|
1. |
The stabilities of Meisenheimer complexes. Part 28. The reactions of 2,2′,4,4′,6,6′-hexanitrobibenzyl with alkoxides |
|
Journal of the Chemical Society, Perkin Transactions 2,
Volume 1,
Issue 1,
1982,
Page 31-34
Michael R. Crampton,
Preview
|
PDF (490KB)
|
|
摘要:
1982 31 The Stabilities of Meisenheimer Complexes. Part 28.1 The Reactions of 2,2‘,4,4‘,6,6‘-Hexanitrobibenzyl with Alkoxides By Michael R. Crampton and Paul J. Routledge, Department of Chemistry, University of Durham, South Road, Durham DH1 3LE George C. Corfield and Roger M. King, Department of Chemistry, Sheffield City Polytechnic, Sheffield S1 1WB Peter Golding, Ministry of Defence, P.E.R.M.E., Waltham Abbey, Essex EN9 1BP lH N.m.r. and visible spectral measurements indicate that in the presence of alkoxide ions 2,2’,4,4‘,6,6’-hexanitro-bibenzyl forms a-adducts by base addition. Attack at the 3-(3’-) positions precedes attack at the 1 -(1’-) positions but gives thermodynamically less stable adducts. Kinetic and equilibrium data are reported for the formation in ethanolic sodium ethoxide of 1 : 1 adducts by attack at the 3-or 1 -position and for the formation of a 1 : 2 adduct by addition at the 1-and 1 ’-positions.Data are also reported for the formation in methanolic sodium methoxide of a I : 1 adduct formed by attack at the 1 -position. These results are compared with values obtained for a-adduct formation of related substrates. ONE method for the production of the commercially important product 2,2‘,4,4’,6,6’-hexanitrostilbeneis the dehydrogenation of 2,2’,4,4’,6,6’-hexanitrobibenzyl(HN-BB) by quinones in basic media.293 Clearly a knowledge of the mode or modes of interaction of HNBB with base is important to the understanding of the mechanism of this reaction. We report here on the reversible reactions of HNBB with alkoxide ions.By analogy with related compounds, such as 2,4,6-trinitrotoluene (TNT) and 2,4,6- t rini trobenz y 1 chloride (TNBCl),4 likely products of (1 : 1) interaction with base are the a-adducts (1) or (2) formed by base attack at the 3-or 1-positions respectively, or the conjugate base (3). Because of the soparation of the aromatic rings by two methylene groups it might be expected that dianionic species would readily be formed. LIL I I CH2 CH2 lH N.m.r. Measurements.-The spectrum of the parent in [2H,]dimethyl sulphoxide shows two singlets at 6 3.35 and 9.0 due respectively to methylene and ring protons. The addition of 2 mol. equiv. of sodium trideuteriomethoxide in [2H,]methan~l resulted in the rapid formation, at the expense of the ring proton band, of two doublets at 6 6.1 and 8.5.These bands decreased in intensity with time and two singlets at 6 8.53 and 8.6 were observed. The band at 6 8.53 eventually took all the intensity due to ring protons and the final spectrum, observed after 6 min 7 and stable over the next 10 min, consisted of this band and a singlet of equal intensity at 6 2.40. This spectrum is consistent only with the di- adduct (4)in which methoxide attack has occurred at the 1-and 1’-positions of the parent. The shift to high field, relative to the resonance positions of the parent, of the ring and methylene proton rescmances is in agreement with similar shifts in related system^.^ The initial observation of doublets at 6 6.1 and 8.5 indicates that methoxide attack at the unsubstituted 3-and/or 3’-positions occurs ra~idly.~ Accordingly the spectrum of the parent in the presence of 1 mol.equiv. of base initially showed doublets at 6 6.1 and 8.5 and a broad band at 6 8.9which may be attributed to the ring protons of (1; R = Me). The broadness of the band (6 8.9) due to protons in the unattacked ring may indicate that some electron transfer to give radical anions OCCU~S.~ The band at 6 8.6 observed at intermediate times in solutions containing 2 equiv. of base may be due to the ring protons in the 1-methoxy-ring of the diadduct (5)which will presumably be present as a transient species. A further possibility is that this band is due to a thermo- dynamically unfavourable rotational isomer of the di- adduct (4).The main conclusion from these measurements is that attack at unsubstituted ring positions is kinetically 1‘ 1 min = 60 s. favoured while attack at the 1-and 1’-positions leads to the thermodynamically more stable products. Reaction with Sodium Methoxide in Methanol.-In the presence of dilute (<O.