Addition reactions of heterocyclic compounds. Part LXIII. New structures for some 2 : 1 molar adducts from dimethyl acetylenedicarboxylate with thiazoles and benzo-imidazoles, -oxazoles, and -thiazoles formed by novel rearrangement. Crystal and molecular structure determinations for tetramethyl 3,8a-dimethylpyrido2,1-bthiazole-5,6,7,8-tetracarboxylate and tetramethyl 5-methylpyrido2,1-bthiazole-6,7,8,8a-tetracarboxylate

摘要

1976 1269Addition Reactions of Heterocyclic Compounds. Part LX1II.l New Struc-tures for Some 2 :I Molar Adducts from Dimethyl Acetylenedicarboxylatewith Thiazoles and Benzo-imidazoles, -oxazoles, and -thiazoles formed byNovel Rearrangement. Crystal and Molecular Structure Determinationsfor Tetramethyl 3,8a- Dimethyl pyrido 2,1 -b t h iazole-5,6,7,8-tetracarb-oxylate and Tetramethyl 5- M ethylpyrido 2,1-b t hiazole-6,7,8,8a-tet ra-car boxylateBy Patrick J. Abbott, R. Morrin Acheson,rsquo; and Ulli Eisner, Department of Biochemistry, South Parks Road,David J. Watkin and J. Robert Carruthers, Department of Chemical Crystallography, South Parks Road,2.4-Dimethylthiazole with dimethyl acetylenedicarboxylate gives tetramethyl 3,8a-dimethylpyrido2,1-6 thiazole-5.6.7.8-tetracarboxylate, whereas thiazole and other alkylthiazoles give tetramethyl pyrido2,1-6 thiazole-6,7,8,8a-tetracarboxylates, rearrangements having taken place.Similar rearrangements can occur in the benzothiazole andbenzimidazole series, and earlier formulations proposed for a number of adducts from several heterocycles with theacetylenic ester have now been revised on the basis of 13C and l H n.m.r. spectra. The structures of the adductsfrom Z-methyl- and 2.4-dimethyl-thiazoles have been established by X-ray diffraction studies.Oxford OX1 3QUOxford OX1 3QSVARIOUS types of product have been obtained2rdquo; bytreating dimethyl acetylenedicarboxylate with thiazoles.The structure of the 2 : 1 molar adducts are the subjectof the present investigation.assigned structuressuch as (1) and (2) to the products from the ester withthiazole and its 2- and 4-methyl derivatives, largely onReid, Skelton, and BonthroneE =C02Me in all formulaethe basis of the presence of a low-field proton (7 ca.1.8) or methyl resonance in their n.m.r.spectra, but wereunable to distinguish between these possibilities. Atthe same time they allocated structure (4) to the cor-responding adduct from 2,4-dimet hylthiazole, as a high-field methyl signal was present. We considered thatall these adducts were best represented by structuresbased on (3), since their U.V. spectra were very similar,and because these spectra closely resembled thoseof t et ramet hyl9aH-quinolizine-l , 2,3,4-t et racarboxylatese.g.