|
1. |
Failure of the antiperiplanar lone pair hypothesis in glycoside hydrolysis. Synthesis, conformation, and hydrolysis of α-D-xylopyranosyl-and α-D-glucopyranosyl-pyridinium salts |
|
Journal of the Chemical Society, Perkin Transactions 2,
Volume 1,
Issue 7,
1984,
Page 1121-1131
Lynn Hosie,
Preview
|
PDF (1246KB)
|
|
摘要:
J. CHEM. SOC. PERKIN TRANS. II 1984 Failure of the Antiperiplanar Lone Pair Hypothesis in Glycoside Hydrolysis. Synthesis, Conformation, and Hydrolysis of a-D-Xylopyranosyl- and a-D-Gluco- pyra nosy1 -pyridi n ium Salts Lynn Hosie, Philip J. Marshall, and Michael L. Sinnott Department of Organic Chemistry, University of Bristol, CantockS Close, BristolBS8 I TS (1) Two series of crystalline a-o-xylopyranosyl- and a-o-glucopyranosyl-pyridinium bromides have been made, by reaction of the acetylated a-bromides with the pyridines in an aprotic solvent in the presence of bromide ion, followed by deacetylation with aqueous HBr at room temperature. (2) 200 MHz H N.m.r. spectra in D,O of representatives of each series show the xylo-compounds to adopt the C, conformation, and the gluco-compounds the S, conformation.(3) The hydrolyses of a-D-1.OM-NaCIO,xylopyranosyl-and a-o-glucopyranosyl-3- bromopyridinium ions in I 0-' + 1.4 x 1O-"/[ at 25.0 "c are = 2 X 1o-' '3Hdescribed = 4.7 and +by kObs/S-l kobs/S-'X 1O--' '/[H'], respectively. (4) The products of pH-independent hydrolysis of the foregoing ions are the sugars and 3-bromopyridine, whereas at pH 12, 10%of both reactions proceed by attack on the pyridine ring: attack on the sugar yields glucose and 1,6-anhydroglucopyranose, but no xylose. (5) The pH -independent hydrolyses display strongly positive entropies of activation: at 25.0 "C, those of five xylo-compounds give a pig value of -1.27 0.06, and those of four gluco-compounds one of -1.06 0.1 2.(6) Arguments are presented, inthe light of the low a:j3 rate ratios for bothxylo and gluco pyridinium salts (8-23 and 80, respectively, at 25 "C), that departure of a pyridine from C(1) of an aldopyranose ring does not require a conformation in which the leaving group is antiperiplanar to a lone pair of electrons on O(5). (7) The antiperiplanar lone pair hypothesis is shown to be a special case of the principle of least nuclear motion, which is known not to apply to reactions with very product- (or 1.4 x reactant-) like transition states. The idea that departure of a leaving group from a tetrahedral carbon atom, substituted with one or more heteroatoms, is favoured in an additive way by the presence of antiperiplanar lone pairs of electrons on the heteroatoms and has achieved some currency of late.This antiperiplanar lone pair hypothesis was originally developed to explain products formed at the acyl level of oxidation,' as was its theoretical rationale.* There are a number of difficulties r+th even its original formulation:' results from the hydrolysis of orthoesters were interpreted on the assumption that the lifetime of a tetrahedral intermediate was short compared with the time of rotation about a carbon- oxygen single bond. In fact hydrogen orthoesters have been directly observed and their lifetimes mea~ured,~~~ and the antiperiplanar lone pair hypothesis questioned on these ground~.~Studies on the hydrolysis of an E-2 pair of imidates showed that likewise the lifetime of a tetrahedral intermediate containing nitrogen was long compared with the time of rotation about a C-0 or C-N bond.' Perrin and Arrhenius6 analysed the logic behind Deslongchamps' work, and con-cluded that it contained no unambiguous evidence for the antiperiplanar lone pair hypothesis.One of the crucial findings in the formulation of the hypothesis, the supposed non-production of 6-valerolactone from hydrolysis of orthoesters (Ia and b), has been shown to be experimentally wrong.7 In the gas phase, there is no preference for the loss of a methoxy group from the axial epimer of orthoester (II).' Kinetic studies of the acid-catalysed hydrolysis of system (Ic) have, however, shown a very modest (2-11 fold), solvent- dependent preference for departure of axial meth~xy.~ At the aldehyde level of oxidation, in simple systems the antiperiplanar lone pair hypothesis lacks the support of even modest kinetic effects in the right direction: in systems (111),7 and (IV) O3 the equatorial leaving group is lost the faster.This faster departure of the equatorial leaving groups could be attributed with some plausibility to reaction through boat conformers, which do indeed have antiperiplanar lone pairs of electrons." An elegant series of investigations by Kirby and co-OMe OR' 01 workers attempted to circumvent the conformational ambiguity inherent in systems such as (111),and demonstrated dramatic effects in system (V) l3 (after correction for ion-pair return), (VI)14 (where the reaction is formally a fragmentation), and (VII) (where an 'anti-Bredt' oxocarbonium ion is generated).The problem is clearly the design of an atomic framework sufficiently rigid to restrict conformational preferences, but not so constraining as to introduce uncertainties into the choice of a suitable reference reaction. We addressed the problem, in a complementary fashion, to Kirby et al. by using a leaving group which was also a powerful conformational determinant, viz., pyridine. The bulk and, more importantly, the reverse anomeric effect,* constrains the leaving ~~~ ~ * This effect, as originally defined," is the additional tendency of positively charged substituents at C(1) of a pyranose ring to adopt the equatorial orientation.It cannot be accounted for by the frontier orbital picture of the ordinary anomeric effect,I6 and has since been re-defined on theoretical grounds as the additional tendency of groups more electropositive than hydrogen in a Q (S. David, 0. Eisenstein, W. J. Hehre, L. Salem, and R. Hoffmann, J. Am. Chem.Soc., 1973,%, 3806) or than oxygen in a x (S. Wolfe, M. H. Whangbo, and D. J. Mitchell, Carbohydr. Res. 1979,69, 1) sense to adopt the equatorial orientation. The effect as so re-defined is wholly without experimental foundation. The experimentally manifested effect is undoubtedly electrostatic in nature, and has nothing to do with any ground-state antiperiplanar lone pair effects. J. CHEM. SOC. PERKIN TRANS. 