The growth of copper oxides on glass by flame assisted chemical vapour deposition

The growth of copper oxides on glass by flame assisted chemical vapour deposition

Thin Solid Films 517 (2008) 517–521 Contents lists available at ScienceDirect Thin Solid Films j o u r n a l h o m e p a g e : w w w. e l s e v i e ...

525KB Sizes 0 Downloads 0 Views

Thin Solid Films 517 (2008) 517–521

Contents lists available at ScienceDirect

Thin Solid Films j o u r n a l h o m e p a g e : w w w. e l s e v i e r. c o m / l o c a t e / t s f

The growth of copper oxides on glass by flame assisted chemical vapour deposition H.M. Yates a,⁎, L.A. Brook a, D.W. Sheel a, I.B. Ditta b, A. Steele b, H.A. Foster b a b

Institute for Materials Research, University of Salford, Manchester, M5 4WT, UK Centre for Parasitology and Disease, Biomedical Sciences Research Institute, University of Salford, Salford, Manchester, M5 4WT, UK



Article history: Received 30 January 2008 Received in revised form 3 June 2008 Accepted 20 June 2008 Available online 26 June 2008 Keywords: Copper oxide Chemical vapour deposition Biocidal Nanostructure

A B S T R A C T Flame assisted chemical vapour deposition is a low cost, relatively simple atmospheric pressure chemical vapour deposition (CVD) technique that is compatible with both small volume, batch, and high volume continuous coating processes. Use of this method with low hazard aqueous solutions of simple metal salts can yield metal oxide thin films, which represents a major advantage in terms of precursor cost and environmental impact compared to alternative CVD methods. In this paper we report the extension of this technique to the growth of copper oxides from aqueous solutions of cupric nitrate (Cu(NO3)2) and discuss the effects on the films of the various growth conditions. It is shown that copper oxide films are produced with nanostructure controlled properties. Furthermore, we report that these films have strong antibacterial activity. © 2008 Elsevier B.V. All rights reserved.

1. Introduction Atmospheric pressure chemical vapour deposition (CVD) processes have a number of advantages over other, reduced pressure CVD techniques such as low pressure chemical vapour deposition and plasma enhanced chemical vapour deposition including their suitability for high volume, continuous growth processes, the high growth rates achievable, the ease with which the CVD equipment may be retro-fitted to an existing, open production line and the conformal nature of the layers deposited [1,2]. Of particular interest is flame assisted chemical vapour deposition (FACVD) [3,4] in which a flame is used to provide the energy required to crack the precursor species into fragments which subsequently form the film upon the substrate. One advantage of this technique is that no closed reaction cell is required, so making it ideal for fitting on open production lines. However, as an open-air process, the FACVD process may be contaminated by atmospheric impurities, which can lead to contamination within the desired film [5]. In previous studies, we have demonstrated that FACVD may be used with precursors in the form of simple, readily available, low cost metal salts in aqueous solution [6] of WO3 and MoO3 films. The use of aqueous solutions in this manner is particularly noteworthy in terms of the likely environmental impact of the FACVD process and its general ease of use. Copper has a long history of use as an antimicrobial agent with its recorded use as early as 2600BC [7]. In the form of its metal or ions has been shown to have excellent antimicrobial activity against a number of microorganisms including bacteria, fungi, algae and viruses yet is relatively safe for humans [8]. Commercial uses of copper include its use in paints in the health care sector [9] and as antifouling surfaces on ⁎ Corresponding author. Tel.: +44 161 295 3115; fax: +44 161 295 5111. E-mail address: [email protected] (H.M. Yates). 0040-6090/$ – see front matter © 2008 Elsevier B.V. All rights reserved. doi:10.1016/j.tsf.2008.06.071

