摘要:

L e t t e r Racemic nematogens having a chiral center in the lateral chain Ceç cile Canlet,a Patrick Judeinstein,a Jean-Pierre Bayle,*,a Freç deç rick Roussel§,b and Bing M. Fungb a L aboratoire de Chimie Structurale Organique, (CNRS URA 1384), Paris XI, Universiteç 410, 91405 Orsay, France Ba� t. b Department of Chemistry and Biochemistry, University of Oklahoma, Norman, Oklahoma, 73019-0370, USA A chiral center can be introduced in the lateral alkoxy chains of mesogens containing four rings in the main core.The compounds present a large nematic range. Comparison with compounds having a linear chain indicates the bifurcate character of the chain does not drastically aÜect the nematic range. 13C NMR spectra in the nematic phase give some indications about the folding of each branch of the chiral chain along the core.Neç matoge` nes raceç miques posseç dant une chaï� ne lateç rale chirale. Nous avons syntheç tiseç des composeç s neç matoge` nes posseç dant une chaï� ne lateç rale alkoxy chirale. Ces composeç s posse` dent un domaine neç matique peu diÜeç rent de celui observeç pour les composeç s ayant une chaï� ne lineç aire posseç dant le me� me nombre de carbones.La RMN C-13 donne des indications sur le repliement des deux fragments de la chaï� ne chirale le long du coeur meç soge` ne dans la phase neç matique. Laterally alkoxy-substituted mesogens deviate from the classical rod-like shape. However, they may exhibit a large nematic range if the core anisotropy is large enough.1h4 Although it has been shown that many —exible lateral substituents can lead to mesogenic properties, the substituent position is quite important for the phase behavior.If the lateral alkoxy chain belongs to one of the outer rings, an enantiotropic smectic phase is enforced by a meta substitution with respect to the link ; meanwhile, a monotropic nematic phase is generally observed when the substitution is in the ortho position with respect to the link.2 If the lateral alkoxy chain is introduced on the inner ring, an enantiotropic nematic phase is always obtained.We have shown that introducing a methyl group near the alkoxy chain in that position enforces the nematic range, because this methyl group –ts the empty space.5 This interesting phenomenon is also observed when two lateral alkoxy chains are displayed on the same inner ring of a main core containing four aromatic rings.In the solid phase and the nematic phase, the two chains are folded back in opposite directions along the core allowing the molecular anisotropy to be preserved.6,7 To assure such a geometry, the –rst two fragments within the chains adopt opposite gauche conformations (Fig. 1a). Therefore, the idea to replace these two alkoxy chains pointing in opposite directions by a single bifurcate one is of interest as we may expect a similar eÜect (Fig. 1b). By molecular modelling, a minimized conformation Fig. 1 Proposed mean conformation of the chains substituted on the inner ring of the mesogen: a two alkoxy chains, b a single chiral chain § On leave from Laboratoire de Dynamique et Structure des Mateç - riaux Moleç culaires, (CNRS URA 801), Universiteç du Littoral, MREID, 59140 Dunkerque, France. can be found with the two parts of the bifurcate chain folded back along the core if the –rst fragment is in a gauche conformation.Thus, this possible conformation will allow the anisotropic arrangement of the molecules inside the liquid crystalline phase. In addition, it is a very elegant way to introduce a chiral center in the vicinity of the core.The lateral chain will have fewer degrees of freedom due to the expected folding of the chain along the core. Thus, we can expect a stronger in—uence of the asymmetric center with comparison to the one introduced in a terminal chain. In this note, we present the synthesis and the mesomorphic properties of the –rst racemic mesogens having a lateral chiral chain.Series I compounds (referred to as Res*n) contain four aromatic rings and the lateral chain is introduced on one of the inner rings (Fig. 2). This series is compared with two other series diÜering by the lateral substitution or the structure of the inner ring. Series II (Pyr*n) has a pyridine ring instead of the aromatic ring.This pyridine ring was chosen to quantify the steric eÜect of the hydrogen atom on the lateral chiral chain alignment. Series III (Res n) has a linear lateral chain, instead of the chiral chain, that contains the same carbon number in order to analyze the perturbation introduced by the bifurcate chain. Three compounds containing 4, 7 or 10 carbons have been synthesized in each series.The chiral chain always has an ethyl fragment. The diÜerent series were prepared according to the scheme shown for series II (Fig. 3). The transition temperatures of the three series are given in Table 1. All compounds exhibit an enantiotropic nematic phase. Fig. 4 shows the comparison of the transition temperatures between the compounds with respect to the total number of carbons in the lateral chain.In each series, the clearing temperatures decrease with the number of carbons in the chain; for the same carbon number in the chain, the clearing temperatures have the following order: Res n [Res*n[Pyr*n. The melting temperatures follow roughly the same order, but are less regular. The chiral chain does not add too much perturbation to the molecular arrangement in comparison with the linear chain.This means that the two branches of the bifurcate chain certainly point towards opposite directions as shown in Fig. 1. The short branch –lls New J. Chem., 1998, Pages 211»213 211C2H5 N N O O C O O C O OC4H9 Series I N N O O Series II N N O O Series III T / °C n C2H5 NH2 N OH KOH PEG 1)NaNO2/HCl HO CnH2n+1 Br C2H5 H + N OH O CnH2n+1 C2H5 + 2)NaOH C2H5 N N OH O CnH2n+1 C2H5 N O2C OC4H9 HOOC + DCC/NPP Toluene C2H5 N N O O CnH2n+1 C2H5 N C OC4H9 C O O O An Bn Pyr*n (n = 1, 4, 7) * Fig. 2 Structure of three series of compounds. In each series three compounds have been synthesized having 4, 7 or 10 carbons in the lateral chain. In series I and II, the bifurcate chain always has an ethyl fragment only the empty space created by the long branch without adding too much conformational disordering. Unfortunately, the introduction of the pyridine ring does not enlarge the nematic range as proposed.This indicates that there is no steric interaction between the ortho hydrogen and the chain. Amazingly, the pyridine ring lowers the melting temperatures, resulting in a nematic compound near room temperature for Pyr*10.As proposed previously,5,6 the appearance of the liquid crystal phase in these laterally substituted compounds is induced by the parallel conformation taken by the lateral Table 1 Transition temperatures (in °C) in the three series Res n, Res*n and Pyr*n. These values are taken with increasing temperature (heating rate 10 °C min~1) Compound K]N N]I Res 4 148 217 Res 7 143.5 185 Res 10 116 171 Res*4 117 201.5 Res*7 95.5 152.5 Res*10 113.5 135.5 Pyr*4 147 176.5 Pyr*7 93 133.5 Pyr*10 67 98 chain with respect to the mesogenic core.In order to check the alignment of the bifurcate chain, we have obtained the evolution of the 13C chemical shifts with temperature for compound Res*10 (Fig. 5). Fig. 4 Nematic range of the three mesogenic series Res n, Res*n and Pyr*n.The transition temperatures were measured by DSC (Mettler FP 52) using a heating rate of 10 °C min~1 Fig. 3 Synthetic scheme for series II. Three compounds have been synthesized with 4, 7 or 10 carbons in the chiral chain 212 New J. Chem., 1998, Pages 211»213Fig. 5 The chemical shifts of the aliphatic signals of Res*10 plotted against temperature as obtained with decreasing temperature from the isotropic melt The jump at the isotropic»nematic transition in the 13C TNI chemical shifts gives some valuable information on the mean conformatiagment.6 Usually, for a terminal chain each fragment experiences a negative jump with decreas- CH2 ing temperature.In many cases, we have shown that the inverse behavior is observed for the –rst in the lateral CH2 linear chain due to the particular mean conformation of this fragment within the nematic phase; the second exhibits a CH2 tiny negative or no obvious jump and the subsequent CH2 groups show the usual negative jump.Fig. 5 presents two interesting features : the expected positive jump for C1, which is proof of the folding of the chain in the nematic phase, and the unexpected positive jump for C2@ and negative jump for C2.This indicates that the two branches in the chiral chain in Res*10 do not act in exactly the same way. This interesting behavior will be analyzed in a forthcoming paper. In conclusion, a chiral chain containing two alkyl branches can be introduced in a lateral position on the inner ring of a main core containing four aromatic rings without a deleterious eÜect on the liquid crystalline properties.Moreover, the liquid crystal range is similar for the compounds with linear or chiral chains. 13C NMR spectroscopy indicates that each branch within the chiral chain is folded along the core. Further investigations dealing with the chiral chain conformations in the solid and liquid crystal phase are under way and will be reported elsewhere. Experimental 2,6-Dihydroxypyridinium chloride is mixed with two equivalents of potassium hydroxide in polyethylene glycol (MW\200) and the solution is stirred at 100 °C for one hour.8 The bromide is then added to the cooled mixture and the temperature is set again to 100 °C for two hours.The mixture is extracted with three volumes of dichloromethane.The monoalkylated compound is obtained through chromatography using dichloromethane as eluent. The diazotization was performed in water with one equivalent of diazonium salt and two equivalents of pyridine, in order to obtain the monodiazotized compound as the major component. The monalkylated compound was then puri–ed by chromatography with as eluent.The esteri–cation step was performed in CHCl3 dichloromethane using the dicyclocarbodiimide (DCC) method.9 Alkoxybenzoic acid was added in slight excess, in order to consume the phenol totally. After –ltration of the amide and evaporation of the solvent, the crude mixture was chromatographed with as eluent and then rec- CH2Cl2 rystallized several times in chloroform»ethanol until constant transition temperatures were obtained.The structures and the purity of the intermediates and –nal products were checked using 1H NMR on an AM 250 Bruker spectrometer, with as solvent. The data CDCl3»[2H5]pyridine are expressed in the form of chemical shifts in ppm (integration, multiplicity). 2-Hydroxy-6-[(1-methyl)propoxy]pyridine (A4) : 0.94 (3H, t), 1.31 (3H, d), 1.7 (2H, m), 4.40 (1H, h), 5.64 (1H, dd), 6.29 (1H, dd), 7.48 (1H, dd). 2-Hydroxy-3-[(4@-ethyl)-azobenzene]-6-[(1-methyl)propoxy]- pyridine (B4) : 1.03 (3H, t), 1.27 (3H, t), 1.45 (3H, d), 1.83 (2H, m), 2.71 (2H, q), 5.33 (1H, h), 6.32 (1H, d), 7.30 (2H, d), 7.70 (2H, d), 8.21 (1H, d). Pyr*4: 1.03 (6H, 2t), 1.29 (3H, t), 1.46 (3H, d), 1.54 (2H, m), 1.76 (2H, m), 2.73 (2H, q), 4.06 (2H, t), 5.28 (1H, h), 6.78 (1H, d), 7.00 (2H, d), 7.33 (2H, d), 7.39 (2H, d), 7.86 (2H, d), 8.06 (1H, d), 8.16 (2H, d), 8.33 (2H, d).The phase transitions were observed and characterized by using an Olympus polarizing microscope –tted with a FP 82 Mettler heating stage and a FP 85 Mettler DSC. The NMR experiments on Res*10 in its isotropic and nematic phases were conducted using a Varian VXR-500 NMR spectrometer T) equipped with an indirect detection probe (B0\11.07 manufactured by Narolac Cryogenic Corporation.The sample was put in a standard 5 mm tube and spun slowly along the magnetic –eld so that the director aligned parallel to the magnetic –eld. To avoid radio frequency overheating, a 0.8% decoupler duty cycle was used. The temperature calibration was made by observing the nematic to isotropic transition.Acknowledgements work of BMF was supported by the U.S. National The Science Foundation under grant number DMR-97000680. References 1 W. Weiss—og and D. Demus, Crystal Res. T ech., 1984, 19, 55. 2 F. Perez, P. Berdagueç , J. P. Bayle, T. Braé uniger, M. A. Khan and B. M. Fung, New J. Chem., 1997, 21, 1283. 3 F. Perez, P. Judeinstein, J.-P. Bayle, F. Roussel and B. M. Fung, L iq. Crystals, 1997, 22, 711. 4 C. T. Imrie and L. Taylor, L iq. Crystals, 1989, 6, 1. 5 P. Berdagueç , P. Judeinstein, F. Perez and J.-P. Bayle, New J. Chem., 1995, 19, 293. 6 F. Perez, J.-P. Bayle and B. M. Fung, New J. Chem., 1996, 20, 537. 7 F. Perez, P. Berdagueç , P. Judeinstein, J.-P. Bayle, H. Alloushi, M. Cotrait and E. Lafontaine, unpublished work. 8 P. Bergagueç , F. Perez, J. Courtieu and J.-P. Bayle, Bull. Soc. Chem. Fr., 1993, 130, 475. 9 A. Hassner and V. Alexanian, T etrahedron L ett., 1978, 46, 4475. Received 17th November 1997; Paper 8/00406D New J. Chem., 1998, Pages 211»213 213

ISSN:1144-0546    

DOI:10.1039/a800406d

出版商:RSC    

年代:1998

数据来源: RSC

摘要:

L e t t e r N N N N N N CH3 N N N N N N H3C (CH2) n [(phen)2RuII(Mebipy)-(CH2) n-(bipyMe)RuII(phen)2]4+ ( n = 5 2a, n = 7 2b or n = 10 2c) Binding of bimetallic 1,10-phenanthroline ruthenium(II) complexes to DNA Fiona M. OœReilly and John M. Kelly* Chemistry Department, T rinity College, University of Dublin, Dublin 2, Ireland The binding of (Mebipy\4-methyl-2,2@-bipyridyl-4@-, [(phen)2RuII(Mebipy)-(CH2)n-(bipyMe)RuII(phen)2]4` n\5 2a, 7 2b or 10 2c) to double-stranded DNA is found to be sensitive to the linker chain length n and to show a DNA-induced stacking interaction at high binding concentrations (e.g.[Nu]/[Ru]\1). There is considerable interest in the DNA binding of metal polypyridyl complexes, because of potential applications as DNA probes1,2 (including the study of electron transfer processes involving DNA3) and as possible antitumour agents.4 A wide range of mononuclear complexes have been studied, with (phen\1,10-phenanthroline), in particular, [Ru(phen)3]2` being much investigated.The relatively weak binding (KB104 dm3 mol~1 basepair~1) of this complex (for which the precise mode has been a matter of controversy1,5) and its displacement from DNA at even medium ionic strengths represent a serious limiting factor for the application of such systems, e.g.in in vivo conditions (150 mM Na`). We6 and others7,8 have suggested that this problem might be overcome by the use of binuclear systems. To demonstrate this eÜect we initially chose to work with complexes of the type where [(bipy)2Ru(Mebipy)-(CH2)n-(bipyMe)Ru(bipy)2]4`, Mebipy\4-methyl-2,2@-bipyridyl-4@-, bipy\2,2@-bipyridyl (n\5 1a or 7 1b), as the parent complex is an [Ru(bipy)3]2` extremely weak binder.6 The bimetallic complexes proved to have much higher affinity for DNA.One might expect that the phenanthroline analogues 2 would bind even more strongly and we present here some DNA binding characteristics of these complexes.We also report that the linker chain length (n) is a crucial factor in determining the binding efficiency. For classic intercalators with polymethylene chains9 it is known that the most eÜective binding occurs with n[8, whereas for the bimetallic complexes 2 we show that the strongest binding occurs with a somewhat shorter chain length, consistent with the partial intercalation model proposed for [Ru(phen)3]2`. 5 It is also found that at low [Nu]/[Ru] ratios (i.e.the ratio of concentrations of DNA nucleotides to ruthenium centres) complexes 2a»2c show unusual photophysical properties, consistent with DNA-induced interaction of the phenanthroline rings. The change in relative intensity of emission with (IDNA/Ifree) increasing DNA concentration, in aerated 10 mM phosphate buÜer, is presented for the series of phen complexes over a range of [Nu]/[Ru] (0»40) in Fig. 1 and Table 1. Under conditions where 100% of the molecules are bound to DNA (i.e. in the plateau region of the plot), only the bimetallic heptanelinked complex 2b exhibited a relative increase of emission intensity comparable to that for the monometallic and consistent with both centres being (IDNA/Ifree\2.40) bound in a similar manner.(We anticipate that the mode of binding is similar to that of which is partially [Ru(phen)3]2` intercalated between the base pairs leading to a kink in the DNA.5) The decane-linked complex 2c exhibits an intensity increase only half of that found for 2b (i.e. IDNA/Ifree\1.68). The longer chain therefore seems to prevent interaction of the second centre with the base pairs of DNA.On the other hand, for 2a (n\5) where it appears that the chain IDNA/Ifree\2.08, is too short to facilitate full interaction of the phen ligand on the second centre. In agreement with this the binding site sizes per complex determined using the McGhee von Hippel approach10 (Table 1) are 3.2 base pairs for the monometallic complex [Ru(phen)2(Me2bipy)]2` (Me2bipy\4,4@-dimethyl- 8.8 base pairs for 2b (n\7), 6.4 for 2a (n\5) 2,2@-bipyridyl), and 6.0 for 2c (n\10).Table 1 Relative emission enhancements binding constants (K) and site sizes measured in 10 mM potassium phosphate buÜer, (IDNA/Ifree), (Bap) and fraction remaining bound upon addition of 500 mM NaCl for and bimetallic phenanthroline complexes 2a»2c [Ru(phen)2(Me2bipy)]2` 10~4 Ka,c Bap a,c Fraction boundd Complex IDNA/Ifree a,b /mol dm~3 basepair~1 /basepair in 500 mM NaCl [Ru(phen)2(Me2bipy)]2` 2.40 7.5 3.2 0.00 2a (n\5) 2.08 240 6.4 0.12 2b (n\7) 2.36 360 8.8 0.21 2c (n\10) 1.68 300 6.0 0.08 a [Ru centre]\1.0]10~5 M in 10mM potassium phosphate buÜer, pH\6.9.b Conditions were [Nu]/[Ru]\80 and correspond to 100% bound ruthenium complexes.c These values were determined from the emission data using the McGhee von Hippel approach. d In buÜer containing 10 mM phosphate and 500 mM NaCl with [Nu]/[Ru]\40 for and 20 for bimetallic complexes 2a»2c. The [Ru(phen)2(Me2bipy)]2` fraction bound is determined from the pre-exponential factor of the long-lived component of the emission decay and assumes that the short-lived component corresponds to ruthenium centres which are not bound to the DNA.New J. Chem., 1998, Pages 215»217 215ADNA / Afree Fig. 1 Variation of relative emission intensity (at k\610 nm) upon addition of salmon sperm DNA to [Ru(phen)2(Me2bipy)]2` and to the bimetallics (K) [(phen)2RuII(Mebipy)-(CH2)n- 2a, n\5; 2b, n\7; 2c, n\10. (bipyMe)RuII(phen)2]4` (L) (=) (+) [Ru centre]\1.0]10~5 M in aerated 10 mM potassium phosphate buÜer, pH\6.9. Insert : expansion of eÜect at low [Nu]/[Ru] ratios (0]1.5) A serious drawback in the application of many mononuclear ruthernium complexes is their poor binding at high ionic strengths.In agreement with this, our steady-state and time-correlated emission studies show that which is 100% bound in 10 mM [Ru(phen)2(Me2bipy)]2` buÜer is completely displaced from DNA upon addition of 250 mM NaCl.By contrast a signi–cant fraction of the bimetallic complexes remain bound even in the presence of 500 mM NaCl, with 2b showing a particularly high affinity (Table 1). This suggests that non-electrostatic binding is considerably more important for 2b and therefore corroborates the results presented above. An additional feature of the bimetallic complexes 2 is observed at high binding ratios (i.e.low [Nu]/[Ru] ratios). Careful examination of the eÜects of binding on the absorption spectra shows that when the absorbance ratio is plotted as a function of [DNA] for both sets of (ADNA/Afree) bimetallic complexes 1 and 2 a two-stage [Nu]/[Ru]-dependent process is revealed, in contrast to that observed for the monometallic species (Fig. 2). The emis- [Ru(L)2(Me2bipy)]2` sion behaviour at low [Nu]/[Ru] (0»1) ratios of the bimetallic phenanthroline complexes 2 however is strikingly diÜerent from that of the bipy analogues 1. Thus whereas titration plots indicated that and the binu- [Ru(phen)2(Me2bipy)]2` clear bipy complexes 1a and 1b each exhibit an increase in luminescence intensity upon addition of DNA, a pronounced decrease in the of 5»10% was observed for each of IDNA/Ifree 2a»2c (inset in Fig. 1). The magnitude of this quenching process varies as a function of chain length according to n\7[5[10. The time-resolved emission data for 2a Fig. 2 Change in the absorbance upon addition of salmon sperm DNA to the ruthenium complex solutions in 10 mM phosphate buÜer.(Absorbance measured at the for the unbound complex). [Ru k' centre]\1.0]10~5 M in 10 mM potassium phosphate buÜer, pH\6.9 (n\5) in deaerated solution at [Nu]/[Ru]\1 indicated the presence of three components including a species with a lifetime signi–cantly shorter ns) than that of the (qDNA\568 complex free in solution ns). These results (qfree\1341 together with the absorption spectral data above correlate well with there being at equimolar concentrations of metal centre and nucleotide, a DNA-induced stacking conformation of the bimetallic phenanthroline complexes, which facilitates a partial emission quenching process.The fact that no emission quenching is observed for the bimetallic bipy complexes may be attributed to their less extended aromatic ligands.In conclusion we have shown that in designing bimetallic systems as DNA probes it is of great importance to optimize both the linker length and the nature of the other interacting ligand. With complexes 2a»2c only the heptane-linker allows full interaction of both centres and this complex also shows the greatest resistance to increased ionic strength of the solution.The signi–cant self-quenching observed for the phen complexes at low [Nu]/[Ru] ratios may also be an important factor. Experimental Synthesis of complexes The ligands (Mebipy)- were prepared by a (CH2)n-(bipyMe) modi–ed version of the procedure used by Furue et al.11 Complexes (L\phen or [L2Ru(Mebipy)-(CH2)n-(bipyMe)RuL2]4` bipy) were prepared by reaction of with the appro- RuL2Cl2 priate ligand.These products, which are mixtures of stereoisomers, were observed as single compounds on HPLC and TLC. 1H NMR spectra were consistent with the proposed structures. An example of the preparative method is given below. (2a). [ (phen)2Ru(Mebipy)-(CH2)5-(bipyMe)Ru(phen)2 ] [PF6 ] 4 An aqueous methanol (1 : 1, v/v) solution (30 cm3) of (235 mg; 0.44 mmol) was re—uxed for 1 h under Ru(phen)2Cl2 nitrogen, and then an aqueous methanol (1 : 1, v/v) solution (25 cm3) of (Mebipy)- (86 mg; 0.21 mmol) was (CH2)5-(bipyMe) added over a period of 30 min.The solution was re—uxed for a further 23 h and monitored by UV/VIS spectroscopy. The complex 2a was separated with an SP-Sephadex (C-25) cation exchange resin (40»120 l) using 0.50 M aqueous NaCl as eluant and precipitated as its salt using a saturated solu- PF6 tion of (yield\158 mg, 41%).Further puri–cation was KPF6 achieved by semi-preparative cationic exchange HPLC using an eluant of (80 : 20, v/v), 0.15 M and CH3CN : H2O KNO3 subsequent precipitation as its salt. PF6 Reactions with DNA Salmon sperm DNA (Sigma) was puri–ed to remove protein and dialysed into 10 mM potassium phosphate buÜer (pH\6.9).Experiments with DNA were carried out in 10mM potassium phosphate buÜer (pH\6.9) using Millipore Milli-Q water, previously sterilized by autoclaving. The ruthenium complexes were converted to their chloride salts using Amberlite IRA 402 ion exchange resin. All ruthenium and DNA solutions were prepared fresh prior to each experiment. The ruthenium centre concentration was –xed at 10 lM, and DNA from a concentrated solution was added to achieve the desired nucleotide to ruthenium centre ratios ([Nu]/[Ru]) of 0»80.Absorption spectra were recorded on a Pye-Unicam SP8-800 or SP8-200 UV/VIS spectrophotometer with data stored, manipulated and analysed using a PC computer. Steady-state emission spectra and luminescence intensity data were recorded on a Perkin-Elmer MPF-44B —uorescence 216 New J.Chem., 1998, Pages 215»217spectrophotometer and spectra were subsequently corrected. Luminescence lifetimes (q) were measured using a timecorrelated single photon counting apparatus (Edinburgh Instruments FL-900) with as the emission gas N2 (kexcit\359 nm, frequency\35 kHz) and using a gated delay generator. All solutions were thermostated to 25 °C.Acknowledgements –nancial support of Forbairt and the European Com- The munity and helpful discussions with Dr. A. Kirsch-De Mesmaeker are gratefully acknowledged. We thank Ms. Louise Marchand for help with preliminary studies. References 1 B. Norden, P. Lincoln, B. Akerman and E. Tuite, in Metal Ions in Biological Systems, eds.A. Sigel and H. Sigel, Dekker, New York 1996, vol. 33, p. 177. 2 J. K. Barton, J. Biomol. Struct. Dynam., 1983, 1, 621; A. B. Tossi and J. M. Kelly, Photochem. Photobiol., 1989, 49, 545; R. M. Hartshorn and J. K. Barton, J. Am. Chem. Soc., 1992, 114, 5919; K. Naing, M. Takahashi, M. Taniguichi and A. Yamagishi, Inorg. Chem., 1995, 34, 350. 3 M. R. Arkin, E. D. A. Stemp, R. E.Holmin, J. K. Barton, A Hoé lmin, E. J. C. Olson and P. F. Barbara, Science, 1996, 273, 475; E. Tuite, P. Lincoln and B. Norden, J. Am. Chem. Soc., 1997, 119, 239; J. P. Lecomte, A. Kirsch-De Mesmaeker, M. M. Feeney and J. M. Kelly, Inorg. Chem., 1995, 34, 6481. 4 M. A. Billadeau and H. Morrison, in Metal Ions in Biological Systems, eds. A. Sigel and H. Sigel, Dekker, New York 1996, vol. 33, p. 269; L. Jacquet, J. M. Kelly and A. Kirsch-De Mesmaeker, J. Chem. Soc., Chem. Commun., 1995, 913. 5 S. Satyanarayana, J. C. Dabrowiak and J. B. Chaires, Biochemistry, 1993, 32, 2573; S.-D. Choi, M.-S. Kim, S. K. Kim, P. Lincoln, E. Tuite and B. Norden, Biochemistry, 1997, 36, 214. 6 F. M. OœReilly, J. M. Kelly and A. Kirsch-De Mesmaeker, Chem. Commun., 1996, 1013. 7 I. Sasaki, M. Imberdis, A. Gaudemer, B. Drahi, D. Azhari and E. Amouyal, New J. Chem., 1994, 18, 759. 8 P. Lincoln and B. Norden, Chem. Commun., 1996, 2145. 9 A. N. Glazer and H. S. Rye, Nature (L ondon), 1992, 359, 859. 10 J. D. McGhee and P. H. von Hippel, J. Mol. Biol., 1974, 86, 469. 11 M. Furue, T. Yoshidzumi, S. Kinoshita, T. Kushida, S.-I. Nozakura and M. Kamachi, Bull. Chem. Soc. Japn., 1991, 64, 1632. Received 21st November 1997; Paper 8/00148K New J. Chem., 1998, Pages 215»217 217

ISSN:1144-0546    

DOI:10.1039/a800148k

出版商:RSC    

年代:1998

数据来源: RSC

摘要:

L e t t e r N N N N N Cl Cl cpqpy A tetranuclear meso-helicate Edwin C. Constable,* Markus Neuburger, Louise A. Whall and Margareta Zehnder Institut Anorganische Chemie, Spitalstrasse 51, CH-4056 Basel, Switzerland fué r A tetranuclear complex formed from the linking together of two double-helical dicobalt species by acetate bridges has been structurally characterised ; the tetranuclear unit consists of two helicates of opposite P and M chirality.Oligopyridines and oligopyridine metal-binding domains are a recurrent feature in metallosupramolecular chemistry1 and it is well-established that the higher oligopyridines can give multiple-helical transition metal complexes.2 Cobalt has been shown to form both mononuclear helical3,4 and dinuclear double-helical complexes5h7 with 2,2@:6@,2A:6A,2”:6”,26- (qpy), in which the precise structures are quinquepyridines de–ned by reaction conditions and substituents.We have utilised interconversions within cobalt(II) complexes for the speci –c assembly of heteronuclear helicates.8,9 In the course of these studies we have further investigated the factors controlling the formation of mononuclear or dinuclear cobalt complexes and now report the formation of a bis(double helicate) as a solid state species.The reaction of cobalt(II) acetate with 4@,4”-bis(4-chlorophenyl)- 2,2@:6@,2A:6A,2”:6”, (cpqpy) in dry 26-quinquepyridine methanol yields an orange solution from which an orange solid of stoichiometry may be pre- MCo2(cpqpy)2(OAc)(PF6)3N cipitated by the addition of ammonium hexa- —uorophosphate.§ The 1H NMR spectrum of a solution of this orange product in revealed one major solution CD3CN species with 19 paramagnetically shifted resonances [d 263, 139, 117, 91, 84, 74, 53, 52, 51.5, 28, 25 (2H), 21, 19, 16 (2H), 7.2, 7 (2H), 5 (2H), [3, [10], as expected for a system in which each of the rings of the cpqpy ligand are nonequivalent.A lower intensity sub-spectrum indicated a system symmetrical about the central pyridine ring of the ligand and corresponded exactly to the mononuclear species that we have previously character- [Co(cpqpy)(MeCN)2]2` ised.5,8 Recrystallisation from acetonitrile»diethyl ether gave orange blocks of the major component.In contrast, recrystallisation from damp solvents yielded predominantly the yellow/cream mononuclear compounds we have previously described.Furthermore, an orange solution of the isolated orange compound in methanol changed to a yellow solution of the mononuclear compound upon treatment with water § Selected data for [M(cpqpy)2Co2N(l-OAc)2MCo2(cpqpy)2N]- mp[250 °C, MALDI-TOF mass spectrum: 1338 [PF6]6 … 12MeCN: 1278 785 728 MCo2(cpqpy)2N`, MCo(cpqpy)2N`, MCo2(cpqpy)OAcN`, IR (KBr): m\1613s, 1574m, 1547m, 1481m, 1450m, MCo2(cpqpy)N`. 1422m, 1386m, 1247w, 1096m, 1012m, 842vs, 792s, 651w, 558s cm~1. and stirring for 30 min at room temperature. The TOF mass spectrum of the orange species exhibited peaks corresponding to MCo2(cpqpy)2N`, MCo(cpqpy)2N`, MCo2(cpqpy)OAcN`, and MCo(cpqpy)N`, and these data combined MCo2(cpqpy)N` with the 1H NMR spectrum suggest the formulation of a dinuclear double helix analogous to the cation [Co2(qpy)2- The IR spectrum exhibits stretches at 1574 and (OAc)]3`.5 1421 cm~1 that were tentatively assigned to a bridging, chelating or semi-chelating acetate ligand.In view of the uncertainty over the precise nature of these cobalt complexes, we have determined the solid state structure of the major orange component of the reaction of cobalt(II) acetate with cpqpy.The crystal structure determination revealed the correct formulation of the orange product to be [M(cpqpy)2Co2N(land the cationic unit OAc)2MCo2(cpqpy)2N][PF6]6 … 12MeCNî is presented in Fig. 1 together with a stereoview in Fig. 2. Firstly, the predicted dinuclear double-helical species has indeed been formed. The presence of two six-coordinate cobalt(II) centres requires a total of twelve donor atoms, which in are provided by the 10 nitrogen atoms [Co2(qpy)2(OAc)]3` of the double-helical ligand array and a didentate chelating acetate.5 In this case, instead of a single chelating acetate ligand providing two oxygen donors per double helix, there are two bridging acetate ligands linking two helicates together.The two helicates are related by an inversion centre and so the overall tetranuclear unit comprises one dinuclear unit of P and one of M chirality (Fig. 3). The end result is an achiral meso-helicate consisting of two linked helicates of opposite chirality. We note that this usage diÜers slightly from the use adopted by Albrect to describe systems in which two metal centres within a dinuclear complex exhibit opposite chirality»in the latter case the meso description is certainly correct but it may be argued whether the complexes should now be called helicates.The central î Crystal data for [M(cpqpy)2Co2N(l-OAc)2MCo2(cpqpy)2N]- orange block, [PF6]6 … 12MeCN: C176H134Co4N32Cl8O4P6F39 . M\4150.35, triclinic, spacegroup a\13.868(2), b\15.271(2), P1 6 , c\22.336(4) a\87.95(1), b\89.65(1), c\75.39(1)°, Aé , U\49 245(9) Z\1, g cm~3, F(000)\2104, Aé 3, Dc\1.51 k\1.541 80 l(Cu-Ka) 5.41 mm~1.Intensity data were collected by Aé , the x/2h scan method (2.52\h\66.74°) ; for a crystal of dimensions 0.22]0.24]0.32 mm at 223 K; of 12 182 (11 884 independent) re—ections measured, 7178 [IP2r(I)] were used in the structure solution.Signi–cant solvent loss occurred over the data collection period (19.47%), even at low temperature. Because of the high decay, absorption corrections were carried out using DIFABS. The structure was solved by direct methods using CRYSTALS [D. J. Watkin, J. R. Carruthers and P. Betteridge, Chemical Crystallography Laboratory, Oxford, UK] to give –nal R and values of 0.0719 and 0.0475, RW respectively [Chebychev polynomial weighting : J.R. Carruthers and D. J. Watkin, Acta Crystallogr., Sect. A, 1979, 35, 698]. CCDC reference number 440/018. New J. Chem., 1998, Pages 219»220 219Fig. 1 The structure of the tetranuclear cation present in showing [M(cpqpy)2Co2N(l-OAc)2MCo2(cpqpy)2N][PF6]6 … 12MeCN the numbering scheme adopted for the rings. Selected bond lengths Co(1)wN(H), 2.195(6) ; Co(1)wN(J), 2.037(6) ; Co(1)wN(L), (Aé ) : 2.204(6) ; Co(1)wN(A), 2.180(6) ; Co(1)wN(B), 2.054(6) ; Co(1)wN(D), 2.218(6) ; Co(2)wN(M), 2.207(6) ; Co(2)wN(P), 2.088(6) ; Co(2)wN(E), 2.251(6) ; Co(2)wN(G), 2.108(6) ; Co(2)wO(1), 2.045(5) ; Co(2)wO(2)@, 2.042(6) CowOwCwOwCowOwCwO eight-membered ring is slightly riffled.The expected p-stacking between the ligand strands is present, with centroid-to-centroid distances of 3.411»3.639 observed between the pairs M-B, L-G, P-D and Aé J-E; each pair is approximately coplanar with least squares planes in the range of 4.03»7.35 The chlorophenyl substit- Aé .uents show no interactions with any other aromatic rings. There are also a number of diÜerences from the dinuclear species previously characterised ; the [Co2(qpy)2(OAc)]3` Co(1)wCo(2) distance is considerably greater in the tetranuclear species than in the dinuclear one (4.776 versus 4.461 Aé This may be traced to changes in the pitch of the helix and Aé ).the cumulative eÜect of changes in the dihedral angles between adjacent rings. The Co(1)wCo(1@) distance is 4.764 In the Aé . case of the lattice consists of equal [Co2(qpy)2(OAc)][PF6]3, numbers of discrete P and M helices, whereas with cpqpy, helices of opposite chirality are linked by acetate bridges.The driving force for this behaviour is unclear, and there are no short contacts between the cation and the anions or the lattice solvent molecules. The similarity of the solution 1H NMR spectrum of this complex with those of and other [Co2(qpy)2(OAc)][PF6]3 complexes with substituted qpy ligands that are unam- Fig. 2 Stereoviews of the structure of the tetranuclear cation present in [M(cpqpy)2Co2N(l-OAc)2MCo2(cpqpy)2N][PF6]6 … 12MeCN Fig. 3 A space-–lling representation of the tetranuclear cation present in [M(cpqpy)2Co2N(l-OAc)2MCo2(cpqpy)2N][PF6]6 … 12MeCN, emphasising the opposite chirality of the two dinuclear components.biguously dinuclear double helicates in the solid state leads us to suggest that the bridging acetates do not persist in solution and that the aggregation is a solid state phenomenon. We are currently extending these studies to co-crystallisation of complexes with chiral oligopyridines.10,11 Acknowledgements should like to thank the Schweizerischer Nationalfonds We zur Foé rderung der wissenschaftlichen Forschung for support.References 1 E. C. Constable, Prog. Inorg. Chem., 1994, 42, 67; T etrahedron, 1992, 48, 10013. 2 E. C. Constable, in Comprehensive Supramolecular Chemistry, ed. J.-M. Lehn, Pergamon, Oxford, 1996, vol. 9, p. 213. 3 K. A. Gheysen, K. T. Potts, H. C. Hurrell and H. D. Abrun8 a, Inorg. Chem., 1990, 29, 1589. 4 E. C. Constable, J.V. Walker, D. A. Tocher and M. A. M. Daniels, J. Chem. Soc., Chem. Commun., 1992, 768. 5 E. C. Constable, S. M. Elder, P. R. Raithby and M. D. Ward, Polyhedron, 1991, 10, 1395. 6 M. Barley, E. C. Constable, S. A. Corr, R. C. S. McQueen, J. C. Nutkins, M. D. Ward and M. G. B. Drew, J. Chem. Soc., Dalton T rans., 1988, 2655. 7 K. T. Potts, K. M. Keshavarz, F. S. Tham, H. D. Abrun8 a and C. Arana, Inorg. Chem., 1993, 32, 4436. 8 E. C. Constable and J. V. Walker, J. Chem. Soc., Chem. Commun., 1992, 884. 9 E. C. Constable, A. J. Edwards, P. R. Raithby and J. V. Walker, Angew. Chem., Int. Ed. Engl., 1993, 32, 1465. 10 E. C. Constable, T. Kulke, M. Neuburger and M. Zehnder, Chem. Commun., 1997, 489 11 E. C. Constable, T. Kulke, M. Neuburger and M. Zehnder, Chem. Commun., 1997, in the press. Received 17th November 1997; Paper 7/09243A 220 New J. Chem., 1998, Pages 219»220New J. Chem., 1998, Pages 219»220 221

