摘要:

L e t t e r Clay catalysed amidation of alcohols with nitriles in dry media Halmuthur M. Sampath Kumar,* Basi V. Subba Reddy, Sheri Anjaneyulu, Etukala Jagan Reddy and Jhillu S. Yadav Organic Division-I, Indian Institute of Chemical T echnology, Hyderabad 500 007, India Received (in Montpellier, France) 3rd June 1999, Accepted 4th August 1999 Montmorillonite KSF catalyses the Ritter reaction in which several benzyl, allyl and tertiary alcohols are converted into amides in high yields, when heated with various nitriles in the absence of solvent. A comparative study of conventional vs.microwave irradiation for this clay catalysed reaction is presented. The reaction of alcohols with nitriles in protic acid, popularly known as the Ritter reaction,1 leads to amides.The carbonium ion formed by the reaction of a strong protic acid and alcohols or ole–ns reacts with nitriles to generate a nitrilium ion, which is subsequently trapped with water resulting in Nsubstituted amides. The classical method, which utilizes conc. (85»90%), is generally suitable for tertiary alcohols, H2SO4 but this strongly acidic medium limits its use, especially for acid sensitive substrates. In order to circumvent serious side reactions2 in the sulfuric acid approach, several modi–cations and improvements have been attempted and the protic acid is often replaced by Lewis acids such as or AlCl3 , FeCl3 SnCl4 and the alcohols by alkyl halides.3 is used as a BF3 Æ OEt24 catalyst for selective amidation of benzyl alcohols ; however, this method is not suitable for substrates bearing electron withdrawing groups (e.g., 4-nitrobenzyl alcohols), 2° and 3° alcohols including aryl-alkyl carbinols and is inefficient in the case of a,b-unsaturated (cinnamyl) alcohols.Recently, rare earth exchanged HY-zeolite5 has been utilized for the Ritter amidation of alcohols ; however, the conversions were only partial even after 24 h of heating.Even though a clayfen (clay supported ferric nitrate) catalysed Ritter reaction has been reported6 under conventional conditions, the method utilizes tert-alkyl halides as substrates and the yields reported are low and require long reaction times. In continuation of our work on eco-friendly organic transformations,7 we present here a convenient Ritter amidation of alcohols catalysed by montmorillonite KSF under solvent-free conditions.We –rst examined the reaction of a number of alcohols and nitriles under microwave heating by irradiating a mixture of clay and equimolar quantitites of the reactants in an open pyrex test tube and complete conversion was asertained by tlc monitoring. In order to precisely assess the efficiency of microwave heating, parallel experiments were conducted under conventional heating, by keeping the test tube containing the dry reaction mass immersed in a hot oil bath (preheated to 110» 120 °C, analogous to the highest observed temperature during irradiation), with occasional mixing.From the results summerized in Table 1, it is evident that both modes of heating aÜord good yields of amides. Obviously with the microwave method conversions are faster (2»5 min), whereas the conventional mode of heating required much longer reaction times in order to attain comparable yields.Further, the clay catalysed Ritter reaction was found to be highly selective for benzylic, allylic and tertiary alcohols, yet general with regard to nitriles, which are the other reaction ingredients. Unlike Lewis acid catalysed Ritter amidation, our method is applicable to electron de–cient benzyl alcohols, 2° and 3° aryl-alkyl carbinols and tertiary alcohols like tert-butanol and all these substrates formed amides in good yields.Furthermore, no appreciable thermal decomposition of the product was observed under either mode of heating. Despite the possibilities for multiple reaction pathways in an overall conversion that proceeds through two high energy formally charged intermediates, the conversions in clay catalysed reactions were generally found to be clean.DiÜerent functional groups could tolerate these reaction conditions and no deallylation, dealkylation or debenzylation were observed, which are otherwise commonly encountered in any acid catalysed reaction.However, minor quantities (5»8%) of dimeric ethers were isolated along with the amides wherever primary benzyl or allyl alcohols were employed, which is in conformity with earlier reports.8 Aryl-alkyl carbinols and tertiary alcohols also exclusively formed amides in high yields without any elimination by-products derived from alcohols. No reaction occurred in the case of primary aliphatic alcohols and no trace of acids derived from hydrolysis of nitriles were detected in the crude products.From these results, it is logical to presume that the selectivity observed during clay catalysed Ritter amidation may possibly be due to the stabilization of cationic intermediates in the interlamellar layers, as encountered in many clay catalysed reactions.9 The catalyst could be eÜectively reused for four cycles (after washing with methanol and drying at 120 °C for 5 h) without any apparent loss of activity ; for example, the benzyl alcohol and benzyl cyanide combination under microwave irradiation6 gave 82%, 80%, 80%, and 81% yields over 4 cycles.These results clearly show the advantages of our method over protic and Lewis acid catalysed Ritter reactions.In conclusion the montmorillonite catalysed Ritter reaction presented in this paper is a convenient and high yielding method for the amidation of alcohols. Other merits such as inexpensive and reusable catalyst, solvent free reaction conditions and application of microwave for achieving quick and clean conversions, make our method eco-friendly and economical and hence it may –nd uses in organic synthesis.Experimental In a typical experiment 1-phenylethanol (1.2 g, 10 mmol) and 3-ethoxypropionitrile (1 g, 10 mmol) were admixed with KSF clay (1.2 g, w/w of alcohol) and subjected to microwave irradiation in a pyrex test tube for 3 min. Then the reaction mass was cooled to room temperature, charged on a short silica gel column (Merck, 200 mesh) and eluted (ethyl acetate»n-hexane, 2: 8) to aÜord the pure amide as a pale yellow liquid (2.05 g, 93%). 1H NMR d 1.15 (t, 3H, J\7.5 Hz), 1.45 (d, (CDCl3) : 3H, J\6.25 Hz), 2.4 (t, 2H, J\6.25 Hz), 3.5 (q, 2H, J\7.5 Hz), 3.65 (t, 2H, J\6.25 Hz), 5.15 (dd, 1H, J\7.5 and 6.5 Hz), 6.85 (s br, 1H), 7.3 (m, 5H). Acknowledgements and S.A. thank CSIR, New Delhi, for the award of B.V.S. fellowship.New J. Chem., 1999, 23, 955»956 955 This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scienti–que 1999 (Table 1 Amidation of alcohols with nitrilesa Microwaveb Conventional irradiation heating Time/ Yieldc/ Time/ Yieldc/ R1 R2 R3 R4 min % h % H H Ph H3CCH2CH2 5 86(6) 10 74(5) H H Ph PhCH2 4 83(5) 8 72(7) H H Ph Ph 4 80(8) 8 70(7) H H 4-NO2C6H4 PhCH2 6 75 12 68 H CH3 Ph EtOCH2CH2 3 93 8 85 H H 4-MeOC6H4 Ph 4 91 8 87 H CH3 Ph CH22CH 3 84 6 73 H H 4-BnOC6H4 PhCH2 4 89 7 78 H CH3 Ph C(CH3)3 4 87 5 75 H Ph Ph CH3 3 85 8 77 H Ph Ph Cl(CH2)2CH2 4 90 7 82 CH3 CH3 CH3 PhCH2 3 87 6 74 CH3 CH3 CH3 Ph 3 83 6 72 CH3 Ph Ph CH3CH2CH2 3 85 8 76 CH3 Ph Ph ClCH2 4 88 7 80 H H PhCH2CH PhCH2 2 78(5) 5 67(6) H H PhCH2CH CH3CH2CH2 3 80(8) 4 70(5) H H CH3 CH2CH PhCH2 2 75(6) 5 65(8) H H CH3 CH2CH CH3CH2CH2 3 77(6) 4 68(8) H H CH3 (CH2)6CH2 CH3CH2CH2 No reaction a All products were characterised by IR, 1H NMR, mass spectra and by comparison with known compounds.b Irradiations were carried out at an output of 650 W in a domestic microwave oven BPL BMO 700T (unoptimised). c Isolated yield after chromatography.Yields indicated in parentheses correspond to dimeric ethers. References 1 (a) J. J. Ritter and P. P. Minier, J. Am. Chem. Soc., 1948, 70, 4045. (b) L. I. Krimen and D. J. Cota, Org. React. (NY ), 1969, 17, 213. 2 (a) S. Top and G. Jaouen, J. Org. Chem., 1981, 46, 78. (b) A. Garcia Martinez, R. Martinez Alvarez, E. Teso Vilar, A. Garcia Fraile, M. Hanack and L. R. Subramanian, T etrahedron L ett., 1989, 30, 581.(c) T. Kiersznicki and R. Mazurkiewicz, ROCZ Chem., 1977, 51, 1021; Chem. Abstr., 1978, 88, 6504f. (d) D. H. R. Barton, P. D. Magnus and R. N. Young, J. Chem. Soc., Chem. Commun., 1973, 331. (e) D. H. R. Barton, P. D. Magnus, J. A. Garbarino and R. N. Young, J. Chem. Soc., Perkin T rans 1., 1974, 1, 2101. ( f) M. C. Cesa, US Pat. 5103055 A, 1992; Chem.Abstr., 1992, 115, 235100k. (g) Y. Takahashi, Y. Fukoaka, K. Sasaki and S. Senco, US Pat. 68»748165, 1968; Chem. Abstr., 1973, 79, 91627e. (h) Israel Min. Ind. Inst. Res. Develop., GB Pat. 1198680, 1970; Chem. Abstr., 1970, 72, 766833c. 3 H. Firouzabadi, A. R. Sardarian and H. Badparva, Synth. Commun., 1994, 24, 601. 4 M. Kacan and A. McKillop, Synth. Commun., 1993, 23, 2185. 5 A. R. A. S. Deshmukh, V. K. Gumaste and B. M. Bhawal, Ind. J. Chem., 1997, 36B, 369. 6 P. Eugenio, Bull. Soc. Chim. Belg., 1985, 94, 81. 7 (a) H. M. Sampath Kumar, P. K. Mohanty, M. S. Kumar and J. S. Yadav, Synth. Commun., 1997, 27, 1327. (b) H. M. Samapth Kumar, B. V. Subba Reddy, P. K. Mohanty and J. S. Yadav, T etrahedron L ett., 1997, 38, 3619. (c) H. M. Sampath Kumar, B. V. Subba Reddy and J. S. Yadav, Chem. L ett., 1998, 637. (d) H. M. Sampath Kumar, B. V. Subba Reddy, S. Anjaneyulu and J. S. Yadav, Synth. Commun., 1998, 28, 3811. 8 T. Li, H. Li, J. Guo and T. Jin, Synth. Commun., 1996, 26, 2497. 9 A. R. A. S. Deshmukh, V. K. Gumaste and B. M. Bhawal, Synth Commun., 1995, 25, 3939 and references cited therein. L etter 9/04464G 956 New J. Chem., 1999, 23, 955»956

ISSN:1144-0546    

DOI:10.1039/a904464g

出版商:RSC    

年代:1999

数据来源: RSC

摘要:

L e t t e r Structural characterization of an ordered crystalline modi–cation of [In(SePh)3 ]= Matthew C. Kuchta,a Arnold L. Rheingoldb and Gerard Parkina a Department of Chemistry, Columbia University, New Y ork, NY 10027, USA b Department of Chemistry and Biochemistry, University of Delaware, Newark, DE 19716-2522, USA Received (in New Haven, USA) 28th June 1999, Accepted 2nd August 1999 A triclinic modi–cation of has been structurally [In(SePh)3 ]= characterized by X-ray diÜraction, demonstrating that it is chemically distinct from a previously reported monoclinic form.Thus, triclinic is crystallographically ordered and [In(SePh)3 ]= contains –ve-coordinate trigonal bipyramidal indium centers with asymmetrically bridging [SePh] ligands, whereas the monoclinic form is disordered and contains six-coordinate octahedral indium centers with symmetrically bridging [SePh] ligands.Until recently, the chemistry of selenolate and tellurolate derivatives had received relatively little attention by comparison with alkoxide and thiolate complexes, but is now an area of much interest.1 For example, with respect to the heavier Group 13 elements, research in chalcogenolate complexes has been prompted by their potential use as single-source precursors to ìì13»16œœ materials.2,3 The structural chemistry of Group 13/16 compounds, however, is still in its infancy,4 and in this paper we report a structure for that [In(SePh)3]= diÜers signi–cantly from that previously reported.5 As a result of the inherent electrophilic nature of the indium center in a three-coordinate molecule, such species [InX3] typically exist as some form of oligomer.For example, the phenylselenolate complexes and [Np2In(l-SePh)]2 6 are dimers with symmetrically bridging [Mes2In(l-SePh)]2 3 [SePh] ligands. Monomeric three-coordinate selenolate complexes can normally only be attained by incorporation of very bulky substituents on either indium, for example (Mes*\2,4,6- or selenium, such Mes*In(SePh)2 Bu3 t C6H2),7 as and (E\C, Si).10 It In(SeMes*)3 8,9 In[SeE(SiMe3)3]3 is, therefore, not surprising that the parent arylselenolate, possesses a polymeric structure (Fig. 1),5 in [In(SePh)3]=,11 which each indium atom relieves its electron de–ciency by octahedral coordination of six bridging [SePh] groups. The phenyl groups of this monoclinic form of [In(SePh)3]=, however, were observed to be severely disordered over two positions. In view of this disorder, we considered it appropriate to attempt to obtain the structure of an ordered material.Signi–cantly, we succeeded in obtaining triclinic crystals of that were devoid of disorder and the structure, [In(SePh)3]= Fig. 1 A portion of the polymeric structure of monoclinic (data taken from ref. 5). For clarity, only the ipso carbon [In(SePh)3]= atoms of the grossly disordered phenyl groups are shown. as determined by X-ray diÜraction, is illustrated in Fig. 2»4. Selected bond lengths and angles are listed in Table 1. More important than simply representing an ordered structure, the molecular structure of the triclinic form of is of [In(SePh)3]= interest because it is substantially diÜerent from that previously reported5 in a number of respects, as summarized in Fig. 2 ORTEP drawing illustrating the asymmetric unit of triclinic [In(SePh)3]= . Fig. 3 A portion of the polymeric structure of triclinic [In(SePh)3]=. Table 1 Selected bond lengths and angles (°) for triclinic (”) [In(SePh)3]= In1»Se1 2.556(1) In2»Se2@ 2.596(1) In1»Se2 2.966(1) In2»Se3@ 2.856(1) In1»Se3 2.610(1) In2»Se4 2.986(1) In1»Se4 2.604(1) In2»Se5 2.624(1) In1»Se5 2.835(1) In2»Se6 2.549(1) In1»Se2»In2@ 90.61(3) In1»Se4»In2 88.88(3) In1»Se3»In2@ 92.80(3) In1»Se5»In2@ 91.82(3) New J.Chem., 1999, 23, 957»959 957 This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scienti–que 1999 (Fig. 4 The trigonal bipyramidal coordination geometry in (carbon and hydrogen atoms are omitted for clarity). [In(SePh)3]= Table 2. Firstly, instead of being six-coordinate, the indium centers of the triclinic structure are –ve-coordinate. Thus, each indium center is coordinated to one terminal and four bridging [SePh] ligands in a trigonal bipyramidal geometry that is most clearly seen in Fig. 4. A second distinction between the two structures is that the [SePh] ligands of the triclinic structure do not bridge the indium centers symmetrically, but are asymmetrically disposed. Speci–cally, whereas the In»Se bond lengths in the monoclinic structure fall in the narrow range 2.77»2.79 those of the bridging ligands in the triclinic form ”,5 span the substantial range of 2.56»2.99 (Table 1).The asym- ” metry of the bridging In»Se interactions is a consequence of each bridging [SePh] ligand occupying an axial position to one indium but an equatorial position to the other. Of the two types of bridging interactions, the equatorial In»Se bonds are considerably shorter than the axial ones (Table 3), as is often observed for trigonal bipyramidal complexes of the main group elements.12 It is interesting to note that the axial In»Se bond lengths also fall into two classes, with In1»Se2 and In2»Se4 (2.98 average) being longer than In1»Se5 and In2» ” Se3@ (2.85 average). Another distinction between the [SePh] ” coordination modes in the triclinic and monoclinic structures Table 2 Comparison of triclinic and monoclinic forms of [In(SePh)3]= Triclinic Monoclinica Ordered/disordered Ordered Disordered In coordination Five-coordinate (T BPY ) Six-coordinate (Oh) [SePh] bridging Asymmetric Symmetric interaction Range of d(In»SePh) 2.56»2.99 ” 2.77»2.79 ” In… … …In separation 3.94 ” 3.63 ” Calcd density 2.09 g cm~3 2.29 g cm~3 a Ref. 5. Table 3 Summary of the average trigonal bipyramidal coordination environment about each indium in triclinic [In(SePh)3]= d(In»Se)/” In»SePh (eq, term) 2.56 In»SePh (eq, bridge) 2.61 In»SePh (axial, short) 2.85 In»SePh (axial, long) 2.98 of is that one of the equatorial ligands attached to In(SePh)3 each indium in the triclinic form remains terminal ; as would be expected, this ligand has the shortest In»Se bond length (2.55 average).It is worthwhile to point out that, in view of ” these diÜerences in bonding, the monoclinic and triclinic forms of are not polymorphs of each other since [In(SePh]3]= they are chemically distinct.13 Finally, the average In………In separation of 3.94 in tri- ”14 clinic is signi–cantly greater than that in its [In(SePh)3]= monoclinic counterpart (3.63 Correspondingly, the calcu- ”).5 lated density of the monoclinic form (2.29 g cm~3)5 is greater than that for the triclinic form (2.09 g cm~3). The shorter separation in the latter complex is presumably due to each [In… … …In] unit being bridged by three [SePh] ligands, which therefore requires a closer approach of the indium centers.In this regard, the complex [MeIn(SePh)(l-SePh)] with only =,15 a single bridging [SePh] ligand, has an even greater In… … …In separation (4.19 than that in doubly bridged triclinic ”) [In(SePh)3]=. It is also worth noting that the In… … …In separation in triclinic is greater than the values for other com- [In(SePh)3]= plexes that possess similar [In(l-SePh) cores (Table 4). 2In] This diÜerence is most probably a result of the fact that the indium centers in triclinic are –ve-coordinate, [In(SePh)3]= whereas those in the other complexes, for example [R2In(lare four-coordinate. Thus, the greater In… … …In SePh)]2 , separation in triclinic is merely a consequence of [In(SePh)3]= the substantially longer In»SePh bond to the axial site.18 In conclusion, has been shown to exist in a [In(SePh)3]= triclinic form that diÜers from the previously reported monoclinic structure in a number of important respects.Speci–cally, triclinic is crystallographically ordered and con- [In(SePh)3]= tains –ve-coordinate trigonal bipyramidal indium centers with asymmetrically bridging [SePh] ligands, whereas the monoclinic form is disordered and contains six-coordinate octahedral indium centers with symmetrically bridging [SePh] ligands. As such, the monoclinic and triclinic forms are chemically distinct.Experimental All manipulations were performed using a combination of glovebox, high-vacuum or Schlenk techniques.19 was [In(SePh)3] prepared according to the literature method11 and crystals suitable for X-ray diÜraction were obtained by evaporation of a toluene solution at room temperature. Table 4 Summary of In»Se and In… … …In distances in complexes [XaIn(SePh)b] d(In»SePh)/” d(In»SePh)/” Terminal Bridging d(In… … …In)/” Reference [In(SePh)3]= (triclinic) 2.56 2.61eq , 2.92ax 3.94av This work [In(SePh)3]= (monoclinic) » 2.78 3.63 5 [Mes2In(l-SePh)]2 » 2.73 3.86 15 [Np2In(l-SePh)]2 » 2.74 3.75 6 [MeIn(SePh)(l-SePh)]= 2.54 2.68 4.19 15 [NpIn]2(l-PBut)(l-SePh) » 2.76 3.95 16 [Ph4P][In(SePh)4] 2.58 » » 17 [Ph4P][In(SePh)3(SeH)] 2.55 » » 17 958 New J.Chem., 1999, 23, 957»959Table 5 Crystal, intensity collection and re–nement data for [In(SePh)3]= Formula C18H15InSe3 Formula weight 583 Lattice Triclinic Space group P1 (no. 2) a/” 7.3207(1) b/” 11.0326(2) c/” 23.3628(4) a/° 81.490(1) b/° 86.265(1) c/° 84.381(1) U/”3 1854.73(5) Z 4 Temperature/K 222 l(Mo-Ka)/mm~1 7.155 No. meas. re—. 7960 No. independ. re—. 7960 R1 0.0533 wR2 0.1048 Crystallographic data for were collected on a In(SePh)3 Siemens P4 diÜractometer equipped with a SMART CCD detector, as summarized in Table 5.The structure was solved using direct methods and standard diÜerence map techniques, and was re–ned by full-matrix least-squares procedures using SHELXTL.20 Hydrogen atoms attached to carbon were included in calculated positions.CCDC reference number 440/136. See http ://www.rsc.org/ suppdata/nj/1999/957/ for crystallographic –les in .cif format. Acknowledgements thank the National Science Foundation (CHE 96-10497) We for support of this research. Dr Brian Bridgewater is thanked for helpful assistance. References 1 For a comprehensive review of selenolate and tellurolate complexes, see : J.Arnold, Prog. Inorg. Chem., 1995, 43, 353. 2 For a review of organometallic Group 13»chalcogen derivatives, see : J. P. Oliver, J. Organomet. Chem., 1995, 500, 269. 3 See, for example: H. Rahbarnoohi, R. Kumar, M. J. Heeg and J. P. Oliver, Organometallics, 1995, 14, 3869. 4 H. Rahbarnoohi, R. L. Wells, L. M. Liable-Sands, G. P. A. Yap and A. L. Rheingold, Organometallics, 1997, 16, 3959. 5. T. A. Annan, R. Kumar, H. E. Mabrouk, D. G. Tuck and R. K. Chadha, Polyhedron, 1989, 8, 865. 6 O. T. Beachley, J. C. Lee, Jr., H. J. Gysling, S. H. L. Chao M. R. Churchill and C. H. Lake, Organometallics, 1992, 11, 3144. 7 H. Rahbarnoohi, R. L. Wells, L. M. Liable-Sands and A. L. Rheingold, Organometallics, 1996, 15, 3898. 8 K. Ruhlandt-Senge, and P. P. Power, Inorg. Chem., 1993, 32, 3478. 9 The gallium analogue is also a monomer. See: K. Ga(SeMes*)3 Ruhlandt-Senge and P. P. Power, Inorg. Chem., 1991, 30, 3683. 10 S. P. Wuller, A. L. Seligson, G. P. Mitchell and J. Arnold, Inorg. Chem., 1995, 34, 4854. 11 R. Kumar, H. E. Mabrouk and D. G. Tuck, J. Chem. Soc., Dalton T rans., 1988, 1045. 12 For example, the axial In»Cl bonds in (2.50 and [Ph4P]2[InCl5] 2.52 are longer than the equatorial bonds (2.41»2.42 See: W.”) ”). Bubenheim, G. Frenzen and U. Mué ller, Acta Crystallogr., Sect. C, 1995, 51, 1120. 13 Speci–cally, polymorphs diÜer only in packing of the repeat units. See, for example: Macmillan Encyclopedia of Chemistry, ed. J. J. Lagowski, Simon & Schuster Macmillan, New York, 1997, vol. 3, p. 1237. 14 The crystallographically unique values are 3.924 and 3.962 ”. 15 H. Rahbarnoohi, R. Kumar, M. J. Heeg and J. P. Oliver, Organometallics, 1995, 14, 3869. 16 O. T. Beachley, Jr., S.-H. L. Chao, M. R. Churchill and C. H. Lake, Organometallics, 1993, 12, 5025. 17 D. M. Smith and J. A. Ibers, Polyhedron, 1998, 17, 2105. 18 It should be noted that the In………In separation also increases when bridging ligands that reduce the degree of pucker in the 4- membered ring, such as are present. See, for example ref. 16. PBu2 t , 19 (a) J. P. McNally, V. S. Leong and N. J. Cooper, in Experimental Organometallic Chemistry, ed. A. L. Wayda and M. Y. Darensbourg, American Chemical Society, Washington, DC, 1987, ch. 2, pp. 6»23. (b) B. J. Burger, and J. E. Bercaw, in reference 19(a), ch. 4, pp. 79»98. (c) D. F. Shriver and M. A. Drezdzon, T he Manipulation of Air-Sensitive Compounds, Wiley-Interscience, New York, 2nd edn. 1986. 20 G. M. Sheldrick, SHELXTL, An Integrated System for Solving, Re–ning and Displaying Crystal Structures from DiÜraction Data, University of Goé ttingen, Goé ttingen, Federal Republic of Germany, 1981. L etter 9/05250J New J. Chem., 1999, 23, 957»959 959

