Elucidation of the mechanism in fluorine-free prepared YBa 2Cu3O7-δ coatings

Article abstract of DOI:10.1021/ic9021799

In this work, the reaction mechanism used in the preparation of fluorine-free superconducting YBa₂Cu₃O_7-δ (YBCO) was investigated. To determine which precursor interactions are dominant, a comprehensive thermal analysis (thermogravimetric analysis-differential thermal analysis) study was performed. The results suggest that a three step reaction mechanism, with a predominant role for BaCO ₃, is responsible for the conversion of the initial state to the superconducting phase. In the presence of CuO, the decarboxylation of BaCO ₃ is kinetically favored with the formation of BaCuO₂ as a result. BaCuO₂ reacts with the remaining CuO to form a liquid which ultimately reacts with Y₂O₃ in a last step to form YBCO. High temperature X-ray diffraction experiments confirm that these results are applicable for thin film synthesis prepared from an aqueous fluorine-free sol-gel precursor.

Full text of DOI:10.1021/ic9021799

Inorg. Chem. 2010, 49, 4471–4477 4471
DOI: 10.1021/ic9021799
Elucidation of the Mechanism in Fluorine-Free Prepared YBa2Cu3O7-δ Coatings
§
†
§
‡
Pieter Vermeir,*,†,‡ Iwein Cardinael, Joseph Schaubroeck, Kim Verbeken, Michael Backer, Petra Lommens,
€
Werner Knaepen,^ Jan D’haen,X Klaartje De Buysser,‡ and Isabel Van Driessche‡
†Faculty of Applied Engineering Sciences, University College Ghent, Schoonmeersstraat 52 (C), 9000 Ghent,
Belgium, ‡Department of Inorganic and Physical Chemistry, Ghent University, Krijgslaan 281 (S3), 9000 Ghent,
Belgium, §Department of Materials Science and Engineering, Ghent University, Technologiepark 903,
9052 Zwijnaarde, Belgium, Zenergy Power GmbH, Heisenbergstrasse 16, 53359 Rheinbach, Germany,
^Department of Solid State Sciences, Ghent University, Krijgslaan 281 (S1), 9000 Ghent, Belgium, and
XImec, Division Imomec, Wetenschapspark 1, 3590 Diepenbeek, Belgium
Received November 4, 2009
In this work, the reaction mechanism used in the preparation of fluorine-free superconducting YBa2Cu3O7-δ (YBCO)
was investigated. To determine which precursor interactions are dominant, a comprehensive thermal analysis
(thermogravimetric analysis-differential thermal analysis) study was performed. The results suggest that a three step
reaction mechanism, with a predominant role for BaCO3, is responsible for the conversion of the initial state to the
superconducting phase. In the presence of CuO, the decarboxylation of BaCO3 is kinetically favored with the formation
of BaCuO2 as a result. BaCuO2 reacts with the remaining CuO to form a liquid which ultimately reacts with Y2O3 in a
last step to form YBCO. High temperature X-ray diffraction experiments confirm that these results are applicable for
thin film synthesis prepared from an aqueous fluorine-free sol-gel precursor.
Introduction
deposition rates with higher yields, and most importantly
(iii) processing under ambient pressure enabling a continuous
process.7
The development of low-cost deposition techniques for
high performance YBa2Cu3O7-δ (YBCO) coated conductors
is one of the major objectives in obtaining a widespread use of
superconductivity in power applications.1-3
Among the different precursor designs described for the
synthesis of YBCO coated conductors using CSD,8-16
During the past decade, various processing techniques
were successfully applied to prepare high critical current
YBCO thin films.4-6 Non-vacuum techniques, overall clas-
sified as Chemical Solution Deposition (CSD) routes, are
preferred over vacuum techniques because they show favor-
able features including (i) lower investments, (ii) faster
(7) Foltyn, S. R.; Civale, L.; Macmanus-Driscoll, J. L.; Jia, Q. X.;
Maiorov, B.; Wang, H.; Maley, M. Nat. Mater. 2007, 6, 631.
