DALTON J. Chem. Soc., Dalton Trans., 1997, Pages 257–261 257 Carbonyl dibromide: a novel reagent for the synthesis of metal bromides and bromide oxides † Michael J. Parkington,a T. Anthony Ryan b and Kenneth R. Seddon *,c a School of Chemistry and Molecular Sciences, University of Sussex, Falmer, Brighton BN1 9QJ, UK b ICI Chemicals and Polymers Ltd., The Heath, Runcorn, Cheshire WA7 4QD, UK c School of Chemistry, The Queen’s University of Belfast, Stranmillis Road, Belfast BT9 5AG, UK Carbonyl dibromide reacted with a wide selection of d- and f-block transition-metal oxides to form either the metal bromide or bromide oxide; the reactions are driven by the elimination of carbon dioxide.In a typical reaction the metal oxide was treated with an eight-fold excess of COBr2 in a sealed Carius tube at 125 8C for 10 d (to ensure complete reaction of the metal oxide). As COBr2 and the reaction by-products (CO2, CO and Br2) are all volatile, the desired products were obtained in essentially quantitative yield and a high degree of purity.Under these conditions V2O5, MoO2, Re2O7, Sm2O3 and UO3 were converted into VOBr2, MoO2Br2, ReOBr4, SmBr3 and UOBr3, respectively. This route offers great potential for the preparation of many known bromide derivatives of the transition metals, lanthanides and actinides, in a very convenient manner, and also for the synthesis of new materials. A modified synthesis of carbonyl dibromide was elaborated, and its 17O NMR and electron impact mass spectra are reported for the first time.Although the routes to pure anhydrous metal chlorides are well established, versatile, and generally convenient,1–3 the analogous routes to metal bromides and bromide oxides are poorly explored.1–3 When appropriate, they can best be prepared by reaction of the element with either dibromine, e.g. equation (1),4 2V + 3Br2 400 8C 2VBr3 (1) or hydrogen bromide, equation (2),5 by bromination of the Cr + 2HBr 750 8C CrBr2 + H2 (2) metal oxide with Br2,6,7 BBr3,8 AlBr3,9 CBr4,10,11 or SOBr2,12 or by halide exchange with HBr13 or BBr3.14 In addition, less general routes include the reduction of high-oxidation-state bromides with the appropriate metal (aluminium or dihydrogen are alternative reductants in some cases),1,2 e.g.equation (3),15 or by 3HfBr4 + Hf 500 8C 4HfBr3 (3) thermal disproportionation, equation (4),15,16 or thermal 2ZrBr3 350 8C ZrBr4 + ZrBr2 (4) decomposition, equation (5),17 of a higher-oxidation-state 2OsBr4 350 8C 2OsBr3 + Br2 (5) binary bromide.Metal-vapour synthesis has also been used to synthesize metal bromides:18 thus co-condensation of rhenium atoms with 1,2-dibromoethane (followed by extraction with tetrahydrofuran, thf) gave [Re3Br9(thf)3]. As is apparent, there is no satisfactory general route to metal bromides and bromide oxides. The two main problems appear to be: (a) many of the synthetic routes require severe experi- * E-Mail: k.seddon@qub.ac.uk WWW: http://www.ch.qub.ac.uk/krs/ krs.html † Non-SI unit employed: mB ª 9.27 × 10224 J T21.mental conditions, and (b) alternative syntheses, performed under milder conditions, frequently lead to product contamination, the contaminant often being extremely difficult to remove (see below). Metal chlorides have long been prepared by treating metal oxides with phosgene, COCl2, equation (6).19 These syntheses MxOy + yCOCl2 æÆ xMCl2y/x + yCO2 (6) are not only clean, high yielding, and performed under mild conditions, but also provide the basis of many patents (e.g.for dealuminating zeolites).20,21 It was somewhat surprising, therefore, that the analogous routes to metal bromides using carbonyl dibromide had not been investigated. The only report in the literature of a reaction between a metal oxide and COBr2 is by Prigent,22 who proposed that heating UO3 with COBr2 in a sealed tube for 2 h at 126 8C produced uranium(V) bromide.In our hands, and those of others,23 however, these