Äîêóìåíò âçÿò èç êýøà ïîèñêîâîé ìàøèíû. Àäðåñ îðèãèíàëüíîãî äîêóìåíòà : http://www.cplire.ru/html/oxide233/paper5.pdf Äàòà èçìåíåíèÿ: Thu Sep 1 13:25:15 2005 Äàòà èíäåêñèðîâàíèÿ: Tue Oct 2 00:53:19 2012 Êîäèðîâêà: Ïîèñêîâûå ñëîâà: ï ï ï ï ï ï ï ï ï ï ï ï ï

Physica C 336 Z2000. 93 ­ 101 www.elsevier.nlrlocaterphysc

Particle formation on the YBCO thin film surface: effect of stoichiometry and substrate material
P.B. Mozhaev
a b

a,b, )

, F. Ronnung a , P.V. Komissinskii a,b, Z.G. Ivanov a , ¨ G.A. Ovsyannikov b

Chalmers Unióersity of Technology, Department of Microelectronics and Nanoscience, SE-41296, Goteborg, Sweden ¨ Institute of Radio Engineering and Electronics, Russian Academy of Sciences, Mokhoóaya Str. 11, 103907, Moscow, Russia Received 29 November 1999; accepted 29 February 2000

Abstract Particle formation on the YBa 2 Cu 3 O 7y x ZYBCO. thin film surface deposited by laser ablation was studied. Ablation of a stoichiometric YBCO target results in Ba-deficient films, probably due to Ba scattering in oxygen atmosphere. Partial melting of the target surface during ablation provided additional Ba flow and resulted in smooth films. Surface quality of the YBCO film with improved element composition depends on lattice mismatch between the growing film and the substrate. Increase of deposition temperature over some threshold level, specific for each tested substrate material, results in particle appearance on the film surface even in the films with improved element composition. Chemical interaction, caused by oxygen depletion of the substrate surface, can be the reason for this effect. q 2000 Elsevier Science B.V. All rights reserved.
PACS: 68.55.y a; 74.76.Bz; 81.15.Fg Keywords: YBCO thin films; Precipitates; Chemical interaction with substrate

1. Introduction Particle formation on the YBa 2 Cu 3 O 7y x ZYBCO. thin film surface seriously limits the application of these superconducting films. A huge amount of papers was published on this topic, see, e.g. Refs. w1 ­ 5x. Two reasons for the particle formation are generally stated. Partial decomposition of the film when the deposition conditions are far from optimal conditions of YBCO formation provides grains of

Corresponding author. Institute of Radio Engineering and Electronics RAS, Mokhovaya Str. 11, 103907, Moscow, Russia. Tel.: q 7-95-203-0935; fax: q 7-95-203-8414. E-mail address: pbmozh@hitech.cplire.ru ZP.B. Mozhaev..

)

individual oxides and decreases superconducting quality of the film. Local or overall non-stoichiometry of the film results in extraction of the excess material into particles of non-superconducting oxide phases. Excess Y is often incorporated into the YBCO film as Y2 O 3 inclusions while CuO, Y2 BaCuO5 , and different barium ­ copper oxides are usually observed as precipitates on the YBCO film surface w1 ­ 5x. Another type of particles often observed on the YBCO film surface are particles of YBCO of different orientations, i.e. Z100.-oriented Z a-oriented. inclusions in Z001.-oriented Z c-oriented. films and coriented inclusions in a-oriented films. Formation of these inclusions is controlled by kinetic factors, leading in general to a-oriented film growth at low deposition temperatures and c-oriented film growth

0921-4534r00r$ - see front matter q 2000 Elsevier Science B.V. All rights reserved. PII: S 0 9 2 1 - 4 5 3 4 Z 0 0 . 0 0 2 4 6 - X


94

P.B. Mozhaeó et al. r Physica C 336 (2000) 93 ­ 101

at high deposition temperatures w6,7x. In some cases, though, formation of a-oriented inclusions is caused by chemical interaction between the substrate and the deposited material in the very beginning of deposition. The products of such a reaction suppress epitaxial growth in their vicinity resulting in a-oriented particle seeding even at high deposition temperatures w8,9x. Laser ablation is one of the most often used techniques for production of high-quality YBCO thin films. The ablation process results in conservation of the target stoichiometry in the initially ablated cloud, but scattering processes during the material propagation can result in deviations from stoichiometry both on the axis of the propagating material Zplume. and in the off-axis regions w10,11x. In the present paper, we study the particle seeding on the surface of the YBCO thin film deposited by laser ablation. Different methods of particle formation suppression are suggested, and the effect of substrate material on the particle formation at different temperatures is investigated.