ISSN:1144-0546    

DOI:10.1039/a709243a

出版商:RSC    

年代:1998

数据来源: RSC

摘要:

L e t t e r O O– HO R O Ph–N=O 2 O R O O N O H O – I O O O R O NH OH + H + 3 4 0 100 200 300 400 0.04 0.08 0.12 0.16 kobs / s–1 kobs / s–1 [H+]–1 [H+] 0 0.05 0.05 0.1 0.10 0.15 Interaction of L-ascorbate with substituted nitrosobenzenes. Role of the ascorbate 2-OH group in antioxidant reactions Stanko Ursó icç , Valerije Vrcó ek, Danijel Ljubas and Ivana Vinkovicç Faculty of Pharmacy and Biochemistry, University of Zagreb, 10 000 Zagreb, A.Kovacica 1, Croatia L-Ascorbate reduces substituted nitrosobenzenes in aqueous solution, in a process that includes a cyclic transition state composed of ascorbate and nitrosobenzene and in which concerted electron and proton transfer leads to the products. The results suggest that the mechanism by which ascorbate interacts with the a-tocopheroxyl radical in the exceptionally important a-tocopherol»ascorbate redox cycle also includes the corresponding cyclic transition state.L-Ascorbic acid is a ìsimple yet mysterious moleculeœ.1 Ascorbate is of exceptional importance in the antioxidant protection of the human and other aerobic organisms. It is considered the most important antioxidant in the human plasma,2 where ascorbate decreases, among others, the oxidative damage of DNA in lymphocytes3 and prevents the metalion- dependent peroxydation of low-density lipoprotein.4 Ascorbate also acts synergistically with other physiological antioxidants5 and is a biological reductant of the human carcinogen CrVI.6 Support of the antioxidant activity of atocopherol by ascorbate in biological membranes, lipoproteins, cells and plasma has been reported.7 In addition, nitric oxide (NO) interacts with the redox cycle involving atocopherol and ascorbate, where ascorbate regenerates atocopherol from the a-tocopheroxyl radical.8 The inactivation of peroxy and similar radical species via the interaction with ascorbate is, in many cases, a fundamental part of the antioxidant action of ascorbate.Ordinarily, an electron and a proton are transferred to the radical, giving the corresponding reduced product and the ascorbyl radical. On the other hand, interaction of ascorbate with the atocopheroxyl radical is of key importance for sustaining the antioxidant capacity of a-tocopherol.7 Recently, it has been observed9a that ascorbate 1 reduces nitrosobenzene 2, giving the corresponding phenylhydroxylamine 3 and dehydroascorbic acid 4 (see Scheme 1).The reaction is of particular interest because of the cytotoxic and antiretroviral properties of the nitrosobenzenes.9 Nitrosobenzenes also can be present in the human blood as products of the reduction of aromatic nitro compounds, introduced Scheme 1 The interaction of ascorbate and nitrosobenzene giving phenylhydroxylamine and dehydroascorbic acid into the organism as a toxin.10 We report here the results of our study on the mechanism of the biochemically important interaction between ascorbate and the aromatic nitroso group11 of substituted nitrosobenzenes. Our observations are as follows : (i) At constant hydronium ion concentration, the observed reaction kinetics are second-order overall, and –rst-order with respect to both ascorbate and nitrosobenzene.(ii) The dependence of the observed rate constants on the hydronium ion concentration indicates (Fig. 1) that the ascorbate anion as well as ascorbic acid interact with nitrosobenzene. 13 (iii) The Hammet plot of log vs. r` (q`\1.31, kobs r\0.990) indicates that the order of reactivity of the substituted nitrosobenzenes is related to the electron-withdrawing Fig. 1 The dependence on the hydronium ion concentration (mol dm~3) of the pseudo-–rst-order rate constants for the reaction of ascorbate/ascorbic acid in water at 25 °C. Rate constants were determined spectrophotometrically by measuring the disappearance of the absorbance of nitrosobenzene at 308 nm. Initial concentrations of total ascorbate/ascorbic acid and nitrosobenzene were 0.0024 and 0.00024 mol dm~3, respectively.The corresponding values of kHA~ and the second-order rate constants for the reaction of nitroso- kH2A , benzene with ascorbate and undissociated ascorbic acid, are 2830(83) mol~1 dm3 s~1 and 2.90(0.49) mol~1 dm3 s~1, respectively New J. Chem., 1998, Pages 221»223 221–0.4 –0.2 0 0.2 0.4 1.8 2.3 2.8 log kobs / mol–1 dm3 s–1 s+ I Ph N OH + Asc• – Ph N OH + Asc – + Asc• – Asc• – + Asc• – H + Asc – + Deasc Net reaction: I + H + Ph NH OH Ph NH OH Deasc + fast • + • O R O H O O O O OH H HOH2C • d– d– II Fig. 2 Plot of log vs. Hammett r`. T \25 °C. Rate constants kobs were determined spectrophotometrically (see Fig. 1), using the secondorder kinetics with mol dm~3 [uNO]\[ascorbate]total\0.00024 and [H`]\0.0008 mol dm~3 properties of the ring substituents, R (Fig. 2). This suggests that the electron transfer could participate in the rate control of the process. For example, the correlation of the rate constants for the reaction of various tocopherols with variously substituted phenoxyl radicals vs. Hammett r parameters was observed, and the proposal was made that charge transfer should play an important role in this reaction.17b On the other hand the enthalpy of activation kJ mol~1 is *HE\20.3 relatively low, similar to the values observed in the reactions of ascorbate with a-tocopheroxyl radicals, for which a concerted electron and proton transfer was proposed.7 (iv) The value of the rate constant ratio between and H2O at pH 3.15, is 8.81, which leads14 to a value of 3.1 for D2O, the kinetic primary deuterium isotope eÜect Although kH/kD .lower than the maximum isotope eÜect expected for the ratecontrolling OwH bond breaking,15 this number is still consistent with a rate-controlling proton transfer in the transition state.15,16 Moreover, similar isotope eÜects were observed in the reaction of ascorbate with the a-tocopheroxyl radical7 and in the reactions of the a-tocopheroxyl radical with alkyl hydroperoxides,17 and have been interpreted as a ratecontrolling proton transfer concerted with an electron transfer in the process.(v) A large negative entropy of activation J *SE\[111 mol~1 K~1 was observed for this reaction (see Table 1). This is consistent with a quite ordered transition state and is strongly in support of a cyclic transition state like I, as proposed for the reaction.In our view, the results are consistent with the mechanism described by Scheme 1 and Scheme 2 (which is added separately for clarity). We believe that the reaction proceeds via a Scheme 2 Proposed mechanism for the conversion of I to the –nal products. (Asc\ascorbate, Deasc\dehydroascorbic acid) cyclic transition state similar to I, where transfer of the 2-OH proton of ascorbate to the nitroso oxygen is concerted with the transfer of an electron from the anionic oxygen of ascorbate to the nitroso nitrogen.18 The signi–cance of the observations could be to help in the further elucidation of the mechanism by which ascorbate interacts with the a-tocopheroxyl radical in the exceptionally important a-tocopherol»ascorbate redox cycle, and perhaps on the interactions of ascorbate with peroxy and similar radical species. There are open questions about the structure of the transition state for the reaction of the a-tocopheroxyl radical with ascorbate.7 Now, the comparison of the activation parameters and kinetic isotope eÜects for this reaction7 and the values obtained in this study (see Table 1) suggest the close similarity between the transition states of the two reactions.Therefore, it seems reasonable to conclude that the reaction of the atocopheroxyl radical with ascorbate proceeds also via a cyclic transition state, similar to structure II (see Scheme 3). In addition, it is normally expected that the anionic oxygen of ascorbate interacts with the a-tocopheroxyl radical centre.Many reactions of ascorbate with oxygen radicals are known,19 and the ascorbate 2-OH group enables the proton transfer concerted with the electron transfer to the radical centre in the transition state. Probably, this conclusion is corroborated by the observation that a proton can be transferred from the solvent molecule to the transition state for the reduction of peroxide radicals with organic reductants.20 In both of the reactions, that of ascorbate with the atocopheroxyl radical and that of ascorbate with nitrosobenzene, formation of a cyclic transition state that includes the 2-OH proton makes the reaction thermodynamically more Scheme 3 Proposed transition state for the reaction of the atocopheroxyl radical with ascorbate Table 1 Comparison of the activation parameters and kinetic isotope eÜects for the reactions of ascorbate with nitrosobenzene and ascorbate with a-tocopheroxyl radicals Oxidant *HE/kJ mol~1 *SE/J mol~1 K~1 kH/kD k2 e/dm3 mol~1 s~1 Nitrosobenzenea 20.3(1.8)c [110.9(7.5)c 3.06d 2.83]103 a-Tocopheroxyl radicalb,f 18.9(0.5) [103.3(1.7) 7.97 4.97]104 a-Tocopheroxyl radicalb,g 28.8(1.3) [57.9(4.1) 3.11 3.05]105g Trolox C radicalb,h 2.2(0.8) [109.0(2.8) 7.00 1.44]107 a This work.b Ref. 7. Trolox C is the a-tocopherol analogue 6-hydroxyl-2,5,7,8-tetramethylchroman-2-carboxylic acid. c From measurements at –ve temperatures, in the range of 15.35»44.45 °C. At least three runs at each temperature were made.In order to obtain values of the k2 , second-order rate constants for the reaction of ascorbate with nitrosobenzene, the temperature dependence of the was determined at the pKas same temperatures. d The ratio of rate constants in corrected for the solvent isotope eÜect on the dissociation of ascorbic acid (see H2O/D2O, above). Rate constants are an average of three paired measurements.e At 25 °C, except otherwise noted. f In sodium dodecyl sulfate micelles. g At 35 °C, in dimirystoylphosphatidylcholine bilayers. h In water. 222 New J. Chem., 1998, Pages 221»223favourable. Moreover, the enthalpy of activation in the reaction of ascorbate with the a-tocopheroxyl radical (Table 1) is small or near zero, and the reaction is entropy-governed. We also have investigated (in part) the reaction of 2-nitroso- 2-methylpropane with ascorbate.The reaction is about three orders of magnitude slower than the one with nitrosobenzene. Another diÜerence is that the pH rate pro–le shows a minimum at about pH 2.4, while the observed rate constants increase above and below this point. We are continuing the investigation of this reaction, as well as the reaction of ascorbic acid with nitrosobenzenes.Acknowledgements gratefully acknowledge the collaboration of Miss Jelena We Buzó ancë icç . We thank the Ministry of Science and Technology of the Republic of Croatia for support (project 006142). References 1 M. Stacey, in V itamin C: Its Chemistry and Biochemistry, ed. M. B. Davies, J. Austin and D. A. Partridge, The Royal Society of Chemistry, Cambridge, 1991. 2 J. M. May, Z. Qu and R. R. Whitesell, Biochemistry, 1995, 34, 12721 and references therein. 3 S. J. Duthie, A. G. Ross and A. R. Collins, Cancer Res., 1996, 56, 1291. 4 K. L. Retsky and B. Frei, Biochim. Biophys. Acta, 1995, 1257, 279. 5 T. Yamaguchi, F. Horio, T. Hashizume, M. Tanaka, S. Ikeda, A. Kakinuma and H. Nakajima, Biochem. Biophys. Res.Commun., 1995, 214, 11; M. Venugopal, J. M. Jamison, J. Giloteaux, J. A. Koch, M. Summers, D. Gianmar, C. Sovick and J. L. Summers, L ife Sci., 1996, 59,1389. 6 L. Zhang and P. A. Lay, J. Am. Chem. Soc., 1996, 118, 12624; D. M. Stearns and K. E. Wetterhahn, Chem. Res. T oxicol., 1997, 10, 271. 7 R. H. Bisby and A. W. Parker, Arch. Biochem. Biophys., 1995, 317, 170 and references therein. 8 N. V. Gorbunov, A. N. Osipov, M. A. Sweetland, B. W. Day, N. M. Elsayed and V. E. Kagan, Biochem. Biophys. Res. Commun., 1996, 219, 835. 9 (a) J. Mendeleyev, E. Kirsten, A. Hakam, K. G. Buki and E. Kun, Biochem. Pharmacol., 1995, 50, 705. (b) W. R. Rice, C. A. SchaÜer, B. Harten, F. Villinger, T. L. South, M. F. Summers, L. E. Henderson, J. W. Bess Jr., L. O. Arthur, J. S. McDougal, S.L. OrloÜ, J. Mendeleyev and E. Kun, Nature (L ondon), 1993, 361, 473. 10 P. Zuman and B. Shah, Chem. Rev., 1994, 1621. 11 This study presents a part of our ongoing investigations on the mechanisms of interactions of C-nitroso compounds with biochemically important molecules.12 To our knowledge, no report on the kinetics or mechanism of the reaction of nitrosobenzene with ascorbate has been reported. 12 (a) O.Kronja, J. Matijevicç -Sosa and S. Ursó icç , J. Chem. Soc., Chem. Commun., 1987, 463. (b) S. Ursó icç , V. Vrcó ek, M. Gabricó evicç and B. Zorc, J. Chem. Soc., Chem. Commun., 1992, 296. (c) S. Ursó icç , Helv. Chim. Acta, 1993, 76, 131. (d) V. Pilepicç and S. Ursó icç , T etrahedron L ett., 1994, 35, 7425. (e) D. Radivoj, V. Pilepicç and S.Ursó icç , Croat. Chem. Acta, 1996, 69, 1633. 13 Under the conditions employed, the reaction of undissociated ascorbic acid with nitrosobenzene is a minor part of the overall process (see also Fig. 1) and the obtained results can be, to a good approximation, referred to the reaction of nitrosobenzene with ascorbate. 14 Taking into account the value of 2.88 for the antilog *pKa between and for ascorbic acid.This is 0.46, D2O H2O *pKa whereas we have determined the for ascorbate to be 4.30 in pKa ìpure waterœ, without any added salt. The concentration of the total ascorbic acid/ascorbate was 1.2]10~3 mol dm~3 throughout. We have determined spectrophotometrically the solvent deuterium isotope eÜect on the dissociation of ascorbic acid (to the best of our knowledge, this isotope eÜect was not determined previously). 15 R. A. More OœFerrall, in Proton T ransfer Reactions, ed. E. Caldin and V. Gold, Chapman and Hall, London, 1975, ch. 8. 16 T. H. Lowry and K. S. Richardson, Mechanism and T heory in Organic Chemistry, Harper Collins, New York, 1987, pp. 232»238. 17 (a) S. Nagaoka, K. Sawada, Y. Fukumoto, U. Nagashima, S. Katsumata and K. Mukai, J. Phys. Chem., 1992, 96, 6663. (b) S. Nagaoka, A. Kuranaka, H. Tsuboi, U. Nagashima and K. Mukai, J. Phys. Chem., 1992, 96, 2754. 18 One referee has suggested some other possible mechanisms: transfer of the proton and two electrons, transfer of a hydride ion, or transfer of a hydrogen atom and one electron. 19 P. Neta, R. E. Huie, S. Mosseri, L. V. Shastri, J. P. Mittal, P. Maruthamuthu and S. Steenken, J. Phys. Chem., 1989, 93, 4099. 20 P. Neta, R. E. Huie, P. Maruthamuthu and S. Steenken, J. Phys. Chem., 1989, 93, 7654. Received 14th October 1997; Paper 7/09241E New J. Chem., 1998, Pages 221»223 223

ISSN:1144-0546    

DOI:10.1039/a709241e

出版商:RSC    

年代:1998

数据来源: RSC

摘要:

N NH N NH COOH COOH E Z Synthesis and aggregation behaviour of two-headed surfactants containing the urocanic acid moiety Sophie Franceschi,a Valeç rie Andreu,a Nancy de Viguerie,a Monique Riviere,*,a Armand Lattesa and Andreç Moisandb a L aboratoire des IMRCP, Paul Sabatier, 118, route de Narbonne, 31062 T oulouse Universiteç cedex, France b Institut de Pharmacologie et Biologie Structurale, CNRS, 205, route de Narbonne, 31077 T oulouse cedex, France Urocanic acid (3-[1H-imidazol-4-yl]propenoic acid) has attracted great interest in photobiology for many years.We describe here the synthesis of three bolaamphiphiles with two urocanic acid heads having the general structure : UAwXwUA, where UA denotes urocanic acid and X is an alkyl chain of varying length. We show that micellisation occurs with the bolaamphiphile whose alkyl chain length is sixteen carbon atoms.For compounds having a shorter chain length, light scattering and electron microscopy suggest the formation of vesicles. Compared to urocanic acid, the bolaamphiphiles, which can form aggregates in aqueous solution, may act diÜerently on membranes and be used in formulations more easily.Synthe` se et agreç gation dœamphiphiles a ` deux te� tes comportant un motif acide urocanique. Lœacide 3-[1H-imidazol- 4-yl]propeç noïé que, ou acide urocanique, posse` de des proprieç teç s photobiologiques inteç ressantes. Des bolaamphiphiles deç riveç s de cet acide sont susceptibles de preç senter des avantages sur le plan de la formulation ainsi quœun comportement diÜeç rent au niveau des membranes.Trois bolaamphiphiles de structure AUwXwAU, ou` AU repreç sente le motif acide urocanique et X est un segment lipophile de longueur variable, ont eç teç syntheç tiseç s et leurs pheç nome` nes dœagreç gation en solution aqueuse eç tudieç s. Seul le composeç preç sentant une longueur de chai� ne de seize atomes de carbone forme des micelles. Pour les bolaamphiphiles ayant des chai� nes hydrocarboneç es plus courtes la formation de veç sicules est observeç e par diÜusion de la lumie` re et microscopie eç lectronique.Urocanic acid (3-[1H-imidazol-4-yl]propenoic acid) has been for a long time of great interest in photobiology.1,2 The two isomers E and Z (Fig. 1) are found in the epidermis. The E isomer, formed by deamination of histidine and not catabolised (absence of urocanase in the skin), accumulates in the Stratum corneum3 and is partly excreted in the sweat.4 Under UV irradiation, (E)-urocanic acid is isomerised in the skin to produce a mixture of the two isomers5 in almost equal quantities. These two isomers, possessing a large absorption band with a maximum at 270 nm and a strong molar extinction coefficient, have remarkable biological properties.Indeed, as a major chromophore present in the skin, urocanic acid acts as a photoprotective agent and for the same reason has potential applications in cosmetology. However, urocanic acid can also lead to harmful eÜects because of [2]2] cycloadditions,6 photooxidations7 and other photochemical interactions8 with various compounds of biological importance.More importantly, due to its immunosuppressive activity9h11 (attributed to the Z isomer) it is susceptible to being involved in the process of skin photocancerisation. The relationship between the formation of (Z)-urocanic acid Fig. 1 E and Z isomers of urocanic acid * Fax: (]33) 5.61.25.17.33 ; E-mail: mriviere=iris.ups-tlse.fr and the eÜects observed on the immunity system are complex and many studies are under way to elucidate the mechanism of action of urocanic acid.12,13 Despite the fact that this naturally produced compound has immunosuppressive eÜects, it remains very interesting because it can have clinical applications against the phenomenon of transplant rejection or for the treatment of skin diseases like psoriasis.Therefore, we have been interested in various derivatives of this compound, particularly long chain analogues.While keeping a photoprotective activity they may, due to their hydrophobic character, operate diÜerently on the membranes. Also, from the galenic point of view they can be introduced more easily into formulations, whereas urocanic acid is practically insoluble in organic and aqueous media and difficult to formulate.In a preceding work we observed that long chain esters of urocanic acid present good photoprotective qualities :14 (i) their spectral properties are analogous to those of urocanic acid ; (ii) they are anchored in aggregates in organized media (micellar solutions and microemulsions), which by extrapolation suggests similar anchoring in biological media; and (iii) their photostability in organized media is good, with urocanic acid being the only degradation product after –ve days of irradiation. Because of these useful properties, we chose to explore other kinds of long chain derivatives, the bolaamphiphiles.A bolaamphiphile15 or bolyte is constituted of two polar heads and a hydrophobic spacer. Bolaamphiphiles are capable of organizing in an aqueous environment to form aggregates of various morphologies:16 spheres (Fig. 2a), small spherocylinders, large cylinders, discs, lamellae and vesicles (Fig. 2b). The shape of the aggregates depends on the length of the spacer and its rigidity or —exibility (in—uenced by the presence of unsaturation or ring structures). The formation of vesicles New J.Chem., 1998, Pages 225»231 225monolayer bilayer (a) sphere (b) vesicles Fig. 2 Aggregation patterns of bolaamphiphiles in (a) spheres and (b) monolayer and bilayer vesicles from such molecules is of interest since their cellular toxicity might be very low, not being able to penetrate through membranes, and therefore they could be used clinically. Such vesicles can be considered as models of stable and functionalized membranes.The synthesis of bolaform molecules and the study of their organization have been described in the literature. In particular, the aggregation behaviour in aqueous solutions of bolaamphiphiles, composed of two ammonium head groups and a single hydrocarbon chain, have been studied.17h20 From surface tension measurements and spectral property changes, the authors conclude that compounds having a spacer of less than 12 carbon atoms behave as simple electrolytes while compounds having spacers of 16 to 22 carbon atoms form micelles.Okahata and Kunitake21 have synthesized a great variety of amphiphiles constituted of two heads and a long chain. They observed by electronic microscopy the formation of globules, ribbons, monolayer discs and vesicles.Fuhrhop and Fritsch22 have investigated the extraction of natural bolaamphiphiles so as to study their cellular function. Among them, the bolaamphiphiles constituting the membrane of a microorganism have been isolated.23 The very rigid membranes allow this microorganism to survive under drastic conditions of pH, temperature and pressure. These authors have also undertaken the synthesis of more than one hundred bolaamphiphiles, for the purpose of forming functionalized vesicles.24,25 Garelli-Calvet et al.26 have synthesized bolaamphiphiles having sugar groups as polar heads (bisgluconamides and bis-lactobionamides). These compounds organize into micelles or vesicles,27 depending upon the length of the hydrocarbon chain.The aggregates formed are able to solubilize hydrophobic compounds like fatty acids and have proven to be not denaturing for lipoxygenases.These authors have also synthesized derivatives of disulfonic naphthalene acids28 as bolaamphiphiles that organise into vesicles. These compounds present an anti-HIV-1 activity more interesting than that of the corresponding monopolar derivatives, as well as a lower toxicity attributed to the lack of a rupturing eÜect on cellular membranes.Other ionic and nonionic bolytes have also been described in the literature. The head group of nonionic bolaamphiphiles includes : aroyl azidede,29 morpholine,30 arborols,31,32 polyoxyethylene,33,34 and in case of an unsymmetrical bolaamphiphile one carboxylic acid head and the other a maleic or succinic anhydride.35 The ionic head groups studied include sodium sulfate,36 pyridinium,37,38 phosphate,39 carboxylate40, an L-lysine and an amino group41 in the case of an unsymmetrical molecule.In this paper, we –rst describe the synthesis of three symmetrical bolaamphiphiles having urocanic acid head groups and a hydrocarbon spacer of 8, 12 or 16 carbon atoms: n\8 UAC8UA n\12 UAC12UA n\16 UAC16UA We then discuss the behaviour of these bolaamphiphiles at the water»air interface as studied by surface tension measurements, and their properties in an aqueous environment as explored by conductometry, spectrophotometry, light scattering and electron microscopy.The results are compared with those obtained for the monopolar derivative, C12UA: C12UA Depending upon the pH of the aqueous medium, the above four compounds can be ionic or nonionic.Therefore the aggregate formed will be in—uenced by both the pH and the length of the alkyl chain. Synthesis We have used for this synthesis the method developed earlier in our laboratory for the alkylation of methyl urocanate.42 Nq Under solid»liquid phase-transfer conditions, the dibromoalkane is allowed to react with (E)-urocanic acid in the presence of potassium carbonate, with a crown ether acting as the catalyst.After puri–cation the methyl esters of the three bolytes are obtained with 52, 45 and 65% yields, respectively. The ester function is then hydrolysed under basic conditions (Fig. 3). The purity of all the compounds has been carefully checked. Only contained signi–cant impurities UAC16UA (3»4%), in the form of the monomethyl ester arising from an incomplete hydrolysis.Fig. 3 Synthesis of urocanic bolaamphiphiles 226 New J. Chem., 1998, Pages 225»231log C S.T. / mN m–1 R–1 / mS R–1 / mS C / mol l–1 C / mol l–1 O.D. O.D. C / mol l–1 C / mol l–1 Study of Self-organization Micelles For each bolaamphiphile we tried to identify which type of aggregates form spontaneously in an aqueous solution. First, the solubility of these four compounds was determined in water at various pH (Table 1).In a neutral solution the solubility is low. Under basic conditions, behaves as a C12UA single-chain ionic surfactant with a KraÜt temperature of 25 °C. behaves similarly and also has a KraÜt tem- UAC16UA perature of 25 °C.For the two other bolytes the solubility is high. Based on these results, self-organization studies have been undertaken at pH 13.2 and 30 °C using three methods: surface tension, conductivity and UV absorbance measurements. Concentrations from 10~5 to 10~2 mol l~1 have been chosen for this study. As a reference the compound C12UA (one-headed amphiphile) has been studied.The results showing the evolution of the surface tension, the conductivity and the optical density with the concentration of amphiphile in aqueous solution at pH 13.2 are given in Figs. 4, 5 and 6. The shape of the surface tension curve (Fig. 4) of solutions of between 10~5 and 10~2 mol l~1 shows that from UAC8UA 5]10~5 to 5]10~4 mol l~1, the surface tension does not decrease, possibly due to the large solubility of this amphiphile in water; indeed, compared to the two other bolaamphiphiles, it is less hydrophobic.Above 10~3 mol l~1, the surface tension decreases strongly until it reaches a value of 35 mN m~1 at 10~2 mol l~1, implying that is posi- UAC8UA tioned at the air»water interface. Above this concentration may form micelles. UAC8UA The curve obtained for the bolaamphiphile UAC12UA shows a break point at a concentration of 4]10~4 mol l~1, which was taken as indicating a change of organization at the air»water interface.However, a real plateau is not observed and the surface tension continues to decrease for higher concentrations, so this behaviour is not characteristic of micelle formation. Fig. 4 Surface tension versus concentration of urocanic acid amphiphiles in (+ UAC12 , = UAC8UA, > UAC12UA, Ö UAC16UA) aqueous solutions (pH 13.2) at 30 °C As indicated above, contains a small amount of UAC16UA monomethyl ester.It seems, therefore, to form micelles at a concentration around 4.7]10~4 mol l~1 (c.m.c.). We notice that the amplitude of variation of the surface tension is rather small (from 72 to 55 mN m~1), compared to the behaviour shown by classical surfactants.The surface tension curve presents a minimum that may be due to the monomethyl ester impurity. Nevertheless, the surface tension curves of UAC12 and also show a slight minimum; to our know- UAC12UA ledge these compounds do not contain the monomethyl ester. Conductometry (Fig. 5) and spectrophotometry (Fig. 6) studies con–rm the results obtained by tensiometry : namely, no micellization phenomenon for a change of UAC8UA, organisation for and a critical micellar concentra- UAC12UA tion (5]10~4 mol l~1) for UAC16UA. The concentrations at which a change of organisation is observed (determined by the above three methods) and the measured surface tension minima are reported in Table 2.Other aggregates : vesicles We also investigated whether these compounds are able to form vesicles in aqueous solution, using two methods that provide information on the shape as well as the size of the aggregates. The –rst method uses light scattering. The measure of the light scattered by particles illuminated with a Fig. 5 Conductivity versus concentration of urocanic acid amphiphile aqueous solutions (pH 13.2) at 30 °C (+ UAC12UA, ) UAC16UA) Fig. 6 Optical density versus concentration of urocanic acid amphiphile aqueous solutions (pH 13.2) at 30 °C. k\322 nm for UAC12UA and k\320 nm for (Ö) UAC16UA (+) Table 1 Values of the solubility (in water and in aqueous NaOH solution), the KraÜt temperature the molar extinction coefficient (e), (Tkrafft), and the maximum wavelength (kmax) Solubility in e in kmax in Solubility in aqueous NaOH aqueous NaOH aqueous NaOH water at 30 °C/ (pH 13.2) at (pH 13.2)/ (pH 13.2)/ Compound mol l~1 30 °C/mol l~1 Tkrafft/°C 1 mol~1 cm~1 nm C12UA 9.3]10~5 [5]10~1 25 13 391 280 UAC8UA 1.5]10~4 7.9]10~2 \5 42 328 280 UAC12UA 3.6]10~5 4.1]10~2 \5 40 533 280 UAC16UA 5.5]10~6 9.6]10~3 25 38 530 280 New J.Chem., 1998, Pages 225»231 227Table 2 Values of the concentrations (mol l~1) (as determined by various methods) at which a change of organization is observed: cmin (mN m~1) is the measured surface tension at these concentrations Compound Tensiometry Conductometry Spectrophotometry cmin C12UA 1.8]10~4 ND ND 37 UAC8UA no no no no UAC12UA 4]10~4 5.4]10~4 5.7]10~4 48 UAC16UA 4.7]10~4 5]10~4 4.7]10~4 55 ND: not determined.no: no change observed. laser beam allows one to determine the size and the distribution of objects in solution. Particles having diameters between 3 and 3000 nm can be detected by this technique. The second method is electron microscopy. It allows one to visualise directly aggregates by colour-staining techniques that provide better contrast. This method allows the observation of aggregates like vesicles, but not micelles because they are dynamic and are of small size.This experimental technique can, however, deform the aggregates. Typical diameters of diÜerent aggregates formed by various one-headed amphiphiles are given in Table 3 and come from the literature.43 For vesicle preparation,44h48 we –rst used a method described for the preparation of giant polydisperse vesicles of phospholipids, namely the drying-rehydratation method.49 In this procedure, phospholipids are dissolved in an organic solvent.The solvent is then removed in order to maximize the –lm area. Water is then slowly added at 70 °C and the —ask is shaken for few seconds. However, no aggregates were obtained with the urocanic bolaamphiphiles when using this procedure.We could suppose that this method, never reported in the literature for synthetic bolaamphiphiles, was not suitable for compounds so diÜerent in structure from phospholipids. Thus we have used the only literature method for bolaamphiphiles. In this procedure, solutions of the bolaamphiphiles at pH 13.2 were sonicated. By light scattering, we observed aggregates of approximately 200 nm diameter for UAC8UA and and also very small objects (B2 nm) difficult UAC12UA to identify.Possibly, these observations can be explained by a competition between micelle and vesicle formation from an homogeneous solution. The eÜect of pH on aggregate formation has been studied, using a pH 10 solution of sodium hydroxyde and pure water. After sonication, the solutions were analysed by light scat- Table 3 Diameters of the various aggregates formed by monopolar amphiphiles.Small Large unilamellar unilamellar Giant Aggregate Micelles vesicles vesicles vesicles Diameter/nm 5 30»50 100»200 5000»200 000 tering. The diameters of all the objects observed were near 200 nm. The eÜect of variations of the power output, the % duty cycle (number of pulses during a –xed time) and the duration of the ultrasonic application have also been studied in the case of a solution of in pure water.Since no diÜerences UAC12UA were observed, the following sonication parameters have been used: 15 min duration of the ultrasonic application ; 110 watt power output; 80% duty cycle. Table 4 shows the diameters and the size distribution of the aggregates (% aggregates of a given size) of the bolaamphiphiles and in pure water and in pH 10 and UAC8UA UAC12UA pH 14 sodium hydroxide solutions, observed by light scattering. The results are the average of 10 to 20 experiments; for each size determination the standard deviation is about 20%.For these two compounds, on going from pH 14 to pH 10 the small-size particles disappeared and the formation of 200 nm vesicles is favoured.An increase in the size of vesicles on going from higher to lower pH has been observed also for double-chain surfactants with two carboxylate head groups by Engberts and co-workers50 and by Jaeger and Brown.51 Knowing that the pK of dodecanoic acid in the bilayer form is about 852 and assuming that in vesicles the second pK of dicarboxylic surfactants is superior to 8.5, the latter authors attributed the pH eÜect on the size of vesicles to the presence of a greater proportion of carboxyl groups among the carboxyl and carboxylate head groups.An incomplete deprotonation can produce hydrogen bonding between the polar heads. Thus pH seems to be an important factor in vesicle formation.53 Our results are consistent with this explanation.In the case of light scattering has not allowed UAC16UA, us to conclude that vesicles are present. The formation of vesicles for all three bolaamphiphiles was also investigated by electron microscopy. The micrographs show more or less spherical particles with well-de–ned outlines. So this technique con–rms that the three urocanic bolytes are able to form vesicles.pH Variations do not aÜect the shape and the size of the observed vesicles. The small particles of 1 or 2 nm size seen by light scattering are not observed by electron microscopy because of their small dimensions. Monolayer vesicles (about 100 nm diameter) are observed with the three urocanic bolaamphiphiles, as well as larger species probably formed by coalescence of vesicles.For electron micrographs show vesicles in UAC16UA, much smaller amounts compared to the two other com- Table 4 Diameters (z) and size distribution (%) of aggregates of the bolaamphiphiles and in pure water and in pH 10 and UAC8UA UAC12UA pH 14 sodium hydroxide solutions (observed by light scattering) UAC8UA UAC12UA NaOH soln NaOH soln NaOH soln NaOH soln Pure water pH 10 pH 14 Pure water pH 10 pH 14 z/nm % z/nm % z/nm % z/nm % z/nm % z/nm % 1 49 2 54 12 1 23 2 64 14 212 99 189 98 201 37 182 100 186 100 204 46 228 New J.Chem., 1998, Pages 225»231Fig. 7 Electron micrograph of (1 cm\55 nm) UAC8UA pounds. A small amount of monomethyl ester may explain in this case the formation of vesicles. Polydisperse multilayered vesicles (50»200 nm diameter) are also formed by (Figs 7 and 8).The thickness of the UAC8UA layer is about 3.4 nm. Conclusion Three bolaamphiphiles with two urocanic acid heads groups have been synthesized. The shape and the size of aggregates in aqueous solution at various pH have been identi–ed by light scattering and electron microscopy. Our results con–rm the importance of the length of the alkyl chain and of pH on the aggregation behaviour of bolytes with carboxylic acid heads.Experimental Solvents were purchased from Prolabo or Carlo Erba and were used after drying and distillation. The reagents were purchased from Aldrich or Acros ([98% purity). 1H and 13C NMR spectra were recorded on Bruker AC 250 and AC 400 spectrometers. The DCI, or mass NH3 CH4 spectra were recorded on a Nermag R10-10 apparatus and the FAB mass spectra on a ZAB-MS apparatus (WG-Analytical, Manchester, UK).Infrared spectra were recorded on a Perkin Elmer 683b spectrophotometer and UV spectra on a Hewlett Packard 8452 A spectrophotometer. Melting points were determined on an Electrothermal apparatus (capillary tubes). The microanalyses were carried out at the ENSCT (Toulouse, France) on a Carlo Erba 1106 instrument.Fig. 8 Electron micrograph of (1 cm\35 nm) UAC8UA Synthesis of urocanic compounds (E)-Methyl urocanate 1 was synthesized according to the method of Lauth-de Viguerie and co-workers.42 Synthesis of A solution of 1-bromododecane (4.91 UAC12 . g, 19.70 mmol) in 10 ml of anhydrous THF was added dropwise to a mixture of 1 (1.50 g, 9.85 mmol), (13.6 g, 98.5 K2CO3 mmol) and 18-crown-6 (0.26 g, 0.985 mmol) in 25 ml of anhydrous THF.The mixture was stirred for 24 h at 60 °C. After –ltration of the remaining the THF was removed by K2CO3 , vacuum evaporation. The residue was recrystallised from hexane and then puri–ed by —ash chromatography (silica gel, chloroform»ethanol, 99 : 1, v : v). methyl ester was UAC12 obtained in 95% yield.Mp\74.5 °C. eluent Rf\0.65 ; (90 : 10, v : v). 1H NMR d) : 0.87 CH2Cl2»EtOH (CDCl3 , (m, 3H, 1.24 (s, 20H, 1.70 (m, 2H, 3.76 CH3) ; CH2) ; CH2bN); (s, 3H, 3.90 (t, J\7 Hz, 2H, 6.52 (AB CH3O); CH2aN); system, 1H, J\15 Hz, CHxCHCO); 7.07 (s, 1H, H5im); 7.45 (s, 1H, H2im); 7.55 (AB system, 1H, J\15 Hz, CHxCHCO). 13C NMR d) : 168.14 (COO); 138.47 (C2im); 138.35 (CDCl3 , (C4im); 136.44 (CHxCHCO); 121.48 (C5im); 115.38 (CHxCHCO); 51.45 47.37 31.91 (CH3O); (CH2aN); 30.95»22.69 Mass spectrum (DCI, (CH2bN); (CH2).NH3) : m/z\321, MH` (100%). UV nm; (CHCl3) : kmax\290 e \19413 l mol~1 cm~1. Anal. calcd (%) for C19H32N2O2 , C 71.21 ; H 10.06 ; N 8.74. Found C 71.47 ; H 10.29 ; N 8.65. A solution of methyl ester (2.1 g, 6.55 mmol) and UAC12 (9.05 g, 65.5 mmol) in a methanol»water mixture (40 K2CO3 ml : 20 ml) was heated at 40 °C for 24 h.After methanol evaporation, the solution was acidi–ed (pH 5) and UAC12 obtained by precipitation in 75% yield. Mp\175 °C. 1H NMR [250 MHz, d] : 0.85 (t, J\6.5 Hz, 3H, (CD3)2SO, 1.23 (m, 18H, 1.77 (m, 2H, 4.12 (t, CH3) ; CH2) ; CH2bN); J\7.1 Hz, 2H, 6.67 (AB system, J\16 Hz, CH2aN); CHxCHCO); 7.45 (AB system, J\16 Hz, CHxCHCO); 8.04 (s, 1H, H5im); 9.02 (s, 1H, H2im). 13C NMR [250 MHz, d] : 166.81 (COOH); 130.37 (C4im); 129.92 (CD3)2SO, (CHxCHCO); 123.21 (C5im); 120.77 (CHxCHCO); 48.27 31.19 29.28»21.99 13.85 (CH2aN); (CH2bN); (CH2) ; (CH3). Mass spectrum (DCI, m/z\307, MH` (100%). IR CH4) : t(cm~1) : 3520 (OH); 3100 (xCH); 2930 2860 (CH2) ; (CH3) ; 1680 (CxO); 1650 (CxC).UV (NaOH solution, pH\13.2) : nm, e\13391 l mol~1 cm~1. Anal. calcd (%) for kmax\280 C 63.13 ; H 10.01 ; N 8.18. Found C C18H30O2N2 … 2H2O, 62.95 ; H 9.66 ; N 8.10. Synthesis of the bolaamphiphiles (5), UAC8UA UAC12UA (6) and (7). 1,16-dibromohexadecane. To a stirred UAC16UA solution of N-bromosuccinimide (2.86 g, 16.1 mmol) in 50 ml of anhydrous THF at 0 °C, was added dropwise a solution of triphenylphosphine (4.22 g, 16.1 mmol) in 50 ml of anhydrous THF.After reaching room temperature a solution of hexadecane-1,16-diol (1.04 g, 4.02 mmol) in THF was added dropwise. The mixture was heated at 55 °C for an additional 2.5 h. The solvent was evaporated under vacuum. Water was added to the residue and the solution extracted with diethyl ether. The organic layer was washed with water, dried with and evaporated under reduced pressure. Silica gel MgSO4 chromatography of the resulting solid (eluent : heptane) gave 1,16-dibromohexadecane (75% yield) as a white solid.Mp\57.3 °C. eluent : heptane. 1H NMR Rf\0.60 ; (CDCl3 , d) : 1.26 (m, 24H, 1.85 (quint, J\7 Hz, 4H, CH2) ; CH2bBr) ; 3.40 (t, J\7 Hz, 4H, 13C NMR d) : 34.10 CH2aBr).(CDCl3 , 32.86 29.63»28.20 Mass spec- (CH2aBr) ; (CH2bBr) ; (CH2). trum (DCI, m/z\383, MH` (25%); 303, MH`[Br CH4) : (100%). IR t(cm~1) : 2841»2857 Anal. calcd (%) for (CH2). C 50.02 ; H 8.39. Found C 50.46 ; H 8.48. C16H32Br2 , Bolaamphiphile methyl esters. A solution of dibromoalkane (4.775 mmol) in 10 ml of anhydrous THF was added dropwise New J. Chem., 1998, Pages 225»231 229to a mixture of 1 (1.45 g, 9.55 mmol), (13.3 g, 95.5 K2CO3 mmol) and 18-crown-6 (0.25 g, 0.955 mmol) in 25 ml of anhydrous THF.The mixture was stirred for –ve days at 60 °C. After –ltration of the remaining the THF was K2CO3, vacuum-evaporated. The resulting product was puri–ed by —ash chromatography (eluent : dichloromethane»ethanol, 99 : 1 v : v) and then recrystallised from ethyl acetate»acetone. methyl ester.Yield\52%. Mp\138 °C. UAC8UA Rf\ eluent : (90 : 10). 1H NMR d) : 0.67 ; CH2Cl2»EtOH (CDCl3, 1.20 (m, 8H, 1.69 (quint, J\7 Hz, 4H, 3.69 (s, CH2) ; CH2bN); 6H, 3.83 (t, J\7 Hz, 4H, 6.46 (AB system, CH3) ; CH2aN); J\15.6 Hz, 2H, CHxCHCO); 7.02 (s, 2H, H5im); 7.38 (s, 2H, H2im); 7.47 (AB system, J\15.6 Hz, 2H, CHxCHCO). 13C NMR d) : 168.23 (COO); 138.45 (C4im); 136.52 (CDCl3 , (CHxCHCO); 121.62 (C5im); 115.51 (CHxCHCO); 51.60 47.39 30.97 28.97»26.48 (CH3O); (CH2aN); (CH2bN); (CH2).Mass spectrum (DCI, m/z\415, MH` (100%). IR NH3) : t(cm~1) : 2877»2809 1695 (CxO); 1635 (CxC). UV (CH2) ; (EtOH): nm, e\44560 l mol~1 cm~1. Anal. calcd kmax\290 (%) for C 63.75 ; H 7.3 ; N 13.52. Found C C22H30O4N4 , 63.33 ; H 7.73 ; N 13.35.methyl ester. Yield\55%. Mp\141 °C. UAC12UA Rf\ eluent : (90 : 10). 1H NMR d) : 0.45 ; CH2Cl2»EtOH (CDCl3 , 1.23 (m, 16H, 1.74 (quint, J\6.9 Hz, 4H, CH2) ; CH2bN); 3.74 (s, 6H, 3.88 (t, J\6.9 Hz, 4H, 6.52 (AB CH3) ; CH2aN); system, J\15.6 Hz, 2H, CHxCHCO); 7.06 (s, 2H, H5im); 7.45 (s, 2H, H2im); 7.52 (AB system, J\15.6 Hz, 2H, CHxCHCO). 13C NMR d) : 168.32 (COO); 138.9 (CDCl3 , (C4im); 136.45 (CHxCHCO); 121.62 (C5im); 115.65 (CHxCHCO); 51.67 47.53 31.06 (CH3) ; (CH2aN); (CH2bN); 29.50»26.61 Mass spectrum (FAB[0, NBA matrix) : (CH2).m/z\471, MH` (100%). IR t(cm~1) : 3125 (CHx) ; 2941» 2857 1700 (CxO); 1635 (CxC). UV (EtOH): (CH2, CH3 ) ; nm, e\44448 l mol~1 cm~1. Anal. calcd (%) for kmax\290 C 61.88 ; H 7.99 ; N 11.10. Found, C C26H38O4N4 , … 2H2O, 62.14 ; H 7.73 ; N 10.97.methyl ester. Yield\65%. Mp\153 °C. UAC16UA Rf\ eluent : (90 : 10). 1H NMR d) : 0.35 ; CH2Cl2»EtOH (CDCl3 , 1.22 (m, 24H, 1.72 (quint, J\7.0 Hz, 4H, CH2) ; CH2bN); 3.73 (s, 6H, 3.87 (t, J\7.0 Hz, 4H, 6.51 (AB CH3) ; CH2aN); system, J\15.7 Hz, 2H, CHxCHCO); 7.06 (s, 2H, H5im); 7.44 (s, 2H, H2im); 7.52 (AB system, J\15.7 Hz, 2H, CHxCHCO). 13C NMR d) : 168.32 (COO); 138.46 (CDCl3 , (C4im); 136.54 (CHxCHCO); 121.64 (C5im); 115.53 (CHxCHCO); 51.64 47.51 31.08 (CH3O); (CH2aN); 29.75»26.63 Mass spectrum (DCI, (CH2bN); (CH2).NH3) : m/z\527, MH` (100%). IR m(cm~1) : 3100 (CHx) ; 2910» 2840 1685 (CxO); 1630 UV (EtOH): (CH2, CH3) ; (CxCconj). nm, e\42879 l mol~1 cm~1. Anal. calcd (%) for kmax\290 C 67.26 ; H 8.84 ; N 10.46. Found C C30H46N4O4 … 1/2 H2O, 67.39 ; H 8.84 ; N 10.33.Bolaamphiphiles and A UAC8UA, UAC12UA UAC16UA. solution of bolaamphiphile methyl ester (2.23 mmol) and (3.08 g, 22.3 mmol) in a methanol»water mixture (40 K2CO3 ml : 20 ml) was heated at 40 °C for 24 h. After vacuum evaporation of the methanol, the solution was acidi–ed by 6 N HCl (pH\5.7) and deprotected bolaforms precipitated. Bolaamphiphile 5, Yield\80%.White solid, UAC8UA. mp\247 °C. 1H NMR [400 MHz, d] : 1.21 (m, (CD3)2SO, 8H, 1.69 (quint, J\7.1 Hz, 4H, 3.94 (t, CH2) ; CH2bN); J\7.1 Hz, 4H, 6.28 (AB system, J\15.6 Hz, 2H, CH2aN); CHxCHCO); 7.41 (AB system, J\15.6 Hz, 2H, CHxCHCO); 7.57 (s, 2H, H5im); 7.72 (s, 2H, H2im); 12 (s, 2H, COOH). 13C NMR [400 MHz, d] : 167.87 (CD3)2SO, (COOH); 136.85 (C4im); 136.60 (CHxCHCO); 122.67 (C5im); 114.88 (CHxCHCO); 46.17 30.13 (CH2aN); 28.15»25.64 Mass spectrum (DCI, (CH2bN); (CH2).NH3) : m/z\387, MH` (100%). IR t(cm~1) : 3448 (OH); 3100»3070 (xCH); 2940 2860 1685 (CxO); 1640 (CxC). (CH2) ; (CH3) ; UV (NaOH solution, pH\13.2) : nm, e\42328 l kmax\280 mol~1 cm~1. Anal. calcd (%) for C C20H27O4N4Cl … 1/2H2O, 55.6 ; H 6.53 ; N 12.97. Found C 55.90 ; H 6.94 ; N 12.97.Bolaamphiphile 6, Yield\70%. White solid, UAC12UA. mp\227 °C. 1H NMR [400 MHz, d] : 1.21 (m, (CD3)2SO, 16H, 1.70 (quint, J\7.0 Hz, 4H, 3.94 (t, CH2) ; CH2bN); J\7.0 Hz, 4H, 6.27 (AB system, J\15.5 Hz, 2H, CH2aN); CHxCHCO); 7.40 (AB, J\15.5 Hz, 2H, CHxCHCO); 7.57 (s, 2H, H5im); 7.70 (s, 2H, H2im). 13C NMR [400 MHz, d] : 167.92 (COOH); 136.98 (C4im); 136.63 (CD3)2SO, (CHxCHCO); 122.62 (C5im); 114.88 (CHxCHCO); 46.15 30.18 28.72»25.72 Mass spectrum (CH2aN); (CH2bN); (CH2). (DCI, m/z\443, MH` (100%).IR t(cm~1) : 3571 NH3) : (OH); 3125 (CHx) ; 2940 2860 1681 (CxO); (CH2) ; (CH3) ; 1653 (CxC). UV (NaOH solution, pH\13.2) : kmax\280 nm; e\40533 l mol~1 cm~1. Anal. calcd (%) for C 62.59 ; H 7.88 ; N 12.16. Found C 62.22 ; H C24H36O5N4 , 7.72 ; N 11.80. Bolaamphiphile 7, Yield\70%. White solid, UAC16UA.mp\201 °C. 1H NMR [400 MHz, d] : 1.20 (m, 24, (CD3)2SO, 1.72 (quint, J\7.0 Hz, 4H, 4.02 (t, J\7.0 CH2) ; CH2bN); Hz, 4H, 6.47 (AB system, J\15.8 Hz, 2H, CH2aN); CHxCHCO); 7.42 (AB, J\15.8 Hz, 2H, CHxCHCO); 7.78 (s, 2H, H5im); 8.30 (s, 2H, H2im). 13C NMR [400 MHz, d] : 167.39 (COOH); 134.01 (C4im); 133.63 (CD3)2SO, (CHxCHCO); 122.86 (C5im); 117.49 (CHxCHCO); 47.09 29.80 28.91»25.60 Mass spectrum (CH2aN); (CH2bN); (CH2).(DCI, m/z\499, MH` (100%). IR t(cm~1) : 3400 NH3) : (OH); 3100 (xCH); 2910 2840 1690 (CxO); (CH2) ; (CH3) ; 1635 (CxC). UV (NaOH solution, pH\13.2) : kmax\280 nm, e\38530 l mol~1 cm~1. Anal. calcd (%) for C 55.35 ; H 7.96 ; N 9.22. Found C 55.88 ; H C28H48O6N4Cl2 , 7.68 ; N 9.14. Molecular aggregation of the bolaforms in aqueous solution Surface tension measurements were made with a Prolabo Tensiomat 3 (equipped with a thermostated water bath; T \30 °C) using the stirrup detachment method.Conductivity measurements were made at 30 °C with a conductivity» resistivity meter CDRV62 (Tacussel Electronique) with a platinium electrode. UV absorbance measurements were made with a Hewlett Packard 8452A spectrometer.A Coulter N4MD was used for the light scattering measurements. The electron micrographs were obtained using a EM-301 PHILIPS electron microscope. The samples were prepared by negative staining with uranyl acetate on a carbon –lm. Vesicles were prepared by the sonication method with a titanium probe (High Intensity Ultrasonic Processor 600-Watt Model).Solutions of 10~3 mol l~1 were sonicated at 0 °C for 15 min using an 80% duty cycle. Titanium (from the probe) and dust were removed by centrifugation (3000 rpm~1 for 10 min) and –ltration through a Millipore 0.45 l –lter. References 1 H. A. Morrison, Photodermatology, 1985, 2, 158. 2 N. K. Gibbs, M. Norval, N. Traynor, M. Wolf, B. E. Johnson and J. Crosby, Photochem.Photobiol., 1994, 60, 280. 3 W. Schwartz, K. Langer, H. Schell and A. Schonberger, Photodermatology, 1986, 3, 239. 4 A. Zenisek and J. A. Kral, Biochem. Biophys. Acta, 1953, 12, 479. 5 P. M. Krein and D. Moyal, Photochem. Photobiol., 1994, 60, 280. 6 J. H. Anglin Jr. and W. H. Batten, Photochem. Photobiol., 1970, 11, 271. 7 H. A. Morrison and R. M. Deibel, Photochem. Photobiol., 1988, 48, 153. 8 S. J. Farrow, C. R. Jones, D. L. Severance, R. M. Deibel, W. M. A. Baird and H. A. Morrison, J. Org. Chem., 1990, 55, 275. 9 E. C. De Fabo and F. P. Nooman, J. Exp. Med., 1983, 158, 84. 10 M. Norval, N. K. Gibbs and J. Gilmour, Photochem. Photobiol., 1995, 62, 209. 11 M. Norval, T. J. Simpson and J. A. Ross, Photochem. Photobiol., 1989, 50, 267. 230 New J. Chem., 1998, Pages 225»23112 K.Kinito, K. Punnonen, J. Toppain and L. Leino, In—ammation, 1996, 20, 451. 13 A. M. MoodycliÜe, C. D. Bucana, M. L. Kripke, M. Norval and S. E. Ullrich, J. Immunology, 1996, 157, 2891. 14 M. C. Monje, A. Lattes and M. Rivie` re, Bull. Soc. Chim. Fr, 1990, 127, 292. 15 J. H. Fuhrhop and J. Mathieu, Angew. Chem., Int. Ed. Engl., 1984, 96, 124. 16 R. Nagarajan, Chem.Eng. Commun., 1987, 55, 251. 17 F. M. Menger and S. Wrenn, J. Phys. Chem., 1974, 78, 1387. 18 S. Yiv, K. M. Kale, J. Lang and R. Zana, J. Phys. Chem., 1976, 80, 2651. 19 S. Yiv and R. Zana, J. Colloid Interface Sci., 1980, 77, 449. 20 R. Zana, S. Yiv and K. M. Kale, J. Colloid Interface Sci., 1980, 77, 456. 21 Y. Okahata and T. Kunitake, J. Am. Chem. Soc., 1979, 101, 5231. 22 J.H. Fuhrhop and D. Fritsch, Acc. Chem. Res., 1986, 19, 130. 23 Z. Mirghani, D. Bertoia, E. Piano and A. Gliozzi, Biochem. Soc. T rans., 1989, 17, 507. 24 J. H. Fuhrhop, H. H. David, J. Mathieu, U. Liman, H. J. Winter and E. Boekema, J. Am. Chem. Soc., 1986, 108, 1785. 25 J. H. Fuhrhop and R. Bach, Advances in Supramolecular Chemistry, 1992, 2, 25. 26 R. Garelli-Calvet, F. Brisset, I. Rico and A.Lattes, Synth. Commun., 1993, 23, 35. 27 F. Brisset, R. Garelli-Calvet, J. Azeç ma, C. Chebli, I. Rico-Lattes and A. Lattes, New J. Chem., 1996, 20, 595. 28 C. Madelaine-Dupuich, B. Guidetti, C. Rico-Lattes and A. Lattes, New J. Chem., 1996, 20, 143. 29 P. Boé hme, H. G. Hicke, C. Boettcher and J. H. Fuhrhop, J. Am. Chem. Soc., 1995, 117, 5824. 30 J. A. McDonald, S.A. Butler and A. R. Rennie, Colloids Surf. A, 1995, 102, 137. 31 M. Jorgensen and K. Bechgaard, J. Org. Chem., 1994, 59, 5877. 32 G. R. Newkome, C. N. Moore–eld, G. R. Baker, R. K. Behera G. H. Escamillia and M. J. Saunders, Angew. Chem., Int. Ed. Engl., 1992, 31, 917. 33 M. Ishikawa, K. I. Matsumura, K. Esumi, K. Meguro, W. Binana- Limbele and R. Zana, J. Colloid Interface Sci., 1991, 151, 70. 34 N. Jayasuriya, S. Bosak and S. L. Regen, J. Am. Chem. Soc., 1990, 112, 5844. 35 D. Tsiourvas, C. M. Paleos and A. Malliaris, Progr. Colloid Polym. Sci., 1994, 97, 163. 36 E. K. Comeau, E. J. Beck, J. F. Caplan, C. V. Howley and D. G. Marangoni, Can. J. Chem., 1995, 73, 1741. 37 R. Festag, V. Hessel, P. Lehmann, H. Ringsdorf and J. H. WendorÜ, Recl. T rav. Chim. Pays Bas, 1994, 113, 222. 38 G. Mao, Y. Tsao, M. Tirrell and H. T. Davis, L angmuir, 1993, 9, 3461. 39 A. Kokkinia, C. M. Paleos, A. Malliaris and A. Xenakis, Progr. Colloid Polym. Sci, 1993, 93, 302. 40 R. Skoé ld and M. A. R. Tunius, J. Colloid Interface Sci., 1992, 152, 183. 41 J. H. Fuhrhop, D. Spiroski and C. Boettcher, J. Am. Chem. Soc., 1993, 115, 1600. 42 N. Lauth-de Viguerie, N. Sergueeva, M. Damiot, H. Mawlawi, M. Rivie` re and A. Lattes, Heterocycles, 1994, 37, 1561. 43 K. Meguro, K. Ikeda, A. Otsuji, M. Taya, M. Yasuda and K. Esumi, J. Colloid Interface Sci., 1987, 118, 372. 44 J. A. McDonald J. A., Butler S. A. and A. R. Rennie, Colloids Surf. A: Physicochem. Eng. Aspects, 1995, 102, 137. 45 F. Menger and K. D. Gabrielson, Angew. Chem., Int. Ed. Engl., 1995, 34, 2091. 46 H. A. Ho—and, J. A. Bouwtra, G. S. Gooris, F. Spies, H. Talsma and H. E. Junginger, J. Interface Sci., 1993, 161, 366. Colloïç d 47 J. H. Fuhrhop, U. Liman and V. Koesling, J. Am. Chem. Soc., 1988, 110, 6840. 48 D. D. Lasic, Biochem. J.,1988, 256, 1. 49 H. H. Hub, U. Zimmerman and H. Ringsdorf, FEBS L ett., 1982, 140, 254. 50 R. W. de Groot, A. Wagenaar, A. Sein and J. B. F. N. Engberts, Recl. T rav. Chim. Pays Bas, 1995, 114, 371. 51 Jaeger D. A. and Brown E. L. G., L angmuir, 1996, 12, 1976. 52 D. P. Cistola, J. A. Hamilton, D. Jackson and D. M. Small, Biochemistry, 1988, 27, 1881. 53 K. Edwards, M. Silvander and G. Karlsson, L angmuir, 1995, 11, 2429. Received 28th May 1997; Paper 7/08326B New J. Chem., 1998, Pages 225»231 231

ISSN:1144-0546    

DOI:10.1039/a708326b

出版商:RSC    

年代:1998

数据来源: RSC

摘要:

The poisoning eÜect of mercury complexes with an anionic exchange membrane used in an electrodialysis process : a Raman study Majid Chaouki, Patrice Huguet, Persin and Jean-Luc Bribes* Franc” oise L aboratoire des et des Membranaires, UMR 5635), Pl. E. Bataillon, Mateç riaux Proceç deç s (CNRS Montpellier II, 34095 Montpellier 5, France Universiteç ceç dex Poisoning of an ion exchange membrane is one of the most important problems encountered in the electrodialysis process.The poisoning of an anion exchange membrane, used in an electrodialysis process to purify and reconcentrate an hydrochloric acid solution containing mercury chloride, has been investigated. Raman spectroscopic analysis of the contaminated material, combined with electrodialysis results, are presented and discussed.The stoichiometry of the poisoning mercury complex and its action as a counter ion inside the membrane are determined. Moreover, a strong interaction between the membrane –xed sites and the mercury complex is pointed out. Etude par spectromeç trie Raman de lœempoisonnement dœune membrane e ç changeuse dœanions, par des complexes mercuriques, au cours dœun proceç deç dœeç lectrodialyse.Lœempoisonnement dœune membrane eç changeuse dœions lorsquœelle est utiliseç e dans un proceç deç dœeç lectrodialyse constitue un handicap majeur. Nous avons eç tudieç lœempoisonnement de la membrane prototype AW11 utiliseç e dans un proceç deç dœeç lectrodialyse destineç a` puri–er et a` reconcentrer de lœacide chlorhydrique a` partir dœune solution aqueuse de ce dernier contenant une faible teneur en chlorure mercurique.Les reç sultats obtenus au cours de lœeç lectrodialyse associeç s a` lœeç tude spectrale de la membrane contamineç e, en diÜusion Raman, sont preç senteç s et discuteç s. Nous avons pu mettre en eç vidence la stoechiomeç trie et le ro� le joueç par le complexe mercurique majoritaire constituant, ici, lœagent empoisonnant.En outre nous montrons aussi que ce complexe entre en interaction forte avec les sites –xes de la membrane contamineç e. Electrodialysis is a powerful technique for the treatment of many waste acids. However, some important problems can be encountered if poisoning of the ionic exchange membrane occurs. This is the case in the puri–cation and concentration process of hydrochloric acid solutions containing zinc, cadmium or mercury chlorides, because metallic complex formation gives rise to an important poisoning eÜect of the membranes used.Our group has studied ionic —uxes across anion exchange membranes and it appears that a considerable loss of permselectivity for the sodium ion takes place in the presence of zinc complexes.1 So we have undertaken an electrodialysis experiment, coupled with a Raman investigation, to obtain information on the poisoning of an anionic exchange membrane by metallic complexes.We have already emphasized the important advantages of Raman spectroscopy in the study of immersed ion exchange membranes in aqueous solutions. By this useful technique we have obtained valuable information on ion exchange membranes.2h8 Here, we have chosen to work with a hydrochloric acid solution containing mercury chloride for the following reasons : (i) mercury chloride complexes are more stable than the zinc or cadmium ones and (ii) there are only anionic complexes with mercury.Experimental Electrodialysis experiment The electrodialysis operation is shown schematically in Fig. 1. Compartment 2 (0.3 M NaCl) and compartment 4 (0.1 M HCl) correspond to the concentrated stream whereas compartment 3 (0.3 M HCl]10~3 M correspond to the HgCl2) diluent stream.Compartments 1 (0.25 M and 5 (0.5 M H2SO4) HCl) are used to rinse oÜ the electrodes. * Fax: (33) 4 67 14 45 53; E-mail: bribes=crit.univ-montp2.fr In compartment 3 is indicated as it is preponder- HgCl42~ ant compared to and taking into account the HgCl3~ HgCl2 , given composition. The apparatus is made of polytri—uorochloroethylene (KEL-F) and each compartment is 12 mm thick.In these compartments the stream speed is about 5.8 cm s~1. The initial volume of the diluent stream is equal to 5 l whereas those of the concentrated streams are equal to 0.5 l. The membrane pieces used in the process have an area of 40 cm2 and the applied electric current density is equal to 0.05 A cm~2.The electrodes are made of platinised titanium and are –xed with nylon screws. Membranes The cationic exchange membranes (CEM) used in the experiment are the CMV selemion membrane from Asahi Glass whereas the anionic exchange membranes (AEM) are the prototype membranes AW11 produced by Solvay. The AW membranes are composed of poly(4-vinylpyridinium) chains (in acidic solution) grafted onto an ETFE (ethylenetri—uoroethylene) polymer matrix and crosslinked with divinylbenzene.This kind of membrane has a low proton leakage and is therefore very well suited for acid reconcentration. Fig. 1 Diagram of the electrodialysis process experiment New J. Chem., 1998, Pages 233»235 233Raman measurements Raman spectra were obtained by excitation with 514.5 nm radiation from an argon ion laser (Spectra Physics model 2020-03) operated at about 300 mW.The spectra were recorded with an OMARS 89 multichannel spectrometer (Dilor, France). The detector is an intensi–ed 1024 photodiode array. The membrane was kept —at between two optical glass plates that were pressed together.Two slots in the plates facing each other allowed the Raman spectrum of the membrane to be recorded without that of the plate material. The glass plates and the membrane can be immersed in various aqueous solutions contained in an optical cell (Hellma). The laser beam was focused onto the edge of the sample and the scattered light was collected at right angles to this beam.The Raman spectra were recorded at 23 °C. Results and Discussion Electrodialysis In Fig. 2A the acid concentration versus time is given for compartments 2 and 4. One can see that a greater increase in acid concentration is observed in compartment 4. Moreover, in Fig. 2B the acid concentration yield has been plotted versus time and a signi–cant diÜerence occurs between these two compartments. The acid concentration yield is about 20% higher in compartment 4 than in compartment 2, whereas the concentration yield is the same in these two compartments when the electrodialysis is carried out under the same conditions without any Finally, Fig. 2C shows that mercury HgCl2 . species are present in compartment 2 while we do not detect this element in compartment 4.Obviously, AEM 2/3 (see Fig. 1) is aÜected by mercury species in contrast to AEM 4/5. The low proton leakage of AEM 2/3 is certainly lost by the poisoning eÜect due to the presence of a mercury chloride complex inside this material. Raman spectra First we have measured Raman spectra of aqueous solutions containing NaCl and at various compositions. The HgCl2 molar fractions of mercury complexes depend on the free chlo- Fig. 2 Main results of the electrodialysis in compartments 2 and (+) 4 (=) ride concentration, the ionic strength of the solution and the diÜerent equilibrium constants of formation.9 So, we have calculated the percentage of and as a HgCl2, HgCl3~ HgCl42~ function of the free ligand concentration10 (see Fig. 3A). We cannot obtain an aqueous solution giving spectra where the bands of are preponderant whereas this is possible for HgCl3~ and However, it is to be noted that in these HgCl42~ HgCl2 .spectra each complex gives rise only to the most intense and characteristic band (i.e., the totally symmetric stretch of m1 HgwCl) as can be seen in Fig. 3B. The observed wavenumbers are in agreement with the work of Waters and coworkers,11 although their spectra were measured in tributylphohate extracts from concentrated aqueous solutions. We have then recorded Raman spectra of the AW11 membrane immersed in water, before and after the electrodialysis operation.The stong band located at 272 cm~1 m1 (Fig. 4A, curve a), which is absent in Fig. 4A, curve b, indicates essentially the presence of inside the anionic HgCl42~ membrane used in the process. Furthermore, while keeping this last membrane immersed in water for several days and recording Raman spectra from time to time we have noticed that the complex remains in this material (see Fig.HgCl42~ 4B). It thus appears that the complex is acting as a HgCl42~ counter ion of the poly(4-vinylpyridinium) –xed sites of the membrane, as we cannot detect co-ions inside this material under these conditions (Donnanœs rule).Moreover, as the concentration of the free chloride ions is very low, it is evident that the stability of the mercury complex involves a strong electrostatic interaction with the –xed sites. In Fig. 4B and Fig. 4C two bands are located by arrows and labeled Pyr and Pyr` since these last two bands are characteristic of the poly(4-vinylpyridine) and poly(4-vinylpyridinium) grafts.4 When the membrane is immersed in water (coming from an acidic solution) deprotonation of the grafts Fig. 3 (A) Percentages of mercury complexes as a function of pKL\ where L is the free ligand concentration. (B) Raman spectra [log(L) of aqueous solutions for diÜerent values, with an (NaCl]HgCl2) pKL ionic strength equal to 1.The intensities are on an arbitrary scale 234 New J. Chem., 1998, Pages 233»235Fig. 4 (A) Raman spectra of the AW11 membrane immersed in water: curve a after 6 h of electrodialysis ; curve b before electrodialysis. (B) Raman spectra versus time of the AW11 membrane containing (C) Raman spectra versus time of the AW11 HgCl42~. membrane without The intensities are on arbitrary scales HgCl42~.occurs with time and the intensity of the Pyr band increases whereas it decreases for the Pyr` band. However, the rate of variation of the intensities with time is very diÜerent between a clean membrane and one containing the mercury complex. If we compare the spectra of Fig. 4B with those of Fig. 4C, it can be seen that the deprotonation rate is much lower when the mercury complex is inside the membrane. Our previous results4 concerning the deprotonation kinetics for this kind of membrane have indicated that the reaction half-life increases markedly when the membrane grafting rate is reduced.Thus, it can be deduced that there are fewer accessible sites in the presence of mercury complex ions. This provides con–rmation of the presence of the mercury complex as counter ions inside the membrane and that some –xed sites continue to be neutralized by chloride ions.Fixed sites neutralized by HgCl42~ ions are then protected against deprotonation by the hydroxide ions present in water. Finally, although some of the initial mercury complex is lost, there is always some present in the used membrane, even after an immersion time as long as 450 days.It is also to be noted that is always involved as the counter ion HgCl42~ because –xed poly(4-vinylpyridinium) sites appear in the Raman spectrum, whereas no more –xed sites are present for a clean membrane immersed in water for the same time. References 1 F. Aouad, A. Lindheimer and C. Gavach, J. Membrane Sci., 1997, 123, 207. 2 M. El Boukari, J. L. Bribes and J. Maillols, J. Raman Spectrosc., 1990, 21, 755. 3 J. L. Bribes, M. El Boukari and J. Maillols, J. Raman Spectrosc., 1991, 22, 275. 4 J. L. Bribes, A. Hasdou and J. Maillols, J. Raman Spectrosc., 1993, 24, 519. 5 I. Tugas, J. M. Lambert, J. Maillols, J. L. Bribes, G. Pourcelly and C. Gavach, J. Membrane Sci., 1993, 78, 25. 6 R. Delimi, J. Maillols, R. Sandeaux, J. Sandeaux and J. L. Bribes, J. Raman Spectrosc., 1995, 26, 313. 7 A. Hasdou, G. Pourcelly, P. Huguet, J. L. Bribes, J. Sandeaux and C. Gavach, New J. Chem., 1996, 20, 515. 8 M. Chaouki, P. Huguet and J. L. Bribes, J. Mol. Struct., 1996, 379, 219. 9 A. E. Martell and R. M. Smith, Critical Stability Constants, V ol. 4 Inorganic Complexes, Plenum, New York and London, 1976. 10 J. Y. Gal and M. Persin, Exercices et programmes en analyse des chimiques en solution, Dunod Universiteç , Bordas, 1986. reç actions 11 D. N. Waters, E. L. Short, M. Tharwat and D. F. C. Morris, J. Mol. Struct., 1973, 17, 389. Received 3rd June 1997; Paper 7/08336J New J. Chem., 1998, Pages 233»235 235

ISSN:1144-0546    

DOI:10.1039/a708336j

出版商:RSC    

年代:1998

数据来源: RSC

摘要:

Electroreduction of and Pd2(dppm)2Cl2 Pd(dppm)Cl2 [dppm= bis(diphenylphosphino)methane ] in aprotic medium under carbon dioxide : electrogeneration of Pd3(l3-CO)(l-dppm)3 Isabelle Gauthron,a Yves Mugnier,*,a Karine Hiersob and Pierre D. Harvey*,b a L aboratoire de et (CNRS UMR 5632), Synthe` se dœEä lectrosynthe` se Organomeç talliques Faculteç des Sciences Gabriel, de Bourgogne, 21000 Dijon, France Universiteç de Chimie, de Sherbrooke, J1K 2R1 Sherbrooke, Canada b Deç partement Universiteç Queç bec, The electroreduction of the d9»d9 and monomeric d8 complexes in aprotic Pd2(dppm)2Cl2 Pd(dppm)Cl2 medium (such as DMF, THF and acetonitrile) has been performed under atmosphere.In all cases the –nal CO2 products are anion, CO and the neutral cluster. This electroreduction is not CO32~ Pd3(l3-CO)(l-dppm)3 catalytic but rather stoichiometric.The electroreduction mechanisms have been addressed experimentally by electrochemical methods and IR spectroscopy, and theoretically by density functional methods via the geometry optimizations of the proposed intermediates. The intermediates and ììPd(dppm)œœ are assumed to ììPd2(dppm)2œœ be active towards the binding of prior to its reduction.CO2 de et de [dppm= bis(dipheç nylphosphino)meç thane ] en milieu E ä lectroreç duction Pd2(dppm)2Cl2 Pd(dppm)Cl2 aprotique sous dioxyde de carbone: de Lœeç lectroreç duction du complexe E ä lectrogeç neç ration Pd3(l3-CO)(l-dppm)3 . d9-d9 et du monome` re d8 en milieu aprotique (DMF, THF et aceç tonitrile) a eç teç Pd2(dppm)2Cl2 Pd(dppm)Cl2 eÜectueç e sous atmosphe` re de Dans tous les cas les produits –naux sont lœanion CO et le cluster CO2.CO3 2~, neutre Cette eç lectroreç duction nœest pas catalytique mais stoechiomeç trique. Les Pd3(l3-CO)(l-dppm)3 . meç canismes dœeç lectroreç duction ont eç teç eç tudieç s expeç rimentalement par des meç thodes eç lectrochimiques et par spectroscopie IR, et theç oriquement par les meç thodes de la fonctionnelle de la densiteç via les optimisations geç ometriques des intermeç diaires proposeç s.Les intermeç diaires et ììPd(dppm)œœ sont supposeç s e� tre ììPd2(dppm)2œœ actifs vis a` vis de la –xation du avant sa reç duction. CO2 The transformation of carbon dioxide into useful chemical derivatives constitutes an attractive goal in the –eld of chemistry.In this way the electrochemical reduction of has CO2 been noted as one practical method.1 Its reduction into its anion radical requires quite a negative potential,2 and a dependence on both the nature of the electrode and the reaction medium has been observed in the products. Considerable eÜorts have been made to –nd catalysts that allow a substantial decrease of the reduction potential and provide CO2 product selectivity.3,4 The most extensively studied homogeneous catalysts so far are transition-metal complexes containing either macrocyclic or bipyridine ligands.5 Only a few reports have appeared describing the electrochemical reduction of using transition-metal phosphine com- CO2 plexes.6h10 Recently DuBois and coworkers showed that complexes can exhibit [Pd(triphosphine)(solvent)](BF4)2 high catalytic rates for the electrochemical reduction of at relatively positive potentials.5,8,10 More recently, CO2 these authors reported that the complex where eHTP is bis[bis(diethyl- Pd2(CH3CN)2(eHTP)(BF4)4 , phosphino)ethyl]phosphinomethane, catalyses the electrochemical reduction of to CO in acidic CO2 dimethylformamide solution.10d In this paper we report the electrochemical behavior of and Pd2(dppm)2Cl2 Pd(dppm)Cl2 in the presence of carbon dioxide [dppm\bis(diphenylphosphino) methane].During the course of this study, the geometry of some of the proposed key intermediates has been optimized using density functional theory, in order to obtain information regarding their structures and the mechanism of activation.CO2 Experimental Materials and have been prepared Pd2(dppm)2Cl2 11a Pd2(dppm)3 11b according to literature procedures. Pd(dppm)Cl2 : PdCl2 (Aldrich ; 0.52 g, 2.9]10~3 M) and dppm (Aldrich ; 1.14 g, 2.9]10~3 M) were suspended in 50 ml of ethanol (95%) and 50 ml of conc. HCl. The solution was heated to re—ux for a period of four hours. A white precipitate with a yellow solution was obtained.The solution was –ltered and the white solid was washed with 50 ml of water and 50 ml of ethanol. Yield 89%. The identity of the product was veri–ed by 1H NMR spectrometry and chemical analysis. IR measurements All IR measurements were performed on a Nicolet 205 spectrophotometer. The electrolysis solutions were transferred into an air-tight IR cell via canular techniques. No attempt to isolate the palladium species was made.Electrochemistry All manipulations were performed using standard Schlenk techniques in an atmosphere of dry, oxygen-free, nitrogen or argon gases. Tetrahydrofuran was distilled under argon from sodium and benzophenone. Acetonitrile (ACN) was puri–ed by simply passing the solvent through a column packed with alumina previously dried at 120 °C; it was deoxygenated by argon bubbling immediately before use. The supporting electrolytes were 0.2 M or LiCl, which were dried (Bun)4NPF6 and degassed before use.In cyclic voltammetry experiments, New J. Chem., 1998, Pages 237»246 237b c –1 –2 a ic ia E / V the concentration of or Pd2(dppm)2Cl2 , Pd(dppm)Cl2 was nearly 10~3 M. Voltammetry analyses were Pd2(dppm)3 carried out in a standard three-electrode cell with a Tacussel UAP4 unit cell. The reference electrode was a saturated calomel electrode separated from the solution by a sintered glass disk.The auxiliary electrode was a platinum wire. For all voltammetric measurements the working electrode was a vitreous carbon electrode. The controlled potential electrolyses were performed with an Amel 552 potentiostat coupled to an Amel 721 electronic integrator.High scale electrolysis was performed in a cell with three compartments separated with fritted glasses of medium porosity. A carbon gauze was used as the cathode, a platinum plate was used as the anode and a saturated calomel electrode was used as the reference electrode. For thin layer cyclic voltammetry the vitreous carbon electrode was lowered until contact with the —at —oor of the electrolytic scale cell so that only a thin layer of solution existed between them.Computational details The reported density functional calculations were all carried out utilizing the Amsterdam Density Functional (ADF) program developed by Baerends et al.12,13 and vectorized by Raveneck.14 The numerical integration procedure applied for the calculations was developed by te Velde and coworkers.15 The geometry optimization procedure was based on the method developed by Versluis and Ziegler.16 The electronic con–gurations of the molecular systems were described by an uncontracted double-f basis set17 on palladium for the 4s, 4p and 5s orbitals, and triple-f for the 5d ones.Double-f STO basis sets18 were used for phosphorous (3s, 3p), oxygen (2s, 2p), carbon (2s, 2p), and hydrogen (1s) orbitals, augmented with a single 4d polarization function for P, a single 3d one for O and C, and a 2p function for H.No polarization function was used for palladium. The 1s22s22p63s23d10 con–guration on palladium, the 1s22s2 con–guration on phosphorous, and the 1s2 con–gurations on oxygen and carbon were treated by the frozen-core approximation.13 A set of auxiliary19 s, p, d, f and g STO functions, centered on all nuclei, was used in order to –t the molecular density and present Coulomb and exchange potentials accurately in each SCF cycle.Energy diÜerences were calculated by including the local exchange-correlati potential of Vosko et al.20 No nonlocal exchange or correlation corrections were made for the geometry optimizations.Results and Discussion Electrochemistry We have examined the electrochemical behavior of 1 and 2 in aprotic medium. The Pd2(dppm)2Cl2 Pd(dppm)Cl2 thin layer cyclic voltammogram of 1 at 0.02 V s~1 in tetrahydrofuran (THF) solution containing 0.2 M as (Bun)4NPF6 supporting electrolyte on a vitreous carbon electrode and at room temperature exhibits a strong reduction peak (D) around [1.2 V vs.SCE. The return potential scan reduction peak D exhibits two oxidation peaks and [D, O1 O2 Ep\ V; V; V vs. saturated [1.23 O1, Ep\[0.47 O2 , Ep\[0.16 calomel electrode (SCE); see Table 1, entry 1]. In diÜusion, at 0.1 V s~1, a similar cyclic voltammogram is obtained except that appears as a shoulder.and correspond to the O2 O1 O2 oxidation of Pd0 complexes. In particular is the oxidation O2 peak of the known complex 4, based upon com- Pd2(dppm)3 parison with an authentic sample. Peak which is situated at a lower anodic potential than O1, peak probably corresponds to the oxidation of an anionic O2 , Pd0 complex 3 formulated as Amatore et Pd2(dppm)2Clx x~.al. showed that anionic Pd0 species such as can be obtained by electroreduction of [Pd0(PPh3)2Clx]n(nx)~ in the absence of To obtain further PdCl2(PPh3)2 PPh3 .21 proof of the formation of 3 we have examined the electrochemical behavior of 1 in DMF containing 0.2 M LiCl as supporting electrolyte salt. The cyclic voltammogram of 1 exhibits peaks D and only a shoulder appears at the O1 ; potential of peak (see Table 1, entry 7).This result indi- O2 cates that an anionic species formulated as Pd2(dppm)2Clx x~ is relatively stable on the cyclic voltammetry timescale ; but under these conditions when an electrolysis is performed at the potential of peak D at 0 °C (see Table 2, entry 1), the two oxidation peaks and are observed from the electrolyzed O1 O2 solution. Increasing the temperature from 0 °C to 20 °C causes peak to decrease and peak to increase (Fig. 1). O1 O2 The above results are in accordance with the following two reactions : Pd2(dppm)2Cl2 1 ]2e~]Pd2(dppm)2Clx x~ 3 ](2[x) Cl~ (1) 2 Pd2(dppm)2Clx x~ 3 ]Pd2(dppm)3 4 ]ììPd2(dppm)Clx x~œœ]x Cl~ (2) The two-electron reduction of 1 gives the anionic species 3, which evolves into a mixture of 4 and an unsaturated derivative of Pd0 formulated as This latter is ììPd2(dppm)Clx x~œœ.found to be very unstable within the timescale of the cyclic voltammogram [i.e. decomposes]. Further- ììPd2(dppm)Clx x~œœ more the addition at 0 °C of dppm to the electrolyzed solution containing 3 and 4 causes peak to disappear completely O1 and peak to appear according to the reaction (3) : O2 Pd2(dppm)2Clx x~ 3 ]dppm]Pd2(dppm)3 4 ]x Cl~ (3) As mentioned above, the formulated species 3 is relatively unstable on the timescale of the electrolysis so that no NMR spectroscopy experiment was possible.In DMF containing or LiCl as supporting elec- (Bun)4NPF6 trolyte the cyclic voltammogram of 1 is not modi–ed in the presence of added Br~ ion (as showing either that Bu4NBr), the halogen-exchange reaction from 3 does not occur, or that the exchange occurs but the voltammograms of the chloro and bromo derivatives are the same.If is bubbled through the solution, a well de–ned CO2 system appears in thin layer voltammetry after several A1/A@1 Fig. 1 Cyclic voltammogram of 1 in DMF/0.2 M LiCl solution : a initial voltammogram; b after two-electron reduction of 1 at [1.3 V at 0 °C; c after evolution at room temperature.Starting potential : 0 V for a, [1.3 V for b and c ; sweep rate : 0.2 V s~1 238 New J. Chem., 1998, Pages 237»246Table 1 Electrochemical data for palladium complexes Solvent/ Sweep rate/ Oxidation Reduction Entry Complex electrolyte salt V s~1 peak/Va peak/Va 1 1 THF/Bu4NPF6 0.02b O1[0.47 D [1.23 O2[0.16 2 1 THF/Bu4NPF6 c 0.02b O1[0.50 D [1.21 O2[0.16 A@1[0.59 A1[0.79 3 1 ACN/Bu4NPF6 0.02b O1[0.35 D [0.89 O2[0.16 4 1 ACN/Bu4NPF6 c 0.02b O1[0.35 D [0.89 O2[0.15 A@1[0.65 A1[0.70 5 1 DMF/Bu4NPF6 0.02b O1[0.53 D [1.18 O2[0.14 6 1 DMF/Bu4NPF6 c 0.02b O1[0.52 D [1.20 O2[0.24 A@1[0.59 A1[0.72 7 1 DMF/LiCl 0.2 O1[0.50 D [1.32 O2[0.20 8 7 ACN/Bu4NPF6 c 0.1 B@ [1.40 B [1.56 9 7 THF/Bu4NPF6 c 0.02b B@1[1.18 B1[1.29 B@2[1.39 B2 [1.52 A@1[0.59 A1[0.76 10 4 THF/Bu4NPF6 0.1 O2[0.12 R [0.48 11 4 DMF/Bu4NPF6 0.1 O2[0.10 R [0.48 12 2 THF/Bu4NPF6 0.02b O@1[0.58 R1 [1.23 O@2[0.23 O@3]0.04 13 2 THF/Bu4NPF6 c 0.02b O@1[0.58 R1[1.19 O@2[0.22 O@3]0.04 A1[0.76 A@1[0.60 14 2 ACN/Bu4NPF6 0.02b O@1[0.73 R1[1.00 O@2[0.34 O@3]0.03 15 2 ACN/Bu4NPF6 c 0.02b O@1[0.73 R1[1.00 O@2[0.34 O@3]0.03 A1[0.71 A@1[0.66 16 13 ACN/Bu4NPF6 c 0.02b A@1[0.68 B*[1.60 a Vitreous carbon electrode is used as working electrode. b In thin layer voltammetry.c Under carbon dioxide. potential scans (Fig. 2), while the intensities of the oxidation peaks and and the reduction peak D decrease O1 O2 (A1, V; V vs. SCE; see Table 1, entry Ep\[0.79 A@1, Ep\[0.59 2; the ratio of the system remains unity).The ipA{1 /ipA1 A1/A@1 system is indicative of the presence of Pd3(l3-CO)(l-dppm)3 5.22 Under these conditions, the bulk electrolysis of 1 at [1.6 V, under at room temperature, leads to solution colour CO2 changes from orange to brown after consumption of over four equivalents of electrons F mol~1; see Table 2, (nexp\4.17 entry 2). In cyclic voltammetry of the resulting solution, the oxidation peaks and appear during the anodic scan.If A@1 O2 the potential scan is reversed after peak two reduction O2 , peaks and D are observed. Peak D is found only if the A1 potential is reversed after peak We have noted that the O2 . intensity of peak relative to that of peak decreases O2 A@1 when the temperature increases. In the presence of added dppm, the relative intensity of peak increases (vide infra).O2 Table 2 Controlled potential electroreductions of palladium complexes on a carbon gauze electrode Solvent/ E/V vs. nexp/ Solution Entry Complex electrolyte salt T SCE F mol~1 colour 1 1 DMF/LiCl 0 °C [1.30 1.84 red 2 1 THF/Bu4NPF6 a r.t. [1.60 4.17 brown 3 1 ACN/Bu4NPF6 a r.t. [1.60 4.25 brown 4 1 ACN/Bu4NPF6 a r.t. [1.15 2.03 green 5 1 THF/Bu4NPF6 a r.t.[1.30 2.08 green 6 1 THF/Bu4NPF6 r.t. [1.20 2.05 brown 7 2 ACN/Bu4NPF6 a r.t. [1.60 3.78 brown 8 2 ACN/Bu4NPF6 a r.t. [1.20 1.90 brown 9 13 ACN/Bu4NPF6 a r.t. [1.60 1.60 brown 10 2 THF/Bu4NPF6 a r.t. [1.05 1.95 brown 11 2 THF/Bu4NPF6 a r.t. [1.60 3.95 brown a Under CO2 . New J. Chem., 1998, Pages 237»246 239E / V ia ic b a –1 –0.3 –1.0 –1.7 E / V ia ic ia ic –1 –2 b a E / V Fig. 2 Thin layer cyclic voltammogram of 1 in THF/0.2 M solution under a –rst scan; b after several scans. (Bun)4NPF6 CO2: Starting potential : 0 V; sweep rate : 0.02 V s~1 Voltammograms similar to those of Fig. 2 are obtained in acetonitrile (ACN) or dimethylformamide (DMF) solutions containing 0.2 M as supporting electrolyte start- (Bun)4NPF6 ing from 1; the system is also observed in thin layer A1/A@1 voltammetry during the second scan.When a bulk electrolysis of 1 is performed at [1.6 V in the presence of carbon dioxide in ACN or DMF solution, the cyclic voltammogram of the resulting solution also exhibits the system. In IR spectroscopy a band at 2358 A1/A@1 mCO cm~1 appears, which is characteristic of the presence of free carbon monoxide in the electrolyzed solution, based on the observation that a DMF solution containing CO exhibits the same band.The addition of water solution to the elec- BaCl2 trolyzed solution of 1 in ACN/0.2 M causes (Bun)4NPF6 to precipitate. After appropriate workup, has BaCO3 BaCO3 been collected and identi–ed by its IR spectrum. As recently mentioned,22 the system corresponds to A1/A@1 the neutral cluster 5, as veri–ed with an Pd3(l3-CO)(l-dppm)3 authentic sample of The dicationic Pd3(l3-CO)(l-dppm)32`.cluster has been initially prepared from palladium(II) acetate with dppm and CO in aqueous acetone containing an excess of tri—uoroacetic acid according to the following literature reaction :23 3 Pd(OAc)2]3 dppm]3 CO]H2O ]2 CF3CO2H]Pd3(l3-CO)(l-dppm)3(O2CCF3)` ]CF3CO2~]2 CO2]6 AcOH (4) The mechanism of formation of this cluster has recently been fully investigated by Puddephatt et al.24 Thus, these results show that the electrochemical reduction of 1 at a potential of [1.6 V in aprotic medium and in the presence of carbon dioxide yields carbon monoxide CO32~, and the neutral cluster 5, according to Pd3(l3-CO)(l-dppm)3 the following global reaction : 3 Pd2(dppm)2Cl2 1 ]6 CO2]12 e~] 2 Pd3(dppm)3CO 5 ]6 Cl~]3 CO32~]CO (5) 5 is then oxidized to the dicationic derivative Pd3(l3-CO)(ldppm) 32`: Pd3(dppm)3CO 5 [2 e~HPd3(dppm)3CO2` A1/A@1 system (6) In all cases, we have veri–ed that the cyclic voltammogram of was not modi–ed in the presence of Pd3(l3-CO)(l-dppm)32` carbon dioxide.25 To gain more insight into the electroreduction process, in particular into the formation of the cluster, we have performed the electrolysis of 1 in the presence of carbon dioxide at potentials higher than [1.6 V in ACN or THF solution containing as supporting electrolyte.The bulk (Bun)4NPF6 electrolysis of 1 in the presence of carbon dioxide at [1.15 V in solution, resulted, after consumption of ACN/(Bun)4NPF6 about two equivalents of electrons F mol~1; see (nexp\2.03 Table 2, entry 4), in a green solution containing complex 7.The latter exhibits a diÜerent cyclic voltammogram as shown in Fig. 3. If the potential scan is reversed after peak B, two oxidation peaks B@ and appear. In thin layer voltammetry, A@1 the system becomes well de–ned after several potential A1/A@1 scans.As this green solution is only stable under a atmo- CO2 sphere, no IR spectroscopic study is possible. When the electrolysis of 1 is performed at the peak potential D in solution, i.e. at [1.3 V, under THF/(Bun)4NPF6 after consumption of about two equivalents of electrons CO2 , F mol~1; see Table 2, entry 5) the cyclic voltam- (nexp\2.08 mogram of the green solution obtained and containing complex 7 exhibits two reduction peaks and at [1.29 V B1 B2 and [1.52 V vs.SCE, respectively (see Table 1, entry 9). When the potential is reversed after peak two oxidation peaks B2 , and appear at [1.18 V and [1.39 V vs. SCE, respec- B@2 B@1 tively (Fig. 4). We can postulate that the initial step corresponds to the formation of a Pd0 complex 6, which is formed Pd2(dppm)2 from the electrogenerated species 3 and 4 according to the two Fig. 3 Cyclic voltammogram of 1 in ACN/0.2 M solu- (Bun)4NPF6 tion after two-electron reduction at [1.15 V under Starting CO2 . potential : [0.3 V; sweep rate : 0.1 V s~1 Fig. 4 Thin layer cyclic voltammograms of 1 in THF/0.2 M solution after two-electron reduction at [1.3 V under (Bun)4NPF6 a –rst scan; b after several scans.Starting potential : [0.2 V; CO2: sweep rate : 0.02 V s~1 240 New J. Chem., 1998, Pages 237»246Pd2(dppm)2(CO2) Pd2 (dppm)2(CO2)– Pd2(dppm)2(CO2)2– +e– +e– 7 7¢ 7¢¢ +CO2 +CO2 +CO2 Pd2(dppm)2(CO2)2 Pd2(dppm)2(CO2)2 – Pd2(dppm)2(CO2)2 2– +e– +e– 8 8¢ 8¢¢ –CO3 2– +L Pd2(dppm)2(CO)L2+ Pd2(dppm)2(CO)L+ Pd2(dppm)2(CO)L +e– +e– 9 9¢ 9¢¢ –CO3 2– +L –CO3 2– +L ic ia a b E / V –1 –2 following equilibrium reactions : Pd2(dppm)2Clx x~ 3 HPd2(dppm)2 6 ]x Cl~ (7) Pd2(dppm)3 4 HPd2(dppm)2 6 ]dppm (8) As the cyclic voltammetric peak disappears and the O1 O2 peak decreases in intensity in the presence of carbon dioxide, we can suggest that the equilibrium reactions (7) and (8) are shifted to the right and the unsaturated electrogenerated Pd0 complex 6 reacts immediately with to give complex 7, CO2 formulated as according to the follow- Pd2(dppm)2(g2-CO2), ing reaction : Pd2(dppm)2 6 ]CO2 ]Pd2(dppm)2(g2-CO2) 7 (9) This behavior has also been observed by electrolysis.After two-electron reduction F mol~1; see Table 2, (nexp\2.05 entry 6) of 1 in solution under argon, deriv- THF/(Bun)4NPF6 atives 3 and 4 are obtained. The addition of causes peak CO2 to disappear, to decrease, and and to appear.O1 O2 B1 B2 Recently, the –rst carbon dioxide coordinated Pd0 complex Pd(g2- has been prepared by the reaction of CO2)(PMePh2)2 with methyl acrylate, followed by treatment PdEt2(PMePh2)2 with The characteristic IR bands due to the coordinated CO2 . ligand are observed at 1658 and 1634 cm~1.26 This CO2 complex is air-sensitive and thermally unstable in the presence of CO2 .The reduction of intermediate 7 in the presence of carbon dioxide gives cluster 5 according to the global reaction : 3 Pd2(dppm)2(g2-CO2) 7 ]3 CO2]6 e~] 2 Pd3(dppm)3CO 5 ]3 CO32~]CO (10) The diverse species obtained by electron or (or CO2 exchange are represented in Scheme 1. From one CO32~) species to the other, one electron is exchanged horizontally and one molecule (or the related reduced species CO2 CO32~) vertically.In order to explain the experimental results we propose that in THF, the uptake of the –rst electron initially gives the anionic species 7@, which reacts quickly Pd2(dppm)2(CO2)~ with to yield the derivative 8@ (EC process). 8@ is oxidized CO2 at the potential of peak The uptake of the second electron B@1.on 7@ to 7A is unlikely, so this hypothesis can be ruled out. 8@ is reduced at the potential of peak to give species 8A, which is B2 relatively stable on the timescale of the voltammetry experiments. Peak corresponds to the oxidation of 8A. B@2 In solution only one B/B@ system is ACN/(Bun)4NPF6 obtained. We suggest that the intermediate 8@ must be reduced at a lower negative potential than 7, according to two diÜerent Scheme 1 possibilities (EC mechanism 8@]8A]9A or CE mechanism 8@]9@]9A).Similar behavior has been observed for palladium22 and ruthenium27 complexes exhibiting an associated single-wave two-electron process. In THF, on the electrolysis timescale (or in cyclic voltammetry at slow sweep rates), complex 8A evolves with elimination of carbonate ion and formation of the neutral derivative 9A containing CO as a ligand and probably a donor solvent molecule (L).It is interesting to note that the proposed mechanism (ECEC process) is similar to that of the electrochemical reduction of in aprotic medium,28 giving a CO2 mixture of CO and [reaction (11)] : CO32~ O~ `CO2 z `e~ CO2]e~(\CO2~~) »»»’ ~CwOwC »»»’ { ~ or CO2’~ O O O~ z ~CwOwC »»»’ CO]CO32~ (11) { ~ O O Moreover, the formation of both CO and from the CO32~ reaction of with a transition-metal complex has already CO2 been described.29 The formation of the neutral cluster 5 from 9A can be explained by the following reaction : 3 Pd2(dppm)2(CO)L 9A ]2 Pd3(dppm)3CO 5 ]CO]3 L (12) Another possibility consists of the reaction between 9A and 6: 2 Pd2(dppm)2(CO)L 9A ]Pd2(dppm)2 6 ] 2 Pd3(dppm)3CO 5 ]2 L (13) To gain more insight into the electrochemical process, we have examined the behavior of the chemically prepared derivative 4.The cyclic voltammogram of 4 in a Pd2(dppm)3 solution exhibits an oxidation peak THF/(Bun)4NPF6 O2 . When the potential scan is reversed after this peak only O2 , one reduction peak R is observed V; R, (O2 , Ep\[0.12 V vs.SCE at 0.1 V s~1; see Table 1, entry 10). Ep\[0.48 Under no modi–cation is observed at room temperature. CO2 Nevertheless, if the solution of THF containing 4 is heated (T \65 °C) before addition of bubbling of carbon CO2 , dioxide induces disappearance of the oxidation wave and O2 apparition of the two de–ned reduction waves and by B1 B2 rotating disk electrode (r.d.e.) voltammetry (Fig. 5) according to the reactions (8) and (9). In the presence of an excess of dppm, the equilibrium of reaction (8) is shifted to the left and no further reactivity of 4 is observed in the presence of CO2 , even at high temperatures. Behavior similar to that of 1 is observed in the case of 2 in solution. The thin layer Pd(dppm)Cl2 ACN/(Bun)4NPF6 Fig. 5 R.d.e. voltammogram of 4 in THF/0.2 M solu- (Bun)4NPF6 tion : a under argon; b under CO2 New J. Chem., 1998, Pages 237»246 241ic ia –1 E / V a b –1 E / V a b ia ic cyclic voltammogram of 2 in ACN or THF exhibits a strong reduction peak After reduction at the potential of peak R1. three oxidation peaks and are observed. The R1, O@1, O@2 O@3 height of peak is comparable to that of peak After O@1 O@3 .several scans, the intensity of peaks and decreases and O@1 O@3 that of peak increases. No oxidation peak is observed, O@2 O2 indicating that derivative 4 is not formed during Pd2(dppm)3 the electroreduction of 2. The above results suggest that the anionic species of Pd0 formulated as 10 is obtained [reaction (14)]. Pd(dppm)Clx x~ 10 is oxidized in two consecutive steps at the potentials of peaks and in turn, giving the PdI intermediate O@1 O@3On the cyclic voltammetry timescale, 10 Pd(dppm)Clx(x~1)~.evolves to give the neutral Pd0 derivative formulated as ììPd0(dppm)Lœœ 11 (L\solvent) [reaction (15)], which is oxidized at the potential peak O@2 . Pd(dppm)Cl2 2 ]2 e~]Pd(dppm)Clx x~ 10 ](2[x) Cl~ (14) Pd(dppm)Clx x~ 10 ]LHPd(dppm)L 11 ]x Cl~ (15) In the presence of the well de–ned system is CO2, A1/A@1 also observed in thin layer voltammetry after several scans if the potential is reversed after peak (Fig. 6). R1 The formation of 5 from the four-electron Pd3(dppm)3CO reduction of 2 F mol~1; see Table 2, entry 7) in (nexp\3.78 the presence of can be explained by the global reaction : CO2 3 Pd(dppm)Cl2 2 ]6 CO2]12 e~] Pd3(dppm)3CO 5 ]6 Cl~]3 CO32~]2 CO (16) To gain more insight into the electroreduction process we have performed the electrolysis of 2 in the presence of carbon dioxide at [1.2 V in solution.After con- ACN/(Bun)4NPF6 sumption of nearly two equivalents of electrons F (nexp\1.90 mol~1; see Table 2, entry 8), a brown solution containing complex 13 was obtained that exhibits the thin layer cyclic voltammogram in Fig. 7; in the cathodic scan, a well de–ned reduction peak B* appears (a shoulder is also observed near [0.9 V). When the scan is reversed after peak B*, the oxidation peak is again observed (see Table 1, entry 16). Under A@1 argon, the height of peak B* decreases and several ill de–ned oxidation peaks are obtained during the anodic scan. In IR spectroscopy, bands located at 1696 and 1634 cm~1 are observed due to the coordinated ligand ; as men- CO2 tioned above, IR bands were observed at 1658 and 1634 cm~1 for the –rst carbon dioxide coordinated Pd0 complex Pd(g2- 13 is only stable under and isolation CO2)(PMePh2)2 .26 CO2 causes its decomposition.We suggest that 13 is a derivative Fig. 6 Thin layer cyclic voltammogram of 2 in ACN/0.2 M solution under a –rst scan; b after several scans.(Bun)4NPF6 CO2: Starting potential : ]0.5 V; sweep rate : 0.02 V s~1 Fig. 7 Thin layer cyclic voltammogram of 2 in ACN/0.2 M solution after two-electron reduction at [1.2 V under (Bun)4NPF6 a –rst scan under b under argon. Starting potential : [0.1 CO2: CO2 ; V; sweep rate : 0.02 V s~1 formulated as Pd(dppm)(g2- which is obtained from CO2), reaction of with the unsaturated Pd0 intermediate 12 CO2 [reaction (19)] ; 12 is formed from complexes 10 and 11 by reactions (17) and (18), respectively.Pd(dppm)Clx x~ 10 HPd(dppm) 12 ]x Cl~ (17) Pd(dppm)L 11 HPd(dppm) 12 ]L (18) Pd(dppm) 12 ]CO2HììPd(dppm)(g2-CO2) œœ 13 (19) The two-electron reduction of intermediate 13 in the presence of carbon dioxide gives cluster 5: 3 ììPd(dppm)(g2-CO2) œœ 13 ]3 CO2]6 e~] Pd3(dppm)3CO 5 ]3 CO32~]2 CO (20) The two-electron process can be rationalized by a mechanism similar to that described in Scheme 1 with the formation, in the initial step, of the anionic species Pd(dppm)(g2-CO2)~. Similar results are obtained in solution.THF/(Bun)4NPF6 In thin layer cyclic voltammetry, the reduction peak R1 decreases, the oxidation peaks and disappear O@1, O@2 O@3 while the system appears after several scans is situ- A1/A@1 (A@1 ated at the same potential as O@1).When the electrolysis of 2 is performed at [1.6 V in solution, cluster 5 is formed after a four- THF/(Bun)4NPF6 electron reduction F mol~1; see Table 2, entry 11) (nexp\3.95 of 2 in the presence of However, when the electrolysis is CO2 .performed at [1.05 V, the current drops to zero after consumption of two electrons F mol~1; see Table 2, (nexp\1.95 entry 10) and no reduction peak is observed by cyclic voltammetry of the resulting solution. In this case, we propose that the equilibrium reactions (18) and (19) are shifted to the left and the electrogenerated species 11 (L\THF) is very unstable on the electrolysis timescale [i.e.Pd0(dppm)L decomposes]. In contrast, when the electrolysis is performed at [1.6 V, the equilibrium reactions are shifted in the opposite way, due to the electrochemical reaction that corresponds to the reduction of 13 [reaction (20)]. The key proposed intermediates in these electrolyses are clearly compounds 6, 7, 12 and 13. They are interesting from a theoretical point of view.For instance, compounds 6 and 7 can exhibit MwM interactions leading to cooperative reactivity. Compounds 8, 8@ and 8A can be used as examples where two ligands are located near each other. Compound 13 is CO2 also important for the preparation of compound 5. One 242 New J. Chem., 1998, Pages 237»246important question is : Can compounds 7 and 13 be reduced at low enough potentials to form their corresponding anionic forms? In the next section, these questions will be addressed theoretically using density functional methods (ADF).Theoretical calculations Both the experimental and theoretical aspects of the binding of on transition metals have been extensively reviewed.30 CO2 Reviews by Gibson31 and Leitner32 have also recently appeared.The bonding of to a single metal atom can CO2 occur via two modes: M(g2- and M(g1- Examples CO2) CO2 ). of the two types are and [Ni(PCy3)2(g2-CO2)]33 respectively and [Ir(dmpe)2Cl(g1-CO2)],34 [Cy\C6H11 Calculations reported in the dmpe\M[(CH3)2P]2CH2N2]. literature35h40 have focused on four basic bonding modes of coordination : (a) carbon»oxygen (g2-CO) ììside-onœœ with CO2 symmetry, (b) carbon (g1-C), with symmetry, (c) oxygen Cs C2v (g1-O) ììend-onœœ and (d) oxygen»oxygen (g2-OO) with C2v symmetry.In general, modes c and d are more energetic than a and b31 and will not be dealt with in this work. The binding of via modes a and b will now be qualitatively described. CO2 has 2 sets of orthogonal p molecular orbitals. The bonding CO2 interactions between and a metal in both modes (g1 and CO2 g2) takes place via two sets of orbitals. The –rst set involves the in-plane p, np and p* molecular orbitals.The second set involves out-of-plane p, np and p* molecular orbitals. The plane is the one. The in-plane molecular orbitals are M(CO2) responsible for the major part of the bonding to the transition metal. In the g1 mode, a signi–cant charge-transfer interaction occurs between the metal orbital and the p* orbital of dZ2 where the metal acts as a two-electron donor and CO2, CO2 as an acceptor. In opposition, the g2 mode uses a p CO2 orbital as the Lewis base (two-electron donor) and an empty d orbital (if available) or a p orbital of the metal atom to generate the r bond.In this case p back bonding occurs using a –lled metal orbital and the empty p* orbital of In dxz CO2 .this work we will concentrate on the g2 form since experimentally the IR data [m(CO)\1696, 1634 cm~1] for 13 indicates the Pd(g2- formulation. Additionally, Salahub and CO2) coworkers40 demonstrated, using ADF in a closely related work, that binds a single palladium atom via a g2 coor- CO2 dination mode. In this case the g1 form was in fact a transition state for the binding of to Pd.Furthermore, based CO2 upon experience with other related Pd compounds, ADF has proven successful in optimizing geometries.41,42 For comparison purposes, and to ensure that the calculation methods are adequate for this work, the geometry of the model compound was optimized and compared to that of the Ni(PH3)2(g2-CO2) X-ray structure and to that of the opti- Ni(PCy3)2(g2-CO2) mized compound.We are interested to Pd(PH3)2(g2-CO2) know what geometry some of the intermediates may have during the various processes encountered in the electroreduction of The geometry was restricted to a sym- CO2 . Cs metry (planar), allowing two possible geometries to occur during optimization (g1-planar, g2-planar).The geometries converged to the g2 form for both Ni and Pd (Fig. 8, Table 3). The comparison of the computed and Ni(PH3)2(g2-CO2) experimental structures is excellent. The Ni(PCy3)2(g2-CO2) largest diÜerence between the two structures is less than 6% [r(MP) 2.160 (calcd.) vs. 2.294 (X-ray)]. The comparison Aé between the calculated and X-ray structure for the Ni(g2- fragment is particularly good with diÜerences not CO2) exceeding 2.5%.We now compare the two optimized geometries for MxNi and Pd. Both geometries exhibit the same coordi- g2-CO2 nation modes with very similar distances and angles. The comparison of the MP, MC and MO indicates that r(Ni)\r(Pd). This diÜerence is simply due to the diÜerence in covalent radii between Ni and Pd.43,44 The signi–cant result is that the r(CO) distance for the coordinated CO is longer in the Ni case.In addition the OCO is smaller in the Ni case n as well. Knowing that activation leads to products in CO2 which the hybridization of the C atom changes from sp to sp2 (carbonate, formate, oxalate, etc. . . .),31,32 these calculations predict that the geometry in the Ni complex is more dis- CO2 torted than in the Pd one.In other words, the CxO bond is Fig. 8 Optimized geometry of the planar model Pd(PH3)2(g2-CO2) compound. The symmetry was restricted to Cs Table 3 Comparison of the calculated planar geometries (M\Ni, Pd) and X-ray structuresa M(PH3)2(g2-CO2) Ni(PCy3)2(g2-CO2) Pd(PH3)2(g2-CO)2 Ni(PH3)2(g2-CO2) Ni(PCy3)2(g2-CO2)b *c r(MP) 2.334 2.160 2.294 0.174 long r(MP) 2.249 2.078 2.163 0.171 short r(MC) 2.147 1.886 1.857 0.261 r(MO) 2.270 1.921 1.967 0.349 r(CO) 1.191 1.195 1.211 [0.004 r(CO) 1.215 1.245 1.257 [0.030 coordinated nOCO 151.0 146.0 136.2 nOMP 120.1 110.3 105.3 nCMP 91.5 97.4 93.9 nPMP 116.6 114.1 122.6 nCMO 31.7 38.2 38.3 a r in and in degrees (°).Only selected data are presented. b From ref. 51. c * is de–ned as ” n r(Pd)calc[r(Ni)calc New J.Chem., 1998, Pages 237»246 243weaker in the Ni complex and more likely to be activated in a thermodynamic sense, than the Pd analogue. The following series of computations concern the reduced species, which is generated by the one- ììPd(dppm)(CO2)~œœ electron reduction of compound 13. The geometry optimizations of the planar model compounds Pd(PH3)2(CO2)~ were performed using the restricted symmetry, and the Cs coordination adopted the g1-bonding form (Fig. 9). CO2 The nature of the frontier orbitals is no diÜerent in the reduced and neutral species. HOMO-1 is a bonding interaction between the mixed with the orbital of the dz2 dx2~y2 metal and the p* orbital of The HOMO is of course CO2 . singly occupied and is mainly composed of the Pd orbital 5pz (consistent with the d10]1 e~ electronic con–guration) and the C orbital (along with some signi–cant O orbitals ; i.e.pz pz p* of the The PdwC interactions are antibonding. The CO2). LUMO is the other p* system. As a result, upon CO2 reduction of 13, the reduced species should exhibit an increased PdwC bond length. The calculated r(PdC), r(CO) and OCO data are as follows : 2.177 1.233 and 136.8°, n ”, ” respectively.The increase in r(PdC) is indeed computed (*\]0.030 but is perhaps not as extensive as one may ”) expect. The reason for this is that the excess electron occupies an orbital that is delocalized in the p* system of the frag- CO2 ment. Indeed, r(CO) shows an increase from 1.191 and 1.215 ” (average 1.203 for the neutral species) to 1.233 (charged ” ” species ; *+0.030 The OCO angle decreases greatly from ”). 151.0° to 136.8°. For comparison purposes, r(CO) and the OCO angle in acetate lie somewhere around (CH3CO2~) 1.24»1.25 and 125»130°, respectively. This favorable com- ” parison strongly suggests that has become a metallated CO2 carboxylate compound upon reduction. (RCO2~; R\PdL2) This observation could explain the reactivity between CO2~ and indicated in reaction (11) (except that is CO2 CO2~ replaced by here).The computed PdwP bond MCO2~ lengths, averaging 2.282 are close to that of the neutral ”, species (Table 3; 2.292 but also indicate an increase in ”) PdwP back bonding (i.e. PdwP bond shortening) due to the increase in electronic density at the metal center upon reduction.The complete computed structures are provided in the Supplementary Material. The following section addresses the com- Pd2(dppm)2 pounds, notably compounds 6, 7 and 8. Recently Fink et al.45 pointed out that d10 monomeric Pd(diphos) compounds (diphos\diphosphine ligand) are in a monomer»dimer equilibrium. In the dppm case described here, the similarities in the electrolysis of 1 and 2 in the presence of is consistent CO2 with the literature : 2 Pd(diphos)HPd2(diphos)2 (19) For convenience the starting geometry of 6 was (in a restricted symmetry with the Pd2(H2PCH2PH2)2 ììC2vœœ Fig. 9 Optimized geometry of the planar model Pd(PH3)2(g1-CO2)~ compound. The symmetry was restricted to Cs methylene groups pointing the same way). The optimizations were performed using the restricted symmetry; this pro- Cs cedure allows some degree of freedom from the molecule.After convergence, the calculated complex does not exhibit a perfectly planar structure. (Fig. 10), but rather the Pd2P4 PPdP angles are 168 and 174°. This slight deviation from linearity is due to the fact that the groups were placed both CH2 pointing the same way instead of in opposite directions (C2h symmetry).This preferred orientation was selected to allow further computations with addition of with a minimum CO2 of steric eÜect (see below). The new feature is, of course, the appearance of Pd… … …Pd interactions. Here the computed distance is 2.854 ” [r(P… … …P)\2.956 The presence of Pd… … …Pd interactions in ”]. d10»d10 complexes is well documented in the literature,46 and can be experimentally addressed by X-ray crystallography (whether there is a chemical bond or just a weak interaction) and by UV-visible and Raman spectroscopy.46h48 Related examples are Finkœs dimer, Pd2(Cy2PCH2CH2PCy2)2 , and reported by Kirss and r(Pd2)\2.7611 ”,45 Pd2(dppm)3 , Eisenberg In this latter case, reso- [r(Pd2)\2.956(1) ”49]. nance Raman spectroscopy established that the m(Pd2), Pd2 stretching frequency, is 120 cm~1.The van der Waals radii is 1.6 Relevant to this work, Sakaki et al.50 have reported a ”.44 theoretical study on the bond energy and the bonding nature of dinuclear d10 metal complexes of the type (M\Pd, (ML2)2 Pt; using ab initio MO methods. The presence of L\PH3), interactions results from bonding interactions with Pd2 ps orbitals.Also, the HOMO mainly includes the anti- ds»ds bonding overlap into which s and orbitals of one Pd mix in ps a bonding way with the orbital of the other Pd atom. ds Because the orbital population decreases slightly, and the s ds orbital population decreases greatly, Sakaki et al.50 concluded that the s and orbitals mix into the antibonding inter- ps dsds action to reduce the exchange repulsion, and the charge ds»ds transfer from the of one M to the of the other M is ds sps weak.As a consequence, a sp]sp2 rehybridization occurs when the geometry changes from the two-coordinate ML2 system to the three-coordinate one. However, the L2MwML2 question is whether the rehybridization process is a complete or partial rehybridization. At a distance of 2.854 which is ”, well above the sum of the covalent radii,44 it is clear that the rehybridization is not as extensive.Indeed, the calculated PdwPd bond energy for is rather small (PH3)2PdwPd(PH3)2 (somewhere between 14 and 4 kcal mol~1)50 for a similar distance of B2.885 (optimized geometry). In conclusion, we ” should consider the PdwPd bonding as weak interactions and Fig. 10 Optimized geometry of the and Pd2(H2PCH2PH2)2 model compounds.The symmetry was Pd2(H2PCH2PH2)2(g2-CO2) restricted to Cs 244 New J. Chem., 1998, Pages 237»246not as a formal coordination bond. The computed r(PdP) data average to 2.262 and are normal. For the ” Pd2(dppm)3 , average X-ray r(PdP) is 2.310 ”.49 The next point addresses the nature of the cooperation between the two Pd atoms.Compound 7 could exist under a l-bridging form of the type PdwOwC(O)wPd (oxidation of the Pd centers). In prior calculations, the was placed in CO2 the bridging position with the same distances between PdwC and PdwO. The symmetry was restricted to During the Cs . optimization, the bridging ligand moved out of its initial CO2 bridging position and stabilized in a g2-conformation (Scheme 2).Here the is still behaving as a two-electron donor CO2 ligand. The conformation is obviously not planar (Fig. PdP2CO 10) as imposed by the computations. Here again some Pd… … …Pd interactions calculated] is also [r(Pd2)\2.821 ”, predicted by theory, but does not signi–cantly diÜer from the model compound described above Pd2(H2PCH2PH2)2 Other structural data of interest are [r(Pd2)\2.854 ”].r(PdC), r(PdO), r(PdP) and the OCO angle : 2.189, 2.579, 2.270^0.015 and 151°, respectively (see the Supplementary ” Material for details). In this case, the bonding is PdwCO2 predicted to be slightly weaker in comparison with the data of Table 3 [i.e. slightly longer r(PdC) and r(PdO) values]. The presence of weak Pd… … …Pd interactions, which induce a localization of part of the electronic density between the two metals, decreases the backbonding interactions between Pd and No computation was performed for the reduced CO2 .species ; it is reasonably assumed that the conclusions drawn for the monomeric model compounds discussed above are the same here. We also anticipate that a Pd(H2PCH2PH2)2Pd(g2- structure with a planar is also possible, but CO2) PdP2CO2 was not optimized.The –nal series of computations deals with compound 8 as regards the nature of the cooperative properties on the skeleton. According to Scheme 1, the conversion Pd2(dppm)2 of two molecules into and CO must pass by a CO2 CO32~ mechanism involving either 8, 8@ or 8A. The geometry of 8 was optimized by placing two face-to-face (in a restricted CO2 Cs symmetry).The optimized geometry (Fig. 11) does keep the face-to-face conformation, but also generates local Pd conformations similar to that shown in Fig. 8. The r(PdC), r(PdO), long r(PdP), short r(PdP) and OCO angle data at convergence are 2.190, 2.284, 2.335 and 2.250 ” and 153.7°, respectively, and compare favorably to that of in Table 3; the PdwC and PdwO dis- Pd(PH3)2(g2-CO2) tances are again slightly longer (same reason: Pd… … …Pd interactions).The calculated Pd… … …Pd separation is now Scheme 2 Fig. 11 Optimized geometry of the Pd2(H2PCH2PH2)2(g2-CO2)2 model compound. The symmetry was restricted to Cs slightly longer (2.955 but close to that reported for 3. More ”) importantly, the C… … …C and O… … …O contacts (between 3.2 and 3.4 are smaller than the sum of the van der Waals radii.43 ”) This result predicts that any intramolecular pro- CO2… … …CO2 cesses (8]9, 8@]9@, 8A]9A; Scheme 1) could be possible in a face-to-face geometry. The limiting step now depends upon the relative ratio of uptake (ex.: 7]8) vs. electron trans- CO2 fer (ex. : 7]7@). These parameters depend upon the applied potential and the concentration.The fact that a second CO2 metal center is located near the –rst one allows one to consider the possibility that a second and diÜerent molecule can be activated (coordinated) simultaneously. This situation brings in the intramolecular coupling of two diÜerent molecules. Monodentate phosphine ligands do not oÜer this ììtemplateœœ opportunity. The geometry for 8@ and 8A has not been calculated ; it is also anticipated that conclusions similar to those obtained for the planar model com- Pd(PH3)2(CO2)~ pound will be drawn.The optimized structural data of this compound are also available in the Supplementary Material. Conclusion The electroreduction of and in Pd2(dppm)2Cl2 Pd(dppm)Cl2 aprotic medium (THF, ACN) under leads to CO2 and The overvoltage of the Pd3(dppm)3CO CO32~.CO2 reduction is decreased by 0.6 V. No catalytic process is observed in these experimental conditions. Studies of this indirect reduction of with the above cited palladium deriv- CO2 atives in the presence of added substrates (i.e. Lewis acid) are in progress. Supplementary material Computed structural data for planar and perpendicular and and for Pd(PH3)2(CO2) Pd(PH3)2(CO2)~, (x\0, 1, 2) are available from the Pd2(H2PCH2PH2)2(CO2)x authors (8 pages).Acknowledgements This research was supported by NSERC (Canada), FCAR (Queç bec), de France (EDF»Club Eä lectriciteç dœEä lectrochimie Organique), the Conseil Reç gional de Bourgogne and the Preç - fecture de la Reç gion Bourgogne. We are grateful to Prof. Jean Lessard and Dr.Jean-Marc Chapuzet for helpful discussions. References 1 (a) B. J. Fisher and R. Eisenberg, J. Am. Chem. Soc., 1980, 102, 7361; (b) M. R. M. Bruce, E. Megehee, B. P. Sullivan, H. Thorp, T. R. OœToole, A. Downard and T. J. Meyer, Organometallics, 1988, 7, 238 and references therein ; (c) K. W. Frese, D. P. Summers and New J. Chem., 1998, Pages 237»246 245M. J. Cinibulk, J.Electrochem.Soc., 1988, 135, 264; (d) C. Amatore and J. M. Saveç ant, J. Am. Chem. Soc., 1981, 103, 5021; (e) E. Lamy, L. Nadjo and J. M. Saveç ant, J. Electroanal. Chem., 1977, 78, 403; ( f ) B. J. Fisher, T. Lehman and E. J. Heitz, Appl. Electrochem., 1981, 11, 743; (g) P. A. Christensen, A. Hammett and A. V. G. Muir, J. Electroanal. Chem., 1990, 288, 197; (h) J. Desilvestro and S.Pons, J. Electroanal. Chem., 1989, 267, 207. 2 The standard potential of the couple in N,N-dimethyl- CO2/CO2~ formamide with 0.1 M as supporting electrolyte has NEt4ClO4 been estimated to be [2.2 V vs. SCE by cyclic voltammetry on a mercury electrode [see ref. 1(e)]. 3 Recent reviews : (a) G. Silvestri, in Carbon Dioxide as a Source of Carbon, ed. M. Aresta and G. Forti, NATO ASI Series : Ser.C, Reidel, Dordrecht, 1987, p. 339. (b) G. Silvestri, S. Gambino and G. Filardo, in Enzymatic and Model Carboxylation and Reduction Reactions for Utilization, NATO ASI Series, Reidel, Dor- CO2 drecht, 1990, 314, 101. 4 J. P. Collin and J. P. Sauvage, Coord. Chem. Rev., 1989, 93, 245. 5 D. L. DuBois and A. Miedaner, J. Am. Chem. Soc., 1987, 109, 113. 6 S.Slater and J. H. Wagenknecht, J. Am. Chem. Soc., 1984, 106, 5367. 7 A. Szymaszek and F. P. Pruchnik, J. Organomet. Chem., 1987, 376, 133. 8 D. L. DuBois, A. Miedaner and R. C. Haltiwanger, J. Am. Chem. Soc., 1991, 113, 8753. 9 K. S. RatcliÜ, R. E. Lentz and C. P. Kubiak, Organometallics, 1992, 11, 1986. 10 (a) D. L. DuBois and A. Miedaner, Inorg. Chem., 1986, 25, 4642; (b) D. L. DuBois and A.Miedaner, Inorg. Chem., 1988, 27, 2479; (c) A. Miedaner, D. L. DuBois and C. J. Curtis, Organometallics, 1993, 12, 299; (d) B. D. Stefey, C. J. Curtis and D. L. DuBois, Organometallics, 1995, 14, 4937. 11 (a) L. S. Benner and A. L. Balch, J. Am. Chem. Soc., 1978, 100, 6099; (b) E. W. Stern and K. Maples, J. Catal., 1972, 27, 120. 12 E. J. Baerends, D. E. Ellis and P.Ros, J. Chem. Phys., 1973, 2, 41. 13 E. J. Baerends, Ph.D. Thesis, Vrije Universiteit, Amsterdam, 1975. 14 W. Raveneck, in Algorithms and Applications on V ector and Parallel Computers, ed. H. J. J. Rigie, Th. J. Dekkor and H. A. van de Vorst, Elsevier, Amsterdam, 1987. 15 P. M. Boerrigter, G. te Velde and E. J. Baerends, Int. J. Quantum Chem., 1988, 33, 87. 16 L. Versluis and T.Ziegler, J. Chem. Phys., 1988, 88, 322. 17 (a) G. J. Snijders, E. J. Baerends and P. Vernooyé s, At. Nucl. Data T ables, 1982, 26, 483; (b) P. Bernooyé s, G. J. Snijders and E. J. Baerends, Int. Slater T ype Basis Functions for the W hole Periodic System, Free University of Amsterdam, Amsterdam, 1981. 18 (a) L. Noodleman and J. G. Norman, J. Chem. Phys., 1979, 70, 4903; (b) L. Noodleman, J.Chem. Phys., 1981, 74, 5737; (c) L. Noodleman and E. J. Baerends, J. Am. Chem. Soc., 1984, 106, 2316. 19 J. Krijn and E. J. Baerends, in Fit Functions in the JFS Method, (in Dutch), Free University of Amsterdam, Amsterdam, 1984. 20 S. D. Vosko, L. Wilk and M. Nusair, Can. J. Phys., 1990, 58, 1200. 21 (a) C. Amatore, M. Azzali and A. Jutand, J. Organomet. Chem., 1989, 363, C41; (b) C.Amatore, M. Azzali and A. Jutand, J. Am. Chem. Soc., 1991, 113, 8375. 22 I. Gauthron, Y. Mugnier, K. Hierso and P. D. Harvey, Can. J. Chem., in the press. 23 (a) L. Manojlovic-Muir, K. W. Muir, B. R. Lloyd and R. J. Puddephatt, J. Chem. Soc., Chem. Commun., 1983, 1336; (b) B. R. Lloyd and R. J. Puddephatt, Inorg. Chim. Acta, 1984, 90, L77; (c) L. Manojlovic-Muir, K. W.Muir, B. R. Lloyd and R. J. Puddephatt, J. Chem. Soc., Chem. Commun., 1985, 536. 24 B. R. Lloyd, L. Manojlovic-Muir, K. W. Muir and R. J. Puddephatt, Organometallics, 1993, 12, 1231. 25 Host-guest reactivity of has been Pd3(l3-CO)(l-dppm)32` observed recently by UV-visible spectroscopy. This reactivity is reversible by bubbling another gas. 26 M. Sakamoto, I. Shimizu and A. Yamamoto, Organometallics, 1994, 13, 407. 27 D. T. Pierce and W. E. Geiger, J. Am. Chem. Soc., 1992, 114, 6063. 28 (a) C. Amatore, L. Nadjo and J. M. Saveç ant, Nouv. J. Chim., 1979, 3, 545; (b) E. Lamy, L. Nadjo and J. M. Saveç ant, J. Electroanal. Chem., 1977, 78, 403; (c) C. Amatore and J. M. Saveç ant, J. Am. Chem. Soc., 1981, 103, 5021; (d) C. Amatore, L. Nadjo and J. M. Saveç ant, Nouv. J. Chim., 1984, 8, 565. 29 See for example: (a) G. R. Lee, J. M. Maher and N. J. Cooper, J. Am. Chem. Soc., 1987, 109, 2956; (b) G. Fachinetti, C. Floriani, A. Chiesi-Villa and C. Guastini, J. Am. Chem. Soc., 1979, 101, 1767; (c) J. Chatt, M. Kubota, G. J. Leigh, T. C. March, R. Mason and D. J. Yarrow, J. Chem. Soc., Chem. Commun., 1974, 1033; (d) H. H. Karsch, Chem. Ber., 1977, 110, 2213; (e) E. Carmona, F. Gonzalez, M. L. Poveda and J. M. Marin, J. Am. Chem. Soc., 1983, 105, 3365; ( f ) R. Alvarez, E. Carmona, M. L. Poveda and R. Sanchez- Delgado, J. Am. Chem. Soc., 1984, 106, 2731. 30 R. H. Crabtree, T he Organometallic Chemistry of the T ransition Metals, Wiley, New York, 2nd edn., 1994, p. 318. 31 D. H. Gibson, Chem. Rev., 1996, 96, 2063. 32 W. Leitner, Coord. Chem. Rev., 1996, 153, 257. 33 M. Aresta, C. F. Nobile, V. G. Albano, E. Forni and M. Manassro, J. Chem. Soc., Chem. Commun., 1975, 636. 34 T. Heckovitz, J. Am. Chem. Soc., 1977, 99, 2391. 35 E. Kaufmann, S. Sieber and P. von Ragueç Schleyer, J. Am. Chem. Soc., 1989, 111, 4005. 36 S. Sakaki and K. Ohkubo, Inorg. Chem., 1988, 27, 2020. 37 S. Sakaki and Y. Musashi, J. Chem. Soc., Dalton T rans., 1994, 3047. 38 C. Bo and A. Dedieu, Inorg. Chem., 1989, 28, 304. 39 M. Sodupe, V. Branchadell and A. Oliva, J. Phys. Chem., 1995, 99, 8567. 40 S. Sirois, M. Castro and D. R. Salahub, Int. J. Quantum Chem. : Quantum Chem. Symp., 1994, 28, 645. 41 R. Provencher and P. D. Harvey, Inorg. Chem., 1996, 35, 2113. 42 P. D. Harvey, R. Provencher, J. Gagnon, T. Zhang, D. Fortin, K. Hierso, M. Drouin and S. M. Socal, Can. J. Chem., 1996, 74, 2268. 43 F. A. Cotton, G. Wilkinson and P. D. Gaus, Basic Inorganic Chemistry, Wiley, Toronto, 3rd edn., 1995, p. 61. The covalent radii for Ni is 1.25 ”. 44 We estimate the covalent radii for Pd to be B1.32 which is half ”, of the average of all known PdwPd bond distances for singly bonded dimers. See, for example: P. D. Harvey and Z. Murtaza, Inorg. Chem., 1993, 32, 4721. 45 M. J. Fink, J. T. Mague and Y. Pan, J. Am. Chem. Soc., 1993, 115, 3842. 46 For a recent review see : P. D. Harvey, J. Cluster Sci., 1993, 4, 377. 47 P. D. Harvey, R. F. Dallinger, W. H. WoodruÜ and H. B. Gray, Inorg. Chem., 1989, 28, 3057. 48 P. D. Harvey and H. B. Gray, J. Am. Chem. Soc., 1988, 110, 2145. 49 R. V. Kirss and R. Eisenberg, Inorg. Chem., 1989, 28, 3372. 50 S. Sakaki, M. Ogawa and Y. Musashi, J. Phys. Chem., 1995, 99, 17134. Received 3rd February 1997; revised M/S received 23rd June 1997; Paper 7/08331I 246 New J. Chem., 1998, Pages 237»246