observations were unrepeatable. Indeed, as uranium(V) bromide decomposes above 80 8C24 it would have been a very surprising result. We report here on the reaction between a wide range of metal oxides with carbonyl dibromide, which offers great potential for the preparation of many known bromide derivatives of the transition metals, lanthanides and actinides, in a very convenient manner, and for the synthesis of new materials.Preliminary observations on this system have been reported previously in a communication 25 and patent applications.26,27 Experimental CAUTION: The physiological effects of carbonyl dibromide were judged (as a result of some rather amateur experiments on white mice) similar to those of phosgene,28 but clearly a modern detailed evaluation is required if COBr2 is to be used more widely. The following safety precautions were adopted on the assumption that its toxicity is similar to that of phosgene.Handling carbonyl dibromide Phosgene is a toxic gas, with a permissable UK Occupational Exposure Limit (OEL) of 0.08 mg m23 of air (0.02 ppm v/v).29 In the event of exposure, the victim may experience chest pain,258 J. Chem. Soc., Dalton Trans., 1997, Pages 257–261 coughing and rapid breathing associated with pulmonary œdema, and it may take over 24 h for symptoms to appear. There is no antidote to phosgene poisoning,19 and hence treatment is usually directed to the main symptom, toxic pulmonary œdema.30 Hence, all manipulations involving carbonyl dibromide were carried out in a well ventilated fume cupboard with a face velocity of >0.75 m s21, and in the presence of at least one other experienced research worker.The vacuum line was constructed within the fume cupboard. The atmosphere both inside and outside the fume cupboard was constantly checked using Dräger tubes 31 and detector tape (Rimon Laboratories Ltd.).All glassware used greaseless taps, and joints were lubricated with Teflon sleeves. After use, carbonyl dibromide was destroyed by passage through a column containing moist activated charcoal. The fume cupboard was fitted with an alarm system, which was activated automatically if the extractor mechanism failed, or manually in the event of an accident. After use all equipment was washed with an aqueous solution of sodium hydroxide before removal from the fume cupboard.Spectroscopic measurements Carbon-13 and 17O NMR spectra were recorded on a Bruker WM360 spectrometer operating at 90.55 and 48.82 MHz, respectively. The 13C and 17O chemical shifts were measured with respect to external tetramethylsilane and water, respectively. Mass spectra were recorded on a Kratos MS80RF spectrometer, and infrared spectra on a Perkin-Elmer 598 spectrometer. Gas-phase infrared spectra were recorded using a 10 cm gas cell fitted with CsI windows, those of solids were recorded as Nujol mulls, using CsI plates.All spectra were calibrated using polystyrene (1601 and 907 cm21) and indene (551.7 and 420.5 cm21). Magnetic susceptibilities were measured at room temperature on a Johnson Matthey magnetic susceptibility balance. Preparation of carbonyl dibromide Concentrated sulfuric acid (20 cm3) was slowly added to molten tetrabromomethane (20 g, 60 mmol) at ca. 90 8C. The reaction vessel, which was connected to a conventional distillation unit, fitted with a high-surface-area trap, was then heated to 150–170 8C for 2 h.The products were collected, as the reaction proceeded, in a 210 8C trap. The deep red impure distillate was then transferred quickly to a vacuum line, held at 295 8C and continuously evacuated for 1 h to remove the small amounts of SO2 present. To remove the considerable quantities of free dibromine, the product was condensed into an ampoule (fitted with a greaseless tap) containing mercury, and allowed to warm to room temperature.The ampoule was then closed, removed from the vacuum line, and vigorously (but carefully) agitated