(21) and (22).5 The n.m.r. spectra, however, werenot explained convincingly. In view of the inconclusiveposition reached, it was decided to reinvestigate thesethiazole derivatives, and some of our new conclusionshave been published in preliminary form.6We have confirmed the findings of Reid et aL3 that theproton n.m.r. spectrum of the 2,4-dimethylthiazolePart LXII, P. J. Abbott, R. M. Acheson, amp;I. Y. Kornilov,and J. K. Stubbs, J.C.S. Perkin I , 1975,2322.R. M. Acheson, M. W. Foxton, and G. R. Miller, J . Chew.Soc., 1965, 3200.D. H. Reid, F. S. Skelton, and W. Bonthrone, TetrahedronLetters, 1964, 1797; W. Bonthrone, F. S. Skelton, and D. H. Reid,lsquo; N.M.R. in Chemistry,rsquo; ed. B. Pesce, Academic Press, New York,1965, p. 263.H.Ogura, H. Takayanagi, K. Furuhata, and Y . Iitaka,J.C.S. Chel-n. Comm., 1974, 759.adduct shows high-field methyl resonances (T 7.98 and8.57) not shown by the analogous thiazole adducts.The 13C n.m.r. spectrum of the 2,4-amp;methylthiazoleadduct was not sufficiently different from those of itsanalogues to be used as a basis for argument, and thestructure of this adduct was finally confirmed as (4) byan X-ray crystallographic analysis.The adducts from thiazole2V3 and its 2-meth~1,~ 4-methyl,rdquo;3 amp;methyl, and 2,5-dimethyl derivatives havevery closely related spectra and it may be assumed thatthey are structurally analogous. The lH signal due tothe methyl group from the 2-position of the originalthiazole appears at T ca. 7.5, in contrast to that in the2,4-dimethylthiazole adduct, but all the adducts show ansfi3 carbon resonance at 8 ca.75 in their 13C spectra.Examination of the 13C spectra leads to the conclusion,confirmed by an X-ray crystal analysis of the adductfrom 2-methylthiazole, that this group of compoundspossesses structures (8)-(12).The above adducts are presumably formed by therearrangement outlined in the Scheme. The initialformation of a compound such as (3) in the reaction isexpected on the basis of many analogies,rsquo; and its re-arrangement to the isomer (8) could take place via aconcerted 1,5 suprafacial sigmatropic shift, whichmight be permitted for a thermal reaction since an atomwith unbonded electrons is involved. An alternativenon-concerted pathway could proceed via the zwitterion(6), for which stable analogues such as (13) are known,*and there are a number of reports of similar openings ofthiazolium rings.9 The proton resonances (T 7.42) dueR.M. Acheson, A. R. Hands, and M. J. Woolven, J . Chem.Soc., 1963, 2082; R. M. Acheson and G. A. Taylor, ibid., 1960,1691.P. J. Abbott, R. M. Acheson, U. Eisner, D. J. Watkin, andJ. R. Carruthers, J.C.S. Chem. Comm., 1975, 155.R. M. Acheson, Adv. Heterocyclic Chem., 1963, 1, 125. * R. M. Acheson and I. A. Selby, J . Chem. Soc. ( C ) , 1971, 691.J. M. Sprague and A. H. Land, in lsquo; Heterocyclic Compounds rsquo;vol. 5, ed. R. C. Elderfield, Wiley, New York, 1957, p. 484;F. Krohnke and W. Friedrich, Chem. Ber., 1963, 96, 1195; cf.G. Bartoli, M.Fiorentino, F. Ciminale, and P. E. Todesco, J.C.S.Chem. Comm., 1974,732; K. T. Potts, D. R. Choudhury, andT. R.Westby, J . Org. Chew., 1976, 41, 1871270 J.C.S. Perkin Ito the C-methyl group of the 1 : 2 molar adduct frommethyl-4,5-dihydrothiazole and the acetylenic ester loshows that this adduct has the rearranged structure( 5) 2-D, 3- C 03,8a - MeH' 'E(7) ( 8 )(9) 2- Me(10) 3 - Me(11) 5 - Me(1212,5-Me2SCHEME(tetramethyl 2,3-dihydro-5-met h yl-pyridol 2, 1 -b t hiazole6,7,8,8a-tetracarboxylate) and not that proposed pre-viously.lo(18) X = S(20)X=NMe(19) x=o(14) (15) X = S(16) X = NMe(17)X = NMe,9-Me(21) 9a-Met 221 2 9 - Me2This general type of structural change involvingapparent ring opening and recyclization has been ob-served before in adducts from dimethyl acetylenedi-carboxylate.The adduct from 4-ethoxyquinazolinelo J. Roggero and C. Divorne, Compt. rend., 1969, 2WC, 870.l1 R. M. Acheson, J. N. Bridson, T. R. Cecil, and A. R. Hands,J.C.S. Perkin I , 1972, 1569.undergoes acid-catalysed isomerisation,l phenanthri-dones are formed from indole,ll and initially formedpyridazino2,3-a benzotriazoles from 1 -alkylbenzotri-azoles yield pyridazin02,3-aquinoxaliriones.~2The 2,4-dimethylthiazole adduct (4) is the only un-rearranged thiazole derivative isolated 50 far. It isstable in 1,2-dichlorobenzene at 150 "C, possibly owingto steric factors. Dreiding models shon- that our pyri-dothiazoles are all crowded, and that there is an ad-ditional interaction between methyl groups when theseare present at both positions correspoiiding to the 2- and4-positions of the original thiazoles, when rotation in theintermediate (6) is considered.One methyl group causeslittle additional hindrance over that already present inthe isomerisation leading to (S), and a methyl group atposition 5 of the original thiazole is well out of the way.The I3C resonances for the adducts (A), ( 5 ) , (S), (9),( l l ) , and (12) (Table 1) fall into a clear pattern; com-pound (10) was not soluble enough for measurement.The sp3 Sa-carbon signals for compounds (S), (9), ( l l ) ,and (12) appear between 6 73.7 and 75.3, and that forcompound (4) ( 6 77.0) shows that replacing the bridgeheadester substituent by a methyl group causes a small down-field shift.The position of these sp3 carbon rewnaiicesis in good accord with that (6 69.9) of the sp3 carbonatom in the thiazoline (14). A comparison of the lHthe 13C resonances for the adduct (4) with those of thedeuteriated derivative (5) enables unambiguous assign-ments to be made for the 13C and IH resonances of themethyl groups and the atoms at position 2. From com-paring the data in Table 1 for these compounds it is clearthat replacing a hydrogen atom by a methyl substituentat positions 2-, 3-, and 5- deshields the adjacent sp2carbon atom by ca. 13, 8, and 10 p.p.m., respectively,whereas 2- compounds (9) and (12)j and 3-methylgroups compound (4) shield the ring carbon atoms atpositions 3 and 2 by smaller amounts.The resonancesdue to the 5J3C atoms were at very low field (6 144), andthose from the 6-, 7-, and 8-carbon atoms, bearing anester group, were in the 6 100-139 range but couldnot beindividually assigned. The 13C resonances for the 2-and amp;methyl groups of compounds (Y), (ll), and (12)correlate well with each other, and the bridgehead-methyl group of compound (4) is much more stronglydeshielded. This last characteristic, however, shouldbe used with caution if employed to locate methylgroups, for the chemical shift of the se3-attached 9a-methyl group of the quinolizine (21) and that of one ofthe sp2-attached methyl groups of (23) are indistin-guishable.In view of the above findings it was necessary toreconsider the structures