ships [10], in alloys for controlling microorganisms on surfaces [11], in dental materials [12] and impregnated fibres [13,14]. Copper can be used to control Legionella in hospital water systems [15] and to control Escherichia coli O157:H7, Listeria monocytogenes and Methicillinresistant Staphylococcus aureus [16] on copper alloy surfaces. Copper ions have also been shown to potentiate the activity of other antimicrobials including mitomycin C [17], chlorine [18] and hexetidine [19]. Copper appears to exert its killing effect by binding to DNA causing single strand breaks and base modification. This effect is enhanced by hydrogen peroxide and free radicals are believed to be involved in the reaction [20–22]. In the present publication we present data for the growth of copper oxide thin films on glass by FACVD by use of aqueous precursor solutions. This metal oxide was selected as target species for thin film deposition because of its widespread use in range of technological applications including, electronics, catalysis, ceramics, pigments and electromagnetic devices [23,24]. In addition the growth of copper oxides by conventional thermal CVD is far from trivial, often requiring costly metal organic compounds such as N,N′-Diisopropylacetamidinatocopper(I) [25] as a precursor. Thus it is appropriate to study alternative methods of achieving CVD growth of these materials. The conditions required in order to achieve growth and the properties of the deposited films are described. Additionally we show that these films are highly biocidal to both Gram negative E. coli and Gram positive Staphylococcus epidermidis. 2. Experimental details 2.1. Growth The FACVD reactor used is of in-house construction and is made up of 4 main sections, these are: the burner head, substrate stage,


H.M. Yates et al. / Thin Solid Films 517 (2008) 517–521

precursor delivery system, and the mass flow controllers. The burner head is a basic design which consists of a hollowed out brass block which contains a removal baffle plate for enhanced gas mixing and a series of closely-spaced jets from which the flame emerges. These jets ensure the flame is distributed uniformly across the substrate. The substrate stage is made up of a carbon block, which can be moved backwards and forwards under the flame. The number of passes relating to the thickness of the film deposited. The carbon block can be heated via three pencil heaters, which run lengthways through the block. Care must be taken to ensure that the carbon block is kept at a temperature of below 500 °C since oxidation of the carbon block can occur at higher temperatures. The distance between the burner and the substrate was fixed at 3 mm. A nebuliser (ultra-neb2000 Devilbiss 200HS-042) was used in order to deliver the precursor to the flame. This is a commercially available ultrasonic system, which uses water as the transmission medium. The precursor solution sits within a cup that in turn is in contact with the water in the nebuliser, such that ultrasonic waves are passed through to the precursor solution producing droplets of precursor solution. Nitrogen carrier gas (1 L min− 1) then passes into and out of the nebuliser, collecting solution droplets as it does so and transporting these to the burner head. A series of mass flow controllers was used to accurately meter the flow of gases to the burner head. Fuelling the burner is a mixture of propane and oxygen gases flowing at 0.99 L min− 1 and 3.65 L min− 1 respectively generating an output power of 1.50 kW these gases are also mixed with 13.90 L min− 1 of nitrogen to ensure no explosive ratios occurs. Cupric Nitrate (0.5 M) (Fisons Scientific Equipment) has a purity of 99.5% and was dissolved in de-ionised water. The feed rate was 0.75 mL min− 1, corresponding to 3.75 × 10− 4 mol min− 1. Experimental conditions such as the substrate temperature were varied between 100 °C and 400 °C, while the number of passes (i.e. thickness) ranged from 16 to 100 so allowing study of both thin and thick films.

3. Results and discussion 3.1. Visual The samples all appeared yellow/brown in reflected light, very uniform and transparent. The films adhered well to the surface, and remained intact even after vigorous rubbing. There was no visible change in the coloration over the length or width of the substrate (9 cm × 22 cm). 3.2. Surface morphology As can clearly be seen from the SEM images, in Fig. 1, the morphology is highly dependent on the substrate growth temperature. The copper oxide films grown at low substrate temperatures, (Fig. 1a) show island growth type morphology, of a tightly packed spherical arrangement. This is controlled by the Volmer–Weber growth mode [30]. The diameter of these spheres varies between approximately 120–173 nm. As the substrate temperatures are increased (Fig. 1b) these spheres coalesce and form a continuous film. However, increasing these temperatures further to 400 °C (Fig. 1c), the film returns to an island type arrangement. Now the particles are larger (229–323 nm), but the separation is greater. This