within the fume cupboard for 5 min. It was then reconnected to the vacuum line, and the liquid was distilled into a storage bulb. The colourless liquid was redistilled into an ampoule fitted with a greaseless tap, and then stored at room temperature in the absence of light. The purity of the product was checked by gasphase infrared,13C and 17O NMR and mass spectrometry.Yield (based on CBr4): 5.8 g (51%). Preparations of metal bromides and metal bromide oxides The procedure for performing the reaction of UO3 with COBr2, and the subsequent isolation of the product, UOBr3, is described in detail. Exactly the same experimental procedures were followed for the other reactions. All reactions were performed at 125 8C for 10 d and in all cases free dibromine was observed during them. Uranium(V) tribromide oxide.Carbonyl dibromide (0.9 g, 4.84 mmol) was condensed into a Carius tube containing uranium( VI) oxide (0.18 g, 0.63 mmol), which was then sealed in vacuo and heated at 125 8C for 10 d. After this time the Carius tube was cooled to 295 8C, and the top (which had been carefully scored with a glass knife) fitted with Portex tubing (which was attached to a ground-glass joint). The Carius tube was then connected to a high-vacuum line, opened carefully and, after removal of the excess of COBr2 and gaseous reaction products, isolated, removed from the high-vacuum line, and taken into an inert-atmosphere dry-box where the contents were transferred into a Schlenk tube.The black powder was subsequently identified as uranium(V) tribromide oxide by bromide analysis (Found: Br, 50.05. Calc. for Br3OU: Br, 48.9%), magnetic measurements [cg = 4.07 × 1028 m3 kg21, meff (296 K) = 2.04 mB], and infrared spectroscopy [960m (br), 812m, 607w, 473m (br), 339s (br) and 281m cm21].Yield (based on UO3): 0.28 g (90%). Samarium(III) bromide. Reaction of carbonyl dibromide (0.95 g, 5.05 mmol) and samarium(III) oxide (0.23 g, 0.66 mmol) at 125 8C for 10 d gave a pale yellow powder which was shown to be samarium(III) bromide by bromide analysis (Found: Br, 61.1. Calc. for Br3Sm: Br, 61.45%), magnetic measurements [cg = 3.29 × 1028 m3 kg21, meff (294 K) = 1.64 mB] and infrared spectroscopy. Yield (based on Sm2O3): 0.47 g (92%). Rhenium(VI) tetrabromide oxide.Reaction of carbonyl dibromide (0.91, 4.84 mmol) and rhenium(VII) oxide (0.31 g, 0.65 mmol) at 125 8C for 10 d gave a deep blue-black solid which was shown to be rhenium(VI) tetrabromide oxide by bromide analysis (Found: Br, 60.3. Calc. for Br4ORe: Br, 61.25%), magnetic measurements [cg = 2.58 × 1028 m3 kg21, meff (296 K) = 1.71 mB], infrared spectroscopy (1003s and 239s cm21), and mass spectrometry {m/z 522 ([ReOBr4]+, 46), 314 ([ReO3Br]+, 64), 283 ([ReOBr]+, 37), 235 ([ReO3]+, 40), 187 (Re+, 58), 160 (Br2 +, 100) and 81 (Br+, 64%)}.Yield (based on Re2O7): 0.59 g (88%). Molybdenum(VI) dibromide dioxide. Reaction of carbonyl dibromide (1.03 g, 5.48 mmol) and molybdenum(IV) oxide (0.09 g, 0.70 mmol) at 125 8C for 10 d gave purple-brown crystals which were shown to be molybdenum(VI) dibromide dioxide by bromide analysis (Found: Br, 56.4. Calc. for Br2MoO2: Br, 55.5%), magnetic measurements (cg = 23.90 × 1028 m3 kg21), infrared spectroscopy [846s (br), 759s (br), 391w, 366w, 340m, 325m, 298m and 261w cm21], and mass spectrometry {m/z: 209 ([MoO2Br]+, 82), 193 ([MoOBr]+, 46), 177 ([MoBr]+, 30), 160 (Br2 +, 15), 130 ([MoO2]+, 12), 114 ([MoO]+, 22), 98 (Mo+, 36) and 79 (Br+, 100%)}. Yield (based on MoO2): 0.18 g (87%).Vanadium(IV) dibromide oxide. Reaction of carbonyl dibromide (0.98 g, 5.21 mmol) and vanadium(V) oxide (0.13 g, 0.71 mmol) at 125 8C for 10 d gave olive-brown leaflets which were shown to be vanadium(IV) dibromide oxide by bromide analysis (Found: Br, 69.7.Calc. for Br2OV: Br, 70.5%), magnetic measurements [cg = 5.34 × 1028 m3 kg21, meff (293 K) = 1.57 mB], and infrared spectroscopy [881s (br), 361m, 290s and 238s cm21]. Yield (based on V2O5): 0.30 g (92%). Results and Discussion Carbonyl dibromide The early attempts 32–37 to prepare COBr2, and the claims and counterclaims of success and failure, are summarized elsewhere. 