Growth of strongly orientated lead sulfide thin films by successive ionic layer adsorption and reaction (SILAR) technique Tapio Kanniainen," Seppo Lindroos, Jarkko Ihanus and Markku Leskela Department of Chemistry, P.O. Box 55, FIN-0001 4 University of Helsinki, Finland Lead sulfide thin films were grown at room temperature by the successive ionic layer adsorption and reaction (SILAR) technique on soda lime glass, IT0 and A1203 covered glass, SO2, ( 1OO)Si and ( 11 l)Si substrates. SILAR utilises sequential treatment of the substrate with aqueous precursor solutions. Dilute solutions of lead acetate and thioacetamide were used as precursors for Pb2+ and S2-, respectively. The lead precursor solution also contained triethanolamine (tea) as a complexing agent, with a Pb: tea mole ratio of 1 :2.On glass the growth rate was 0.12 nm per cycle with 0.2 mol dmP3 lead and 0.4 mol dm-3 thioacetamide solution. The appearance of the films was metallic. X-Ray diffraction studies revealed a strong [200) orientation of the films. According to the Rutherford back-scattering (RBS) and nuclear reaction analysis (NRA) results the films were stoichiometric PbS and contained small amounts of some lighter impurities, possibly 0 and H. Scanning electron microscope (SEM) images revealed that the films were rather rough and consisted of grains with a diameter approximately corresponding to the thickness of the film. Lead sulfide thin films have been used during recent years in photoconductive devices as detectors. Owing to its suitable bandgap of 0.41 eV, PbS can be used as an IR detector in various applications such as scientific instruments and indus- trial and military equipment. PbS thin films can be grown from solution either at ambient or at elevated temperatures and in the gas phase.Techniques utilizing aqueous conditions are chemical bath deposition (CBD),lP5 the 'thin liquid film' method, (which closely resembles CBD)6, successive ionic layer adsorption and reaction (SILAR)7*8 and electrodepo~ition.~ Gas phase methods used for the growth of PbS thin films are atomic layer epitaxy (ALE)"," and molecular beam epitaxy (MBE).I2 CBD is a widely used technique due to the simplicity of the procedure and the high quality of the produced films. Furthermore, the equipment is uncomplicated and the precur- sors are in most cases common laboratory chemicals.On the other hand, the procedure is a bulk reaction, where adjustment of the process is not an easy task. The advantages of the aqueous techniques like CBD and SILAR are: simple pro- cedure, low temperature and low cost. In addition, SILAR utilises a sequential growth mechanism, which makes process control straightforward. The SILAR technique for the deposition of thin films was introduced by Nicolau in the mid-1980s.I3 It is based on a heterogeneous reaction between adsorbed ions and solvated ions on the solid-liquid interface. The substrate is treated separately with each aqueous precursor solution so that the individual steps, adsorption and reaction, can take place.During the first step, cations are adsorbed onto the substrate, and in the next step all the excess unadsorbed cations are washed away by rinsing the substrate with purified water. The rinsing is followed by the reaction step, during which the substrate is immersed into a vessel containing the aqueous anion precursor solution. When the solvated anions enter the diffusion layer they react with the adsorbed cations, and a solid adsorbed compound is formed on the surface. Again the ions in the diffusion layer are washed away with a rinsing pulse. By repeating these deposition steps a thin smooth film can be grown layer by layer. The thickness of the growing film is controlled by the number of deposition cycles.So far the SILAR technique has been mainly used to grow 11-VI compound thin films on various Our previous study on the growth of PbS thin films by the SILAR technique from lead acetate and nitrate [Pb(CH,COO), and Pb(N03),] together with sodium sulfide and thioacetamide (Na,S and CH3CSNH2) precursors showed that this deposition method allows one to affect the crystallization and the crystal orientation, as well as the grain size of the films.7 The purpose of this work was to further investigate the deposition and properties of PbS thin films grown with the SILAR method using complexation of lead. Experimental For the deposition a new apparatus, shown schematically in Fig. 1, was used. The new system was based on a modified Gilson Model 232 XL automatic sample processor, which was controlled by a PC via an RS232 interface.A substrate holder was mounted in the place of the needle. The xyz robot allowed full control of the substrate positioning as well as the dipping and rinsing times. The reaction vessels were standard 250 cm3 glass laboratory beakers. The rinsing vessels were constructed in the same way as in the older eq~ipment.'