2.2. Characterisation X-ray diffraction (Siemens D5000) with a Cu Kα source was used for structural characterisation of the films. The scans were run over the 2θ range 15–60 ° using 5 s/step at 0.02 increments. Data was compared to that from the PDF powder diffraction database (International Centre for Diffraction Data, formally the Joint Committee on Power Diffraction Standards, JCPDS) [26]. Surface morphology was obtained by SEM (Philips XL30), using an accelerating voltage of 10 kV and atomic force microscopy in tapping mode (Nanoscope llla, Digital Instruments Ltd). The silicon tips used had a resonant frequency round 190 kHz. Surface composition of the films produced was analysed by X-ray photoelectron spectroscopy (XPS). The XPS system used consists of an Alpha 110 hemispherical analyser and a monochromatic AlKα radiation (hv = 1486.27 eV) X-ray source (Vacuum Instruments). Prior to entry into the XPS system the samples were ultrasonically cleaned in spectrophotometric grade methanol (Sigma-Aldrich). The analysis conditions were for 10 (summed) scans, with a dwell time of 20 ms and step size of 500 meV and 50 meV respectively for the survey and high resolution elemental scans. Curve fitting with CASA XP software was used to deconvolute spectra. A Gaussian–Lorentzian function (30% Lorentzian) was used to define curve shape with a Shirley background [27]. Restraints for area and relative position to each other were placed on the related interdependent 2p Cu fitted curves (2p1/2 and 2p3/2). Biocidal activity was determined by monitoring the survival rate of E. coli ATCC 10536 and S. epidermidis NCTC 11047. The test used was a modification of the standard test described by BS EN 13, 697:2001, which has been described previously [28,29]. Prepared cultures of E. coli ATCC 10536 and S. epidermidis (2 × 108 108 colony forming units (cfu) mL− 1) were used. Each experiment was performed in triplicate.

Fig. 1. SEM images of thin copper oxide films produced by FACVD. (a) 30 passes deposited at 200 °C, (b) 30 passes deposited at 300 °C, (c) 30 passes deposited at 400 °C.

H.M. Yates et al. / Thin Solid Films 517 (2008) 517–521


change in morphology may also relate to the nature of the main species, which as shown later, changes from mostly Cu2O at low growth temperatures to mostly CuO at higher temperatures. For a closer look at the morphology atomic force microscopy was used. As shown in Fig. 2 the images (2 µm × 2 µm) confirm the particulate nature of the copper oxide deposition. Firstly, taking samples deposited at 30 passes the effect of temperature was assessed. The sample deposited at 200 °C consisted of smaller nanoparticles at closer distances than those formed at 400 °C, as also shown in the SEM. Quantitative values of the roughness can be obtained from the Ra value. That of the 200 °C sample is much lower at 12 nm than that of the 400 °C sample at 41 nm. As would be expected, growth at the much higher temperature led to an increased growth rate, changing the thickness from 34 nm to 61 nm. In this we define thickness to be that of the average peak height of the individual island growths. The change in morphology is linked to the thicker film, along with the growth temperature and change of oxidation state of the copper oxide. Secondly, to study the change of thickness further, the growth temperature was fixed while the number of passes was varied. The particle size increased in samples prepared with 4 passes (Fig. 3a), through 16 passes (Fig. 3b) to 30 passes at 400 °C (Fig. 2b) or with 30 passes (Fig. 2a) to 100 passes at 200 °C (Fig. 3c). Also noticeable in the case of 100 passes the particles were larger and had coalesced. For the series at 400 °C the roughness (Ra) increased from 7 nm (4 passes), through 14 nm (16 passes) to 41 nm (30 passes). For the set at 200 °C the roughness increased considerably from 12 nm to 41 nm where it gave an almost continuous coating for the 100 pass sample. Due to the discontinuous nature of the films growth rate can only be approximately calculated. Using the values for the maximum heights of the individual island deposition, the equivalent static growth rate was

Fig. 3. AFM images of thin copper oxide films (a) 4 passes deposited at 400 °C, (b) 16 passes deposited at 400 °C, (c) 100 passes deposited at 200 °C.

calculated to range from 115 nm min− 1 to 202 nm min− 1 at deposition temperatures of 200 °C and 400 °C respectively. This can be obtained from the growth rate per pass along with the residence time of the flame (flame width/rate of translation = 2 cm/3.6 cm s− 1). 3.3. Film crystal structure

Fig. 2. AFM images of thin copper oxide films (a) deposition at 200 °C, (b) deposition at 400 °C.

