IGR Report on Grants GR M and GR M

IGR Report on Grants GR/M98623 and GR/M98630 Chemistry, University of Bath, Dr K. Sanderson Pilkington Glass and Dr S. Rushworth, Epichem. 1 Prof. I. P. Parkin and Dr C. J. Carmalt, Department of Chemistry, University College London; Prof. K. C. Molloy, Department of Chemical Vapour Deposition of Transition and Main Group Metal Phosphides 1.1 Background/ Context This project produced new metal phosphide thin films. It involved a close collaboration between UCL and Bath University on film growth, characterisation and precursor synthesis. The work also involved Pilkington Glass and EpiChem Ltd, who provided precursors, substrates, equipment and measurement facilities. Pilkington are looking closely at the commercialisation of one of the phosphide films found in the project for solar control applications. The project enabled the growth of a range of metal phosphide films – a number of which have never been grown before by any technique. The functional properties of the metal phosphides were assessed – this revealed that many were hard, metallic conductors that showed exceptional resistance to a wide range of solvents and to oxidation. Some of the films showed excellent heat–mirror properties for use in solar-control, and for use in VLSI devices for the microelectronics industry. In addition, some fundamental yet unexpected chemistry of main group metal phosphide "adducts" has been elucidated. 1.2 Aim The primary aim of the project was to prepare new metal phosphide coatings by CVD. This involved direct deposition from available dual-source precursors in combination with more speculative single-source routes. 1.3 Introduction Prior to this project virtually no work had been conducted into making metal phosphide thin-film coatings. This lack of development was due perhaps to the perceived paucity of suitable phosphorus containing precursors. In metal nitride synthesis NH3 is often used as a nitrogen source, the corresponding phosphorus analogue, PH3, is extremely toxic (tlc 0.2 ppm), pyrophoric and difficult to handle. The syntheses of metal phosphide coatings prior to this work were limited, element specific, and usually involve the use of phosphine (PH3) or a PCl3/H2 mix at extremely high temperatures.1 The synthesis of TiP coatings has achieved some attention from decomposition of molecular precursors.2 It has also been prepared by reduction of TiCl4/PCl3 vapour at 1050°C 3 and from reaction of white phosphorus vapour with titanium plates at 900°C.4 Because of the lack of synthetic routes, the functional properties of metal phosphides, particularly in thin-film form, are largely unknown, and thus the technological potential of this class of materials remains under exploited. Our research program has helped to fill this vacuum. 1.4 Overview of Project All of the initial objectives in the original proposal were thoroughly investigated. It was found that the single-source strategy of making homoleptic transition and main group metal phosphides for use in CVD required modification. Attempts at making homoleptic transition metal phosphides led to elimination reactions of the diphosphine and reduction at the metal centre, this was novel chemistry and is explained below. To circumvent this elimination reaction an additional goal of making phosphine adducts of metal halides was developed.5 This led to new chemistry and also the ready formation of metal phosphide films.6 The APCVD reactions worked exceptionally well and were extended beyond that of the original project. 7-18 The single-source precursors for germanium and tin phosphides were targeted towards adducts of primary and secondary phosphines, in the light of initial successes at UCL in analogous transition metal chemistry. This led to a reappraisal of compounds described in the older literature as "adducts", but where the fundamental chemistry we have found is far more complex. Some success at depositing GeP and SnP from single-source precursors was achieved in the later stages of the project. 