19 By 1906, von Bartal 34 had demonstrated that COBr2 could be prepared in 50–60% yield by the oxidation of CBr4 with concentrated sulfuric acid at 150–170 8C, equations (7) and (8), although oleum is too vigorous a reagent, oxidizing the CBr4 + H2SO4 150–170 8C COBr2 + 2HBr + SO3 (7) 2HBr + SO3 æÆ SO2 + H2O + Br2 (8)J.Chem. Soc., Dalton Trans., 1997, Pages 257–261 259 CBr4 through to CO2 and Br2. By the nature of all the known synthetic routes, COBr2 is always produced contaminated with elemental bromine, and von Bartal 34 proposed a two-step puri- fication technique. Crude COBr2 is initially shaken with mercury at 0 8C, and then distilled, collecting the 62–65 8C fraction.This distillate is then treated with powdered antimony, and redistilled to yield colourless COBr2. If the first stage of the reaction with mercury is omitted the reaction with antimony is too vigorous, and some COBr2 is lost through decomposition. Slight modifications of this procedure were later published by Schumacher and Lenher,28 and this has become the most commonly used procedure.38 The procedures used here are derived from von Bartal’s preparation,34 followed by Schumacher and Lenher’s purification, 28 but they differ in some significant details (especially in the procedure for the removal of Br2).The antimony step has been eliminated, as the heat generated was observed to cause decomposition of the carbonyl dibromide. The infrared spectrum of gaseous COBr2 did not differ significantly from that reported elsewhere,38 and showed no detectable traces of CO2, CO, COCl2 or COBrCl. The 13C and 17O NMR spectra (in CD2Cl2 at 250 8C) of COBr2 gave chemical shifts at d 106.9 and 549.2, respectively [cf.d(C) 103.4 in CCl3F],39,40 and its mass spectrum (Table 1) is discussed in the preceding paper.41 These data highlight the purity of the product produced. The pure COBr2 was stored in the dark, since it was found that, in the presence of light, the colourless liquid became straw-coloured within 1 d due to decomposition to carbon monoxide and dibromine, equation (9).Over a prolonged COBr2 CO + Br2 (9) period this would result in a hazardous build-up of pressure in the storage vessel. Reactions of carbonyl dibromide with metal oxides The yields of the metal-containing products from the reactions of UO3, Sm2O3, Re2O7, MoO2, or V2O5 with COBr2 at 125 8C were all greater than 87%, and it can be assumed that, neglecting manipulative losses, conversion of the oxide was essentially quantitative.Attempted reactions with WO3, PbO2, Al2O3 and CaO led to incomplete reaction, products being heavily contaminated with unreacted metal oxide; as convenient syntheses of the desired products already existed, the use of alternative reaction conditions was not explored, although the reaction with WO3 had clearly produced significant amounts of WO2Br2. Although free Br2 was observed in all the reactions, its presence can give no information concerning the stoichiometry of the reactions, since pure COBr2, if heated to 125 8C, undergoes some dissociation to CO and Br2, equation (9).28 The presence of Br2 raises the possibility of the formation of [Br3]2; however, the satisfactory bromide analyses together with the appropriate magnetic moments mean that [Br3]2 contamination of the product can be safely discounted.Uranium(V) tribromide oxide. The reaction of UO3 and COBr2 at 125 8C gave UOBr3, as a black powder, presumably Table 1 Mass spectral data for COBr2 m/z Relative intensity Assignment 190, 188, 186 162, 160, 158 109, 107 93, 91 81, 79 28 — 37 100 12 86 4 M+ [Br2]+ [COBr] + [CBr] + Br + [CO] + according to equation (10).Unfortunately, the colour of UOBr3 2UO3 + 4COBr2 125 8C 2UOBr3 + 