11 1984 /-\ i-l HO HO A (IV) a; X = HZPO[(ref.lO) b; X = F (ref.11) H H H (VII) group in the ground state to the equatorial orientation. (Studies with acetylated a-D-glucopyranosyl- l7 and a-D-xylopyranosyl- imidazoles that change their conformational preference on protonation, establish conclusively that the reverse anomeric effect is both powerful, and not trivially caused by the bulk of the pyridine ring.) The ground-state conformational require- ments of a pyridinium leaving group, therefore, are directly opposite to the requirements imposed by the antiperiplanar lone pair hypothesis for its departure. By working with aldopyranose derivatives in aqueous solution, for which there exist a reasonably reliable set of empirical conformational energies,' it is possible to make predictions as to what the ratio of S,l hydrolysis of anomeric pairs of aldopyranosylpyridinium salts would be, if the antiperiplanar lone pair hypothesis were correct.The ground- state conformational preference of the pyridine leaving group is exactly opposite to the supposed requirement for its departure. Therefore we assume that the most accessible conformation of the pyranose ring of a pyridinium salt that has a lone pair of electrons antiperiplanar to the leaving pyridine is the ground- state conformation of the pyranose ring in the pyridinium salt of opposite anomeric configuration. The further advantage of carbohydrate systems is that they are directly relevant to biological glycosyl-transferring systems, which have been subjected to theoretical consideration in the light of the antiperiplanar lone pair hypothesis." A dis-advantage of using aldopyranosyl derivatives is that in ethanol or trifluoroethanol certainly,* the intermediate oxocarbonium ions are too unstable to exist 22 and the reactions are therefore bimolecular. However, in water, although there are salt effects on the hydrolysis of glycosyl derivatives, these parallel basicity rather than the nucleophilicity of the salt.Rates in 1.0~salt, for instance, are in the order NaF > NaCl > NaI. Moreover, the neutral nucleophile thiourea has no effect.l2 This provides reassurance that the reactions are indeed unimolecular in water, even if only just so; therefore the trajectory of approach of a nucleophile does not have to be considered.Experimental Tri-0-acetyl-a-D-xylopyranosylpyridinium Bromides.-Generally, tri-0-acetyl-a-D-xylopyranosylbromide (5 g) was dissolved in a mixture of dry acetone (10 ml), dry m-cresol (1.0 ml), and the pyridine (5 g) and allowed to stand overnight at room temperature. The product was isolated by pouring the reaction mixture into ether. Yields, characterisation data, and recrystallisation solvents are given in Table 1. The conditions were slightly altered for the unsubstituted pyridine [glycosyl halide (2.0 g), acetone (75 ml), and pyridine (25 ml); product crystallised out of the reaction mixture] and nicotinamide [glycosyl halide (5 g), dry nicotinamide (2.0 g), in dry acetone (125 ml); after 3 h the abundant precipitate of nicotinamide hydrobromide was filtered off, and the mixture allowed to stand overnight at room temperature: on lowering the temperature to 4 "C the product crystallised].Tri-0-acetyl-p-D-xylopyranosylpyridinium Bromides.-Generally, the pyridine (3 g) was warmed with tri-0-acetyl-a-D- xylopyranosyl bromide (5 g) and m-cresol(4.5-6.5 ml) until a homogeneous melt was obtained, then left overnight at room temperature. The product was isolated by pouring the reaction mixture into ether. Yields, characterisation data, and recrystallisation solvents are given in Table 1. However, the unsubstituted pyridine compound was prepared by reaction of the glycosyl halide (2.0 g) with pyridine (25 ml) in dry etbanol (75 ml) at room temperature. Tetra-0-acetyl-a-D-glucopyranosylpyridinium Bromides.-These were prepared by the method of Lemieux and Morgan 23 for the 4-methylpyridinium salt, and were isolated by pouring into ether.Yields, characterisation data, and recrystallisation solvents are given in Table 1. De-0-acetylation of Labile (Largely a) Glycopyranosyl- pyridinium Salts.-The fully acetylated compounds (1.O g) were dissolved in 8% aqueous hydrogen bromide (10 ml) and left at room temperature in the dark for 6 days. Any small amount of insoluble material was removed by filtration through Celite, and the hydrogen bromide was removed from the filtrate by extraction with a 33% (v/v) solution of tri-n-octylamine in dichloromethane until the pH of the aqueous layer increased to between 5 and 6 (no higher).The aqueous layer was then extracted with dichloromethane (3 x 10 ml), and was evaporated to a pale yellow syrup. Portions of dry ethanol (3 x 10 ml) were evaporated from the syrup to give a very stiff white gum which, on trituration with a 3: 1 (v/v) mixture of ethanol and methanol (3.5 ml) gave a crystalline product. The comparatively stable 2,3,4-tri-O-acetyl-P-~-xylopyrano-sylpyridinum salt was taken to pH 8 after deacetylation: in this case the acetate salt of the p-D-xylopyranosylpyridinium ion was isolated. De-0-acetylation of some of the more robust tri- 0-acetyl-m- and p-xylopyranosylpyridinium salts was achieved by the method of Lemieux and Morgan.23 Kinetic Measurements.-The changes in absorbance were followed at the wavelengths of maximum absorbance for the stated aglycones: pyridine and 4-methylpyridine, 265 nm; 4- bromoisoquinoline, 344 nm; 3-~arbamoylpyridine, 275 nm; and 3-bromopyridine, 280 nm.Runs below 80 "Cwere followed in 1 cm cells in the thermostatted cell block of a Unicam SP 1800 spectrometer system, the temperature of the cell block being kept constant by a Techne C-100closed-system water-bath. The temperature in the cell was measured by means of Radiospares miniature bead thermistors enclosed in glass capillaries (sealed at one end), inserted through holes drilled in the Teflon stoppers of the cells and glued in with epoxy resin.The thermistor resistances were calibrated by immersing them in a water-bath J. CHEM. SOC. PERKIN TRANS. II 1984 maintained at a constant temperature by a circulating thermostatic pump. At temperatures above 90 "C, the reactions were followed by immersing sealed ampoules in a Tamson thermostatic oil-bath, withdrawing them at suitable invervals, cooling, and taking the optical density reading on a Unicam SP 1800 or Perkin-Elmer 5 5 5 spectrometer. In general, rate constants