~.'~ The rinsing was carried out in water with 18 MQcm resistivity in order to thoroughly rinse the diffusion layer. The water was provided by a Millipore equipment (Milli-Q) and the flow rate was 250 cm3 min-'. The immersion times for the cation adsorption and the anion reaction were 20 and 40 s, respectively. The rinsing time was 70s.During the immersion and the rinsing the substrates were continuously moved by 1mm side-to-side lN21 1 I[--I I I 1 GILSON x/z Fl REACTION VESSELS Fig. 1 Schematic diagram of the new SILAR deposition machine J. Mater. Chem., 1996, 6(2), 161-164 161 in order to enhance convection around the substrate. The total time required for one deposition cycle was 215 s; these cycles were repeated as many times as required. The depositions were carried out under an N, atmosphere. Aqueous solutions of lead acetate [Pb(CH,COO),, Merck p.a.1 complexed with tea [N(CH,CH,OH),, Riedel-deHaen p.a.1 for lead, and with taa (CH3CSNH2, Merck, p.a.) for the sulfide ion were used as precursors. The concentrations of the solutions were 0.05-0.60 and 0.10-1.20 mol drn-,, respectively (ie.Pb: S mole ratio 1 : 2). The Pb :tea mole ratio was 1 :2.l' The pH values were 7.0-7.8 for Pb(tea);+ and 4.7 for the taa solution. The substrate materials used were soda lime glass, ITO-covered ( 150 nm) soda lime glass, A1,03-covered ( 150 nm) soda lime glass, SiO,, ( 1OO)Si and ( 11 1)Si. The substrates were cleaned ultrasonically, rinsed with acetone and ethanol and dried overnight before deposition. The Si substrates were etched with hydrofluoric acid for 2min and washed with 18 MR cm water for 30 min inside the N2 gas chamber just before the deposition. The amount of lead in the films was determined by dissolving the film in concentrated nitric acid and measuring the amount of Pb2+ ions, either using a Techron atomic absorption spectro- photometer or in basic media with a Metrohm 626 polaro- graph. The nominal thickness of the film was calculated using the bulk density of PbS (7.5 g cm-,).The crystal structure, crystallite orientation and crystal size were determined using a Philips MPD 1880 X-ray powder diffractometer using Cu-Ka radiation. The surface morphology of the films was characterized with a Zeiss DSM 962 scanning electron microscope (SEM). The chemical composition and thickness of the films were also studied by Rutherford back-scattering spectrometry (RBS) of 2.0 MeV 4He+ ions from the 2.5 MV Van de Graaf acceler- ator of the Accelerator Laboratory, University of Helsinki.,' Hydrogen profiling was carried out by the nuclear reaction analysis (NRA) technique using a 15N2+ beam from the 5 MV tandem accelerator EGP-10-11 of the Accelerator Laboratory to excite the 6.385 MeV resonance of the 'H(15N, a, y)12C reaction.,' Results and Discussion Growth rate According to the principles of SILAR, film growth proceeds by the adsorption of lead ions and the consecutive reaction of the sulfide precursor with the adsorbed lead ions, hence the theoretical maximum growth rate would be one monolayer of lead sulfide per deposition cycle.On [2OO]PbS this can be defined as the interatomic distance of the atoms (cu. 0.30 nm). The maximum growth rate i.e. the nominal film thickness divided by the number of cycles, was 0.12 nm per cycle, corresponding to an average surface coverage of 40%.This value has already been achieved with the 0.2 mol dm-, lead and 0.4 mol dm-, sulfur precursor solutions. More concen- trated solutions did not increase the growth rate. On the other hand, smaller concentrations than these resulted in a slower growth rate. The growth rate may be limited sterically by both the size of the adsorbed Pb species and the size of the acetate counterion occupying the outer Helmholtz layer, and therefore the density of the smaller lead ions in the inner Helmholtz layer is limited. The growth is also limited by the dissolution of PbS in acidic solutions. The anionic precursor solution had a pH value of 4.7, which is low enough to dissolve some PbS.22v23 This was also indicated by some small black PbS precipitates in the taa solution. Furthermore, a longer dipping time than 40s in anionic solutions resulted in a lower growth rate.On IT0 and SiOz substrates the growth rate was roughly the same as that on the soda lime glass. The growth rate on 162 J. Muter. Chem., 1996,6(2), 161-164 both ( 1OO)Si and (1 11)Si was slightly lower than that on glass, similar to the results achieved for PbS films grown by CBD.24 On A1,03-covered glass the attachment of the film was poor. The growth rate observed using Pb(tea),,+ as the lead precursor was higher (0.12nm per cycle) than that from a solution without the complexation (0.04 nm per ~ycle).