4CO2 + Br2 (10) is not reported in the literature, nor are there any reports of its magnetic moment or infrared spectrum. The effective magnetic moment of 2.04 mB (at 296 K) reported here is similar to values obtained for other uranium(V) compounds, e.g.UO2Cl [meff (295 K) = 1.86 mB]42 and UCl5 [meff (300 K) = 2.00 mB].43 The infrared spectrum of UOCl3 has been reported twice (1000–450 cm21 only),44,45 with the bands at 965, 845, 615 and 450 cm21 analogous to the bands at 960, 812, 607 and 473 cm21 for UOBr3. Attempts to record the electron impact (EI) mass spectrum of UOBr3 were unsuccessful due to its involatility, and the positive-ion fast-atom bombardment (FAB) technique failed to give a spectrum due to reaction of the UOBr3 with the matrix.Interestingly, the proposal by Russian workers 46 that UOBr3 slowly evolved Br2 at room temperature was not vindicated. The only reproducible synthesis of UOBr3 in the literature is by Prigent,10 who heated UO3 in a stream of N2 and CBr4 vapour at 110 8C. It has been reported, also by Prigent,22,47 that reaction of UO3 and COBr2 in a sealed tube at 126 8C (i.e. the same conditions as used here) gave UBr5, although attempts to repeat this by other workers have been unsuccessful.23 Furthermore, work by Blair and Ihle 24 has shown that UBr5 readily decomposes at >80 8C, and hence Prigent’s claim 22,47 to have prepared UBr5 must be regarded as incorrect.It was hoped that performing the reaction of UO3 and COBr2 at a lower temperature, viz. 70 8C, might give a different product (perhaps even UBr5); unfortunately, under these milder conditions, no reaction occurred.Samarium(III) bromide. The reaction of Sm2O3 and COBr2 at 125 8C gave SmBr3, as a pale yellow powder (the same colour as reported in the literature),3 presumably according to equation (11). The effective magnetic moment of 1.64 mB (at 294 K) was in Sm2O3 + 3COBr2 125 8C 2SmBr3 + 3CO2 (11) reasonable agreement with the 1.51 mB (at 293 K) obtained by Selwood.48 The infrared spectrum showed no bands in the range 1000–200 cm21, indicating the absence of Sm2O3 and SmOBr.The existing syntheses of anhydrous SmBr3 involve either dehydration of SmBr3?6H2O in the presence of HBr at high temperature (>640 8C),49 or reaction of Sm2O3 and NH4Br, again at high temperature.50–52 The synthesis reported here required less severe conditions, and more importantly did not produce unwanted SmBr2 and SmOBr, the latter being a frequent contaminant when synthesizing SmBr3 from SmBr3? 6H2O.3,53 Rhenium(VI) tetrabromide oxide. The reaction of Re2O7 and COBr2 at 125 8C gave ReOBr4, as a deep blue-black solid (the same colour as reported in the literature 54,55), presumably according to equation (12).The effective magnetic moment of Re2O7 + 5COBr2 125 8C 2ReOBr4 + 5CO2 + Br2 (12) 1.71 mB (at 296 K) and infrared spectral bands at 1003s and 239s cm21 were in reasonable agreement with those reported by Edwards and Ward [meff = 1.80 ± 0.1 mB (at 294 K), infrared bands at 1005s, 364m and 242s cm21],55 although they report a band at 364 cm21 in their infrared spectrum which was not observed here.The previously unrecorded mass spectrum of ReOBr4 shows a strong molecular ion. The most recent synthesis of ReOBr4 was by the reaction of rhenium metal, Br2 and SO2 in a sealed tube at 400 8C:55 the preparation reported here was performed under far milder conditions.260 J. Chem. Soc., Dalton Trans., 1997, Pages 257–261 Molybdenum(VI) dibromide dioxide. The reaction of MoO2 and COBr2 at 125 8C gave MoO2Br2, as purple-brown crystals (the same colour as reported in the literature),2 presumably according to equation (13).The diamagnetism of the product is MoO2 + COBr2 125 8C MoO2Br2 + CO (13) consistent with a d0 molybdenum(VI) compound. The infrared and mass spectra were in good agreement with those reported by Barraclough and Stals,56 the only significant difference being the absence of the