were calculated from linear least- squares treatments of ln(A -A,) versus time plots using experimental infinity points. However, degradative browning reactions at high pH or above 90°C made these values unreliable, so in these cases A, values calculated from Kezdy- Swinbourne 24 plots were used.All data pertain to 5(hn~-buffer solutions in 1.0M-sodium perchlorate: pH values were measured at room temperature. The following buffer systems were used: pH 3.5-5.5, acetic acid- sodium hydroxide; pH 5.5-8.5, disodium hydrogen phosphate- sodium dihydrogen phosphate; pH 8.5-10.5, EDTA-sodium hydroxide; pH 1 1-1 2.5, disodium hydrogen phosphate-sodium hydroxide; pH 12.95 was obtained using sodium hydroxide alone. Buffers were made up to the pH values and measured with a standard combination electrode, used in conjunction with a Radiometer PHM 62 pH meter, which had been freshly calibrated with BDH standard buffers. Note on the Use of Standard Combination Electrodes and NaCIO, Solutions.-After a period of time a standard combination pH electrode immersed in 1 .OM-sodium perchlorate will become sluggish, owing, no doubt, to the precipitation of potassium perchlorate in the porous plug of the potassium chloride bridge.On soaking in dilute buffer, a quick response is re-established, and the electrode, when calibrated with standard buffer, reads correctly in dilute solutions and in 1 .OM-sodium perchlorate. Insidiously, however, in 1 .Owpotassium chloride solutions the electrode can now read up to 0.8 pH units too low. H Nmr. Spectra.-For the determination of conformations, these were measured on a JEOL 200 MHz Fourier transform instrument in D20 with 3-trimethylsilylpropane-1-sulphonate as internal standard. Signals were assigned by irradiation at H(l), H(2), H(3), and H(4) in the case of a-~-xylopyranosyl-4- methylpyridinium ion, H( 1) and H(5) (pro-R) in the case of P-D-xylopyranosylpyridinium ion, and at H( 1) and H(2) in the case of the a-glucosylpyridinium ions.The H(3), H(4), H(5), H(6), H(6') region of a-~-g~ucopyranosy~-4-bromoisoquinoliniumion was simulated using the NUMARIT program. Analysis of inversion :retention ratios in the reaction of acetylglyco-pyranosyl bromides with pyridine were made on a JEOL PMX60 continuous wave instrument. Product Analyses.-The glycones were analysed by g.1.c. of their trimethylsilyl ethers using 1,6-anhydrogalactose as an internal standard for products from glucosylpyridinium salts and 1,6-anhydro-~-glucose as an internal standard for xylosylpyridinium salts.Analyses were performed on a 2 m column of 5% Dexsil 300SC (on 100-20 mesh Gas Chrom Q) at 170 "C for xylose and 190 "C for glucose. Relative retention times were 1,6-anhydrogalactose:1,6-anhydro-glucose :a-glucose :j3-glucose, 1:1.26:1.97:2.96, and 1,6-anhydroglucose :a-xylose:P-xylose, 1: 0.84: 1.16; the number of theoretical plates was 1 520 (P-glucose) and 1 280 (P-xylose). Portions (5 ml) of a solution of the pyridinium salt in the appropriate buffer were heated for 6 half-lives at 100"C, neutralised (where appropriate), and extracted with ether (3 x 2 ml). The ethereal solution was blown down with nitrogen, the residue taken up in pH 6.0 buffer, and the pyridine estimated from its U.V. absorbance.To the aqueous layer was added an internal standard, the solution was shell-frozen and 1123 freeze-dried, and then derivatised for g.1.c. with a pyridine-hexa- methyldisilazane-trimethylchlorosilane mixture (5 :1:2 v/v). Aglycones liberated in alkaline solution were also analysed without ether extraction. Results and Discussion Reaction of Acetylglycopyranosyl Bromides with Pyridines.- This work partly has its origin in our finding that conditions for reaction of 2,3,4,6-tetra-0-acetyl-a-~-galactopyranosyl~~ 26or 2,3,4-tri-O-acetyl-P-~-arabinopyranosylbromides with pyridines, which gave pyridinium salts of inverted anomeric configuration, when applied to 2,3,4-tri-O-acetyl-a-D-xylo-pyranosyl bromide, gave 2,3,4-tri-O-acetyl-a-~-xylopyranosyl-pyridinium salts.A previous, related, preparative observation had been that 2,3,4,6-tetra-0-acetyl-P-~-glucopyranosy1-4-bromoisoquinolinium bromide, prepared in the same way as its C(4) e~imer,~~ was, as a crude material, contaminated with the a-an~mer.~'Some quantitation of these preparative impressions can be obtained from the inversion :retention ratio, as measured by n.m.r. integration of the anomeric hydrogen resonance, of the pyridinium salt products formed when acetylglycopyranosyl bromide (0.2 g) was dissolved in dry deuteriopyridine (0.5 ml) and left overnight at room temperature: a-D-galacto, 8; a-D-gluco, 3.6; a-D-XylO, 0 and P-L-arabino, 3. The basic proposition, advanced by Lemieux and Morgan to explain their results in the glu~o-series,~ that apparently retained pyridinium salts arise from SN2 reactions of the pyridine on inverted acylglycopyranosyl bromide, itself the product of SN2attack by bromide ion on the original halide, is clearly correct.The halide-ion-catalysed anomerisation of glycosyl halides, discovered in that in~estigation,~~was subsequently used to develop one of the few successful and widely applied a-glycoside syntheses2' This synthesis, and the original reactions with pyridine, have been rationalised in terms of a complex reaction pathway, under stereoelectronic control, involving ion pairs and ion triplet^.^' It is, however, difficult to reconcile this pathway, complex though it is, with either the qualitatively different results obtained with acylglycopyranosyl bromides epimeric at C(4), or with recent evidence against intimate ion pairs as intermediates in glycosyl transfer reactions.2' We now suggest that our results, and those of Lemieux and Morgan,23 can be rationalised by a series of classical sN2 reactions, on two supplementary hypotheses.The first is that 'naked' bromide ion is a better nucleophile than a pyridine unless its approach is opposed by the negative end of the dipole of an acyloxy group, when the neutral nucleophile is favoured. The second is that the anomerised, thermodynamically less stable glycopyranosyl halides are the more susceptible to nucleophilic attack. Our proposals are set out in Scheme 1. Initial nucleophilic attack by pyridine on starting bromide (VIII) gives 'inverted' pyridinium ion (IX) and bromide ion, this last serving to anomerise (VIII) to (X) rapidly unless opposed by an axial acetoxy group at C(4).