~M) sodium methoxide in methanol a rapid reversible reaction is observed giving a species with A,,,. 430 and 500 nm (shoulder). This is followed fairly quickly (minutes) by irreversible decomposition of the substrate. Examination of the system by stopped flow spectrophotometry indicated that a single rapid colour-forming reaction was present.Data are in Table 1. These data are best interpreted as TABLE1 Kinetic and equilibrium data for the reaction of 2,2’4,4’,-6,6’-hexanitrobibenzyl with sodium methoxide in methanol at 25 “C * [NaoMe]/~ 0.01 kobs./s-l 1.4 i0.1 OD (480 nm) (I 0.0041 K, b/l mo1-l 23 0.02 1.6 0.0066 21 0.04 2.1 0.0098 20 0.07 2.8 0.0114 16 0.10 3.4 0.0146 20 * “C = K - 273.15. Fo r 1 x 10-6~ parent measured with a 2 mm cell. A Benesi-Hildebrand plot gives a value of 0.022 for complete conversion, corresponding to a value for c of 1.1 x lo* 1 mol-1 cm-1. b Calculated using the expression OD(480)/[0.022 -OD(480)][NaOMe]. shown in equation (1)as reaction of the parent (P)with methoxide ions to give the adduct (2; R = Me).Since the base concentration is in large excess over the parent equation (2) will apply. k(P) + OMe-+ (2; R = Me) (1) k-1 kobs. = K,[OMe-] + K-, (2) A linear plot of kobs. versm base concentration gave values for k, of 23 1mol-l s-l f and k-, of 1.15 s-l. Combin-ation of these values gives a value for K, of 20 1 mol-l in good agreement with that obtained from equilibrium optical density measurements. The visible spectrum is consistent 495 with the form- ation of a o-adduct rather than formation of the con- jugate base (3). Two facts indicating that, at the base concentrations used, there is no substantial conversion of the parent into the diadduct (4; R = Me) are the observ- h/nm Visible spectra of HNBB (2 x 10-5~) and sodium ethoxide in ethanol.A, [NaOEt] 2 x 10-3~after 2 min; B, same as A after 30 min (indicates irreversible reaction of substrate) ; C, [NaOEt] 4 x 10-2~after 2 min t 1 1 = 10-3 m3. J.C.S. Perkin I1 ation of only one rate process and the value of the extinction coefficient which is too low for a diadduct (cf. ethoxide addition). It should be noted that the presence of dimethyl sulphoxide, used in the n.m.r. work, greatly increases the basicity of the medium enabling formation of the diadduct (4;R = Me) to occur. The n.m.r. work also indicates that in dimethyl sulphoxide addition at the 3-position to give (1; R = Me) precedes addition at the l-position. Our failure to observe addition at the 3-position in methanol can be attributed to the low equilibrium constant expected for formation of this adduct.For comparison the value of the equilibrium constant for methoxide attack at the 3-position of 2,4,6-trinitrotoluene in methanol is 0.07 1 mol-l. TABLE2 Equilibrium data for reaction of 2,2’,4,4’,6,6’-hexanitrobi-benzyl a with sodium ethoxide in ethanol at 25 “C [NaOEt] b / ~ OD (500 nm) observed calculated OD (500 nm)c 1 0.001 01 0.192 0.190 2 0.002 02 0.245 0.253 3 0.004 04 0.297 0.311 4 0.007 06 0.357 0.356 6 0.0099 0.388 0.384 6 0.0199 0.438 0.445 7 0.0398 0.509 0.509 8 0.0597 0.538 0.544 9 0.0796 0.560 0.566 10 0.099 0.582 0.581 11 0.149 0.607 0.604 12 0.198 0.623 0.617 a Concentration is 2.0 x 10-6~.In items 1-10 the solu-tions were made up to constant ionic strength, I O.IM, with sodium perchlorate. Calculated from equations (3)-(5) f with values of K, 1200 1 mold’, K,30 1 mol-l, ~(2; R = Et)1.67 x lo4,44; R = Et) 3.33 x lo4 1 mol-1 cm-’. Reaction with Sodium Ethoxide in Ethanol.-Ethoxide in ethanol is a more basic medium than methoxide in methanol and here there is evidence that the formation of (2; R = Et) is preceded by the formation of (1; R = Et), and for the presence at equilibrium of the diadduct (4;R = Et). Visible spectra were recorded for solutions containing 2 x 10-5~-parent and 0.001-0.2~-sodium ethoxide. The spectra recorded after two minutes showed the double absorption maxima characteristic of a-adduct f~rmation.