(8) Risse, G.; Schlobach, B.; Hassler, W.; Stephan, D.; Fahr, T.; Fischer,
K. J. Eur. Ceram. Soc. 1999, 19, 125.
(9) Xu, Y.; Goyal, A.; Rutter, N. A.; Shi, D.; Paranthaman, M.;
Sathyamurthy, S.; Martin, P. M.; Kroeger, D. M. J. Mater. Res. 2003, 18,
677.
(10) Rice, C. E.; Vandover, R. B.; Fisanick, G. J. Appl. Phys. Lett. 1987,
51, 1842.
(11) Hamdi, A. H.; Mantese, J. V.; Micheli, A. L.; Laugal, R. C. O.;
Dungan, D. F.; Zhang, Z. H.; Padmanabhan, K. R. Appl. Phys. Lett. 1987,
51, 2152.
(12) Gross, M. E.; Hong, M.; Liou, S. H.; Gallagher, P. K.; Kwo, J. Appl.
Phys. Lett. 1988, 52, 160.
(13) Falter, M.; Hassler, W.; Schlobach, B.; Holzapfel, B. Phys. C
(Amsterdam, Neth.) 2002, 372, 46.
(14) Manabe, T.; Sohma, M.; Yamaguchi, I.; Tsukada, K.; Kondo, W.;
Kamiya, K.; Tsuchiya, T.; Mizuta, S.; Kumagai, T. Phys. C (Amsterdam,
Neth.) 2006, 445, 823.
(15) Obradors, X.; Puig, T.; Pomar, A.; Sandiumenge, F.; Mestres, N.;
Coll, M.; Cavallaro, A.; Roma, N.; Gazquez, J.; Gonzalez, J. C.; Castano,
O.; Gutierrez, J.; Palau, A.; Zalamova, K.; Morlens, S.; Hassini, A.; Gibert,
M.; Ricart, S.; Moreto, J. M.; Pinol, S.; Isfort, D.; Bock, J. Supercond. Sci.
Technol. 2006, 19, S13.
(16) Li, X.; Rupich, M. W.; Kodenkandath, T.; Huang, Y.; Zhang, W.;
Siegal, E.; Verebelyi, D. T.; Schoop, U.; Nguyen, N.; Thieme, C.; Chen, Z.;
Feldman, D. M.; Larbalestier, D. C.; Holesinger, T. G.; Civale, L.; Jia, Q. X.;
Maroni, V.; Rane, M. V. IEEE Trans. Appl. Supercond. 2007, 17, 3553.
*To whom correspondence should be addressed. E-mail: pieter.vermeir@
ugent.be. Fax: þ3292644983. Phone: þ3292644550.
(1) Holesinger, T. G.; Civale, L.; Maiorov, B.; Feldmann, D. M.; Coulter,
J. Y.; Miller, J.; Maroni, V. A.; Chen, Z. J.; Larbalestier, D. C.; Feenstra, R.;
Li, X. P.; Huang, M. B.; Kodenkandath, T.; Zhang, W.; Rupich, M. W.;
Malozemoff, A. P. Adv. Mater. 2008, 20, 391.
(2) Schoofs, B.; Cloet, V.; Vermeir, P.; Schaubroeck, J.; Hoste, S.;
Van Driessche, I. Supercond. Sci. Technol. 2006, 19, 1178.
(3) Zhang, W.; Rupich, M. W.; Schoop, U.; Verebelyi, D. T.; Thieme,
C. L. H.; Li, X.; Kodenkandath, T.; Huang, Y.; Siegal, E.; Buczek, D.;
Carter, W.; Nguyen, N.; Schreiber, J.; Prasova, M.; Lynch, J.; Tucker, D.;
Fleshler, S. Phys. C (Amsterdam, Neth.) 2007, 463, 505.
(4) Bhuiyan, M. S.; Paranthaman, M.; Salama, K. Supercond. Sci.
Technol. 2006, 19, R1.
(5) Kashima, N.; Niwa, T.; Mori, M.; Nagaya, S.; Muroga, T.; Miyata, S.;
Watanabe, T.; Yamada, Y.; Izumi, T.; Shiohara, Y. Appl. Supercond. Conf.
2004, 2763.