2.1 Preparation of Metal Phosphide Films by APCVD The APCVD reactions to form metal phosphide films were examined in detail. In the initial proposal we were to focus on making Ti, V and Sn phosphides from reactions of the metal halides and various phosphines (PRH2, PR3). All of these reactions were successful and we extended the programme to the synthesis of Ge, Zr, Hf, Nb, Ta, Mo and W phosphides (see Eq. (1)).7-18 Detailed investigations were undertaken of growth rates, processes gas velocity, precursor concentrations and bubbler and substrate temperatures. This meant that we were able to optimise the CVD growth conditions and explore in detail the functional properties of the films. In the course of the extension studies we discovered a new method of making phosphide films from the use of metal amides. This was unexpected and beyond the scope of the initial proposal. The reaction of metal halides with PCyH2 was investigated for a wide range of metals. It was found that in all systems there was film growth onset between 450-500°C. Depositions were primarily on glass, however silicon, tin dioxide, silica, alumina and metal substrates were also coated. There was no substrate selective deposition - the films grew equally well on all substrates with the same onset temperature. The phase of transition metal phosphide obtained in the reactions was surprisingly insensitive to precursor flow rates or substrate temperature. Even using very high temperatures (650 °C) or by using vast excess of phosphine (10000 fold excess) the same phosphides- h-TiP, β-NbP, β-TaP, VP, MoP2 or WP 2 were formed.7-10 The main group phosphides - Sn and Ge proved to be different and a number of discrete phases were made - Sn3P4, SnP, Sn3P2 and Sn5P2; GeP2, GeP, Ge2 P3 and Ge2 P.16,18 The phase of main group phosphide formed was directly related to substrate temperature and precursor flow rates - higher phosphine concentrations and lower substrate temperatures led to a more P rich phase. The transition metal phosphides tended to form the most thermodynamically favoured phosphide phase at the substrate temperatures investigated whilst the main group phosphides showed a compositional variation dependent on reaction conditions. MCln + RPH2 à MP + HCl + RH n = 4, M = Ti, V, Sn; n = 5, M = Ta, Nb (1) 2 The effect of using different phosphines (PButH2, PCyH2, PPhH2, PCy2H, PCy3, PPh2Ph, PPh3, P(SiMe3) 3) as the phosphorus source in the APCVD was investigated in detail for TiCl4 (corresponding XRD and Raman patterns are shown in Figure 1a and 1b respectively), VCl4 and SnCl4.6, 7 In general it was found that all of the phosphines would form the same metal phosphide phase for titanium and vanadium; but that some compositional variation was noted with tin. Primary phosphines were found to be better than secondary phosphines than tertiary phosphines in terms of both growth rates (300 vs 25 nm min-1) and minimising carbon incorporation (primary phosphines produced metal phosphide films with undetectable carbon contamination, whilst the tertiary phosphines could incorporate upto 10 atom%.).17, 18 The one exception was P(SiMe3) 3 which produced essentially carbon free metal phosphide films (see Eq. (2)). This was put down to the facile elimination of Me3SiCl. 3 TiCl4 + 4P(SiMe3) 3 à 3 TiP + 13 Me3SiCl + P (2) In order to extend the range of metal phosphides films available from APCVD - other metal sources for Hf and Zr were investigated as the metal halides were too involatile for use as precursors.13 The homoleptic metal amide precursors - [M(NMe2) 4] (M = Ti, Zr, Hf) proved to be an excellent source material (see Eq. (3)). These homoleptic metal amides have been used previously as single-source precursors to metal nitride and carbonitride films. Surprisingly, even when the gas phase ratio of [M(NMe2) 4] to PCyH2 was close to unity - MP films were formed with undetectable nitrogen and very low carbon contamination. In these experiments we have discovered a new reaction and shown that there is a strong tendency for metal phosphide formation even in the presence of effective single-source precursors for nitrides/carbonitrides (this work was described as 'ground breaking’ by the Chem. Mater. referees).13 Attempts at making CrP from CrO2C l2 led to formation of chromia films. However both CrP and FeP films were made by the APCVD reaction of the metal carbonyl and PCyH2 (shown schematically in Eq. (4)). M(NMe)4 + 2PCyH2 à MP + 4HNMe2 + 1/2Cy2P-PCy2 (M = Ti, Zr, Hf) M(CO) n + 2PCyH2 à MP + nCO + 1/2PCyH-PCyH + H2 (M = Fe, Cr) (a) (b) (3) (4) Figure 1. (a) Powder XRD of h-TiP and (b) Raman pattern of h-TiP grown from APCVD of TiCl4 and CyPH2. The functional properties of the metal phosphide films formed by APCVD are summarised in section 3.1 - the transition metal phosphides all had an aesthetic gold, silver or blue appearance; they were hard films with metallic conductivity and exceptional resistance