Deposited films were also studied using X-ray diffraction (XRD). Fig. 4 shows the structures obtained at different growth temperatures and thicknesses. Analysis of the XRD (Fig. 4d) spectra for 30 passes at 200 °C showed a material of very low crystallinity with mixed copper oxides. Only very small broad peaks were seen, 35° and 38° which correspond to CuO (JCPDS 05-0661) and two others at 36° and 42°, which correspond to Cu2O (JCPDS 05-0667). On increasing the


H.M. Yates et al. / Thin Solid Films 517 (2008) 517–521

Fig. 4. XRD of copper oxide films deposited on glass (a) 100 passes deposited at 200 °C, (b) 30 passes deposited at 400 °C, (c) 30 passes deposited at 300 °C, (d) 30 passes deposited at 200 °C.

temperature to 300 °C (Fig. 4c) the material was still of low crystallinity with mixed CuO (35° and 38°) and Cu2O (36°). Now there was no peak at 42° relating to Cu2O. At a temperature of 400 °C (Fig. 4b) the only peaks which were seen are those corresponding to CuO reflections observed at: 32.5°, 33.2°, 35.6° (most intense), 38.8° and 48.6°. Finally if the number of passes and hence thickness was increased at a low temperature (Fig. 4a), the main reflections observed were those corresponding to Cu2O with reflections observed at: 29.7°, 36.5° (most intense) and 42.6°. Also present were small shoulders at 35° and 38° indicating the presence of a small amount of CuO. The increased intensity here could be due to the increased crystallinity of the sample, but is much more likely to relate to the increased layer thickness (34 nm to 106 nm). In summary the XRD suggests that as the growth temperature is increased the copper oxide films change from mainly polycrystalline Cu2O to polycrystalline CuO. 3.4. X-ray photoelectron spectroscopy The elemental composition of the deposited film at 400 °C (and 30 passes) was determined to mainly contain silicon, carbon, oxygen, and copper. The carbon 1s peak (285 eV) relates to a mixture of surface

carbon (used as reference for binding energy shift) and that incorporated into the sample. The silicon 2p peak can be seen at 101.5 eV, which is bonded to oxygen in the form of SiO2 and is from the barrier layer on the glass. This was confirmed (Fig. 5a) by the presence of an O 1s peak at 533.24 eV, which is assigned to O2− within SiO2. Apart from the SiO2 related peak the O 1s scan also showed a peak at 530.60 eV, which relates to bonded oxygen in the copper oxides. The final, smallest, resolved peak is related to absorbed water. Fig. 5b shows the high resolution Cu 2p scan shows the major 2p3/2 peak at 936.21 eV and the 2p1/2 at 956.17 eV. The splitting of these (Δ = 19.9 eV), along with their position establishes the presence of CuO [31]. The peaks shown at 944.82 eV and 964.76 eV are shake-up satellite peaks. The presence of these high intensity satellites is another indication of CuO being the main component, rather than Cu metal or Cu2O. The 2p peaks assigned to CuO are also broader than those associated with Cu2O, which is also to be expected [32]. Using curve fitting (with constraints for peak area) it can be established that there are also small peaks at 935.52 eV and 955.42 eV, which are assigned to the minor copper oxide Cu2O [33]. Calculations from this fit suggest about 10% is Cu2O. The positions of all these Cu 2p peaks are shifted by approximately 2 eV higher binding energies than that of a standard bulk material, despite corrections for charging (due mainly to deposition on glass substrates). This is considered due to the nanoparticulate nature of the films. Similar shifts to higher binding energy, as the nanoparticle size decreases, have been previously cited for silver [34]. A similar analysis was carried out for samples grown at lower temperatures and various thicknesses. In all cases the XPS confirmed that mainly CuO was present on the surface. 3.5. Bioactivity To give examples of the photo-induced biocidal nature of these CuO films a relatively thick sample (30 passes) and a thin sample (4 passes) of polycrystalline films deposited at 400 °C were chosen. It can be seen in Fig. 6 that the control samples (UV blocked CuO films identical to the respective test films) showed no significant change even after 60 min. The error bars show the standard deviation over 3 separate runs.

Fig. 5. XPS High resolution scans for (a) O 1s, (b) Cu 2p. The black line shows the experimental data, while the grey lines show the fitted, resolved curves.