(6) Ibi, A.; Iwai, H.; Takahashi, K.; Muroga, T.; Miyata, S.; Watanabe,
T.; Yamada, Y.; Shiohara, Y. Phys. C 2005, 426, 910.
r
2010 American Chemical Society
Published on Web 04/21/2010
pubs.acs.org/IC

4472 Inorganic Chemistry, Vol. 49, No. 10, 2010
Vermeir et al.
metalorganic deposition using trifluoroacatetes (TFA) is the
most widespread. This originates from the fact that TFA
solutions decompose to carbonate-free precursor films while
in fluorine-free CSD processes, the organic precursors lead to
the formation of BaCO3. Until recently, it was generally
accepted that the presence of this BaCO3 kinetically hinders
the formation of the YBCO phase, leading to superconduct-
ing films of poor quality.17,18
Nevertheless, several papers report on the successful pre-
paration of high critical current superconducting films from
fluorine-free CSD precursors.19-23 While most of these
methods use organic solvents, we reported in a previous
paper the synthesis of YBCO coatings starting from an
aqueous solution.24 In this work, we present a study of
(i) the reaction mechanism responsible for the formation of
YBCO from fluorine-free CSD precursors and (ii) the role of
BaCO3 in this.
Figure 1. TGA-DTA spectrum in air for a mixture containingY(OAc)3,
.
Ba(OAc)2, and Cu(OAc)2 (Y/Ba/Cu = 1:2:3) heated at 10 ꢀC min-1
Experimental Section
Precursor Preparation. Precursor solutions for thin film
deposition were fabricated from readily available Y-, Ba-, and
Cu-acetates (Sigma-Aldrich). These powders are dissolved in an
acetic acid/water mixture (ratio 1:4) to a total metal-ion con-
centration of 0.6 M. After refluxing at 90 ꢀC for 1 h, triethanol-
amine (>99%, Sigma-Aldrich) is added under continuous stir-
ring until a Cu to ligand concentration of 1:5 is obtained.
Finally, ammonia (25 wt %, Sigma-Aldrich) is added to adjust
the pH between 5 and 7. The complexity to stabilize metal-ions
in a water-based environment is described in a previous paper.25
For TGA-DTA (Thermogravimetric Analysis-Differential
Thermal Analysis) powders of Y2O3, BaCO3, and CuO (Sigma
Aldrich) were used. The YBCO sample, supplied by Zenergy
Power GmbH, was produced using the melt-grown technique.
Small impurities of Y2BaCuO5 and CuO were detected using
XRD (X-ray Diffraction).
Figure 2. XRD spectrum of a mixture containing Y(OAc)3, Ba(OAc)2,
and Cu(OAc)2 combusted at 450 ꢀC in air for 24 h (9) Y2O3 (*) BaCO3
and (o) CuO.
Results and Discussion
Part A: TGA-DTA Study. Fluorine-free CSD precursor
solutions all consist of metal-salts that are stabilized in water
and/or other solvents.19-24 Upon transformation of the
precursor state to the actual superconducting phase, the
respective metal-salts are combusted to Y2O3, BaCO3, and
CuO. Depending on the type of precursor and the atmo-
sphere, combustion normally takes place below 500 ꢀC.
As our precursor solutions are prepared from metal-
acetates, a TGA-DTA spectrum (Figure 1) was recorded
Characterization Techniques. The thermal decomposition
behavior of the various powders was investigated using a TGA-
DTA setup from Stanton-Redcroft STA 1500. All powders
were ball-milled for 12 h to ensure a homogeneous mixture
of small grains. Powders of different metal combinations
were mixed in a metal-ratio referring to the stoichiometry in
YBa2Cu3O7-δ
Micrograph images were taken with a normal light micro-
scope.
.
HT-XRD (High Temperature X-ray Diffraction) measure-
ments of the BaCO3-CuO and Y2O3-BaCO3-CuO films were
performed using a Bruker D8 Discover. Using a parallel beam
geometry in a θ-2θ position, counts were registered with (i) a
linear Vantec detector for BaCO3-CuO films and (ii) a position
sensitive Si(Li) detector for Y2O3-BaCO3-CuO films.
for a powder mixture of Y(OAc)3 2.3H2O, Ba(OAc)2,
and Cu(OAc)2 in a Y/Ba/Cu ratio of 1:2:3.