ISSN:1144-0546    

DOI:10.1039/a905250j

出版商:RSC    

年代:1999

数据来源: RSC

摘要:

L e t t e r Synthesis and molecular structure of bis(pyrazolyl)(3,5-di-tertbutylpyrazolyl) hydroborato thallium : a hetero-tris(pyrazolyl)- hydroborato ligand derived from two diÜerent pyrazoles Prasenjit Ghosh, David G. Churchill, Mark Rubinshtein and Gerard Parkin* Department of Chemistry, Columbia University, New Y ork, NY 10027, USA Received (in New Haven, USA) 20th July 1999, Accepted 20th August 1999 The hetero-tris(pyrazolyl)hydroborato thallium complex derived from two diÜerent pyrazoles, may [HB(pz)2(pzBu2 t ) ] Tl, be obtained by the reaction of with a mixture of pyra- LiBH4 zole and 3,5-di-tert-butylpyrazole, followed by treatment with Tl(O2CMe).Tris(pyrazolyl)hydroborato ligands are presently a prominent feature in coordination chemistry.1 In part, the interest in such ligands derives from the facile ability to modify the steric and electronic properties by incorporation of a variety of different substituents. Without exception, however, structurally characterized tris(pyrazolyl)hydroborato ligands are presently restricted to derivatives in which all three pyrazolyl groups are derived from the same pyrazole.2,3 In this paper, we report the synthesis and structure of a hetero-tris(pyrazolyl)hydroborato thallium complex that is derived from two diÜerently substituted pyrazoles.We have recently described that asymmetrically substituted hetero-bis(pyrazolyl)hydroborato ligands may be obtained by the direct reaction of with a mixture of two diÜerent LiBH4 pyrazoles.4 Ligands that have been constructed using this method include and [H2B(pz)(pzBu2t )]~, [H2 B(pzMe2)(pzBu2t )]~ We now report that hetero- [H2B(pzTrip)(pzBu2t )]~.tris(pyrazolyl)hydroborato complexes derived from two diÜerent pyrazoles may also be obtained by this method. Speci–- cally, the mixed bis(pyrazolyl)(3,5-di-tert-butylpyrazolyl)- hydroborato complex may be synthesized [HB(pz)2(pzBu2t )]Li by reaction of with pyrazole followed by 3,5-di-tert- LiBH4 butylpyrazole (Scheme 1).Subsequent treatment of with yields the thallium [HB(pz)2(pzBu2t )]Li Tl(O2CMe) complex which has been structurally char- [HB(pz)2(pzBu2t )]Tl, acterized by X-ray diÜraction. The molecular structure of is shown in [HB(pz)2(pzBu2t )]Tl Fig. 1, with selected bond lengths and angles listed in Table 1. Inspection of Table 1 indicates that the pyrazolyl groups are Scheme 1 not coordinated identically, with the bond length Tl»N(pzBu2t ) [2.741(4) being considerably longer than those for the ”] unsubstituted pyrazolyl groups [2.526(4) and 2.588(4) ”].While such a diÜerence may be attributed to steric factors, it is worthwhile noting that an opposite trend is observed for the homoleptic counterparts, and [Tp]Tl6 (Table 2).7 [TpBu2t ]Tl5 Thus, the unsubstituted [Tp]Tl complex exhibits both a greater range of Tl»N bond lengths, and a greater average Tl»N bond length, than does On the other hand, [TpBu2t ]Tl.8 the Tl»N bond lengths in the hetero-bis(pyrazolyl)hydroborato complex are comparable: d[Tl» [H2B(pz)(pzBu2t )]Tl N(pz)]\2.65 and It is, therefore, ” d[Tl»N(pzBu2t )]\2.68 ”.evident that the Tl»N bond lengths in these complexes must be in—uenced by rather subtle factors, one of which is likely to be ìì crystal packing eÜectsœœ,9 especially since deriv- [TpRR{]Tl atives are known to exhibit a variety of packing arrangements. 10 In this regard, the molecules pack [HB(pz)2(pzBu2t )]Tl in the crystal such that each thallium atom exhibits weak interactions with four nitrogen atoms of the pyrazolyl groups of an adjacent molecule, as illustrated in Fig. 2. As would be Fig. 1 Molecular structure of [HB(pz)2(pzBu2t )]Tl. Table 1 Selected bond lengths and angles (°) (”) Tl»N12 2.526(4) N12»Tl»N22 77.7(1) Tl»N22 2.588(4) N12»Tl»N32 71.0(1) Tl»N32 2.741(4) N22»Tl»N32 71.1(1) New J. Chem., 1999, 23, 961»963 961 This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scienti–que 1999 (Table 2 Comparison of Tl»N bond lengths in [Tp]Tl and The range is given in parentheses [HB(pz)2(pzBu2t )]Tl, [TpBu2t ]Tl.[HB(pz)2(pzBu2t )]Tla [Tp]Tlb [TpBu2t ]Tlc d[Tl»N(pz)]/” 2.56 (2.53»2.59) 2.63 (2.55»2.73) d[Tl»N(pzBu2t )]/” 2.74 2.57 (2.54»2.59) a This work. b Ref. 6. c Ref. 5. expected, these intermolecular Tl… … …N interactions, which are in the range 3.65»3.85 are considerably longer than the cor- ”, responding intramolecular bond lengths (2.53»2.74 ”).In summary, a hetero-tris(pyrazolyl)hydroborato ligand derived from two diÜerent pyrazoles may be obtained by the direct reaction of with a mixture of pyrazole and 3,5- LiBH4 di-tert-butylpyrazole. The ability to synthesize such ligands is noteworthy since it provides a means by which the steric environment about a metal center may be modulated in an asymmetric manner.Finally, it is likely that this approach may be adapted for other pyrazole derivatives. Experimental All manipulations were performed using a combination of glovebox, high-vacuum and Schlenk techniques.11 Solvents were puri–ed and degassed by standard procedures. 1H and 13C chemical shifts are reported in ppm relative to (d\0) SiMe4 and were referenced internally with respect to the protio solvent impurity or the 13C resonances, respectively. All coupling constants are reported in Hz. C, H, and N elemental analyses were measured using a Perkin-Elmer 2400 CHN Elemental Analyzer. Di-tert-butylpyrazole was prepared by the literature method.12 Synthesis of and [HB(pz)2(pzBu2 t ) ]Li [HB(pz)2(pzBu2 t ) ]Tl A solution of in THF (7.4 mL of 2M, 14.8 mmol) was LiBH4 added to a suspension of pyrazole (2.00 g, 29.4 mmol) and toluene (ca. 20 mL) in a large glass ampoule. The mixture was stirred overnight at room temperature, after which the volatile components were removed in vacuo. 3,5-Di-tert-butylpyrazole (2.65 g, 14.7 mmol) and toluene (ca. 30 mL) were added and the mixture was heated in the glass ampoule at ca. 120 °C overnight (CARE! Pressures of several atmospheres may be generated). The mixture was allowed to cool to room tem- Fig. 2 Interaction of the Tl atom of with an adja- [HB(pz)2(pzBu2t )]Tl cent molecule. Selected separations Tl»N21@ 3.77, Tl»N22@ 3.85, (”) : Tl»N31@ 3.65 Tl»N32@ 3.71 ”, ”.perature, depositing crude as a white solid, [HB(pz)2(pzBu2t )]Li which was isolated by –ltration. The volatile components were removed from the –ltrate and the residue obtained was washed with pentane, giving pure (0.30 g). [HB(pz)2(pzBu2t )]Li 1H NMR 400 MHz): 1.08 [s, 9H, 1.54 [s, (C6D6 , 1C(CH3)3], 9H, 6.04 [br s, 1H, 6.10 [2H of 1C(CH3)3], C3N2HBu2 t ], 7.69 [2H of (2H of obscured 2C3N2H3], 2C3N2H3] 2C3 N2H3 by solvent).(3.87 g, 14.7 mmol) and THF (ca. 60 mL) were Tl(O2CMe) added to the crude obtained above and the [HB(pz)2(pzBu2t )]Li mixture was stirred overnight at room temperature and –ltered. The volatile components were removed in vacuo and the residue obtained was washed with pentane to give as a white solid (1.30 g). Total yield of [HB(pz)2(pzBu2t )]Tl and 25%.The thal- [HB(pz)2(pzBu2t )]Li [HB(pz)2(pzBu2t )]Tl : lium complex was also obtained by an analogous procedure, but reversing the order of addition of pyrazole and 3,5-di-tertbutylpyrazole. Anal. calcd for C, 38.6%; H, C17H26N6BTl: 5.0%; N, 15.9%. Found: C, 37.7%; H, 4.7%; N, 15.9%. IR (cm~1), KBr pellet : 2532 [m(BH)]. 1H NMR 300 (C6D6 , MHz): 1.29 [s, 9H, 1.49 [s, 9H, 6.05 [s, 1C(CH3)3], 1C(CH3)3], 1H, 6.10 [t, J\2, 2H of 7.36 [d, C3N2HBu2 t ], 2C3N2H3], J\1,13 2H of 7.69 [d, J\2, 2H of 2C3N2H3], 2C3N2H3]. 13C NMR 125.76 MHz): 31.2 [q, (C6D6 , 1JCH\125, 32.1 [q, 32.2 [s, 1C(CH3)3], 1JCH\125, 1C(CH3)3], 32.3 [s, 100.7 [d, 1C of 1C(CH3)3], 1C(CH3)3], 1JCH\171, 104.5 [d, 2C of 136.3 [d, C3N2HBu2 t ], 1JCH\175, 2C3N2H3], 2C of 138.9 [d, 2C of 1JCH\184, 2C3N2H3], 1JCH\182, 156.8 [s, 1C of 161.2 [s, 1C of 2C3N2H3], C3N2HBu2 t ], C3N2HBu2 t ].X-Ray structure determination Crystal data, data collection and re–nement parameters for are summarized in Table 3. X-Ray diÜrac- [HB(pz)2(pzBu2t )]Tl tion data were collected on a Bruker P4 diÜractometer equipped with a SMART CCD detector using graphite monochromated Mo-Ka X-radiation (k\0.710 73 The struc- ”).Table 3 Crystal, intensity collection and re–nement data for [HB(pz)2(pzBu2t )]Tl Formula C17H26BN6Tl Formula weight 529.62 Lattice Monoclinic Space group P21/n (no. 14) a/” 10.300(3) b/” 10.286(3) c/” 18.917(5) a/° 90 b/° 99.770(7) c/° 90 U/”3 1975(1) Z 4 T /K 233 l(Mo-Ka)/mm~1 8.189 No. of data 4224 No. of parameters 237 R1 a 0.0375 wR2 a 0.0849 and a R1\& Mp Fo o [ o Fc p N/& o Fo o wR2\ for [I[2p(I)].M&[w(Fo2[Fc2)2]/&[w(Fo2)2]N1@2 962 New J. Chem., 1999, 23, 961»963ture was solved using direct methods and standard diÜerence map techniques, and was re–ned by full-matrix least-squares procedures using SHELXTL.14 Hydrogen atoms on carbon were included in calculated positions, but the hydrogen attached to boron was freely re–ned. Systematic absences were consistent uniquely with (no. 14). P21/n CCDC reference number 440/139. See http ://www.rsc.org/ suppdata/nj/1999/961/ for crystallographic –les in .cif format. Acknowledgements thank the National Science Foundation (CHE 96-10497) We for support of this research. References 1 For recent reviews, see : (a) S. Tro–menko, Chem.Rev., 1993, 93, 943; (b) G. Parkin, Adv. Inorg. Chem., 1995, 42, 291; (c) N. Kitajima and W. B. Tolman, Prog. Inorg. Chem., 1995, 43, 419; (d) I. Santos and N. Marques, New. J. Chem., 1995, 19, 551; (e) D. L. Reger, Coord. Chem. Rev., 1996, 147, 571; ( f ) M. Etienne, Coord. Chem. Rev., 1997, 156, 201; (g) P. K. Byers, A. J. Canty and R. T. Honeyman, Adv. Organomet. Chem., 1992, 34, 1. 2 However, a nickel complex, [HB(pz)2(4-cyano-1-pyrazolyl)]2Ni, has been cited as unpublished results in a 1971 review article, but no experimental details, spectroscopic or structural data have been reported to date. See: S. Tro–menko, Acc. Chem. Res., 1971, 4, 17. 3 Tris(pyrazolyl)hydroborato ligands with diÜerently substituted pyrazolyl groups have been obtained in a few instances where a 1,2-borotropic rearrangement results in the 3- and 5-substituents exchanging positions.Nevertheless, these ligands are still derived from the same pyrazolyl fragment. For examples, see ref. 1(a) and (a) A. L. Rheingold, L. M. Liable»Sands, G. P. A. Yap and S. Tro–menko, Chem. Commun., 1996, 1233; (b) M. H. Chisholm, N. W. Eilerts and J. C. HuÜman, Inorg. Chem., 1996, 35, 445; (c) M.Cano, J. V. Heras, A. Monge, E. Pinilla, E. Santamaria, H. A. Hinton, C. J. Jones and J. A. McCleverty, J. Chem. Soc., Dalton T rans., 1995, 2281; (d) A. L. Rheingold, C. B. White and S. Tro–- menko, Inorg. Chem., 1993, 32, 3471; (e) M. Cano, J. V. Heras, C. J. Jones, J. A. McCleverty and S. Tro–menko, Polyhedron, 1990, 9, 619; ( f ) S. Tro–menko, J. C. Calabrese, P.J. Domaille and J. S. Thompson, Inorg. Chem., 1989, 28, 1091; (g) J. C. Calabrese and S. Tro–menko, Inorg. Chem., 1992, 31, 4810; (h) M. Cano, J. V. Heras, S. Tro–menko, A. Monge, E. Gutierrez, C. J. Jones and J. A. McCleverty, J. Chem. Soc., Dalton T rans., 1990, 3577. 4 P. Ghosh, T. Hascall, C. Dowling and G. Parkin, J. Chem. Soc., Dalton T rans., 1998, 3355. 5 C. M. Dowling, D.Leslie, M. H. Chisholm and G. Parkin, Main Group Chem., 1995, 1, 29. 6 C. Janiak, S. Temizdemir and T. G. Scharmann, Z. Anorg. Allg. Chem., 1998, 624, 755. 7 For further reviews of thallium tris(pyrazolyl)hydroborato complexes, see ref. 1(b) and (a) C. Janiak, Coord. Chem. Rev., 1997, 163, 107; (b) C. Janiak, Main Group Met. Chem., 1998, 21, 33. 8 Furthermore, the average Tl»N bond length in is com- [TpBu2t ]Tl parable to that in (2.58 with tert-butyl substituents [TpBu2t ]Tl ”), on only the 3-position.See: A. H. Cowley, R. L. Geerts, C. M. Nunn and S. Tro–menko, J. Organomet. Chem., 1989, 365, 19. 9 A. Martïç n and A. G. Orpen, J. Am. Chem. Soc., 1996, 118, 1464. 10 See, for example, ref. 7. 11 (a) J. P. McNally, V. S. Leong and N. J. Cooper, in Experimental Organometallic Chemistry, ed A.L. Wayda and M. Y. Darensbourg, American Chemical Society, Washington, DC, 1987, ch. 2, pp. 6»23; (b) B. J. Burger and J. E. Bercaw, in Experimental Organometallic Chemistry, ed. A. L. Wayda and M. Y. Darensbourg, American Chemical Society, Washington, DC, 1987, ch. 4, pp. 79»98; (c) D. F. Shriver and M. A. Drezdzon, T he Manipulation of Air-Sensitive Compounds, Wiley-Interscience, New York, 2nd edn., 1986. 12 J. E. CosgriÜ, G. B. Deacon, B. M. Gatehouse, H. Hemling and H. Schumann, Aust. J. Chem., 1994, 47, 1223. 13 The apparent coupling constant is reduced from the value of 2 Hz for the other signals due to facile thallium relaxation as a consequence of chemical shift anisotropy. See: P. Ghosh, P. J. Desrosiers and G. Parkin, J. Am. Chem. Soc., 1998, 120, 10416. 14 G. M. Sheldrick, SHELXTL, An Integrated System for Solving, Re–ning and Displaying Crystal Structures from DiÜraction Data, University of Goé ttingen, Goé ttingen, Federal Republic of Germany, 1981. L etter 9/06292K New J. Chem., 1999, 23, 961»963 963

ISSN:1144-0546    

DOI:10.1039/a906292k

出版商:RSC    

年代:1999

数据来源: RSC

摘要:

