J. Chem. Soc., Perkin Trans. 2, 1998 25 1H and 13C NMR studies of para-substituted benzaldoximes for evaluation of the electron donor properties of substituted amino groups † Ryszard Gawinecki,*,a Erkki Kolehmainen b and Reijo Kauppinen b a Department of Chemistry, Technical and Agricultural University, Seminaryjna 3 PL-85-326 Bydgoszcz, Poland b Department of Chemistry, University of Jyväskylä, PO Box 35, FIN-40351 Jyväskylä, Finland 1H and 13C NMR spectra of seventeen (E)-benzaldoximes and three acetophenone oximes, both carrying substituted p-amino groups, have been recorded and discussed from the point of view of substituent effect.The resonance effect of these substituents is not transmitted strongly to CH]] NOH group. However, it is found that the chemical shift of Cpara depends linearly on sR o values. This dependence has been used to calculate the resonance substituent constants for the less common amino groups and the 1-pyrrolidine group is found to be the most powerful electron donor among the substituents studied.Introduction Substituted amino groups, which are very interesting from the point of view of differentiation of their electron donor properties, comprise a wide range of members. Electronic effects of aromatic amino substituents depend on the geometry of the CAr]NR2 fragment and it is determined by CAr]N distance, R]N]R valence angle, twist (torsion) 2,3 and bend angles,4,5 both referring to mutual orientation of nN and pAr orbitals, and also to the dihedral angle between the NR2 and the ring planes (the pyramidalization or tilt angle 6).The s substituent constants of various amino groups differ almost exclusively due to the extent of benzene ring–nitrogen atom resonance.7 Since the pKa values of aniline derivatives involve contribution of steric inhibition to solvation,8 basicity does not show properly the resonance between such groups and the aromatic ring. Instead, other data such as intensities of the UV 8,9 and IR bands,2,10 dipole moments,11,12 polarographic data 12,13 and exaltations of molar refraction 4,14 are more useful in prediction of the electron donor strength of amino groups.The sR8 constant of the p-dimethylamino group based on the fluorine chemical shift in the spectrum of N,N-dimethyl-p-fluoroaniline, which is sensitive to very small perturbations in the p-charge density at fluorine atom produced by the substituent,15 is equal to 20.52.15 That group in N,N,2,6-tetramethyl-4-fluoroaniline is much twisted out of the ring plane but there is still some resonance interaction between the two parts of the molecule.16 Both the 19F chemical shift in the NMR spectrum of N,N,2-trimethyl-4- fluoroaniline 15 and intensity of the UV spectrum of ethyl N,N,3-trimethyl-4-aminobenzoate 17 show that the ortho methyl group produces a 56% steric inhibition of the resonance in the Ar]N fragment and this results in a reduction of the sR8 constant of this p-dimethylamino group to 20.24.15 The chemical shift of the para carbon predicts well the electron donor strength of the amino groups.8,18 It shows N-phenylaziridine to be the least conjugated of all N-phenyl cyclic amines studied.4,8,19 It seems noteworthy that two ortho methyl groups decrease the N]methyl one-bond 13C]1H coupling constant in the spectra of N-methyl- and N,N-dimethyl-aniline by about 0.5 and 1.5 Hz, respectively.20 UV spectral data of p-aminonitrobenzenes,9 p-aminoazobenzenes, 21 p-aminobenzenes 4 and