ISSN:1144-0546    

DOI:10.1039/a708331i

出版商:RSC    

年代:1998

数据来源: RSC

摘要:

O O OH O O OH HO –H2O +H2O Regioselective or enantiogenic electrochemical and microbial reductions of 1,2-diketones Patrick Boutoute, Guy Mousset* and Henri Veschambre L aboratoire de et Etude de Biologique (CNRS Synthe` se, Electrosynthe` se Syste` mes a` Inteç re� t UMR 6504), Blaise Pascal, 24, Avenue des L andais, 63177 Aubiere, France Universiteç The electrochemical and microbial reduction of 1,2-diketones have been studied and the results compared among themselves and with those of previous works.The electrochemical reduction is highly selective and allows a-ketols to be obtained. The latter, being unreducible at the –xed potential used, never form diols. The anaerobic microbial reduction by Proteus mirabilis proves to be more chemo- and regioselective than that previously performed with other microorganisms.After two hours of incubation, the enantiomerically pure (2S)- a-ketols are isolated in high yield except in the case of 1,2-cyclohexanedione, for which the (1S,2S) diol is formed for all reaction times. The microbial reduction of the electrochemically generated racemic 2-hydroxycyclohexanone yields the sole (1S,2S) diol. Reç duction e ç lectrochimique ou microbiologique reç gioselective ou e ç nantiogeç nique de 1,2-diceç tones.Les reç sultats de la reç duction eç lectrochimique et microbiologique par Proteus mirabilis de six 1,2-diceç tones ont eç teç compareç s entre eux et avec ceux obtenus dans des travaux anteç rieurs pour dœautres microorganismes. La meç thode eç lectrochimique, qui oÜre la possibiliteç dœimposer un pouvoir reç ducteur, conduit aux a-ceç tols en absence de toute trace de diols. Lœaction du microorganisme Proteus mirabilis sœest aveç reç e beaucoup plus reç gioseç lective que celle observeç e dans des eç tudes preç ceç dentes faisant intervenir dœautres microorganismes.Apre` s deux heures dœincubation, les a-ceç tols de con–guration (2S) sont obtenus eç nantiomeç riquement purs avec de bons rendements sauf pour la 1,2 cyclohexanedione qui donne directement le diol (1S,2S).La reç duction microbiologique de la 2-hydroxycyclohexanone raceç mique geç neç reç e eç lectrochimiquement fournit le diol (1S,2S) par reç duction eç nantioseç lective du ceç tol (2S). Regioselective or enantiogenic reactions have been frequently used in syntheses of biologically active compounds or natural product precursors.1h4 In the –eld of regioselective and asymmetric synthesis, considerable eÜort has been made to obtain ketols and diols from prochiral dicarbonyl derivatives or ole- –nes.5h9 The present paper deals with the electrochemical and microbial reduction of the diketones : In aqueous medium, the a-diketones 1»6 may exist under three forms in equilibrium : The hydrated species is largely prevalent for alkyl diketones10 whereas the enolic form is predominant in the case of cyclic a-diones.11 Electrochemistry can be successfully used for speci–c reductions (or oxidations) of organic molecules possessing more than one electroactive group.12 On the other hand, microbial reactions are often the best way to prepare optically active products with very high enantiomeric excesses.13 In this work we will assess the capability of the electrochemical method to reduce selectively one of the carbonyl groups and the possibility of microorganisms either to produce enantiomerically pure compounds from a prochiral center or to perform an enantiomeric recognition in the case of electrogenerated racemic mixtures.Results and Discussion Electrochemical reduction of the a-diketones 1ñ6 Numerous analytical studies have led to the establishment of detailed mechanisms concerning the electrochemical reduction of aliphatic or aromatic a-diketones according to the pH of aqueous solutions.14h21 We have carried out all our electrolyses in buÜered solutions at pH 7, thus allowing a comparison with microbial reductions.Under these conditions the cyclic voltammograms of dicarbonyl compounds 1»5 show identical two-electron reduction waves and on the reverse scan, a smaller anodic wave attributed to the oxidation of a dienolic structure RwC(OH)xC(OH)wR@ formed in the reduction process.15 The aromatic diketone 6 is reduced in two steps : the –rst one V vs.SCE) corresponds (Ep1\[0.72 to the reduction of the diketone15 and the second, of weak intensity V vs. SCE), to that of an a-ketol15 or a (Ep2\[1.15 hydrated form of the initial dicarbonyl derivative.16,17 As a comparison, the electrochemical reduction of a-diketones in an aprotic solvent takes place in two steps.22 Table 1 illustrates the results obtained from the electrolyses of compounds 1»6 in aqueous medium.In all the cases, the a-ketols are the main reaction products. We have never observed the formation of a-diols. The asymmetrical diketones New J. Chem., 1998, Pages 247»251 247O OH HO O + O O OH OH C XH R¢ R¢ C X R R¢ H reduction O R O OH R O O R OH OH R OH + ketol a ( S) ketol b ( S) erythro + threo Table 1 Electrochemical reduction of a-diketones 1»6 at a –xed potential a-Ketols/% Side products/ a-Diketone Ep/vs.SCE CH3CHOHCOR CH3COCHOHR % 1 [0.86 28 52 20 2 [0.93 32 58 10 3 [1.05 38 57 5 6 [0.72 0.8 92 » 4 [0.88 86 14 5 [1.50 85 15 yielded a mixture of the two isomeric ketols, in which the 3- hydroxyketones are always the major product. Except for the aromatic diketone, we have observed the formation of side products.These, derived from diketones 1»4, are very unstable and did not allow for puri–cation on a silica gel column or by preparative gas chromatography. Their retention times in analytical gas chromatography and the fact that the crude reaction mixtures show, for the side products, the presence of three isomers in the case of the asymmetrical diketones 1, 2, 3 and only one for the symmetrical 4 lets us presume the formation of diastereomeric keto-pinacols by dimerization of the intermediate free radical.In the case of 1,2-cyclohexanedione, the impurity has been identi–ed as cyclohexanone. The electroreductive cleavage of the CwOH bond in a-ketols has previously been observed.14h18 We can also note that the a-ketol of 1,2-cyclohexanedione (adipoine) can slowly dimerize in aqueous solution :23 This reaction is favored by silica gel and therefore prevents any puri–cation by silica gel chromatography.The dimer is, however, completely insoluble in diethyl ether and cannot be extracted from the aqueous solutions. Microbial reduction of the a-diketones 1ñ6 by Proteus mirabilis The synthesis of optically active natural products and biologically active compounds often requires chiral synthons as starting materials, most of them being obtained by enantiogenic reduction of prochiral moieties.24h27 The carbonyl function (X\O) is certainly one of the most studied and the use of enzymatic biocatalytic systems constituted of puri–ed enzymes or whole cell microorganisms generally oÜers the advantage of producing enantiomerically pure compounds.The microbial reduction of dicarbonyl compounds by Bakerœs yeast, microorganisms or bacteria has been previously approached in our group7 but in all cases, the asymmetrical a-diketones 1, 3 and 6 have been reduced to a mixture of a-ketols and a-diols, even for very short reaction times (1 h, Table 2). The diols and a-ketols obtained were enantiomerically pure.Unfortunately, the reaction was never regioselective. In the particular case of 1,2-cyclohexanedione 5, the formation of the a-ketol was never observed, regardless of the microorganism or reaction time used.7 The diastereomeric diols were the sole products isolated. In continuation of our research, we present some results obtained under anaerobic conditions with Proteus mirabilis, a more selective microoanism allowing the formation of enantiomerically pure aketols using short reaction times (Table 3). In a –rst step the a-diketones 1»4 and 6 are reduced to the corresponding aketols, which can be further reduced to diols with longer reaction times.In contrast to the former microorganisms, it has been observed that Proteus mirabilis can induce a regiospeci–c reaction with the less hindered carbonyl group.Only the 2- hydroxyketones are obtained in high yields. In the case of 1,2- cyclohexanedione, it is unfortunately impossible to restrict the reaction to the reduction of only one carbonyl function. Even for short reaction times, the (1S,2S)-cyclohexanediol is isolated as the main reaction product. We can hypothesize that the (2S)-ketol is rapidly reduced to the (1S,2S)-diol.Electrochem- Table 2 Microbial reduction of a-diketones 1, 3 and 6a a-Diketones 1 R\C2H5 3 R\C5H11 6 R\C6H5 Reaction products/% Ketol a Ketol b Diols Ketol a Ketol b Diols Ketol a Ketol b Diols Bakerœs yeast 17 14 69 35 7 58 70 » » (30% ketone) Aspergillus niger 4 12 84 2 4 94 » 50 50 Geotrichum candidum 25 15 60 29 7 64 60 » 17 (23% ketone) Rodhotorula rubra 42 10 48 38 12 50 62 » 13 (25% ketone) a Incubation time: 1h.Results for 6 from ref. 7. 248 New J. Chem., 1998, Pages 247»251OH O OH O OH O OH O OH O O O OH OH 1 S,2 S Proteus mirabilis OH O ( R + S) Microbial Reduction Electrochem reduction Table 3 Microbial reduction of a-diketones 1»6 by Proteus mirabilis Reaction Compound time/h a-Diketone/% a-Ketol/% Diols/% 1 2 8 86 (2S,3S) 4 (2S,3R) 2 24 » 9 (2S,3S) 84 (2S,3R) 7 2 2 1 95 (2S,3S) 3 (2S,3R) 1 24 » 52 (2S,3S) 45 (2S,3R) 3 3 2 1 99 » 24 » 99 » 6 2 18 82 » 24 7 98 » 4 2 11 87 (3S,4S) 2 24 » 2 (3S,4S) 98 5 2 70 3a (1S,2S) 27 24 11 2 (1S,2S) 87 a The ketol has not been isolated but only characterized by gas chromatography and 1H and 13C NMR spectrometry.istry is able to produce the racemic (R]S)-2-hydroxycyclohexanone.We have submitted the racemic form of this a-ketol to an enantiospeci–c reduction by Proteus mirabilis. The reaction is controlled by GC analyses on chiral and nonchiral columns. We have observed the formation of the sole (1S,2S)-a-diol at all reaction times with an enantiomeric excess (ee) higher than 96%. For a reaction time of 24 h, 45% of the a-ketol was consumed.If the reaction is carried out for a longer time, the ketol progressively disappears without increasing the quantity of pure diastereomeric (1S,2S)-a-diol. We can tentatively explain this fact by an enantioselective recognition of the (2S)-ketol by the enzyme reductase with formation of the enantiomerically pure (1S,2S)-diol and the progressive dimerization of the (2R)-ketol to a nonbiologically reducible form.All the isolated compounds have been characterized by comparison of their optical activity with that of authentic samples.7,27 The enantiomeric excesses are de–ned by gas chromatography on a chiral column (Chirasil L-Valine or Lipodex E). In conclusion, electrochemistry is a good chemospeci–c method for reducing a-diketones in aqueous medium owing to the possibility to regulate the reduction efficiency by applying the appropriate redox potentials.Only a-ketols are isolated, even in the case of cyclohexanedione. In the case of asymmetrical diketones, the technique appeared slightly regioselective for compounds 1»5 and highly regioselective for diketone 6. The microbial reduction of the same diketones by Proteus mirabilis is particularly interesting compared to the results obtained with other microorganisms because of its high regioand enantioselectivity.For short incubation times (1 h), contrary to electrochemistry, we observed only the reduction of the –rst carbonyl group (less hindered) and the isolated ketols are enantiomerically pure. These results complement those previously published because, in our –rst studies, we have never obtained pure ketols of 2,3-diketones with short alkyl chains and even for an incubation time of one hour.(C2 C3), The reaction products were always a mixture of isomeric ketols and diols. So, we now possess a choice of microorganisms giving simple access to a-ketols or a-diols. Experimental General methods Cyclic voltammetry measurements were performed with a stationary mercury drop electrode and a Tacussel PRT 20-2X potentiostat. The reference electrode was a saturated calomel electrode (SCE).A Tacussel PRT 100-1X potentiostat coupled with a Tacussel IG5N integrator was used for controlled potential electrolyses, which were performed in a threecompartment glass cell joined by two glass frits.All the New J. Chem., 1998, Pages 247»251 249macroscale reductions were carried out at a stirred mercury pool electrode (total area approximately 45 cm2) with 400 ll (or 400 mg for solids) of substrate in 100 ml of a pH 7 Mac Ilvain buÜered solution. The microorganisms were all laboratory grown except Bakerœs yeast, which was purchased (Hirondelle, SI LesaÜre, Paris).Proteus mirabilis is obtained from Institut Pasteur (Paris, CIP 75.15). After culturing at 37 °C in a sterilized (20 min at 120°) medium [composed of yeast extract (Difco, 5 g), Tryptone (Difco, 20 g), glucose (5 g), (5 g) and dis- K2HPO4 tilled water to make one litre of solution] for 9 h under anaerobic conditions, 50 ll of dione (or 50 mg for solids) were syringed through a septum.After incubation at 37 °C on a rotating table for the time appropriate to the formation of the desired product (2 h for a-ketones or longer times for diols), the mixture was –ltered on sintered glass or centrifuged for ten minutes at 8000 rpm. The solution was then continually extracted with diethyl ether for 24 h. The overall yields were quite similar for electrochemical or microbial reductions.The culture of the other microorganisms mentioned have been previously described.27 Gas chromatography was performed using a Shimadzu GC 14 instrument equipped with a —ame ionisation detector and DB1, Carbowax 20 M or Chirazil L-Valine and Lipodex E capillary columns for the measurement of enantiomeric excesses. Substrates and solutions The diketones 1, 2, 4, 5 and 6 were purchased from Aldrich ; 2,3-octanedione 3 was synthesized from 2-octanone according to a previously published method.28 The pH 7 Mac Ilvaine buÜered solution is obtained by dissolving 58.9 g of 3.7 g of citric acid monohydrate and 5.44 Na2HPO4 Æ 12 H2O, g of KCl in distilled water to make one litre of solution.The a-ketols and diols were characterized by their optical rotation, 1H and 13C NMR spectra and comparison with authentic samples previously obtained.7,27,29 Reaction products Microbial reduction by Proteus mirabilis. (a) Reduction of 2,3-pentanedione, 1.After 2 h of incubation, the reaction products were chromatographed on a silica gel column. The eluent was pentane»ether, 80 : 20. The 2S-(])-2- hydroxypentane-3-one is obtained in 63% yield.[a]578 25 \ (c\0.03 eeP98%. 1H NMR (400 MHz, ]44° CHCl3), d: 1.10 (t, 3H), 1.38 (d, 3H), 2.47 (m, 2H), 3.5 (s, 1H CDCl3) exchangeable with 4.22 (q, 1H). 13C NMR (100 MHz, D2O), d: 9 (C5), 20.2 (C1), 31 (C4), 77.9 (C2), 210 (C3). CDCl3) By incubating for 24 h, the (2S,3S) and (2S,3R) diols are formed with an overall yield of 85%. They are puri–ed by chromatography on a silica gel column: eluent ethyl acetate.The diastereoisomer (2S,3S) is predominent (de 88%). (c\0.03 eeP98%. (2S,3S) : 1H NMR [a]578 25 \[9° CHCl3), (400 MHz, d: 0.92 (t, 3H), 1.08 (d, 3H), 1.2»1.6 (m, 2H), CDCl3) 3.1»3.6 (m, 1H), 3.8 (s, 2H exchangeable with 13C NMR D2O). (100 MHz, d: 7.6 (C5), 16.6 (C1), 23.5 (C4), 67.7 (C2), CDCl3) 74.7 (C3). (b) Reduction of 2,3-hexanedione, 2.By following the same procedure, (2S)-(])-2-hydroxyhexane-3-one is isolated in 66% yield after 2 h of incubation. The a-ketol is isolated by chromatography on a silica gel column (eluent ether»pentane, 40 : 60). (c\0.03 eeP98%. 1H [a]578 25 \]50° CHCl3), NMR (400 MHz, d: 0.9 (t, 3H), 1.3 (d, 3H), 1.6 (m, CDCl3) 2H), 2.4 (m, 2H), 3.5 (s, 1H exchangeable with 4.18 D2O), (q, 1H). 13C NMR (100 MHz, d: 13.8 (C6), 17 (C5), CDCl3) 19.7 (C1), 39.3 (C4), 76.7 (C2), 210 (C3). The (2S,3S) diol is obtained in 41% yield after 24 h of incubation and puri–cation on a silica gel column with ethyl acetate as eluent. The diastereomeric and enantiomeric excesses were 88% and 98%. (c\0.03 [a]578 25 \[11° 1H NMR (400 MHz, d: 0.88 (t, 3H), 1.07 (d, CHCl3). CDCl3) 3H), 1.2»1.6 (m, 2H), 3.2»3.6 (m, 2H), 3.8 (s, 2H exchangeable with 13C NMR (100 MHz, d: 13.5 (C6), 17.2 D2O).CDCl3) (C5), 18.6 (C1), 34.9 (C4), 70.0 (C2), 74.9 (C3). (c) Reduction of 2,3-octanedione, 3. After 2 h of incubation, (2S)-(])-2-hydroxyoctane-3-one is isolated in 75% yield after puri–cation on a silica gel column (eluent ether»pentane, 20 : 80). (c\0.03 The diols are never [a]578 25 \]63° CHCl3).formed, whatever the reaction time used. 1H NMR (400 MHz, d: 0.85 (t, 3H), 1.2»1.7 (m, 6H), 1.35 (d, 3H), 2.6 (m, CDCl3) 2H), 3.5 (s, 1H exchangeable with 4.18 (q, 1H). 13C D2O), NMR (100 MHz, d: 14 (C8), 22.6, 23.5, 20 (C1), 31.8 CDCl3) (C7, C6, C5), 37.7 (C4), 76.6 (C2), 210.2 (C3). (d) Reduction of 1-phenylpropane-1,2-dione, 6. The microbial reduction allows (2S)-([)-2-hydroxy-1-phenylpropane-1- one to be obtained in 78% yield.(c\0.03 [a]578 25 \[78° eeP98%. As observed for the 2,3-octanedione, the CHCl3), diols are never formed, even after 24 h of incubation. (e) Reduction of 3,4-hexanedione, 4. Incubation of the diketone for 2 h gave (4S)-(])-4-hydroxyhexane-3-one in 67% yield. The a-ketol is isolated by chromatography on a silica gel column (eluent ether»pentane, 40 : 60).[a]578 25 \]70° (c\0.03 eeP98%. 1H NMR (400 MHz, d: CHCl3), CDCl3) 0.85 (t, 3H), 1.05 (t, 3H), 1.56»1.82 (m, 2H), 2.44 (m, 2H), 3.35 (s, 1H exchangeable with 4.41 (dd, 1H). 13C NMR (100 D2O), MHz, d: 7.6 (C6), 8.9 (C1), 27 (C5), 31.2 (C2), 77.2 (C4), CDCl3) 213 (C3). (3S,4S)-([)-Hexane-3,4-diol is obtained in 85% yield after 24 h of incubation.(c\0.03 [a]578 25 \[16.5 CHCl3), eeP98%. The reaction product was chromatographed on a silica gel column (eluent ethyl acetate). 1H NMR (400 MHz, d: 0.9 (t, 6H), 1.42»1.54 (m, 4H), 3.30 (m, 2H), 4.10 (s, CDCl3) 2H exchangeable with 13C NMR (100 MHz, d: D2O). CDCl3) 10.2 (C1, C6), 26.4 (C2, C5), 75.2 (C3, C4). ( f ) Reduction 1,2-cyclohexanedione, 5. The microbial reduction gave, after 24 h of incubation, (1S,2S)-(])- cyclohexane-1,2-diol as the sole reaction product in 85% yield.Only 3% of the a-ketol has been formed and characterized by GC chromatography for a short reaction time (2 h). The diol is puri–ed by chromatography on a silica gel column with ethyl acetate as eluent. (c\0.03 [a]578 25 \]32° CHCl3), eeP98%. 1H NMR (400 MHz, d: 1.2 (m, 4H), 1.6» CDCl3) 1.82 (m, 4H), 3.2 (m, 2H), 3.85 (2H exchangeable with D2O). 13C NMR (100 MHz, d: 24.5 (C4, C5), 33.1 (C3, C6), CDCl3) 75.3 (C1, C2). Electrochemical reductions. As examples, we will mention the electrochemical reduction of 1-phenylpropane-1,2-dione and cyclohexane-1,2-dione. At the –xed potential of [0.72 V vs. SCE, 1-phenylpropane-1,2-dione gives a mixture of 2-hydroxy-1-phenylpropane-1-one and 1-hydroxy-1-phenylpropane- 2-one in a ratio 8 : 92 and in an overall yield of 85%.The reaction products were chromatographed on a silica gel column (eluent ethyl acetate»cyclohexane, 15 : 85). 2-Hydroxy- 1-phenylpropane-1-one: 1H NMR (400 MHz, d: 1.5 CDCl3) (d, 3H), 3.9 (s, 1H exchangeable with 5.2 (q, 1H), 7.3»8.1 D2O), (m, 5H). 13C NMR (100 MHz, d: 22.3 (C3), 69.4 (C2), CDCl3) 129.1, 129.5, 129.9, 133 (C arom), 202.5 (C1). 1-Hydroxy-1- phenylpropane-2-one: 1H NMR (400 MHz, d: 2.1 CDCl3) (s, 3H), 4.3 (s, 1H exchangeable with 5.1 (s, 1H), 7.3»7.8 D2O), (m, 5H). 13C NMR (100 MHz, d: 23.5 (C3), 80.2 (C1), CDCl3) 127.4, 128.8, 129.1, 136.2 (C arom), 207.3 (C2). (R]S)-2-Hydroxycyclohexane-1-one (adipoine) has been obtained by electrolysis at the –xed potential of [1.50 V vs.SCE. The monomeric a-ketol could not be chromatographed because of the dimerization reaction, which is catalysed by silica gel. 1H NMR (400 MHz, d: 1.63, 2.13 (H5, CDCl3) H5{), 1.50, 2.47 (H3, 1.74, 1.91 (H4, 2.37, 2.58 (H6, H3{), H4{), H6{), 250 New J. Chem., 1998, Pages 247»2513.85 (1H exchangeable with 4.14 (H2). 13C NMR (100 D2O), MHz, d: 23.3 (C4), 27.4 (C5), 36.6 (C3), 39.4 (C6), 75.3 CDCl3) (C2), 211.3 (C1).Dimer: mp\130»131 °C. HRMS: calcd 228.2, exptal 228.1379. 1H NMR [400 MHz, d: (CD3)2SO] 1.1»1.75 (m, 8H), 3.35 (s, 2H exchangeable with 3.87 (dd D2O), 2H). 13C NMR [100 MHz, d: 22.2, 24.1, 27.6, 35.4 (CD3)2SO] (C4, C5, C6, 72.2 (C2, 94.3 (C1, C4{, C5{, C6{), C2{), C1{). References 1 M. P. Doyle and C. T. West, Stereoselective Reductions, Halsted Press, New York, 1976. 2 A. Hajos, Complex Hydrides, Elsevier, Amsterdam, 1979. 3 P. A. McNeil, N. K. Roberts and B. Bosnich, J. Am. Chem. Soc., 1981, 103, 2273 and references therein. 4 M. M. Midland, in Asymmetric Synthesis, ed. J. D. Morrison, Academic Press, San Diego, 1983, vol. 2, p. 45. 5 M. Imuya and H. ZiÜer, J. Org. Chem., 1978, 43, 3319. 6 A. Fauve and H.Veschambre, T etrahedron L ett., 1987, 28, 5037. 7 R. Bel-Rhlid, Thesis, University of Clermont-Ferrand, 1990. 8 A. M. Martre, G. Mousset, S. Fabre and M. Prudhomme, New J. Chem., 1983, 17, 207. 9 S. Torii, P. Liu, N. Bhuvaneswari, C. Amatore and A. Jutand, J. Org. Chem., 1996, 61, 3055 and references therein. 10 N. Sleszynski and P. Zuman, J. Org. Chem., 1987, 52, 2622. 11 (a) J. P. Segretario, N. Sleszynski and P. Zuman, J. Electroanal Chem., 1986, 214, 259. (b) F. A. Long and R. Bekule, J. Am. Chem. Soc., 1963, 85, 2313. 12 (a) M. M. Baizer and H. Lund, Organic Electrochemistry, Marcel Dekker, 3rd edn., 1983. (b) M. M. Baizer, Pure Appl. Chem., 1986, 58, 889 and references therein. 13 (a) S. Servi, Synthesis, 1990, 1. (b) H. L. Holland, Organic Synthesis with Oxidative Enzymes, VCH, New York, 1992. 14 S. Letellier, Electrochim. Acta., 1980, 25,1051. 15 J. P. Segretario and P. Zuman, J. Electroanal. Chem., 1986, 214. 16 J. M. Rodriguez-Mellado, J. L. Avila and J. J. Ruiz, Can. J. Chem., 1985, 63, 891. 17 M. A. Zon and J. M. Rodriguez-Mellado, J. Electroanal. Chem., 1991, 318, 283. 18 M. Fedoronko, J. Konigstein and K. Linek, Collect Czech. Chem. Commun., 1967, 32, 3998. 19 M. R. Montoya, M. A. Zon and J. M. Rodriguez-Mellado, J. Electroanal. Chem., 1993, 353, 217. 20 J. M. Rodriguez-Mellado and M. R. Montoya, J. Electroanal. Chem., 1994, 365, 71; ibid., 1994, 371, 215. 21 M. A. Zon and J. M. Rodriguez-Mellado, J. Electroanal. Chem., 1992, 338, 229. 22 (a) M. D. Ryan and D. H. Evans, J. Electroanal. Chem., 1976, 67, 333. (b) K. Boujlel and J. Simonet, T etrahedron L ett., 1979, 12, 1063. 23 Beilstein, 1st edn., Edwards Brothers Inc., Michigan, 1944, vol. 8, part 2, p. 504. 24 J. Retey and J. A. Robinson, Stereospeci–city in Organic Chemistry and Enzymology, VCH, Weinheim, 1982. 25 C. J. Sih and J. P. Rosazza, in Applications of Biochemical Systems in Organic Chemistry, ed. J. B. Jones, C. J. Sik and D. Perlman, Wiley, New York, 1976, vol. 10, pp. 69»106. 26 H. Izuka and A. Naito, Microbial Conversion of Steroids and Alkaloids, Springer, New York, 1981. 27 R. Bel-Rhlid, A. Fauve, M. F. Renard and H.Veschambre, Biocatalysis, 1992, 6, 319. 28 R. Bel-Rhlid, A. Fauve and H. Veschambre, J. Org. Chem., 1989, 54, 3221. 29 R. Bel-Rhlid, M. F. Renard and H. Veschambre, Bull. Soc. Chim. Fr., 1996, 133, 1011. Received 12th May 1997; Paper 7/08327K New J. Chem., 1998, Pages 247»251 251