molecular ion in the mass spectrum reported here. This compound is usually prepared by passing a mixture of O2 and Br2, diluted with N2, over the metal at 300 8C.57 The method reported here was performed under milder conditions, and may be considered a more accessible synthesis.Vanadium(IV) dibromide oxide. The reaction of V2O5 and COBr2 at 125 8C gave VOBr2, as olive-brown leaflets (the same colour as reported in the literature),6,58 presumably according to equation (14). The magnetic moment of 1.57 mB (at 293 K) was V2O5 + 3COBr2 125 8C 2VOBr2 + 3CO2 + Br2 (14) reasonable for a d1 halide oxide with an extended lattice.The infrared spectrum was in very good agreement with that reported by Dehnicke 6 [bands at 871s (br), 360m and 293m cm21], although he did not report the spectrum below 250 cm21 and thus did not observe the band at 238 cm21. There are several syntheses of VOBr2 reported in the literature, 1,58 the two most widely used being bromination of V2O3 at 600 8C in a flow system 6 and thermal decomposition of VOBr3 at 180 8C.59 The synthesis employed here has an obvious advantage over the bromination reaction, and is also preferable to the alternative method, since synthesis of VOBr3 is itself not trivial.1,58 Thermodynamic comparison of brominating agents The thermodynamics of the reactions of EBr3 (E = B or Al), CBr4 and EOBr2 (E = S or C) with metal oxides [equations (15)–(17)], derived from the JANAF Thermochemical Tables 60 3M2On + 2(n22)EBr3 æÆ 6MOBrn22 + (n22)E2O3 (15) M2On + (n 2 2)CBr4 æÆ 2MOBrn22 + (n22)COBr2 (16) M2On + (n22)EOBr2 æÆ 2MOBrn22 + (n22)EO2 (17) and the NBS Tables,61 are compared in Table 2.As the metal oxide, M2On, and metal-containing product, MOBrn22, are assumed to be the same in each case, only the differences in free energy of formation, DGdiff (and enthalpy of formation, DHdiff) of the brominating agent and the product derived from the brominating agent are listed, expressed per mol of MOBrn22 formed. Dibromine was not included in this table since no thermodynamic data were available for Br2O (the ‘expected’ byproduct of the reaction of Br2 with metal oxides).However, as Br2O is unstable above 240 8C62 it is unlikely to provide a significant thermodynamic driving force, and this is reflected in the observation that conversion of metal oxides into metal bromide oxides using Br2 often requires the use of very high temperatures and/or the presence of reducing agents.1–3 Thermodynamically, SOBr2 (which decomposes above 80 8C)12 and CBr4 are the poorest brominating agents listed in Table 2 and, not surprisingly, are rarely used in this way (cf.CCl4, which is a significantly better halogenating agent, and is commonly used in the synthesis of metal chlorides and chloride oxides 1–3). The remaining brominating agents listed in Table 2, BBr3, AlBr3 and COBr2, are all thermodynamically excellent, with COBr2 being the best. The driving force for the first two reactions is the large enthalpy of formation of the extended solids B2O3 and Al2O3, respectively, whilst for COBr2 both the enthalpy of formation of CO2 and the concomitant favourable increase in entropy provides a significant part of the driving force.However, although BBr3 and AlBr3 are thermodynamically excellent brominating agents, the generation of E2O3 (E = B or Al) as by-products often causes experimental difficulties, viz. separation of the E2O3 from the metal bromide or bromide oxide. Sublimation (providing, of course, the product is volatile) often leads to decomposition (e.g.FeBr3 and TaOBr3),63 while other separation techniques, such as dissolution in methanol (often used to remove B2O3),8,63 are often unsuitable since many metal bromides and bromide oxides react with donor solvents (e.g. UOBr3 10,64,65 and TiBr4 58) giving both solvation and solvolysis products. Alternative syntheses of metal bromides and bromide oxides usually involve the use