~ The pH value of the complexed precursor solution was 7.8, whereas that of the uncomplexed precursor was 6.5.7 On natural PbS (galena surface), the adsorption of the Pb(0H) + ion increased with increasing pH due to an increase in the number of negatively charged surface sites available for ad~orption.,~ Hence, the more effective adsorption of the lead complex from the precursor solution with the higher pH value is reasonable.Compared to CBD which is the method used most often to grow PbS thin films, the growth rate using the SILAR tech- nique was still lower. In CBD, all the reagents are present in the reaction vessel at the same time, hence a high growth rate is achieved. The average growth rate achieved by CBD has been up to several tens of nm min-l, independent of the precursors ~sed.'-~.~ On the other hand, in the gas phase analogy to SILAR, i.e.atomic layer epitaxy (ALE; both methods utilize sequential introduction of precursors), the overall growth rate varied between 0.01 and 0.09 nm per cycle depending on the precursor and the substrate temperature." X-ray diffraction As indicated in Fig. 2, where an XRD pattern of an as-grown 200nm thick PbS film on a soda lime glass substrate is depicted, the films were strongly [200] orientated. The intensit- ies of the other existing peaks in Fig. 2 were very small (Table 1). It has previously been stated that the lead counter- ion affects the orientation of the SILAR-grown PbS thin films.7 Compared to those results, it can be seen that the complexed lead precursor remarkably enhances the orientation. This obviously occurs by guiding the central lead atoms during adsorption to match the right distance of lead atoms on the PbS(200) plane.The Pb(tea)2+ complex is rather large I3' c.30. 9 .-25. v).c 8-C.-2e /degrees Fig.2 X-Ray diffraction pattern of a 200nm thick PbS thin film grown by SILAR on soda lime glass at room temperature Table 1 Six strongest reflections in the XRD patterns of the PbS thin films grown on glass, SO2, (100)Si, (1 11)Si and IT0 covered glass substrates (the pattern for PbS powder is presented for comparison26) substrate h k 1 glass SiO, (1OO)Si (111)Si IT0 powder 111 0 1 1 1 50-80 84 200 100 100 100 100 100 100 220 0 2 1 0 30 57 311 0 0 0 0 20 35 222 0 0 0 0 10 16 400 4 4 5 5 a- 10 a This reflection was masked by an IT0 reflection.0s OPb Fig. 3 Representation of the diagonal orientation of the Pb(tea),'+ complex on the (200)PbS surface. The PbS(200) plane is tilted through 66". (ca. 0.7 xO.5 nm2),27 and therefore it cannot bring its lead atom very close to the adjacent lead atom. The two tea groups are bound to the central lead atom via oxygen atoms. In an ideal case, i.e. the complex approaches an ideal PbS(200) plane, the ligands are parallel to the substrate surface and the oxygen atoms of the opposite ethanol groups may also affect the orientation of the complex as they are closest to the substrate surface and have an affinity to interact with the Pb atoms of the surface.On the PbS(200) plane this leads to a configuration where the complex is along the diagonal on the (200) plane, and the negative oxygen atoms, by reaching the positive lead atoms of the surface, guide the lead atom of the complex accurately on top of a sulfur on the surface. Hence the [2001 orientation is favourable. The possible approach of the Pb(tea)2f complex to the (200)PbS surface is illustrated in Fig. 3. The intensities of the six strongest reflections in the XRD patterns of the as-grown PbS thin films on various substrates are listed in Table 1. On amorphous glass and Si02 as well as on single-crystal Si substrates the [2001 direction was strongly dominating. One clear exception from this tendency was the XRD pattern of a film grown on polycrystalline ITO.Besides resembling the PbS powder pattern with the (111) reflection as the strongest peak, the intensities of the peaks were very low compared to those observed for films on other substrates. This indicates a lower crystallinity of PbS films on ITO. Annealing for 2 h at 120°C in air had no effect on the XRD patterns of the films. The CBD method employs aqueous solutions as precursors and is carried out at room temperature, similarly to the SILAR method. However, the XRD pattern of the PbS thin films grown on amorphous glass by CBD are in general powder- like, regardless of the lead precursors used, and no orientation has been found,2 even with the same precursors as used in this st~dy.