L e t t e r Fluoride ligands exhibit marked departures from the hydrogen bond acceptor behavior of their heavier halogen congeners§ Lee Brammer,*a Eric A. Brutona and Paul Sherwoodb a Department of Chemistry, University of Missouri-St. L ouis, 8001 Natural Bridge Road, St. L ouis, MO 63121-4499, USA. E-mail : lee.brammer=umsl.edu b Computational Science and Engineering Department, CL RC Daresbury L aboratory, Daresbury, W arrington, UK WA4 4AD Received (in New Haven, USA) 5th August 1999 Much shorter hydrogen bonds with unexpected directional properties are formed by —uoride ligands in comparison to their heavier halogen congeners (X= Cl, Br, I).Ab initio calculations elucidate the underlying electronic origin of this geometric behavior. The viability of hydrogen bonding is clearly established for all donorñacceptor combinations, DñHÆ Æ ÆXñM (D= C, N, O; X= F, Cl, Br, I).The hydrogen bond acceptor capability of halogens has attracted attention in –elds as diverse as supramolecular chemistry,1h3 biochemistry,4 coordination and organometallic chemistry.5 Recent studies have examined both the ability1 and inability4 of speci–c halogens to form strong hydrogen bonds depending upon the coordination environment of the halogen.Interest in organometallic chemistry has centered upon the protonation and protonolysis of metal halides.5b,e Halide ions and halogen-containing ions have been shown to exhibit a templating eÜect via hydrogen bonding to form cavities2a,b or channels2c in supramolecular assemblies. Hydrogen bonding has been used strategically in the design of receptors for halide ions,2d but perhaps most pertinent to the present work are the possibilities of using metal halides in the design of receptors for molecules that can serve as proton donors.5a This idea is reinforced by recent work demonstrating the potential utility of D»H… … …Cl»M hydrogen bonds (D\N, O; M\transition metal) in the self-assembly of tape and sheet structures.3 Previously we have shown that metal-bound chlorine is a very good hydrogen bond acceptor in contrast to its organochlorine counterpart.1a Comparisons have also been made between the hydrogen bond acceptor capability of the halide ions.1b,c Here we focus speci–cally on intermolecular hydrogen bonds involving terminal, metal-bound halogens and show that the acceptor groups involving —uorine (F»M) exhibit substantially diÜerent behavior from that of their heavier halogen counterparts.Much shorter hydrogen bonds are formed and quite diÜerent directionality is exhibited by the F»M acceptors, viz. angles M»F… … …H?M»X… … …H (X\Cl, Br, I). This can be rationalized by consideration of the electronic structure of metal halides as it manifests itself in the electrostatic potential around the halogen.To compare the relative hydrogen bond acceptor capabilities of the halogens the Cambridge Structural Database6 has § Supplementary material available : model geometries used in ab initio calculations ; vs. 1[cos(180[(D»H… … …X)) plots ; vs. (RHX)3 (RHX)3 1[cos(180[H… … …X»M) plots ; calculated electrostatic potentials.For direct electronic access see http ://www.rsc.org/suppdata/nj/1999/ 965/, otherwise available from BLDSC (No. SUP 57619, 5 pp.) or the RSC Library. See Instructions for Authors, 1999, Issue 1 (http :// www.rsc.org/njc). been used to examine the geometries of D»H………X»M hydrogen bonds7 (D\C, N, O; X\F, Cl, Br, I). The hydrogen bond distances for all donor»acceptor pairs have been put on a common scale by use of the normalized distance function This allows a qualitative assess- RHX\d(H… … …X)/(rH]rX).7 ment of the relative strengths8 of the diÜerent hydrogen bond types as summarized in Table 1 (and SUP 57619).The ability to form viable hydrogen bonds is indicated for all hydrogen bond donor (D»H) and acceptor (M»X) combinations. The shortest mean arises for the combination of the most RHX polar bonds in the donor and acceptor components, i.e.O» H… … …F»M; the longest arises for C»H… … …I»M. Hydrogen RHX bonds involving metal —uorides are markedly shorter than those of their heavier halogen congeners, for all donor groups considered. The appropriateness of a hydrogen bonding description is con–rmed by the clear preference for D»H… … …X angles approaching 180° at shorter separations for all RHX donor»acceptor combinations.Fig. 1 illustrates these observations for the case of N»H… … …X»M hydrogen bonds. Spatially normalized distance vs. angle plots are used to avoid the biases inherent in simple distance vs. angle plots.9 The preference for linearity at short H… … …X separations is most pronounced for acceptors M»F[M»Cl[M»Br[M»I (Fig. 1) and donors O»H[N»H[C»H (SUP 57619 Fig. S1). The most remarkable aspect of the behavior of transition metal —uoride acceptors is in the directionality of approach of the donor groups to the —uoride. Considering N»H… … …X»M hydrogen bonds, for example, H… … …X»M angles in the range 90»130° account for 66.0% (Cl), 62.3% (Br), and 75.4% (I) of observations, but only 32.4% where X\F.10 Rather, for N» H… … …F»M 64.1% of the hydrogen bonds adopt larger H… … …F» M angles in the range 130»160°. These angular preferences are illustrated in the histograms of Fig. 2; more detailed spatially normalized distance vs. angle plots are also provided in SUP 57619 (Fig. S2). Qualitatively similar trends in H… … …X»M angles are seen for O»H… … …X»M hydrogen bonds, whereas C» H… … …X»M hydrogen bonds exhibit no discernible directional preference for approach of the C»H donor to the X»M acceptor. Table 1 Mean distances from H… … …X contacts7 with RHX (RHX)3O (ca. 1.150 RHX\1.048) Mean normalized distance, RHX (number of observations) C»H… … …X»M N»H… … …X»M O»H… … …X»M X\F 0.943 (374) 0.776 (73) 0.703 (37) X\Cl 0.975 (7943) 0.853 (1341) 0.799 (416) X\Br 0.982 (3269) 0.879 (205) 0.820 (30) X\I 0.997 (2429) 0.923 (83) 0.868 (8) New J.Chem., 1999, 23, 965»968 965 This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scienti–que 1999 (Fig. 1 Spatially normalized plots9 of hydrogen bond distances (represented as against angle at the hydrogen Mrepresented as (RHX)3) 1[cos(180[(N»H… … …X))N for hydrogen bonded N»H… … …X»M, X\F (a), Cl (b), Br (c), I (d).Fig. 2 Normalized distribution of H… … …X»M angles10 expressed as percentages of total number of observed N»H… … …X»M hydrogen bonds. The diÜerence between the angular behavior of the —uoride acceptors and those of the other halogens could in principle be electronic in origin or merely the eÜect of greater steric repulsions due to cis ligands being in closer proximity to the smaller —uoride ligand.As the electrostatic attraction between donors and acceptors is typically the largest contributor to hydrogen bond energies, the possible electronic origin of the geometric trend was investigated by calculating the electrostatic potentials of model compounds trans-PdX(Me)(PH3)2 using ab initio methods.11 These potentials are shown in Fig. 3. The angle subtended at the halogens by the points of potential energy minimum should indicate the preferred angle of approach of a hydrogen bond donor assuming the electrostatic contribution to the hydrogen bond energy is the dominant one.5c The angles are in excellent agreement with the database results. (Compare Table 2 with Fig. 3.) The depths of the potential minima (Table 2) also illustrate why the —uoride acceptors form much stronger hydrogen bonds than the other halogens. The calculations clearly support an electronic rather than steric rationale for the observed geometries. Examination of the molecular orbitals for the model systems indicates the source of angular discrimination at the halogen.Involvement of the halogen p-orbitals in M»X r-bonding will reduce the Table 2 Data for calculated electrostatic potential in the vicinity of the halogen for trans-PdX(Me) model systems (PH3)2 In molecular plane (cf. Fig. 3) In orthogonal plane through C»Pd»X (cf. Fig. S3) Potential Potential minimum/ minimum/ kcal mol~1 R(X… … …min)a Pd»X… … …min/° b kcal mol~1 R(X… … …min)a Pd»X… … …min/° X\F [81.3 0.77 162 [82.9 0.77 137 (140»180)c (120»155)c X\Cl [47.6 0.94 124 [51.7 0.94 110 X\Br [39.9 1.02 124 [44.3 1.03 109 X\I [32.3 1.09 122 [35.5 1.11 106 (i.e. distance of potential minimum from halogen nucleus as a fraction of halogen van der Waals radius).b The a R(X… … …min)\d(X… … …min)/rvdW(X) larger angles in the molecular plane arise from overlap of the positive potential of the ligands cis to the halogen with the negative potential PH3 generated by the halogen. c Potential well is —at-bottomed, and varies by less than 1 kcal mol~1 over this angular range. 966 New J. Chem., 1999, 23, 965»968Fig. 3 Calculated electrostatic potential in the metal coordination plane of the model compounds trans-PdX(Me) X\F, Cl, Br, I. (PH3)2 ; Electrostatic potential represented with color gradation and contoured at 10 kcal mol~1 intervals.Only negative and zero contours shown; atoms that lie below the coordination plane are not shown. p-orbital contribution to the lone pair trans to the metal,12 depleting the charge density in this direction relative to that of a spherical halide ion13 and favoring approaches towards the p-orbital lone-pairs orthogonal to the M»X bond.This eÜect is much more pronounced for X\Cl, Br, and I than for X\F, due to the smaller p-orbital contribution to M»F bonding. In summary, the overall order of hydrogen bond strength is D»H… … …F»MAD»H… … …Cl»MPD»H… … …Br»M[D»H… … …I»M, as implied by the H… … …X distance distributions and the magnitude of the halogen potential minima from the model complex calculations.The observations agree remarkably well with the trend in intramolecular N»H… … …X»Ir bond strengths determined experimentally using NMR methods by Crabtree, Eisenstein and coworkers for mer-[IrH2X(pyNH2)(PPh3)2] (viz., X\F, 5.2 ; X\Cl, 2.1 ; X\Br, 1.8 ; X\1, \1.3 kcal mol~1).14 The marked diÜerence in angular preference for the approach of hydrogen bond donors to the terminal halogens can be attributed to the greater contribution to M»X bonding of the halogen axial p-orbital exhibited by the heavier halogens.We are presently completing similar studies of D»H… … …X»C and D»H… … …X~ hydrogen bonds, as well as comparing simple and bifurcated hydrogen bonds, in an eÜort to provide a comprehensive analysis of the hydrogen bond acceptor behavior of halogens.Acknowledgements and E.A.B. are grateful to Karl Kedrovsky for pro- L.B. gramming assistance, and the University of Missouri-St. Louis for –nancial support. L.B. and P.S. thank NATO for support of their collaboration (grant CRG-920164). Notes and references 1 (a) G. Aulloç n, D. Bellamy, L. Brammer, E. A. Bruton and A. G. Orpen, Chem. Commun., 1998, 653; (b) M.Mascal, J. Chem. Soc., Perkin T rans. 2, 1997, 1999; (c) T. Steiner, Acta Crystallogr., Sect. New J. Chem., 1999, 23, 965»968 967B, 1998, 54, 456; (d) C. B. Aakeroé y, T. A. Evans, K. R. Seddon and I. Paç linkoç , New J. Chem., 1999, 23, 145. 2 (a) R. Vilar, D. M. P. Mingos, A. J. P. White and D. J. Williams, Angew. Chem., Int. Ed., 1998, 37, 1258; (b) J. S.Fleming, K. L. V. Mann, C.-A. Carraz, E. Psillakis, J. C. JeÜrey, J. A. McCleverty and M. D. Ward, Angew. Chem., Int. Ed., 1998, 37, 1279; (c) J. C. Mareque Rivas and L. Brammer, New J. Chem., 1998, 22, 1315; (d) K. Kavallieratos, S. R. de Gala, D. J. Austin and R. H. Crabtree, J. Am. Chem. Soc., 1997, 119, 2325. 3 (a) J. C. Mareque Rivas and L. Brammer, Inorg. Chem., 1998, 37, 4756; (b) P.J. Davies, N. Veldman, D. M. Grove, A. L. Spek, B. T. G. Lutz and G. van Koten, Angew. Chem., Int. Ed. Engl., 1996, 35, 1959; (c) G. R. Lewis and A. G. Orpen, Chem. Commun., 1998, 1873. 4 J. D. Dunitz and R. Taylor, Chem. Eur. J., 1997, 3, 89. 5 (a) T. G. Richmond, Coord. Chem. Rev., 1990, 105, 221; (b) For a discussion of the protonation of metal halides, see R. Kuhlman, Coord. Chem.Rev., 1997, 167, 205; (c) D. V. Yandulov, K. G. Caulton, N. V. Belkova, E. S. Shubina, L. M. Epstein, D. V. Khoroshum, D. G. Musaev and K. Morokuma, J. Am. Chem. Soc., 1998, 120, 12553; (d) V. J. Murphy, T. Hascall, J. Y. Chen and G. Parkin, J. Am. Chem. Soc., 1996, 118, 7428; (e) D.-H. Lee, H. J. Kwon, B. P. Patel, L. M. Liable-Sands, A. L. Rheingold and R. H. Crabtree, Organometallics, 1999, 18, 1615. 6 F. H. Allen and O. Kennard, Chem. Des. Automat. News, 1993, 8, 1 and 31. 7 (a) All N»H and O»H distances were normalized to neutron lengths, and only those structures containing intermolecular contacts with D»H… … …X angles P110° and were (RHX)3O1.15 included in subsequent analyses. Normalized distances, RHX\ are calculated based upon van der Waals radii, d(H… … …X)/(rH]rX), and Only rH\1.20, rF\1.47, rCl\1.75, rBr\1.85 rI\1.96 ”.7b terminally bound halogens were considered.In the case of duplicate structure determinations the structure of highest overall quality was retained following manual inspection. Data for bifurcated hydrogen bonds were removed after the initial search using locally written programs. (b) A. J. Bondi, J.Chem. Phys., 1964, 68, 441. 8 (a) While inferring a correlation between hydrogen bond length and strength is inappropriate for individual crystal structures, large populations of geometric data allow a qualitative assessment of bond strengths to be made. It is recognized that the relative values of mean normalized distances will be aÜected by the proportion of neutral and charge-assisted (i.e., interionic) hydrogen bonds in each sample population.8b However, it is anticipated that these eÜects will be similar for each donor»acceptor pair studied.(b) D. Braga and F. Grepioni, New. J. Chem., 1998, 22, 1159. 9 (a) Spatially normalized plots using the transformed coordinate system vs. 1[cos T (T \180[(D»H… … …X)), as described (RHX)3 by Allen and Taylor,9b removes the inherent statistical biases of conventional distance vs.angle plots by ensuring that equal volumes of space are mapped onto equal areas of the twodimensional plot. Some useful points of reference are : (RHX)3\1.0 corresponds to (RHX\1.0) d(H… … …X)\&(rH]rX) ; 1[cos T \0.0 corresponds to N»H… … …X\180°, 1[cos T \0.5 to N»H… … …X\120°, 1[cos T \0.75 to N»H… … …X\104.5°, and 1[cos T \1.0 to N»H… … …X\90° (Fig. 1 and S1) ; the same correspondences arise between the H………X»M angle and the function 1[cos A (A\180[(H… … …X»M)) (see SUP 57619, Fig. S2). (b) J. P. M. Lommerse, A. J. Stone, R. Taylor and F. H. Allen, J. Am. Chem. Soc., 1996, 118, 3108. 10 Histograms of the H… … …X»M angles were plotted in 10° intervals and corrected for the geometric error in the frequency of observations that naturally arises in sampling such angle data from crystal structures, which depends upon the sine of the angle ; see J.Kroon and J. A. Kanters, Nature (L ondon), 1974, 248, 667. 11 (a) Electrostatic potentials were calculated at the Hartree»Fock level using the GAMESS-UK11b package, employing a CEP (compact eÜective potential) basis set for Pd11c,e and pVTZ (polarized valence triple-zeta)11d,e basis sets on all other atoms.Details of the model geometries used are provided in SUP 57619; (b) GAMESS-UK is a package of ab initio programs written by M. F. Guest, J. H. van Lenthe, J. Kendrick, K. SchoÜel and P. Sherwood, with contributions from R. D. Amos, R. J. Buenker, H. J. J. van Dam, M. Dupuis, N. C. Handy, I. H.Hillier, P. J. Knowles, V. Bonacic-Koutecky, W. von Niessen, R. J. Harrison, A. P. Rendell, V. R. Saunders, A. J. Stone and A. H. de Vries. The package is derived from the original GAMESS code due to M. Dupuis, D. Spangler and J. Wendoloski, NRCC Software Catalog, 1980, vol. 1, Program No. QG01 (GAMESS), Lawrence Berkely Laboratory, University of California, Berkeley; (c) W. J. Stevens, M.Krauss, H. Basch and P. G. Jasien, Can. J. Chem., 1992, 70, 612; (d) A. J. Sadlej, T heor. Chim. Acta, 1992, 81, 45 and refs. therein ; (e) Basis sets were obtained from the Extensible Computational Chemistry Environment Basis Set Database, Version 1.0, as developed and distributed by the Molecular Science Computing Facility, Environmental and Molecular Sciences Laboratory which is part of the Paci–c Northwest Laboratory, P.O. Box 999, Richland, Washington 99352, USA, and funded by the U.S. Department of Energy. The Paci–c Northwest Laboratory is a multi-program laboratory operated by Battelle Memorial Institute for the U.S. Department of Energy under contract DE-AC06-76RLO 1830. Contact David Feller, Karen Schuchardt, or Don Jones for further information. 12 This explanation is related to a previously proposed orbital model for N»H… … …Cl»M hydrogen bonds that considered the relative basicities of sp-hydrid orbital and p-orbital lone pairs on the chloride ligands, though trends with respect to the series of halogens were not considered. See: G. P. A. Yap, A. L. Rheingold, P. Das and R. H. Crabtree, Inorg. Chem., 1995, 34, 3474. 13 The depopulation (electrons) of the halogen p-orbital trans to the metal relative to the average of the two orthogonal halogen porbitals is calculated to be: F (0.093 e), Cl (0.226 e), Br (0.270 e), I(0.305 e) using the Lowdin population analysis.11b 14 E. Peris, J. C. Lee, J. R. Rambo, O. Eisenstein and R. H. Crabtree, J. Am. Soc., 1995, 117, 3485. L etter 9/06512A 968 New J. Chem., 1999, 23, 965»968

ISSN:1144-0546    

DOI:10.1039/a906512a

出版商:RSC    

年代:1999

数据来源: RSC

摘要:

L e t t e r Secondary bonding as a potential design element for crystal engineering Jonathan Starbuck, Nicholas C. Norman and A. Guy Orpen* School of Chemistry, University of Bristol, Bristol, UK BS8 1T S Received (in Montpellier, France) 27th May 1999, Accepted 4th August 1999 Crystallographic data imply that intermolecular hypervalent interactions, ììsecondary bondsœœ, involving heavy p-block elements such as Bi(III), may form robust supramolecular synthons with many features in common with organic hydrogen bonds, including the ability to form polymeric networks in the solid state.The synthesis of crystalline solids in which molecular units are arranged in predictable ways»crystal engineering»is a major and a realistic goal for modern chemistry.1 One means to this end is the identi–cation and application of reliable supramolecular synthons,2 which control molecular aggregation and lead to crystal structures with at least partly controlled structures, containing sheets, ribbons and other desired motifs.The synthons identi–ed and exploited to date1h3 have been for the most part taken from organic and biological chemistry and use the directionality of hydrogen bonds to aÜord the desired control of aggregation.There is a clear need for alternatives to this strategy in order to aÜord chemical and functional diversity in these solids. Among the areas most pro–tably explored in this context is that of dative coordinate bonded networks, usually of the late d-block elements (Groups 9»12, see, for example, the work of Robson, Ciani, Iwamoto, Schroé der, Zaworotko and others).4 In this paper we show that secondary bonding,5 that is, intermolecular hypervalent interactions, has many structural features in common with the hydrogen bond and may aÜord a reliable new class of supramolecular synthon.In 1972 Alcock de–ned secondary bonding in a seminal paper5a as involving intermolecular hypervalent interactions of length less than the sum of van der Waals radii between a heavy p-block element (E, say) and an electron pair donor (typically halogen, O, N or S).Such secondary bonding is ubiquitous in the solid state chemistry of the compounds of Pb(II), Bi(III), Te(IV), I(III) and other related species in which the electron con–guration at E is formally nd10(n]1)s2(n]1)px, x\0, 2, 4.These hypervalent interactions are electronically and structurally (see below) distinct from systems in which the Lewis acid centre has no lone pair [e.g., in aluminium(III) or antimony(V) complexes]. To date there has been no systematic attempt to exploit secondary bonding in the design of crystal structures (but see reference 6 for examples in iodine chemistry). The description of the hypervalent interactions involved in secondary bonding has attracted controversy and interest for much of the past 40 years.The current consensus is that the best simple bonding model treats such interactions as involv- Scheme 1 Formation of from and Cl~. [BiCl2Ph2]~ BiClPh2 ing delocalised 3c-4e r bonds.7 An archetypal example of such an interaction is the anion8 Mwhich is formally [BiCl2Ph2]~ and practically obtained by addition of Cl~ to the Bi(III) [5d106s2] centre in see Scheme 1N.Here the second BiClPh2 , Bi»Cl bond is formed trans to the Bi»Cl bond in pyramidal leading to the familiar disphenoidal (equatorially BiClPh2 , vacant trigonal bipyramidal) symmetric geometry of C2l In this case the Lewis base (Cl~) donates an [BiCl2Ph2]~. electron pair into a Bi»Cl r* orbital, thereby forming the 3c-4e interaction, and a linear or near-linear Cl»Bi»Cl moiety.Many similar species are formed in Bi(III) chemistry [and related systems based on Pb(II), Sb(III), Te(IV), I(III), etc.]. Most frequently the r* orbital is associated with an E»X bond involving an electronegative substituent X [X\halide, alkoxide or other oxygen ligand, etc., see Scheme 2(a) and 2(b)] since this lowers the energy of the r* orbital, ensures it is mainly bismuth-centred and that the bismuth is more positively charged.In many cases of secondary bonding the incoming Lewis base is the electronegative atom X in an adjacent molecule and the interaction leads to the formation of polymeric or oligomeric structures. In a number of respects the intermolecular structural chemistry of such Bi(III) species (and the related systems from the p-block noted above) might therefore be expected to resemble that of the more familiar hydrogen bond.9 Taking the and anions as our archetypes, the [BiCl2Ph2]~ [HCl2]~ replacement of H` by (for example) should yield [BiPh2]` systems with similar structural properties.The similarities in the distribution of angles at hydrogen in O»H… … …O hydrogen bonds and the Cl»Bi… … …Cl angles in trans-Bi(III) fragments Cl2 [see Scheme 2(c) ; Fig. 1]§ is striking. Despite the diÜerences in the details of the 3c-4e bonding7 in these systems the preference for linearity at the central atom is clear. The bond length variability in O»H… … …O bonds is familiar from many neutron diÜraction studies and is illustrated in Fig. 2(a). The two O»H distances are coupled and follow a well-de–ned hyperbolic curve.10,11 The corresponding plot for the trans-Bi(III)Cl2 system [Fig. 2(b)] shows some resemblance in having many points around a similarly curved trajectory. It is noteworthy that similar plots for other Bi(III) and Sb(III) dihalides have essentially identical forms, while those for Sb(V) dihalides have points only very near the symmetrical geometry.7b,8 (d1\d2) The Cl»Bi… … …Cl plot shows much more scatter perpendicular to the ideal hyperbolic path and somewhat more concentration of points near the symmetrical geometry than (d1\d2) Scheme 2 (a) moiety (X\uninegative electronegative substit- BiXZ2 uent, Z\X or R, R\uninegative less electronegative substituent.(b) Bi»X r* orbital. (c) fragment. trans-BiCl2 New J. Chem., 1999, 23, 969»972 969 This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scienti–que 1999 (Fig. 1 Histograms of angles (deg) in (a) O»H… … …O and (b) Cl»Bi… … …Cl 3c-4e systems in the CSD.î the O»H… … …O plot. The former diÜerence may be ascribed to the presence of (and variation in) substituents on the central atom (Bi) cis to the 3c-4e system, which are of course not present in the O»H… … …O hydrogen bonds.We have commented on the mechanisms by which cis substituents interact with the geometry of the moiety.12 Analysis of the BiCl2 second aspect will be addressed in a full paper, but surely relates to the diÜerence in the 3c-4e bonding in the two systems.The third way in which the O»H… … …O and Cl»Bi… … …Cl plots diÜer is metric ; the Bi»Cl bonds are much longer (ca. 2.7 for the symmetrical species, cf. ca. 1.2 for O»H in sym- ” ” metrical O»H»O bonds). The pyramidal shape of and molecules and their BiX3 SbX3 derivatives and the model given above for hypervalent bonding in these systems allow qualitative predictions of the likely secondary bonding motifs in the solid state structures of such species. A selection of relatively common bismuth or antimony species for which such predictions may be made is given in Table 1.In these species the formula is given by (Scheme 3) where E\Sb or Bi; X, as above, is a EXxR3~xLy uninegative ligand with an electronegative contact atom (e.g., Table 1 Secondary bonding capabilities of bismuth(III) and antimony(III) moieties Moietya nb xb yb Ns c Nw c ER2X 3 1 0 1 2 ERX2 3 2 0 2 1 EX3L2 5 3 2 1 0 EX3L 4 3 1 2 0 ER2XL 4 1 1 0 2 a Moiety formula: E\Sb, Bi; X\halide, alkoxide, etc.; L\2e donor ligand, including cases where L\X~; R\alkyl, aryl, etc. b n\coordination number of E; x\number of substitutents X; y\number of ligands L.and are the number of strong and c Ns Nw weak secondary bonding sites at E, respectively. halide), R is a uninegative ligand with a less electronegative contact atom (e.g., aryl), and L is a 2e donor ligand (e.g., pyridine, THF, Cl~). Therefore, the coordination number at E is given by n\3]y; the number of strong Lewis acid sites at E is given by while there are weak acceptor sites Ns\x[y, Nw trans to the R substituents (assuming the maximum coordination number for E to be 6 as is usually but not universally the case in these species).Recently,3 Allen and co-workers reported a crystallographic database analysis that tested the efficacy of important cyclic bimolecular hydrogen bond motifs by comparing the number of occurrences of the component fragments in the CSD with the number of occasions on which the desired H-bonding pattern is observed.We have carried out a similar analysisî for the systems listed in Table 1 in order to establish the efficacy and reliability of their secondary bonding capabilities. Table 2 lists the percentage occurrence of secondary bonds having intermolecular contacts less than the sum of the van der Waals radii in the crystal structures of these species. In addition Table 2 lists the networks present in structures they form.In this analysis the graph set nomenclature of Etter and Bernstein is used13 to de–ne the networks formed. In order to exploit this established method of analysis we have denoted the X groups ììacceptorsœœ and the E atoms ììdonorsœœ by analogy with jargon of the hydrogen bonding literature in which the Lewis acid (H) is the ììdonorœœ of the hydrogen bond and the Lewis base (Cl~ or similar) the ììacceptorœœ.The frequency with which secondary bonds are formed by these various moieties is given in Table 2. It is striking that the strong Lewis acid sites (trans to the electronegative substit- Scheme 3 Molecular species for which secondary bonding has been studied [E\Sb(III) or Bi(III)] (see Table 1). Fig. 2 Scattergrams of bond lengths in (a) O»H… … …O and (b) Cl»Bi… … …Cl 3c-4e systems in the CSD.î (”) 970 New J. Chem., 1999, 23, 969»972Table 2 Secondary bonding interactions in bismuth(III) and antimony(III) structuresî Moiety n(CSD)a n(frag)b Ns Fs c Nw Fw c Network familyd ER2X 18 24 1 71 2 8 44% C(2) ; 11% R3 3 (6) ERX2 20 21 2 81 1 19 50% C(2) EX3L2 67 75 1 85 0 » 39% C(2) ; 39% R2 2 (4) EX3L 60 90 2 85 0 » 40% C(2) ; 18% R2 2 (4) ER2XL 27 28 0 » 2 18 » a Number of unique crystal structures located from the CSD.b Number of unique moieties located from the CSD. c Percentage of possible secondary bonds actually formed. d Family to which graph set13 of secondary bond network belongs. uents, X) are consistently involved in secondary interactions at a frequency of between 70 and 90%.This contrasts with frequencies of \20% for the weaker secondary bonding sites. The marked diÜerence is therefore in accord with the postulate that the electronegativity of the trans substituent is a major determinant in the likelihood and strength of secondary bond formation.This parallels the familiar9 in—uence of the electronegativity of the element to which a hydrogen atom is bonded on the strength of the hydrogen bonds it forms. The high frequency of secondary bond formation in the ììstrongœœ Lewis acid sites implies that such sites have considerable potential as components of robust supramolecular synthons; that is, they can be relied upon to form secondary bonds. The predominant mode of association in these species is that in which chains of molecules are formed (see Table 2), albeit by a variety of speci–c networks in which considerable complexity may be present.Rings are also formed in a substantial number of cases. For example, in the structures of [BiClMMo(CO)3(gand (i.e., structures C5H5)N2]14 [BiClMFe(CO)2(g-C5H5)N2]15 of type polymeric [C(2)] and ring (n\3)] ER2X) [Rn n (2n) structures are observed in which secondary bonds with near linear Cl»Bi… … …Cl moieties form the links between the molecular units (see Scheme 4).It is noteworthy that in an analogous class of organic crystals of molecules with exactly one hydrogen bond donor and one hydrogen bond acceptor (the monoalcohols) Brock and Duncan have shown16 that unusual space groups are often encountered and that in many cases (n\4, 6) ring and C(2) chain networks are formed.The Rn n (2n) same classes of network occur in the category of Bi(III) species with exactly one Lewis acid site and one available secondary bond acceptor (the set). ER2X Previously we have shown that the networks formed by these species in the solid state may be disrupted by dissolution in Lewis basic solvents such as THF.12 In this respect again there is an analogy with the behaviour of H-bonded solids in which the intermolecular H-bonds are frequently lost on dissolving in strongly hydrogen bonding solvents such as water.In summary hydrogen bonds and secondary bonds have the following similarities (a) The preference for linearity at the central atom in 3c-4e interactions (b) The softness of the bond lengths in the 3c-4e interactions (c) The formation of solid state networks of these interactions, describable by graph set nomenclature.Scheme 4 Typical (a) ring and (b) C(2) chain networks formed R3 3 (6) by species. ER2X (d) The disruption of these networks in solution. (e) The percentages of intermolecular interactions formed are high for the stronger Lewis acid sites.In other respects the two classes of intermolecular bond do diÜer, as listed below. (a) The metric. E… … …X bonds are typically about twice as long as comparable X… … …H contacts (b) The variability of E… … …X bond lengths. In particular, there is less strong coupling between the two lengths. In hydrogen bonding, bifurcated bonds in which the hydrogen interacts with two acceptors, are relatively common if often difficult to predict or control.Similar interactions are possible in secondary bonding systems [e.g., for cases in which the E centre is [6 coordinate, as in BiCl3(THF)217]. However, more signi–cant ììbifurcationœœ or branching of the intermolecular bond network is possible in this class of supramolecular synthon because of the ability of the Lewis acidic E centre to be simultaneously involved in two or even three 3c-4e interactions in approximately perpendicular directions.This and the other properties noted above may oÜer new opportunities for the preparation of novel solids based on secondary bonding with characteristics that complement those currently available.It is noteworthy that the work of Mitzi and others18 has already demonstrated the potentially important properties that such solids might possess. Acknowledgements support of the EPSRC is gratefully acknowledged. Financial The help of the staÜ of the Cambridge Crystallographic Data Centre including Drs F. H. Allen and G. P. Shields is much appreciated by the authors.Notes and references § Data for relevant structures (neutron diÜraction studies only for O» H… … …O hydrogen bonds) were retrieved from the CSD19 and bond lengths and angles calculated. The data sets for Figs. 1 and 2 were permuted10 to re—ect the symmetry of the potential energy hypersurface for the reaction path in question. î Data for relevant structures were retrieved from the CSD19 and bond lengths and angles calculated.Intermolecular contacts involving E shorter than the sum of the van der Waals radii20 were recorded as being secondary bonds and the networks and frequencies computed accordingly. 1 (a) G. R. Desiraju, Chem. Commun., 1997, 1475; (b) J. C. MacDonald and G. M. Whitesides, Acc. Chem. Res., 1995, 28, 37; (c) C. B. Aakeroé y, Acta Crystallogr., Sect.B, 1997, 53, 569; (d) I. G. Dance, in T he Crystal as a Supramolecular Entity, ed. G. R. Desiraju, Wiley, Chichester, 1996, ch. 5. 2 (a) G. R. Desiraju, Angew Chem., Int. Ed. Engl., 1995, 34, 2311; (b) A. Nangia and G. R. Desiraju, Acta Crystallogr., Sect. A, 1998, 54, 934. 3 F. H. Allen, W. D. S. Motherwell, P. R. Raithby, G. P. Shields and R. Taylor, New J.Chem., 1999, 23, 25. 4 (a) S. R. Batten and R. Robson, Angew. Chem., Int. Ed. Engl., 1998, 37, 1461; (b) M. J. Zaworotko, Chem. Soc. Rev., 1994, 23, 283; (c) L. Carlucci, G. Ciani, P. Macchi, D. M. Proserpio and S. Rizzato, Chem. Eur. J., 1999, 5, 237; (d) L. Tei, V. Lippolis, A. J. Blake, P. A. Cooke and M. Schroé der, Chem. Commun., 1998, 2633; (e) T. Iwamoto, S. Nishikiori, T.Kitazawa and H. Yuge, J. Chem. Soc., Dalton T rans., 1997, 4127. New J. Chem., 1999, 23, 969»972 9715 (a) N. W. Alcock, Adv. Inorg. Chem. Radiochem., 1972, 15, 1; (b) N. W. Alcock, Bonding and Structure, Ellis Horwood, Chichester, 1990. 6 (a) F. H. Allen, B. S. Goud, V. J. Hoy, J. A. K. Howard and G. R. Desiraju, J. Chem. Soc., Chem. Commun., 1994, 2729; (b) V. R. Thalladi, B.S. Goud, V. J. Hoy, F. H. Allen, J. A. K. Howard and G. R. Desiraju, Chem. Commun., 1996, 401; (c) J.-P. M. Lommerse, A. J. Stone, R. Taylor and F. H. Allen, J. Am. Chem. Soc., 1996, 118, 3108; (d) A. J. Blake, F. A. Devillanova, R. O. Gould, W. S. Li, V. Lippolis, S. Parsons, C. Radek and M. Schroé der, Chem. Soc. Rev., 1998, 27, 195; (e) M. Boucher, D. Macikenas, T. Ren and J.D. Protasiewicz, J. Am. Chem. Soc., 1997, 119, 9366. 7 (a) G. A. Landrum, N. Goldberg and R. HoÜmann, J. Chem. Soc., Dalton T rans., 1997, 3605; (b) G. A. Landrum and R. HoÜmann, Angew. Chem., Int. Ed. Engl., 1998, 37, 1887; (c) G. A. Landrum, N. Goldberg, R. HoÜmann and R. M. Minyaev, New J. Chem., 1998, 22, 883; (d) P. Pyykkoé , Chem. Rev., 1997, 97, 597. 8 J. Starbuck, S. C.James, N. C. Norman and A. G. Orpen, unpublished results. 9 (a) G. A. JeÜery, Introduction to Hydrogen Bonding, Wiley, Chichester, 1997; (b) G. A. JeÜery and W. Saenger, Hydrogen Bonding in Biology and Chemistry, Springer Verlag, Berlin, 1993. 10 Structure Correlation, eds. H.-B. Bué rgi and J. Dunitz, VCH, Weinheim, 1994. 11 I. Olofsson and P. G. Joé nsson, in T he Hydrogen Bond, eds.P. Schuster, G. Zundel and C. Sandorfy, North-Holland, Amsterdam, 1976, vol. II, ch. 8. 12 W. Clegg, N. A. Compton, R. J. Errington, G. A. Fisher, D. C. R. Hockless, N. C. Norman, A. G. Orpen and S. E. Stratford, J. Chem. Soc., Dalton T rans., 1992, 3515. 13 (a) M. C. Etter, Acc. Chem. Res., 1990, 23, 120; (b) M. C. Etter, J. C. MacDonald and J. Bernstein, Acta Crystallogr., Sect. B, 1990, 46, 256; (c) M. C. Etter, J. Phys. Chem., 1991, 95, 4601; (d) J. Bernstein, R. E. Davis, L. Shimoni and N.-L. Chang, Angew. Chem., Int. Ed. Engl., 1995, 34, 1555. 14 W. Clegg, N. A. Compton, R. J. Errington, N. C. Norman, A. J. Tucker and M. J. Winter, J. Chem. Soc., Dalton T rans., 1988, 2941. 15 W. Clegg, N. A. Compton, R. J. Errington and N. C. Norman, J. Chem. Soc., Dalton T rans., 1988, 1671. 16 C. P. Brock and L. L. Duncan, Chem. Mater., 1994, 6, 1307. 17 C. J. Carmalt, W. Clegg, M. R. J. Elsegood, R. J. Errington, J. Havelock, P. Lightfoot, N. C. Norman and A. J. Scott, Inorg. Chem., 1996, 35, 3709. 18 (a) D. B. Mitzi, Prog. Inorg. Chem., 1999, 48, 1; (b) L. Sobczyk, R. Jakubas and J. Zaleski, Pol. J. Chem., 1997, 71, 265. 19 (a) F. H. Allen, J. E. Davies, J. J. Galloy, O. Johnson, O. Kennard, C. F. Macrae, E. M. Mitchell, G. F. Mitchell, J. M. Smith and D. G. Watson, J. Chem. Inf. Comput. Sci., 1987, 31, 187; (b) F. H. Allen and O. Kennard, Chem. Des. Automation News, 1993, 8, 1 & 31. 20 A. J. Bondi, J. Chem. Phys., 1964, 68, 441. L etter 9/06352H 972 New J. Chem., 1999, 23, 969»972