~In the most dilute base concentration used the maxima were at 432 and 500 nm; in more concen- trated base solutions the higher energy band shifted to 438 nm.Acidification of the solutions at this stage resulted in the regeneration of the parent. With time the bands due to a-adducts were replaced by a band at 380 nm with broad shoulder at longer wavelength; acidification failed to regenerate the parent indicating that irreversible reaction had occurred. Values of optical density after completion of the re- action giving a-adducts but before decomposition had occurred are in Table 2. A Benesi-Hildebrand plot of the reciprocal of optical density versus the reciprocal of base concentration was curved indicating that the reaction is more complicated than simple 1 : 1 adduct 1982 formation. However a short extrapolation of the plot gave a value of 0.666 at 500 nm for complete conversion.From this we are able to calculate the values for the extinction coefficients for the species formed at the higher base concentrations. The values E 6.0 x lo4 (438 nm) and 3.33 x lo41 mol-l cm-l (500 nm) are approximately double those normally found for adducts of 1: 1 stoi-cheiometry.8 This strongly suggests that in the more concentrated base solutions used a diadduct is formed by ethoxide attack on the two benzene rings and in view of the n.m.r. evidence the most likely structure is (4;R = Et). Since the benzene rings are separated by two methylene groups it would be expected that the absorp- tion would be twice that of a 1 : 1adduct.Our data can be accommodated by the presence of two equilibria involving attack of ethoxide on one [equation (3)] or two rings [equation (4)]. Initially we obtained an approxi- mate value for K, by assuming that in items 1-3 of K(P) + OEt-4(2; R = Et) (3) (2; R = Et) + OEt-& K (4;R = Et) (4) (P) + (2; R = Et) + (4;R = Et) = 2.0 x (5) Table 2 there would be little of the diadduct present, and an approximate value for K, by assuming that in items 8-12 little parent would remain. By iteration we obtained values for K, of 1 200 200 1 mol-l and K, of 30 & 10 1 mol-l which gave a good fit of calculated with observed optical densities. The considerably lower value for K, indicates that even though the 2,4,6- trinitrobenzene rings are separated by two methylene groups attack of alkoxide on one ring to give a a-adduct inhibits attack on the second ring.Examination by stopped flow spectrophotornetry of solutions containing low (<0.01~) concentrations of base indicates the presence of two processes attributable to 1 : 1 interactions. The faster process which is of relatively low intensity gives rise to a species with maxima at 430 and 490 nm which is likely to be the (1; R = Et) (P) + OEt-(6) (2; R = Et) kfast = k,[OEt-] -tk-3 (7) kslow = k-I + k,[OEt-]/(l K3LOEt-I) (8) adduct (1; R = Et). The slower process giving an increase in intensity represents formation of the adduct (2; R = Et).Since the base concentration is in large excess over the parent concentration and the rates of the two processes are well separated it is readily shown by standard methods that equations (7) and (8) will apply. A linear plot of kfast versus base concentration gave values for k, 4 000 1 mot1 s-l and k-, 32 s-l. Combination of these values gives a value for K3of 125 1 mol-l in good agreement with that obtained from the optical densities at the completion of the fast process. The values of the slower rate process are accommodated by equation (8) with the values k-, 0.070 s-l, k, 84 1 mol-l s-l, and K, 125 1 mol-l. TABLE3 Kinetic data for 1 : 1 interaction of 2,2’,4,4‘,6,6’-hexanitro-bibenzyl(1 x 10-6~)with sodium ethoxide in ethanol at 25 “C [NaOEt]/ krastl OD K, bl kelowl kslor M s-l (430 nm) 1 mol-1 s-1 (calc) 0.0010 36.5 0.0039 110 0.14 0.14 0.0015 0.18 0.18 0.0020 41 0.0092 150 0.21 0.20 0.0025 0.23 0.23 0.0040 48 0.0137 130 0.30 0.30 0.0060 52 0.0164 115 0.33 0.36 0.0084 66 0.0209 130 0.0100 72 0.0218 120 After completion of the fast process. 2 mm path length cell.A Benesi-Hildebrand plot gives a value of 0.040 for complete conversion. b Calculated from OD (430)/[0.040 -OD (430)][NaOEt]. c Calculated from equation (8) with k-, 0.07 SO,k, 84 1 mol-l s-l, and K, 125 1 mol-l. Comparison with Related Compounds.--Rate and equilibrium constants for 1 : 1interaction are summarised in Table 4 where they are compared with data for 2,4,6- t rini trobenz y1 chloride (TNBCl), 2,4,6-t rinit rot oluene (TNT), and lJ3,5-trinitrobenzene (TNB).The data for 2,2’,4,4’,6,6’-hexanitrobibenzyl(HNBB) have not been statistically adjusted. The value of K, for ethoxide attack at the l-position of HNBB is ca. 10 times higher than the value of K, for attack at the 3-position. Similarly TNBCl shows a thermodynamic preference for attack at the l-position.* This contrasts with the behaviour of TNT where ts-adducts are only observed from attack at the 3-p0sition.