(X) reacts competitively with bromide ion to re-form (VIII) and pyridine to form (XI): because of opposition from the acetoxy dipole on C(2), pyridine is a better nucleophile than bromide ion: this is particularly pronounced in the xylose case since it is likely that tri-0-acetyl-P-D-xylopyranosyl bromide, like the ~hloride,~' adopts the 'C4 conformation as shown. Addition of bromide ion thus favours production of compound (X)and hence pyridinium salt (XI); addition of hydrogen-bonding solvents, such as phenols or alcohols, solvates the bromide ion and decelerates the anomerisation of the glycosyl bromide.A similar argument with alcohols (or a hydrogen-bonded complex of alcohol and amine base) explains the success of the halide-ion-catalysed a-glycosylation reaction; 29 in this case c. c. h,P ~~ ~ Table 1. Yields, characterisation data, and recrystallisation data Required (%) Found (%) Empirical 7--Yield Recrystallis-7 Anomeric proton (0)Sugar residue Agl ycone (%I ation solvent M.P. (oc) caiDZ5 formula C H N Br C H N Br 6 (solvent) (J/Hz) 2,3,4-Tri-O-acetyl-3-Bromopyridine 71 Me,CO 114-115 C16H 19°7NBr2 38.65 3.85 2.8 32.15 38.5 4.0 2.8 32.3 7.50 (CDCl,) s a-D-xylop yranos yl (dec.) 4-Methylpyridine 37 Me, CO-Et, 0 156-158 -38.9 C, ,H,,O,NBr 47.25 5.15 3.25 18.5 47.05 5.05 3.35 18.45 7.32 (CDCI,) s (c 1 in CHCl,) Isoquinoline 30 Me,CO 1 38- 143 -46.0 C2,H,,07NBr 51.3 4.75 3.0 17.05 5 1.20 5.3 2.55 17.1 7.50 (CDCI,) s (c 0.8 in CHCI,) 4-Bromoisoquinoline 50 122-126 C20H2107NBr2 43.9 3.85 2.55 29.2 43.5 3.85 2.6 29.75 7.72 (CDCI,) s Pyridine 42 169-1 70 -45.7 C16H,,07NBr 45.95 4.8 3.35 19.1 46.25 5.05 3.4 20.25 7.41 (CDCI,) s (c 0.8 in CHCl,) 3-Carbamo ylp y ridine 16 114-1 18 C17H2108N2Br' 45.3 4.95 5.7 16.3 44.9 5.6 4.95 17.7 6.62 (D,O) s (dec.) OSMe,CO 3-Methylpyridine 24 Me,CO 131-133 -41.7 C,,H,,O,NBr 47.25 5.15 3.25 18.5 47.0 5.25 3.3 18.4 7.40 (CDCI,) s (c 1 in CHCI,) 3,4-Dimet h ylpyridine 61 Et,O-Me,CO 164-168 -36.6 C,,H,,O,NBr 48.45 5.4 3.15 17.9 48.4 5.7 3.05 17.9 7.40 (CDCI,) s (dec.) (c 1 in CHCI,) 3,5-Dimethylpyridine 39 Et,O-Me,CO 159-161 18H2,07NBr 48.45 5.4 3.15 17.9 48.45 5.6 3.1 17.8 7.52 (CDCI,) s 2,3,4-Tn-O-acetyl-Isoquinoline 23 EtOH 185-186 -39.8 C,,H,,O,NBr 51.3 4.75 3.0 17.05 51.45 5.1 2.95 16.95 6.29 (D,O) (8) 0-D-xy lop yranos yl (dec.) (c 1 in CHCI,) CBromoisoquinoline 19 EtOH 169-1 7 1 -71.7 C,,H,,O,NBr, 43.9 3.85 2.55 29.2 43.4 4.3 2.55 29.25 6.38 (CD,OD)(8) (dec.) (c 1 in CHCl,) 3-Carbamo ylpyridine 21 144-147 C17H2108N2Br 44.25 4.6 6.05 17.3 44.4 4.8 6.15 17.2 6.18 (D,O) (8) (dec.) Pyridine 21 Me,CO 167-1 69 -21.9 C16H,o0,NBr 45.95 4.8 3.35 19.1 45.0 4.9 3.3 19.3 7.41 (CDCI,) (8) (dec.) (c 1 in CHCI,) 3-Bromopyridine 20 EtOH 155-1 56 C16H1907NBr2 38.65 3.85 2.8 32.15 38.85 3.85 2.95 31.85 6.16 (D20)(8) (dec.) 2,3,4,6-Tetra-O-4-Methylpyridine 24 Et ,O-Me,CO 168-169 + 25.3 C,,H,,O,NBr 47.65 5.15 2.75 15.85 47.0 5.8 2.8 15.5 7.42 (CDCl,) (3) acety l-a-D-ghco-(c 1 in CHC1,) pyranosyl Isoquinoline 12 Et,O-Me,CO 154-155 i-21.0 C,,H,,O,NBr 51.1 4.85 2.6 14.8 5 1.05 5.1 2.35 14.85 7.74 (CDCI,) (3) (c 1 in CHCI,) 3,5-Dimethylpyridine 41 Et,O-Me,CO 177 +21.2 C2 H,,O,NBr 48.65 5.45 2.7 15.4 48.3 5.7 2.45 15.4 7.67 (CDCl,) (3) (c 1 in CHCI,) 3-Bromopyridine 18 Me,CO-Et,O 137-1 38 +31.0 C,,H,,O,NBr, 40.05 4.05 2.45 28.1 39.9 4.15 1.95 28.5 7.7 (CDCI,) (2.8) (c 0.8 in CHCl,) 4-Bromoisoquinoline 43 Et,O-Me,CO 161-162 +25.1 C2,Hz5O,NBr2 44.6 4.05 2.25 25.8 44.0 4.35 1.9 25.55 7.9 (CDCI,) (1.7) (c 1 in CHCl,) Pyridine 14 Et,O-Me,CO 127-128 + 19.2 C,,H,,O,NBr 46.55 4.9 2.85 16.3 45.0 4.85 2.65 14.1 7.5 (CDCI,) (2.6) (c 1 in CHCI,) 3-Carbamo ylpyndine 36 Et,O-Me,CO 162-165 + 21.1 C2,HZ5O ,N, Br 45.05 4.75 5.25 15.0 45.05 4.85 5.6 15.1 6.78 (DzO) (2.4) (c 1.0 in H,O) Required (%) Found (%) Yield Recrystallis-Empirical 7u r-*-,-Anomeric proton Sugar residue Aglycone (%I ation solvent M.p.("C) [a],," (") formula C HNBrC H N Br 1 6 (solvent) (J/Hz) a-D-Xylopyranosyl 3-Bromopyridine 60 See text 108-1 11 -38.7 1OH 1,04NBr2 32.35 3.55 3.8 43.05 32.45 3.6 3.45 43.3 6.25 (D20)s (c 0.9 in H20) CBromoisoquinoline 76 107-1 15 -40.6 C14H1 5°4NBr2-6.33 (DzO)s: 1/2 u (dec.) (c 1.0 in H20) 0.5C2H50H* mole P H2O 39.0 4.35 3.05 34.6 38.8 4.35 3.0 34.1 Equivalent ofethanol detected 3-Carbamo ylp yridine 52 127-129 -43.4 39.4 4.5 8.35 23.85 39.8 4.6 7.9 23.8 6.34 (D2O) s (c 0.9 in H20) Pyridine 79 157-159 -58.5 41.1 4.85 4.8 27.35 41.15 5.0 4.95 27.7 6.19 (D,O) s (dec.) (c 1.4 in H20) Isoquinoline 41 146-149 -39.4 49.15 4.7 4.1 23.35 48.85 4.9 4.05 23.15 6.29 (D20)s (c 1.5 in H20) 4-Methylpyridine 68 155-1 57 -52.6 43.15 5.25 4.55 26.1 43.45 5.45 4.6 25.7 6.09 (D20)s (c 0.9 in H20) P-D-Xylopyranosyl 4-Bromoisoquinoline 50 148-150 -8.4 39.95 3.6 3.35 37.95 40.35 3.65 3.35 37.5 5.84 (D20), 8.5 (+ (dec.) (c 0.8 in H20) virtual coupling) 3-Carbamoylpyridine 74 157-160 39.4 4.5 8.35 23.85 40.1 5.05 7.9 23.65 5.79 (D20)(8)cl 1H1505N2BrPyridine 92 149-150 -5.6 1OH 14°4NBr 41.1 4.85 4.8 27.35 41.15 5.0 4.75 27.2 5.70 (D20)(8) (c 1.0 in H20) Isoquinoline 51 142-145 -1.7 49.15 4.7 4.1 23.35 49.15 5.0 3.9 23.25 5.87 (D20)(8.3) (c 1.2 in H20) a-D-Glucopyranosyl 3-Bromopyridine 28 121-122 +40.2 cl 1 1 5°5NBr2 32.95 3.8 3.5 39.85 32.75 3.9 3.2 40.05 6.39 (D20)(3.8) (c 1 in H20) CBromoisoquinoline 40 140-143 +48.9 39.95 3.8 3.1 34.45 39.75 3.95 2.9 35.05 6.60 (D2O) (3.9) (c 0.9 in H20) Pyridine 54 145-147 +52.5 41.0 5.01 4.35 24.8 40.7 5.3 4.1 25.25 6.28 (DZO) (3.8) (c 1 in H,O) Isoquinoline 21 138-140 +51.5 48.4 4.85 3.75 21.5 48.25 5.05 3.55 21.65 6.6 (D20)(3) (c 0.9 in H20) CMethylpyridine 46 127-128 +57.3 1ZH 18°5NBr 42.85 5.4 4.15 23.75 43.1 5.35 4.1 23.45 6.32 (D20)(3.9) (c 0.8 in H20) 1126 J.CHEM. SOC. PERKIN TRANS. II 1984 R’ =OAc, R2=H, fast R2 R*=OAc, R’=H, slow BrAcO ‘T’ACO AcO Br 1(VIII) R1 +QAcO .N AcO (1x1 (XI 1 R’=OAc, R2= R3=H, 0-xylo R2 =OAc, R’ = R3 =HI L -orabino R’ =OAC, RZ =HI R3= CH~OAC, D -gluco R2=OAc, R1 = HI d =CH20Ac, 0 -galact0 Scheme 1.Table 2. Chemical shifts and coupling constants (D,O) Chemical shifts (6) , a-D-XylOpyranOSyl-4-meth ylp yridinium bromide P-D-Xylop yranos yl-pyridinium acetate , a-D-Glucop yranos yl-4bromoisoquinolinium bromide H(1) 6.19 5.70 6.60 H(2) H(3) 4.01 4.23 3.63 3.72 4.49 4.19 H(4) 3.86 3.91 3.90 H(5) (pro-R) 4.35 4.32 H(5) br0-S) 4.25 3.65 Coupling constants (Hz) 1.4 8.3 3.7 2.7 10.0 6.8 2.5 8.9 6.2 1.8 5.4 1.5 10.4 8.3 13.4 11.2 0.9 1.2 1.2 however the 2-acetoxy group of halide (X) competes favourably that the a-~-xylopyranosyl-4-methylpyridiniumion adopts the as an intramolecular nucleophile with alcohol as an external C, conformation in D20.Indeed, the coupling constants nucleophile, and orthoesters are eventually produced unless quoted by Paulsen et aZ.I8 for the tri-0-acetylpyridinium non-participating protecting groups are used.chloride in deuterionitromethane can be used to give coupling patterns that correspond to the observed spectrum, if an Conformation of Deprotected a-D-Xylopyranosyl- and a-D-additional 0.9 Hz coupling between H(l) and H(3) is assumed. Glucopyranosyl-pyridiniumSaltsin Water.