ISSN:1144-0546    

DOI:10.1039/a906352h

出版商:RSC    

年代:1999

数据来源: RSC

摘要:

L e t t e r Generation of a highly enantioenriched a-phenylthio-substituted Grignard-reagent Peter G. Nell Fachbereich Chemie, Marburg, Hans-Meerwein- D-35032 Marburg, Philipps-Universitaé t Straêe, Germany. E-mail : pnell=ps1515.chemie.uni-marburg.de Received (in Freiburg, Germany) 26th July 1999, Accepted 3rd August 1999 The a-phenylthio-substituted Grignard reagent 8 was generated in > 95% ee by a sulfoxide/magnesium exchange reaction starting from the enantio- and diasteromerically pure aphenylthio sulfoxide 6a.The a-thio-substituted Grignard reagent has been trapped with benzaldehyde to give the syn-bphenylthio- substituted alcohol 9b in very high diastereoselectivity and enantiomeric purity. a-Heterosubstituted alkylmetal reagents (1) are potentially useful chiral d1-synthons in organic synthesis.1 On reaction with aldehydes two diastereomeric alcohols 2a and 2b are formed (Scheme 1).Simple diastereoselectivity is generally low using organolithium compounds.2 There are indications that Grignard reagents of the type of 1 lead to higher diastereoselectivity, 3,4 possibly due to a change in the addition mechanism. 5 Con–gurational stability of the organometallic reagents 1 is another issue.For instance, in the case of the arylthiosubstituted derivatives of 1 the organolithium compounds are con–gurationally labile, racemizing rapidly even at [110 °C.6h9 Substantially higher con–gurational stability can be expected for the corresponding Grignard reagents.3,10 Thus, for both reasons Grignard reagents such as 8 would be of interest, provided they can be generated in enantiomerically pure form.Grignard reagents can be generated by a sulfoxide/ magnesium exchange, a reaction studied extensively by Satoh and Takano.11 We used this reaction previously12 to generate enantiomerically pure a-chloroalkyl Grignard reagents. This led us to investigate the sulfoxide/magnesium exchange reaction of the a-arylthio alkyl sulfoxides 6.The starting sulfoxide 6 was prepared (Scheme 2) by reacting a-phenylthio methyllithium (4) with a menthyl aryl sul- –nate following precedent from the Gennari group.13 We chose the p-chlorobenzene sul–nate 314 instead of the more common p-toluene sul–nates in order to facilitate puri–cation of the starting sulfoxides 6 by crystallization. Treatment of the menthyl sul–nate 3 with 4 at [78 °C thus generated the (])- (S)-a-phenylthio sulfoxide 5 as white crystals (mp 71»72 °C) in 70% yield.The sulfoxide 5 was benzylated in 80% yield to furnish a 1 : 1 mixture of the diastereomeric sulfoxides 6a and 6b. The mixture was separated by column chromatography on silica gel to give 6a Mmp 103 °C, (c\1.09, [a]D20\]17.9 acetone)N and 6b Mmp 75 °C, (c\1.56, [a]D20\[164.1 Scheme 1 acetone)N.The relative and absolute con–guration of 6a was assigned as (S,S) by X-ray structure analysis,§ cf. Fig. 1. When the stereochemically homogeneous sulfoxide 6a was treated with ethylmagnesium bromide at [78 °C in THF (Scheme 3), rapid sulfoxide/magnesium exchange ensued to give the a-phenylthio alkyl Grignard reagent 8.The latter was trapped after 10 min by addition of benzaldehyde, providing 58% of the b-hydroxy thioether 9 along with 99% of the sulfoxide 7 of 99% ee. The b-hydroxy thioether was obtained as a single diastereomer (syn : anti [98 : \2). Scheme 2 Fig. 1 Molecular structure of 6a. New J. Chem., 1999, 23, 973»975 973 This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scienti–que 1999 (Scheme 3 Its relative and absolute con–guration could be assigned as 9b by the following transformation.Methylation of the thioether with Meerweinœs salt15 generated a sulfonium ion, which on treatment with base (Scheme 4) furnished the epoxide 10 (67%, cis : trans [98 : \2). The enantiomeric purity of the epoxide was determined by 1H NMR spectroscopy in the presence of [hfc\3-(hepta—uoropropyl hydroxy- Eu(hfc)3 methylene)-D-camphorate] to be [95%.Since the absolute con–guration of the epoxide 1016 and the sulfoxide 717 is known, the stereochemical course of the sulfoxide/magnesium exchange can be shown to proceed with inversion at the sulfoxide sulfur atom and with retention of con–guration in the formation of the carbon magnesium bond.This corresponds to our recent results found in the sulfoxide/magnesium exchange on a-chloroalkyl sulfoxides.12 The present study shows that a-arylthio alkyl Grignard reagents of type 8 can be generated in enantiomerically pure form by a sulfoxide/magnesium exchange reaction, that these reagents are con–gurationally stable at [78 °C for a sufficient period of time to be added to an aldehyde, and that this addition proceeds with very high ([98%) simple diastereoselectivity to provide the syn-diastereomer of 9.Experimental All NMR experiments were performed on a Bruker AC300 spectrometer operating at 300 MHz (1H) or 75.5 MHz (13C). Scheme 4 All manipulations were carried out under an argon atmosphere. All solvents were dried and deoxygenated prior to use.The ([)-(1R,3R,4S)-menthyl (S)-p-chlorobenzenesul–nate 3 was prepared according to the procedure of Klunder and Sharpless.14 Syntheses (+ )-(S)-p-Chlorophenyl 1-phenylthiomethyl sulfoxide, 5. An n-BuLi solution (18.3 mL, 26.6 mmol) in hexane was added to a solution of methyl phenyl sul–de (3.00 g, 24.2 mmol) and DABCO (2.71 g, 24.2 mmol) in 35 mL of dry THF at 0 °C. After stirring for 15 min and subsequently cooling to [78 °C a solution of ([)-(1R,3R,4S)-menthyl (S)-p-chlorobenzene sul- –nate 3 (3.81 g, 12.1 mmol) in 40 ml of dry THF was added dropwise.Stirring overnight and the usual workup aÜorded 5 in 70% yield (2.40 g), which was puri–ed by crystallization from acetone (c\2.17, acetone), mp 71» M[a]D20\]103.0 72 °CN. 1H NMR d 7.64»7.21 (m, 9H, CH of Ar), 4.22 (CDCl3) : (d, 2J\13.6 Hz, 1H, CHH), 4.07 (d, 2J\13.5 Hz, 1H, CHH); 13C NMR d 141.07, 137.91, 133.25, 130.92 (2C), (CDCl3) : 129.29 (2C), 129.27 (2C), 127.77, 126.31 (2C), 60.77 Calc.(CH2). for C, 55.21 ; H, 3.92. Found: C, 55.13 ; H, C13H11ClOS2 : 3.85%. p-Chlorophenyl (2-phenyl-1-phenylthio-ethyl) sulfoxide, 6. A lithium diisopropylamide solution (1.5 M, 3.07 mL, 4.60 mmol) in THF was added to a solution of 5 (1.15 g, 4.06 mmol) in 5 mL of dry THF at [78 °C.After 20 min, addition of benzyl bromide (0.90 g, 5.29 mmol), further stirring overnight, and the usual workup yielded 6 as a mixture of diastereomers (1 : 1) in 80% yield (1.22 g). Separation by column chromatography on silica gel (pentane : tert-butyl methyl ether 15 : 1) yielded 0.60 g of diastereomer (S,S)-6 M[a]D20\]17.9 (c\1.09, acetone), mp 103 °CN. 1H NMR d 7.62» (CDCl3) : 7.33 (m, 4H, CH of Ar), 7.32»6.89 (m, 10H, CH of Ar), 4.01 (dd, 3J\3.3 Hz, 3J\10.8 Hz, 1H, CHS), 3.53 (dd, 2J\14.3 Hz, 3J\3.2 Hz, 1H, CHH), 2.94 (dd, 2J\14.3 Hz, 3J\10.9 Hz, 1H, CHH); 13C NMR d 140.27, 137.88, 136.41, (CDCl3) : 132.73, 132.46 (2C), 129.68 (2C), 129.07 (2C), 128.95 (2C), 128.56 (2C), 127.96 (2C), 127.39, 127.10, 75.95, 34.20.Diastereomer (S,R)-6 (0.59 g) (c\1.56, acetone), M[a]D20\[164.1 mp 75 °CN. 1H NMR d 7.77»7.48 (m, 4H, CH of Ar), (CDCl3) : 7.30»7.10 (m, 10H, CH of Ar), 4.20 (dd, 3J\3.5 Hz, 3J\11.2 Hz, 1H, CHS), 3.58 (dd, 2J\14.0 Hz, 3J\3.4 Hz, 1H, CHH), 2.18 (dd, 2J\14.0 Hz, 3J\11.3 Hz, 1H, CHH); 13C NMR d 137.92, 136.73, 132.14 (2C), 129.62 (2C), 129.21 (CDCl3) : (2C), 128.80 (2C), 128.63 (2C), 128.14 (2C), 127.70 (2C), 127.08, 73.57, 32.63, 1C not detected.Calc. for C, C20H17ClOS2: 64.41 ; H, 4.59. Found: C, 64.41 ; H, 4.64%. Typical procedure for sulfoxide/magnesium exchange. The lower compartment of a two-compartment reaction vessel18 was charged with 109 lL of a solution of ethylmagnesium bromide (2.05 M, 0.22 mmol, diethyl ether) in 2 mL of dry THF.The upper compartment was charged with (S,S)-6 (64 mg, 0.17 mmol) and 2 mL of dry THF. After cooling of both compartments to [78 °C the content of the upper chamber was added to the lower one and the reaction mixture was stirred for 10 min. Addition of precooled ([78 °C) benzaldehyde (54 mg, 0.51 mmol), stirring for 30 min, and the usual workup aÜorded the b-phenylthio-substituted alcohol 9b in 58% yield (31 mg) as a colorless oil (syn : anti [98 : \2). 1H NMR d 7.31»7.04 (m, 15H, CH of Ar), 4.58 (dd, (CDCl3) : 3J\3.4 Hz, 3J\6.8 Hz, 1H, CHOH), 3.39 (ddd, 3J\5.3 Hz, 3J\6.7 Hz, 3J\9.5 Hz, 1H, CHS), 3.12 (d, 3J\3.4 Hz, 1H, OH), 2.86 (dd, 2J\14.3 Hz, 3J\5.3 Hz, 1H, CHH), 2.61 (dd, 2J\14.3 Hz, 3J\9.5 Hz, 1H, CHH); 13C NMR d (CDCl3) : 142.05, 139.00, 133.10 (2C), 129.25 (2C), 128.82 (2C), 128.36 974 New J.Chem., 1999, 23, 973»975(2C), 128.31 (2C), 127.64 (2C), 127.49 (2C), 126.83, 126.41, 74.95, 61.33, 29.69. HRMS Calc. for 320.1235. C21H20OS: Found: 320.1232^0.3 amu. Acknowledgements work was supported by the Deutsche Forschungsgemein- This schaft (SFB 260) and the Fonds der Chemischen Industrie.I would like to thank Dr. K. Harms for solving the crystal structure and I gratefully acknowledge Professor R. W. Hoffmann for generous support. Notes and references § Crystal structure analysis : (S,S)-6: M\372.91, C20H17ClOS2 : monoclinic, space group a\925.1 (1), b\904.0 (1), c\1140.1 P21, (1) pm, a\90, b\111.656 (5), c\90°, U\886.1(1)]10~30 m3, Z\2, k(Cu-Ka)\4.129 mm~1, T \213(2) K, 2767 re—ections measured, re—ections with I[2r(I)\2598, –nal Rint\0.0719, R(F)\0.0486, wR(F2)\0.1336.Re–nement of an inversion twin parameter19 [x\0.00(2), where x\0 for the correct absolute structure and]1 for the inverted structure] con–rmed the absolute structure of 6a. CCDC reference number 440/138. See http ://www.rsc.org/ suppdata/nj/1999/973/ for crystallographic –les in .cif format. 1 Houben-W eyl, Methoden der Organischen Chemie, 4th edn., Carbanionen, ed. M. Hanack, Thieme, Stuttgart, 1993, E19d. 2 D. Seebach and M. A. Syfrig, Angew. Chem., 1984, 96, 235; Angew. Chem., Int. Ed. Engl., 1984, 23, 248. 3 R. W. HoÜmann, M. Julius, F. Chemla, T. Ruhland and G. Frenzen, T etrahedron, 1994, 50, 6049. 4 D. Seebach, J. Hansen, P. Seiler and J.M. Gromek, J. Organomet. Chem., 1985, 285, 1. 5 K. S. Rein, Z.-H. Chen, P. T. Perumal, L. Echegoyen and R. E. Gawley, T etrahedron L ett., 1991, 32, 1941. 6 (a) H. J. Reich and M. D. Bowe, J. Am. Chem. Soc., 1990, 112, 8994; (b) T. Shinozuka, Y. Kikori, M. Asaoka and H. Takei, J. Chem. Soc., Perkin T rans. 1, 1996, 119; (c) P. G. McDougal, B. D. Condon, M. D. LaÜosse, Jr., A.M. Lauro and D. VanDerveer, T etrahedron L ett., 1988, 29, 2547; (d) K. Brickmann and R. Brué ckner, Chem. Ber., 1993, 126, 1227. 7 (a) R. W. HoÜmann, R. K. Dress, T. Ruhland and A. Wenzel, Chem. Ber., 1995, 128, 861; (b) R. Hirsch and R. W. HoÜmann, Chem. Ber., 1992, 125, 975. 8 R. W. HoÜmann and W. Klute, Chem. Eur. J., 1996, 2, 694. 9 B. Kaiser and D. Hoppe, Angew. Chem., 1995, 107, 344; Angew.Chem., Int. Ed. Engl., 1995, 34, 323. 10 W. Klute, M. Krué ger and R. W. HoÜmann, Chem. Ber., 1996, 129, 633. 11 T. Satoh and K. Takano, T etrahedron, 1996, 52, 2349. 12 R. W. HoÜmann and P. G. Nell, Angew. Chem., 1999, 111, 354; Angew. Chem., Int. Ed., 1999, 38, 338. 13 L. Colombo, C. Gennari and E. Narisano, T etrahedron L ett., 1978, 40, 3861. 14 J. M. Klunder and K. B. Sharpless, J. Org. Chem., 1987, 52, 2598. 15 J. R. Shanklin, C. R. Johnson, J. Ollinger and R. M. Coates, J. Am. Chem. Soc., 1973, 95, 3429. 16 V. Schulze and R. W. HoÜmann, Chem. Eur. J., 1999, 5, 337. 17 H. Takeuchi, H. Minato, M. Kobayashi, M. Yoshida, H. Matsuyama and N. Kamigata, Sulfur Silicon Relat. Elem., 1990, 47, 165. 18 H. C. Stiasny and R. W. HoÜmann, Chem. Eur. J., 1995, 1, 619. 19 (a) H. D. Flack, Acta Crystallogr., Sect. A, 1983, 39, 876; (b) G. Bernadinelli and H. D. Flack, Acta Crystallogr., Sect. A, 1985, 41, 500; (c) K. Harms, personal communication. L etter 9/06179G New J. Chem., 1999, 23, 973»975 975

ISSN:1144-0546    

DOI:10.1039/a906179g

出版商:RSC    

年代:1999

数据来源: RSC

摘要:

L e t t e r A non-covalent assembly for electron transfer based on a calixareneñporphyrin conjugate : tweezers for a quinone Takashi Arimura,a* Seiji Ide,a Hideki Sugihara,a Shigeo Murataa and Jonathan L. Sesslerb a COE L aboratory, National Institute of Materials and Chemical Research, T sukuba 305-8565, Japan b Department of Chemistry and Biochemistry, University of T exas at Austin, Austin, T exas 78712, USA Received 22nd July 1999, Accepted 9th August 1999 The synthesis and characterization of a new supramolecular assembly, a calix [4] arene substituted Zn(II) metalloporphyrin 4 and benzoquinone 6, wherein photoinduced electron transfer through non-covalent interactions may be probed, is reported.Considerable debate within the electron transfer modeling community continues to be devoted to the question of how speci–c protein pathways might, or might not, be in—uencing long-range biological electron transfer events.1 One way in which this critical issue is being addressed is through the synthesis and study of simple, non-covalently constructed model systems.2,3 In general, these have consisted of a photo donor and one or more electron acceptors held together by neutral or charged hydrogen-bonding interactions.On the other hand, few, if any, systems are known wherein other kinds of non-covalent (e.g., van der Waals) contacts serve to de–ne the key donor-to-acceptor supramolecular interactions. In this paper, therefore, we report a new calixarene-based donor» acceptor system, ensemble I, in which hydrogen bonding interactions (between the two phenolic OH groups on the calixarene and the carbonyl of the quinone) serve as tweezers to complex non-covalently a quinone acceptor.Synthesisof5-formyl-17-nitro-25,27-dimethoxycalix[4]arene- 26,28-diol 1 was communicated previously.4 Its elaboration into the calixarene-substituted porphyrin derivative 4 is shown in Scheme 1. Brie—y, the target compound 4 was prepared in 30% yield from the cross-condensation of calixarene 1, dipyrromethane, 2, and benzaldehyde 3. 5-Phenyl-15-(4- hydroxyphenyl)-2,8,12,18-tetraethyl-3,7,13,17-tetramethylporphyrin 5 was also prepared in the usual manner.5 Evidence for the formation of ensemble I in came CDCl3 from 1H NMR spectroscopic studies. In the absence of benzoquinone 6, the calix[4]arene-substituted zinc porphyrin 4 displays two types of phenolic hydroxyl groups in its 1H NMR spectrum, OHa at 9.03 ppm and OHb at 7.97 ppm.Upon addition of 6, the Ha, Hb and meso proton signals of 4 are shifted up–eld slightly, while the other porphyrinic signals remain unperturbed. Interestingly, no spectral shifts of 1H NMR or UV/vis that could be attributed to p stacking between the porphyrin moiety and quinone subunit, even at the highest available concentration (60 mmol dm~3).Moreover, 1H NMR shift changes of the control system 5 were not observed in the presence of 6. Analysis of the up–eld shift for the Ha and Hb protons as a function of increasing quinone concentration by standard curve –tting methods6 provided support for a 1 : 1 binding model and yielded an association constant of Ka 70^10 dm3 mol~1.The binding constant is lower than Ka that of usual hydrogen-bonded complexes, perhaps as a result of strong intramolecular hydrogen-bonding interactions between the two hydroxy groups present in the calixarene. Prior to testing ensemble I, analyses of the control system 5, a Zn(II) mono-hydroxy diphenylporphyrin, and 6 were carried out.In this instance, steady state —uorescence quenching studies aÜorded a linear Stern»Volmer plot (Fig. 1), a –nding that was explained by —uorescence of the porphyrin 5 being Scheme 1 New J. Chem., 1999, 23, 977»979 977 This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scienti–que 1999 (Fig. 1 Stern»Volmer plots for the —uorescence quenching of 4 (L) and 5 with benzoquinone 6 in [4]\[5]\5.0]10~6 (Ö) CH2Cl2 .mol dm~3. The sample was excited at 400 nm with emission being integrated from 540 to 700 nm. quenched only by a diÜusional, as opposed to a static, mechanism. 7 The diÜusional quenching constant (kq\6.31]109 dm3 mol~1 s~1) was calculated from the Stern»Volmer constant. 8 In the case of ensemble I, the species produced by mixing 4 with 6, the corresponding plot was found to be curved upward (Fig. 1). Such non-linear curvature suggests that the porphyrin excited state of 4 is being quenched by both static and dynamic processes. In other words, we interpret the —uorescence of 4 in ensemble I as being partially quenched by an intra-ensemble electron-transfer process involving a complexed quinone.Addition of methanol disrupts the non-covalent interactions in the ensemble I and restores the porphyrin —uorescence to the control value. To provide further support for the above interpretation, time resolved —uorescence studies were performed. First, the control system 5 (5.0]10~6 mol dm~3) was examined in degassed In the absence of 6, the decay of the singlet CH2Cl2 .excited state of this species (excitation wavelength 400 nm with the emission monitored at 580 nm), was found to be monoexponential with a lifetime of 1.3 ns. When 6 was added (0»60 mmol dm~3) to a solution of 5 (5.0]10~6 mol dm~3), the —uorescence decay pro–le remained monoexponential in character. It did, however, display a decreased dynamic lifetime as would be expected for a concentration-dependent bimolecular quenching process.9 In the case of the calixarene-substituted porphyrin 4, studied under conditions identical to those above, a single exponential decay with a lifetime of 1.3 ns was observed in the absence of 6.Adding increasing quantities of 6 (0»60 mmol dm~3), however, resulted in a —uorescence decay pro–le which could be best analyzed in terms of two components, a long lived component with a variable lifetime and a short lived component with a constant lifetime of 30 ps.The fractional amplitude of the shorter lived component increased as the benzoquinone concentration was increased from 0 to 60 mmol dm~3. In spite of this increase in fractional amplitude, the lifetime of this shorter lived component remained essentially unchanged. The shorter lived component disappeared upon adding methanol.Under these conditions, the —uorescence decay pro–le of 4 could be analyzed in terms of a single exponential even in the presence of 6. By contrast, both the fractional amplitude and the lifetime of the longer lived component was found to decrease as the concentration of benzoquinone was increased from 0 to 60 mmol dm~3.This decrease from 1.3 ns to 550 ps, was similar to that seen in the control system consisting of 5 and 6. Based on previous work with other H-bonded donor» acceptor assemblies,2 the shorter lived component is attributed to a quenching process involving unidirectional singlet»singlet electron-transfer from calixarene»Zn(II) porphyrin 4 to 6 bound within the supramolecular assembly I.The longer lived component, on the other hand, is readily ascribed to ììnormalœœ deactivation of the excited state of uncomplexed (i.e. quinone-free) 4. The present paper outlines that two phenolic hydroxy groups of the calix[4]arene serve as tweezers to capture the benzoquinone by two-point hydrogen-bonding –xation. While hydrogen bonding interactions between 4 and 6 serve to bring the donor and acceptor into close contact, the observed electron transfer process is thought to result from a through-bond (including H-bonds) pathway.Experimental Materials All chemicals were reagent grade and used without further puri–cation. Previously published procedures10 were used for the synthesis of 3,3@-dimethyl-4,4@-diethyldipyrromethane 2. Apparatus 1H NMR spectra were measured on a Varian XL-300 spectrometer.FAB mass spectra were recorded on JEOL-DX303. Steady-state —uorescence spectra were measured on a Hitachi F-4500 spectrophotometer. Fluorescence lifetimes were measured by time-correlated single photon counting using a mode-locked Ti : sapphire laser for excitation. The full width at half maximum height of the instrument response function was B60 ps.Synthesis Zn(II) complex of 5-phenyl-15-(5-(25,27-dihydroxy-26,28- dimethoxy-17-nitrocalix [4] arene))-2,8,12,18-tetraethyl-3,7,13, 17-tetramethylporphyrin (4). Calix[4]arene 1 (390 mg, 0.75 mmol), dipyrromethane 2 (520 mg, 2.3 mmol) and benzaldehyde 3 (160 mg, 1.5 mmol) were dissolved in (80»110 mL). After the addition of trichloro- CH2Cl2»CH3CN acetic acid (110 mg, 0.68 mmol) in 10 mL of the CH3CN, mixture was stirred for 19 h under a nitrogen atmosphere at which time chloranil (1.1 g, 4.6 mmol) in 60 mL of CH2Cl2 was added and the reaction allowed to stir for an additional 2.5 h.This was then washed with aqueous sodium bicarbonate followed by The organic layer was then dried over H2O. After the solvent was removed, the residue was dis- Na2SO4 .solved in 200 mL of and 2.0 mL of saturated zinc CH2Cl2 acetate in methanol was added. After stirring for 30 min, the solvent was evaporated in vacuo. Puri–cation by column chromatography on silica gel (eluting with gave 4 (250 CH2Cl2) mg, 30%) as a red powder; mp[300 °C; 1H NMR (300 MHz, d 10.21, 10.17 (each 1H, s, meso-H), 9.03, 7.97 CDCl3) (each 1H, s, OH), 8.13 (2H, s, ArH), 8.09 (2H, d, J\7. 3 Hz, ArH), 7.79 (2H, s, ArH), 7.80»7.74 (3H, m, ArH), 6.98»6.94 (4H, m, ArH), 6.81 (2H, d, J\7.3 Hz, ArH), 4.56, 4.34 (each 2H, d, J\13.1 Hz, 4.06 (6H, s, 3.64»3.49 ArCH2Ar), OCH3), (4H, m, 4.05»3.95 (8H, m, 2.63 (3H, s, ArCH2Ar), CH2CH3), 2.46 (6H, s, 1.96 (3H, s, 1.75 (12H, m, CH3), CH3), CH3), MS (FAB): m/z 1111 (M`). (Found: C, 72.58 ; H, CH2CH3) ; 5.77 ; N, 6.13.requires C, 72.34 ; H, C68H63N5O6Zn ÆH2O 5.80 ; N, 6.20%). References 1 (a) C. C. Moser, J. M. Keske, K. Warncke, R. S. Farid and P. L. Dutton, Nature, 1992, 355, 796. (b) D. N. Beratan, J. N. Onuchic, J. R. Winkler and H. B. Gray, Science, 1992, 258, 1740. (c) H. Pelletier and J. Kraut, Science, 1992, 258, 1748. 2 (a) A. Harriman, D. Magda and J. L. Sessler, J.Chem. Soc., Chem Commun., 1991, 345. (b) C. M. Drain, R. Fischer, E. G. Nolten and J.-M. Lehn, J. Chem. Soc., Chem. Commun., 1993, 243. (c) J. A. Roberts, J. P. Kirby and D. G. Nocera, J. Am. Chem. Soc., 1995, 117, 8051. (d) P. J. F. de Rege, S. A. Williams and M. J. Therien, 978 New J. Chem., 1999, 23, 977»979Science, 1995, 269, 1409. (e) C. A. Hunter and R. J. Shannon, Chem.Commun., 1996, 1361. ( f ) J. L. Sessler, B. Wang, S. L. Springs and C. T. Brown, Electron and Energy T ransfer Reactions in Non-Covalently L inked, Supramolecular Model Systems in Comprehensive Superamolecular Chemistry, Y. Murakami ed., Pergamon Press Ltd, UK, 1996, vol. 4, Ch. 9, pp. 311»335 and references therein. 3 (a) C. A. Hunter, J. M. Sanders, G. S. Beddard and S. Evans, J. Chem. Soc., Chem. Commun., 1989, 1765. (b) C. A. Hunter and R. K. Hyde, Angew. Chem., Int. Ed. Engl., 1996, 35, 1936. (c) R. Milbradt and J. Weiss, T etrahedron L ett., 1995, 36, 2999. 4 T. Arimura, S. Ide, H. Sugihara, S. Murata and M. Sato, J. Jpn. Oil Chem. Soc., 1999, 48, 775. 5 A. Kajiwara and M. Kamachi, Chem. Express, 1989, 4, 105. 6 (a) M. Hynes, J. Chem. Soc., Dalton T rans., 1993, 311. (b) B. J. Whitlock and H. W. Whitlock Jr., J. Am. Chem. Soc., 1990, 112, 3910. 7 V. O. Stern and M. Volmer, Phys. Z., 1919, 20, 183. 8 J. R. Lakowicz, Principle of Fluorescence Spectroscopy, Plenum Press, New York, 1986, pp. 260»273. 9 N. J. Turro, Modern Molecular Photochemistry, University Science Books, Mill Valley, 1991, pp. 246»252. 10 R. Young and C. K. Chang, J. Am. Chem. Soc., 1985, 107, 898. L etter 9/06474E New J. Chem., 1999, 23, 977»979 979

ISSN:1144-0546    

DOI:10.1039/a906474e

出版商:RSC    

年代:1999

数据来源: RSC

摘要:

New design of vesicle-forming potential anti-cancer agent§ Philippe Coutrot,*a Patrick Oliger,b Claude Grison,a Steç phane Joliez,a Marc Heç brantb and Christian Tondre*b a L aboratoire de Chimie Organique II and b L aboratoire de Chimie Physique Organique et UMR CNRS 7565, Institut de Chimie Henri Colloïé dale, Nanceç ien Moleç culaire, Universiteç I, B.P. 239, 54506 V andoeuvre-les-Nancy cedex, France Poincareç -Nancy Received (in Montpellier, France) 14th June 1999, Accepted 3 August 1999 A novel amphiphile derivative of the anti-cancer agent PALA (N-(phosphonoacetyl)-L-aspartate), has been synthesized : dioctadecyl N-(phosphonoacetyl)-L-aspartate (PALAmod).Its aptitude to form bilayers or vesicles has been investigated. The melting temperature of the alkyl chains of PALAmod is comparable with those of the Tm bilayers of distearoyl phosphatidylcholine (DSPC) or distearoyl phosphatidylglycerol (DSPG) which have fatty chains of the same length.Encapsulation of two diÜerent hydrophilic probes, carboxy—uorescein (CF) and glucose, was examined, demonstrating the likely existence of vesicular structures either in lecithin»PALAmod mixtures or in pure PALAmod, respectively.The size of the particles has been measured from quasielastic light scattering experiments. The particle diameter at –xed pH (6.9), goes from 74 nm at [PALAmod] 3.6 mM to 52 nm at 14.8 mM. Nouvelle conception dœun agent anticanceç reux potentiel sous forme de ve ç sicule. Le diester carboxylique steç arique de lœacide N-(phosphonoacetyl)-L-aspartique (PALAmod) a eç teç syntheç tiseç en vue dœexaminer les capaciteç s de ce tensioactif a` sœautoorganiser en bicouches ou en veç sicules.La tempeç rature de fusion des alkyles du Tm chaï� nes PALAmod est comparable a` celles des bicouches de la distearoyl phosphatidyl choline (DSPC) ou du distearoyl phosphatidyl glycerol (DSPG), qui posse` dent des grasses de me� me longueur. La formation de structures chaï� nes veç siculaires constitueç es soit dœun meç lange leç cithine»PALAmod, soit de PALAmod pur a eç teç eç tablie a` partir de mesures dœencapsulation de sondes hydrophiles, carboxy—uoresceç ine (CF) ou glucose.La taille des particules a eç teç eç valueç e par des expeç riences de diÜusion quasieç lastique de la lumie` re. Le diame` tre des particules a` pH –xeç (6.9), varie de 74 nm, a` une concentration en PALAmod de 3.6 mM, a` 52 nm, a` une concentration de 14.8 mM.Introduction Research in the –eld of anti-cancer chemotherapeutic agents includes diÜerent conceptual approaches.1 One of them is based on the inhibition of nucleotide synthesis. Nucleotides are indeed essential metabolites involved in the building of DNA and RNA.Decreasing their production is considered as one of the possible ways to prevent the growth of tumorous cells. The biosynthesis pathway of pyrimidinic nucleotides (uracil, cytosine, thymine) is controlled by an allosteric enzyme, aspartate transcarbamylase (ATCase).2 Inhibition of this enzyme may be obtained by using substrate analogues showing competitive binding to the catalytic site with respect to the natural substrates (carbamyl phosphate and L-aspartate in the present case).A bisubstrate analogue, with a chemical structure close to that of the assumed transition-state of the ATCase-catalyzed reaction, has been synthesized by Collins and Stark.3 This molecule, N-(phosphonacetyl)-L-aspartate (PALA), has proved to be a good competitive inhibitor of carbamyl phosphate, as demonstrated by in vitro studies4,5 and then, has in—uenced numerous other preparations.6 Unfortunately, the anti-cancer activity of free PALA, which was established from in vitro investigations, was not as large as expected when in vivo tests were conducted.5,6e This was at least partly attributed to the proposed mechanism for cellular uptake and especially to the § Part of this work was presented at the 13th Annual Meeting of the GTRV, Paris (December 1998).role of endocytosis and lysosomal pH.7 A major research objective is aimed at improving the transport of PALA towards its desired target, ATCase. At the lysosomal pH, which is of the order of 5, the PALA molecule is expected to be essentially in the form of a trianion, when considering the of the phosphonic group (2.43 and pKaœs 7.85 in the case of ethylphosphonic acid) and of the carboxylic functions (2.09 and 3.86 for aspartic acid), respectively.7 Whether the global electric charge carried by the molecule should be decreased or increased to facilitate its penetration in the cytoplasm is a controversial question.7,8 Attempts have recently been made to answer this question by modifying the global charge of the molecule.This was achieved, for instance, by substituting the hydrogens in the a-position of the phosphonate group by —uorine atoms, in order to decrease the second of the phosphonic acid and thus to ensure a full pKa ionization of the molecule at the lysosomal pH. Unfortunately this new molecule has shown, in vitro, a decreased affinity for ATCase.9 So, getting a better insight into the part played by the electric charges in the translocation mechanism (passive diÜusion or endocytosis) appears to be a difficult task.On the other hand, endocytosis cannot be considered a very efficient process. It has been demonstrated that, even with a large extracellular concentration of PALA, only a very small quantity is transferred across the cell walls.7 Any improvement of the efficiency of the uptake mechanism should obviously be bene–cial to the development of the therapeutic applications.In this perspective, very encouraging results have been obtained by Sharma et al.10 who have demonstrated that New J. Chem., 1999, 23, 981»987 981 This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scienti–que 1999 (liposome-encapsulated PALA shows enhanced antitumor activity.They have examined the inhibitory eÜects on tumor cell growth of diÜerent PALA-containing liposome formulations. Depending on the cell line considered, the optimal formulation indicated a 22- to 570-fold greater sensitivity compared to free PALA. Additional work in this direction has been pursued by Kim and Heath11 who have encapsulated PALA in antibody-directed liposomes, able to recognize speci –c antigens present at the surface of tumor cells.A spectacular efficiency of PALA was obtained in that case, giving further encouragement for developing the use of that kind of vector. However, a major drawback of the vesicular encapsulation method is that only a very small amount of the initially synthesized PALA (usually less than 5%) is encapsulated, the largest part of the drug being lost.This has prompted us to imagine a new formulation in which the loss of PALA would be limited and its local concentration would be high. The idea was to confer to the PALA molecule itself an amphiphilic character, so that the modi–ed molecule (PALAmod) can mimic the behaviour of phospholipids in making vesicular structures.These structures could either result from a self-organization of PALAmod or from the association of this molecule with vesicle-forming amphiphiles like phospholipids. In the latter case the local concentration of PALAmod could be tuned by adjusting the composition of the vesicles.Of course the inhibition of ATCase could only be obtained after the release of free PALA, which could be regenerated in situ from the esterolysis of PALAmod by lipases present in the cytoplasm.12 We are not yet at this point, which will have to be examined in a forthcoming step. At the present stage of this work, we had two objectives : (i) to synthesize a modi–ed PALA molecule presenting a strong analogy with natural phospholipids (see Scheme 1), and (ii) to demonstrate that this new molecule has the expecsicle-forming capacity.Experimental Materials All the chemicals used were commercially available guaranteed reagents unless otherwise stated. 1-Octadecanol, egg yolk lecithins (L-a-phosphatidylcholine type XVI), cholesterol and HEPES buÜer were obtained from Sigma.Dibenzyl-L-aspartate was purchased from Bachem, n-BuLi and bromotrimethylsilane were purchased from Aldrich. 5(6)-Carboxy—uorescein (CF) was from Kodak, Sephadex G50 (medium) from Pharmacia Biotech, Triton X-100 from Fluka. The —uorescence polarization probe, 1,6- diphenyl-1,3,5-hexatriene (DPH) and D-glucose were purchased from Fluka and Sigma, respectively. Diethyl methylphosphonate was prepared from triethylphosphite and iodomethane,13 and benzotriazol-1-yloxytris( dimethylamino)phosphonium hexa—uorophosphate (BOP) was prepared according to the previously reported method.14 Melting points were determined with an Electrothermal digital apparatus and are uncorrected.Optical rotations were measured with a Bellingam-Stanley Ltd. ADP 220 polarimeter (10 cm, 5 ml, at 25 °C, c in g per 100 ml).IR spectra CHCl3 were taken on a Nicolet 210 infrared spectrometer. 1H, 13C and 31P NMR spectra were recorded on a Bruker AC 250 spectrometer. Deuteriochloroform (99.8% atom enriched, Eurisotop) was used as NMR solvent. The NMR chemical shifts are reported in d (ppm) based on internal SiMe4 (dH\0) or the solvent signal or external 1% (CDCl3 , dC\77.0) in as references.FAB mass spectra were H3PO4 D2O (dP\0) measured on a Micromass Manchester Autospec Fitted Cesium gun spectrometer. All reactions, except for the catalytic hydrogenation were carried out in an inert atmosphere. Tetrahydrofuran (THF) was distilled from sodium» benzophenone, diethyl ether and dichloromethane were distilled from Column chromatography was carried out P2O5 .on Merck Silicagel 60 (particle size 0.040»0.063 mm). Syntheses Dibenzyl N-(diethylphosphonoacetyl)-L-aspartate 1.15 To a stirred solution of the p-tosylate salt of H-Asp(OBn)-OBn (10 g, 20 mmol) in dichloromethane (100 ml) was added triethylamine (2 g, 20 mmol), followed by the successive addition of diethyl phosphonoacetic acid16 (3.9 g, 20 mmol) in dichloromethane (100 ml), BOP (8.83 g, 20 mmol) in dichloromethane (50 ml) and triethylamine (2 g, 20 mmol).After stirring for 1.5 h at room temperature the reaction mixture was diluted with dichloromethane (100 ml). The organic layer was washed sequentially with 2 N (3]100 ml), 5% H2SO4 (3]100 ml), and saturated NaCl (100 ml) solutions. NaHCO3 The dichloromethane layer was dried –ltered and (MgSO4), concentrated.The crude oil was puri–ed by rapid –ltration on silica gel with pure ethyl acetate as eluent to give the product 1 (8.55 g, 82%) as a yellow oil. 1H NMR (250 MHz, d CDCl3) : 1.30 (t, J\7.0 Hz, 3H), 1.31 (t, J\7.0 Hz, 3H), 2.90 (dd, J\4.70 Hz, J\17.1 Hz, 1H), 2.87 (d, J\20.7 Hz, 2H), 3.08 (dd, J\4.67 Hz, J\17.1 Hz, 1H), 4.08»4.18 (m, 4H), 4.85» 5.00 (m, 1H), 5.08 (s, 2H), 5.13 (s, 2H), 7.29»7.35 (m, 10H), 7.43 (d, J\8.0 Hz, 1H). 13C NMR (62.53 MHz, d 16.28 CDCl3) : (d, J\7.1 Hz), 35.31 (d, J\142.0 Hz), 36.27, 49.00, 62.71 (d, J\5.6 Hz), 62.80 (d, J\5.6 Hz), 66.82, 67.53, 128.27, 128.40, 128.56, 164.07, 170.03, 170.36. 31P NMR (101.25 MHz, d 19.23. IR (neat) : 1025, 1250, 1675, 1740 cm~1. CDCl3) : N-(Diethylphosphonoacetyl)-L-aspartic acid 2.To a solution of 1 (2.5 g, 5 mmol) in AcOEt (20 ml) was added 10% Pd»C (0.5 g). The solution was stirred for 20 h under 30 bars of hydrogen at room temperature. The mixture was –ltered through Celite and the Celite was washed with AcOEt. The –ltrate was concentrated and the resulting N- (diethylphosphonoacetyl)-L-aspartic acid 2 was used directly in the next step.The product 2 (1.52 g) was obtained as a Scheme 1 982 New J. Chem., 1999, 23, 981»987Scheme 2 brown oil in 98% yield. 1H NMR (250 MHz, d 1.32 CDCl3) : (t, J\7.0 Hz, 6H), 2.80»3.20 (m, 4H), 4.09»4.22 (m, 4H), 4.80» 5.00 (m, 1H), 7.88 (d, J\8.0 Hz, 1H). 13C NMR (62.53 MHz, d 16.13 (d, J\6.0 Hz), 34.99 (d, J\131.4 Hz), 35.83, CDCl3) : 48.82, 63.49 (d, J\5.5 Hz), 63.57 (d, J\5.5 Hz), 164.66, 173.38, 174.17. 31P NMR (101.25 MHz, d 20.48. IR CDCl3) : (neat) : 1025, 1200, 1660, 1730, 3100»3600 cm~1. Dioctadecyl N-(diethylphosphonoacetyl)-L-aspartate 3. 1- Octadecanol (stearic alcohol) (3.5 g, 13 mmol), BOP (5.7 g, 13 mmol), and triethylamine (658 mg, 6.5 mmol) in dichloromethane were stirred for 1 h at room temperature followed by the slow addition (4 h) of N-(diethylphosphonoacetyl)-L-aspartic acid 2 (1.0 g, 3.2 mmol) diluted in dichloromethane (75 ml).After stirring for 15 h the reaction mixture was diluted with dichloromethane (75 ml). The organic layer was washed sequentially with 2 N (3]50 ml), 5% H2SO4 NaHCO3 (3]50 ml), and saturated NaCl (100 ml) solutions. The dichloromethane layer was dried –ltered and con- (MgSO4), centrated. The crude oil was puri–ed by column chromatography eluting with hexane»ethyl acetate (100 : 0»10 : 90).Dioctadecyl N-(diethylphosphonoacetyl)-L-aspartate 3 (1.93 g, 74%) was obtained as a white powder 0.75, hexane»ethyl (Rf acetate 10 : 90). 1H NMR (250 MHz, d 0.88 (t, CDCl3) : J\6.5 Hz, 6H), 1.20»1.80 (m, 64H), 1.32 (t, J\7.3 Hz, 3H), 1.33 (t, J\7.3 Hz, 3H), 2.83 (dd, J\4.6 Hz, J\17.1 Hz, 1H), 2.90 (d, J\20.5 Hz, 2H), 2.99 (dd, J\5.2 Hz, J\17.1 Hz, 1H), 4.07 (t, J\7.0 Hz, 4H), 4.08»4.23 (m, 4H), 4.89»5.16 (m, 1H), 7.38 (d, J\8.0 Hz, 1H). 13C NMR (62.53 MHz, d 14.07, 16.25 (d, J\6.0 Hz), 22.65, 25.74, 25.82, CDCl3) : 28.41, 28.46, 29.24, 29.32, 29.51, 29.66, 31.89, 35.18 (d, J\131.0 Hz), 36.08, 48.95, 62.77 (d, J\5.5 Hz), 62.85 (d, J\5.5 Hz), 65.29, 66.04, 164.14, 170.38, 170.77. 31P NMR (101.25 MHz, d 19.32. IR (neat) : 1025, 1240, 1660, CDCl3) : 1735, 2850, 2920, 2955 cm~1. Dioctadecyl N-(phosphonoacetyl)-L-aspartate 4 (PALAmod). Dioctadecyl N-(diethylphosphonoacetyl)-L-aspartate 3 (2.0 g, 2.5 mmol) and bromotrimethylsilane (5 ml, 38.6 mmol) in 1,2- dichloroethane (10 ml) were stirred for 3 h at room temperature, then the solvent was evaporated under vacuum.The resulting brown powder was washed with wet acetone, then dried in vacuo. Compound 4 (1.85 g, 97%) was obtained as a white solid. 1H NMR (250 MHz, d 0.88 (t, J\6.0 CDCl3) : Hz, 6H), 1.10»1.80 (m, 64H), 2.90»3.30 (m, 4H), 4.00»4.30 (m, 4H), 4.80»5.00 (m, 1H). 13C NMR (62.53 MHz, d CDCl3) : 14.08, 22.67, 25.82, 25.89, 28.41, 28.49, 29.36, 29.74, 31.91, 35.99, 36.12 (d, J\140.0 Hz), 49.32, 65.49, 66.23, 167.07, 170.49, 171.00. 31P NMR (101.25 MHz, d 18.44.MS CDCl3) : (negative FAB): m/z 758.5 (M`[1). IR (neat) : 1200, 1655, 1735, 2850, 2920, 2950 cm~1. Mp\74»76 °C. [a]D\]7.5° (c\1). Calc. for C, 66.37 ; H, 10.87 ; N, 1.84 ; P, C42H82NO8P: 4.08. Found: C, 65.7 ; H, 10.96 ; N, 1.86 ; P, 3.88%. Techniques The vesicle dispersions were prepared following a procedure derived from that described by Cleij et al.17 The surfactants (either pure PALAmod, or a lecithin»PALAmod mixture, eventually including cholesterol, depending on the case) were dissolved in methylene chloride or in a solution of DPH in methylene chloride for the —uorescence polarization experiments.The solvent was evaporated and the resulting solid was dried for 4 h under vacuum.HEPES buÜer (10 mM, pH 7.4 unless otherwise indicated) and NaOH (usually 1 equivalent per PALAmod molecule) were then added to the –lm which was left over night under agitation to allow its hydration. The obtained dispersion was sonicated for 5 min at 40 °C (Branson soni–er 200 W, tip 13 mm, 40»45% power output). The vesicle dispersion was –ltered at room temperature through 0.45 lm Millipore membranes (HATF type, i.e.surfactant-free) in order to remove the titanium particles coming from the immersion probe. For the encapsulation experiments either CF (50 mM) or glucose (0.5 M) were introduced into the dispersion before sonication. The size of the particles was measured at 25 °C using a home-assembled quasielastic light scattering (QELS) apparatus coupled with a Malvern autocorrelator.When necessary the dispersions were diluted with the HEPES buÜer, so as to adjust the scattering intensity. Filtration through 0.45 lm Millipore membranes was systematically performed just before the scattering experiments. We checked that dilution, after sonication, does not aÜect signi–cantly the value or the hydrodynamic radius of the particles. For the encapsulation experiments, steric exclusion chromatography (SEC) on Sephadex G50 was used to separate the vesicle-encapsulated probes from those which remained free.Some of the preliminary experiments involving the CF probe were carried out with visual examination of the eluting solution, which was strongly colored, followed by —uorescence analysis of the collected fractions. The majority of the experiments (especially those involving glucose) were performed with a Biologic LP chromatography system (Bio-Rad) equipped with a model 2128 fraction collector and a double trace recorder (Tracelab BD 41).Both the absorbance (or turbidity) measured at 254 nm and the conductivity of the eluted solution were recorded. The void volume of the Sephadex column was calibrated using Blue Dextran 2000 kDa and found to be of the order of 13 ml for 4 g of Sephadex.During the preequilibration of the column, the absence of change of conductivity of the eluent was used as a test to ensure a good conditioning of the column. The eluent used for the chromatography of the dispersions of vesicle-encapsulated CF was a solution containing 10 mM HEPES buÜer (pH 7.0) and 0.1 M NaCl.This was assumed to limit osmotic pressure variations between the inner and outer vesicle compartments and thus to avoid osmotic swelling. In the case of vesicle-encapsulated glucose, HEPES buÜer was used to maintain the pH, without any other added solute. Under these conditions a partial breaking of the vesicles, due to osmotic stress, cannot be completely ruled out.The —uorescence measurements were carried out on a Shimadzu FR 540 spectro—uorimeter. For carboxy—uorescein the excitation and emission wavelengths were 490 and 520 nm, respectively. To measure the concentration of encapsulated New J. Chem., 1999, 23, 981»987 983CF, 0.5 ml of Triton X-100 at concentration 0.16 M were added to 0.5 ml of the vesicular dispersion and diluted with the buÜer (HEPES 0.01 M, NaCl 0.1 M, pH 7.4) to a –nal volume of 3 ml.The CF concentration was determined from a standard curve, with all the caution necessary to be sure that the dilution is large enough so that there is no —uorescence quenching. For polarization measurements with DPH (3]10~5 M), polarizers were introduced in the light beam before and after the —uorescence cell.The excitation and emission wavelengths were 360 and 430 nm, respectively. The chain melting temperature was taken to be equal to the temperature at midtransition. For glucose trapping experiments the analysis of the encapsulated and free glucose respectively was based on the usual enzymatic reaction.18 The glucose was oxidized by glucose oxidase leading to a production of hydrogen peroxide.The latter reacts with o-dianisidine (Sigma) to form a coloured product absorbing at 435 nm. A Varian Cary 3E spectrophotometer was used for the measurements. The glucose content was determined from a calibration curve obtained with known concentrations of glucose. Triton X-100 was added to the dispersions, in conditions similar to those reported above in the case of CF, to ascertain the breaking of the vesicles.Results and discussion Synthesis of PALAmod 4 The synthesis of PALAmod 4 was based on the preparation of the precursor 1 from a classical coupling between diethylphosphonoacetic acid16 and H-Asp(OBn)-OBn using BOP as the coupling reagent (82%).15 Benzyl esters of 1 could be chemoselectively hydrogenated with a Pd»C catalyst to give 2 keeping the diethylphosphonic ester intact (98%).Reaction of 2 with an excess of stearic alcohol (4 equiv.) and BOP as coupling reagent led to dioctadecyl N-(diethylphosphonoacetyl)- L-aspartate 3 (74%). Chemoselective deprotection of the diethylphosphonic ester was then achieved with a large excess of bromotrimethylsilane (15 equiv.) followed by washing of the resulting brown solid with wet acetone and aÜorded the target compound 4 (97%).Preparation and characterization of vesicles Stability. DiÜerent kinds of dispersions including PALAmod or a mixture of PALAmod and lecithin have been prepared by sonication as indicated in the Experimental section. The composition of these dispersions are indicated in the diÜerent Tables.We have –rst checked that, with PALAmod alone, stable dispersions presenting a slight opalescence, are obtained in a pH range of 2.4 to 11.8. The following experiments were performed at neutral pH, so as to meet physiological conditions. The in—uence of the salinity was also investigated. With monovalent salts (0.1 M of NaCl or KBr), added before the sonication in a 3 mM dispersion of PALAmod, a spectroscopic observation of the turbidity indicates that the dispersions are stable for more than a week.However, a salt addition after the sonication induces a destabilization of the dispersion. This can be assumed to be due to the osmotic stress resulting from the diÜerence between the intra- and extra-vesicular electrolyte concentrations, respectively.Transition temperature. A well-known property of vesicular systems is the melting temperature of the alkyl chains of Tm the bilayer, which characterizes the transition between the gel (rigid) phase b and the —uid (liquid crystalline) phase a. The value of was measured from the —uorescence polarization Tm Fig. 1 DPH —uorescence polarization vs. temperature in vesicles of PALAmod.[DPH]\3]10~5 M; [PALAmod]\1]10~3 M. of DPH,19 a lipophilic —uorescent probe whose motion is very sensitive to the —uidity of the amphiphilic membrane. The variation of —uorescence polarization as a function of temperature is represented in Fig. 1. The results indicate a transition temperature Taking into account the fact Tm\62 °C. that the polar heads are not identical, this value is roughly comparable to the values measured20 for distearoyl phosphatidylcholine (DSPC) or distearoyl phosphatidylglycerol (DSPG), 55 °C, which have fatty chains of the same length.This result is in line with the idea that vesicles are formed. However we cannot completely rule out at this point the possibility of having open fragments of bilayers, as reported in other circumstances.21 For this reason, in order to demonstrate the existence of closed structures, we have undertaken encapsulation experiments. Encapsulation.We have considered two diÜerent hydrophilic probes which have been extensively used in the literature : carboxy—uorescein22 (CF) and glucose.23,24 The former is very convenient because its —uorescence is quenched when its concentration is above 1.5]10~5 M.So if we are able to encapsulate a 50 mM concentration of CF, purify the obtained vesicles and provoke their rupture in one way or another, we should see a —uorescent enhancement due to the dilution. With the second probe the determination of the amount of encapsulated glucose is more tedious and an enzymatic reaction associated with a colorimetric method must be used.However, the fact that the latter probe is neutral constitutes a signi–cant advantage compared to CF (the double negative charge may cause problems when electrostatic interactions with the vesicular particle can occur). Whatever the type of molecule used to probe the encapsulation, steric exclusion chromatography (SEC) has been used to separate the encapsulated product from the free one.Once the vesicles have been separated, the amount of encapsulated product was determined after breaking of the vesicles by addition of Triton X-100.22 An illustration of the results obtained is given in Fig. 2, which is relative to glucose. We have collected in Table 1 some of the information gained from CF encapsulation. DiÜerent compositions have been tested. When PALAmod was used alone (C dispersion) or in a mixture containing 25% lecithin (D dispersion), no encapsulation could be measured because the systems were unstable.Increasing the proportion of lecithin to 50% (E dispersion) or 75% (F dispersion) gives stable dispersions with an unchanged percentage of encapsulated CF of 0.21. The addition in the D dispersion of cholesterol, which is expected to rigidify the membrane25 (G dispersion) improves the stability and leads to a quite similar encapsulation : 0.19%.These 984 New J. Chem., 1999, 23, 981»987Fig. 2 Gel permeation chromatographic separation of glucosecontaining vesicles of PALAmod from free glucose. results can be compared with those obtained with pure lecithin (B dispersion) or with a mixture of lecithin and cholesterol (A dispersion) : 0.25 and 0.44% respectively.The negative result obtained with PALAmod alone is obviously related to the problems of electrostatic interactions between the negatively charged surfactant and CF. A dilution of PALAmod in lecithin (zwitterionic surfactant) favors the formation of stable vesicles, provided that the mixture contains a sufficient amount of lecithin.This amount can be reduced in the presence of cholesterol. We also notice that the encapsulated amount is independent of the molar ratio between PALAmod and lecithin when stable dispersions are obtained (see E and F dispersions). This can be taken as proof that the encapsulation cannot be due to vesicles of pure lecithin should they coexist with particles of PALAmod.We have tested the stability of the G dispersion over a period of two weeks. The measured variations of the amount of encapsulated CF are almost within the accuracy of the experiments: starting from a value of 0.19% we found 0.16 after 24 hours, 0.17 after 48 hours and 0.16 after 2 weeks. These small variations indicate that leakage of the encapsulated probe is almost negligible.Fig. 3 Variation of the amount of encapsulated glucose (%) vs. PALAmod concentration. The results obtained with the second probe (glucose) demonstrate that vesicles of pure PALAmod are truly formed in the absence of CF. As shown in Table 2 and in Fig. 3, the percentage of encapsulated glucose varies from 0.09 to 0.24% when the PALAmod concentration goes from 3.7 to 14.8 mM.We have also indicated in Table 2 the size of the particles measured from quasielastic light scattering experiments. The decrease of the size is perfectly consistent with the non-linear behaviour observed in Fig. 3. Indeed a rough estimation of the encapsulated volume can be obtained as follows : EV EV\ 4 3 p Dw3 8 N (1) where is the diameter of the water core of the particles and Dw N their number.N\ c … NA … ao 2pDw2 (2) where c is the amphiphile concentration, the Avogadro NA number, the surface per polar head and the factor takes ao 12 into account the distribution of the amphiphile between the outside and inside layers, respectively (this is only a crude approximation as indicated above, since for geometrical reasons the partitioning of the amphiphiles between the two layers is in the order of 60 : 40 rather than 50 : 50).26,27 Introducing eqn.(2) into eqn. (1) and multiplying by 100 to express the value in percent of the total volume leads to EV (%)\ 25 3 c … NA … ao … Dw (3) The values experimentally measured are approximately six times smaller than those calculated from this eqn. (see below), but a more interesting thing is the variation of experimen- EV tally observed when the PALAmod concentration is varied.From eqn. (3) we expect a linear dependence of with c only EV if remains unchanged. This is not the case according to the Dw particle size measurements reported in Table 2. At –xed pH (6.9^0.2), the particle diameter D goes from 74 nm at [PALAmod] 3.6 mM to 52 nm at 14.8 mM. For a –xed concentration the ratio of the values obtained for two diÜerent EV sizes of particles should be such that EV1 EV2 \ Dw1 Dw2 (4) This is almost exactly veri–ed with the results reported in Fig. 3, if we compare the value obtained from a linear extrapo- EV lation of the results at low concentration we obtain at [PALAmod] 14.8 mM, whereas the experimen- EV1\0.36%, tal value is 0.24%.The ratio between the two values is of EV2 the order of 1.5 : 1. If now we estimate the value of from Dw Dw\D[2l (5) where l is the thickness of the bilayer (about 50 we obtain ”), So the drop of compared to the Dw1/Dw2\64/42\1.52. EV linear behaviour is almost exactly what can be expected from the variation of the size of the particles. This can be taken as additional proof that PALAmod vesicles are really formed.Table 1 Percentages of encapsulated carboxy—uorescein Composition of dispersions A B C D E F G [PALAmod]/mM 0 0 5 3.9 2.6 1.3 3.9 [a-Lecithin]/mM 5.5 5 0 1.3 2.6 3.9 1.3 [Cholesterol]/mM 3.5 0 0 0 0 0 1.2 Stability Stable Stable Unstable Unstable Stable Stable Stable (%) Encapsulated 0.44 0.25 0 0 0.21 0.21 0.19 New J. Chem., 1999, 23, 981»987 985Table 2 Percentage of glucose encapsulation and particle diameters Amphiphile nature Final pH Glucose Particle Experiment and concentration/ Sonicated (after encapsulation size/ number mM solution sonication) Eluent composition (%) nm B2 PALAmod HEPES 10 mM 4.80 HEPES 10 mM 0.09 95 3.7 mM Glucose 0.5 M (pH\7) B3 PALAmod HEPES 10 mM 6.70 HEPES 10 mM 0.09 74 3.6 mM Glucose 0.5 M (pH\7) NaOH 2.7 mM B4 PALAmod HEPES 10 mM 7.00 HEPES 10 mM » 72 3.7 mM Glucose 0.5 M (pH\7) NaOH 3.7 mM B7 PALAmod HEPES 10 mM 6.80 HEPES 10 mM 0.17 65.5a 7.4 mM Glucose 0.5 M (pH\7) NaOH 7.4 mM B8a PALAmod HEPES 10 mM 6.90 HEPES 10 mM 0.17 58 7 mM Glucose 0.5 M (pH\7) B8b NaOH 7.4 mM 6.90 HEPES 10 mM 0.22 58 (pH\7)]0.13 mM Lecithin B8c 6.90 HEPES 10 mM 0.17 58 (pH\7)]0.13 mM PALAmod B9a PALAmod HEPES 10 mM 7.09 HEPES 10 mM 0.19 62 11.3 mM Glucose 0.5 M (pH\7) B9b NaOH 11.3 mM 7.09 HEPES 10 mM 0.32 62 (pH\7)]0.13 mM Lecithin B10 PALAmod HEPES 10 mM 7.00 HEPES 10 mM 0.24 52 14.8 mM Glucose 0.5 M (pH\7) NaOH 14.8 mM B12b PALAmod HEPES 10 mM 7.00 HEPES 10 mM 0.07 65 7.2 mM Glucose 0.5 M (pH\7) NaOH 7.2 mM U Phosphatidylcholine HEPES 10 mM 7.00 HEPES 10 mM 0.25 » 3.6 mM Glucose 0.5 M (pH\7)]0.13 mM Lecithin A3 Phosphatidylcholine HEPES 10 mM 7.00 HEPES 10 mM 0.31 100 7.2 mM Glucose 0.5 M (pH\7)]0.13 mM Lecithin a Average over two diÜerent experiments.b Sonication time was 15 minutes instead of 5 as for the other experiments. However the low absolute value of compared to the EV theoretical prediction remains a puzzling problem. Indeed, assuming as previously reported for phospha- ao\61 ”2, tidylcholine at the inner surface of vesicles,26 one obtains 0.58% at PALAmod concentration 3 mM nm) and (Dw\64 1.9% at 14.8 mM nm), when the experimental values (Dw\42 are 0.09 and 0.24% respectively.Taking into account the dissymmetry between the two amphiphile layers would reduce the theoretical values by less than 20%.For the sake of comparison we have performed glucose encapsulation experi- Table 3 Analysis of a series of collected fractions : glucose concentration, turbidity, size of particles. Initial condition : [PALAmod]\7.4 mM, HEPES buÜer 10 mM, pH\7, glucose 0.5 M (B7 in Table 2) Glucose Test tube Eluted concentration/ Turbidity Particle size/ number volume/ml lmole l~1 (254 nm) nm 9 9.9 0 0 No diÜusion 10 11 0 0 No diÜusion 11 12.1 0 0 No diÜusion 12 13.2 0 0 No diÜusion 13 14.3 0 0 No diÜusion 14 15.4 0 0 No diÜusion 15 16.5 19 0.003 No diÜusion 16 17.6 67 0.017 65 17 18.7 78 0.032 52 18 19.8 67 0.024 58 19 20.9 57 0.019 62 20 22 41 0.013 58 » » » » » 24 26.4 23 0.003 50 25 27.5 17 0.002 65 26 28.6 10 0.001 No diÜusion 27 29.7 8 0.001 No diÜusion 28 30.8 0 0.001 No diÜusion 29 31.9 0 0 No diÜusion 30 33 0 0 No diÜusion 986 New J.Chem., 1999, 23, 981»987ments, in similar conditions, using lecithin instead of PALAmod. The measured values were 0.25% at concentration 3.6 mM and 0.31% at 7.2 mM. We are closer to the theoretical value, but still far from this value by a factor of 2 to 3. The same kind of gap was also observed by Walde et al.23 in the case of mono-n-alkylphosphate vesicles. DiÜerent explanations can be put forward to explain our results : (i) closed vesicles could coexist with fragments of bilayer or with open vesicles presenting some leakage due for instance to the existence of pores across the bilayer ; (ii) the calculation is valid only for single-wall vesicles, but in the case of multilayer vesicles, the encapsulated volume could be considerably decreased : experiments are presently in progress to check this point ; (iii) we cannot exclude the fact that the closing of lamellar fragments under sonication may be accompanied by some expulsion of the entrapped probe due to volume exclusion or steric eÜects, resulting in diÜerent concentrations of the probe inside and outside the particle ; (iv) the volume occupied by the polar heads in the inner water core is not taken into account in the calculation and may also contribute to a reduction of the entrapped volume; (v) –nally the light scattering experiments give a hydrodynamic radius so that the particle diameter introduced in eqn.(3) may be slightly overestimated. In addition, it is an average value and it is well known that the largest particles are contributing more than the smallest ones to the scattering intensity.Eqn. (3) strictly applies only to monodisperse systems with a well de–ned particle diameter. The polydispersity of the system may also contribute to lowering the value of experimentally measured EV because the particles having diameters much smaller than average oÜer reduced encapsulation capability.We should also mention here that we have run some additional gel exclusion chromatographic experiments in which the column was preequilibrated not only with the HEPES buÜer, but also with a vesicular dispersion. This was assumed to avoid (or at least reduce) possible interactions between the vesicles under study and the solid phase, which could have destabilized the colloidal particles.The results shown in Table 2 indicate a total absence of eÜect of this procedure when PALAmod was used in the eluent. A small increase of from EV 0.17 to 0.22 was however observed when eluting with lecithin, but this would simply indicate a possible exchange of amphiphile molecules during the migration. We have also checked that the average particle diameter measured is independent of the eluted volume and that the values are thus very similar whatever the fraction collected during the recording of the turbidity peak (see Fig. 2). The results obtained for a PALAmod concentration 7.4 mM, in the presence of encapsulated glucose, are shown in Table 3 for the whole series of collected fractions. Conclusion We have shown in this work that the amphiphile derivative of PALA (dioctadecyl N-(phosphonoacetyl)-L-aspartate, PALAmod) is able to self-organize in bilayer structures.In addition, the formation of vesicles of either pure PALAmod or of mixtures of PALAmod with lecithin was demonstrated from encapsulation experiments and the average size of the particles was determined. However, we cannot exclude at the present stage of this work the coexistence of other types of objects in the dispersions. These preliminary results encourage us to carry on with research along these lines.Acknowledgements –nancial support of this work through a speci–c action Partial of Institut Nanceç ien de Chimie Moleç culaire (INCM) is greatly acknowledged. We also wish to thank Dr M.-J. Steç beç (UMR CNRS 7565) for the use of the light scattering and —uorescence equipment, G.Raval (Laboratoire de Biochimie UHPNancy I) for suggesting the method of analysis for glucose determination, and Dr A. Van Dorsselaer for the mass spectral measurements (UMR 7509, Strasbourg University). References 1 J. Chauvergne and B. Hoerni, Chimiotheç rapie anticanceç reuse, Masson, Paris, 1992, p. 1. 2 M. Le Maire, R. Chabaud and G. Herveç , Biochimie, Un mode` le lœAspartate transcarbamylase, Masson, Paris, 1990, p. 1. dœeç tude : 3 (a) K. D. Collins and G. R. Stark, J. Biol. Chem., 1971, 246, 6599; (b) E. A. Swyryd, S. S. Seaver and G. R. Stark, J. Biol. Chem., 1974, 249, 6945. 4 K. K. Tsuboi, H. N. Edmunds and L. K. Kwong, Cancer Res., 1977, 37, 3080. 5 J. L. Grem, S.A. King, P. J. OœDwyer and B. Leyland-Jones, Cancer Res., 1988, 48, 4441. 6 (a) A. D. Morris and A. A. Cordi, Synth. Commun., 1997, 27, 1259; (b) P. Henklein and J. Gloede, Z. Chem., 1989, 29, 19; (c) J. Gloede, H. Gross, P. Henklein, H. Niedrich, St. Tanneberger and B. Tschiersch, Pharmazie, 1988, 43, 434; (d) J. L. Montero and J. L. Imbach, Eur. J. Med. Chem.-Chim. T her., 1982, 17, 97; (e) B.B. Natale, A. Yagoda, D. P. Kelsen, R. J. Gralla and R. C. Watson, Cancer T reat. Rep., 1982, 66, 2091; ( f ) P. Kafarski and M. Soroka, Synthesis, 1982, 219; (g) J. J. Goodson, C. J. Wharton and R. J. Wrigglesworth, J. Chem. Soc., Perkin T rans. 1, 1980, 2721; (h) T. D Kempe, E. A. Swyryd, M. Bruist and G. R. Stark, Cell, 1976, 9, 541. 7 J. Courtland White and L. H. Hines, Cancer Res., 1984, 44, 507 and refs. therein. 8 T. L. Loo, J. Friedman, E. C. Moore, M. Valdivieso, J. R. Marti and D. Stewart, Cancer Res., 1980, 40, 86. 9 (a) S. D. Lindell and R. M. Turner, T etrahedron L ett., 1990, 5381; (b) S. Joliez, Thesis, Universiteç Henri Poincareç -Nancy 1, 1998. 10 A. Sharma, N. L. Straubinger and R. M. Straubinger, Pharm. Res., 1993, 10, 1434. 11 J.-S. Kim and T. D. Heath, J. Controlled Release, 1996, 40, 101. 12 N. Garti, D. Lichtenberg and T. Silberstein, Colloids Surf., 1997, 128, 17. 13 M. P. Teulade, P. Savignac, E. E. Aboujaoude and N. Collignon, J. Organomet. Chem., 1986, 312, 283. 14 B. Castro, J. R. Dormoy, G. Evin and C. Selve, T etrahedron L ett., 1975, 1219. 15 P. Coutrot, C. Grison and C. Charbonnier-Geç rardin, T etrahedron, 1992, 48, 9841. 16 (a) P. Coutrot and A. Ghribi, Synthesis, 1986, 661; (b) P. Coutrot, M. Snoussi and P. Savignac, Synthesis, 1978, 133. 17 M. C. Cleij, P. Scrimin, P. Tecilla and U. Tonellato, L angmuir, 1996, 12, 2956. 18 A. Huggett and D. Nixon, Biochem J., 1957, 66, 12. 19 M. P. Andrich and J. M. Vanderkoi, Biochemistry, 1976, 15, 1257. 20 F. Szoka and D. Papahadjopoulos, Annu. Rev. Biophys. Bioeng., 1980, 9, 467. 21 R. P. Pansu, B. Arrio, J. Roncin and J. Faure, J. Phys. Chem., 1990, 94, 796. 22 Y. Liu and S. L. Regen, J. Am. Chem. Soc., 1993, 115, 708. 23 P. Walde, M. Wessicken, U. Raé dler, N. Berclaz, K. Conde-Friebos and P. L. Luisi, J. Phys. Chem. B, 1997, 101, 7390. 24 Y. Kondo, H. Uchiyama, N. Yoshino, K. Nishiyama and M. Abe, L angmuir, 1995, 11, 2380. 25 M. Shinitzky and Y. Barenholz, Biochim. Biophys. Acta, 1978, 515, 367. 26 J. H. Fendler, Membrane Mimetic Chemistry, John Wiley & Sons, New York, 1982, pp. 129»130. 27 G. Ghirlanda, P. Scrimin, P. Tecilla and U. Tonellato, J. Org. Chem., 1993, 58, 3025. Paper 9/04759J New J. Chem., 1999, 23, 981»987 987