of high temperatures and pressures, one of the few exceptions being the halogen-exchange reaction with BBr3.14 The advantage of this halogen-exchange method is that the reaction can be carried out under mild conditions, and, more importantly, since the by-products are volatile puri- fication is straightforward.The only problem with it is the possibility of mixed-halide formation (e.g. WOCl3Br and WCl3Br2 are well known,66 and are possible products of the reaction of BBr3 with WOCl4 and WCl5, respectively).In the light of this discussion, it is apparent that existing syntheses of metal bromides and bromide oxides are, on the whole, performed under very forcing conditions, and in many cases yield impure products. The use of COBr2 offers a new synthetic route under mild conditions. The synthesis of a pure 3d, 4d, 5d and 5f bromide oxide, together with a pure 4f bromide illustrates the widespread applicability of COBr2 as a brominating agent. There is a strong thermodynamic driving force (viz.formation of CO2) and, more importantly, puri- fication of the metal-containing product is trivial providing reaction has gone to completion. Given the efficacy of COBr2 in synthesizing metal bromides and bromide oxides, its toxicity is not of major significance. Indeed, current synthetic routes frequently involve the use of toxic (but less emotive) compounds, and COBr2 appears to be no more toxic that O3, and is considerably less toxic than [Ni(CO)4].Acknowledgements We are grateful to the EPSRC and ICI plc for the award of a Table 2 Thermodynamic comparison of some brominating agents, at 600 Ka Brominating agent Product derived from the brominating agent DHdiff b/kJ mol 2 1 DGdiff b/kJ mol 2 1 BBr3 (g) Al2Br6 (g) CBr4 (g) SOBr2 (g) COBr2 (g) B2O3 (s) a-Al2O3 (s) COBr2 (g) SO2 (g) CO2 (g) 2 128c 2 108c 2 53d 2 91e 2 140e 2 112c 2 101c 2 79d 2 75f 2 154e a Enthalpies and free energies of formation of the brominating agents and the products derived from the brominating agent were obtained from the JANAF Thermochemical Tables,60 except for COBr2 19 and SOBr2 61.b DHdiff is the difference in enthalpy of formation (DHf) of the product derived from the brominating agent and the brominating agent itself, expressed per mol of MOBrn 2 2 formed. Thus, for COBr2, DHdiff = ��� [DHf (CO2) 2 DHf (COBr2)]. DGdiff is the analogous freeenergy difference. c Calculated for a general reaction (15). d Calculated for a general reaction (16).e Calculated for a general reaction (17). f The free energy of formation of SOBr2 was estimated assuming that DHf (SOBr2) is independent of temperature.J. Chem. Soc., Dalton Trans., 1997, Pages 257–261 261 CASE studentship (to M. J. P.), to the EPSRC and Royal Academy of Engineering for the award of a Clean Technology Fellowship (to K. R. S.), and to Drs. A. K. Abdul-Sada and A. G. Avent for spectroscopic assistance. References 1 R.Colton and J. H. Canterford, Halides of the First Row Transition Metals, Wiley, London, 1969. 2 J. H. Canterford and R. Colton, Halides of the Second and Third Row Transition Metals, Wiley, London, 1968. 3 D. Brown, Halides of the Lanthanides and Actinides, Wiley, London, 1968. 4 R. E. McCarley and J. W. Roddy, Inorg. Chem., 1964, 3, 60. 5 R. J. Sime and N. W. Gregory, J. Am. Chem. Soc., 1960, 82, 800. 6 K. Dehnicke, Chem. Ber., 1965, 98, 290. 7 S. A. Shchukarev and G. A.Kokovin, Russ. J. Inorg. Chem., 1964, 9, 849. 8 P. M. Druce and M. F. Lappert, J. Chem. Soc. A, 1971, 3595. 9 M. Chaigneau, Compt. Rend., 1956, 243, 957. 10 J. Prigent, Ann. Chim. (Paris), 1960, 5, 65. 11 S. A. Shchukarev, E. K. Smirnova, I. V. Vasil’kova and N. I. Borovkova, Russ. J. Inorg. Chem., 1962, 7, 625. 12 R. C. Paul, M. Singh and S. K. Vasisht, J. Indian Chem. Soc., 1970, 47, 3. 13 A. Anagnostopoulos, D. Nicholls and M. E. Pettifer, J. Chem. Soc., Dalton Trans., 1974, 569. 14 P. M. Druce, M. F. Lappert and P. N. K. Riley, Chem. Commun., 1967, 486. 