to oxidation and attack by concentrated acids/ alkalis and common solvents. Many of the metal phosphide films were crystalline - h-TiP, βNbP, β-TaP, V12P7, FeP and CrP however the main group phosphides and MoP2 and WP 2 were not. The transition metal phosphides grew via an island growth mechanism. Tin and germanium phsphide seemed to grow with spherical microstructure with evidence of poor wetting of the surface. Film growth rates varied from ca 500 nm min-1 for primary phosphines to 50 nm min-1 for tertiary phosphines.7-17 All of the transition metal phosphide films were very adhesive to the substrate. It was noted that all of the APCVD reactions of metal halides and phosphines proceeded via the formation of a gas phase adduct (as this adduct was observed in the exhaust of the reactor), for example for TiCl4 and PCyH2 reaction went via the formation of [TiCl4.(PCyH2) 2]. It was possible to grow solid solution phosphides such as Hfx TiyP and Hfx TiyP from the dual source APCVD reactions using amide precursors. These materials had similar properties – gold colour, hard, conductive and resistant to oxidation to the parent metal phosphides. 2.3 Synthesis of Phosphide Precursors Transition metal precursors A number of strategies were employed for the synthesis of single-source precursors for use in CVD of metal phosphides. A range of complexes of the type [TiCl4L2] (where L = PhPH2, Ph2PH, Ph3P, CyPH2, Cy2PH and Cy3P) or [TiCl4L'] (where L' = DPPM, DPPE and DPPP) were synthesised from the reaction of TiCl4 with 2 equivalents of L, or one equivalent of L', in hexane, Eq. (5) .5 The reaction occurred instantly depositing (generally) highly coloured solids in relatively high yield (> 60 %). The crystal structures of the adducts formed using Cy 2PH and DPPE are shown in Figure 2. TiCl4 + 2PRn H3-n à TiCl4(PRn H3-n) 2 (5) n = 1-3, R = cyclohexyl or phenyl These adducts were effective precursors to thin films of TiP, with the level of film contamination increasing with carbon content of the precursor5, 17. Further adducts were formed from the reaction of NbCl5 with the same phosphine ligands - the cyclohexylphosphine adducts were suitable for use in CVD and formed phosphide films - NbP. 3 The adduct methodology was extended to the other group IVb members, Zr and Hf, but no obvious reaction took place in hexane, CH2Cl2 or toluene. Identical reaction of VCl4 with various phosphines reliably deposited blue-purple solids from hexane solution in high yield. These solids were unsuitable for CVD, displaying no transport properties (AP and LPCVD) and being insoluble (AACVD). This indicates that reaction of VCl4 with organophosphines proceeds via a different route to the reaction of TiCl4, forming involatile salts similar to the phosphonium salt formed as a minor product in the NbCl5 reactions. Reaction of TaCl5 with 2 equivalents of phosphine deposited small amounts of a white crystalline solid. X-ray crystallography showed that the ionic complex [TaCl4(PCyH2) 4][TaCl6] has formed. The cation consists of an 8-coordinate tantalum centre with a square antiprismatic geometry comprised of four chloride and four phosphine (PCyH2) ligands (the structure of [TaCl4(PCyH2) 4]+ is shown in Figure 2c). This was not a viable precursor for CVD. (a) (b) (c) Figure 2. Molecular structure of (a) [TiCl4(PCy2H)2], (b) [TiCl4(DPPE)] and (c) the cation in [TaCl4(PCyH2) 4][TaCl6]. Much effort was spent trying to synthesise homoleptic phosphides of the type M(PR2) n. Following the work of Baker19 the reaction of metal halide with lithiated diphenyl and dicylohexylphosphine was examined. Reaction was tried with TiCl4, NbCl5 and MoCl5. In all cases an instant reaction took place depositing black/brown solids from solution. Analysis of the reaction products was complicated due to low solubility and involatility. It was found that in all cases R2P-PR2 was formed as a major product indicating that reduction of the metal had taken place in all cases (confirmed by magnetic measurements). Reaction of TiCl3 with LiPPh2 also deposited an intractable black solid. Efforts were made to try and protect the metal centre from reduction during reaction by using solvent adducts of the metal halides, such as [TiCl4(THF) 2] and [MoCl4(CH3CN)2], but the result of reaction was identical. Further attempts were made to prepare the neutral homoleptic phosphides by reacting the appropriate metal halide/phosphine adduct with a lithiated phosphide, i.e. [TiCl4(Ph2PH)2] + 4 