3
After heating this metal-acetate mixture to 440 ꢀC, an
experimental mass loss of 45% of the total weight is
found. This corresponds well to the theoretical value of
45.2% that can be calculated for the combustion reaction
that transforms the respective metal-salts into Y2O3,
BaCO3, and CuO, as given below.
(17) Parmigiani, F.; Chiarello, G.; Ripamonti, N.; Goretzki, H.; Roll, U.
Phys. Rev. B 1987, 36, 7148.
(18) Hirano, S.; Hayashi, T.; Miura, M. J. Am. Ceram. Soc. 1990, 73, 885.
(19) Kim, B. J.; Yi, K. Y.; Kim, H. J.; Ahn, J. H.; Kim, J. G.; Hong, S. K.;
Hong, G. W.; Lee, H. G. Supercond. Sci. Technol. 2007, 20, 428.
(20) Long, N.; Campbell, L.; Kemmitt, T.; Kennedy, V. J.; Markwitz, A.;
Bubendorfer, A. J. Electroceram. 2004, 13, 361.
2YðOAcÞ þ 12O2ðgÞ f Y2O3ðsÞ þ 12CO2ðgÞ þ 9H2O
ðgÞ
3
(21) Patta, Y. R.; Wesolowski, D. E.; Cima, M. J. Phys. C (Amsterdam,
Neth.) 2009, 469, 129.
(22) Tsukada, K.; Yamaguchi, I.; Sohma, M.; Kondo, W.; Kamiya, K.;
Kumagai, T.; Manabe, T. Phys. C (Amsterdam, Neth.) 2007, 458, 29.
(23) Xu, Y.; Goyal, A.; Leonard, K.; Martin, P. Phys. C (Amsterdam,
Neth.) 2005, 421, 67.
BaðOAcÞ þ 4O2ðgÞ f BaCO3ðsÞ þ 3CO2ðgÞ þ 3H2O
ðgÞ
2
CuðOAcÞ þ 4O2ðgÞ f CuOðsÞ þ 4CO2ðgÞ þ 3H2O
ðgÞ
2
(24) Vermeir, P.; Cardinael, I.; Backer, M.; Schaubroeck, J.; Schacht, E.;
Hoste, S.; Van Driessche, I. Supercond. Sci. Technol. 2009, 22, 075009.
(25) Schoofs, B.; Van de Vyver, D.; Vermeir, P.; Schaubroeck, J.; Hoste,
S.; Herman, G.; Van Driessche, I. J. Mater. Chem. 2007, 17, 1714.
XRD analysis of the powder mixture combusted at
450 ꢀC in air for 24 h, as shown in Figure 2, confirms the
presence of Y2O3, BaCO3, and CuO.

Article
Inorganic Chemistry, Vol. 49, No. 10, 2010 4473
released during reduction, this reaction is atmosphere
dependent.
1
2CuO a Cu2O þ O2
ðAÞ
2
The maximum of the DTA-peak shifts from 1060 ꢀC in
air to 940 ꢀC in Ar.
Figure 3b represents data from thermal analysis for
BaCO3 powder in air and Ar respectively. Both TGA
spectra show that BaCO3 is stable until 900 ꢀC, and
decarboxylation to BaO takes place over a broad temp-
erature range. Even at 1100 ꢀC, no complete combustion
of the precursor is observed (mass loss <22.3%). During
combustion, no substantial DTA-deviation is observed.
DTA-analysis shows two small endothermic peaks at
821 ꢀC (B) and 989 ꢀC (C), corresponding to crystallographic
phase transitions of BaCO328 from γ-orthorhombic (natural
witherite) to R-cubic over β-hexagonal.
BaCO3ðγÞ f BaCO3ðβÞ
BaCO3ðβÞ f BaCO3ðRÞ
BaCO3 f BaO þ CO2
ðBÞ
ðCÞ
ðDÞ
Figure 3c represents the TGA-DTA spectra of Y2O3 in
air and Ar. Both TGA and DTA analysis show that Y2O3
is stable until 1100 ꢀC.