ISSN:1144-0546    

DOI:10.1039/a904759j

出版商:RSC    

年代:1999

数据来源: RSC

摘要:

Cycloaddition of some arylnitriloxides with 3-methylenephthalide : electrocatalytic opening of the adducts Christophe Roussel,a Kabula Ciamala,*a Pierre Audebert*b and J. Vebrela a L aboratoire de Chimie et Electrochimie de Franche Moleç culaire, Universiteç Comteç , 16 route de Gray, F-25030 cedex, France Besanc” on b L aboratoire de Photophysique et Photochimie et Supramoleç culaires Macromoleç culaire, Ecole Normale de Cachan, 71 Av.du Pdt W ilson, F-94235 Cachan cedex, France. Supeç rieure E-mail : audebert=ppsm.ens-cachan.fr Received (in Montpellier, France) 14th June 1999, Accepted 3rd August 1999 Some 4-substituted arylnitriloxides undergo 1,3 dipolar cycloadditions with 3-methylenephthalide with formation of spirodihydroisoxazoles. These spiroadducts can be opened to the corresponding 2-(3-arylisoxazol-5-yl)benzoic acid by various methods, including thermal and acidic treatments, as well as electrooxidation.In the latter case, the ring-opening mechanism is shown to involve a catalytic electron exchange between the substrate and an intermediate radical. ArCO2~ Cycloaddïç tion de quelques arylnitiloxydes avec le 3-meç thyle` nephthalate : ouverture e ç lectrocatalytique des adduits.Nous avons fait reç agir quelques arylnitriloxydes substitueç s en position 4 sur le cycle aromatique avec le 3-meç thyle` nephthalide via une reç action de cycloaddition [3]2]. La reç action conduit a` lœobtention dœun spiroheç teç rocycle unique. Ces spiroadduits peuvent se reç arranger en acide 2-(3-arylisoxazol-5-yl)benzoïé que en milieu acide ou basique, par thermolyse et par eç lectrooxydation.Dans ce dernier cas, lœouverture sœeÜectue par un eç change eç lectronique catalytique entre le substrat et un radical de type produit intermeç diairement. ArCO2~ The recurrent interest in bicyclic molecules with a spirannic junction leads us to reinvestigate the reactivity of some amethylene- and c-methylene-c-butyrolactones with selected 1,3-dipoles.1 We present here the reaction of 3-methylenephthalide with various 4-substituted arylnitriloxides.This reaction was previously investigated by Liu and Howe2 with diÜerently substituted substrates ; it is shown here that their conclusion can be generalized, whatever the electronic eÜect of the substituent in the 4 position of the aromatic ring of the dipole, with the same regio- and stereochemistry.In addition, we investigated extensively the opening of the cycloadducts. We show that acid catalysed opening takes place much more easily in polar solvents like acetonitrile at room temperature. It is also possible to electrooxidize the resulting spiroisoxazoles, which occur in high yields with ring Scheme 1 opening of the benzocondensed ring and formation of the same benzoic acid as that produced by thermal or acid treatment of the adducts (Scheme 1).Moreover, we suggest that an original electrocatalytic mechanism is responsible in this particular case, through the formation of an intermediate that undergoes an elec- ArCO2~ tron transfer reaction rather than the classical Kolbe decarboxylation, which is known to be disfavoured in the case of aromatic acids.3,4 Results and discussion Cycloadditions We have submitted to cycloaddition reactions with 3- methylenephthalide 1 the arylnitriloxides 2a»d, respectively unsubstituted or bearing a methyl, methoxy or nitro group on the para position of the ring, according to Scheme 1.In each case the conclusion of Liu and Howe was veri–ed ; the regioand stereochemistry of the cycloadduct are in favour of the sole formation of the adduct 3 (Scheme 1).This is proved unambiguously by the 13C NMR displacement of the spirannic C, which is observed at 112 ppm, while in the case of the formation of the other isomer (with inversion of the spirannic carbon) it would have been observed at 60»70 ppm. Chemical opening of the adducts As was observed by Liu and Howe2 in the case of 3a only the adducts can be opened by thermal (200»230 °C) treatment, as well as by the action of a diluted acid (chlorhydric acid) or alkali (NaOH), giving the benzoic acids 4a»d.This reaction is general and insensitive to the substituent on the phenyl ring, which is not unexpected in this case given that the most probable mechanism involves an attack on the lactone, and not on the isoxazole ring.However, we found that acetonitrile was a much better solvent for acidic opening of the adducts. In fact, New J. Chem., 1999, 23, 989»992 989 This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scienti–que 1999 (the ring-opening reaction was complete with a catalytic amount of chlorhydric acid at room temperature in 15»20 min instead of 1 h at re—ux in dioxane as previously described.2 Electrochemical opening of the adducts Although chemical oxidation of some isoxazolines has been described in the literature,5h9 the electrooxidation of rings such as the adducts 3 is unknown to date.The cyclic voltammograms (CVs) of the adducts 3 have been performed and are relatively similar, as exempli–ed by the case of 3a (Fig. 1, curves a»d). The voltammograms feature two irreversible peaks, the –rst being in the 1.1»1.2 V region and the second towards 1.8 V (except for 3d) ; the potentials are reported in Table 1. In addition, the –rst peak displays an anomalous behaviour since it vanishes at low scan rates, while it grows in at higher scan rates to reach an intensity approximately equal to that of the second peak.By simple comparison with the CVs of the resulting acids 4 (Fig. 2), it is clear that the second peak must be ascribed to the irreversible oxidation of the isoxazolic ring (Table 2). Electrolyses have been performed, at potentials given in Table 1. The electrolyses are complete after ca. 5% of the theoretical amount of charge has been passed (on the basis of one electron per mol). The results of the constant potential electrolyses are independent of the applied potential in the 1.2»2.0 V range. On the basis of these results, we propose the electrocatalytic mechanism pictured in Scheme 2. Since the Kolbe reaction is known not to occur with aromatic acids,4 the radical ArCO2~ is likely to undergo further electron transfer to the substrate.Fig. 1 Electrochemical oxidation of the spiroadduct 3a in CH3CN with 0.1 M (a) Scan rate 0.05 V s~1, C\5 mM, on a 1 NEt4ClO4 . mm diameter Pt electrode. (b) Scan rate 0.27 V s~1, C\5 mM, on a 1 mm diameter Pt electrode. (c) Scan rate 2.70 V s~1, C\1 mM, on a 0.5 mm diameter Pt electrode. (d) Scan rate 27.0 V s~1, C\1 mM, on a 0.5 mm diameter Pt electrode.Table 1 Electrochemical data for the spiroadducts 3a Product Ev b/V Ew b/V 3a 1.15»1.55 1.40 3b 1.15»1.60 1.40 3c 1.15»1.55 1.40 3d 1.80»2.35 2.10 a Electrolysis time of 10 min. and are the catalytic wave b Ev Ew potential and working electrode potential during the electrolysis, respectively, vs. SCE. Fig. 2 Electrochemical oxidation of the spiroadduct 3a (left) and resulting acid 4a (right) in with 0.1 M performed CH3CN NEt4ClO4 on a 3 mm diameter carbon electrode, scan rate 0.05 V s~1, C\1 mM.There is no data on the formal redox potentials of the couple (its determination is almost impossible RCO2~/RCO2~ due to the fast follow-up reactions). However, the peak potential for the oxidation of benzylic acids is in the ]1.0 V range; therefore, it is likely that the formal potential E° in our case lies quite above 1.1 V, which is high enough to allow the oxidation of the substrate by the radical.ArCO2~ Scheme 2 Another possible explanation could be that the opening is catalysed by protons coming from the –rst oxidation reaction ; however, in the case of acidic treatment the reaction requires a longer time and much more acidic conditions to reach completion.TLC monitoring shows that the ring-opening reactions in acetonitrile with HCl require 20 min with a ca. –ve-fold excess of hydrochloric acid, while the electrochemical opening of the adducts is complete within 10 min. A further explanation could involve hydrogen atom abstraction (on the isoxazolinic ring) from substrates 3 by the electrogenerated radical, followed by reoxidation and aromatization.In order to rule out this hypothesis, we have re—uxed compounds 3 in methanol in the presence of benzoyl peroxide, which is known to form phenyl radicals under these conditions. Compounds 3 are insensitive to such treatment, since they are recovered unchanged after the re—uxing treatment. Therefore, it can be concluded that the substrates do not react with radicals, Ph~ which therefore rules out any radical based H atom abstraction by the intermediate.Table 2 Electrochemical data for the ring opening products 4 Product Ep a/V 4a 2.33 4b 2.31 4c 2.29 4d 2.60 is the peak potential vs. SCE. a Ep 990 New J. Chem., 1999, 23, 989»992Experimental Melting points were determined with a digital Electrothermal IA 9200 apparatus and are uncorrected.IR spectra (KBr) were recorded on a Bio-Rad FTS-7 spectrophotometer. 1H and 13C NMR spectra were recorded on a Bruker Spectrospin AC 200 spectrometer operating at 200 MHz for 1H and at 50 MHz for 13C spectra. Chemical shifts were measured relative to TMS. Analytical data were obtained by the CNRS Vernaison (France) and were satisfactory (C, H, N ^0.30% from theoretical). Synthesis of compounds 1, 2a»d and 3a»d The 3-methylenephthalide 1 and arylnitriloxides 2 were prepared according to literature procedures.2,10h14 The cycloadditions giving 3 were realized in diethyl oxide at room temperature, under nitrogen atmosphere with 50 mg of hydroquinone added in order to prevent polymerization of the dipolarophile, according to our previously published procedure. 1 3a.Yield : 1 g (76%), mp 131 °C (ethanol), green solid, mp lit 120»122 °C. Anal. Calcd. % for C, 72.44 ; H, C16H11NO3 : 4.18 ; N, 5.28 ; Found C, 72.31 ; H, 4.20 ; N, 5.34. IR (KBr) m 8 1595, 1770 cm~1; H NMR d 4.10»4.50 (AB (d6-DMSO) system, J\18.5), 7.45»8.15 (m, 9H, arom); 13C NMR (d6- d 44.3, 112.3, 123.9»143.3, 159.1, 166.4.DMSO) 3b. Yield : 0.71 g (51%), mp 135 °C (ethanol), white solid. Anal. Calcd. % for C, 73.11 ; H, 4.69 ; N, 5.06 ; C17H13NO3 : Found C, 73.03 ; H, 4.74 ; N, 5.12. IR (KBr) 1595, 1770 m 8 cm~1; 1H NMR d 2.40 (s, 3H, 4.10»4.50 (d6-DMSO) CH3), (AB system, J\18.5), 6.95»8.25 (m, 8H, arom); 13C NMR d 20.8, 44.4, 112.1, 123.8»143.3, 158.9, 166.4 (d6-DMSO) 3c.Yield : 0.93 g (63%), mp 168»169 °C (ethanol), beige solid. Anal. Calcd. % for C, 69.14 ; H, 4.44 ; N, C17H13NO4 : 4.74 ; Found C, 69.22 ; H, 4.51 ; N, 4.79. IR (KBr) 1595, 1770 m 8 cm~1; 1H NMR d 3.85 (s, 3H, 4.10»4.50 (d6-DMSO) OCH3), (AB system, J\18.6), 6.95»8.25 (m, 8H, arom); 13C NMR d 44.6, 55.2, 112.2, 114.4»161.4, 158.5, 166.4 (d6-DMSO) 3d. Yield : 0.34 g (22%), mp 181»183 °C (ethanol), yellow solid.Anal. Calcd. % for C, 61.94 ; H, 3.25 ; N, C16H10N2O5 : 9.03 ; Found C, 62.06 ; H, 3.29 ; N, 9.06. IR (KBr) 1590, 1770 m 8 cm~1; 1H NMR d 4.20»4.60 (AB system, (d6-DMSO) J\18.6), 7.70»8.60 (m, 8H, arom); 13C NMR d (d6-DMSO) 43.8, 112.6, 124.2»148.7, 158.3, 167.3. Ring-opening reactions Opening through acidic treatment. To a stirred solution of spiroadducts 3 (0.25 mmol) in acetonitrile (4.85 mL) was added slowly, at room temperature, 0.15 mL of concentrated HCl.Stirring was continued for 15»20 min, the solvent was removed in vacuo and the obtained residue 4 was recrystallized from carbon tetrachloride. Yields were quantitative. Opening through electrochemical oxidations. In the anodic compartment of a two-compartment cell –tted with a saturated calomel reference electrode (SCE), was placed a spiroadduct 3 (0.25 mmol) dissolved in acetonitrile (10 mL) and 0.1 M of tetraethylammonium perchlorate.The cathodic compartment was –lled with 10 mL of pure electrolyte. A 3.5 cm2 platinum plate was used as the working electrode, and the electrolyses were performed under argon atmosphere at a constant potential of ]1.4 V (]2.1 in the case of 3d). The electrolyses were stopped when the current reached 1.5 times the background current of the cell (separately measured with no substrate added).After the end of the electrolysis, acetonitrile was evaporated in vacuo and the compounds 4 were then extracted and puri–ed as before, with thorough washing to eliminate residual traces of the electrolyte salt. 4a. Yield : 0.050 g (75%), mp 190»191 °C white solid. (CCl4), Anal. Calcd. % for C, 72.44 ; H, 4.18 ; N, 5.28 ; C16H11NO3 : Found C, 72.31 ; H, 4.20 ; N, 5.34. IR (KBr) 3095»2885, 1690, m 8 1610 cm~1. 1H NMR d 4.65 (s, 1H, OH), 6.80 (s, 1H, (CDCl3) H isoxazole), 7.15»8.10 (m, 9H, arom); 13C NMR d (CDCl3) 101.2, 126.8»132.1, 159.9, 167.5, 169.9. 4b. Yield : 0.055 g (78%), mp 160»162 °C orange (CCl4), solid.Anal. Calcd. % for C, 73.11 ; H, 4.69 ; N, C17H13NO3 : 5.06 ; Found C, 73.03 ; H, 4.74 ; N, 5.12. IR (KBr) 3100»2865, m 8 1690, 1605 cm~1. 1H NMR d 2.40 (s, 3H, 3.65 (CDCl3) CH3 ), (s, 1H, OH), 6.80 (s, 1H, H isoxazole), 7.15»8.10 (m, 8H, arom); 13C NMR d 21.2, 101.0, 125.9»139.9, 162.4, (CDCl3) 169.1, 171.1. HRMS (m/z) : 279, 278, 188, 158, 77. 4c.Yield : 0.055 g (74%), mp 151»152 °C beige solid. (CCl4), Anal. Calcd. % for C, 69.14 ; H, 4.44 ; N, 4.74 ; C17H13NO4 : Found C, 62.22 ; H, 4.51 ; N, 4.79. IR (KBr) 3100»2870, 1695, m 8 1615 cm~1. 1H NMR d 2.90 (s, 1H, OH), 3.85 (s, 3H, (CDCl3) 6.75 (s, 1H, H isoxazole), 7.15»8.10 (m, 8H, arom); OCH3), 13C NMR d 55.2, 100.9, 114.1»161.8, 160.8, 169.0, (CDCl3) 170.7. 4d. Yield : 0.060 g (77%), mp 199»200 °C yellow (CCl4), solid.Anal. Calcd. % for C, 61.94 ; H, 3.25 ; N, C16H10N2O5 : 9.03 ; Found C, 62.06 ; H, 3.29 ; N, 9.06. IR (KBr) 3115»2840, m 8 1690, 1600 cm~1. 1H NMR d 2.90 (s, 1H, OH), (d6-acetone) 7.30 (s, 1H, H isoxazole), 7.60»8.50 (m, 8H, arom); 13C NMR d 101.8, 124.8»149.4, 161.8, 167.9, 171.7. (d6-acetone) Electrochemical setup Analytical experiments were performed in a threecompartment cell –tted with an SCE, a glassy carbon (diameter 3 mm) or platinum (diameter 1 or 0.5 mm) electrode and a platinum counter electrode.The electrochemical apparatus was composed of a home-made potentiostat15 (equipped with an ohmic drop compensation system), a PAR 173 Universal programmer, a Nicolet digital oscilloscope and a Sefram 164 plotter.The solvent was spectroscopic grade acetonitrile [distilled over and stored on 3 molecular CaCl2 ” sieves with 0.1 M tetraethylamonium perchlorate (Fluka puriss, recrystallized once in acetonitrile»diethyl oxide)] as supporting electrolyte. The concentration of product 3 or 4 was usually 1 or 5]10~3 M and the cell was —ushed with argon throughout the experiment. Ohmic compensation was used when necessary (i.e., for scan rates over 1 V s~1).Electrosyntheses were performed in a two-compartment cell –tted with a SCE as reference electrode, a platinum work electrode (diameter 15 mm) and platinum wire counter electrode. The electrochemical apparatus was a radiometer PGP 201 potentiostat. Conclusion The 1,3 dipolar cycloadditions of 4-substituted arylnitriloxides with 3-methylenephthalide occurs according to a classical scheme.However, an original feature of the adducts is their electrooxidative conversion, which occurs according to an electrocatalytic mechanism. Further investigations are currently being done in this area. References 1 R. Fihi, K. Ciamala, J. Vebrel and N. Rodier, Bull. Soc. Chim. Belg., 1995, 104, 55. 2 K. C.Liu and R. K. Howe, J. Org. Chem., 1983, 48, 4590. 3 J. March, in Advanced Organic Chemistry, J. Wiley and Sons, New York, 4th edn., 1992, ch. 14, pp. 729»730. 4 H. J. Schaé fer, T op. Curr. Chem., 1990, 152, 90. 5 A. A. Akhrem, F. A. Lakhvich, V. A. Khripach and I. B. Klebanovich, T etrahedron L ett., 1976, 44, 3983. 6 A. Barco, S. Benetti, G. P. Pollini and P. G. Baraldi, Synthesis, 1977, 837. 7 A. Barco, S. Benetti, G. P. Pollini, P. G. Baraldi, B. Veronesi, M. Guarneri and G. B. Vincentini, Synth. Commun., 1978, 8, 219. New J. Chem., 1999, 23, 989»992 9918 G. Bianchi and M. De Amici, J. Chem. Res (S), 1979, 9, 311. 9 G. Menozzi, P. Schenone and L. Nosti, J. Heterocycl. Chem., 1983, 20, 645. 10 C. Grundman and R. Richter, J. Org. Chem., 1967, 32, 2308. 11 R. Huisgen and N. Mack, T etrahedron L ett., 1961, 17, 583. 12 K. C. Liu, B. R. Schelton and R. K. Howe, J. Org. Chem., 1980, 45, 3916. 13 R. H. Wiley and B. J. Wake–eld, J. Org. Chem., 1960, 25, 546. 14 Y. H. Chiang, J. Org. Chem., 1971, 36, 2146. 15 D. Garreau and J. M. Saveç ant, J. Electroanal. Chem., 1972, 35, 309. Paper 9/04758A 992 New J. Chem., 1999, 23, 989»992

ISSN:1144-0546    

DOI:10.1039/a904758a

出版商:RSC    

年代:1999

数据来源: RSC

摘要:

Electrochemically induced chain reactions. Reactions of aromatic aldehydes induced by electrogenerated bases Michaeé l Bernard,a Dominique Lucas,a Bernard Hanquet,a Yves Mugniera,* and Jean Lessardb,* a L aboratoire de et (CNRS UMR 5632), Synthe` se dœElectrosynthe` se Organomeç talliques Faculteç des Sciences Gabriel, de Bourgogne, 21000 Dijon, France. Universiteç E-mail : Y ves.Mugnier=u-bourgogne.fr b L aboratoire dœElectrochimie et dœElectrocatalyse, de Chimie, de Deç partement Universiteç Sherbrooke, Sherbrooke, QUE, J1K 2R1, Canada. E-mail : JL essard=courrier.usherb.ca.Received (in Montpellier, France) 19th May 1999, Accepted 3rd August 1999 The electrochemical reduction of aromatic aldehydes (ArCHO) such as 2-naphthaldehyde (1), 4-dimethylaminobenzaldehyde (2), 3-methoxybenzaldehyde (3) and 2-pyrrole carboxaldehyde (4) in N,Ndimethylformamide with tetrabutylammonium hexa—uorophosphate as supporting electrolyte, under an argon atmosphere and in the presence of protons donors such as —uorene or indene gives products (AH2) (FlH2) (InH2), such as (ArCH2A) and/or (AH\FlH or InH).Three consecutive chain reactions are initiated by [ArCH(AH)2] proton abstraction from —uorene or indene by a base electrogenerated by the reduction of ArCHO.The propagation steps of the –rst chain reaction involves addition of AH~ to ArCHO and protonation of the resulting alcoholate with regeneration of AH~ to give the carbinol ArCH(OH)AH. The second chain reaction is the base-catalyzed dehydration of ArCH(OH)AH to ArCH2A. The third is the addition of AH~ to ArCH2A, giving with regeneration of AH~.ArCH(AH)2 Reç actions en chaï� ne induites e ç lectrochimiquement. Re ç actions dœaldeç hydes aromatiques induites par des bases e ç lectrogeç neç reç es. La reç duction eç lectrochimique dœaldeç hydes aromatiques (ArCHO) tels que le 2-naphtaldeç hydes (1), le 4-dimeç thylaminobenzaldeç hyde (2), le 3-meç thoxybenzaldeç hyde (3) et le 2-pyrrolecarboxaldehyde (4) en preç sence dœhexa—uorophosphate de teç trabutylammonium comme eç lectrolyte support sous lœin—uence de donneurs de protons tels que le —uore` ne ou lœinde` ne conduit aux composeç s ArCH2A et (ou) (AH\FlH AH2 (FlH2) (InH2) ArCH(AH)2 ou InH).Trois reç actions en chaï� ne conseç cutives sont initieç es par abstraction dœun proton de par la base AH2 eç lectrogeç neç reç e par reç duction de ArCHO.Lœeç tape de propagation de la premie` re reç action en chaï� ne correspond a` lœaddition de lœanion AH~ sur lœaldeç hyde ArCHO suivie de la protonation de lœalcoolate reç sultant par pour AH2 conduire a` lœalcool ArCH(OH)AH avec reç geç neç ration de AH~. La seconde reç action en chaï� ne correspond a` la deç shydratation de ArCH(OH)AH en ArCH2A sous lœin—uence de la base AH~; en–n la formation de ArCH(AH)2 sœeÜectue par addition de AH~ sur ArCH2A suivie dœune protonation par AH2 .Reaction of organic products in suitable medium aÜords highly basic and/or nucleophilic reagents that may –nd some use in organic synthesis. Particularly, electrogenerated bases have been reported to induce chain reactions.The addition of geminal polyhalocarbanions to activated ole–ns,1 the reaction of oxygen with electrogenerated —uorenyl anion to give —uorenone, 2 the electrocatalyzed elimination of methanol from 9-methoxybi—uorenyl,3 and Michael reactions4 are representative examples. In the latter case the donor Don-H containing an acidic hydrogen atom is deprotonated by the base giving the anion Don~, which adds to the acceptor, for example an activated ole–n.The resulting anionic adduct deprotonates a new molecule of Don-H with regeneration of Don~. The base can be generated electrochemically using as probase the acceptor itself4 or an auxiliary substance like azobenzene.5 The base only needs to be present in very small amounts and accordingly the process requires only a catalytic quantity of electricity.A similar mechanism has been recently postulated in the case of the addition of indene or —uorene to (InH2) (FlH2) nitrosobenzene;6,7 the base corresponds to the anion radical of nitrosobenzene which is protonated by the proton ArNO~~, donor or to give AH~. Addition of AH~ to AH2 (InH2 FlH2) nitrosobenzene yields the anionic derivative Ar(NO)(AH)~, which is also protonated by to complete the cycle. AH2 Finally, Ar(NOH)(AH) is converted to the corresponding imine by dehydration at low temperature.More recently, we have found that the same mechanism was involved with an aromatic aldehyde instead of nitrosobenzene. 8h10 Three consecutive chain reactions occur under the in—uence of the electrogenerated base The –rst ArCHO~~.step is the formation of the corresponding carbinol ArCH(OH)(AH) (AH\FlH or InH), the second the formation of ArCH2A and the third the addition of or FlH2 InH2 to the latter to give There is much interest in ArCH(AH)2 . –nding new synthetic routes to the latter bis(indenyl) or bis(—uorenyl) methanes since they may serve as starting ligands to prepare the corresponding ansa-metallocenes.11 The alkenyl products (ArCH2A), which can be regarded as substituted fulvenes, can themselves be used in the synthesis of the bridged metallocenes.12 In recent years the main focus has been directed towards such complexes in the Group 4 series due to their polymerization catalytic properties.13,14 It must be noted that or can add to aldehydes in InH2 FlH2 the presence of a basic catalyst (potassium hydroxide,15,16 sodium hydroxide,17,18 or benzyltrimethyl KF»Al2O6 19 hydroxide-triton B 4).Nevertheless, the studies have been New J. Chem., 1999, 23, 993»999 993 This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scienti–que 1999 (limited to a few aldehydes only. Moreover, the yields were found to depend on the kind of base as well as its ratio.By using the electrochemical method the quantity of catalyst is limited to what is needed to complete the reaction since the initiator is generated from controlled potential electrolysis of the aldehydic reactives.10 In addition, some kinetic and mechanistic information can be obtained from the quantity of electrocatalyst, which is itself directly related to the charge consumed.All these reasons prompted us to try to evaluate the scope of the electroinduced reaction for the synthesis of ArCH(AH)2 and we present in this paper the results of a study on the electrochemical reduction of 2-naphthaldehyde (1), 4- dimethylaminobenzaldehyde (2), 3-methoxybenzaldehyde (3) and 2-pyrrole carboxaldehyde (4) in the presence of —uorene or indene The reaction is found to be selective : (FlH2) (InH2).the additional organic function in 2, 3 and 4 is not aÜected by the chain process. Results and discussion Electrochemistry and reaction mechanisms Fig. 1, curve a, shows the cyclic voltammogram of naphthaldehyde in DMF»0.2 M at room temperature. n-Bu4NPF6 Reduction peak A is observed in the cathodic sweep (Ep\ Fig. 1 Cyclic voltammogram of 1 in at room tem- DMF»Bu4NPF6 perature : sweep rate 0.1 V s~1, starting potential ]0.4 V. (a) 1 alone. (b) In the presence of 10 equiv. of FlH2 . V vs. SCE). If the potential is reversed after peak A, [1.8 three oxidation peaks are found: peak A@ V), (Ep\[1.70 peak V) and peak V). Peak A A2 @ (Ep\[0.5 A3 @ (Ep\[0.15 corresponds to the reduction of 1 to its anion radical and peak A@ to the oxidation of this anion radical.Peaks and A2 @ can be reasonably attributed to the oxidation of A3 @ (Np\2-naphthyl), which is obtained from (NpCHO)22~ dimerization of NpCHO~~.20 In the presence of 10 equiv. of (Fig. 1, curvek A FlH2 is not modi–ed, oxidation peaks A@ and disappear and a A3 @ new oxidation peak V) appears. The latter A1 @ (Ep\[0.72 corresponds to the oxidation of the conjugate base FlH~, which is formed from deprotonation of by either FlH2 or On the other hand, at low tem- NpCHO~~ (NpCHO)22~.perature ([20 °C) and in the presence of 10 equiv. of FlH2 (Fig. 2), peak A decreases and has a diÜerent shape (more rapid drop of current after reaching the peak potential). This is characteristic of a catalytic reaction occurring out of the diÜusion layer,7 a reaction that may be described as follows : NpCHO 1 ]e~]NpCHO~~ (1) NpCHO~~]FlH2 ]NpCHOH~]FlH~ (2) produced in reaction (1) is protonated by NpCHO~~ FlH2 , forming anion FlH~ and neutral radical Alterna- NpCHOH~.tively, the dianionic dimer may also deproton- (NpCHO)22~ ate FlH~ diÜuses towards the bulk of the solution and FlH2 .comes in contact with ArCHO, which moves to the electrode where reaction (3) takes place : NpCHO]FlH~HNpCHO(FlH)~ (3) Interestingly, such a chain reaction does not occur at room temperature, at least on the time scale of the cyclic voltammetry, since the intensity of peak A is unmodi–ed by adding reaction (3) must be slower. It has long been FlH2 : demonstrated15 that temperature has a determining in—uence on this type of reaction : raising the temperature shifts equilibrium (3) to the left.The chain reaction also does not occur on the coulometry time scale. Indeed, at room temperature, under the same conditions as the voltammetric studies, the current dropped to zero after the consumption of one electron per molecule of 1. This is in contrast with the electroreduction of benzaldehyde, 2,6-dichlorobenzaldehyde and terephthalaldehyde in the presence of —uorene or indene for which the consumption of electricity was catalytic (10~3 to 10~2 equiv.of electrons per molecule of aromatic aldehyde).9,10 Thus, the addition of Fig. 2 Cyclic voltammogram of 1 in at [20 °C: DMF»Bu4NPF6 sweep rate 0.1 V s~1, starting potential ]0.4 V.(a) 1 alone. (b) In the presence of 10 equiv. of FlH2 . 994 New J. Chem., 1999, 23, 993»999FlH~ to naphthaldehyde must be much slower than its addition to the previously studied benzaldehydes and terephthalaldehyde. In contrast, when the electrolysis of 1 in the presence of excess was performed as described above, except that FlH2 it was stopped after a charge corresponding to 2.8]10~2 equiv.of electrons per mol of 1 had been consumed and the electrolysis solution was then kept at room temperature under argon for 70 h, products [9-(2-naphthylmethylene)-9H- 1a1 —uorene] and M9-[9H-9-—uorenyl(2-naphthyl)methyl]-9H- 1a2 —uoreneN (see Scheme 1) were isolated in 17% and 41% yield, respectively (Table 1, entry 1), and identi–ed by 1H NMR spectroscopy, mass spectrometry and microanalysis.The formation of can be explained by equilibrium (3) followed by 1a1 reaction (4) and (5) : NpCHO(FlH)~]FlH2 ]NpCH(OH)(FlH)]FlH~ (4) NpCH(OH)(FlH)]NpCH2Fl 1a1 ]H2O (5) Reaction (5) consists in the dehydration of the intermediate carbinol NpCH(OH)FlH under the in—uence of FlH~, as we have previously mentioned.10 Scheme 1 New J. Chem., 1999, 23, 993»999 995Table 1 Controlled potential electrolysis of ArCHO in the presence of or on a mercury electrode in (0.2 M) at FlH2 InH2 DMF»n-Bu4NPF6 room temperature followed by a variable standing time Mol e~/mol Product(s) (% yield) [Ew a Molar ratio ArCHO Solution Standing Entry ArCHO /V AH2 AH2/ArCHO (]102) color time/h ArCH2A ArCH(AH)2 1 1 1.8 FlH2 15 2.8b Green 70 1a1(17) 1a2(41) 2 1 1.8 InH2 10 4.2c Grey 23 1b2(50) 3 2 1.95 FlH2 10 8b Red 23 2a1(50) 4 2 2.15 InH2 10 26c Green 20 2b1(3) 2b2(34) 5 3 1.85 FlH2 4 10b Blue 24 3a1(33) 3a2(39) 6 3 1.85 InH2 8 10b Purple 20 3b2(45) 7 4 1.9 FlH2 10 5b Green 48 4a1(32) 8 4 1.9 InH2 10 5b Purple 72 4b2(72) a Working potential vs.SCE. b Quantity of electricity imposed. c Quantity of electricity consumed by the time current had dropped to zero.NpCH2Fl is a highly conjugated p-electron system so that it is electrophilic enough to react with excess in a base- FlH2 catalyzed process : NpCH2Fl]FlH2½NpCH(FlH)2 1a2 (6) The –rst step is the addition of FlH~ to followed by 1a1, proton transfer from to the resulting anionic adduct and FlH2 release of FlH~, which is able to react again. With the aim to demonstrate the occurrence of reaction (6), compound was 1a1 reacted with a mixture of and FlH~: was obtained FlH2 1a2 in quantitative yield.As outlined above the overall process is rather slow compared with earlier studied aldehydes so that at least one of the propagation steps [reactions(1)»(6)] must be slower. We believe that the slower steps are the addition of FlH~ to 1 and to (NpCH2Fl) and not the base-catalyzed dehydration of 1a1 the carbinol.We already pointed out that the addition of FlH~ to 1 is slower than the addition to benzaldehyde and terephthalaldehyde.10 But the slowest step, for steric reasons, must be the addition of FlH~ to 1a1. These two addition steps proved to be faster with InH~, which is less bulky than FlH~, as the nucleophile.Indeed, when replacing by (entry 2 of Table 1), the current FlH2 InH2 dropped to zero after the consumption of a catalytic quantity of electricity (4.2]10~2 equiv. of electron per mol of 1), which shows that the addition of InH~ to 1 is faster than the corresponding addition of FlH~. The addition of InH~ on the alkenyl intermediate NpCH2In [2-(1H-3-indenylmethyl)naphthalene] is also faster than the addition of FlH~ to As 1a1.evidence we did not succeed in isolating this intermediate, owing to its short lifetime. The bis(indenyl) adduct was 1b2 obtained in 50% yield. can also be obtained from in the presence of a 1b2 1a1 mixture of InH~ and according to the global reaction : InH2 A possible mechanism for this reaction can be rationalized in terms of the following reaction scheme: NpCH2Fl 1a1 ]InH~ A88B NpCH(Fl)(InH)~ (7) NpCH(Fl)(InH)~ A88B transfer proton NpCH(FlH)(In)~ (8) NpCH(FlH)(In)~ A88B NpCH2In 1b1 ]FlH~ (9) NpCH2In]InH2 »»»’ NpCH(InH)2 1b2 (10) The –rst step consists in the nucleophilic addition of InH~ to to give the anionic intermediate NpCH(Fl)(InH)~ that, 1a1 after an intramolecular proton transfer, eliminates FlH~ to give which immediately reacts with to yield 1b1, InH2 1b2 .In addition, we have found that reacts easily with InH~ 1a2 to give 1b2: A mechanism can be postulated in which InH~ successively displaces the two —uorenyl moieties. This reaction involves two base-catalyzed elimination-addition sequences. The –rst elimination gives and [reaction (12)]. Then InH~ 1a1 FlH2 adds to [reaction (7)].The second sequence consists in 1a1 reactions (8), (9) and (10). NpCH(FlH)2 1a2 HNpCH2Fl 1a1 ]FlH2 (12) The cyclic voltammetric behavior of aldehydes 2, 3 and 4 at room temperature is similar to that of naphthaldehyde 1. There is no modi–cation of the reduction peak A upon adding or indicating that no chain reaction occurs on the FlH2 InH2 , voltammetric time scale at room temperature.The results of the electrolysis are given in Table 1, entries 3 to 8. The same types of products [ArCH2A and/or as those ArCH(AH)2] obtained with naphthaldehyde 1 were isolated. As in the case of the electrolysis of aldehyde 1 (entries 1 and 2), the reaction of FlH~ with ArCH2Fl is slower than the reaction of InH~ with ArCH2In. Indeed, with aldehyde 2, only MN- 2a1 [4-(9H-9-—uorenylidenemethyl)phenyl]-N,N-dimethylamineN (ArCH2Fl) was isolated after 23 h from the electrolysis in the presence of whereas MN-[4-(di-1H-3-indenylmethyl) FlH2 2b2 phenyl]-N,N-dimethylamineN was the main [ArCH(InH)2] product isolated after 20 h from the electrolysis in the presence of A similar observation can be made for the elec- InH2 . 996 New J. Chem., 1999, 23, 993»999trolyses of aldehyde 3 (compare entries 5 and 6) and of aldehyde 4 (compare entries 7 and 8).Structure of products To our knowledge all compounds are new compounds. They were characterized by 1H NMR spectroscopy (Table 2), mass spectrometry (Table 3), microanalysis (Table 4) and infrared spectroscopy (Table 5). Only one isomer of product MN-[4-(1H-1-indenylidene- 2b1 methyl)phenyl]-N,N-dimethylamineN was isolated.We have been unable to determine, on the basis of the NMR spectroscopic data, if it was the E or Z isomer, but the E isomer should be more easily formed than the Z isomer since it is less hindered. For product only one of the six most probable isomers 1b2 shown in Scheme 2 has been isolated. Isomers and are i1 i3 interconvertible through two 1,5-sigmatropic rearrangements21 or a base-catalyzed proton transfer22,23 and isomers and through base-catalyzed transfers.i4 i5 The same types of isomers can be drawn for 2b2 , 3b2 M3-[1H-3-indenyl(3-methoxyphenyl)methyl]-1H-indeneN and M2-(di-1H-3-indenylmethyl)-1H-pyrroleN but, as for 4b2 1b2 , only one product was formed. The 1H NMR spectra of all these compounds (1b2 , 2b2 , 3b2 and show that the indenyl moiety represented by InH 4b2) Table 2 1H NMR data Product d Multiplicity Integration J/Hz Proton assignment 1a1 8.11 br s 1H » Ole–nic 7.05»7.96 a 15H » Aromatic 1a2 3.41»3.49 t 1H 7.9 ArCH(AH)2 5.12, 5.16 d 2H 7.9 ArCH(AH)2 6.84»7.79 m » » Aromatic 1b2 3.53 br s 4H » Methylene 5.70 br s 1H » ArCH(AH)2 6.21, 6.22 ììdœœ 2H 1.5 Ole–nic 7.29»7.98 m » » Aromatic 2a1 3.05 s 6H » Dimethylamino 6.74»6.82 pert d 2H 8.9 p-Substituted benzene 7.54»7.61 pert d 2H 8.9 p-Substituted benzene 7.67 s 1H » Ole–nic 7.10»8.02 d and several m » » Aromatic —uorene 2b1 3.03 s 6H » Dimethylamino 6.70»6.97 pert d 2H 8.9 p-Substituted benzene 7.55»7.62 pert d 2H 8.9 p-Substituted benzene 7.43 s 1H » Ole–nic 6.97»7.71 dd and several m » » Indenyl 2b2 3.12 s 6H » Dimethylamino 3.59 br s 4H » Methylene 5.55 br s 1H » ArCH(AH)2 6.29, 6.30 ììdœœ 2H 1.3 Ole–nic 6.92»6.96 pert d 2H 8.9 p-Substituted benzene 7.40»7.72 two m » » Aromatic 3a1 3.84 s 3H » Methoxy 7.68 s 1H » Ole–nic 6.93»7.81 two m » » Aromatic 3a2 3.19»3.27 t 1H 7.9 ArCH(AH)2 3.37 s 3H » Methoxy 4.99, 5.03 d 2H 7.9 ArCH(AH)2 6.21»7.77 a » » Aromatic 3b2 3.43 br s 4H » Methylene 3.81 s 3H » Methoxy 5.40 br s 1H » ArCH(AH)2 6.11, 6.12 pert d 2H 1.6 Ole–nic 6.83»7.53 m » Aromatic 4a1 6.39»6.43 sex 1H 2.6 Pyrrolic 6.73»6.77 pert s 1H » Pyrrolic 6.90»6.94 sex 1H 2.7, 1.3 Pyrrolic 7.24»7.81 two m » » Aromatic 8.19, 8.23 pert d 1H 7.3 8.47 very br s 1H » NH 4b2 3.42 br s 4H » Methylene 5.49 br s 1H » ArCH(AH)2 6.11»6.14 pert s 1H » Pyrrolic 6.18»6.24 m 3H » Ole–nic and one pyrrolic 6.68»6.71 sex 1H 2.9, 1.6 Pyrrolic 7.19»7.53 two m » » Aromatic 8.04 very br s 1H » NH a Several diÜerent multiplicities (m, t, dd, .. .). New J. Chem., 1999, 23, 993»999 997Table 3 Mass spectral data m/z (% rel int) Product Molecular ion Other peaks 1a1 304(100) 165(2.2) » FlH 1a2 470(38.8) 305(100) 165(52.5) [FlH FlH 1b2 370(100) 255(72.1) 115(8.2) [InH InH 2a1 297(100) 252(16.4) [NH(CH3)2 » 2b1 247(100) 202(25.4) [NH(CH3)2 » 2b2 363(100) 248(54.1) 203(13.9) [InH [InH[NH(CH3)2 3a1 284(100) 253(40.2) [OCH3 3a2 450(20.3) 285(100) 165(14.5) [FlH FlH 3b2 350(100) 235(57.4) 115(12.3) [InH InH 4a1 243(100) 176(5) 163(7.2) [C4H5N [C5H6N 4b2 309(100) 242(16.5) 194(58.2) [C4H5N [InH exists in a form where the double bond is trisubstituted (cform) since a four-proton singlet in the 3.42»3.59 ppm range appears in all spectra and only structure is in agreement i3 with the 1H NMR data.In the mass spectra (Table 3) the molecular ion peak was observed for all derivatives. Compound M9-[3-methoxyphenyl)methylene]-9H- 3a1 —uoreneN was reported to be crystalline in the literature (mp Scheme 2 145»146 °C24). However, it was obtained as an oil and has been fully characterized by elemental microanalysis, 1H NMR spectroscopy and mass spectrometry.Table 4 Eluents for chromatographic separation, solvents for recrystallization, melting points and microanalytical data Microanalytical data C/% H/% N/% Chromatography Recrystallization Melting Molecular Product eluenta solvent point/°C formula Calcd Found Calcd Found Calcd Found 1a1 1 : 4 Hexane 98 C24H16 94.70 94.35 5.30 5.36 1a2 1 : 4 Hexane» 229 C37H26 94.43 94.19 5.57 6.07 CH2Cl2 1b2 1 : 4 Hexane» 152 C29H22 94.01 93.89 5.99 6.00 CH2Cl2 2a1 1 : 3 Hexane» 132 C22H19N 88.85 88.83 6.44 6.36 4.71 4.68 CH2Cl2 2b1 1 : 1 Hexane» 158 C18H17N 87.41 87.02 6.93 6.89 5.66 5.47 CH2Cl2 2b2 1 : 1 Hexane» 154 C27H25N 89.21 88.41 6.93 6.97 3.86 3.60 CH2Cl2 3a1 1 : 4 Oilb C21H16O 88.73 88.46 5.63 5.66 3a2 1 : 4 Hexane» 205 C34H26O 90.63 90.66 5.81 5.82 CH2Cl2 3b2 1 : 3 C2H5OH» 52»54 C26H22O 89.11 88.74 6.33 6.37 H2O 4a1 2 : 1 Hexane» 94 C18H13N 88.86 88.77 5.38 5.48 5.76 5.76 CH2Cl2 4b2 1 : 1 Hexane» 101 C23H19N 89.28 89.10 6.19 6.38 4.53 4.22 CH2Cl2 b Reported as crystalline in ref. 24. a CH2Cl2»hexane. 998 New J.Chem., 1999, 23, 993»999Table 5 Assignment of characteristic bands in the IR spectra of compounds to 1a1 4b2 Product l/cm~1 Intensity Assignment 1a1 3013 m Ole–nic C»H stretch 1a2 2892 w Aliphatic C»H stretch 1b2 2891 w Aliphatic C»H stretch 2a1 2805 m N»CH3, C»H stretch 2b1 2814 w N»CH3, C»H stretch 2b2 2805 m N»CH3, C»H stretch 3a1 2832 m O»CH3, C»H stretch 3a2 2832 m O»CH3, C»H stretch 3b2 2833 m O»CH3, C»H stretch 4a1 3406 s (sharp) N»H stretch 4b2 3419 s (sharp) N»H stretch Experimental The apparatus and techniques used have been described previously. 10 1H NMR spectra were taken on a Bruker 200 MHz spectrometer in Chemical shifts are reported in d CDCl3 . down–eld from internal tetramethylsilane. Infrared data were obtained on a Nicolet 205 spectrophotometer in the solid state using KBr as matrix. 2-Naphthaldehyde (1), 4-dimethylaminobenzadehyde (2) and 3-methoxybenzaldehyde (3) were commercially available. 2-Pyrrole carboxaldehyde (4) was synthesized according to a literature procedure.25 Typical electrolysis procedure p-Dimethylaminobenzaldehyde 2 (608 mg, 4.08 mmol) was electrolyzed on a mercury electrode at [2.15 V vs. SCE with indene (5.1 ml, 43.6 mmol) in DMF (30 ml) containing n-Bu4 (0.2 M) at room temperature (see Table 1, entry 3).The NPF6 electrolysis was stopped when the current dropped to zero (0.26 electrons per molecule of 2). The electrolyzed solution was stirred for 20 h (the color turned from red to green) then poured into water and extracted with dichloromethane. Chromatography on silica gel with hexane»dichloromethane (1 : 1) as eluent gave two products (less polar) 2b1 and in 3% and 34% yield, respectively.Compounds 2b2 2b1 and were characterized by 1H NMR spectroscopy (Table 2b2 2), mass spectrometry (Table 3), elemental microanalysis (Table 4) and IR spectroscopy (Table 5). Acknowledgements are grateful to M. T. Compain for her technical assistance. We We acknowledge –nancial support from lœElectriciteç de France.EDF (Novelect Bourgogne and Club dœElectrochimie Organique), the Ministe` re des Relations Internationales du Queç bec, the Fonds pour la Formation de Chercheurs et dœAide a` la Recherche (FCAR) du Queç bec and the Natural Sciences and Engineering Research Council of Canada. References 1 M. M. Baizer and J. L. Chruma, J.Org. Chem., 1972, 37, 1951. 2 K. J. Borhani and M. D. Hawley, J. Electroanal. Chem., 1979, 101, 407. 3 C. Nuntnarumit, F. M. Triebe and M. D. Hawley, J. Electroanal. Chem., 1981, 126, 145. 4 M. M. Baizer, J. L. Chruma and D. A. White, T etrahedron. L ett., 1973, 5209. 5 D. A. Monteç , M. M. Baizer and R. D. Little, J. Org. Chem., 1983, 48, 803. 6 Y. Mugnier, E. Laviron, Y. Huang and J.Lessard, J. Electroanal. Chem., 1991, 297, 529. 7 J. C. Gard, Y. Mugnier, J. Lessard, Y. Huang, Y. Couture and A. Lasia, J. Org. Chem., 1993, 58, 5329. 8 J. C. Gard, B. Hanquet, Y. Mugnier and J. Lessard, J. Electroanal. Chem., 1994, 365, 299. 9 J. C. Gard, B. Hanquet, Y. Mugnier, J. Lessard and L. Roullier, Can J. Chem., 1996, 74, 55. 10 J. C. Gard, F. Williot, M. Bernard, P.Richard, M. Kubicki, D. Lucas, Y. Mugnier and J. Lessard, New J. Chem., 1997, 21, 929. 11 (a) I. E. Nifantœev, P. V. Ivchenko, L. G. Kuzœmina, Y. N. Lizikov, A. A. Sitnikov and O. E. Sizan, Synthesis, 1997, 469; (b) W. Spaleck, M. Antberg, V. Dolle, R. Klein, J. Rohrmann and A. Winter, New J. Chem., 1990, 14, 449; (c) D. T. Mallin, M. D. Rausch, Y. Lin, S. Dong and J. C. V. Chein, J. Am. Chem. Soc., 1990, 112, 2030. 12 J. J. Eisch, X. Shi and F. A. Owuor, Organometallics, 1998, 17, 5219 and references therein. 13 W. Kaminski, K. Kulper, H. H. Brintzinger and F. R. W. P. Wild, Angew. Chem., Int. Ed. Engl., 1985, 24, 507. 14 H. H. Brintzinger, D. Fischer, R. Mulhaupt, B. Rieger and R. M. Waymouth, Angew. Chem., Int. Ed. Engl., 1995, 34, 1134. 15 E. Ghera and Y. Sprinzak, J. Am. Chem. Soc., 1960, 82, 4945. 16 G. Kresze, H. G. Henkel and H. Goetz, Ann. Chem., 1964, 18, 674. 17 M. AvranoÜ and Y. Sprinsak, J. Am. Chem. Soc., 1960, 82, 4953. 18 A. Sieglitz, Ber. Dtsch. Chem. Ges. B., 1919, 52, 1513. 20 M. M. Baizer, in Organic Electrochemistry, ed. H. Lund and M. M. Baizer, Marcel Dekker, New York, 1991, ch. 22, p. 879. 21 (a) D. J. Field, D. W. Jones and G. Kneen, J. Chem. Soc., Chem. Commun., 1975, 754; (b) D. J. Field, D. W. Jones and G. Kneen, J. Chem. Soc., Perkin T rans. I, 1978, 1050. 22 C. Courtot, C. R. Acad. Sci. Paris, 1915, 160, 523. 23 D. B. Taggart and E. C. Friedrich, J. Org. Chem., 1975, 40, 720. 24 R. S. Tewari, K. Nirmal and P. S. Kendurkar, J. Indian Chem. Soc., 1977, 54, 443. 25 (a) G. F. Smith, J. Chem. Soc., 1954, 3842; (b) R. M. Silverstein, E. E. Rtskiewicz, C. Willard and R. C. Koehler, J. Org. Chem., 1955, 20, 668. Paper 9/04036F New J. Chem., 1999, 23, 993»999 999

ISSN:1144-0546    

DOI:10.1039/a904036f

出版商:RSC    

年代:1999

数据来源: RSC