~The elevated deposition temperature (up to 50 "C) using a 'thin liquid film' method was found to enhance crystallinity, but the diffraction pattern was still powder-like.6 On the other hand, PbS thin films grown by CBD on (1OO)Si were (100) ~rientated,~~.~*whereas on ( 11 1)Si the 220 and 31 1 reflections dominated.24 The sequential introduction of reactants in the SILAR method involves heterogeneous reactions between the adsorbed species and the solvated ions in the solution.This layer-by-layer growth affects the crystal structure of the grow- ing film. Similarly to SILAR, the ALE method also utilizes sequential reactant pulses followed by inert rinsing/purging. Although ALE utilizes higher temperatures (300-350 "C for PbS) and reduced pressure, the XRD pattern of the PbS films grown on glass closely resembled the powder pattern. However, by suitable choice of the lead precursor, orientated films were achieved on alumina covered glass, where the (200) orientation was prevailing.lO*ll Electrodeposition of PbS thin films from an acidic solution containing Pb(N03)2 and Na2S203 on titanium sheets resulted in a powder-like XRD at tern.^ The crystallite size derived from the Scherrer equation for a 200 nm thick film was ca.100 nm. By comparison, CBD-grown PbS thin films on glass (thickness also ca. 200 nm)2 and electrodeposited PbS on a titanium sheetg resulted in crystallite sizes of 40 and 13 nm, respectively. However, larger crystallite sizes have also been observed in CBD films, uiz.crystallite sizes of 200nm in 400 nm thick films.24 The ratio of film thickness :crystallite size is in this case similar to that observed in this work.SEM studies The topography of an as-grown PbS thin film deposited on a soda lime glass is presented in Fig. 4(a), where the top view of a SEM image is depicted. The appearance of the film is rather rough and inhomogeneous, however, no cracks or voids can be detected. The roughness is caused by the preferred crystalline growth and the cubic crystallization is clearly seen in some of the particles. A side view [Fig. 4(b)] reveals that the particles are columnar and reach down to the substrate, and that the particles are well separated. The sizes of these particles, as estimated from Fig.4, were ca. 200 nm, which is similar to the thickness of the film. The grain size is approximately the same as those produced from PbS thin films grown with SILAR using uncomplexed lead acetate solution as the lead precursor. Fig. 4 SEM image of a 200 nm thick PbS film grown on soda lime glass. (a) Top view, (b)tilted view (45"). J. Mater. Chem., 1996, 6(2), 161-164 163 The grain sizes are much smaller than the 1 pm found in thicker PbS films grown by ALE'' and CBD3 with the same precursors, i.e. lead acetate, tea and taa precursors. When lead nitrate, NaOH and thiourea were used to grow PbS thin films by CBD, the grain size varied from values close to our results' up to 1-10 ~m.'~ RBS and NRA According to RBS analysis, the films contained equal amounts of lead and sulfide.In addition, 6-8 atom% of lighter impurities distributed evenly throughout the film could be detected. These elements (H, C, N or 0)could not be detected individually due to the limited resolution of the RBS method and the roughness of the film, which makes the fitting of the RBS results difficult. The amount of impurities present is roughtly the same as that in ZnS thin films grown by SILAR, which contained water and possibly zinc hydr0~ide.l~ Accordingly, because in NRA 6-8 atom% of hydrogen was found, it is likely that the PbS films contain water. Natural PbS (galena) was found to react with air to give lead carbonate and/or lead hydroxycarbonate species on the galena surface,23 and after several days of exposure to air sulfate species were also dete~ted.