complexes of amino- † For a preliminary report of this work, see ref. 1. benzenes with 1,3,5-trinitrobenzene,14 IR spectral data of aminobenzenes,2,10 19F chemical shifts of fluoroanilines,15 acidities of p-aminobenzoic acids,13,22 rates of reduction of p-nitroanilines,13 polarographic halfwave potentials of pnitroanilines, 13 polarographic oxidation potentials of aminobenzenes 14 and dipole moments of aromatic amines 11,12 all show the following order of electron-donor strength of amino substituents: (CH2)4N > Et2N>Me2N > (CH2)5N > NHMe > NH2 > N(CH2)2.Alkane bridges between Cortho and Namino, e.g. in julolidine derivatives, cause the nitrogen atom to reveal stronger donor properties than that in the 1-pyrrolidino group.23 The published 1H, 13C, 15N and 17O NMR spectra of aromatic oximes 24–30 have been mainly concerned with the configuration and with transmission of substituent effects to the CH]] NOH group through the benzene ring.As found,30 there are poor or very poor correlations between the 17O NMR chemical shifts of the oximino oxygen and s, s1 and s1 substituent constants for substituted benzaldoximes. Electron acceptor properties of the CH]] NOH group (sR = 0.10) 7 preclude cross-conjugation to be important in p-aminobenzaldoximes. Thus, such compounds seem to be a very good model series to study the substituent effect. Since electronic properties of the less common amino groups are not known, the 1H and 13C NMR chemical shifts of p-aminobenzaldoximes are used in the present paper to evaluate their resonance substituent effects.The compounds under study have the formulae 1–20. Experimental Syntheses All melting points are uncorrected. The boiling points of all reaction products are expressed in the 8C/mmHg units. Most amines used in formylations are commercially available and known procedures were used to prepare others.31,32 Lilolidine‡ bp 139–141/10 (lit.bp 90–100/0.5,33 112–113/5 34) was prepared in overall 36% yield by reduction (LiAlH4, standard procedure) of 4-oxolilolidine which, in turn, was obtained by ‡ The IUPAC name for lilolidine is 1,2,5,6-tetrahydro-4H-pyrrolo[3,2,1- ij]quinoline. The name benzo[h,i]indolizidine has previously been used in the literature for lilolidine.33,3426 J. Chem. Soc., Perkin Trans. 2, 1998 cyclization–acylation of 1-(b-chloropropionyl)indoline.34 To synthesise 1-methylindoline, 1-formylindoline (bp 168–173/15, lit.bp 115–117/2 35) was first prepared in 94% yield from indoline and formic acid following the procedure used in formylation of N-methylaniline.36 It was then reduced (LiAlH4, standard procedure, 85% yield) to a product of bp 91–94/9 (lit. bps of 1-methylindoline are 68–73/1,37 120/12 38). Kairoline (1- methyl-1,2,3,4-tetrahydroquinoline), bp 102–104/4 (lit. bp 160/ 12,38 80–81/0.4 39) was obtained in 92% yield by reduction (LiAlH4, standard procedure) of 1-formyl-1,2,3,4-tetrahydroquinoline which, in turn, was prepared in 80% yield by reductive formylation of quinoline with formic acid according to ref. 40. 