2. Experimental The laser deposition system was described in detail in Ref. w12x. In short, the KrF excimer laser provided an energy density on the target of up to 3 Jrcm2 with a pulse duration less than 30 ns. A one-lens spot formation optical system allowed the change of spot size and spot position on the target. Two spot positions were used ZFig. 1.: ``donut'' position I, covering a donut area during target rotation and thus improving the lifetime of the target, and ``centred'' position II with overlapping of spots

Fig. 1. Two possible positions of the laser beam on the target and areas, covered on the target during rotation. Sequential spots overlap in the target centre for position II.

in the target centre. The deposition pressure in the chamber was set in the 0.05 ­ 1 mbar range by adjustment of the oxygen flow at constant pumping rate of the evacuation system. The ThermoCoax heating element provided heating of the sample holder up to 8508C. The desired temperature and heating rate were controlled by a EuroTherm controller. The substrate was glued on the holder with silver paint, providing good thermal contact. The target ­ substrate distance was about 60 mm. Standard deposition conditions were: substrate temperature, 750 ­ 8008C; oxygen pressure, 0.8 mbar; laser beam energy density on the target, 1.7 Jrcm2 ; repetition rate, 10 pulsesrs; resulting in deposition rate about 0.6 ° Arpulse. The visible plume length at standard deposition conditions was about 45 mm, placing the substrate out of the plume. After deposition, the chamber was filled with oxygen to 900 mbar pressure and the samples were left at 400 ­ 4508C for 1 h providing necessary oxidation of the YBCO film. The superconducting properties of the deposited films were determined by measuring the magnetic susceptibility dependence on temperature. The critical temperature Tc was determined as the highest temperature at which the screening effect was observed. The width of the superconducting transition DTc , showing the degree of uniformity of the superconducting properties of the film, was measured between 10% and 90% levels of the maximal screening effect. The structure of the films was investigated using X-ray diffractional ur2 u-scans, providing information on phase composition and lattice parameters. Precise calculations of the interplanar distance were done using all available peaks w13x. The dependence of the ur2 u-scan peak width on 2 u was analysed to evaluate strain in the studied films w5x. Element composition of some films was studied using X-ray microanalysis. The electron beam scanned over a 100 = 100 m m2 , providing overall element composition of the film and particles on its surface. Film surface quality was characterised by two parameters: density of particles and surface roughness. The density of particles was determined using optical microscope, so only particles greater than 0.5 m m in size were taken into account for this calculation. The roughness parameter R a Zmean square of vertical distances between neighbour peak and valley. was determined using profile meter. The


P.B. Mozhaeó et al. r Physica C 336 (2000) 93 ­ 101

95

scan velocity was 2 m mrs, vertical accuracy was 5 ° A, and scan distance was 50 m m. 3. Results and discussion 3.1. Donut spot position The YBCO films deposited at standard deposition conditions on the LaAlO 3 Z001. substrates showed Tc of 88 ­ 908C and DTc less than 1.5 K; the critical current density at 77 K exceeded 10 6 Arcm2 . The X-ray investigations showed that the films were completely c-oriented with c lattice constant about ° 11.7 A and strain along substrate normal less than 0.5% Zsee Table 1.. No additional phase peaks were observed. The optical micrographs of typical YBCO films, deposited from a stoichiometric YBCO target, are shown in Fig. 2. The density of particles on the film surface varies from 5 = 10 6 to 10 8 cmy2 . A typical particle is shown on the upper-left inset of Fig. 2c; clear facets with hexagonal symmetry are observed. A precise tuning of the geometrical factors of deposition, such as target ­ substrate distance, substrate shift from the plume axis, laser beam energy density on the target and deposition pressure, decreases the particle density to 10 6 cmy2 , but even a slight change in the stated deposition conditions results in a rapid increase of particle density. Good superconducting properties of the deposited films imply deposition conditions close to the optimal for the YBCO film formation excluding particle formation due to decomposition of the film. This supposition is confirmed also by low Z- 0.5%. strain of the films ZTable 1.. The particle formation, hence, takes place mainly due to composition deviations from stoichiometric metal ratio 1:2:3. Comparison of the element composition of laser deposited films with