LiPPh2. Again the reduced phosphine (Ph2P-PPh2) was detected as the major P- containing species. A number of other reactions were examined to see if they afforded suitable precursors for CVD. The reaction of TiCl4 with P(SiMe3) 3 was examined, yielding the reduced species [TiCl3(P(SiMe3) 3) 2]. This precursor decomposed without transport yielding a black pyrophoric powder (bulk TiP). The reaction of [Cp2TiCl2] with LiPPh2 yielded [Cp2Ti(PPh2) 2] which did transport but did not deposit a film. The precursor molecule [Cp2Ti(PMe3) 2] was synthesised from the reaction [Cp2TiCl2] + Mg + 2PMe3 but again this transformed to bulk amorphous bulk TiP without transport. A further strategy of making mixed phosphide-phosphine complexes such as TiCl2(P(SiMe3) 2)(P(SiMe3) 3) was successful at making bulk TiP but did not prove to be a good CVD precursor as the molecules too readily formed bulk TiP – in effective it was too thermally unstable to give good CVD transport. Hence the single source approach to metal phosphide thin films was most successful for the metal halide adducts. These afforded TiP and NbP films by LPCVD. The more sophisticated homoleptic phosphide precursors proved to be ineffective precursors for thin films because of a generic problem of reduction at the metal center and elimination of diphosphine - PR2-PR2. However both the adduct and homolpetic phosphide approaches did enable bulk TiP to be made – this was one of the subsidiary goals in the original proposal. Main group precursors In view of the initial successful synthesis and exploitation of transition metal adducts of various phosphines, the main group chemistry concentrated initially on the synthesis of adducts of MX4 (M = Ge, Sn; X = Cl, Br, I) with predominantly primary and secondary phosphines. The reaction chemistry proved to be extremely complex, and many species reported in the older literature were found to be erroneous. The availability of both M(II) and M(IV) oxidation states, particularly for tin, meant that redox reactions proved common. For example, reaction of SnI4 with Cy3P yielded [Cy3PI]+[SnI3]- rather than an adduct of the same empirical formula. In addition, when primary or secondary phosphines were used, products consistent with the elimination of HX (presumably from the adduct initially formed) were isolated. Thus, 1:1 reaction of GeCl4 with CyPH2 afforded [Cy(H)PGeCl3], along with [CyPH3]+[GeCl3]-, Cy(H)PPCy(Cl) and further coupling products. [Cy(H)PGeCl3] could be generated as the sole product only when an excess of GeCl4 (4:1) was used; this compound was used for the deposition of GeP films by both AA- and LPCVD, though the film showed post-deposition oxidation.1 4 In the case of tin, redox reactions proved to be commonplace and in many instances Mössbauer spectroscopy revealed mixtures of products involving both oxidation states. In the case of reactions of SnI4 with either CyPH2 or Ph2PH, novel ionic species such as [Sn3I12]6-, [SnI5]- and [SnI6]2- have been isolated for the first time along with known anions [SnI3]- and [SnI4]2-; all these species have been crystallographically characterised (examples shown below). These involve both reduction at tin (though not always) and the formation of HI. Unfortunately, the lack of any Sn-P bond in these species rendered then unsuitable for single source CVD purposes. We were also able to synthesise covalent species such as [R3GePPh2] (R = Et, Ph), [EtGe(Cl)PPh2] and [R3SnPPh2] (R = Ph, Bu), though these species proved sensitive to oxidation to, for example, [Ph3SnO2PPh2]. However, freshly prepared samples were successfully used to generate M-P films under conditions where mixtures of R3MX and phosphine do not i.e. the establishment of a M-P bond in a single-source precursor is requirement for successful film deposition. Some of the precursors e.g. [Et3GePPh2], [EtGe(Cl)PPh2] were sufficiently volatile for LPCVD deposition protocols. In all cases, the Ge-P films deposited showed oxygen incorporation, which decreased as the film is penetrated, indicative of post-deposition oxidation. Figure 3a Single crystal X-ray structure of [Ph2PH2]+6 [Sn3I12]6- Figure 3b Single crystal X-ray structure of [Cy3PI]+ [SnI5]- 3.1 Functional Analysis - Overview The functional properties of the films prepared by dual source APCVD and single source LPCVD were investigated in detail. This was given great emphasis because many of the phosphide materials had never been prepared in thin film form. Solar Control