-Chemical shifts and A slight change in the shape of the pyranose ring is expected on coupling constants of protons attached to the pyranose rings deacetylation, since acylation is known to reduce the repulsion of representative a-and P-D-xylopyranosyl- and a-D-g1UCO- between 1,3-diaxial hydroxy groups. 3*3* pyranosyl-pyridinium salts are given in Table 2.These indicate This additional four-bond coupling makes the anomeric 1127J. CHEM. SOC. PERKIN TRANS. 11 1984 HO OH a-D-Xylopyranosylpyridinium ion, 1c, HO p-0 -Xylopyranosylpyridiniurn ion, 4C1 HO a-0-Glucopyranosylpyridinium ion,'S3 Figure 1. Preferred conformations of glycopyranosylpyridinium salts protons of a-D-xylopyranosylpyridinium salts appear as broad singlets. The P-D-xylopyranosylpyridinium salts are clearly in the ,C, conformation; the ,C1 and the 'C4 conformations are illustrated in Figure 1.The conformation of the a-D-glucopyranosylpyridinium salts is less obvious. Lemieux ' examined the H n.m.r. spectrum of 2,3,4,6-tetra-O-acetyl-a-~-glucopyranosyl-4-methylpyridiniumbromide in D20, found J,,2 2.8, J2,3 3.1,J3,43.2,and J4,55.7 Hz, and concluded that the same B2,5conformation was adopted as in the crystal.32 A full analysis of the deacetylated compound in D20was not possible because of chemical shift coincidences, but we find Jl,23.9, J2.3 7.1, and J3,45.6 Hz: with the 4-bromoisoquinolinium salt J4,5is also determinable (Table 2). It is clear that dihedral angles are radically different from those of the tetra-0-acetyl derivative, a possible consequence of the apparent attraction between acyloxy groups mentioned earlier.31 Comparison with the data for a-and P-xylopyranosylpyridinium salts (Table 2) shows that the deacetylated compound is neither in the ,C, nor the 'C, conformation, nor an equilibrating mixture of just those two conformations.Clearly, as with the acetylated compound, the axial substituent at C(5) makes the 'C, conformation prohibitively energetic, and boat conformations are adopted. [The full positive charge on the nitrogen of the pyridinium salt thus has a bigger effect than the half positive charge on N( 1) of protonated a-D-glucopyranosylimidazole, which remains in the ,C, conf~rmation.~~] Although the possibility that observed coupling constants are a time average of several different conformations cannot be ruled out, one single conformation at an energy minimum on the twist-boat pseudorotational itinerary can account for the observed coupling constants in a general way (cf.ref.34). In this, the 'S, conformation (Figure l), all substituents are in energetically favourable positions. The C(1)-N bond is pseudoequatorial exactly eclipsing a line bisecting the two oxygen lone pair orbitals, thus allowing maximum operation of the reverse anomeric effect. The hydroxymethyl group and 2-OH are isoclinal, and the remaining substituents pseudo- -4.0 I h -5.0I -8O'i 5' 6' 7' 8 9 10 11 12 13' ' I PH Figure 2. Dependence of rate of hydrolysis of (A) a-D-glUCOpyranOSyl-and (B) a-~-xylopyranosyl-3-bromopyridiniumions on pH.Conditions: 25.0 "C, l.0M-sodium perchlorate. Rate constants < l(3' s-' were estimated from extrapolation of two rate constants obtained at two higher temperatures. The solid lines describe the following equations, for the a-D-xylopyranosylpyridinium salts: kobo/S-'= 4.7 x 1P + 1.4 x lo-'* CH'I and for the a-D-glucopyranosylpyridinium salts: 1017 kobr/s = 2 x 10 + 1.4 CH 'I equatorial. This is thus likely to be the most populated conformation of a-D-glucopyranosylpyridinium salts. Hydrolysis of a-and p-, D-Glucopyranosyj- and D-XylO- pyranosylpyridinium Ions.-Figure 2 shows the dependence on pH of the first-order rate constant for disappearance of a-D-xylopyranosyl-and ~-~-glucopyranosyl-3-bromopyridinium ions.It is clear that there are two pathways for these hydrolyses, a pH-independent and a base-catalysed one. The pH-independent pathway is a simple unimolecular heterolysis of the glycosyl-pyridine bond, as is shown by the positive entropies of activation (15 & 1 for the a-xylo- and 13 & 2 cal mol-' K-' for the a-giuco-compound) and, in the xylu-case, the absence of a detectable rate enhancement in the presence of the neutral nucleophile thiourea.'* At pH 6.0, glucose (80%) and 3-bromopyridine (78%) or xylose (45%) and 3-bromopyridine (92%) are the products. The low yield of xylose is almost certainly due to degradation of the labile pentose: products were isolated after six half-lives, and from p-~-xylopyranosyl-3-carbamoylpyridiniumion, a slower sub- strate, only 17% xylose could be detected, despite a 67% yield of aglycone.Heating D-xylose and nicotinamide under the conditions, and at the concentrations used for product isolation, results in extensive degradation, browning, and 70% loss of optical activity; the nicotinamide is not affected. The degradation of reducing sugars under the fairly drastic conditions used for complete hydrolysis of the pyridinium salts (pH 6 at 100 "C for several hours) is well recognised: the process is acid- and base-dependent, having a minimum at pH 3.5,35 and gives complex, non-carbohydrate products.36 We assume that the p-~-glucopyranosyl-4-bromoiso-quinolinium salt gives the sugar and the aglycone, by analogy Table3.Kinetic data for hydrolysis of glycosylpyridinium salts in 5Om~- NaH,PO,-Na,HPO,, buffer 1 .OM-NaClO,, pH 6.0 AH/ AS/Pyridine T/"C lo5 k/s-' kcal/mol-' cal mol-'K-' (i) a-D-Xylopyranosylpyndiniumsalts 3-Bromopyridine 89.8 74.3 31.2 f0.8 13.0 f 2.8 80.5 26.9 (31.7 f0.4) (14.6 f 1.2)" 50.2 0.35 CBromoisoquinoline 100.1 85.7 33.1 f0.3 15.5 f0.9 89.9 23.1 80.7 6.98 Nicotinamide 100. 