that of films deposited by DC-sputtering at high pressure w14x showed 5 ­ 8% Ba-deficiency in the laser deposited films. Evaluation of the particle volume ratio to the overall film volume gives close numbers ZTable 1.. Deposition at standard parameters from a target with excess Ba ZY1 Ba 2.2 Cu 3 O x . resulted in a smooth film with particle density less than 2 = 10 6 cmy2 without special deposition parameter tuning. Tc of the obtained films was about 88.5 K and DTc was less than 2.0 K. Ba-deficiency is often observed type of YBCO film non-stoichiometry, resulting in copper-rich particles on the film surface. Excess Y is incorporated into the film as Y2 O 3 inclusions and cannot be seen in SEM or optics w3x. The hexagonal shape of particles Zinset of Fig. 2c. can be produced by a Z111.-oriented phase with cubic lattice, like CuO. The X-ray ur2 u-scans showed some diffraction intensity increase at CuO Z111. peak position Z38.738., but close vicinity of a strong Z005. YBCO peak makes impossible clear peak observation. Two possible reasons for Ba-deficiency in the films deposited from stoichiometric YBCO targets can be suggested: Ba scattering in oxygen atmosphere and Ba re-evaporation from the growing film. Strong dependence of the particle density on geometrical factors suggests that scattering is the main cause. Both particle size and density depend on the deposition temperature ZFig. 2, Table 1.. At deposition temperatures less than 7608C, particles of 0.7 ­ 1.2 m m size were observed with densities Z5 ­ 10. = 10 6 cmy2 , while at temperatures higher than 7708C, typical particles were 0.5 ­ 0.8 m m in diameter and their density was up to 10 8 cmy2 . At 7658C, both types of particles could be observed, supposing presence of two different seeding mechanisms. Evaluation of the particle volume gives the same factors for

Table 1 Properties of the YBCO thin films deposited on LaAlO 3 Z001. substrates by laser ablation with donut spot position, and parameters of numerical simulation of particle seeding Deposition temperature Z8C. 750 765 780 C lattice constant ° ZA. 11.714 11.698 11.702 Strain Z%. Average particle size Zm m. 0.95 1.1 0.75 Average distance between particles Zm m. 6 8 1.5 Particles volume Z% of the film volume. 6.9 6.6 6.5 Simulation parameters Excess material Z%. 5 5 5 Diffusion length Zm m. 4.5 10 32 Stoichiometry fluctuations Z%. 5 5 5

0.18 0.12 0.31


96

P.B. Mozhaeó et al. r Physica C 336 (2000) 93 ­ 101

rate on temperature excludes re-evaporation as possible reason for the film overall non-stoichiometry. We performed computer simulations of the particle seeding with the following suppositions: Zi. only excess material is incorporated into particles; Zii. excess material amount fluctuates along the surface; Ziii. seeding of a particle becomes possible when one monolayer of excess material Ze.g. CuO. is present on the film surface. Seeding probability was set to 1r2 for simplicity. Excess material is added at moments t N Zlaser pulses. c Z x , t ) t N . s c Z x , t N . q cp Z 1 q k j . ,

Z 1.

where cZ x , t . is excess material amount on the film surface, c p is the excess material amount in the plume, k is the fluctuations intensity, and j is a stochastic variable uniformly distributed in wy1r2, 1r2x. If the particle was seeded in some point x p on the surface, the excess material around the particle is supposed completely consumed: cZ x p , t ) t N . s 0. The excess material distribution between the pulses was ``smoothened'' by diffusion E crE t syDE 2 cr Z E x . ,
2

Z 2.