Figures 4a and 4b show the TGA-DTA spectra of YBCO
during heating and cooling, respectively. The DTA-heating
curves show a first exothermic reaction E at 960 ꢀC in air
and at 880 ꢀC in Ar, which can be assigned according to the
literature to the following peritectic transformation:29-31
YBa2Cu3O7 - δ þ CuO f Y2BaCuO5 þ L
ðEÞ
The detection of a small amount of CuO in the melt-
grown YBCO sample using XRD endorses this assumption.
Further heating gives rise to two reversible reactions F and
G respectively at 1040 and 1065 ꢀC in air, 970 and 990 ꢀC
in Ar. Both reactions are accompanied by a small weight
loss. The presence of exothermic peaks during cooling both
in air and Ar suggests that these reactions are solidification
reactions. Furthermore, only the sample heated in air shows
a weight increase during cooling. This suggests that these
reactions involve a change in oxygen stoichiometry.
Binary Precursor Interactions. Figure 5a represents the
TGA-DTA spectra of a binary powder mixture of Y2O3 and
BaCO3. The two compounds were mixed in a Y/Ba-ratio of
1:2. No evidence of any chemical reaction between Y2O3
and BaCO3 is found.32 Only the DTA peak (820 ꢀC) related
to the BaCO3 crystal transition (B) is present. Because of the
limited resolution of the experiment, the βfR transition (C)
is not visible here.
Figure 3. TGA-DTA spectra heated at 10 ꢀC min-1 in air and Ar for
(a) CuO, (b) BaCO3, and (c) Y2O3.
We restrict our study to the reactions between Y2O3,
BaCO3, and CuO. Since it is generally accepted that solid
state reactions are sensitive to the atmosphere in which
they occur, and more specifically to O2 partial pressures,
all TGA-DTA measurements were done both in air and
Ar.26,27
Separate Metal-Oxides and YBCO. Figure 3a shows
the TGA-DTA spectra obtained for CuO powder in air
and Ar respectively. In both cases a mass drop of ∼10% is
found which can be correlated to the reduction of CuO
to Cu2O. DTA analysis reveals that the reduction reac-
tion is endothermic in both atmospheres. Since oxygen is
(28) Pasierb, P.; Gajerski, R.; Rokita, M.; Rekas, M. Phys. B (Amsterdam,
Neth.) 2001, 304, 463.
(29) Gervais, M.; Odier, P.; Coutures, J. P. Mater. Sci. Eng., B 1991,
8, 287.
(30) Kosmynin, A. S.; Shter, G. E.; Garkushin, I. K.; Trunin, A. S.;
Balashov, V. A.; Fotiev, A. A. Supercond.: Phys. Chem. Technol. 1990,
3, S271.
(31) Krabbes, G.; Bieger, W.; Wiesner, U.; Ritschel, M.; Teresiak, A.
J. Solid State Chem. 1993, 103, 420.
(32) Zhang, W.; Osamura, K. Mater. Trans., JIM 1991, 32, 1048.
(26) Feenstra, R.; Lindemer, T. B.; Budai, J. D.; Galloway, M. D. J. Appl.
Phys. 1991, 69, 6569.
(27) Morris, D. E.; Markelz, A. G.; Fayn, B.; Nickel, J. H. Phys. C
(Amsterdam, Neth.) 1990, 168, 153.

4474 Inorganic Chemistry, Vol. 49, No. 10, 2010
Vermeir et al.
Figure 4. TGA-DTA spectra of YBa2Cu3O7-δ in air and Ar using (a) a heating ramp of 10 ꢀC min-1 and (b) a cooling ramp of 20 ꢀC min-1
.
Figure 5. TGA-DTA spectra in air and Ar for binary mixtures of (a) Y2O3:BaCO3 (Y/Ba = 1:2) and (b) Y2O3:CuO (Y/Cu=1:3) both heated at 10 ꢀCmin-1
.
the first mass losses start to appear at higher tempera-
tures, 1050 and 900 ꢀC respectively. This suggests that a
reaction between BaCO3 and CuO takes place.
At 990 ꢀC, a final mass loss of 15.2% is observed while
for a separated decarboxylation and reduction a theore-
tical mass loss of 17.7% is expected.