1-Methyl-2,3-benzohexamethyleneimine (bp 120–125/16, lit. bp 59–60/0.2,41 160/12 38) was obtained in a five-step synthesis (overall yield 34.5%) involving oximation of a-tetralone, 42 tosylation of a-tetralone oxime,43,44 Beckmann rearrangement of the O-tosyl derivative of a-tetralone oxime (for details see synthesis of its O-benzenesulfonyl derivative 44), methylation of homohydrocarbostyril 41,45 and reduction (LiAlH4) of 1-methyl-homohydrocarbostyril.41,45 Some aldehydes and ketones were commercial products.p-(N,N-Dimethylamino)acetophenone was a gift from Dr Tomasz Ba�k. Other aldehydes were obtained by the Vilsmeier– Haack46 or Duff 47 methods and were purified by vacuum distillation or crystallization from aqueous ethanol. Detailed synthetic procedures will be given in another paper.48 p-Aminobenzaldoxime was obtained in 47% yield by reduction of p-nitrobenzaldoxime, according to the procedure used in synthesis of p-aminoacetophenone oxime.49 The product was purified by crystallization from aqueous ethanol.Other aldoximes were obtained from their respective aldehydes by the standard procedure 50 and crystallized from aqueous ethanol. The yields were 26–91% (no attempts were made to improve the reaction efficiency).The synthetic procedure for ketoximes was slightly different. Thus, the mixture of the appropriate acetophenone (0.08 mmol), hydroxylamine hydrochloride (11.1 g, 0.16 mol), 96% aqueous ethanol (60 ml) and conc. hydrochloric acid (few drops) was refluxed for 1.5 h. The reaction mixture was then diluted with water (1 ml) and extracted with diethyl ether. Ketoximes were prepared in 51–69% yield by evaporation of solvent from the extract and recrystallization of the residue from 96% ethanol.Melting points (8C) of oximes: 1, 126–130 (127–128 51); 2, 96.5–96.9; 3, 147–148 (145–147 52); 4, N R5 R6 C R1 R2 R3 R4 R7 NOH 3 4 5 6 1 7 2 1 2 3 4 5 6 7 8 R1 H H H H H Me Me H R2 H H H H H H H Me R3 H H Me Me Et Me Me Me R4 H Me Me Et Et Me Me Me R5 H H H H H H H H R6 H H H H H H Me H R7 H H H H H H H H 9 10 11 H H H H H H (CH2)4 (CH2)5 (CH2)6 H H H H H H H H H 12 13 14 15 H H H H (CH2)2 (CH2)3 (CH2)3 (CH2)4 Me H Me Me H H H H H H H H H H H H 16 17 H H (CH2)2 (CH2)3 (CH2)3 (CH2)3 H H H H 18 19 H H H H H Me H Me H H H H Me Me 20 H H (CH2)5 H H Me 117–121; 5, 89–91 (93 53); 6, 105–109 (107–109 52); 7, 134–135; 8, 97–98 (97–98 52); 9, 190–192; 10, 161–163; 11, 110–114; 12, 97–99; 13, 147–151; 14, 87–88; 15, 107–109; 16, 104–106.5; 17, 126–128 (127–128 54); 18, 150–152 (153–154 55); 19, 218–222; 20, 163–116. NMR spectroscopy NMR spectra of the saturated solutions of oximes in [2H6] acetone were recorded on a JEOL JNM GSX-270 FT NMR spectrometer working at 270.17 and 67.94 MHz for 1H and 13C observation, respectively.TMS (internal reference) and [2H6] acetone (lock) were used both in 1H and 13C NMR experiments. Other conditions are: 1H: spectral width 3500 Hz, 32 K data points, digital resolution 0.21 Hz/point, pulse width 9.4 ms, flip angle 90 deg, number of scans 4, pulse delay 1 s, pulse sequence SGNON; 13C: spectral width 15 000 Hz (1H decoupled)/10 000 Hz (1H coupled), 32 K data points (1H decoupled)/64 K data points (1H coupled), digital resolution 0.92 Hz/point (1H decoupled) 0.32 Hz/point (1H coupled), pulse width 7.8 ms, flip angle 90 deg, number of scans 100–400 (1H decoupled)/ca. 9000 (1H coupled), pulse delay 4 s, pulse sequence/decoupling SGBCM/continuous bilevel SGNON (1H coupled). In order to analyse accurately the 1H NMR spectra, resolution enhancement was performed by combined exponential and trapezoidal windowing (T2 = 5% and T3 = 50%) and zero filling until the digital resolution was <0.05 Hz.The data matrix for 13C–1H HETCOR was as follows: 10 000 Hz and 1024 points for the 13C-axis and 1800–2000 Hz and 256 points for the 1H-axis. An average value of 1J(C,H) = 125 Hz was used for the correlation