Fig. 2. Micrographs of YBCO thin films deposited on LaAlO 3 Z001. substrates from a stoichiometric YBa 2 Cu 3 O 7y x target using laser ablation with donut laser beam spot position. Substrate temperature: Za. 7508C; Zb. 7658C; Zc. 7808C. On the right-bottom insets: simulated particle pattern, simulation parameters in Table 1. On the upper-left inset Zc.: SEM micrograph of a typical particle.

all deposition temperatures, proving the same level of overall non-stoichiometry at all deposition temperatures. Strong dependence of Ba re-evaporation

2 where D s l d rt is the diffusion constant: l d is the characteristic diffusion length and t is the characteristic diffusion time. Diffusion ``smoothening'' makes seeding of the particles possible only immediately after laser pulse; additional surface fluctuations of the excess material were not taken into account. The simulation parameters were chosen using the observed particle patterns: c p should be approximately the observed volume of particles relative to the overall film volume and l d should be close to the observed distance between big particles ZTable 1.. To obtain necessary l d numerically, the time scale of the simulation Znumber of smoothing steps between pulses. was changed. Characteristic diffusion time t was chosen as time of diffusion transport of material from the target to the substrate being about 0.01 s w15x. Fluctuation intensity k was used as tuning parameter. One-dimensional test simulations were performed first and after that Eqs. Z1. and Z2. were generalised for the two-dimensional case. The simulated patterns similar to the observed ones are shown in the bottom-right corners of Fig. 2 and the simulation parameters used are stated in Table 1.


P.B. Mozhaeó et al. r Physica C 336 (2000) 93 ­ 101

97

Possible reason for particle seeding at low deposition temperatures is fluctuation of the element distribution on the substrate surface. The threshold amount of excess material is accumulated on the film surface during some initial period ZFig. 3 inset, 75 ­ 90 laser pulses.. After seeding the growing particle acts as a sink for excess material, suppressing seeding of new particles in its vicinity ZFig. 3 inset, 95 ­ 110 laser pulses.. Far from the particle, new seeding still can occur, prolonging the period of particle seeding ZFig. 3, dotted line.. The resulting particles are big and sparse; the particle size depends on the distance to the neighbour particles. Experimental investigations of particle seeding on the early stages of YBCO film formation showed no particles on the film surface Zsee, e.g. Ref w16x.. The particles appear during a short period of deposition and suppress further seeding w16x. These observations are in good agreement with suggested simple particle seeding model. `` Fluctuational'' seeding mechanism explains YBCO film surface morphology at low deposition temperatures, with increase of particle size and distance between them with increase of deposition temperature. The small distance between particles at high deposition temperature ZFig. 2c. can be explained as sudden decrease of diffusion intensity

along the surface, for example, as a result of characteristic time increase, but presence of both dense and sparse particles at intermediate temperature ZFig. 2b. demands second seeding mechanism. Such a mechanism can be provided by an increase of diffusion at high deposition temperatures, making it sufficient to smoothen all fluctuations before a particle can be seeded due to fluctuation. In this case, the amount of excess material increases almost uniformly along the film surface and seeding of particles occurs in many places simultaneously ZFig. 3, solid line.. The area per particle decreases, and the size of resulting particles is smaller than in the case of fluctuational seeding. Both mechanisms coincide at intermediate diffusion intensities, as can be seen on inset Fig. 2b, with fluctuational seeding on earlier stages of deposition and accumulational seeding on later stages ZFig. 3, dashed line.. 3.2. Centred spot position We found out that shifting the laser beam spot position to the centre of the stoichiometric YBCO target Zposition II, Fig. 1. results in a dramatic decrease of the particle density on the film surface to less than 10 6 cmy2 . This effect is observed for both

Fig. 3. Simulation of the particle seeding during deposition: increase of diffusion length decreases the seeding period. Lines correspond to insets Fig. 2a ­ c. On the inset: cross-section of the excess material distribution after 75 ­ 110 laser pulses, crosses mark positions of the particles seeded in vicinity of the presented cross-section. Simulation parameters same as on the Fig. 2b inset.