Properties: The titanium phosphide films showed some interesting solar control properties. They showed good reflectance in the infra-red and good transmittance in the visible region – ideal for window use. Further the films had an aesthetic silver mirror-like appearance. The commercial potential of these films is currently under evaluation by Pilkington Glass who have instigated trial scale up experiments on commercial float glass lines. The other transition metal phosphide films – NbP, TaP and VP also show marked solar control behaviour. Hardness The early transition metal phosphide films were all hard, passing the brass and steel scratch tests. Indentation experiments were conducted on selected samples (in conjunction with Dr P. Hatto formally of Multiarc UK) these showed that the metal phosphides were again hard and similar in hardness values to the analogous metal nitride. (Vickers 900 –1400) Reactivity: The transition metal phosphide coatings were insoluble in all common solvents (acetone, water, THF, toluene) and surprisingly resistant to mineral acids (concentrated nitric and hydrochloric acids) and alkalis (concentrated NaOH). Hot concentrated hydrochloric acid digested a TiP coating in 48 h. This chemical inertness we attribute to a tenacious oxide surface coating. It was noted that the oxide coating became self-limiting, due to reaction of the metal phosphide with air. However the oxidation did not progress further than ca 50 nm from the surface. Monitoring of the optical properties of the films showed that they were indefinitely stable in air (ca 2 years, see Figure 4a and 4b). Ti 1 .0 P0 .5 O0 .5 C0 .5 Cl2 Th = 140 nm Ti 1 .0 P1 .0 5 O0 .0 1C0 .0 1Cl0 .03 Th = 1700 nm QuickTime™ and a TIFF (LZW) see this picture. are needed todecompressor Figure 4a XPS depth profile of a TiP film showing the oxide layer. Each level is ca 4 nm. Figure 4b RBS profile for a TiP film – modelled on an oxide overlayer Contact Angle: The metal phosphide films showed contact angles for water droplets in the range of 50-60°. They did not show any photoinduced hydrophilicity. 5 Electrical Conductivity: The transition metal phosphide coatings were found to behave as metallic conductors with resistivities that approached that of the parent metal (100 - 1000 µΩ cm). The changes in conductivity with temperature were measured for a range of samples and all showed typical metallic behaviour with an increase of conductance with a drop in temperature. The SnP and GeP2 coatings were semiconductors and had bandgaps of 2.2 and 1.1 eV respectively. They showed n-type conductivity. MOSFET devices: The TiP coatings have been incorporated into a new generation of MOSFET devices. The excellent electrical conductivity of TiP – combined with its remarkable oxidation resistance and stability at high temperature (1000 °C) have made this an ideal candidate for use as conductive gate electrodes in MOS field effect transistors in VLSI circuits. TiP, HfP and ZrP thin films have been layed down on MOSFET devices – these devices have been patterned and are currently undergoing testing (in association with Prof W. in particular to determine work function and conductivity. Preliminary results are very encouraging. Photocatalysis : The metal phosphide coatings were assessed as part of an ongoing survey to see if they showed any photocatalaytic effects for the degradation of a test organic material stearic acid. None of the coatings showed any photocatalyic destruction of the test organic. Colour/Appearance: The metal phosphide films were exceptionally reflective in the visible and when suitably thick (500 nm +) were of commercial mirror like appearance. The titanium phosphide films on glass have been supplied to the Bartlett School of architecture (to assess usage in a building environment), the Slade art school and also to ceramic artists. They have been incorporated in some glass collages. The metal phosphide films are very aesthetically pleasing – silver, gold or dark blue. Gas Sensing: The gas sensing behaviour of the main group metal phosphides has been determined for the stochiometric SnP films. The films were grown on commercial sensor substrates. The sensors derived from the films show changes in base line resistance when exposed to trace hydrocarbons in air (ppm concentration, n-type behaviour). Analysis of the sensors show that the gas sensing response is largely a function of the oxide overlayer – SnO2 . Commercialisation of these films is being considered by Novosense – a UCL spin out company established by Prof. Parkin and Dr K. Pratt. Charaterisation: This project has provided the first measured XPS and Raman spectra of many transition metal phosphides. These spectra are in the process of being deposited in the appropriate reference library. 