H.M. Yates et al. / Thin Solid Films 517 (2008) 517–521


copper nitrate by FACVD. The exact chemical composition depended on the growth conditions used to produce the films. In addition the film morphology obtained was also surprising. While it might be expected that a continuous film structure would result, in fact the films obtained consisted of nanoparticles ranging in size from 120 nm to 323 nm depending on the growth conditions. This work has demonstrated a route to the production of noncontinuous CuO films that uses low cost, relatively low environmental impact CVD chemistry. We have also shown that the films can be remarkably bioactive—and this property, combined with the CVD method used, which could be relatively simply transferred to commercial production, offers possibilities for future applications. Acknowledgements The authors wish to acknowledge the EPSRC and Corus PLC for financial support for LB. We would like to thank M. Faulkner from the University of Manchester Materials Science Centre for the SEM images and C. Liptrot from the Biomedical Sciences Research Institute, University of Salford for the technical assistance with the biological tests. References

Fig. 6. The biocidal activity of (a) copper oxide film (30 passes, 400 °C) on Escherichia coli, (b) copper oxide film, 4 passes, 400 °C on Escherichia coli and Staphylococcus epidermidis.

Firstly, when using gram negative E. coli the films were found to be highly bioactive and achieved high kill rates of N5 log reduction over 12 min, as seen in Fig. 6a. This example is for a film deposited at 400 °C, with 30 passes, which consists of mainly CuO, as shown earlier. In comparison to other bioactive surfaces, such as films of TiO2 containing nanocrystalline silver, this activity is much greater. Photoactive films (e.g. crystalline titania) can achieve similar kill levels but over significantly longer periods (e.g. several hours under comparable test conditions.)[35]. Silver films have also been shown to be bioactive [36]. However, some bacteria are, or can develop, silver resistance [37]. Copper offers an attractive alternative option as although bacteria can become resistance to copper [38] (although less common than for silver) it would be unusual for the same bacteria to be resistant to both silver and copper. A much thinner sample of CuO, again with the island growth, noncontinuous film was also tested. As can be seen (Fig. 6b) the kill rate is much slower taking 80 min for a 3 log reduction for E. coli. The much slower growth rate is considered to relate to presence of much less CuO. Interestingly the rate of kill is similar for both E. coli and S. epidermidis. Generally, gram positive bacteria such as S. epidermidis are considered to be more resistant than gram negative bacteria such as E. coli [39,40] but in this case there was little difference. The results for the S. epidermidis were obtained from samples which had already been used 3 times for the E. coli tests and had been cleaned stringently between tests to allow accurate biocidal experimentation. That the samples are still bioactive after this treatment suggests that the samples have biocidal regeneration. 4. Conclusions The work presented here, showed that it was possible to deposit polycrystalline copper oxide films on glass from aqueous solutions of