From the literature it is established that if BaCO3 and
CuO are mixed in a Ba/Cu 1:1 ratio, (i) CuO can act as a
catalyst in the decarboxylation of BaCO3 and (ii) decarboxy-
lation of BaCO3 proceeds simultaneously with the forma-
tion of BaCuO2.33 XRD analysis (Figure 8) confirms the
formation of BaCuO2, which is in agreement with the
BaCO3(BaO)-CuO phase diagrams in the literature.34-36
Figure 6. XRD spectrumofanY2O3:CuO (Y/Cu = 1:3) mixture heated
at 1000 ꢀC for 24 h in N2 (9) Y2O3, (b) Cu2O and (8) yttriumcuprate.
BaCO3 þ CuO f BaCuO2 þ CO2v
ðHÞ
In our case, excess CuO is present in the mixture so that
BaCuO2 can further react with the remaining CuO form-
ing a eutectic liquid.37
Figure 5b shows the TGA-DTA spectra of the binary
mixture of Y2O3 and CuO. The two compounds were mixed
in a Y/Cu-ratio of 1:3. As for pure CuO, two endothermic
peaks are found at 1060 and 940 ꢀC, respectively. TGA
analysis in Ar shows an extra mass drop of 1.5% which is
associated with a slightly endothermic deviation in the
DTA-curve. XRD analysis (Figure 6) of the final product
identified the existence of an yttriumcuprate.
2BaCuO2 þ CuO f liquid þ xO2v
ðIÞ
Since reaction H is quite slow,33 a low heating rate of
1 ꢀC min-1 was used in the TGA-DTA experiment to
improve conversion of reaction H, before reaction I starts.
Finally, Figure 7a represents the TGA-DTA spectra of
the binary mixture of BaCO3 and CuO. The two com-
pounds were mixed in a Ba/Cu-ratio of 2:3.
Under air, a first decrease in mass is detected at 700 ꢀC,
while for pure CuO (Figure 3a) and BaCO3 (Figure 3b),
(33) Itoh, T. J. Mater. Sci. Lett. 2003, 22, 185.
(34) Voronin, G. F.; Degterov, S. A. J. Solid State Chem. 1994, 110, 50.
(35) Wong-Ng, W.; Cook, L. P. J. Am. Ceram. Soc. 1994, 77, 1883.
(36) Zhang, W.; Osamura, K.; Ochiai, S. J. Am. Ceram. Soc. 1990, 73,
1958.
(37) Chu, P. Y.; Buchanan, R. C. J. Mater. Res. 1993, 8, 2134.

Article
Inorganic Chemistry, Vol. 49, No. 10, 2010 4475
Figure 7. TGA-DTA spectra of a binary mixture of BaCO3 and CuO (Ba/Cu = 2:3) in air and Ar when (a) heated at 1ꢀCmin-1 and (b) cooled at 20 ꢀCmin-1
.
When the TGA-DTA spectra are recorded using a
heating ramp of 10 ꢀC min-1 instead of 1 ꢀC min-1
(Figure 10), the formation of BaCuO2 is still incomplete
before CuO starts to reduce and liquid formation starts.
For this reason, BaCO3 phase transition (B) and CuO
reduction (A) are visible in the DTA-curves. As all
reactions takes place simultaneously, only the character-
istic peaks A and B are assigned.
The above results for binary precursor interactions
suggest that the reaction between BaCO3 and CuO has
a predominant role in the formation of YBCO starting
from non-fluorine precursors. To verify this, ternary
precursor interactions were studied.
Ternary Precursor Interactions. The TGA-DTA spec-
tra for a powder mixture of Y2O3, BaCO3, and CuO in
a Y/Ba/Cu-ratio of 1:2:3 are shown in Figure 11. This
ternary mixture starts to lose weight from 800 ꢀC. In Ar,
the mass loss until 910 ꢀC can be associated to the
simultaneous reactions involving (i) decarboxylation of
BaCO3, (ii) reduction of CuO, and (iii) formation of
BaCuO2. The theoretical mass loss of 12.9% matches
the experimental mass loss of 13%. This is in agreement
with the observations obtained for the binary BaCO3:
CuO mixture. At 920 ꢀC a small mass loss is observed in
TGA, probably resulting from the eutectic reaction be-
tween BaCuO2 and CuO. Subsequently, at 935 ꢀC, an
exothermic reaction J takes place: Y2O3 reacts most likely
with the eutectic liquid obtained in reaction I to form
Figure 8. XRD spectrum of a BaCO3:CuO mixture heated at 1000 ꢀC
for 1 h in N2, (*) BaCO3, (o) CuO, and (|) BaCuO2.
