Figure 3). We also anncalcri some of the fluorine-treated samples in either air or 02. The critical temperature T, changes after this air/02 treatment, approaching the critical tcmpcraturc of the starting material treated normally with oxygen. Fluoride ion selective elcctrodc analysis reveals that all, or nearly all, of the fluoride ion is lost from the treated samples during the air/02 anncal process. Figure 4 shows the inductance versus temperature plots for sample 6A, 6B, and GC, which correspond, respectively, to the untreated, F2-trcatcd, and F2- then air-trcatcd rcsponsc of sample 6.
The fluorine-treated sample 6ti has the nominal formula YBa2Cu307_xF0,2 and a T, onset of 96 K. After a l-h air anncal at 500°C, this sample (labeled 6C) contains less than 0.005 mole F-/mole compound, and the T, dccreascs to 93.5 K after the air anncal. Figure 4 also shows the cxtrapolation method for the T, onsets. 4.
Our treatment employs fluorine gas as a potent oxidizing agent, which is itself reduced upon entcring the superconductor material. The reaction conditions for cnhanccmcnt arc quite mild with regard to trcatmcnt tcmperaturc and total amounts of fluoride ion introduced into the Iatticc, but
40 2 6
Figure 2. X-ray
powder diffraction pattern for (a) untrcatcd YBa2Cu307_x and (b) F2-trcatcd YBa2Cu307_x. The formula for the F2-treated sample is approximately YBa2Cu307_xF0.1.
Plot of ac. susceptibility versus tempcraturc for sample 6. Sample corresponds to the untreated YBa2Cu307_x starting material, 68 corresponds to the F2-treated sample, and 6C corresponds to the F2-, then air-trcatcd sample.
maximum Tc rcachcd by the F2 trcntmcnt of YBa2C~307.~ is the normal value of T, for untreated EuBa2Cu307_,. The subtlc but reproducible T,
L I!,.; 30
X-ray powder diffraction pnttcrn an ovcrtrentcd sample (Sample 14, YBa2Cu307_xFl) Showing multiphasc character and destruction of the orthorhombic, Pmmm structure.
they dramatically nffcct rcdox propcrtics in the superconducting matcrinls. WC obscrvc a 2-4 K T’, cnhanccmcnt upon oxidntivc fluot inc trcntmcnt of YBa2Cu307.x. The ac. susceptibility mcasuremcnts arc particularly suited for observing small changes in T,. Although great care was taken to insure accurate absolute mcasurcments of T,, it is the rclativc chnngcs that arc important in thus work. The T, of YBa2Cu307.x incrcascd rcpreducibly with exposure to small amounts of F2 and returned to the initial value nftcr subscqucnt treatment in 02. Similar trcntmcnt of EuBn2Cu307-x caused a dccrcnsc in T,. lntcrcstingly, the
enhancement produced by the oxidative fluorine gas treatment is vcrv similar to the enhancement seen by Reilly el rrl. I3 in their work with H2 gas treatment of YBa2Cu3O7_x. The cffccts of oxidative fluorine gas treatment hnvc also been observed by Tissue (‘1 nl.’ for La2CuO4_y. This scmiconducting matcrinl is convcrtcd to La2Cu04_gFZ which exhibits a superconducting onset T, of approximately 35 K. Air/02 anneals of YBa2Cu307_,Fz rcvcrse the effect of fluorine trcntmcnt in as little as 2 h at 400°C, as shown in Figure 4. The T, returns to its untreated value and F- ion is rcmovcd from the lattice indicating that this fluorine is incorporated into the superconductor in a rcvcrsible manner. Work on EuBa2Cu307_, rcvcals that it too returns to its original onset T, = 95 K form. However, for fluorinated superconductors prepared by using BaF2 as one of the starting mntcrials2.3, we find that heat trcatmcnt rn air or 02 at 900°C for 24 h does not rcducc the F- content. This diffcrcncc in the reversibility of the F- ion addition upon air/02 trcatmcnt clearly indicates the different roles F2 and BaF2 play in superconductor processing. Destruction of the Pmmm, orthorhombic lattice is not seen for fluorine trcatmcnt leading to YBa2Cu307_xFz and EuBa2Cu307_xFz whcrc z is below 0.5. Although x-ray powder diffraction is only scnsitivc to changes in the lattice greater than a few percent. inductive mcasurcmcnts indicate that a large fraction of the material is superconducting at the post F2-trcatmcnt T,. The reason for the cnhnnccmcnt of Tc in fluorine-trcatcd YBn2Cu307.x but the dcprcssion
FLUORINE OXIDATION OF HIGH-TEMPERATURE
in fluorine-trcatcd EuBa2Cu307., is not clear. One should note that EuBa2Cu307_, had a supcrconducting onset critical tcmpcraturc of 95 K starting mntcrials had an while YBa2Cu307_x onset T, of only 92 (i 0.5) K. Perhaps the EuBa2. Cu307.x compound already contained the optimum [Cu-O]+/[CU-O]~ ratio14-19 so that oxidativc fluorine treatments upset the dclicatc balance. The YBa2Cu307_x starting mntcrials, on the other hand, may have had too low a [Cu-O]+/[Cu-O]O ratio. Thus oxidative fluorine trcntmcnts. result-
iug in low concentrations of F‘, rmprovc the conduction propertics of YBa?Cu307_, and lcad to the observed cnhanccmcnt Tn the superconducting ’
Acknowlcdgcmcnt -- This work was supported by the National Scicncc Foundation, Division of Materials Research under grant DMR-8645405 and by the Dcpartmcnt of Energy, Office of Fusion Energy and Division of High Energy Physics.