[1] M.L. Hitchman, K.F. Jensen, Chemical Vapor Deposition, Academic Press, New-York NY, 1993. [2] J.J. Schmitt, in: M. Pickering, B.W. Sheldon, W.Y. Lee, R.N. Johnson (Eds.), MRS Symposium Proc., 555, 1999, p. 255. [3] A.T. Hunt, J.K. Cochran, W.B. Carter, U.S. Patent No. 5 652 021, 29 Jul.1997. [4] A.T. Hunt, W.B. Carter, J.K. Cochran, Appl. Phys. Lett. 63 (1993) 266. [5] T.A. Polley, W.B. Carter, Thin Solid Films 384 (2001) 177. [6] M.J. Davis, G. Benito, D.W. Sheel, M.E. Pemble, Chem. Vapor Depos. 10 (2004) 29. [7] H.H.A. Dollwet, J.R.J. Sorenson, Trace elem. med. 2 (1985) 80. [8] G. Borkow, J. Gabbay, Curr. Med. Chem. 12 (2005) 2163. [9] T.E. Cooney, Infect. Control Hosp. Epidemiol. 16 (1995) 444. [10] T.E. Cooney, R.J. Tang, Methods Enzymol. 310 (1999) 637. [11] H.T. Michels, S.A. Wilks, J.O. Noyce, C.W. Kneevil, Materals Science. And Technololgy Conference, September 2005, Pittsburg, PA, U.S.A, vol 1, 2005, p. 3. [12] A.M. Mulligan, M. Wilson, J.C. Knowles, Biomaterials 24 (2003) 1797. [13] J. Grzybowski, E.A. Trafny, Polim. Med. 29 (1999) 27. [14] G. Borkow, J. Gabbay, FASEB J. 18 (2004) 1728. [15] J.E. Stout, Y.S. Lin, A.M. Goetz, R.R. Muder, Infect. Control Hosp. Epidemiol. 19 (1998) 911. [16] J.O. Noyce, H. Michels, C.W. Keevil, Appl. Environ. Microbiol. 72 (2006) 4239. [17] K. Ueda, J. Morita, K. Yamishita, T. Komano, Chem.-Biol. Interact. 29 (1980) 145. [18] T.M. Straub, C.P. Gerba, X. Zhou, R. Price, M.T. Yahiya, Water Res. 29 (1995) 811. [19] J. Grytten, A.A. Scheie, E. Giertsen, Acta odontol. Scand. 46 (1988) 181. [20] R.B. Thurman, C.P. Gerba, CRC Crit. Rev. Environ. Control 18 (1988) 295. [21] O.L. Aruoma, B. Halliwell, E. Gajewski, M. Dizdaroglou, Biochem. J. 273 (1991) 601. [22] J.L. Sagripanti, L.B. Routson, A.C. Bonifacino, C.D. Lytle, Antimicrob. Agents Chemother. 41 (1997) 812. [23] T.L. Alford, L. Chen, K.S. Gadre, Thin Solid Films 429 (2003) 248. [24] T. Maruyama, Sol. Energy Mater. Sol. Cells 56 (1998) 85. [25] Z. Li, R.G. Gordon, D.B. Farmer, Y.B. Lin, J. Vlassak, Electrochem. Solid-State Lett. 8 (2005) G182. [26] Powder Diffraction File, Joint Committee on Powder Diffraction Standards, ASTM, Philadelphia, PA, 1967. [27] D.A. Shirley, Phys. Rev. 55 (1972) 4709. [28] L.A. Brook, P. Evans, H.A. Foster, M.E. Pemble, A. Steele, D.W. Sheel, H.M. Yates, J. Photochem. Photobiol., A 187 (2007) 53. [29] Anon BS N 697:2001 Chemical Disinfectants and Antiseptics – Quantitative Nonporous Surface Test for the Evaluation of Bactericidal and/or Fungicidal Activity of Chemical Disinfectants Used in Food, Industrial, Domestic and Institutional Areas – Test Method and Requirements Without Mechanical Action (phase 2/step 2), British Standards Institution, London, 2001. [30] C. Friesen, S.C. Seel, C.V. Thompson, J. Appl. Phys. 95 (2004) 1011. [31] J. Chastain, R.C. King, Handbook of X-ray photoelectron spectroscopy, Physical Electronic Inc, New York, 1995. [32] K.S. Kim, J. Electron Spectrosc. Relat. Phenom. 3 (1974) 217. [33] S. Poulston, P.M. Parlett, P. Stone, M. Bowker, Surf. Interface Anal. 24 (1996) 811. [34] K. Luo, T.P. St. Clair, X. Lai, D.W. Goodman, J. Phys. Chem. B 104 (2000) 3050. [35] L.A. Brook, P. Evans, H.A. Foster, M.E. Pemble, A. Steele, D.W. Sheel, H.M. Yates, J. Photochem. Photobiol. A 187 (2007) 53. [36] H. Kim, H. Jeong, K.S. Kim, S.H. Yoon, S.S. Lee, K.W. Seo, I. Shim, Thin Solid Films 2005 (2005) 72. [37] I. Chopra, J. Antimicrob. Chemother. 59 (2007) 587. [38] R.P. Ryan, D.J. Ryan, Y.C. Sun, F.M. Li, Y. Wang, D.N. Dowling, FEMS Microbiol. Lett. 268 (2007) 40. [39] A. Pal, S.O. Pekhonen, L.E. Yu, M.B. Ray, J. Photochem. Photobiol., A 186 (2007) 335. [40] H.L. Liang Liu, T.C.K. Yang, Process Biochem. 39 (2003) 475.