4. Summary This project synthesised of a range of new metal phosphide thin-film coatings on a variety of substrates. This has enabled functional testing for a variety of solar control, MOSFET and gas sensor applications. 5. Project Plan Review The project plan did involve some changes. The success of the dual source APCVD approach to metal phosphide films and the difficulties in precursor synthesis meant that more emphasis was placed in the programme on both film growth studies and on functional characterisation than initially planned. Further the range of films investigated was also expanded and some new CVD reactions developed. The synthesis programme was also changed in direction - the difficulties encountered in making homoleptic metal phosphides was shown to be due to a diphosphide elimination reaction. The synthesis section was augmented by the formation of novel metal halide - phosphine adducts. Many of these proved to be good single source precursors for the formation of metal phosphide coatings. In the light of the latter finding, the main group chemistry was re-oriented towards the synthesis of germanium/tin halides adducts of primary and secondary phophines. This proved more complex than anticipated, as halogen-transfer and HX elimination products were found. Main group metal phosphides also proved difficult precursors to handle due to their tendency to oxidise to metal phosphinates, though both AA- and LPCVD of freshly prepared samples yielded Ge/Sn-P films. The project partners (including a mixture of industrial collaborators – Dr L. Smith and Dr S. Rushworth Epichem: Dr S. Hurst and K. Sanderson, Pilkington) met at regular intervals during the course of the project – every 3-4 months and held 12 joint meetings during the course of the project. Regular e-mail bulletins were sent around to keep the project participants up to date with the results of the study. Further samples were regularly exchanged from Bath to UCL – in particular for detailed characterisation and functional analysis testing. 6. Research Impact and Benefits to Society To date the research has yielded thirteen publications in high quality international journals (Chem. Mater., J. Mater Chem., Adv. Mater. Dalton) and six refereed conference papers that were published in book. A further six publications are planned. Work has been presented as a talk or poster presentation at eight international conferences including Euro CVD, MC-5, ACS and ICMCTF and a number of national and international Universities (15), including Bochum, Dortmund, UCLA, University of Southern California, Caltech, Liverpool, Manchester, Loughborough, Herriot Watt, Cambridge, Oxford and Imperial College. It has also been presented at the EPSRC CVD network meetings. The titanium phosphide films on glass show potential as solar control coatings. Pilkingtons have devoted considerable in house resources into the evaluation and on-line testing of these materials. These materials could form a commercial product - for example the TiP coatings have an aesthetic colour - silver or gold (ideal for window applications) coupled with good reflection properties in the IR. This together with the facts that they are extremely hard wearing and show excellent resistance to solvents makes them good candidates for wide spread solar-control usage. This could have a direct impact on the consumer who would gain better heat insulation from a window. The coated glass has also been supplied to various ceramic artists for incorporation into new pieces of art. The work has also been incorporated into the lecture demonstration that Prof. Parkin gives to school-children (6 presentations during the course of the project). The metal phosphide films have also been grown on MOSFET (supplied to Prof. W. Wosik) and solid-state gas sensor devices (supplied to Novosense plc). At present they are also under evaluation for use as inert barrier layers and as gas sensitive resistors. 