The first mass drop of 13.2% in the TGA-DTA spec-
trum corresponds with the formation of BaCuO2 accord-
ing to reaction H (13.9%). The small deviation in mass
loss can be associated to an incomplete reaction. Second,
BaCuO2 reacts eutectically with CuO in accordance with
reaction I. A gradual exponential mass loss in combina-
tion with an endothermic DTA signal during heating and
a microscopic image showing a typical pattern of solidifi-
cation after cooling (Figure 9) endorse this assumption.
Furthermore, this is in good agreement with the pub-
lished phase diagrams.34-36
A third mass loss can be correlated to a deviation in
oxygen content. During cooling (Figure 7b) an equivalent
amount of oxygen is absorbed. Furthermore, the DTA
analysis during cooling shows an exothermic deviation,
which can be associated with the solidification of the
liquid. Therefore, the exothermic peak in Figure 7b is
associated with the backward reaction of I and forms an
extra endorsement of the existence of reaction I.
Under Ar, a similar pattern is obtained. Here, the first
mass drop is 14.5% instead of 13.2%. This difference
corresponds to the reduction of the remaining CuO. In
air, this reduction does not take place because tempera-
tures of 1050 ꢀC and more are necessary to induce reduc-
tion. Yet, at 980 ꢀC all CuO present is already consumed
in reactions H and I.
The presence of a mass increase under Ar around
950 ꢀC is probably an anomaly because of an incomplete
adaptation of the furnace atmosphere.
YBa2Cu3O6þx
.
1
Y2O3 þ liquid f YBa2Cu3O6 þ xðsÞ
ðJÞ
2
At 970 and 990 ꢀC, the characteristic endothermic
peaks F and G correlate to the decomposition of YBCO.
The temperatures are similar to the decomposition temp-
eratures of pure YBCO-powder (Figure 4a). In air a
similar pattern is observed.
Part B: Thin Films. Since coated conductors are composed
of thin YBa2Cu3O7-δ films, it was verified if the YBCO reac-
tion mechanism presented here is also valid in the case of thin
films prepared from an aqueous fluorine-free sol-gel pre-
cursor. Therefore a HT-XRD experiment was performed. A
first experiment was set up to examine the reaction behavior in
BaCO3-CuO films. Therefore, a sol-gel precursor contain-
ing Ba- and Cu-acetate in a Ba/Cu ratio of 2:3 was dip-coated

4476 Inorganic Chemistry, Vol. 49, No. 10, 2010
Vermeir et al.
Figure 9. Microscopic image of the BaCO3-CuO powder after TGA-DTA analysis.
Figure 10. TGA-DTA spectra of a binary mixture of BaCO3:CuO
(Ba/Cu = 2:3) in air and Ar when heated at 10 ꢀC min-1
.
Figure 11. TGA-DTA spectra of a ternary mixture of Y2O3, BaCO3,
and CuO (Y/Ba/Cu = 1:2:3) in air and Ar when heated at 10 ꢀC min-1
.
on a STO substrate. The film was burned out at 500 ꢀC
resulting in a film containing BaCO3 and CuO. This tempera-
ture was chosen based on the TGA-DTA spectrum of the
separated acetates (Figure 1).
Subsequently, the film was placed in the furnace of the
HT-XRD and heated at 10 ꢀC min-1 until 750 ꢀC under a
flowing He atmosphere.
Figure 12. HT-XRD measurement of a BaCO3-CuO film under He
with a heating ramp of 10 ꢀC min-1 (a) full 2θ-scan plot (b) log intensity
plot of important compounds.