ISSN:1144-0546    

DOI:10.1039/a708327k

出版商:RSC    

年代:1998

数据来源: RSC

摘要:

SO2R X 1 SO2R SO2R 2 a R = Me b R = Et c R = Pr d R = Bu e R = Pentyl f R = Hexyl g R = Octyl Electrochemical behaviour of aromatic polysulfones II Cathodic reduction of o-bis(alkylsulfonyl)benzenes in aprotic media in the presence of aliphatic halides Pascal Cauliez, Mohammed Benaskar, Ahmed Ghanimi and Jacques Simonet* L aboratoire dœElectrochimie et (CNRS UMR 6510), de Moleç culaire Macromoleç culaire Universiteç Rennes 1, Beaulieu, 35042 Rennes France ceç dex, o-Bis(alkylsulfonyl)benzenes taken as disulfonylbenzene model substrates lead unexpectedly, under cathodic reduction performed in aprotic media in the presence of an excess of alkyl halide, to an alkylation concomitantly with a monocleavage reaction.The formation mechanism of these alkylated aromatic monosulfones is discussed on the basis of an electron-transfer reaction between the sulfone anion radical and the organic halide followed by a radical coupling.Comportement e ç lectrochimique des polysulfones aromatiques. II. Reç duction cathodique des o-bis(alkylsulfonyl)benze` nes en milieu aprotique en preç sence dœhalogeç nures aliphatiques. Les o-bis(alkylsulfonyl) benze` nes pris comme substrats mode` les des disulfonylbenze` nes conduisent, de inattendue, lors de leur fac” on reç duction cathodique en milieu aprotique en preç sence dœun exce` s dœhalogeç nure dœalkyle, a` une alkylation concomitante a` la reç action de monoclivage.Le meç canisme de formation de ces monosulfones aromatiques alkyleç es, baseç sur une reç action de transfert dœeç lectron entre le radical anion de la sulfone et lœhalogeç nure organique suivi par un couplage radicalaire, est discuteç .Generally, it can be admitted that the sulfonyl group allows some strongly activating properties to be introduced into a given molecular structure.1,2 The interest of sulfone chemistry appears to be reinforced owing to the ability of the arylsulfonyl moiety to be easily introduced and cleaved3,4 afterwards.Thus, the sulfone function allows polarity inversion (umpolung). The deprotection chemistry necessarily includes the cleavage of the right CwS bond for sulfones. Desulfonylation reactions can be carried out using diÜerent methods. Let us quote mild cleavage methods such as hydrogenolysis by means of Raney nickel,5 sodium amalgam,6 as well as alkali metals in amines.7 As a matter of fact, reduction by sodium amalgam in protic media of a phenyl alkyl sulfone led to the alkane and phenylsul–nate.Data relative to the cathodic reduction of regular aromatic sulfones (1, X\H) are now numerous and well documented. 8h10 They all conclude that the sulfone transient radical anion is cleaved. Papers dealing with more elaborate and complex sulfones are much rarer. o-Disulfonylbenzenes, (1, appeared to be of special interest and resemble X\SO2R) the disulfonamides and disul–nic esters already studied by Horner and Schmitt.11 More particularly, the novelty of the work reported by Novi et al.,12 which showed the catalytic conversion in the presence of a strong base of a disubstituted o-bis(alkylsulfonyl)benzene into a cyclic monosulfone, prompted us to test the general behaviour of sulfones 2 where the R substituents are now exclusively alkyl groups.The work reported here aims to demonstrate the unexpected reactivity of sulfones 2a»g, where besides the expected cleavage reaction, a possible radical coupling implying R~ could be taken into account in the reduction process. In the present full paper, and similarly to the work already reported13 concerning a complete series of o-disulfones, classic experimental conditions were chosen: aprotic dipolar solvents, essentially dimethylformamide (DMF) as it is known not to be a strong H-atom donor solvent, quaternary ammonium salts as the electrolyte and a mercury pool cathode.Results Voltammetry Linear sweep voltammetry achieved with the whole 2 series at slow sweep rate (v) exhibited a –rst reversible step within the range of [1.0 V to [1.5 V vs. the Ag/AgI/0.1 M I~ system, followed by two other steps (see Fig. 1). Voltammetric data concerning the two main steps are gathered in Table 1. Coulometry Coulometry experiments were performed to determine accurately the number of electrons involved in the overall electrochemical process (see Table 2).For a total consumption of the starting material an electricity amount between 1.25 and 1.65 moles of electrons per mole of sulfone appeared to be necessary. In Fig. 2, it can be seen that micropotentiostatic electrolysis at a small area mercury pool cathode did not bring about the disappearance of the second and the third step.The in situ addition of methyl iodide»currently used in sulfone chemistry to trap a possible sul–nic anion issued from the two-electron cleavage of the relevant sulfone»led to the emergence of a reversible step that strongly resembles that of the starting material. However, its current intensity is less than half of that already observed with 2 before the coulometry experiments. It is noticeable that step 2 also increased after the methylation process. New J.Chem., 1998, Pages 253»261 25312 9 6 3 0 i / mA 2e a b c E / V –2.0 –1.5 –1.0 –0.5 0 2 4 6 8 10 i / mA 2e 1 2 3 a b c E / V –2.0 –1.5 –0.5 Fig. 1 Cyclic voltammetry of 2e at various concentrations in DMF» 0.1 M Stationary mercury microelectrode (area 0.8 mm2). Bu4NBF4 . Reference system: Ag/AgI/0.1 M I~.Sweep rate : v\100 mV s~1. a: C\2.41]10~3 M, b: C\3.85]10~3 M, c : C\5.29]10~3 M Macroelectrolysis All macroelectrolyses, performed on the entire 2 series at a stirred mercury pool cathode, aÜorded only four major products, as shown in Scheme 1. Data concerning the macro- Fig. 2 Cyclic voltammetry of 2e in DMF»0.1 M Station- Bu4NBF4 . ary mercury microelectrode (area 0.8 mm2).Reference system: Ag/ AgI/0.1 M I~1. Sweep rate : v\100 mV s~1. a: C\4.8]10~3 M, b: solution a after coulometry at [1.19 V until null current, c : solution b after addition of an excess of methyl iodide reductions of 2 are gathered in Table 2. In all cases, compounds 3 were produced in larger amounts than 4. Yields of 5 (or at least that of the corresponding methyl sulfone by means of methyl iodide during the work-up) were generally found to be roughly equivalent to the sum of the yields of 3 and 4.Totally cleaved compounds 6 could be considered as minor side products. Their relative yields were less than 15% in most cases. Table 1 Features of the cyclic voltammetric response with disulfones 2a First step Second step Concentration Substrate /103 M E°/Vb Ep1 /V i1/lA Ep2 /V i2/lA 2a 4.3 [1.30 [1.33 9.5 [1.89 7.1 2b 4.8 [1.12 [1.15 7.6 [1.70 7.4 2c 4.92 [1.14 [1.17 9.1 [1.76 7.5 2d 4.0 [1.13 [1.16 5.3 [1.72 5.1 2e 4.62 [1.15 [1.18 8.5 [1.75 7.8 2f 4.28 [1.15 [1.18 8.2 [1.77 6.0 2g 4.70 [1.19 [1.22 8.6 [1.78 6.8 a Stationary mercury microelectrode (area 0.8 mm2).Electrolyte : DMF]0.1 M Reference system: Ag/AgI/0.1 M I~. Sweep rate : 0.1 Bu4NBF4 V s~1.b Standard potential relative to the –rst electron transfer in DMF. Table 2 Macroelectrolysis of disulfones 2 at a stirred mercury pool cathodea Fixed Isolated yields/% Substrate potential n/ Entry 2b /V F mol~1 Solvent 3 4 5c 6 1 2ad [1.32 1.42 aprotic DMF \1 [ 75 10 2 2ae [1.32 1.25 aprotic DMF 15 4 20 12 3 2b [1.18 1.34 aprotic DMF 33 2 30 13 4 2b [1.18 1.12 DMF]PhSNBu4 f 42 \1 » » 5 2b [1.18 1.45 DMF]Phenolg 22 2 28 28 6 2c [1.20 1.40 aprotic DMF 30 2 30 12 7 2d [1.21 1.31 aprotic DMF 32 2 34 11 8 2d [1.21 1.10 DMF]AcONBu4 h 44 \1 » » 9 2e [1.19 1.65 aprotic DMF 32 2 30 10 10 2f [1.17 1.52 aprotic DMF 34 4 33 7 a Electrode area : 10 cm2.Solvent : DMF or DMSO containing 0.1 M Potentials are referred to the Ag/Ag I/0.1 M I~system in DMF.Bu4NBF4 . b Concentration of 10~2 mol l~1 except as noted. c Sul–nate was alkylated with methyl iodide in excess added directly to the catholyte. Yields reported in this column are those relevant to the obtained disulfone. d Concentration of 8.6]10~3 mol l~1. e Concentration of 2.2]10~2 mol l~1. f Molar excess factor towards 2b is 20. g Molar excess factor towards 2b is 4.h Molar excess factor towards 2d is 6. 254 New J. Chem., 1998, Pages 253»261SO2R R SO2R R 3 4 cathodic reduction 1.10–1.65 F mole–1 2 SO2R SO2 R H 5 6 SO2 (–) Bu SO2Bu Et SO2Bu Bu SO2Et Et SO2Et 8% 9% 8% 8% SO2R R¢ SO2R R¢ 8 9 cathodic reduction 2.20–3.90 F mole–1 2 + R¢X SO2R SO2 R H 10 6 SO2R R¢ Scheme 1 Mixed electrolysis of two diÜerent disulfones A mixture of 2b and 2d in equal amounts in the catholyte (C\1.2]10~2 M) was electrolysed at [1.2 V and consumed 1.4 moles of electrons per total sulfone amount. A mixture of four main compounds in more or less equal proportions could be isolated : Reduction of 2 in the presence of alkyl halides Preliminary experiments13d demonstrated that disulfones 2 also led to the alkylation products 8 and 9 in high yields when reduced in the presence of alkyl halides (R@X) in large excess.The chosen R@X were not electrochemically reduced prior to the disulfones. Alkylation in the meta position was found to be strongly favored. A minor ortho-alkylation reaction was concomitantly observed. Here, in most of the cases, the yield for the overall alkylation was quite high (at least 70»80% of isolated compounds).Under these experimental conditions (large excess of R@X), cleavage of disulfones leading to sul–nate 5 was not observed (see Table 3) with nonbulky R@ groups. Additionally, it is worth mentioning that greater electricity consumptions are readily consistent with a homogeneous indirect reduction of the alkyl halide by the 2 anion radical. The scope of this study has been extended to a large palette of primary and secondary alkyl halides as reported in Table 4 and all results are approximately the same.However, when the alkyl group R@ was particularly bulky (like the ButCH2), alkylation yield dropped drastically and self-alkylation with Scheme 2 the alkyl group R of the disulfone, then yielding 3 and 4, became predominant. Secondary products (such as 10, see Scheme 2) were also isolated (see Table 4, entries 1, 3 and 5). Alkyl a,x-dihalides have also been used as alkylating agents in these experiments. Such compounds might have been expected to undergo a second coupling, either intramolecularly to yield bicyclic derivatives or by reacting with another disulfone.Such reactions were absolutely not observed and monoalkylations remained the only reaction pathways (see Table 5, entries 2»6).Additional important information about the alkylation mechanism could be obtained from the selectivity observed in the case of an optically active halide. The inversion-toracemization ratio for the reaction between aromatic radical anions and chiral alkyl halides has been discussed in terms of a competition between a classical pathway (leading to SN2 complete con–guration inversion) and an electron transfer followed by a coupling between the alkyl radical and the aromatic anion radical.14 In our case (see Table 5, entry 1), a total racemization occurred, ruling out the reaction.SN2-type Hexen-5-en-1-yl radical was –rst reported in 1963 to undergo an intramolecular ring closure leading to the cyclopentylmethyl radical.15 Since then, this species and its derivatives, often referred to as radical probes or radical clocks, have gained widespread acceptance as mechanistic tools to provide evidence of the radical character of substitution reactions. 16 x-Bromoalk-1-enes, potential precursors of x-alk-1- enyl radicals, behaved similarly to alkyl halides when they were added during the electrochemical reduction of disulfones 2: when added in excess in cyclic voltammetry, they induced the loss of reversibility of the disulfone peaks without noticeable increase of current, and preparative scale electrolyses yielded meta- and ortho-alkylated sulfones of type 8 and 9 (see Table 5, entry 7).It is also worth noting that the cyclized derivative 12 was not obtained in this experiment, as well as in similar ones13c not reported in the present paper.Table 3 Electrolysis of 2b in the presence of some R@X compoundsa Isolated Substrate R@X cb n/F mol~1 electrolysis products/% 2b BuBr 20 4.1 2b BuI 5 4.3 2b Octyl-I 3 3.3 a Solvent : DMF containing 0.1 M Cathode: stirred mercury pool (area 10 cm2). Substrate concentration at the start of the electro- Bu4NBF4 .lysis : 1.2]10~2 M. Applied voltage : [1.20 V.b Excess factor (molar ratio [R@X]/[2]). New J. Chem., 1998, Pages 253»261 255SO2R SO2R ( ) n 8, 9 12 Table 4 Macroelectrolysis of various disulfones 2 in the presence of alkyl halides at a stirred mercury pool cathodea Substrate Alkyl halide Fixed n/ Entry 2 R@X (cb) potential/V F mol~1 Isolated yields/% 1 2e Octyl-I [1.20 2.8 (3) 2 2d Heptyl-Br [1.25 2.2 (16) 3 2e C18H37Br [1.30 2.5 (8) 4 2d ButCH2Br [1.30 2.0 (20) 5 2e PriBr [1.20 2.3 (10) 6 2f BuiBr [1.28 2.5 (4) a Electrode area : 10 cm2.Solvent : DMF containing 0.1 M Potentials are referred to the Ag/AgI/0.1 M I~ system in DMF. b Molar Bu4NBF4 . excess factor towards 2. Discussion The cathodic decomposition of disulfones 2 appears to be an alkylation reaction concomitant with a cleavage reaction, leading to a global scheme that is really unexpected.Kinetic studies13c in UV-visible spectroscopy that followed the concentration decay of the anion radical of 2 in solution have demonstrated that this decomposition obeys a second-order kinetic law, likely with a charge-transfer complex between the anion radical and the substrate prior to decomposition.In contrast, when the reduction of disulfones 2 was performed in the presence of alkyl halides, the global mechanism seems to –t quite well with previously reported results17 concerning the indirect reduction of alkyl halides (RX) in the case of homogeneous electron transfer (ET) by means of the anion radical of the mediator (A).Thus, the reaction scheme below describes a nondissociative ET but dissociative ET can also be taken into account. A A8B e~ A~~ A~~]RX A8B A][RX]~~ [RX]~~ »»»’ kC R~]X~ A radical can be formed in solution and then further couple R~ with the mediator anion radical in the general case. R~]A~~]AR~ In the present case, issued from the cathodic cleavage of R~ couple with The mechanism is obviously rendered 2~~can 2~~.more complex since 2 intrinsically possesses two potential leaving groups. 2~~]R~ »»»’ [2wR]~ »»»’ 3]4 å Tentatively, a general mechanism can be proposed (see Scheme 3). It takes into account the fact that the alkylation cannot be intramolecular as shown by the results of mixed electrolyses (example described with 2b]2d), which are in full agreement with the statistical coupling of four radical species leading to four dimers in equal amounts.Therefore, the ì self-alkylation œ observed in the decomposition of 2 under electron transfer (the total alkylation yield cannot exceed 50% owing to formation of the sul–nate 5) can also be achieved with success in the presence of an excess of alkyl halide R@X (obviously R@ can be equal to R). Thus, compounds 3 and 4 on the one hand and 8 and 9 on the other hand could be obtained but in much higher yields.With structure 2 only two sites seem to be suitable to yield a radical coupling: the ipso-type coupling may explain, owing to the bulkiness of the sulfonyl group, the much lower amount of –nal ortho-substituted compounds 4 and 9 than those observed with the meta-substituted compounds 3 and 8.With neopentyl bromide, only the meta derivative could be obtained (Table 4, entry 4), suggesting (which is reasonable) that the neopentyl radical could not reach the carbon bearing 256 New J. Chem., 1998, Pages 253»261SO2R SO2R SO2R SO2R •– SO2R SO2R •– SO2 – SO2R + R • R • + SO2R SO2R •– R H SO2R SO2R + SO2R SO2R R 1 2 3 4 5 6 R SO2R SO2R H – – –SO2R – 3 10 –SO2R – 4 – e– kC Table 5 Macroelectrolysis of disulfones 2 in the presence of alkyl halides or dihalides at a stirred mercury pool cathodea Substrate Alkyl halide R@X Fixed n/ Entry 2 (cb) potential/V F mol~1 Isolated yields/% 1 2e (R)- [1.20 2.3 C6H13CH*(CH3)Br (2) 2 2b Br(CH2)4Br [1.15 3.9 (4) 3 2c Br(CH2)4Br [1.15 2.3 (16) 4 2b Br(CH2)6Br [1.18 1.7 (8) 5 2c Br(CH2)10Br [1.20 3.3 (10) 6 2d (^)-BrCH2CH(CH3)CH2Cl [1.20 3.7 (8) 7 2b [1.25 2.8 (20) a Electrode area : 10 cm2.Solvent : DMF containing 0.1 M Potentials are referred to the Ag/AgI/0.1 M I~ system in DMF. b Molar Bu4NBF4 . excess factor towards 2. Scheme 3 a sulfonyl moiety. In this case, the coupling is anyway so slow that the regular decomposition of 2 may occur. Optically active R@X (Table 5, entry 1) allowed us to check the radical-coupling nature of the key reaction14 since 100% racemization was observed.Similarly (entry 7), the use of radical probes15,16 did not allow any cyclization of the expected hex-5-enyl primary free radical to be seen. This suggests a very fast deactivation by coupling of this radical. Unexpectedly, entries 2»6 (Table 5) relative to the use of a,xdibromoalkanes as R@X demonstrated the impossibility of getting twinned alkylation reactions.All these previous results could suggest the existence of an intermediary complex between and R@X [the existence of 2~~ which was con–rmed by the disappearance of reversibility in the CV of 2 when R@X (see Fig. 3) was added, although no current increase was observed13d].Under these conditions, the alkylation process could occur essentially inside a solvent cage: 2~~]R@X A8B [2, R@X]~~ »»»’ [2, R@~] »»»’ fast 2’~ 2wR@~]2 New J. Chem., 1998, Pages 253»261 257E / V ab c 2f i / mA –0.5 –1.0 –1.4 0 2 4 6 Br Br + 2 RSCu SR SR + 2 CuBr quinoline pyridine Fig. 3 Cyclic voltammetry of 2f in DMF»0.1 M Station- Bu4NBF4 . ary mercury microelectrode (area 0.8 mm2).Reference system: Ag/ AgI/0.1 M I~. Sweep rate : v\100 m V s~1. Curve morphology in the presence of increasing concentration of R@X\Pri Br with c the molar excess factor toward 2f. a c\0, b c\2, c c\20 Lastly, the general formation, besides 8 and 9, of compounds 10 (see Table 4, entries 1, 3 and 5) appears fully compatible with the main coupling in the 4 position between 2~~and R@~ (see Scheme 3).Under these conditions, the anion is protonated and readily rearomatized during the work-up. Experimental The products were identi–ed mainly by 60 MHz (Varian EM 360 A), 200 MHz (Bruker ARX 200) or 300 MHz (Bruker AC 300 P) 1H NMR and by high resolution mass spectrometry (Varian MAT 311 at the Centre de Mesures Physiques de lœOuest, Universiteç de Rennes 1).In the cases where an additional structural proof was essential, 13C NMR (75 MHz) spectra were obtained and/or selective irradiation performed. All the NMR spectra were recorded in with TMS as CDCl3 an internal reference. Details on the determination of the alkylation position in compounds 3 and 4 have already been reported.13d Syntheses of disulfones 2 Several modes of synthesis could be used to get the whole 2 series ; the oxidation of the corresponding dithioethers either by a solution (30%) of hydrogen peroxide in acetic acid18 or by using the method of Trost and Curran19 (50% oxone solution in methanol) was chosen.Most of the dithioethers were known and have been prepared according to the following procedure.20 The –rst step consists of forming cuprous salts from the chosen thiols : 2 RSH]Cu2O]2 RSCu]H2O followed by the second synthesis step corresponding to the substitution of o-dibromobenzene to lead to the dithioethers as key products: However, with disulfone 2a, the use of another method appeared necessary : 1,2-benzenedithiol was alkylated with methyl iodide to yield the corresponding dithioether, as previously described.18c,21 Physical constants of compound 2a22 have been reported.Those of the unknown disulfones 2 are detailed below: 2b: mp: 123»125 °C. MS: exact mass measurement: obsd, 262.0333; calcd for 262.0333. 200 MHz 1H C10H14O4S2 , NMR d: 1.29 (6H, t, J\7.5 Hz); 3.69 (4H, q, J\7.5 Hz); 7.83»7.94 (2H, m) and 8.26»8.38 (2H, m). 2c: mp: 107»108 °C. 200 MHz 1H NMR d: 1.03 (6H, t, J\7.4 Hz); 1.77 (4H, tq, J\7.4 and 7.9 Hz); 3.65 (4H, t, J\7.9 Hz); 7.86»8.02 (2H, m); 8.27»8.42 (2H, m). 2d: mp: 86»87 °C. MS: exact mass measurement: obsd, 318.1108; calcd for 318.0960. 200 MHz 1H C14H22O4S2 , NMR d: 0.90 (6H, m, J\7.2 Hz); 1.42 (4H, tq, J\6.9 and 7.2 Hz); 1.70 (4H, tt, J\6.9 and 8.0 Hz); 3.65 (4H, t, J\8.0 Hz); 7.80»7.92 (2H, m); 8.25»8.37 (2H, m). 2e: mp: 69»70 °C.MS: exact mass measurement: obsd, 346.1266; calcd for 346.1272. 300 MHz 1H C16H26O4S2 , NMR d: 0.88 (6H, t, J\7.0 Hz); 1.25»1.42 (8H, m); 1.66»1.79 (4H, m); 3.63 (4H, t, J\8.0 Hz); 7.81»7.91 (2H, m); 8.26»8.37 (2H, m). 2f: mp: 38»39 °C. MS: exact mass measurement: obsd, 374.1595; calcd for 374.1585. 200 MHz 1H C18H30O4S2 , NMR d: 0.85 (6H, t, J\6.5 Hz); 1.11»1.50 (12H, m); 1.60» 1.82 (4H, m); 3.64 (4H, t, J\8.0 Hz); 7.80»7.91 (2H, m); 8.25» 8.37 (2H, m). 2g: liquid. MS: exact mass measurement: obsd, 430.2592; calcd for 430.2212. 60 MHz 1H NMR d: 1.3 (30 C22H38O4S2 , H, m); 3.7 (4H, t) ; 7.9 (2H, m); 8.3 (2H, m). Electrolyses All the electrolyses were achieved with a large H-shaped cell (total volume: 150 mL) equipped with a G4 glass frit separator.The stirred mercury pool cathode potential was controlled with an EGG potentiostat Model 173. The reference electrode was in all cases an Ag/AgI/0.1 M system in INBu4 DMF. The counter electrode was always a glassy carbon plate (total area : about 15 cm2). An argon atmosphere was maintained over the electrolysis solution in all cases. Amounts of disulfones 2 electrolysed until null current were of the order of 200 to 300 mg. An alkyl halide R@X could be (or not) added in the cathode compartment.Excess factors (c) as described in Tables 3 and 4 aim to quantify the relative amount of alkylating reagents. After electrolysis completion, the entire contents of the cathodic compartment were poured into 150 mL of distilled water. The resulting solution was extracted with diethyl ether.The ethereal phase was then washed twice with water and dried over magnesium sulfate. After evaporation under vacuum, the dried extract was separated by column chromatography (silica gel, eluent : ethyl acetate»pentane). Electrolysis products Electrolysis of 2a. Compound 3a was identi–ed as m-tolyl methyl sulfone by comparison of the NMR data with those reported in the literature.23 Compound 4a was identi–ed as o-tolyl methyl sulfone by comparison of the NMR data with those reported in the literature. 