98

P.B. Mozhaeó et al. r Physica C 336 (2000) 93 ­ 101

spot position crossing the centre of the target and for spots just overlapping in the centre. The shape and size of the plume do not differ significantly from the case of donut ablation. The central part of the target after ablation is partially molten in contrast to the outer regions ZFig. 4.. The typical for laser ablation columnar hillocks on the target surface are tilted to the laser in case of donut ablation and normal to the target surface when the spot is centred. Melting of the target in the area of spot overlapping seems to be the reason for the improved surface quality of the film. The post-ablation Ba flow from the target surface was observed in Ref. w11x; possible reason for such additional Ba emission can be heating of the target by the ablated cloud. To test heating effect, we decreased the laser beam repetition rate two times resulting in increase of the particle density to 2 = 10 6 cmy2 . Further decrease of the repetition rate restored the particle density to values typical for donut abla-

Fig. 4. Micrographs of the stoichiometric YBa 2 Cu 3 O 7y x target after 770 laser shots with centred laser spot position: Za. outer region of the target; Zb. central part of the target.

tion. The melting temperature of any combination of target elements is higher than 8008C w17x. Ba evaporation rate at these temperatures is very high, so one can suppose additional Ba flow from the molten parts of the target, compensating Ba deficiency due to scattering. Using the centre ablation technique, we studied deposition of smooth YBCO films on different substrates in temperature range 760 ­ 8208C. The deposition was performed simultaneously on four substrates: Z001. LaAlO 3 , Z110. NdGaO 3 , Z001. SrTiO 3 , and Z001. CeO 2-buffered Z1102. sapphire. The YBCO film parameters at deposition temperature 7808C are given in Table 2. The surface smoothness depends strongly on the lattice mismatch between the growing film and substrate ZFig. 5.: both density of particles and roughness of the film monotonously increase with the lattice mismatch for deposition temperatures less than 7908C. Lattice mismatch was calculated as Z d F y d S .rd S = 100%, where d F s Z a 2 ° q b 2 .1r 2 s 5.454 A is the w110x translation distance of YBCO and d S is the corresponding translation distance of the substrate surface, namely w110x LaAlO 3 , w001x NdGaO 3 , w110x SrTiO 3 , and w100x CeO 2 , all taken at room temperature. Dependence of surface morphology on lattice mismatch, probably, results from prolonged layer-by-layer growth on the substrate with smaller lattice mismatch w18x. The superconducting parameters and crystal quality of the films show generally opposite behaviour: films on LaAlO 3 have highest critical temperature and lowest lattice strain compared with films on substrates with smaller lattice mismatch. Lattice strain of the YBCO films on LaAlO 3 substrates deposited using donut ablation is even smaller Zsee Table 1., supposing dependence of the film strain on particle density. More uniform oxygenation of the films along the particles w19,20x and strain relaxation at the particles sites can be the reasons for this effect. With increase of deposition temperature over some threshold level, many small particles appear on the film surface ZFig. 6.. The threshold level was about 7908C for CeO 2-buffered sapphire and about 8108C for SrTiO 3 , LaAlO 3 , and NdGaO 3 . Formation of a-oriented particles is observed usually at temperatures 10 ­ 208C below threshold Zsee inset Fig. 6.. Three possible reasons for particle formation at high deposition temperatures can be supposed: Zi. devia-