of similar adducts which havebeen reported to be formed from benzo-thiazole, -oxazole,-imidazole, etc.Benzothiazole with the ester in theabsence of solvent ,2 or in acetonitrile, dimethylforma-mide or toluene, gives as major product a 1 : 2 molaradduct described earlier as tetramethyl 5aH-dibenzo-12 P. J. Abbott, R. ill. Acheson, XI. W. Foxton, N. R. Raulins,and G. E. Robinson, J.C.S. Pevkzpt I, 1972, 21821976 1271TABLE 1l3C N.m.r. spectra (22.63 MHz; internal Me,Si or CD,NO, as reference; shifts, 6c, with respect to Me,Si)SolventCD,NO,CD,NO,CD,NO,CD,NO,CD,NO,-(CDJ,SOCD,NO,CD,XO,CD,NO,CD,NO,CDCI,CD,NO,CD,NO,CDCI,CDCI,CDC1,(25)O CD,NO,t/"C4545332740502780502780692525252540No. of scans4.00x 10-33.401.083.004.004.100.510.0021.0055.214.847.581.01.002.002.007.21Carbon assignments3-CH3, 15.1; 5-C, 144.4;8a-C, 77.0; 8a-CH3, 25.43-CD, not observed ; 5-C,144.4; 8a-C, 76.9; 8a-CH3,25.4144.7; 8a-C, 74.52-C, 108.7; 3-C, 133.5;'2-C, 108.9;' 3-C, 133.6;'2-C, 112.0; 3-C, 126.9; 5-C,2-C, 125.6; Z-CH,, 12.7;3-C, 122.7; 5-C, 144.3; 8a-C,74.92-C, 111.4; 3-C, 124.6; 5-C,154.5; 5-CH3, 18.7; 8a-C,73.72-C, 125.0; Me, 12.8; 3-C,18.4; 8a-C, 75.33-CH3, 33.2; Ar-CH, 126.2,122.2, 119.4, 108.3; 3a, 7a-C,,124.8, 149.2Ar-CH, 127.7, 127.6, 123.9,110.3; 4a, 1Oa-C,, 140.3, 129.4;5a-C, 75.6; 9-C, 141.7Ar-CH, 127.3, 121.3,110.7,108.2; 4a, 1Oa-C,, 144.1,131.7; 5-CH3, 34.9; 5a-C,Ar-CH, 125.0, 119.8, 113.6,109.0; 4a, 1Oa-C,, 143.9,"131.4; 5-CH,, 33.8; 5a-C,Ar-CH, 128.5, 126.1, 123.6,120.6; 5-C, 154.3; 5-CH3,2-C, 69.9; 2-CH,, 20.8 ;86.5; 9-C, 136.484.9; 9-C,143.9;' 9-CH,, 17.7113.1; 9-C, 57.2Ar-CH, 127.0, 126.2, 112.0,112.0; 9-C, 56.8Ar-CH, 123.6, 123.4, 110.0,109.4; 5-CH3, 34.9; 9-C, 53.56,7,8,9-C4, 125.7, 123.0, 120.6,106.8; 9a-C, 60.2; 9a-CH,,22.06,8-C,,122.6, 119.5; 7-CH3,17.9;' 9-CH3, 17.5;' 9a-C,57.74-C, 65.5; 6,8-C,, 142.6, 134.8;7-CH3, 17.3;' 9-CH3, 22.1 'l-CH,, 35.4; 2-C, 132.9,' 3-C,107.1;' 8a-C, 83.6; 8a-CH3,15.0sp2-CC0,Meandunidentified132.8,'113.3,103.5132.8,'113.3,103.5138.0,106.0,101.8139.1,104.6,100.6138.2,106.5,103.0138.8,105.9,102.1sp2-c138.8,110.3,103.9136.4,g110.6,104.0150.5 ?,104.8 f140.7,129.6,103.0 f161.9,149.1,145.8,130.8,126.1,102.4149.0,144.9,133.9,129.0,123.6,g97.1146.1,130.4,118.8,100.5148.5,134.9,131.6,127.4,107.6,97.1147.7,145.2,133.8,126.4,93.7,92.9114.0fc=o168.3,165.8,165.3,165.0168.3,165.8,165.3,165.0169.9,168.3,165.2,165.2170.1,168.4,165.4,166.2170.5,168.8,166.8,165.3170.8,168.9,167.0,165.4170.2,168.0,165.3168.5,e165.5,165.0168.1 ,c167.6,165.8169.9,168.3,165.6,165.2169.6,168.5,165.4,164.6167.9,h167.2,163.2,'162.4167.3,165.0,163.9,163.9167.6,164.4,163.8,162.9169.6,168.6,165.0,164.2169.3,168.9,166.6,165.4OCH,54.1,53.1,52.7,52.754.1,53.1,52.7,52.754.4,53.2,53.0,52.354.7,53.4,53.2,52.654.2,63.1,52.8,52.154.2,53.0,52.7,52.054.4,53.0,"52.453.5,63.2,62.8,50.852.9,52.6,52.2,51.653.8,53.0,52.7.52.253.8,53.0,52.7,51.752.6,51.8,51.3,50.353.4,52.4,52.2,52.053.2,52.3,52.0,51.853.3,52.5,51.8,50.854.2,52.8,52.4,52.1272 J.C.S. Perkin ITABLE 1 (Continued)sp2-CC0,MeandNo.of scans unidentifiedCompd. Solvent t/"C x 10-3 Carbon assignments sp2-c c=o OCH,(26)" CD3N02 27 8.58 Ar-CH,127.3,120.7, 112.1, 139.8, 168.3, 54.4,111.0; 4a, 10a-C,, 143.9, 130.4, 165.9, 53.3,129.8; 6-CH3, 33.2; 5a-C, 119.0, 165.2, 53.1,85.7; 5a-CH,, 17.2 103.9 165.0 52.60 All W-lH attachments confirmed by off-resonance experiments.b Trisacetylacetonatochromium added to reduce relaxatione Probably due to two coincident reson- time.ances. f All $92-C atoms not located. 