~'Hence, as the films were stored in air it is possible that the impurities are of atmospheric origin.Annealing for 2 h at 120 "C in air did not change the Pb :S ratio, nor did it affect the amount of lighter impurities. According to NRA, the amount of hydrogen decreased by 2-5 atom% after the annealing. Auger analysis of CBD-grown PbS thin films revealed that the films also contain oxygen.24 However, in contrast to this work, oxygen was detected on the surface, down to 20nm thickness, as well as on the interface between the film and the Si(100) substrate. Conclusions High quality PbSthin films were synthesized at room tempera- ture and normal pressure by using the SILAR technique.The film surfaces were homogeneous, smooth and metallic. A complexed lead ion [bis(triethanolamine)lead(II), Pb(tea),'+ 3 was used as the lead precursor. The complexation of the lead precursor enhanced the growth rate. Furthermore, the orien- tation of the films was strongly affected by the precursor. The PbS thin films grown on several different substrates were found to be cubic and showed remarkable [200] orientation. The films also exhibited a rather large crystallite size. On the other hand, the morphology and the grain size of the film were not affected by the complexation of the lead precursor when compared to the films deposited from the uncomplexed lead acetate precursor.The stoichiometry of the films was PbS, but according to RBS the films contained some light-atom impurities. Dr. E. Rauhala and Mr. P. Haussalo (the Accelerator Laboratory of the University of Helsinki) are acknowledged for the RBS and NRA measurements. Facilities provided by the Department of Electron Microscopy at the University of Helsinki were exploited for SEM characterization. The work was supported financially by the Academy of Finland and Technology Development Center, TEKES, Helsinki, Finland. References 1 M. Isshiki, T. Endo, K. Masumoto and Y. Usui, J. Electrochem. SOC.,1990, 137,2697. 2 Y. S. Sarma, N. K. Misra and H. N. Acharya, Znd. J. Phys. A, 1989, 63,445. 3 P. K. Basu, T. K. Chaudhuri, K. C. Nandi, R. S. Saraswat and H.N. Acharya, J. Muter. Sci., 1990,25,4014. 4 P. K. Nair, V. M. Garcia, A. B. Hernandez and M. T. S. Nair, J. Phys. D: Appl. Phys., 1991,24, 1466. 5 P. K. Nair, M. T. S. Nair, A. Fernanadez and M. Combo, J. Phys. D: Appl. Phys., 1989,22,829. 6 K. Ito and S. Tamaki, Tech. Char. Synth. Inorg. Muter. Lett., 1991, 10, 1395. 7 T. Kanniainen, S. Lindroos and M. Leskela, in Advances in Inorganic Films and Coatings, vol. 5, Advances in Science and Technology. Proceedings of the Topical Symposium 1 of the Forum on New Materials of the 8th CIMTEC World Ceramic Congress and Forum on New Materials, ed. P. Vincenzini, Techna Srl, Faenza, 1995, p. 291. 8 V. P. Tolstoi, Russ.Chem. Rev. (Engl. Transl.), 1993,62,237. 9 M. Takahashi, Y. Ohshima, K.Nagata and S. Furuta, J. Electroanal. Chem., 1993,359,291. 10 M. Leskela, L. Niinisto, P. Niemela, E. Nykanen, P. Soininen, M. Tiita and J. Vahakangas, Vacuum, 1990,41,1457. 11 E. Nykanen, J. Laine-Ylijoki, P. Soininen, L. Niinisto, M. Leskela and L. G. Hubert-Pfalzgraf, J. Muter. Chem., 1994,4, 1409. 12 H. Zogg, C. Maissen, J. Masek, T. Hoshino, S. Blunier and A. N. Tiwari, Semicond. Sci. Technol., 1991,6, C36. 13 Y. F. Nicolau, Appl. Surf. Sci., 1985,22/23, 1061. 14 S. Lindroos, T. Kanniainen and M. Leskela, Appl. Surf. Sci., 1994, 75,70. 15 Y. F. Nicolau and J. C. Menard, J. Cryst. Growth, 1988,92, 128. 16 Y. F. Nicolau and J. C. Menard, J. Appl. Electrochem., 1990, 20, 1063. 17 Y. F. Nicolau, US Pat., 4675207, 1987. 18 V. V. Klechkovskaya, V. M. Maslov, M. B. Muradov and S. A. Semiletov, Sov. Phys. Crystallogr. (Engl. Transl.), 1989, 34, 105. 19 K. Singh and M. Singh, Ind. J. Chem. A, 1982,21, 595. 20 E. Rauhala, J. Keinonen, K. Rakennus and M. Pessa, Appl. Phys. Lett., 1987,51,973. 21 H. J. Whitlow, J. Keinonen, M. Hautala and A. Hautojarvi, Nucl. Instr. Methods Phys. Res; Sect. B, 1984,5, 505. 22 Z. X. Sun, W. Forsling, L. Ronngren, S. Sjoberg and P. W. Schindler, Colloids Surf., 1991,59, 243. 23 D. Fornasiero, F. Li, J. Ralston and R. St. C. Smart, J. Colloid Interface Sci., 1994, 164,333. 24 H. Elabd and A. J. Steckl, J. Appl. Phys., 1980,51, 726. 25 D. Fornasiero, F. Li and J. Ralston, J. Colloid Znterface Sci., 1994, 164, 345. 26 Joint Committee on Powder Diffraction Standards, Card 5-592 27 V. Kokozay and A. Sienkiewicz,J. Coord. Chem., 1993,30,245. 28 H. Rahmanai, H. J. Gray and J. N. Zemel, Thin Solid Films, 1980, 69, 347. 29 A. N. Buckley and R. Woods, Appl. Surf. Sci., 1984,17,401. Paper 5/04104J;Received 26th June, 1995 164 J. Muter. Chem., 1996, 6(2), 161-164