1 126.2 32.9 k0.9 15.8 f 3.1 89.9 33.6 80.6 10.5 Py ridine 130 111.6 38.9 k 1.7 24 f6 120 28.9 110 8.56 CMethylpyridine 140 77.8 39 f2 21 _+ 7 130 26.1 120 6.66 (ii) P-D-Xylopyranosylpyndiniumsalts 4-Bromoisoquinoline 109.8 116.9 39 i-2 30 f7 100.1 24.5 81.1 1.621 Nicotinamide 104.7 75.0 37.6 1.0 25 f3 100.1 36.8 80.8 2.58 79.4 1.88 Pyridine 136 119.4 42.3 f0.8 31 _+ 3 131 59.3 116 7.81 (iii) ~-~-Glucopyranosyl-4-bromoisoquinoliniumion 127 164 118 62.9 36.5 k 1.8 18 f6 108 16.7 (iv) a-D-Glucopyranosylpyridiniumsalts 3-Bromopyridine 80.4 59.7 80.4 62.0 80.8 64.4 29.5 f 0.2 9.9 & 0.8 65.7 9.98 (30.4 f 0.8) (13.0 i-2.4)" 65.8 10.67 65.7 10.00 65.75 10.22 49.9 1.017 50.2 1.20 4-Bromoisoquinoline 89.9 57.6 89.9 56.1 81.3 17.6 33.2 f0.9 17.6 f3.1 81.4 16.9 66.8 2.14 66.8 2.14 67.0 2.61 66.6 2.44 Pyridine 120.0 102.5 119.5 88.5 34.6 f1.0 15.2 k 3.3 110.0 27.4 110.5 30.0 100.2 9.43 100.2 8.58 4-Methylpyridine 136.0 80.1 136.0 89.5 128.0 36.5 35.6 k 1.5 13.9 & 5.1 126.7 33.0 115.0 10.0 118.0 11.9 Mean and S.D. of data used to construct pH-rate profile, pH 3-9. J. CHEM. SOC. PERKIN TRANS. 11 1984 with the galacto-compounds, where this was demonstrated 37 in a representative case: The base-catalysed pathways are more complex, since in principle attack can occur either on the pyridinium ring or at the anomeric centre, and attack at either of these sites can be by either hydroxide ion or by one of the sugar hydroxy groups, ionised in a pre-equilibrium process. Attack at the anomeric centre is predominant; at pH 12.2 both a-D-ghlcopyranosyl- and a-~-xylopyranosyl-3-bromopyridiniumbromides give an 80% yield of free pyridine, and the glucosyl salt gives 35% 1,6- anhydro-D-glucopyranose and 17% glucose.This last figure is almost certainly a lower limit; glucose, like all reducing sugars is destroyed in alkali, and indeed from the xylo salt no xylose could be detected (although a fast-moving peak on the g.1.c. could conceivably be the trimethylsilylated 1,3-anhydro-~-xylopyranose). However, attack on the pyridine also takes place: both the a-D-glucopyranosyl- and a-~-xylopyranosyl-3- bromopyridinium ions at pH 12.2 give an intense, non-ether- extractable absorbance at Amax. 366, probably due to the sodium salt of bromoglutaconic dialdehyde: its intensity, and the literature extinction ~oefficient,~' indicates about 8% of the reaction goes this way in both cases. Since the absolute rates of the a-gluco- and a-xylo-salts in alkali differ by an order of magnitude (Figure 2), and since the rate of attack of hydroxide ion on the pyridinium ring in the two ions should be approximately the same, the attack on the pyridine ring is probably initiated by intramolecular attack of an ionised sugar hydroxy Kinetic data for the hydrolyses of various glycosylpyridinium salts in 1.0M-sodium perchlorate at pH 6.0, in the clearly pH-independent region, are given in Table 3.Rates, extrapolated from these data to 25.0°C, correlate with the pKa of the departing pyridine as shown in equations (l), (2), and (3) for the a-D-xylopyranosylpyridinium salts, the three p-D-xylopyranosylpyridinium salts, and the a-D-glucopyranosyl- pyridinium salts, respectively. It has been shown previously 37 that for pH-independent hydrolysis of P-D-galactopyranosyl- pyridinium salts in 1 .OM-potassium chloride equation (4) results.log(k/s-') = -(3.67 f0.26) -(1.28 k0.06) PK, (1) log(k/s-') = -(3.88 & 0.53)-(1.06 +_ 0.12) pK, (3) log(k/s-') = -(4.33 f0.49) -(1.26 f0.12) pK, (4) The ratio of rates of departure of pyridines from a-and p-xylopyranosyl residues is very modest, about an order of magnitude if the data are extrapolated to 25 "C (24 for the 4-bromoisoquinolinium, 1 3 for the 3-carbamoylpyridinium, and 8 for the pyridinium salts), and 3.5 if measured rates for the first two compounds are directly compared at 100.1 "C. It is what would have been expected before the advent of the antiperiplanar lone pair hypothesis for conversion of three equatorial to three axial substituents: it was then considered that relief of strain on going to the half-chair conformation of the glycosyl cation resulted in about a three-fold increase in the rate of hydrolysis of glycopyranosyl derivatives with one axial hydroxy group, compared with the corresponding all-equatorial deri~ative.~'The calculations set out in Figure 3, using Angyal's instability factors," indicate that, according to the antiperi- planar lone pair hypothesis, the a-xylosyl salts should hydrolyse at least 150 times faster than their anomers, even in the unlikely event that repulsive 1,3-interactions between a pyridine and a hydroxy group were no greater than between a pyridine and a hydrogen.It could doubtless be argued that the structure of the J.CHEM. SOC. PERKIN TRANS. 11 1984 Hb i)H 2 vicinal diequatorial OH 1 1-3 OH-OH diaxial inter- interactions = 0.7 kcal action = 1.5 kcal 1 1-3 OH-H diaxial inter- action = 0.45 kcal 1 axial pyridine = X 1 1-3 OH-pyridine diaxial interaction = Y Total change in conformational energy on going from ,C1 to 'C, conformation = 1.25 + X + Y 2 1-3 OH-H diaxial inter- 2 vicinal diequatorial OH actions = 0.9 kcal interactions = 0.7 kcal t 1-3 OH-OH diaxial inter- 1 axial pyridine = X action = 1.5 kcal Total change in conformational energy on going from 'C,to ,C, conformation = X -1.7 kcal Difference in energy required to provide an antiplanar lone pair = 2.95 + Y kcal Figure 3. Estimation of the ratio of hydrolysis rates of a-to p-xylopyranosylpyridinium salts, if the antiperiplanar lone pair hypothesis were correct transition state is different for the two sets of anomeric pyridinium salts: there is no evidence from p,, values * that this is so.