23,24 Compound 5a: In order to avoid confusion with the starting compound, sul–nate was trapped with ethyl iodide. It has been checked that addition of methyl iodide to the electrolysis mixture yielded 2a. 5a was identi–ed as o-(methylsulfonyl)- phenyl ethyl sulfone by comparison of the NMR data with those reported in the literature.25 Mp: 105»106 °C (lit.25 100» 102 °C).Compound 6a was identi–ed as methyl phenyl sulfone by comparison with an authentic sample. Electrolysis of 2b. Compound 3b: liquid. MS: exact mass measurement: obsd, 198.0721; calcd for 198.0715. C10H14O2S, 200 MHz 1H NMR d: 1.28 (3H, t, J\7.6 Hz); 1.28 (3H, t, J\7.5 Hz); 2.75 (2H, q, J\7.6 Hz); 3.10 (2H, q, J\7.5 Hz); 258 New J.Chem., 1998, Pages 253»2617.40»7.53 (2H, m); 7.66»7.77 (m, 2H). Compound 4b: liquid. MS: exact mass measurement: obsd, 198.0702; calcd for 198.0715. 200 MHz 1H NMR C10H14O2S, d: 1.30 (3H, t, J\7.4 Hz); 1.33 (3H, t, J\7.5 Hz); 3.06 (2H, q, J\7.4 Hz); 3.17 (2H, q, J\7.5 Hz); 7.33 (1H, ddd, 3J\7.7 and 7.7 Hz, 4J\1.7 Hz); 7.36 (1H, dd, 3J\7.6 Hz, 4J\1.7 Hz); 7.54 (1H, ddd, 3J\7.7 and 7.7 Hz, 4J\1.3 Hz); 7.98 (1H, dd, 3J\7.7 Hz, 4J\1.3 Hz).Compound 5b is the same as 5a above. Compound 6b was identi–ed as ethyl phenyl sulfone by comparison with an authentic sample. Electrolysis of 2c. Compound 3c: liquid. 60 MHz 1H NMR d: 0.8»1.2 (6H, m); 1.3»2.1 (m, 4H); 2.7 (2H, t) ; 3.1 (2H, t) ; 7.4»7.9 (4H, m).Compound 4c: liquid. 300 MHz 1H NMR d: 0.86»0.95 (6H, m); 1.5»1.8 (4H, m); 3.0 (2H, t, J\8 Hz) ; 3.10 (2H, t, J\8 Hz); 7.30»8.05 (4H, m). Compound 5c: mp: 86»87 °C. MS: exact mass measurement: obsd, 262.0333; calcd for 262.0334. 60 C10H14O4S2 , MHz 1H NMR d: 1.2 (3H, t, J\7.5 Hz); 1.8»1.9 (2H, m); 3.45 (3H, s) ; 3.65 (2H, t, J\7.5 Hz); 8.0»8.5 (4H, m). Compound 6c was identi–ed as propyl phenyl sulfone by comparison with an authentic sample. Electrolysis of 2d.Compound 3d: liquid. MS: exact mass measurement: obsd, 254.1340; calcd for 254.1341. C14H22O2S, 200 MHz 1H NMR d: 0.89 (3H, t, J\7.3 Hz); 0.93 (3H, t, J\7.3 Hz); 1.37 (4H, m); 1.62 (m, 4H); 2.71 (2H, t, J\7.5 Hz); 3.09 (2H, t, J\8.0 Hz); 7.43»7.50 (2H, m); 7.65»7.77 (m, 2H).Compound 4d: liquid. 200 MHz 1H NMR d: 0.89 (3H, t, J\7.3 Hz); 0.96 (3H, t, J\7.3 Hz); 1.37 (4H, m); 1.63 (m, 4H); 2.70 (2H, t, J\7.5 Hz); 3.10 (2H, t, J\8.0 Hz); 7.35 (1H, ddd, 3J\7.7 and 7.7 Hz, 4J\1.7 Hz); 7.39 (1H, dd, 3J\7.7 Hz, 4J\1.7 Hz); 7.56 (1H, ddd, 3J\7.7 and 7.7 Hz, 4J\1.3 Hz); 7.99 (1H, dd, 3J\7.7 Hz, 4J\1.3 Hz). Compound 5d: mp: 72.5»75 °C.18c MS: exact mass measurement: obsd, 276.0864; calcd for 276.0490.C11H16O4S2 , 60 MHz 1H NMR d: 1.2 (7H, m); 3.45 (3H, s) ; 3.6 (2H, t) ; 8.0 (4H, m). Compound 6d was identi–ed as butyl phenyl sulfone by comparison with an authentic sample. Electrolysis of 2e. Compound 3e: liquid. MS: exact mass measurement: obsd, 282.1654; calcd for 282.1654. C16H26O2S, 60 MHz 1H NMR d: 0.7»1.1 (6H, m); 1.2»2.0 (12H, m); 2.75 (2H, t) ; 3.15 (2H, t) ; 7.35»7.90 (4H, m).Compound 4e: liquid. MS: exact mass measurement: obsd, 282.1654; calcd for 282.1654. 300 MHz 1H NMR C16H26O2S, d: 0.82»0.94 (6H, m); 1.22»1.43 (m, 8H); 1.62»1.79 (m, 4H); 3.0 (t, 2H, J\8.0 Hz); 3.11 (t, 2H, J\8.0 Hz); 7.3»8.02 (m, 4H). Compound 5e: liquid. MS: exact mass measurement: obsd, 290.1182; calcd for 290.0647. 60 MHz 1H NMR C12H18O4S2 , d: 1.15 (9H, m); 3.45 (3H, s) ; 3.65 (2H, t) ; 7.8»8.2 (4H, m).Compound 6e was identi–ed as pentyl phenyl sulfone by comparison with an authentic sample. Electrolysis of 2f. Compound 3f : liquid. MS: exact mass measurement: obsd, 310.1975; calcd for 310.1967. C18H30O2S, 300 MHz 1H NMR d: 0.85 (3H, t, J\6.9 Hz); 0.88 (3H, t, J\6.8 Hz); 1.17»1.42 (12H, m); 1.56»1.77 (4H, m); 2.69 (2H, t, J\7.7 Hz); 3.07 (2H, t, J\8.0 Hz); 7.42»7.50 (2H, m); 7.67»7.75 (2H, m). 75 MHz 13C NMR d: 13.91 ; 14.05 ; 22.29 ; 22.56 ; 22.62 ; 27.96 ; 28.84 ; 31.18 (2C) ; 31.61 ; 35.73 ; 56.37 ; 125.35 ; 127.74 ; 129.14 ; 133.74 ; 139.13 ; 133.58. Compound 4f : liquid. MS: exact mass measurement: obsd, 310.1975; calcd for 310.1967. 300 MHz 1H NMR C18H30O2S, d: 0.85 (3H, t, J\6.8 Hz); 0.89 (3H, t, J\7.0 Hz); 1.16»1.49 (12H, m); 1.59»1.77 (4H, m); 3.00 (2H, t, J\8.0 Hz); 3.11 (2H, t, J\8.0 Hz); 7.36 (1H, ddd, 3J\7.7 and 7.9 Hz; 4J\1.4 Hz); 7.38 (1H, dd, 3J\7.7 Hz, 4J\1.4 Hz); 7.54 (1H, ddd, 3J\7.7 and 7.7 Hz, 4J\1.3 Hz); 7.98 (1H, dd, 3J\7.9 Hz, 4J\1.3 Hz).Compound 5f : liquid. 60 MHz 1H NMR d: 1.15 (11H, m); 3.50 (3H, s) ; 3.75 (2H, t) ; 8.0»8.3 (4H, m).Compound 6f was identi–ed as hexyl phenyl sulfone by comparison with an authentic sample. Electrolysis of 2g. Compound 3g: liquid. MS: exact mass measurement: obsd, 366.2602; calcd for 366.2593. C22H38O2S, 60 MHz 1H NMR d: 1.30 (30H, m); 2.90 (4H, m); 7.60 (4H, m). Compound 5g was identi–ed as o-(methylsulfonyl)phenyl octyl sulfone by comparison of the NMR data with those reported in the literature.26 Compound 6g was identi–ed as octyl phenyl sulfone by comparison with an authentic sample.Electrolysis of 2e in the presence of octyl iodide. Compound 8e (R@\octyl) : liquid. MS: exact mass measurement: obsd, 324.2129; calcd for 324.2123. 300 MHz 1H NMR C19H32O2S, d: 0.86 (3H, t, J\7.1 Hz); 0.88 (3H, t, J\6.3 Hz); 1.19»1.40 (14H, m); 1.58»1.78 (4H, m); 2.69 (2H, t, J\7.7 Hz); 3.07 (2H, t, J\8.1 Hz); 7.40»7.52 (2H, m); 7.66»7.77 (2H, m). 75 MHz 13C NMR d: 13.68 ; 14.08 ; 22.11 ; 22.33 ; 22.66 ; 29.21 (3C) ; 30.38 ; 31.20 ; 31.85 ; 35.74 ; 56.34 ; 125.35 ; 127.74 ; 129.14 ; 133.74 ; 139.15 ; 144.58. Compound 9e (R@\octyl) : liquid. MS: exact mass measurement: obsd, 324.2129; calcd for 324.2123. 200 C19H32O2S, MHz 1H NMR d: 0.86 (3H, t, J\6.9 Hz); 0.88 (3H, t, J\6.5 Hz); 1.15»1.44 (14H, m); 1.54»1.80 (4H, m); 3.00 (2H, t, J\7.7 Hz); 3.11 (2H, t, J\8.0 Hz); 7.36 (1H, ddd, 3J\7.6 and 8.1 Hz, 4J\1.3 Hz); 7.38 (1H, dd, 3J\7.6 Hz, 4J\1.3 Hz); 7.51 (1H, ddd, 3J\7.6 and 7.6 Hz, 4J\1.5 Hz), 7.98 (1H, dd, 3J\8.1 Hz, 4J\1.5 Hz). 50 MHz 13C NMR d: 13.70 ; 14.10 ; 22.12 ; 22.19 ; 22.67 ; 29.25 ; 29.46 ; 29.80 ; 30.43 ; 31.87 ; 32.24 ; 33.06 ; 56.43 ; 126.34 ; 130.37 ; 131.60 ; 133.46 ; 137.02 ; 143.22.Compound 10e (R@\octyl) : liquid. 300 MHz 1H NMR d: 0.84 (3H, t, J\6.8 Hz); 0.88 (6H, t, J\6.8 Hz); 1.15»1.47 (18H, m); 1.51»1.78 (6H, m); 2.79 (2H, t, J\7.6 Hz); 3.56» 3.68 (4H, m); 7.61 (1H, dd, 3J\8.0 Hz, 4J\1.8 Hz); 8.09 (1H, d, 4J\1.8 Hz); 8.19 (1H, d, 3J\8.0 Hz). Electrolysis of 2d in the presence of heptyl bromide.Compound 8d (R@\heptyl) : liquid. 300 MHz 1H NMR d: 0.88 (3H, t, J\6.9 Hz); 0.89 (3H, t, J\7.3 Hz); 1.19»1.50 (10H, m); 1.58»1.75 (4H, m); 3.08 (2H, t, J\8.0 Hz); 3.32 (2H, t, J\7.7 Hz); 7.67»7.75 (2H, m); 7.43»7.51 (2H, m). 75 MHz 13C NMR d: 13.50 ; 14.07 ; 21.56 ; 22.63 ; 24.65 ; 29.07 ; 29.14 ; 31.19 ; 31.76 ; 35.73 ; 56.13 ; 125.36 ; 127.74 ; 129.14 ; 133.74 ; 139.15 ; 144.59.Compound 9d (R@\heptyl) : liquid. 300 MHz 1H NMR d: 0.88 (3H, t, J\6.8 Hz); 0.89 (3H, t, J\7.3 Hz); 1.20»1.51 (10H, m); 1.57»1.80 (4H, m); 3.00 (2H, t, J\8.0 Hz); 3.12 (2H, t, J\8.1 Hz); 7.36 (1H, ddd, 3J\7.5 and 7.9 Hz, 4J\1.3 Hz); 7.38 (1H, dd, 3J\7.5 Hz, 4J\1.3 Hz); 7.54 (1H, ddd, 3J\7.5 and 7.5 Hz, 4J\1.4 Hz), 7.98 (1H, dd, 3J\7.9 Hz, 4J\1.3 Hz). 75 MHz 13C NMR d: 13.68 ; 14.08 ; 22.11 ; 22.33 ; 22.66 ; 29.21 ; 30.38 ; 31.20 ; 31.85 ; 35.74 ; 56.34 ; 125.35 ; 127.74 ; 129.14 ; 133.74 ; 139.15 ; 144.58.Electrolysis of 2e in the presence of octadecyl bromide. Compound 8e liquid. MS: exact mass measure- (R@\C18H37) : ment: obsd, 464.3684; calcd for 464.3688. 300 C29H52O2S, MHz 1H NMR d: 0.86 (3H, t, J\7.0 Hz); 0.88 (3H, t, J\6.7 Hz); 1.12»1.50 (36H, m); 1.57»1.78 (4H, m); 2.69 (2H, t, J\7.7 Hz); 3.07 (2H, t, J\8.1 Hz); 7.42»7.50 (2H, m); 7.67» 7.75 (2H, m). 75 MHz 13C NMR d: 13.69 ; 14.12 ; 22.11 ; 22.33 ; New J. Chem., 1998, Pages 253»261 25922.70 ; 29.21 ; 29.38 ; 29.44 ; 29.58 ; 29.68 (3C) ; 29.72 (6C) ; 30.38 ; 31.21 ; 31.95 ; 35.74 ; 56.74 ; 125.34 ; 127.74 ; 129.13 ; 133.73 ; 139.15 ; 144.58. Compound 9e liquid. 200 MHz 1H NMR d: (R@\C18H37) : 0.86 (3H, t, J\6.8 Hz); 0.89 (3H, t, J\6.8 Hz); 1.13»1.39 (36H, m); 1.53»1.79 (4H, m); 3.00 (2H, t, J\8.1 Hz); 3.07 (2H, t, J\8.0 Hz); 7.36 (1H, ddd, 3J\7.5 and 7.9 Hz, 4J\1.4 Hz); 7.38 (1H, dd, 3J\7.5 Hz, 4J\1.4 Hz); 7.51 (1H, ddd, 3J\7.5 and 7.5 Hz, 4J\1.7 Hz), 7.98 (1H, dd, 3J\7.9 Hz, 4J\1.7 Hz).Compound 10e liquid. MS: exact mass mea- (R@\C18H37) : surement: obsd, 598.4106; calcd for 598.4089.C34H63O4S2 , 200 MHz 1H NMR d: 0.86 (3H, t, J\6.8 Hz); 0.89 (6H, t, J\6.8 Hz); 1.13»1.39 (38H, m); 1.53»1.79 (6H, m); 2.78 (2H, t, J\7.6 Hz); 3.61 (2H, t, J\7.8 Hz); 3.63 (2H, t, J\7.8 Hz); 7.60 (1H, dd, 3J\8.0 Hz, 4J\1.8 Hz); 8.08 (1H, d, 4J\1.6 Hz); 8.18 (1H, d, 3J\8.0 Hz). Electrolysis of 2d in the presence of neopentyl bromide.Compound 8d liquid. 200 MHz 1H NMR d: 0.88 (R@\CH2But) : (3H, t, J\7.2 Hz); 0.91 (9H, s) ; 1.22»1.52 (2H, m); 1.54»1.78 (2H, m); 2.59 (2H, s) ; 3.09 (1H, t, J\8.0 Hz); 7.35»7.52 (2H, m); 7.65»7.78 (2H, m). Compound 3d, 4d and 6d: see above. Electrolysis of 2e in the presence of isopropyl bromide. Compound 8e (R@\Pri) : liquid. MS: exact mass measurement: obsd, 254.1343; calcd for 254.1340. 300 MHz 1H C14H22O2S, NMR d: 0.86 (3H, t, J\6.8 Hz); 1.17»1.46 (4H, m); 1.29 (6H, d, J\7.0 Hz); 1.66»1.79 (2H, m); 3.01 (1H, hept, J\6.9 Hz); 3.07 (2H, t, J\8.1 Hz); 7.48 (1H, dd, 3J\7.7 and 7.7 Hz); 7.52 (1H, ddd, 3J\7.7 Hz, 4J\1.9 and 1.9) ; 7.72 (1H, ddd, 3J\7.7, 4J\1.9 and 1.9 Hz), 7.76 (1H, dd, 4J\1.9 and 1.9 Hz). 75 MHz 13C NMR d: 13.69 ; 22.11 ; 22.25 ; 23.77 (2C) ; 30.341 ; 34,12 ; 56.30 ; 125.50 ; 125.87 ; 129.27 ; 131.89 ; 139.15 ; 150.46.Compound 9e (R@\Pri) : liquid. MS: exact mass measurement: obsd, 254.1343; calcd for 254.1340. 300 C14H22O2S, MHz 1H NMR d: 0.86 (3H, t, J\7.0 Hz); 1.15»1.43 (4H, m); 1.32 (6H, d, J\6.8 Hz); 1.56»1.82 (2H, m); 3.13 (2H, t, J\8.0 Hz); 3.89 (1H, hept, J\6.8 Hz); 7.35 (1H, ddd, 3J\7.6 and 8.0 Hz, 4J\1.3 Hz); 7.51 (1H, dd, 3J\7.9 Hz, 4J\1.3 Hz); 7.58 (1H, ddd, 3J\7.6 and 7.9 Hz, 4J\1.3 Hz), 8.0 (1H, dd, 3J\8.0 Hz, 4J\1.3 Hz). Compound 10e (R@\Pri) : liquid. 300 MHz 1H NMR d: 0.86 (6H, t, J\7.0 Hz); 1.15»1.43 (8H, m); 1.33 (6H, d, J\6.9 Hz); 1.56»1.82 (12H, m); 3.09 (1H, hept, J\6.8 Hz); 3.61 (2H, t, J\8.0 Hz); 3.64 (2H, t, J\8.0 Hz); 7.60 (1H, dd, 3J\8.0 Hz, 4J\1.8 Hz); 8.08 (1H, d, 4J\1.6 Hz); 8.18 (1H, d, 3J\8.0 Hz) Compound 6e: see above.Electrolysis of 2f in the presence of isobutyl bromide. Compound 8f (R@\Bui) : liquid. MS: exact mass measurement: obsd, 282.1640; calcd for 282.1653. 300 MHz 1H C16H26O2S, NMR d: 0.82 (3H, t, J\7.4 Hz); 0.85 (3H, t, J\7.7 Hz); 1.16»1.49 (6H, m); 1.53»1.78 (4H, m); 1.279 (3H, d, J\6.9 Hz); 2.71 (1H, tq, J\6.9 and 7.1 Hz); 3.08 (2H, t, J\8.0 Hz); 7.43»7.55 (2H, m); 7.67»7.76 (2H, m). 75 MHz 13C NMR d: 13.88 ; 13.95 ; 21.58 ; 22.25 ; 22.60 ; 27.93 ; 31.01 ; 31.16 ; 41.64 ; 56.38 ; 125.54 ; 126.54 ; 129.21 ; 132.43 ; 139.20 ; 149.35. Compound 9f (R@\Bui) : liquid. MS: exact mass measurement: obsd, 282.1640; calcd for 282.1653. 300 C16H26O2S, MHz 1H NMR d: 0.85 (3H, t, J\6.8 Hz); 0.87 (3H, t, J\7.3 Hz); 1.17»1.46 (4H, m); 1.54»1.78 (6H, m); 1.29 (3H, d, J\6.8 Hz); 3.13 (2H, t, J\8.1 Hz); 3.64 (1H, tq, J\6.8 and 7.1 Hz); 7.35 (1H, ddd, 3J\7.5 and 8.0 Hz, 4J\1.2 Hz); 7.46 (1H, dd, 3J\7.9 Hz, 4J\1.2) ; 7.59 (1H, ddd, 3J\7.5 and 7.9 Hz, 4J\1.2 Hz), 8.01 (1H, dd, 3J\8.0 Hz, 4J\1.2 Hz). 75 MHz 13C NMR d: 12.29 ; 13.89 ; 22.28 ; 22.57 (2C) ; 28.02 ; 31.12 ; 31.17 ; 36.11 ; 56.82 ; 126.20 ; 128.04 ; 130.19 ; 133.80 ; 137.05 ; 148.42.Electrolysis of 2e in the presence of (R)-2-bromooctane. Compound 8e racemic: liquid. 300 [R@\CH(CH3)(C6H13)], MHz 1H NMR d: 0.86 (6H, t, J\7.1 Hz); 1.26 (3H, t, J\6.9 Hz); 1.15»1.43 (12H, m); 1.53»1.64 (2H, m); 1.65»1.78 (2H, m); 2.77 (1H, tq, J\6.9 and 7.1 Hz); 3.07 (2H, t, J\8.1 Hz); 7.43»7.50 (2H, m); 7.68»7.76 (2H, m). 75 MHz 13C NMR d: 13.67 ; 14.06 ; 22.07 ; 22.11 ; 22.36 ; 22.61 ; 27.57 ; 29.27 ; 30.37 ; 31.75 ; 38.22 ; 39.96 ; 56.36 ; 125.51 ; 126.49 ; 129.23 ; 132.38 ; 139.14 ; 149.67.Compound 9e racemic: liquid. [R@\CH(CH3)(C6H13)], 300 MHz 1H NMR d: 0.86 (9H, t, J\6.89 Hz); 1.29 (3H, t, J\7.0 Hz); 1.14»1.45 (14H, m); 1.53»1.79 (8H, m); 2.89 (1H, tq, J\7.1 and 7.1 Hz); 3.62 (4H, t, J\7.8 Hz); 7.61 (1H, dd, 3J\8.1 Hz, 4J\1.8 Hz); 8.08 (1H, d, 4J\1.8 Hz); 8.20 (1H, d, 3J\8.1 Hz). Compound 3e: see above.Compound 10e racemic: liquid. [R@\CH(CH3)(C6H13)] ; 300 MHz 1H NMR d: 0.86 (6H, t, J\6.8 Hz); 1.27 (3H, t, J\7.0 Hz); 1.15»1.45 (12H, m); 1.53»1.79 (4H, m); 3.13 (2H, t, J\8.1 Hz); 3.69 (1H, tq, J\6.8 and 6.8 Hz); 7.35 (1H, ddd, 3J\8.0 and 8.0 Hz, 4J\1.1 Hz); 7.47 (1H, dd, 3J\7.8, 4J\1.1 Hz); 7.59 (1H, ddd, 3J\7.8 and 8.0 Hz, 4J\1.3 Hz), 8.00 (1H, dd, 3J\8.0 Hz, 4J\1.3 Hz).Electrolysis of 2b in the presence of 1,4-dibromobutane. Compound 8b liquid. MS: exact mass mea- [R@\(CH2)4Br]: surement: obsd, 304.0141; calcd for 304.0132. C12H17O2SBr, 300 MHz 1H NMR d: 1.28 (3H, t, J\7.4 Hz); 1.76»1.97 (4H, m); 2.74 (2H, t, J\7.4 Hz); 3.12 (2H, q, J\7.4 Hz); 3.43 (2H, t, J\6.4 Hz); 7.44»7.54 (2H, m); 7.70»7.78 (2H, m). 75 MHz 13C NMR d: 7.49 ; 29.56 ; 32.11 ; 33.30 ; 34.79 ; 50.65 ; 125.90 ; 127.88 ; 129.39 ; 133.80 ; 138.70 ; 143.53. Compound 9b liquid. MS: exact mass [R@\(CH2)4Br]: measurement: obsd, 304.0140; calcd for C12H17O2SBr, 304.0132. 300 MHz 1H NMR d: 1.27 (3H, t, J\7.4 Hz); 1.78» 1.93 (2H, m); 1.92»2.06 (2H, m); 3.05 (2H, t, J\7.8 Hz); 3.15 (2H, q, J\7.4 Hz); 3.46 (2H, t, J\6.6 Hz); 7.39 (1H, ddd, 3J\7.5 and 8.3 Hz, 4J\1.3 Hz); 7.40 (1H, dd, 3J\7.5 Hz, 4J\1.3 Hz); 7.57 (1H, ddd, 3J\7.5 and 7.5 Hz, 4J\1.4 Hz), 7.99 (1H, dd, 3J\8.3 Hz, 4J\1.4 Hz).Compound 6b: see above. Electrolysis of 2c in the presence of 1,4-dibromobutane. Compound 8c liquid. MS: exact mass mea- [R@\(CH2)4Br]: surement: obsd, 318.0291; calcd for 318.0289. C13H19O2SBr, 300 MHz 1H NMR d: 1.00 (3H, t, J\7.4 Hz); 1.68»1.96 (6H, m); 2.74 (2H, t, J\7.4 Hz); 3.07 (2H, t, J\8.0 Hz); 3.43 (2H, t, J\6.4 Hz); 7.44»7.53 (2H, m); 7.70»7.78 (2H, m). 75 MHz 13C NMR d: 13.10 ; 16.52 ; 29.52 ; 32.08 ; 33.22 ; 34.76 ; 58.05 ; 125.73 ; 127.71 ; 129.34 ; 133.67 ; 138.36 ; 143.48. Compound 9c liquid. 300 MHz 1H NMR [R@\(CH2)4Br]: d: 1.00 (3H, t, J\7.5 Hz); 1.66»1.81 (2H, m); 1.78»1.93 (2H, m); 1.92»2.06 (2H, m); 3.04 (2H, t, J\7.9 Hz); 3.10 (2H, t, J\8.0 Hz); 3.46 (2H, t, J\6.5 Hz); 7.38 (1H, ddd, 3J\7.5 and 7.7 Hz, 4J\1.3 Hz); 7.40 (1H, dd, 3J\7.5 Hz, 4J\1.3 Hz); 7.57 (1H, ddd, 3J\7.5 and 7.5 Hz, 4J\1.4 Hz), 8.00 (1H, dd, 3J\7.7 Hz, 4J\1.4 Hz). 75 MHz 13C NMR d: 13.04 ; 16.43 ; 30.68 ; 32.18 ; 32.53 ; 33.49 ; 58.36 ; 126.80 ; 130.60 ; 131.71 ; 133.70 ; 137.16 ; 142.26.Electrolysis of 2b in the presence of 1,6-dibromohexane. Compound 8b liquid. MS: exact mass mea- [R@\(CH2)6Br]: surement: obsd, 332.0450; calcd for 332.0445. C14H21O2SBr, 300 MHz 1H NMR d: 1.28 (3H, t, J\7.5 Hz); 1.36»1.58 (4H, m); 1.59»1.79 (2H, m); 1.80»1.94 (2H, m); 2.70 (2H, t, J\7.6 Hz); 3.11 (2H, t, J\7.5 Hz); 3.40 (2H, t, J\6.7 Hz); 7.44» 7.51 (2H, m); 7.68»7.76 (2H, m). 260 New J. Chem., 1998, Pages 253»261Compound 9b liquid. MS: exact mass [R@\(CH2)6Br]: measurement: obsd, 332.0450; calcd for C14H21O2SBr, 332.0445. 300 MHz 1H NMR d: 1.24 (3H, t, J\7.5 Hz); 1.35»1.58 (4H, m); 1.59»1.79 (2H, m); 1.80»2.94 (2H, m); 3.00 (2H, t, J\8.0 Hz); 3.14 (2H, t, J\7.5 Hz); 3.41 (2H, t, J\6.8 Hz); 7.36 (1H, ddd, 3J\7.5 and 8.2 Hz, 4J\1.3 Hz); 7.38 (1H, dd, 3J\7.4 Hz, 4J\1.3 Hz); 7.55 (1H, ddd, 3J\7.4 and 7.5 Hz, 4J\1.4 Hz), 7.99 (1H, dd, 3J\8.2 Hz, 4J\1.4 Hz).Compound 3b, 4b and 6b: see above. Electrolysis of 2c in the presence of 1,10-dibromodecane. Compound 8c liquid. MS: exact mass mea- [R@\(CH2)10Br]: surement: obsd, 402.1206; calcd for 402.1228. C19H31O2SBr, 200 MHz 1H NMR d: 1.00 (3H, t, J\7.4 Hz); 1.17»1.51 (12H, m); 1.54»1.93 (6H, m); 2.70 (2H, t, J\7.7 Hz); 3.07 (2H, t, J\8.0 Hz); 3.41 (2H, t, J\6.8 Hz); 7.43»7.52 (2H, m); 7.66»7.76 (2H, m). 75 MHz 13C NMR d: 12.97 ; 16.51 ; 28.12 ; 28.71 ; 29.14 ; 29.34 ; 29.37 (2C) ; 31.16 ; 32.80 ; 34.09 ; 35.66 ; 57.98 ; 125.34 ; 127.69 ; 129.15 ; 133.75 ; 139.02 ; 144.49. Compound 9c liquid. MS: exact mass [R@\(CH2)10Br]: measurement: obsd, 402.1206; calcd for C19H31O2SBr, 402.1228. 200 MHz 1H NMR d: 0.99 (3H, t, J\6.8 Hz); 1.13»1.52 (12H, m); 1.52»1.96 (6H, m); 3.00 (2H, t, J\7.9 Hz); 3.10 (2H, t, J\8.0 Hz); 3.41 (2H, t, J\6.8 Hz); 7.36 (1H, ddd, 3J\7.5 and 7.5 Hz, 4J\1.3 Hz); 7.38 (1H, dd, 3J\7.5 Hz, 4J\1.3 Hz); 7.55 (1H, ddd, 3J\7.5 and 8.1 Hz, 4J\1.3 Hz), 7.99 (1H, dd, 3J\8.1 Hz, 4J\1.3 Hz). 75 MHz 13C NMR d: 13.02 ; 16.41 ; 28.14 ; 28.73 ; 29.41 (3C) ; 29.73 ; 32.25 ; 32.76 ; 33.09 ; 34.10 ; 58.13 ; 126.36 ; 130.39 ; 131.62 ; 133.49 ; 136.89 ; 143.18.Electrolysis of 2d in the presence of (ª)-1-bromo-3-chloro-2- methylpropane. Compound 8d [R@\CH2CH(CH3)(CH2Cl)] : liquid. MS: exact mass measurement: 288.0949; calcd for 288.0951. 300 MHz 1H NMR d: 0.89 (3H, t, C14H21O2SCl, J\7.3 Hz); 1.03 (3H, d, J\6.7 Hz); 1.39 (2H, m); 1.69 (2H, m); 2.18 (1H, m); 2.77 (2H, Hz); 3.09 (2H, t, ABX , JAB\13.6 J\8.0 Hz); 3.43 (2H, Hz); 7.45»7.54 (2H, ABX , JAB\10.8 m); 7.72»7.79 (2H, m). 75 MHz 13C NMR d: 13.48 ; 17.56 ; 21.56 ; 24.71 ; 37.29 ; 39.80 ; 49.90 ; 56.16 ; 125.99 ; 128.50 ; 129.39 ; 134.44 ; 139.58 ; 141.65. Compound 9d liquid. MS: [R@\CH2CH(CH3)(CH2Cl)] : exact mass measurement: obsd, 288.0951; calcd for 288.0951. 300 MHz 1H NMR d: 0.89 (3H, t, C14H21O2SCl, J\7.3 Hz); 1.07 (3H, d, J\6.7 Hz); 1.39 (2H, m); 1.57»1.75 (2H, m); 2.39 (1H, m); 3.03 (2H, Hz); 3.10 ABX , JAB\13.4 (2H, t, J\8.1 Hz); 3.54 (2H, Hz); 7.40 (1H, ABX , JAB\4.6 dd, 3J\7.4, 4J\1.4 Hz); 7.45 (1H, ddd, 3J\7.5 and 7.9 Hz, 4J\1.4 Hz); 7.57 (1H, ddd, 3J\7.4 and 7.5 Hz, 4J\1.5 Hz), 8.02 (1H, dd, 3J\7.9 Hz, 4J\1.5 Hz). 75 MHz 13C NMR d: 13.52 ; 17.61 ; 21.57 ; 24.57 ; 37.08 ; 37.36 ; 50.66 ; 56.56 ; 127.15 ; 130.93 ; 132.36 ; 133.44 ; 137.75 ; 140.39. Electrolysis of 2b in the presence of 6-bromohex-1-ene. Compound 8b liquid. 300 MHz 1H [R@\(CH2)4CHxCH2] : NMR d: 1.29 (3H, t, J\7.4 Hz); 1.37»1.77 (4H, m); 2.08 (2H, dt, J\6.7 and 6.7 Hz); 2.71 (2H, m, J\7.6 Hz); 3.11 (2H, q, J\7.4 Hz); 4.94 (1H, ddt, Hz, 2Jgem\2.1 3Jcis\10.3, Hz); 5.00 (1H, ddt, Hz, 4Jcis\1.1 2Jgem\2.1 3Jtrans\17.06 Hz, Hz); 5.79 (1H, ddt, 3J\6.7 Hz, 4Jtrans\1.6 3Jcis\10.3 Hz and Hz); 7.42»7.49 (2H, m); 7.67»7.76 (2H, 3Jtrans\17.1 m). 75 MHz 13C NMR d: 6.59 ; 24.94 ; 28.43 ; 30.59 ; 33.50 ; 50.66 ; 77.36 ; 114.74 ; 125.66 ; 127.89 ; 129.20 ; 133.79 ; 138.46 ; 144.36. Compound 9b liquid. 300 MHz [R@\(CH2)4CHxCH2] : 1H NMR d: 1.26 (3H, t, J\7.5 Hz); 1.45»1.81 (4H, m); 2.11 (2H, dt, J\6.7 and 7.3 Hz); 3.01 (2H, t J\7.8 Hz); 3.15 (2H, q, J\7.5 Hz); 4.95 (1H, ddt, Hz, 2Jgem\2.1 3Jcis\10.2, Hz); 5.02 (1H, ddt, Hz, 4Jcis\1.1 2Jgem\2.1 3Jtrans\17.06 Hz, Hz); 5.82 (1H, ddt, 3J\6.7 Hz, 4Jtrans\1.8 3Jcis\10.2 Hz and Hz); 7.38 (1H, dd, 3J\7.5, 4J\1.3 3Jtrans\17.1 Hz); 7.37 (1H, ddd, 3J\7.5 and 8.3 Hz, 4J\1.3 Hz); 7.55 (1H, ddd, 3J\7.5 and 7.5 Hz, 4J\1.3 Hz), 7.98 (1H, dd, 3J\8.3 Hz, 4J\1.3 Hz). 75 MHz 13C NMR d: 7.27 ; 24.85 ; 28.90 ; 31.64 ; 33.50 ; 50.73 ; 77.23 ; 114.61 ; 126.42 ; 130.66 ; 131.67 ; 133.56 ; 138.65 ; 143.13. Acknowledgements authors wish to thank CNRS and University of Rennes 1 The for –nancial and technical support. References 1 K. Tanaka and A. Kaji, in T he Chemistry of Sulphones and Sulphoxides, ed.S. Pataïé , Z. Rappoport and C. J. M. Stirling, Wiley, New York, 1988, p. 759 and references cited therein. 2 P. A. Grieco, Y. Mazaki and D. Boxler, J. Org. Chem., 1975, 40, 2261. 3 K. Schank, Methoden der Organischen Chemie (Houben-W eyl), Thieme Verlag, Stuttgart, 1985, vol. E11, p. 1132. 4 J. L. Fabre, M. Julia and J. N. Verpeaux, Bull. Soc. Chim.Fr., 1985, 762. 5 R. Mozingo, D. E. Wolf, A. Stanton, S. A. Harris and K. Folkers, J. Am. Chem. Soc., 1943, 65, 1013. 6 R. E. Dabby, J. Kenyon and R. F. Mason, J. Chem. Soc., 1952, 481. 7 (a) W. E. Truce, D. P. Tate and D. N. Burdge, J. Am. Chem. Soc., 1960, 82, 2872. (b) L. Reggel, R. A. Friedel and I. Wender, J. Org. Chem., 1957, 22, 891. 8 J. Simonet and G. Jeminet, Bull. Soc.Chim. Fr., 1971, 2754. 9 L. Horner, in Organic Electrochemistry, ed. M. M. Baizer, Marcel Dekker, New York, 1973, p. 746 and cited references. 10 J. Simonet, in T he Chemistry of Sulphones and Sulphoxides, ed. S. Pataïé , Z. Rappoport and C. J. M. Stirling, Wiley, New York, 1988, p. 1001 and cited references. 11 L. Horner and R. E. Schmitt, Z. Naturforsch Anorg., 1982, 37B, 1332. 12 M. Novi, G. Garbarino, C. DellœErba and G. Petrillo, J. Chem. Soc., Chem. Commun., 1984, 1205. 13 Preliminary results were already reported : (a) A. Belkasmioui and J. Simonet, T etrahedron L ett., 1991, 32, 2481. (b) A. Belkasmioui, Thesis, University of Rennes 1, 1991. (c) M. Benaskar, Thesis, University of Rennes, 1, 1994. (d) P. Cauliez, M. Benaskar and J. Simonet, Electrochim. Acta, 1996, 42, 2191. 14 (a) M. Malissard, J. P. Mazaleyrat and Z. Welvart, J. Am. Chem. Soc., 1977, 99, 6933. (b) E. Hebert, J. P. Mazaleyrat and Z. Welvart, Nouv. J. Chim., 1985, 9, 75. 15 R. C. Lamb, P. W. Ayers and M. K. Toney, J. Am. Chem. Soc., 1963, 85, 3482. 16 See for example: (a) J. F. Garst and F. E. Barton, J. Am. Chem. Soc., 1974, 96, 523. (b) K. Daasbjerg, T. Lund and H. Lund, T etrahedron L ett., 1989, 30, 493. (c) E. C. Ashby, T. N. Pham and A. Amrollah-Madjabadi, J. Org. Chem., 1991, 56, 1596. 17 See for example: (a) J. Simonet, M. A. Michel and H. Lund, Acta Chem. Scand., 1975, 29B, 489. (b) J. M. Saveç ant, J. Am. Chem. Soc., 1987, 109, 6788. (c) R. FuhlendorÜ, D. Occhialini, S. U. Pedersen and H. Lund, Acta Chem. Scand., 1989, 43, 803. 18 (a) M. W. Cronym and E. Zavarin, J. Org. Chem., 1954, 19, 139. (b) P. Cogolli, L. Testaferri, M. Tingoli and M. Tiecco, J. Org. Chem., 1979, 44, 2636. (c) W. E. Parham and P. L. Stright, J. Am. Chem. Soc., 1956, 78, 4783. 19 B. M. Trost and D. P. Curran, T etrahedron L ett., 1981, 22, 1287. 20 (a) R. Adams, W. Reifschneider and M. D. Nair, Org. Synth., 1961, 5, 106. (b) R. Adams and A. Ferreti, J. Am. Chem. Soc., 1959, 81, 4927. 21 A. M. Bernard, P. P. Piras, A. Plumitallo, S. Melis and F. Sotgiu, Gazz. Chim. Ital., 1982, 112, 443. 22 F. Maiolo, L. Testaferri, M. Tiecco and M. Tingoli, J. Org. Chem., 1981, 46, 3070. 23 G. W. Buchanam, C. Reyes-Zamora and D. E. Clarke, Can. J. Chem., 1974, 52, 3895. 24 J. A. Hyatt and A. W. White, Synthesis, 1984, 214. 25 M. Tiecco, M. Tingoli, L. Testaferri, D. Chianelli and F. Maiolo, Synthesis, 1982, 478. 26 J. Clayden, A. J. Cooney and M. Julia, J. Chem. Soc., Perkin T rans. 1, 1995, 7. Received 14th April 1997; revised M/S received 15th July 1997; Paper 7/08333E New J. Chem., 1998, Pages 253»261 261