P.B. Mozhaeó et al. r Physica C 336 (2000) 93 ­ 101

99

Table 2 Properties of the YBCO thin films, simultaneously deposited on different substrates by laser ablation with centred position of the spot at standard deposition conditions; deposition temperature 7808C Parameter Substrate NGOZ110. Critical temperature ZK. Superconducting transition width ZK. ° Surface roughness ZA. Particle density on the film surface Z=10 5 cmy 2 . ° Lattice constant c ZA. FWHM Z005. Z8. Film strain Z%. 89.6 1.2 25 0.7 11.70 0.24 0.8 CeO 2 Z001. I Al 2 O 3 Z1102. 89.4 0.3 30 3 11.68 0.17 0.6 STOZ001. 89.6 0.5 40 5 11.685 0.22 0.75 LAOZ001. 90.0 0.9 100 11 11.70 0.2 0.6

tion from optimal YBCO formation conditions and partial decomposition of the film; Zii. Ba re-evaporation from the substrate resulting in overall non-stoichiometry of the film; Ziii. chemical interaction with the substrate suppressing epitaxial film growth. Both superconducting properties and crystal quality of the deposited films improve with increase of temperature even after appearance of the particles ZFig. 7.. This improvement implies close vicinity of the deposition conditions to the optimal conditions of YBCO formation and thus excludes partial decomposition of the film. On the other hand, if Ba re-evaporation is a

main reason for particle appearance, the threshold temperature should be the same for all studied substrates. Chemical reactions of the deposited YBCO film with substrates were observed for many substrate materials w21 ­ 23x. a-oriented inclusions, as signs of chemical interaction, were observed for NdGaO 3 substrates w8,9x. Appearance of a-oriented particles at temperatures slightly lower than threshold deposition temperature allows supposition of the chemical nature of particle formation at high deposition temperatures. The reason for chemical interaction can be oxygen depletion of the substrate surface after heating to deposition temperature in vacuum w24x. We

Fig. 5. Surface quality dependence on the lattice mismatch between the substrate and the YBCO thin film. The YBCO films on four different substrates were deposited simultaneously by laser ablation using the centred laser spot position, deposition temperature 7808C.

Fig. 6. Micrograph of the YBCO thin films deposited on NdGaO 3 Z110. substrate using laser ablation of a stoichiometric YBa 2 Cu 3 O 7y x target with centred laser beam spot position, deposition temperature 8208C. On the inset: SEM micrograph of the a-oriented particles appearing with the increase of deposition temperature.


100

P.B. Mozhaeó et al. r Physica C 336 (2000) 93 ­ 101

element composition. Chemical interaction due to oxygen depletion of the substrate surface can be the reason for particle formation at high deposition temperatures.

Acknowledgements Authors would like to thank Dr. P. Larsson, Chalmers University of Technology, Gothenburg, Sweden, for the help in experiments, Dr. T. Claeson and Dr. D. Winkler, Chalmers University of Technology, Gothenburg, Sweden, for the useful discussion. The work was supported in part by the ESPRIT contract, 23429 HTS-RSFQ, Swedish Material Consortium on superconductivity, Russian State Program `` Modern Problems of the Solid State Physics'', ``Superconductivity'' division, Russian Foundation for Basic Research and INTAS program of EU.

Fig. 7. Dependence of the YBCO thin film properties on the deposition temperature. Films deposited on SrTiO 3 Z001. substrate by laser ablation with the centred spot position. The a-oriented particles appear on the film surface at about 7908C, precipitates observed at temperatures above 8108C.

deposited films on Z001. SrTiO 3 and Z001. LaAlO 3 substrates after leaving them for 30 min in vacuum at 7808C. The resulting films were polycrystalline and showed no superconductive transition. Even 10 min dwell of the substrate in vacuum at deposition temperature decreased critical temperature of the resulting film to 86 ­ 87 K. No effect of dwell at 760 ­ 7808C in 0.6 mbar oxygen was observed.