34.5ꢀ) and CuO (2θ = 35.5ꢀ) are present in the layer. At
640 ꢀC BaCO3 decomposes and BaCuO2 (2θ = 29.3ꢀ) is
formed. At that time, CuO reduces to Cu2O (2θ = 36.5ꢀ).
Figure 12 shows the HT-XRD measurement of a
BaCO3-CuO thin film. Up to 640 ꢀC, BaCO3 (2θ = 33.7ꢀ,

Article
Inorganic Chemistry, Vol. 49, No. 10, 2010 4477
was already successfully used in previous research to
prepare superconducting YBCO films on STO substrates
with the same precursor solution.24
Figure 13a shows the HT-XRD measurement of a
Y2O3-BaCO3-CuO thin film. Strong (00 L) STO reflec-
tions are present at 2θ values of 22.8ꢀ and 46.5ꢀ.
BaCO3 (2θ = 34.5ꢀ) and CuO (2θ = 35.5ꢀ and 38.7ꢀ)
reflections started to disappear around 650 ꢀC, with the
formation of BaCuO2 (2θ = 29.3ꢀ) as a result. Sub-
sequently, the formation of preferentially c-axis oriented
YBCO started (2θ = 14.9ꢀ, 37.9ꢀ and 54.2ꢀ).
This HT-XRD experiment confirms the results ob-
tained for bulk materials: (i) the formation mechanism
of non-fluorine prepared YBCO is dominated by the
Ba-Cu interactions, and (ii) BaCuO2 is an intermediate
in the formation mechanism of YBCO.
Conclusions
Within this study, the objective was to investigate (i) the
YBCO reaction mechanism for fluorine-free approaches and
(ii) the role of BaCO3 in the production of superconducting
materials. First, an intensive TGA-DTA study on bulk
materials suggests a three step mechanism for the formation
of YBCO.
In a first step BaCO3 and CuO react to form BaCuO2:
BaCO3 þ CuO f BaCuO2 þ CO2v
In a next step, which can be simultaneous with the previous
step, the intermediate BaCuO2 will react with CuO to form a
liquid:
2BaCuO2 þ CuO f liquid þ xO2v
Finally, the liquid will react with Y2O3 to form YBCO:
1
Y2O3 þ liquid f YBa2Cu3O6 þ x
2
Figure 13. HT-XRD measurement of a Y2O3-BaCO3-CuO film
under a 200 ppm O2:N2 atmosphere with a heating ramp of 10 ꢀC min-1
(a) full 2θ-scan plot and (b) log intensity plot of important compounds.
Within this mechanism, BaCO3 has clearly a predominant
role.
HT-XRD experiments confirmed that the results obtained
for bulk mixtures are also applicable in the case of thin films.
This mechanism explains the presence and consumption of
BaCO3 during thin film synthesis starting from non-fluorine
precursors.
This study establishes clearly that the presence of BaCO3 is
not fatal in obtaining an YBCO compound. Moreover, this
paper gives a suitable explanation to the fact why BaCO3 was
a desired intermediate phase in the production of YBCO thin
films starting from an aqueous non-fluorine precursor, as
published before.24
Small deviations from theoretical 2θ-values at higher
temperatures are the result of thermal expansion of the
material. The decomposition temperature of BaCO3 is
much lower in thin films (640 ꢀC) than in powders because
the grain size is much smaller for the particles in the
precursor solution than those obtained from powder
mixing.37
Second, a ternary Y2O3-BaCO3-CuO film was exa-
mined with HT-XRD. Therefore, a sol-gel precursor con-
taining Y-, Ba-, and Cu-acetate in a Y/Ba/Cu ratio of
1:2:3 was dip-coated on a STO substrate. Again, the film
was burned out at 500 ꢀC resulting in a film containing
Y2O3, BaCO3, and CuO. Subsequently, the film was
placed in the HT-XRD where it was heated until 900 ꢀC
at a heating rate of 10 ꢀC min-1. The ruling atmosphere
was a flowing 200 ppm O2:N2 mixture. This atmosphere
Acknowledgment. Bart Ruttens is gratefully acknow-
ledged for performing HT-XRD measurements. K.V. is
a postdoctoral fellow with the Fund for Scientific
Research-Flanders (Belgium) (FWO Vlaanderen).