(Strictly, the difference in conformational energies between ground state and transition states, rather than reactive ground state conformations, should be estimated: however, the antiperiplanar lone pair hypothesis as formulated' is a statement about reactive ground state conformations and thus, strictu sensu, violates the Curtin-Hammett principle. As the chair conformation with an oxygen lone pair antiperiplanar to the leaving group moves towards the half-chair of the oxocarbonium ion, the disfavouring of axial OH groups will decrease somewhat, but Y will still be substantial.) Similar considerations apply to the gfuco-series.The a:p rate ratio here, determined on the 4-bromoisoquinolinium salts,27 is 80 at 25.0 "C. It is highly unlikely that the p,, value for the p salts will vary as between glucose and galactose, and so the value of -1.26 0.12 from equation (4) can be compared with the value of -1.06 & 0.12 for the a-glucosyl salts [equation (3)]: there is little evidence for a substantial change of transition state structure in the two sets of compounds. A similar calculation to that of Figure 3 can be performed, but conformational energies for substituents in boat-form pyranose rings are not available.A value of 5.5 kcal mol-' for the chair-twist boat conformational enthalpy difference in cyclohexane has been estimated.41 If we assume that a 0-glucopyranosylpyridinium salt has to adopt the 'S, conformation to present the leaving group with an antiperiplanar lone pair, whereas the a salt has to adopt a 4C1 * PI, is the gradient of the plot of log,, (rate) against the pK, of the leaving group. chair from the 'S, conformation, this value alone, neglecting any additional steric interactions of substituents in the ISg conformation, predicts a rate ratio of around 10'. This is clearly not observed: indeed, the a:p rate ratio, corresponding to departure of a leaving group from a boat, or from the equatorial position of a chair, is surprisingly close to that observed by Hanack and Heinz 42 for ethanolysis of toluene-p-sulphonates (XII) and (XIII) (137 at 50 "C).The possibility that a: p ratios in either the gluco- or the xylo- series arise from the selective operation of eliminative pathways or of neighbouring group participation by OH is unlikely.Eliminative pathways would not give any reducing sugar at all, and pH-independence of the reaction would require H20 tQ act as a base towards unactivated C-H bonds. Participation in reactions at C( 1) of a sugar by un-ionised OH groups in water has never been observed. In the most favourable case, the a-glucopyridinium salts, it would have produced 1,6-anhydro- glucose, which was not detected. We conclude that the antiperiplanar lone pair hypothesis has no predictive value in the area of glycosyl transfer.In order to clarify the reasons for the failure of the hypothesis in this area, it is instructive to consider its predictions in terms of nuclear motion. Chemical reactions necessarily involve nuclear motion, and, in accord with the Born-Oppenheimer approximation, the motions of nuclei can be considered independently of the motion of electrons.43 As applied to S,l reactions of aldopyranosyl derivatives, the antiperiplanar lone pair hypothesis predicts trajectory (a) to be favoured over trajectory (b) (Scheme 2). If we discuss motion along these pathways relative to the plane defined by C(2), 0(5),C(5), C(1), and H(l) of the oxocarbonium ion, the only difference in motion of the glycone portion of the molecule is a 90", rather than a 30", rotation of the C(1)-H( 1) bond.In this case, then, the predictions of the antiperiplanar lone pair hypothesis are equivalent to the predictions of the principle of least nuclear motion.44 Least motion effects are recognised to be of variable imp~rtance.~~In particular, because of the variation of internal energy as (approximately) the square of nuclear displacement, they are of limited importance for very reactant-like or very product-like transition states. A substantial body of evidence indicates that the transition state in glycosyl transfer reactions is very late: 40,45 in the case of the glycosylpyridinium salts, the hydrolyses of which are considered here, this is shown by PI, values of less than -1.Therefore, since the transition state for glycosyl transfer strongly resembles the immediate product, a glycosyl cation, least motion effects are not expected to be important; indeed they are not. The failure of the antiperiplanar lone pair hypothesis in glycosyl transfer is therefore in accord with the re-interpretation of the hypothesis as a special case of the principle of least nuclear motion. This principle also accounts for the 'first unambiguous evidence for stereoelectronic control', published by Perrin and Arrhenius in 1982.This was the production of o-amino amides as kinetic products, and lactams as thermo- dynamic products, in the base-catalysed hydrolysis of cyclic amidines (XIV) (Scheme 3).The tetrahedral intermediate was considered to have a lifetime that was short compared with the time of inversion about nitrogen, but long compared with ring inversion or rotation about C-O or C-N bonds6 Under these conditions least motion arguments also predict the observed results, in the following way. The tetrahedral intermediate will LG Scheme 2. LG = (XYa) HO HO H (XV b) (XVC) To'H Scheme 3. be generated in conformation (XVa) first, since this conformation alone is accessible by only 30"rotations of bonds out of the plane defined by the amidine function. This species has various rotamers Leg. (XVb and c)] and the ring-flipped conformation (XVI)and its rotamers, available to it, but does not live long enough for nitrogen inversion to convert the ring NH of (XV) or (XVI) into the equatorial orientation.Consequently, there is no accessible conformation that can give lactam without a 90" motion, similar to that of H( 1) in pathway (b) in Scheme 2, either of the OH group in conformation (XV) or of the hydrogen bound to the ring nitrogen in conformation (XVI).However, opening of the ring from either conformation (XVb or c) is possible with only 30" bond motions. Therefore, least motion predicts the exclusive formation of a-amino amide. Re-interpretation of the corpus of data held to support the antiperiplanar lone pair hypothesis in terms of least motion removes a puzzling feature of this supposedly electronic phenomenon, namely, that it is more readily observed with compounds at the acyl level of oxidation than at the aldehyde level.The electronic interactions with a developing cationic centre are of decreasing importance in the presence of another electron-donating conjugating By contrast, the transition state for reactions of acetals is more oxocarbonium- ion like than the transition state for the hydrolysis of ~rthoesters,~'and this is consistent with the greater importance of least motion effects in the latter case. However, if molecules, such as acetal (VII),are constructed in which motion analogous to that of H(l) in pathway (b) of Scheme 2 is made wholly impossible, least motion predictions of reactivity are necessarily fulfilled. J. CHEM. SOC. PERKIN TRANS.11 1984 leaving group Conclusion The antiperiplanar lone pair hypothesis has been promoted as a fundamental chemical principle, of widespread appli~ation.