between coupled nuclei. Assignments of the signals in the aliphatic part of the spectra were possible from their homonuclear 1H–1H DFQ–COSY experiment. The signals of aromatic quaternary carbon atoms for compounds 9, 13, 15, 16 and 17 were assigned with the help of the 2D 13C–13C INADEQUATE spectra (solutions in [2H6] acetone).Data matrix: 10 000 Hz and 1024 points in the f1-axis and 20 000 Hz and 256 points in the f2-axis. The number of scans was 256 (8*32). An average value of 1J(C,C) = 36 Hz was used for the correlation between coupled nuclei. The 2D 13C–13C INADEQUATE correlation map of oxime 16 was recorded in saturated [2H6]DMSO solution at 30 8C with a Bruker Avance DRX500 spectrometer working at 125.76 MHz equipped with a 5 mm broadband direct detection probehead.The spectral width was 23 000 Hz (180 ppm), number of scans 160 and composite pulse decoupling (WALTZ-16) was used to decouple protons during the pulse sequence. The delay transmitting the correlation between coupled neighbouring 13Cnuclei was set to correspond the direct coupling constants of 1J(13C,13C) = 36 Hz. Results and discussion All oximes studied are the syn, i.e. E isomers. Although a coplanar arrangement of the ring and the CH]] NOH group in compounds carrying two methyls ortho to the oxime group may lead to serious steric repulsion, oxime 7 has also been assigned the E configuration.24 The 1H and 13C NMR chemical shifts for the oximes 1–20 are collected in Tables 1, 2, 4 and 5.As seen, the chemical shift of C7 changes in a narrow range (1.48 ppm) from 148.65 for 6 to 150.13 ppm for 16. The values of dC7 in the spectra of ketoximes 18–20 are >154 ppm. There is no linear relationship between the chemical shift of C7 and sR o substituent constants of the different amino groups for the compounds studied.It is known27,29,30 that the chemical shift of C7 in the spectra of p-substituted benzaldoximes depends mainly on the substituent inductive effect and its resonance is of reduced importance. Moreover, multiparameter correlations of dC7 with inductive and resonance substituent constants 56 and with semiempirical parameters that represent the paramagnetic interactionJ.Chem. Soc., Perkin Trans. 2, 1998 27 Table 1 13C chemical shifts of aromatic and a-methine carbons in the spectra of oximes 1–20 (d in ppm from TMS, in [2H6]acetone) 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 16 c 17 18 19 20 C1 122.57 122.05 121.86 121.43 121.01 120.10 119.55 128.02 121.10 123.98 121.01 123.29 121.05 a 123.22 126.09 124.10 122.84 121.00 126.61 126.01 128.05 C2 128.74 128.73 128.52 128.70 128.86 138.07 139.07 130.05 128.67 128.47 128.87 122.55 128.67 127.54 126.34 126.74 125.68 d 126.34 127.58 127.42 127.33 C3 114.91 112.51 112.75 112.58 112.17 114.55 113.14 132.53 112.35 116.11 111.83 131.52 121.56 a 121.41 135.44 129.57 128.45 121.80 114.65 122.62 115.95 C4 150.73 152.06 152.24 150.81 149.35 151.91 151.21 154.92 146.45 153.59 150.59 155.42 147.62 148.46 154.64 b 151.01 144.70 149.97 151.89 153.14 C5 114.91 112.51 112.75 112.58 112.17 110.78 113.14 119.07 112.35 116.11 111.83 106.75 114.25 111.08 116.76 119.16 118.03 121.80 114.65 112.62 115.95 C6 128.74 128.73 128.52 128.70 128.86 128.75 139.07 126.05 128.67 128.47 128.87 128.41 126.43 127.07 128.79 120.90 119.83 e 126.34 127.58 127.42 127.33 C7 149.61 149.62 149.42 149.46 149.50 148.65 149.16 149.20 149.63 149.23 149.46 149.80 149.68 149.62 149.31 150.13 148.88 f 149.61 154.32 