References 1. B. M. Tissue, K. M. Cirillo, J. C. Wright, M. M. Daeumling, and D. C. Larbalcstier, Solid State Commun. a, in press. 2. S. R. Ovshinsky, R. T. Young, D. D. Allred, G. DeMaggio, and G. A. Van dcr Lecdcn, Phys. Rev. Lett. x, 2579 (1987). 3. R. N. Bhargava, S. P. Herko, and W. N. Osborne, Phys. Rev. Lctt. 59, 1468 (1987). 4. M. Xian-Ren, R. Yan-Ru, L. Ming-Zhu, T. Qing-Yun, L. Zhen-Jin, S. Li-Hun, D. WeiQing, F. Min-Hua, M. Qing-Yun, L. ChangJiang, L. Xiu-Hai, Q. Guan-Liang, C. MouYuan, Solid State Commun. 64, 325 (1987). 5. K. M. Cirillo and J. C. Wright, J. Crystal Growth & 453 (1987). V. Y. Lee, M. 6. R. Beyers, G. Lim, E. M. Englei, L. Ramirez, R. J. Savoy, R. D. Jacowitz, T. M. Shaw, S. La Placa, R. Boehme, C. C. Tsuei, S. I. Park, M. W. Shafer, and W. J. Gallagher, Appl. Phys. Lett. 51, 614 (1987). 7. J. B. Torrance, E. M. Englcr, V. Y. Lee, A. I. Nazzal, Y. Tokura, M. L. Ramirez, J. E. Vazquez, R. D. Jacowitz, and P. M. Grant, in Chemistrv of Hiah-Temoc’rature Suoerconductors, edited by D. I.. Nelson, M. S. Whittingham, and T. F. George, Chapter 9; ACS Symposium Series 351, American Chemical Society, Washington DC (1987). M. Dncumling, X. Cai, J. 8. D. C. Larbalesticr, Seuntjens, J. McKinncll, D. Hampshire, P. Lee, C. Meingast, T. Willis, 1-i. Muller, R. D. Ray, R. G. Dillcnbcrg. E. F. Hcllstrom, and R. Joynt, J. Appl. Phys. 62. 3308 (1987). 9. R. A. Hcin, Phys. Rev. B 33, 7539 (19’86). !!?. See, for cxnmplc. matrix cl’fccls. internal standard, or matrix matching in (a) R. D.
16. 17. 18.
Braun, Introduction to Instrumental Analvsis; hlcGraw-Hill, New York, (1987); (b) R. A. Day, Jr., and A. L. Underwood, Ouantitative Analvsis, 5th Ed.; Prentice-Hall, Englewood Cliffs, NJ (1986); (c) Standard Methods for the Examination of Water and Wastcwater edited by A. E. -I Greenberg, J. J. Connors, D. Jenkins, and M. A. H Franson; R. R. Donnellcy & Sons, Washington, DC (198 I). Fluoride Ion Sclectivc Electrode Manual, Orion Research Instruction Manual for Fluoride Electrodes Models 94-09-00, 96-09-00. J. M. Tarascon, W. R. McKinnon, L. H. Greene, G. W. Hull, and E. M. Vogel, Phys. Rev. B 5, 226 (1987). J. J. Reilly, M. Sucnaga, J. R. Johnson, P. Thompson, and A. R. Moodenbaugh, Phys. Rev. B x, 5694 (1987). M. W. Shafer, T. Penney, and B. L. Olson, Phys. Rev. B x, 4047 (1987). B. G. Banley, L. H. Greene. J.-M. Tarascon. and G. W Hull, Appl. Phys: Lctt. s, 622 (1987). L. F. Matthciss and D. R. Hamann, Solid State Commun. 63, 395 (1987). A. P. Malozemoff and P. M. Grant. Z. Phvs. . B - Condensed Matter 67, 275 (1987). S. Hcrn, J. Cai, S. A. Shahecn, Y. Jcon, M. Croft, C. L. Chang, and M. L. denBocr, Phys. Rev. B 2, 3895 (1987). B. Batlogg, R. J. Cava, A. Jayaraman, R. B. van Dover, G. A. Kourouklis, S. Sunshine, D. W. Murphy, L. W. Rupp, H. S. Chen, A. White, K. T. Short, A. M. Mujsce, and E. A. Rietman, Phys. Rev. Lctt. &?, 2333 (1987).