15 E. M. Larson and J. J. Leddy, J. Am. Chem. Soc., 1956, 78, 5983. 16 R. C. Young, J. Am. Chem. Soc., 1931, 53, 2148. 17 I. N. Semenov and N. I. Kolbin, Russ. J. Inorg. Chem., 1962, 7, 111. 18 P. R. Brown, F. G. N. Cloke, M. L. H. Green and R. C. Tovey, J. Chem. Soc., Chem. Commun., 1982, 519. 19 T. A. Ryan, C. Ryan, E. A. Seddon and K. R. Seddon, Phosgene and Related Carbonyl Halides, Elsevier, Amsterdam, 1996. 20 B. Fahlke, G. Blumenthal, W. Wieker, G. Wegner, K. Kintscher, W. Roscher and P. Knop, Ger. (East) Pat., DD 222 271, 1985. 21 D. Parris, Can. Pat., CA 1 209 976, 1986. 22 J. Prigent, C. R. Acad. Sci., 1954, 239, 424. 23 F. Lux, G. Wirth and K. W. Bagnall, Chem. Ber., 1970, 103, 2807. 24 A. Blair and H. Ihle, J. Inorg. Nucl. Chem., 1973, 35, 3795. 25 M. J. Parkington, K. R. Seddon and T. A. Ryan, J. Chem. Soc., Chem. Commun., 1989, 1823. 26 M. J. Parkington, T. A. Ryan and K.R. Seddon, UK Pat., GB 8 728 883, 1988. 27 M. J. Parkington, T. A. Ryan and K. R. Seddon, Eur. Pat., 320 161, 1988. 28 H.-J. Schumacher and S. Lenher, Ber. Dtsch. Chem. Ges., B, 1928, 61, 1671. 29 Occupational Exposure Limits 1995, Guidance Note EH40, Health and Safety Executive, 1995. 30 W. F. Diller, Toxicol. Ind. Health, 1985, 1, 93. 31 K. Leichnitz, Dräger Detector Tube Handbook, Drägerwerk, Lubeck, 1979. 32 J. Schiel, Liebig’s Ann. Chem. Pharm., Suppl., 1863, 2, 311. 33 A.Emmerling, Ber. Dtsch. Chem. Ges., 1880, 13, 873. 34 A. von Bartal, Liebigs Ann. Chem., 1906, 345, 334. 35 A. Besson, Compt. Rend., 1895, 120, 190. 36 A. Besson, Compt. Rend., 1896, 122, 140. 37 A. Brochet, Bull. Soc. Chim., 1897, 13, 221. 38 J. Overend and J. C. Evans, Trans. Faraday Soc., 1959, 55, 1817. 39 W. Gombler, Z. Naturforsch., Teil B, 1981, 36, 1561. 40 W. Gombler, Spectrochim. Acta, Part A, 1981, 37, 57. 41 M. P. Parkington, T. A. Ryan and K. R. Seddon, preceding paper. 42 J. C. Levet, Compt. Rend., 1969, 268, 703. 43 P. Handler and C. A. Hutchinson, J. Chem. Phys., 1956, 25, 1210. 44 I. A. Glukhov, S. S. Eliseev and E. E. Vozhdaeva, Russ. J. Inorg. Chem., 1969, 13, 483. 45 S. Sostero, O. Traverso, C. Bartocci, P. DiBernado, L. Magon and V. Carassiti, Inorg. Chim. Acta, 1976, 19, 229. 46 S. A. Shchukarev, I. V. Vasil’kova and V. M. Drozdova, Russ. J. Inorg. Chem., 1958, 3, 75. 47 J. Prigent, Compt. Rend., 1954, 238, 102. 48 P. W. Selwood, J. Am. Chem. Soc., 1934, 56, 2392. 49 E. R. Harrison, J. Appl. Chem. (London), 1952, 2, 601. 50 N. H. Kiess, J. Res. Nat. Bur. Stand., Sect. A, 1963, 67, 343. 51 G. Jantsch, N. Skalla and H. Jawarek, Z. Anorg. Allg. Chem., 1931, 201, 219. 52 A. S. Dworkin, H. R. Bronstein and M. A. Bredig, J. Phys. Chem., 1963, 67, 2715. 53 F. H. Spedding and A. H. Daane, The Rare Earths, Wiley, New York, 1961. 54 R. Colton, J. Chem. Soc., 1962, 2078. 55 D. A. Edwards and R. T. Ward, Inorg. Nucl. Chem. Lett., 1973, 9, 145. 56 C. G. Barraclough and J. Stals, Aust. J. Chem., 1966, 19, 741. 57 R. Colton and I. B. Tomkins, Aust. J. Chem., 1965, 18, 447. 58 R. J. H. Clark, The Chemistry of Titanium and Vanadium, Elsevier, Amsterdam, 1968. 59 H. E. Roscoe, Proc. R. Soc. London, 1870, 18, 316. 60 M. W. Chase jun., C. A. Davies, J. R. Downey, jun., D. J. Frurip, R. A. McDonald and A. N. Syverud, JANAF Thermochemical Tables, 3rd edn., J. Phys. Chem. Ref. Data, 1985, 14 (suppl. 1), 1. 61 D. D. Wagman, W. H. Evans, V. B. Parker, R. H. Schumm, I. Halow, S. M. Bailey, K. L. Churney and R. L. Nuttall, The NBS Tables of Chemical Thermodynamic Properties, J. Phys. Chem. Ref. Data, 1982, 11 (suppl. 2), 1. 62 C. Campbell, J. P. M. Jones and J. J. Turner, Chem. Commun., 1968, 888. 63 P. M. Druce, D. Phil Thesis, University of Sussex, 1968. 64 G. Kaufmann and R. Rohmer, Bull. Soc. Chim. Fr., 1961, 1969. 65 G. Kaufmann, Rev. Chim. Miner., 1964, 1, 129. 66 P. M. Boorman, N. N. Greenwood and H. J. Whitsfield, J. Chem. Soc. A, 1968, 2256. Received 6th June 1996; Paper 6/