154.07 154.01 a Signals may be interchanged. b Due to limited solubility of this compound in acetone, the quality of its INADEQUATE spectra is poor and this signal is not seen in the spectrum. c In [2H6]DMSO at 125.758 MHz.d 1J(C2,H2) = 155.2 Hz. e 1J(C6,H6) = 158.14 Hz. f 1J(C7,H7) = 160.4 Hz. between the substituent and carbon atom57 are of better quality.27 The chemical shift of C1, i.e. the para carbon atom with respect to the amino substituent is the most appropriate from the point of view of electron donor strength of the amino substituent. Since sensitivity of 2D 13C–13C INADEQUATE experiment for compounds 9, 13, 15, 16 and 17 was good enough (see Fig. 1), the respective spectra were recorded to distinguish between the signals of different quaternary carbon atoms. The 13C chemical shifts for both aldoximes 1–17 and ketoximes 18–20 vary from 119.55 for 7 and 120.10 for 6 to 128.02 for 8 and 128.05 ppm for 20. Thus, the chemical shift dispersion of dC1, i.e. 8.50 ppm, is wider than that for dC7. Although, resonance and inductive effects of the para substituent on dC1 in the spectra of benzaldoximes are comparable by their magni- Table 2 13C chemical shifts of the side-chain carbons in the spectra of oximes 1–21 (d in ppm from TMS, in [2H6]acetone) 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 16 g 17 18 19 20 NC 29.83 40.27 46.93 a 37.41 b 44.85 40.20 40.22 44.10 48.12 50.17 49.81 56.13 a 35.41 b 42.17 51.53 a 38.81 b 57.18 a 43.01 b 47.20 c 55.08 e 46.02 c 54.13 e 50.49 — 40.38 50.12 NCC — — 11.44 12.86 — — — 26.01 26.31 28.12 28.67 27.79 22.81 30.12 23.54 d 28.80 f 22.40 d 27.85 f 28.40 — — 26.26 NC2C and NC3C — — — — — — — — 25.07 27.56 — 22.56 28.39 26.13 35.52 24.54 23.45 22.62 — — 25.07 CCH3 — — — — 20.63 22.27 18.68 — — — — — — — — — — 11.31 11.22 11.26 a NCH2.b NCH3. c NCH2CH2CH2. d NCH2CH2CH2. e NCH2CH2. f NCH2CH2. g In [2H6]DMSO at 125.758 MHz. tudes,27 only dCl values in the spectra of p-aminobenzaldoximes studied were found to be linearly dependent on sR o. The correlation obtained is: dC1 = 132.44 1 19.34sR o for compounds 1–3, 5 and 8–10 (correlation coefficient = 0.974, standard deviation = 2.45, standard error = 0.93).From its decreased value of sR o it is clear that the dimethylamino group in 8 is considerably twisted out of the ring plane. In order to determine the resonance effect of the less common amino groups, the above equation was used to calculate the respective sR o values. They are given in Table 3. The chemical shift of C4, i.e. the ipso carbon atom with respect to the amino substituent, for compounds 1–20 changes in a very wide range from 130.88 to 155.42 ppm (Dd = 24.54 ppm) but no simple dependence between dC4 and the type of substituent was found.It should be mentioned that multiparameter correlation between the chemical shift of C4, i.e. the ipso carbon, and the substituent constants in the NMR spectra of p-substituted benzaldoximes was observed.27 Although the contribution of the inductive effect is more than seven times that of the resonance effect, the paramagnetic interaction between the substituent and C4 is that which most contributes to dC4.57 Fig. 1 The 2D 13C–13C INADEQUATE correlation map of oxime 16 in saturated [2H6]DMSO solution at 30 8C28 J. Chem. Soc., Perkin Trans. 