ISSN:1144-0546    

DOI:10.1039/a708333e

出版商:RSC    

年代:1998

数据来源: RSC

摘要:

Influence du pH sur la vitesse dœoxydation de composeç s organiques par Meç canismes re ç actionnels et modeç lisation FeII/H2O2 . Herveç Gallard, Joseph de Laat* and Bernard Legube L aboratoire de Chimie de lœEau et de lœEnvironnement (CNRS UPRES A 6008), ESIP, 40, avenue du Recteur Pineau, 86022 Poitiers, France Cette eç tude a eu pour but de modeç liser la vitesse dœoxydation du fer ferreux et de composeç s organiques (atrazine, 1,2,4-trichlorobenze` ne) en solution aqueuse dilueç e par le syste` me Les expeç riences ont eç teç reç aliseç es a` FeII/H2O2 . 25 °C, a` pH initial compris entre 1 et 8, et en preç sence dœun exce` s de fer ferreux Les ([FeII]0P2 [H2O2]0). reç sultats obtenus a` pH acide (pHO3) ont permis de montrer que le mode` le cineç tique baseç sur le meç canisme reç actionnel de Barb et al.1 (1951) deç crit parfaitement la vitesse dœoxydation du fer ferreux et des composeç s organiques.Lœapplication de ce mode` le a permis de mesurer les valeurs des constantes cineç tiques des radicaux hydroxyle sur lœatrazine (forme moleç culaire : k\3,0]109 L mol~1 s~1; forme protoneç e : k\1,2]109 L mol~1 s~1) et du 1,2,4-trichlorobenze` ne (k\6,2]109 L mol~1 s~1).Pour les expeç riences reç aliseç es a` pH[4 et en absence dœoxyge` ne dissous, les reç sultats ont montreç que le mode` le de Barb et al.1 conduit a` une surestimation de la vitesse dœoxydation des composeç s organiques. Ces reç sultats ont mis en eç vidence que la reç action de H2O2 sur FeII ne conduit pas directement a` la libeç ration de radicaux hydroxyle mais a` la formation dœun intermeç diaire qui peut soit libeç rer des radicaux hydroxyle et FeIII (reç action favoriseç e a` pH acide) soit former un autre intermeç diaire qui peut reç agir avec FeII pour donner FeIII sans libeç ration de radicaux hydroxyle.EÜect of pH on the oxidation rate of organic compounds by Mechanisms and simulation. In this work, FeII/H2O2 . the kinetics of oxidation of FeII and of organic compounds (atrazine, 1,2,4-trichlorobenzene) in dilute aqueous solution by have been investigated.Experiments were carried out at 25 °C, at initial pH between 1 FeII/H2O2 and 8, and in the presence of an excess of FeII Data obtained at pHO3 showed that the ([FeII]0P2 [H2O2]0). rates of oxidation of FeII and of organic compounds were predicted well by a kinetic model based on the mechanism proposed by Barb et al.1 (1951).The following values were found for the kinetic rate FeII/H2O2 constants for the reaction of with atrazine (molecular form: k\3.0]109 L mol~1 s~1; protonated form: OH~ k\1.2]109 L mol~1 s~1) and 1,2,4-trichlorobenzene (k\6.2]109 L mol~1 s~1). At pH[4 and in the absence of dissolved oxygen, the kinetic model was found to overestimate the rate of oxidation of both atrazine and 1,2,4-trichlorobenzene.These data suggest that the reaction of with FeII probably goes through the H2O2 formation of an intermediate. This intermediate leads to the formation of radicals and FeIII (major route OH~ under acidic pH) or leads to the formation of another intermediate which may react with FeII to produce FeIII without formation of OH~.Le syste` me connu sous le terme de reç actif de FeII/H2O2 , Fenton, a fait lœobjet de tre` s nombreux travaux tant au niveau de lœeç tude du meç canisme reç actionnel1,2 quœau niveau de ses applications dans le domaine de lœoxydation de polluants organiques en milieux aqueux.3 Plus reç cemment, un certain nombre de travaux consacreç s a` la chimie de lœatmosphe` re et des milieux aquatiques (oceç ans, eaux de surface) ont montreç que les reç actions de type Fenton peuvent aussi participer a` lœoxydation de composeç s organiques et mineç raux et avoir une in—uence sur le cycle du fer et sur les concentrations des espe` ces oxydantes etc.) dans les aeç rosols et (O2, H2O2 , OH~, dans les eaux naturelles.4,5 A pH acide (pH\3 a` 4), il est geç neç ralement admis que lœeç tape initiale du meç canisme dœoxydation des composeç s organiques en milieux aqueux par est la suivante : FeII/H2O2 FeII]H2O2 ]FeIII]OH~]OH~ Les radicaux hydroxyle libeç reç s dans le milieu reç actionnel peuvent reç agir avec les ions ferreux, le peroxyde dœhydroge` ne et les composeç s organiques.Les ions ferrique formeç s peuvent e� tre hydrolyseç s et preç cipiteç s ou conduire a` une reç geç neç ration de fer ferreux par des reç actions secondaires avec le peroxyde dœhydroge` ne ou avec des entiteç s radicalaires (O2~~, HO2~, R~, etc.).Lœefficaciteç du syste` me pour lœoxydation de FeII/H2O2 * Tel: (]33) 5 49 45 39 21; Fax: (]33) 5 49 45 37 68. composeç s organiques est optimale a` pH acide (pH+3) et la diminution de lœefficaciteç pour des pH supeç rieurs est geç neç ralement attribueç e a` la preç cipitation de fer ferrique.Sur le plan cineç tique, la vitesse de lœeç tape initiale du meç canisme de deç composition de par FeII augmente avec le H2O2 pH car lœion (k\5,9]106 L mol~1 s~1) [(H2O)5FeIIOH]` est beaucoup plus reç actif que lœion (k\63 L [(H2O)6FeII]2` mol~1 s~1) vis-a` -vis du peroxyde dœhydroge` ne.6 Compte tenu de cette diÜeç rence importante de reç activiteç entre et une augmentation sig- [(H2O)5FeIIOH]` [(H2O)6FeII]2`, ni–cative de la vitesse dœoxydation de composeç s organiques par peut e� tre attendue lorsque le pH augmente et si FeII/H2O2 le fer ferreux nœest pas en concentration limitante.Cette eç tude a donc eu pour objectif dœexaminer lœin—uence du pH sur la vitesse dœoxydation de composeç s organiques en milieux aqueux tre` s dilueç par Cet article preç sen- FeII/H2O2 .tera uniquement les reç sultats dœexpeç riences dœoxydation reç alise ç es avec des concentrations initiales en fer ferreux supeç rieures a` la stoechiomeç trie de manie` re a` neç gli- ([FeII]0/[H2O2]0P2) ger les reç actions faisant intervenir les ions FeIII.Partie expeç rimentale Preç paration des solutions Les solutions me` res dœatrazine (produit Merck, pureteç P98%), de 1,2,4-trichlorobenze` ne (TCB) (Fluka 99%), de peroxyde New J. Chem., 1998, Pages 263»268 263dœhydroge` ne (Fluka solution commerciale stabiliseç e a` 30% massique), dœions ferreux CarloErba 98%] [Fe(ClO4)2 … 6H2O, ont eç teç preç pareç es dans lœeau ultra-pure (eau ììMilliQœœ : COTO0,2 mg L~1; 18 M) cm).Le perchlorate de sodium, lœacide perchlorique et la soude ont eç teç utiliseç s pour –xer la force ionique et le pH des solutions. Toutes les solutions ont eç teç preç pareç es dans lœeau ultra-pure. Conditions díoxydation Le montage expeç rimental est constitueç dœun reç acteur a` double enveloppe opaque a` la lumie` re, de volume utile eç gal a` 2 L.Lœoxydation a eç teç reç aliseç e a` 25^0,2 °C gra� ce a` une recirculation dœeau thermostateç e dans lœenveloppe externe, en preç sence dœoxyge` ne dissous mg L~1, +0,26 mmol L~1) ou ([O2]+8,5 en absence dœoxyge` ne dissous gra� ce a` un deç gazage de la solution par de lœazote mg L~1, \3]10~3 m]\0,1 L~1). Pour les expeç riences reç aliseç es a` pHP7, cette concentration reç siduelle en oxyge` ne dissous contribue a` lœoxydation instantaneç e de quelques micromoles de fer ferreux a` pH neutre avant lœajout de peroxyde dœhydroge` ne.Toutefois, la concentration reç elle est doseç e pour chaque expeç rience avant [FeII]0 lœaddition de peroxyde dœhydroge` ne et est utiliseç e comme valeur initiale dans le calcul.Au temps t\0, le peroxyde dœhydroge` ne est ajouteç a` la solution (volume: 1 L) contenant le sel de fer et (ou) le(s) composeç (s) organique(s). Lœinjection de peroxyde dœhydroge` ne est reç aliseç e, sous agitation intense, a` lœaide dœune micropipette de preç cision (volume introduit :O10 mL). Une eç tude preç liminaire a montreç aucune in—uence de la proceç dure dœajout des reç actifs a` pH acide et a` pH neutre.En cours de reç action, les preç le` vements ont eç teç eÜectueç s avec une micropipette preç alablement eç talonneç e (3 mL). Pour lœeç tude cineç tique de la vitesse dœoxydation des composeç s organiques, ces preç le` vements ont eç teç aussito� t meç langeç s a` 1 mL de meç thanol a–n dœarre� ter les reç actions dœoxydation.7 Des expeç riences preç liminaires ont montreç une stabiliteç parfaite de la concentration en composeç organique durant plusieurs heures en preç sence de fer ferreux, de peroxyde dœhydroge` ne et de meç thanol. Meç thodes díanalyses Les concentrations en peroxyde dœhydroge` ne dans les solutions me` res mol L~1) ont eç teç doseç es par ([H2O2]\5]10~3 iodomeç trie.Les concentrations en peroxyde dœhydroge` ne nœont pas eç teç mesureç es en cours de reç action a` ([H2O2]\0 50]10~6 mol L~1) en raison des interfeç rences des ions ferreux ([FeII]\0 a` 100]10~6 mol L~1) sur le dosage par la meç thode spectrophotomeç trique au titane.Les analyses de fer ferreux ont eç teç eÜectueç es par la meç thode spectrophotomeç trique a` lœo-pheç nantroline (norme AFNOR NFT 90-017).Le coefficient dœextinction molaire du complexe est eç gal a` 10 300 L mol~1 cm~1. Aux concentrations eç tudieç es mol L~1), il nœa pas eç teç noteç ([FeII]0O100]10~6 dœinterfeç rences du fer ferrique sur le dosage du fer ferreux. Les composeç s organiques ont eç teç analyseç s par chromatographie liquide haute performance (CLHP): colonne Resolve C18, 90 5 lm, 3,9]150 mm; phase meç thanol»eau 45: 55 ”, ou 70: 30 selon les composeç s ; deç bit 0,9 mL min~1.La deç tection des composeç s organiques est reç aliseç e a` lœaide dœun deç tecteur UV caleç a` 210 nm ou a` 220 nm. Lœeç talonnage a eç teç reç aliseç a` lœaide de solutions preç pareç es dans un meç lange meç thanol» eau. Les limites de deç tection analytique (volume injecteç 200 lL) sont de lœordre de 0,02 lmol L~1 pour lœatrazine et de 0,077 lmol L~1 pour le 1,2,4-trichlorobenze` ne.Mode` le cineç tique A pH infeç rieur a` 4 et en absence dœoxyge` ne dissous, lœensemble des auteurs sœaccorde sur le meç canisme reç actionnel geç neç ral de Barb et al.1 pour deç crire le reç actif de Fenton. Le tableau 1 preç sente les diÜeç rentes reç actions ainsi que les valeurs des constantes cineç tiques.En consideç rant les reç actions preç senteç es dans le tableau 1, les eç quations deç crivant la variation de la concentration en fer ferreux et en composeç organique en fonction du temps sont les suivantes : d[FeII] dt \[k1[FeII][H2O2][k2[FeII][OH~] [k5[FeII][HO2~][k9[FeII][O2~~] ]k6[FeIII][HO2~]]k10[FeIII][O2~~] ]k4[FeIII][H2O2] (1) k1\k1a [Fe2`] [FeII] ]k1b [FeIIOH`] [FeII] (1@) d[M] dt \[kM[M][OH~] (2) En milieu non tamponneç et a` pH proche de la neutraliteç , lœhydrolyse du fer ferrique provoque une diminution du pH en Tableau 1 Reç actions prises en compte dans le mode` le cineç tique dœoxydation de composeç s organiques par FeII/H2O2 Constante cineç tique/ L mol~1 s~1 Reç feç rence No Reç action (sauf mention contraire) bibliographique (1a) Fe2`]H2O]FeIII]OH~]OH~ 63 notre eç tude (1b) FeIIOH`]H2O2 ]FeIII]OH~]OH~ 5,9]106 6 (1c) Fe2`]OH~¢FeIIOH` pK1\9,87 8 (2) FeII]OH~]FeIII]OH~ 3]108 9 (3) H2O2]OH~]HO2~]H2O 3,3]107 10 (4) FeIII]H2O2 ]FeII]HO2~]H` O3]10~3 eç tude en cours (5) FeII]HO2~]H`]FeIII]H2O2 1,2]106 11 (6) FeIII]HO2~]H`]FeII]O2]H2O 1]103 11 HO2~¢O2~~]H` pKa\4,8 12 (7) HO2~]O2~~]H` 1,58]105 s~1 (8) O2~~]H2 ]HO2~ 1]1010 (9) FeII]O2~~]2H`]FeIII]H2O2 1]107 11 (10) FeIII]O2~~]H`]FeII]O2]OH~ 1,5]108 11 (11) HO2~]O2~~]O2]HO2~ 9,7]107 12 (12) HO2~]HO2~]H2O2]O2 8,3]105 12 (13) HO2~]OH~]H2O]O2 7,1]109 13 (14) OH~]O2~~]OH~]O2 1]1010 13 (15) OH~]OH~]H2O2 5,2]109 13 Oxydation de composeç s organiques: (M) M]OH~]R~ notre eç tude 264 New J.Chem., 1998, Pages 263»268[FeII] / [FeII]0 t / s t / s t / s [FeII] / [FeII]0 pH a b cours de reç action.En neç gligeant la consommation des radicaux hydroxyle par les composeç s organiques et par le peroxyde dœhydroge` ne et en consideç rant que le fer ferrique est preç sent seulement sous la forme dœhydroxydes ferriques pour des valeurs de pH proches de la neutraliteç , les eç quations globales dœoxydation du fer ferreux sont : 2 Fe2`]H2O2]4 H2O]2 FeIII(OH)3]4 H` 2 FeIIOH`]H2O2]2 H2O]2 FeIII(OH)3]2 H` et lœeç volution du pH en cours de reç action peut e� tre deç crite par d[H`] dt \4k1a[Fe2`][H2O2] ]2k1b[FeIIOH`][H2O2] (3) Les diÜeç rentes eç quations donnant lœeç volution du pH, de la concentration totale en FeII et en composeç organique en fonction du temps ont eç teç reç solues par analyse numeç rique en utilisant la meç thode de Runge»Kutta a` lœordre 4 et en consideç rant lœeç tat stationnaire pour les espe` ces radicalaires.Reç sultats et Discussion La premie` re partie de cette eç tude a eç teç consacreç e a` la cineç tique dœoxydation du fer ferreux par le peroxyde dœhydroge` ne en absence de composeç organique. La vitesse dœoxydation des deux composeç s organiques, lœatrazine et le 1,2,4-trichlorobenze` ne, par a ensuite eç teç eç tudieç e.FeII/H2O2 Vitesse dœoxydation du fer ferreux en absence de composeç organique Les reç sultats expeç rimentaux obtenus a` pHO3,0 et pour des forces ioniques comprises entre 10~3 et 10~1 mol L~1 et avec diÜeç rentes concentrations initiales en fer (NaClO4) ferreux (10 a` 100 lmol L~1) et en peroxyde dœhydroge` ne (5 a` 50 lmol L~1) ont eç teç correctement modeç liseç s en prenant une valeur de constante cineç tique apparente dœoxydation du fer ferreux par le peroxyde dœhydroge` ne eç gale a` 63 L mol~1 s~1 a` 25 °C (Fig. 1). En milieu perchlorate, cette valeur peut e� tre assimileç e a` la constante cineç tique de reç action entre lœion Fe2` et le peroxyde dœhydroge` ne Pour des pH compe (k1a). 1,0 et 3,0, le pH est resteç inchangeç en cours de reç action. La concentration reç siduelle en fer ferreux mesureç e en –n de reç action est en accord avec la concentration theç orique calculeç e par le mode` le, quœelles que soient les concentrations initiales en fer ferreux et en peroxyde dœhydroge` ne. Pour des rapports molaires la stoechiomeç trie de reç action [FeII]0/[H2O2]0A2, correspond a` la consommation de deux moles de FeII par mole de peroxyde dœhydroge` ne.Pour des rapports molaires la stoechiomeç trie est leç ge` rement infeç - [FeII]0/[H2O2]0+2, rieure a` deux moles de FeII par mole de en raison H2O2 , Fig. 1 Evolution des concentrations expeç rimentales (symboles) et theç oriques (courbes) en fer ferreux en fonction du temps.Ö [FeII]0\ lmol L~1, lmol L~1; lmol L~1, 100 [H2O2]0\50 K [FeII]0\50 lmol L~1; lmol L~1, [H2O2]0\25 L [FeII]0\10 [H2O2]0\5 lmol L~1; lmol L~1, lmol L~1; > [FeII]0\100 [H2O2]0\25 = lmol L~1, lmol L~1 [FeII]0\100 [H2O2]0\10 dœune consommation dœune partie du peroxyde dœhydroge` ne par les radicaux hydroxyle (reç action 3, tableau 1). Lœeç tude de la cineç tique dœoxydation du fer ferreux en milieu perchlorique nœa montreç aucune in—uence perceptible du pH dans le domaine de valeurs de pH compris entre 1,0 et 3,0 en accord avec la litteç rature et la chimie du fer dans ce domaine de pH.Lœeç tude de lœin—uence du pH a eç teç compleç teç e en reç alisant des expeç riences en milieu plus basique (pH initial 7), en absence dœoxyge` ne dissous, a–n dœeç viter une oxydation du fer ferreux par lœoxyge` ne dissous.Lœoxydation a eç teç eÜectueç e en milieu non tamponneç et avec une force ionique –xeç e a` 0,1 mol L~1 par du perchlorate de sodium. Les –gures 2a et 2b preç sentent lœeç volution du fer ferreux et du pH en fonction du temps. Les reç sultats obtenus ont permis de montrer que le mode` le cineç tique deç crit parfaitement lœeç volution de la concentration en fer ferreux en fonction du temps (Fig. 2a). La vitesse dœoxydation du fer ferreux est plus rapide a` pH neutre quœa` pH acide car le peroxyde dœhydroge` ne reç agit plus rapidement sur FeIIOH` L mol~1 (k1b\5,9]106 s~1) que sur Fe2` L mol~1 s~1). (k1a\63 La diminution du pH en cours de reç action (Fig. 2b) qui reç sulte de lœhydrolyse du fer ferrique est eç galement correctement deç crite par le mode` le.Cette diminution du pH deç pend de la quantiteç totale de fer ferrique produit et du pH initial. En conclusion, les reç sultats expeç rimentaux pour lœoxydation du fer ferreux par le peroxyde dœhydroge` ne en milieu perchlorate et en absence de composeç organique sont en accord avec le mode` le cineç tique (tableau 1) et ont permis de valider les valeurs des diÜeç rentes constantes cineç tiques et des constantes dœhydrolyses utiliseç es dans le mode` le.Modeç lisation de la vitesse dœoxydation de micropolluants organiques La modeç lisation de la vitesse dœoxydation de composeç s organiques par le syste` me a eç teç reç aliseç e en milieu FeII/H2O2 Fig. 2 In—uence du pH initial (K pHi\6,85 ; + pHi\6,27 ; Ö et des concentrations initiales en FeII et et pHi\3,0) H2O2 (+ Ö: lmol L~1, lmol L~1; [FeII]0\10 [H2O2]0\5 K [FeII]0\100 lmol L~1, lmol L~1) sur la vitesse dœoxydation de FeII [H2O2]0\50 (symboles: reç sultats expeç rimentaux; trait plein : modeç lisation cineç tique).a Evolution de FeII. b Evolution du pH New J. Chem., 1998, Pages 263»268 265t / s t / s [TCB] / [TCB]0 [At] / [At]0 a b t / s t / s [At] / mM [TCB] / mM a b aqueux tre` s dilueç a` 2 lmol L~1) dans le but de ([M]0+0,6 limiter les reç actions secondaires faisant intervenir les radicaux organiques sur la vitesse dœoxydation du fer ferreux.(R~, RO2~) Lœatrazine et le 1,2,4-trichlorobenze` ne ont eç teç utiliseç s comme moleç cules organiques mode` les car des expeç riences preç liminaires lmol L~1; lmol ([FeII]0\100 [H2O2]0\25 L~1; pH\3,0) ont montreç que les vitesses dœoxydation de ces deux composeç s nœeç taient pas aÜecteç es par la preç sence dœoxyge` ne dissous.Les reç sultats obtenus avec dœautres moleç cules organiques avaient en particulier montreç que la vitesse dœoxydation de pheç nylureç es (diuron et isoproturon) et dœun composeç polyaromatique (naphthale` ne) eç tait fortement augmente ç e en preç sence dœoxyge` ne dissous et que le 2,5-dichloronitrobenze` ne eç tait deç composeç en milieu neutre en preç sence de fer ferreux.Modeç lisation de la vitesse dœoxydation du 1,2,4- trichlorobenze` ne et de lœatrazine a ` pHO3. Lœin—uence du pH sur la vitesse dœoxydation de lœatrazine lmol L~1) ([At]0B0,6 ou du trichlorobenze` ne lmol L~1) a eç teç eç tu- ([TCB]0B1,4 dieç e, en absence dœoxyge` ne dissous, dans un domaine de valeurs de pH compris entre 1,0 et 3,0 avec des concentrations initiales en fer ferreux et en peroxyde dœhydroge` ne respectivement eç gales a` 50 lmol L~1 et 10 lmol L~1.Les –gures 3a et 3b preç sentent quelques exemples de reç sultats expeç rimentaux concernant lœeç volution de la concentration en composeç organique en fonction du temps de reç action.Elles montrent que la vitesse de disparition du TCB nœest pas in—uence ç e par le pH entre 1,0 et 3,0 alors que la vitesse dœeç limination de lœatrazine augmente lorsque le pH passe de 1,0 a` 3,0, vraisemblablement en raison dœune diÜeç rence de reç activite ç des radicaux hydroxyles vis-a` -vis de la forme protoneç e et moleç culaire de lœatrazine (pKa\1,6).Les eç volutions de la concentration en TCB et en atrazine observeç es sous diÜeç rentes conditions expeç rimentales ont pu e� tre correctement deç crites par le mode` le cineç tique en prenant comme constantes cineç tiques de reç action des radicaux Fig. 3 In—uence du pH pH\1; pH\1,3 ; pH\3,0) sur (L + Ö la vitesse dœoxydation de lœatrazine lmol L~1) et du tri- ([At]0\0,6 chlorobenze` ne lmol L~1) par ([TCB]0\1,4 FeII/H2O2 ([FeII]0\50 lmol L~1; lmol L~1; symboles: points expeç rimen- [H2O2]0\10 taux; trait plein : modeç lisation cineç tique).a Evolution de lœatrazine. b Evolution du TCB hydroxyle sur les composeç s organiques a` 25 °C les valeurs suivantes : 1,2,4-trichlorobenze` ne, L mol~1 s~1; kM\6,2]109 atrazine (forme protoneç e), L mol~1 s~1; atra- kM\1,2]109 zine (forme moleç culaire), L mol~1 s~1.En kM\3,0]109 conclusion, entre pH 1,0 et 3,0 et en absence dœoxyge` ne, le mode` le cineç tique deç crit parfaitement lœeç volution des concentrations en composeç s organiques et du fer ferreux en fonction du temps et quœelles que soient les concentrations initiales en reç actifs utiliseç es.Modeç lisation de la vitesse dœoxydation du 1,2,4- trichlorobenze` ne et de lœatrazine a ` pHP3. La cineç tique dœoxydation du 1,2,4-ichlorobenze` ne et de lœatrazine a eç teç eç tudieç e pour des pH initiaux compris entre 3,0 et 8,0 et des concentrations initiales en fer ferreux et en peroxyde dœhydroge` ne respectivement eç gales a` 50 lmol L~1 et 10 lmol L~1 (Fig. 4a et 4b). Les reç sultats expeç rimentaux montrent que la vitesse dœoxydation des deux composeç s organiques diminue lorsque le pH initial de la solution augmente alors que le mode` le cineç - tique preç voit une augmentation de la vitesse dœoxydation des composeç s organiques en raison dœune augmentation de la vitesse de production des radicaux hydroxyle. Par ailleurs, les expeç riences de cineç tique compeç titive reç aliseç es en oxydant des solutions contenant a` la fois lœatrazine et le trichlorobenze` ne ont montreç que le rapport des constantes cineç tiques kTCB/kAt ne varie pas avec le pH (pHi[4,0 : reç - kTCB/kAt\2,07^0,05, sultats non preç senteç s).Ce rapport, qui correspond au rapport des constantes cineç tiques de reç action des radicaux hydroxyle sur ces deux composeç s L mol~1 s~1; (kTCB\6,2]109 kAt\ L mol~1 s~1), indique que lœoxydation de lœatrazine 3,0]109 et du trichlorobenze` ne a` pH neutre reç sulte uniquement de reç actions avec le radical hydroxyle.Pour des valeurs de pH supeç rieures a` 3,0, lœeç cart entre les reç sultats expeç rimentaux et les reç sultats calculeç s a` lœaide du Fig. 4 Cineç tique dœoxydation de lœatrazine et du TCB pour des pHiP3,0 (Ö pHi\3,0 ; = pHi\5,0 ; K pHi\6,0 ; L pHi\7,7 ; + lmol L~1; lmol L~1; symbo- pHi\8,0 ; [FeII]0\50 [H2O2]0\10 les : points expeç rimentaux; trait plein : modeç lisation cineç tique).a Evolution de lœatrazine. b Evolution du TCB 266 New J. Chem., 1998, Pages 263»268t / s [At] / mM [TCB] / mM t / s Fig. 5 Etude cineç tique de lœoxydation de lœatrazine par a` FeII/H2O2 pHiP3,0 (Ö pHi\3,0 ; = pHi\5,0 ; K pHi\6,0 ; L pHi\7,7 ; lmol L~1; lmol L~1; symboles: points [FeII]0\50 [H2O2]0\10 expeç rimentaux; trait plein ; mode` le cineç tique modi–eç ) mode` le (Fig. 4) peut e� tre expliqueç soit par une production moins importante de radicaux hydroxyle a` pH neutre quœa` pH acide, soit par une consommation plus rapide des radicaux hydroxyle par FeIIOH` que par Fe2`.Cette deuxie` me hypothe` se a pu e� tre eç carteç e car des approches successives par modeç lisation ont montreç que la constante cineç tique de reç action des radicaux hydroxyle sur FeIIOH` doit alors e� tre au moins eç gale a` 1014 L mol~1 s~1. Les reç actions faisant intervenir les radicaux et HO2~/O2~~ les ions ferrique eç tant tre` s neç gligeables dans les conditions expeç rimentales utiliseç es au cours de cette eç tude, la diminution de la vitesse dœoxydation des composeç s organiques qui est observeç e lorsque le pH augmente ne peut e� tre expliqueç e que par une diminution de la vitesse de production des radicaux hydroxyle.Cette hypothe` se indiquerait que la reç action du peroxyde dœhydroge` ne sur les ions ferreux (Fe2` et FeIIOH`) ne conduirait pas directement a` la formation de radicaux hydroxyle (reç actions 1a et 1b, tableau 1) mais a` la formation dœun intermeç diaire noteç Ia : FeII]H2O2 ]Ia (A) Cet intermeç diaire se deç composerait tre` s rapidement en milieu acide pour donner des radicaux hydroxyle selon la reç action : Ia]H`]FeIII]OH~ (B) Cet intermeç diaire Ia pourrait aussi donner les reç actions hypothe ç tiques suivantes : Ia]FeII]2 FeIII (C) Ia]Ib (]H`) (D) Ib]FeII]2 FeIII (E) Ces reç actions, qui joueraient un ro� le plus important a` pH[3, permettraient dœexpliquer la production moindre de radicaux Fig. 6 Etude cineç tique de lœoxydation du TCB par a` FeII/H2O2 pHiP3,0 (Ö pHi\3,0 ; = pHi\5,0 ; + pHi\8,0 ; [FeII]0\50 lmol L~1; lmol L~1; symboles: points expeç rimen- [H2O2]0\10 taux; trait plein : mode` le cineç tique modi–eç ) hydroxyle et ne modi–eraient pas la stoechiomeç trie globale dœoxydation de FeII par (2 moles de FeII oxydeç es par H2O2 mole de consommeç e).Parmi ces reç actions secondaires, H2O2 lœexistence de la reç action C impliquerait quœune augmentation de la concentration en ions ferreux conduirait me� me, a` pH acide, a` une diminution de la vitesse de production des radicaux hydroxyle (compeç tition entre les reç actions B et C). La reç action C a eçteç eç carteç e car la diminution du rendement dœeç limination de lœatrazine observeç e lors dœexpeç riences reç aliseç es avec des concentrations croissantes en FeII a` pH 3 ([H`]\10~3 mol L~1; lmol L~1 a` 10~3 mol [FeII]0\50 L~1; lmol L~1 a` 50 lmol L~1) ne reç sulte [H2O2]0\2,5 uniquement que dœune consommation plus importante des radicaux hydroxyle par les ions ferreux (reç action 2, tableau 1) et pas dœune diminution de la concentration en radicaux hydroxyle par consommation de lœintermeç diaire Ia par le fer ferreux selon la reç action C.La formation dœintermeç diaires reç actionnels lors de lœoxydation de FeII par a eç galement eç teç eç voqueç e dans la H2O2 litteç rature mais le meç canisme nœest pas encore parfaitement eç tabli.2,14,15 Pour lœeç tape initiale, certains auteurs proposent la formation dœun complexe par [(H2O)5 `FeIIOOH, H3O`] addition de sur FeII, complexe qui se deç composerait tre` s H2O2 rapidement en milieu acide pour libeç rer des radicaux hydroxyle. Par ailleurs, dœautres auteurs proposent la formation de lœion ferryl comme espe` ce intermeç diaire.15 Lœoxydation des composeç s organiques est geç neç ralement attribueç e a` lœaction du radical hydroxyle (preç sent a` lœeç tat libre ou lieç a` un intermeç diaire) ou de lœion ferryl.Les reç sultats obtenus au cours de cette eç tude ne permettent pas de preç ciser la nature des intermeç diaires reç actionnels (Ia et Ib).Cependant, les valeurs de constantes cineç tiques obtenues par modeç lisation indiquent que le radical hydroxyle repreç sente la principale espe` ce responsable de lœoxydation de lœatrazine et du trichlorobenze` ne dans les conditions expeç rimentales utiliseç es au cours de cette eç tude. La diminution de la vitesse dœoxydation des composeç s organiques a` pH[4 montre par ailleurs que les reç actions entre FeII et en milieu neutre conduisent a` des espe` ces H2O2 intermeç diaires beaucoup moins reç actives que le radical hydroxyle ou ne conduisent pas a` une libeç ration de radicaux hydroxyle.En la reç action 1 (tableau 1) par la reç action A remplac” ant (avec et en prenant en compte les reç actions B, D et E, kA\k1) la modeç lisation a alors permis de deç crire dœune manie` re convenable lœeç volution de la concentration en composeç organique (Fig. 5 et 6) et en ions ferreux ainsi que lœeç volution du pH pour toutes les expeç riences reç aliseç es au cours de cette eç tude. La simulation qui a eç teç reç aliseç e avec et avec L kA103 mol~1 s~1 a conduit a` un rapport de constantes cineç tiques eç gal a` 2,3]10~5 mol L~1.kD/kB Il convient eç galement de signaler que des expeç riences dœoxydation par reç aliseç es en preç sence dœions FeIII FeII/H2O2 (avec nœont montreç aucune in—uence de la [FeIII]0\[FeII]0) preç sence de FeIII sur la vitesse dœoxydation des composeç s organiques et des ions ferreux. Ces reç sultats indiquent que dans les conditions expeç rimentales utiliseç es au cours de cette eç tude les reç actions secondaires faisant ([FeII]0/[H2O2]0P2), intervenir les ions ferriques jouent un ro� le tre` s neç gligeable.Les travaux actuellement en cours ont pour but dœeç tudier le meç canisme et la cineç tique de reç action du peroxyde dœhydroge` ne sur le fer ferrique en preç sence et en absence de composeç organique. Conclusion La modeç lisation de la vitesse dœoxydation de composeç s organiques (atrazine, trichlorobenze` ne) par reç aliseç e FeII/H2O2 en milieu aqueux tre` s dilueç , a` diÜeç rents pH (pH initial 1,0 a` 8,0) et en preç sence dœun large exce` s de FeII a permis de New J. Chem., 1998, Pages 263»268 267montrer que la reç action initiale de reç action du peroxyde dœhydroge` ne sur le fer ferreux ne conduirait pas directement a` la formation de radicaux hydroxyle (reç action geç neç ralement adopteç e par la plupart des auteurs) mais a` la formation dœun intermeç diaire reç actionnel qui peut soit libeç rer des radicaux hydroxyle (reç action favoriseç e pour des pHO3,0) soit OH~ donner des espe` ces non reç actives vis-a` -vis des composeç s organiques eç tudieç s (pH neutre).Pour des rapports molaires tre` s supeç rieurs a` 2, la stoechiomeç trie de la [FeII]0/[H2O2]0 reç action globale entre les ions ferreux et le peroxyde dœhydroge` ne nœest pas in—uenceç e par le pH (2 moles de FeII oxydeç es par mole de consommeç e). H2O2 Reç feç rences 1 W. G. Barb, J. H. Baxendale, P. George and K. R. Hargrave, T rans. Faraday Soc., 1951, 47, 462. 2 C. Walling, Acc. Chem. Res., 1975, 8, 125. 3 A. Aguiar, F. Carbonnier, H. Paillard and B. Legube, Wat. Supply, 11, 1993, 129; A. R. Bowers, P. Gadipati, W. W. Eckenfelder and R. M. Monsen, Wat. Sci. T ech., 1989, 21, 447; O. Koyama, Y. Kamagata and K. Nakamura, Wat. Res., 1994, 28, 895. 4 D. L. Sedlak and J. Hoigne, Environ. Sci. T echnol., 1994, 28, 1898; T. E. Graedel, M. L. Mandich and C. J. Weschler, J. Geophys Res., 1986, 91, 5205. 5 W. L. Miller, W. D. King, J. Lin and D. R. Kester, Mar. Chem., 1995, 50, 63. 6 F. J. Millero and S. Sotolongo, Geochim. Cosmochim. Acta., 1989, 53, 1867; J. W. MoÜett and R. G. Zika, Environ. Sci. T echnol., 1987, 21, 804. 7 S. M. Arnold, W. J. Hickey and R. F. Harris, Environ. Sci. T echnol., 1995, 29, 2083. 8 W. Stumm and J. J. Morgan, Aquatic Chemistry, 2nd ed., Wiley, New York, 1981 9 L. M. Dorfman and G. E. Adams, Reactivity of the Hydroxyl Radical in Aqueous Solution, NSRDS-NBS Report No. 46, U.S. Government Printing Office, Washington, D.C., 1973. 10 G. V. Buxton, C. L. Greenstock, W. P. Helman and A. B. Ross, J. Phys. Chem., 1988, 17, 513. 11 J. D. Rush and B. H. J. Bielski, J. Phys. Chem., 1985, 89, 5062. 12 B. H. J. Bielski, Photochem. Photobiol., 1978, 28, 645. 13 K. Sehested, O. L. Rasmussen and H. J. Fricke, J. Phys. Chem., 1968, 72, 626. 14 S. Rahal and H. W. Richter, J. Am. Chem. Soc., 1988, 110, 3126; I. Yamazaki and L. H. Piette, ibid., 1991, 113, 7588; H.-C. Tung, C. Kang and D. T. Sawyer, ibid., 1992, 114, 3345. 15 C. Walling and K. Amarnath, J. Am. Chem. Soc., 1982, 104, 1185. Received 5th May 1997; Paper 7/08335A 268 New J. Chem., 1998, Pages 263»2

ISSN:1144-0546    

DOI:10.1039/a708335a

出版商:RSC    

年代:1998

数据来源: RSC