References
w1x S. Proyer, E. Stangl, M. Borz, B. Hellebrand, D. Bauerle, Physica C 257 Z1996. 1. w2x D. Bhatt, Physica C 222 Z1994. 283. w3x K. Verbist, A. Kuhle, A.L. Vasiliev, Physica C 269 Z1996. 131. w4x M. Kawasaki, M. Nantoh, MRS Bull. XIX Z9. Z1994. 33. w5x J.P. Gong, M. Kawasaki, K. Fujito, R. Tsuchiya, M. Yoshimoto, H. Koinuma, Phys. Rev. B 50 Z5. Z1994. 3280. w6x F. Miletto Granozio, U. Scotti di Uccio, J. Cryst. Growth 175 Z1997. . w7x U. Jeschke, R. Schneider, G. Ulmer, G. Linker, Physica C 243 Z1995. 243. w8x Yu.A. Boikov, Z.G. Ivanov, E. Olsson, V.A. Danilov, T. Claeson, M. Shcheglov, D. Erts, Phys. Solid State Fizika Tverdogo Tela 37 Z3. Z1995. 880, in Russian. w9x P.B. Mozhaev, G.A. Ovsyannikov, S.N. Polyakov, E.K. Kov'ev, N.P. Kukhta, Supercond.: Phys., Chem., Tech. 9 Z1996. 304, in Russian. w10x Z. Trajanovich, L. Senapati, R.P. Sharma, T. Venkatesan, Appl. Phys. Lett. 66 Z18. Z1995. 2418. w11x Y. Nakata, W.K.A. Kumuduni, T. Okada, M. Maeda, Appl. Phys. Lett. 66 Z23. Z1995. 3206. w12x P. Larsson, Sub-m icron structures in laser ablated YBa 2 Cu 3 O 7y x thin films, PhD Thesis, Chalmers University of Technology, 1999. w13x M.S. Raven, E.E. Inametti, Y.M. Wan, B.G. Murray, Supercond. Sci. Technol. 7 Z1994. 462. w14x P.B. Mozhaev, P.V. Komissinski, N.P. Kukhta, G.A. Ovsyannikov, J.L. Skov, J. Supercond. 10 Z3. Z1997. 221.

4. Conclusions We studied particle formation on the YBCO thin film surface deposited by laser ablation technique. Films deposited by ablation of a stoichiometric YBCO target were Ba-deficient, probably due to Ba scattering in oxygen atmosphere. Partial melting of the target surface during ablation resulted in smooth films; Ba evaporation from molten parts can be the reason for this effect. Decrease of lattice mismatch between the growing film and the substrate improves smoothness of the film surface. Deposition of YBCO films at different temperatures showed the presence of a threshold temperature, specific for each substrate material. At a temperature higher than threshold, particles appear even in the films with improved


P.B. Mozhaeó et al. r Physica C 336 (2000) 93 ­ 101 w15x P. Kubat, P. Engst, Z. Zelinger, J. Wild, P. Bonacek, J. Appl. Phys. 77 Z6. Z1995. 2822. w16x M.G. Norton, R.R. Biggers, I. Maartense, E.K. Moser, J.L. Brown, Physica C 233 Z1994. 321. w17x I.E. Graboy, A.R. Kaul', Yu.G. Metlin, Results of science and technique VINITI AS USSR, Chem. Solid State 6 Z1989. 3, in Russian. w18x M. Ece, E.G. Gonzalez, H.-U. Habermeier, B. Oral, J. Appl. Phys. 77 Z4. Z1995. 1646. w19x G. Ockenfuss, R. Wordenweber, T.A. Scherer, R. Unger, W. Jutzi, Physica C 243 Z1-2. Z1995. 24.

101

w20x S.C. Tidrow, W.D. Wilber, A. Tauber, S.N. Schauer, D.W. Eckart, R.D. Finnegan, R.L. Pfeffer, J. Mater. Res. 10 Z7. Z1995. 1622. w21x C.T. Cheung, E. Ruckenstein, J. Mater. Res. 4 Z1. Z1989. 1. w2 2 x G . L . S k o fro n ic k , A . H . C a rim , S . R . F o lty n , R . E . Munechausen, J. Appl. Phys. 76 Z8. Z1994. 4753. w23x A.D. Mashtakov, I.M. Kotelyanskii, V.A. Lusanov, P.B. Mozhaev, G.A. Ovsyannikov, I.K. Bdikin, Tech. Phys. Lett. Pis'ma v ZhTF 23 Z19. Z1997. 8, in Russian. w24x D.M. Hill, H.M. Meyer, J.H. Weaver, J. Appl. Phys. 65 Z12. Z1989. 4943.