2, 1998 The chemical shift of H7 and that of the oxime proton, NOH, in the spectra of oximes studied changes in a similar narrow range, i.e. 0.49 and 0.43 ppm, respectively. Although, dC7 for both E and Z ring-substituted benzaldoximes correlates well with the s constants,29 no linear relationship between the shift of H7 and NOH, and sR o values was found for the compounds studied.The results obtained show that the resonance effect of p-amino groups is not transmitted strongly to the CH]] NOH group which is in disagreement with the X-ray studies on p-dimethylaminobenzaldoxime.58 Since the inductive substituent constants of amino groups are scarce and those that are available differ only slightly from each other, no correlation between the shift of C7 and s1 values can be obtained (such a procedure was used for some other p-substituted benzaldoximes 28).X-Ray determination 58 shows that the molecule p-(N,Ndimethylamino) benzaldoxime is planar with angles /CMeNCMe9 and /CMeNCAr equal to 115.58, and 121.78 (120.78), respectively. The CAr]N distance (1.380 Å) indicates this bond to have significant double bond character. The angle /CMeNCMe shows the amine nitrogen atom to have sp2 hybridization. Other bond lengths and valence angles confirm also that there is an electron Table 3 sR o substituent constants of amino groups R NH2 NHMe NMe2 N(Me)Et NEt2 1-NMe2; 3-Me 1-NMe2; 3,5-Me2 1-NMe2; 2-Me N(CH2)4 N(CH2)5 N(CH2)6 1-N(Me)[2-(CH2)2] 1-NH[2-(CH2)3] 1-N(Me)[2-(CH2)3] 1-N(Me)[2-(CH2)4] 1-N[2-(CH2)2][6-(CH2)3] 1-N{2,6-[(CH2)3]2} sR o 20.48 a 20.52 a 20.53 a 20.57 b 20.57 a 20.64 b 20.67 b 20.24 a 20.63 a 20.47 a 20.59 b 20.47 b — 20.48 b 20.33 b 20.43 b 20.59 b a Literature values.2,10,15 b Calculated from equation sR o = (dC1 2 132.44)/ 19.34.Table 4 1H chemical shifts of aromatic, a-methine and oximino protons in the spectra of oximes 1–20 (d in ppm from TMS, in [2H6] acetone) 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 H2 7.32 7.37 7.44 7.42 7.40 — — 7.40 7.41 7.44 7.40 7.34 7.11 7.16 7.35 6.96 6.95 7.42 7.54 7.54 H3 6.66 6.59 6.71 6.69 6.68 6.54 6.44 — 6.56 6.92 6.71 — — — — — — 6.64 6.71 6.90 H5 6.66 6.59 6.71 6.69 6.68 6.56 6.44 7.01 6.56 6.92 6.71 6.41 6.45 6.53 6.88 — — 6.64 6.71 6.90 H6 7.32 7.37 7.44 7.42 7.40 7.49 — 7.35 7.41 7.44 7.40 7.21 7.13 7.22 7.33 7.16 6.95 7.42 7.54 7.54 H7 7.96 7.96 7.99 7.99 7.96 8.23 8.34 8.02 7.97 8.00 7.96 7.98 7.90 7.93 8.03 7.93 7.85 — — — NOH 9.67 9.63 9.71 9.69 9.62 9.73 9.83 9.98 9.63 9.80 9.63 9.73 9.56 9.65 9.95 9.63 9.55 ~9.7 a 9.75 9.87 a Very weak signal.transfer from the amino nitrogen to oximino oxygen. However, the NMR results presented in Tables 1 and 4 show this is not the case for p-(N,N-dimethylamino)benzaldoxime and the other oximes studied.Moreover, those data indicate that hybridization of the amino nitrogen atom in the compounds studied is more sp3 than sp2-like. The conformation of indoline and of its homologs is not known. Absorption bands in the spectra of N-alkylindolines have significantly reduced intensity, and are red shifted, as compared to the spectra of N,N-dialkylanilines.59 Both differences in hybridization of nitrogen atoms in those compounds and conformational equilibria of the five-membered ring in indoline can account for this behaviour but no definitive explanation of extent of the benzene ring–nitrogen resonance in that compound was given.59 On the other hand, molecular models show that hybridization of the N atom in indolines can be both of sp2 and sp3 type, and the five-membered ring can be both planar and puckered.Moreover, deformation of the benzene ring in anilines