~~ Our results with glycosyl derivatives show unambiguously that such large claims are unwarranted: a theory that claims to be fundamental need only fail once to be disproved, no matter how useful it is as a rule-of-thumb. The predictions of the antiperiplanar lone pair hypothesis are in fact equivalent to the predictions of the principle of least nuclear motion; this principle accounts, for example, for the positive result of 'the first unambiguous test of the theory of stereoelectronic control'. The limitations of the principle of least nuclear motion are well known.44 Whatever value the antiperiplanar lone pair hypothesis may have as the simple pictorial illustration of a rule-of-thumb, its limitations must be recognised.Its application, as dogma, to the interpretation of ambiguous preparative experiments (cf: the discussion in ref. 6) or mechanistically inaccessible steps in enzymic reactions,4' can only hinder proper understanding. Acknowledgements We thank the S.E.R.C. for research studentships (to L. H. and P. J. M.), Dr. M. Murray for runningn.m.r. spectra, and Dr. R.W. Alder for helpful discussions. References 1 Reviews: P. Deslongchamps, Pure Appl. Chem., 1975, 43, 351; Tetrahedron, 1975, 31, 2463; Heterocycles, 1977, 7, 1271. 2 J. M. Lehn and G. Wipff, J.Am. Chem. SOC.,1974,%, 4048; 1976,98, 7498; 1980, 102, 1347.3 B. Capon, A. K. Ghosh, and D. McL. A. Grieve, Ace. Chem. Res., 1981, 14, 306. 4 (a) M. Ahmad, R. G. Bergstrom, M. J. Cashen, Y. Chiang, A. J. Kresge, R. A. McClelland, and M. F. Powell,J.Am. Chem. SOC.,1979, 101,2669; (b)B. Capon and D. McL. A. Grieve, J. Chem SOC.,Perkin Trans. 2, 1980, 300; (c) R. A. McClelland and M. Alibhai, Can. J. Chem., 1981, 59, 1169. 5 M. Caswell and G. L. Schmir, J. Am. Chem. SOC.,1979, 101, 7323. 6 C. L. Perrin and G. M. L. Arrhenius, J. Am. Chem. Soc., 1982, 104, 2839. 7 B. Capon and D. McL. A. Grieve, Tetrahedron Lett., 1982,23,4823. 8 M. C. Caserio, Y. Sooma, and J. K. Kim, J. Am. Chem. Soc., 1981, 103,6712. 9 0.E. Desvard and A. J. Kirby, Tetrahedron Lett., 1982, 23, 4163. 0 J. V. OConnor and R.Barker, Curbohydr. Res., 1979, 73, 227. 1 A. J. Bennet and M. L. Sinnott, unpublished work. 2 A. J. Kirby, 'The Anomeric Effect and Related Stereoelectronic Effects at Oxygen,' Springer, Berlin, 1983. 3 A. J. Kirby and R. J. Martin J.Chem. SOC.,Chem. Commun., 1978,803. 4 A. J. Kirby and R. J. Martin, J. Chem SOC.,Chem. Commun., 1979, 1079. 5 C. M. Evans, R. Glenn, and A. J. Kirby, J. Am. Chern. SOC.,1982,104, 4706. 6 S. Chandrasekharand A. J. Kirby, J. Chem. Soc., Chem. Commun., 1978, 171. 7 R. U. Lemieux, Pure Appl. Chem., 1971,25, 527. 8 H. Paulsen. Z. Gyorgydeak, and M. Friedmann, Chem. Ber., 1974, 107, 1590. 9 S. J. Angyal, Angew. Chem., Int. Ed. Engl., 1969, 8, 157. J. CHEM. SOC. PERKIN TRANS. XI 1984 20 D.G. Gorenstein, J. B. Findlay, B. A. Luxon, and D. Kar, J. Am. Chem. SOC., 1977,99, 3473. 21 M. L. Sinnott and W. P. Jencks, J. Am. Chem. SOC., 1980,102,2026. 22 (a)W. P. Jencks, Chem. SOC.Rev., 1981,10,345; (b)P. R. Young and W. P.Jencks, J. Am. Chem. SOC., 1977,99, 8238; (c) B. L. Knier and W. P. Jencks, ibid., 1980, 102, 6789. 23 R. U. Lemieux and A. R. Morgan, Can.J. Chem., 1965,43, (a) 2205; (b) 2214. 24 E. S. Swinbourne, ‘Analysis of Kinetic Data,’ Nelson, London, 1971. 25 M. L. Sinnott and S. G. Withers, Biochem. J., 1974, 143, 751. 26 P. Marshall, C. G. Reed, M. L. Sinnott, and I. J. L. Souchard, J. Chem. SOC.,Perkin Trans. 2, 1977, 1 198. 27 M. L. Sinnott and S. G. Withers, unpublished work; G. Legler, M. L. Sinnott, and S.G. Withers, J.Chem. SOC., Perkin Trans. 2,1980,1376. 28 R. U. Lemieux, Chem. SOC. Rev., 1978, 7,423. 29 R. U. Lemieux, K. B. Hendriks, R. V. Stick, and K. James, J. Am. Chem. SOC., 1975,97,4056. 30 C. V. Holland, D. Horton, and J. S. Jewell, J. Org. Chem., 1967, 32, 1818. 31 P. Luger, G. Kothe, K. Vangehr, H. Paulsen, and F. R. Heiker, Carbohydr. Res., 1979,68, 207. 32 M. N. G. James, Proc. Can. Fed. Biol. Sci., 1969,13,71, quoted in ref. 16, p. 15. 33 P. Finch and A. G. Nagpurkar, Carbohydr. Res., 1976,49, 275. 34 G. Kotowycz and R.U. Lemieux, Chem. Rev., 1973,73,669. These authors caution against other than qualitative reliance on the various modifications of the Karplus equation for polar molecules. For a multi-parameter treatment see C.A. G. Haasnodt, F. A. M. de Leeuw, and C. Altona, Bull. SOC. Chim. Belg., 1980, 89, 125. 35 A. R. Sapronov and S. E. Kharin, Sakh. Prom., 1967,41,55 (Chem. Abstr., 1968, 67, 108864g). 36 T. Popoff and 0.Theander, Carbohydr. Res., 1972, 22, 135. 37 C. C. Jones, M. L. Sinnott,and I. J. L. Souchard, J.Chem. SOC., Perkin Trans. 2, 1977, 1191. 38 J. Becher and M. C. Christensen, Tetrahedron, 1979,35, 1523. 39 S. G. Withers, M. Jullien, M. L. Sinnott, 0.M. Viratelle, and J. M. Yon, Eur. J. Biochem., 1978, 87, 249. 40 B. Capon, Chem Rev., 1969,69,407. 41 M. Squillacote, R. S. Sheridan, 0.L. Chapman, and F. A. L. Anet, J. Am. Chem. SOC.,1975,97, 3244. 42 M Hanack and K.-W. Heinz, Annalen, 1965,682,75. 43 (a) M. Born and J. R. Oppenheimer, Ann. Phys. (Leipzig), 1927,84, 457; (b) P. W. Atkins, ‘Quanta, a Handbook of Concepts,’ Oxford University Press, Oxford, 1974, p. 29. 44 Review: J. Hine, Adv. Phys. Org. Chem., 1977, 15, 1. 45 Review: M. L. Sinnott, in ‘The Chemistry of Enzyme Action,’ ed. M.I. Page, Elsevier, Amsterdam, 1984, p. 389. 46 G. M. Loudon and C. Berke, J. Am. Chem. SOC., 1974, %, 4508. 47 E. H. Cordes and H. G. Bull, Chem. Rev., 1974,74, 581. 48 P. Deslongchamps, ‘Stereoelectronic Effects in Organic Chemistry,’ Pergamon, Oxford, 1983. 49 B. Asboth and L. Polgar, Biochemistry, 1983, 22, 117, report a predictive failure of the antiperiplanar lone pair hypothesis in serine proteinase catalysis, but invoke ‘interactions with the protein environment (which) may endow the tetrahedral intermediate with properties different from those in simple solution reactions.’ Received 24th March 1983; Paper 31465
ISSN:1472-779X
DOI:10.1039/P29840001121
出版商:RSC
年代:1984
数据来源: RSC
|
|