6 7. Explanation of Expenditure The grant expenditure did not deviate from that stated in the original proposal within the virement allowed. Pilkington provided more analytical support for the project than initially envisaged and a directly funded CASE studentship. They also helped with direct cash purchases of equipment (contact angle measurement and UV accessories). Epichem provided a range of commercial phosphines for the project (t-BuPH2,) and also samples of metal amides [Ti(NMe2) 4]. 8. Further Research and Dissemination Activities Dr C. Blackman (UCL) is employed on a further Post-doctoral contract at UCL. Leo Apostolico (Bath Student) has been awarded his Ph.D (Jan 2004) and has taken up a post doctoral position with Prof. A. Cowley at Texas A + M. Links to the final grant report and papers can be found from Prof. Parkin web home page – http://www.che.ucl.ac.uk/people/ipparkin. 9. References 1. M. Fuji, H. Iwanaga and S. Motojima, J. Crystal Growth, 1996, 166, 99. 2. T. S. Lewkebandara, J. W. Proscia and C. H. Winter, Chem. Mater., 1995, 7, 1053. 3. S. Motojima, T. Wakamatsu and K. Sugiyama, J. Less Common Metals, 1981, 82, 379. 4. Y. Saski and M. Hirohashi, Denki Kagaku Oyobi Kogyo Butsuri Kagaku, 1984, 52, 203. 5. C.S. Blackman, C.J. Carmalt, I.P. Parkin, S.A. O’Neill, K.C. Molloy, L. Apostolico, 'Single-source CVD routes to titanium phosphide"J. Chem. Soc. Dalton Trans., 2002, 13, 2702. 6. C.S. Blackman, C.J. Carmalt, I.P. Parkin, S.A. O’Neill, K.C. Molloy, L. Apostolico, 'Dual source atmospheric pressure chemical vapour deposition of TiP films on glass using TiCl4 and PH2But', J. Mater. Chem., 2001, 11(10), 2408. 7. C.S. Blackman, C.J. Carmalt, I.P. Parkin, S.A. O’Neill, K.C. Molloy, L. Apostolico, 'CVD of Group Vb metal phosphide thin films', J. Mater. Chem., 2003, 13, 1930. 8. C.S. Blackman, C.J. Carmalt, I.P. Parkin, S.A. O’Neill, K.C. Molloy, L. Apostolico, 'Dual-source chemical vapour deposition of titanium (III) phosphide from titanium tetrachloride and tristrimethylsilylphosphine' Appl. Surface Sci., 2003, 211, 2. 9. C.S. Blackman, C.J. Carmalt, I.P. Parkin, S.A. O’Neill, K.C. Molloy, L. Apostolico, 'CVD of crystalline thin films of tantalum phosphide', Mater. Lett., 2003, 57, 2634. 10. C.S. Blackman, C.J. Carmalt, T.D. Manning, I.P. Parkin, S.A. O’Neill, K.C. Molloy, L. Apostolico, 'Dual source APCVD of amorphous molybdenum phosphide films on glass using molybdenum(V) chloride and cyclohexylphosphine', Chem. Vap. Deposition, 2003, 9(1), 10. 11. C.S. Blackman, C.J. Carmalt, I.P. Parkin, S.A. O’Neill, K.C. Molloy, L. Apostolico, 'Dual-source APCVD of chromium phosphide thin films using chromium hexacarbonyl as a precursor', Appl. Surface Sci., Submitted for publication. 12. R. Binions, C.S. Blackman, C.J. Carmalt, I.P. Parkin, S.A. O’Neill, K.C. Molloy, L. Apostolico, 'Tin phosphide coatings from the APCVD of SnX4 (X = Cl or Br) and PRx H3-x (R = Cyc(hex) or phenyl)', Polyhedron, 2002, 21(19), 1943. 13. C.S. Blackman, C.J. Carmalt, I.P. Parkin, S.A. O’Neill, K.C. Molloy, L. Apostolico, 'Dual-source APCVD of Group IVb metal phosphide thin films using tetrakisdimethylamido metal precursors', Chem. Mater., in press 14. L. Apostolico, M.F. Mahon, K.C. Molloy, R. Binions, C.S. Blackman, C.J. Carmalt, I.P. Parkin, 'The reaction of GeCl4 with primary and secondary phosphines', J. Chem. Soc. Dalton Trans., 2004, 470. 15. C.S. Blackman, C.J. Carmalt, I.P. Parkin, K.C. Molloy, L. Apostolico, 'Production of thin films of vanadium phosphide by dual-source CVD of tetrakisdimethylamidovanadium and cyclohexylphosphine' Chem. Vap. Dep., in press. 16. R. Binions, C. S. Blackman, C. J. Carmalt, S. A. O’Neill, I. P. Parkin, K. C. Molloy and L. Apostolico, 'APCVD of germanium phosphide thin films on glass', Polyhedron, 2002, 14, 3808-3816 17. C. S. Blackman, C. J. Carmalt, S. A. O'Niell, I. P. Parkin, K. C. Molloy and L. Apostolico;. Allendorf, M.D.,Mary,F., and Teyssandier,F., Editors. 'Titanium Phosphide Coatings From the Atmospheric Pressure CVD Reaction of TiCl4 with PRx H3-x (R = Cy hex ; or R = SiMe3 where x-3)'. In Chemical Vapour Deposition XVI and EuroCVD 14. 2003; Pennington, New Jersey: The Electrochemical Society, Inc.; vol. 2, 1387-1394. ISBN 1-56677-378-4. 18. R. Binions., C. J. Carmalt and I. P. Parkin; Allendorf,M.D., Maury,F., and Teyssandier,F., Editors. 'Tin Phosphide Coatings from the Atmospheric Pressure Chemical Vapour Depoaition of SnCl4 and PCychexxH3-x '. In Chemical Vapour Deposition XVI and EuroCVD 14. 2003; Pennington, New Jersey: The Electrochemical Society, Inc.; vol. 2, 1426-1433. ISBN 1-56677-378-4. 19. R. T. Baker, P. J. Krusic, T. H. Tulip, J. C. Calabrese and S. S. Wreford, J. Am. Chem. Soc., 1983, 105, 6763.