carrying short CAr]N bridges can also occur.The derived sR o values show piperidino to be the most powerful donor among the studied amino substituents. The propane bridge between the amino N atom and ortho position was found to enable the resonance to be more effective than that in the systems that contain shorter and longer bridges. Finally, in disagreement with earlier work,23 the sR o values show that the amino nitrogen atom in julolidine is a weaker electron donor than that in pyrrolidine.It is noteworthy that the observed order of the substituents resembles, in general, that based on values of EHOMO and positions of the CT bands in the spectra of complexes of aminobenzenes with 1,3,5-trinitrobenzene,14 UV spectral data of p-aminoazobenzenes,21 dipole moments of aromatic amines,11,12 polarographic oxidation potentials of aminobenzenes 14 and polarographic halfwave potentials of p-nitroanilines.13 Other spectral studies, now in progress, are expected to show the accuracy of estimation of the s values obtained.Acknowledgements Financial support provided by the (Polish) Committee for Table 5 1H chemical shifts of the side-chain protons in the spectra of oximes 1–20 (d in ppm from TMS, in [2H6]acetone) 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 NH 4.90 5.24 — — — — — — — — — — 5.30 — — — — ~4.4 h — — NCH — 2.83 2.96 3.43 a 2.93 b 3.41 2.95 2.94 2.69 3.28 3.23 3.51 3.31 a 2.74 b 3.29 3.23 a 2.87 b 2.92 a 2.84 b 2.98 d 3.28 f 3.18 — 2.96 3.20 NC2H — — — 1.09 1.15 — — — 2.00 1.64 c 1.78 2.89 1.86 1.91 1.71 2.04 e 2.87 g 1.92 — — 1.63 d NC3H and NC4H — — — — — — — — — 1.64 c 1.53 — 2.71 2.69 1.58 2.74 2.62 2.69 — — 1.63 d CCH3 — — — — — 2.37 2.35 2.29 — — — — — — — — — 2.14 2.15 2.16 a NCH2.b NCH3. c Centre of the multiplet. d NCH2CH2CH2. e NCH2- CH2CH2. f NCH2CH2. g NCH2CH2. h Very weak signal.J. Chem. Soc., Perkin Trans. 2, 1998 29 Scientific Research [KBN, grant 3 T09A 147 09] (R.G.) and the Academy of Finland (E. K.) is gratefully acknowledged. References 1 Preliminary communication: 13th IUPAC Conference on Physical Organic Chemistry, (Inchon, Korea), 1996, p. 74. 2 A. R. Katritzky and R. D. Topsom, Angew. Chem., Int. Ed. Engl., 1970, 9, 87. 3 M. D. Rozeboom, K. N. Houk, S. Searles and S. E. Seyedrezai, J. Am. Chem. Soc., 1982, 104, 3448. 4 A. T. Bottini and Ch. P. Nash, J. Am. Chem. Soc., 1962, 84, 734. 5 C. Cauletti, G.Cerichelli, F. Grandinetti, L. Luchetti and M. Speranza, J. Phys. Chem., 1988, 92, 2751. 6 A. Hastie, D. G. Lister, R. L. McNeil and J. K. Tyler, J. Chem. Soc., Chem. Commun., 1970, 108. 7 O. Exner, Critical Compilation of Substituent Constants in Correlation Analysis in Chemistry: Recent Advances, ed. N. B. Chapman and J. Shorter, Plenum Press, New York, 1978, p. 439. 8 C. P. Nash and G. E. Maciel, J. Phys. Chem., 1964, 68, 832. 9 M. J. Kamlet, J. L. Abboud and R. W. Taft, J.Am. Chem. Soc., 1977, 99, 6027. 10 R. T. C. Brownlee, R. E. J. Hutchinson, A. R. Katritzky, T. T. Tidwell and R. D. Topsom, J. Am. Chem. Soc., 1968, 90, 1757. 11 K. S. F. Beach, J. D. Hepworth, J. Sawyer, G. Hallas, R. Marsden and M. M. Mitchel, J. Chem. Soc., Perkin Trans. 2, 1984, 217. 12 D. Mazet, W. D. Weringa and H. Lumbroso, C. R. Seances Acad. Sci., 1970, 270C, 1537. 13 W. D. Weringa and M. J. Janssen, Recl. Trav. Chim. Pays-Bas, 1968, 87, 1372. 14 F. Effenberger, P.Fischer, W. W. Schoeller and W.-D. Stohrer, Tetrahedron, 1978, 34, 2409. 