~9~ One factor favouring alkoxide addition at the l-position of HNBB or TNBC1, relative to TNT, is the inductive electron withdrawal of the 2,4,6-trinitro- benzyl or chlorine substituents. However steric factors are likely to be of importance. Thus the greater the size of the group at the l-position the greater the pos- sibility of relief of steric strain as the group is bent from the ring-plane in the l-alkoxy-adduct.Kinetically, attack at the unsubstituted ring positions is favoured over attack at the substituted positions. In particular the values for the rate coefficients for the reverse reactions appear to be characteristic of the posi- tion of addition. Thus, values of k-, are at least 100 times lower than values of k,. The values of K, for attack at unsubstituted positions are lowered by the presence of a bulky substituent at the l-position. Thus for ethoxide additions the values decrease from TNB to TNBCl to HNBB. Nitro-groups will display their maximum electron-withdrawing ability when they are coplanar with the benzene ring and the presence of bulky substituents which cause their rotation from the ring-plane will be expected to cause a destabilis- ing effect on addition at unsubstituted ring-positions.A major reaction of both TNT and TNBCl with alk- J.C.S. Perkin I1 TABLE4 Comparison of rate and equilibrium data for HNBB with those for related compounds HNBB-methoxide k,/1 mol-1 s-1 k_,/s-l K,/1 molP k,/1 molP s-1 23 K-,/S-~ 1.15 KJ1 rnol-' 20 TNBC1-methoxide a <20 770 2.2 350 TNT-methoxide 6 280 3 000 0.07 TNB-methoxide C 7 300 330 20 HNBB-ethoxide 4 000 32 125 84 0.07 1 200 TNBCl-ethoxide a TNB-ethoxide c 10 000 40 000 14 20 700 2 000 7 000 (1 >10 000 a Ref.4. b Ref. 6. c C. F. Bernasconi, J. Awz. Chem. Soc., 1970, 92, 4682 (refers to addition at an uiisubstituted positioii) oxide ions is formation of the conjugate base by transfer of a side-chain pr~ton.~ With both these substrates proton-transfer is a much slower process than base addition. Our data for HNBB do not provide any evidence for formation of the conjugate base (3),the lH n.m.r. and visible spectra being in accord with a-adduct formation. We think it unlikely that the previously reported loassignment of bands at 460 and 500 nm to the vcarbanion (3) is correct. We do however have evi- dence l1for the presence in dimethyl sulphoxide contain- ing the tertiary base 1,4-diazabicyclo[2.2.2]octaneof a blue species with maximum at 650 nm which may be the anion (3) or some species derived from it.The failure to observe (3)in solutions of base in alcohol is probably due to the slowness of its formation rather than its thermodynamic instability. We have shown that irreversible reaction of HNBB occurs fairly rapidly so that decomposition may occur before there is appreci- able formation of (3). EXPERIMENTAL HNBB, m.p, 218-220 "C (lit3218--220"), was provided by the Ministry of Defence. Solutions of sodium alkoxides were prepared by solution of clean pieces of sodium in AnalaR alcohol and were titrated against standard acid. 1H N.m.r. measurements were made with a Varian EM 360L instrument using tetramethylsilane as internal refer- ence. Visible spectral measurements were made with Unicaiii SP 500 or 8000 instruments, or a Canterbury stopped-flow spectrophotometer.All kinetic measure-ments were made at 35 "cwith base in large excess over sub-strate concentration. We thank the Ministry of Defence for a, maintenance grant (1'. J. li.)and the S.1C.C. for a grant to purchase the stopped-flow [1/878 Itcceived, 1st Jitne, 19811 REFERENCES Part 27, 31. I<. Crampton, 1'. J. Routledge, and 1'. hl. Wilson, J. Chetn. Hes., 2981, (S)153, (M)1972. G. P. Sollott, M. Warman, and E. E. Gilbert, J. Ovg.Chem., 1979, 44, 3329. I<. G. Shipp and L. A. liaplan, J. Org. Chem., 1966,31, 857. 1).N. Brooke, M. K. Crampton, G. C. Corfield, P. Golding,and G. I;.Hayes, J. Ckem. Soc., Perktn Trans. 2, 1981, 526, and references therein. M. R. Crampton, Adv. Phys. Ovg. Chem., 1969, 7, 211; M. J. Strauss, Chem. Rev.,1970, 70, 667. D. N. Brooke and M. R. Crampton, J. Chem. Hes. 1980, (S)340, (M)4401. 7 H. A. Benesi and J. H. Hildebrand, J. Am. Chem. Soc., 1949, 71, 2703. 8 R. Foster and C. A. Fyfe, Rev. Pawe Appl. Chem., 1966, 16, 61. 8 C. A. Fyfe, C. D. Malkiewich, S. W. H. Damji, and A. H. Norris, J. Am. Chem. SOC.,1976, 98, 6983. lo C. Capellos, quoted in ref. 2. 11 M. R. Crampton and P. J. Routlcdge, unpublished observ- ation.
ISSN:1472-779X
DOI:10.1039/P29820000031
出版商:RSC
年代:1982
数据来源: RSC
|
|