15 R. W. Taft, E. Price, I. R. Fox, I. C. Lewis, K. K. Andersen and G. T. Davis, J. Am. Chem. Soc., 1963, 85, 3146. 16 M. J. S. Dewar and Y. Takeuchi, J. Am. Chem. Soc., 1967, 89, 390. 17 R. W. Taft and H. D. Evans, J. Chem. Phys., 1957, 27, 1427. 18 P. C. Lauterbur, J. Chem. Phys., 1963, 38, 1415. 19 R. G. Pews, J. Am. Chem. Soc., 1967, 89, 5605. 20 C. H. Yoder, B. A. Kaduk and R. E. Hess, Tetrahedron Lett., 1970, 3711. 21 G. Hallas, R. Marsden, J. D. Hepworth and D. Mason, J. Chem. Soc., Perkin Trans. 2, 1984, 149. 22 B. van de Graaf, H. J. Hoefnagel and B. M. Wepster, J. Org. Chem., 1981, 46, 653. 23 R. W. Castelino and G. Hallas, J. Chem. Soc., B, 1971, 793. 24 M. Christl, J. P. Warren, B. L. Hawkins and J. D. Roberts, J. Am. Chem. Soc., 1973, 95, 4392. 25 C. P. J. Vuik, M. ul Hassan and C. E. Holloway, J. Chem. Soc., Perkin Trans. 2, 1979, 1214. 26 R. E. Botto, P. W. Westerman and J.D. Roberts, Org. Magn. Reson., 1978, 11, 510. 27 G. V. Rutkovskii and V. P. Zmeikov, Zh. Org. Khim., 1987, 23, 142. 28 M. S. Gordon, S. A. Sojka and J. G. Krause, J. Org. Chem., 1984, 49, 97. 29 I. Pejkovic�-Tadic�, M. Hranislavljevic�-Jakovljevic�, S. Nes¡ic�, C. Pascual and W. Simon, Helv. Chim. Acta, 1965, 48, 1157. 30 G. Cerioni and A. Plumitallo, Magn. Reson. Chem., 1993, 31, 320. 31 C. F. van Duin, Rec. Trav. Chim., 1932, 51, 878. 32 L. C. Craig and R. M. Hixon, J.Am. Chem. Soc., 1930, 52, 804. 33 B. D. Astill and V. Boekelheide, J. Org. Chem., 1958, 23, 316. 34 G. Hallas and D. C. Taylor, J. Chem. Soc., 1964, 1518. 35 J. T. Shaw and F. T. Tyson, J. Am. Chem. Soc., 1956, 78, 2538. 36 L. F. Fieser and J. E. Jones, Org. Synth., 1955, Coll. Vol. 3, 590. 37 G. W. Gribble and Ch. F. Nutaitis, Synthesis, 1987, 709. 38 H. Ahlbrecht, E. O. Düber, J. Epsztajn and R. M. K. Marcinkowski, Tetrahedron, 1984, 40, 1157. 39 F. Minisci, R. Galli and M. Perchinunno, Org. Prep. Proced. Int., 1969, 1, 77. 40 A. N. Kost and L. G. Yudin, Zh. Obshch. Khim., 1955, 25, 1947. 41 L. A. Paquette, D. E. Kuhla and J. H. Barrett, J. Org. Chem., 1969, 34, 2879. 42 R. Huisgen, W. Rapp, I. Ugi, H. Walz and E. Mergenthaler, Liebigs Ann. Chem., 1954, 586, 1. 43 G. Schroeter, Ber., 1930, 63, 1308. 44 R. Huisgen, I. Ugi, H. Brade and E. Rauenbusch, Liebigs Ann. Chem., 1954, 586, 30. 45 G. Baddeley, J. Chadwick and H. T. Taylor, J. Chem. Soc., 1956, 451. 46 E. Campaigne and W. L. Archer, Org. Synth., 1963, Coll. Vol. 4, 331. 47 C. Berres and W. Müller, Swiss Patent 430 677/1967; C. Berres and W. Müeller, French Patent 1, 377 226/1964; Chem. Abstr., 1965, 62, P7691f. 48 R. Gawinecki, S. Andrzejak and A. Pucha�a, in preparation. 49 T. Satoh, N. Mitsuo, M. Nishiki, Y. Inoue and Y. Ooi, Chem. Pharm. Bull., 1981, 29, 1443. 50 Vogel’s Textbook of Practical Organic Chemistry, 4th edn., Longman, London, 1978, p. 810. 51 S. Witek, J. Bielawski and A. Bielawska, Polish J. Chem., 1981, 55, 2589. 52 D. J. Gale and J. F. K. Wilshire, Aust. J. Chem., 1975, 28, 2447. 53 J. C. Duff, J. Chem. Soc., 1945, 276. 54 P. A. S. Smith and T.-Y. Yu, J. Org. Chem., 1952, 17, 1281. 55 J. Sykulski, K. Lesz and U. Tomera, Polish J. Chem., 1979, 53, 2013. 56 C. G. Swain and E. C. Lupton, J. Am. Chem. Soc., 1968, 90, 4328. 57 N. B. Smith and T. W. Proulx, Org. Magn. Reson., 1976, 8, 567. 58 F. Bachechi and L. Zambonelli, Acta Crystallogr., Sect. B, 1972, 28, 2489. 59 W. Rettig, K. Rotkiewicz and W. Rubaszewska, Spectrochim. Acta, Part A, 1984, 40, 241. Paper 7/05668K Received 4th August 1997 Accepted 25th Se