0 Observed in off-resonance decoupling experiment. h 9-13C0,Me. 8-13C0,Me.e These assignments may be interchanged. d Broad, confirming 13C-2H coupling.bd thiazole-5,6,7,8-t et racarbox ylat e . The low-fieldposition of the 9-H n.m.r. signal and the correspondenceof the 13C n.m.r. spectrum with those of the thiazoleadducts (Table 1) leave little doubt that rearrangementto the isomeric structure (15) has taken place. On thisnew premise, t et rame t h yl 1,243 hydro - 1 -phen y 1 p yridine-2,3,4,5-tetracarboxylate is the expected product ofRaney nickel desulphurisation, and the structure of(15) has now been fully established by X-ray crystal-10graphy.l~In methanol, benzothiazole with the ester yields a smallamount of the pyridobenzothiazole (15) and 34 of anisomer with a different U.V.spectrum, a single proton(7 3.60), and an sp3 carbon atom bearing one hydrogenMe EH 'E(23)H E(24)1 E(28) R1=Me,R2= H(29) R' =H,R* =Eatom (6 57.2). These data correspond well to structure(IS), which could be obtained from the benzo-analogueof (3) by a concerted 1,5 hydrogen shift similar to thatestablished l4 in the quinolizine series, and X-ray studieshave recently shown that this structure is correct.13l3 H. Ogura, K.Kikuchi, H. Takayanagi, K. Furuhata, Y.Iitaka, and R. M. Acheson, J.C.S. Perkin I , 1976, 2316.l4 R. M. Acheson and B. J . Jones, J . Chem. Soc. (C), 1970,1301.IE. R. M. Acheson, M. W. Foxton, P. J. Abbott, and K. J. Mills,J . Chem. SOC. (C), 1967, 882.Benzoxazole gives a single adduct with dimethylacetylenedicarboxylate,2 originally formulated with ahydrogen atom at a bridgehead cf. (3). However, theresonance position of the single proton (7 3.82) and thechemical shift of the sp3 carbon atom bearing a singlehydrogen atom ( 6 56.8) strongly suggest that a protonshift has occurred to give structure (19).It is significant that the 13C resonance positions for the9-carbon atoms of compounds (18)-(20) are very similar,which would not be the case if the single hydrogen atomwas at the bridgehead position cf.(3), for then the ~ $ 3carbon atom would be adjacent to different heteroatomsin the different compounds.1-Methylbenzimidazole and the ester in acetonitrilegave mainly a 1 : 2 molar yellow adduct, the wide-bandand off-resonance decoupled 13C spectra for which con-firm the structure (20) already proposed.15 The un-decoupled spectrum for the carbonyl carbon atoms showsquartets with J(13C,0CH3) 3.5 Hz for two of these carbonatoms, which must be at positions 6 and 7 as a conse-quence of the subsequent argument, and more complexmultiplets for the other two. Selective decoupling ofthe protons of all the ester-methyl groups caused thecollapse of the multiplets due to the carbonyl carbonatoms at positions 6 and 7 to singlets, showing that thesecarbonyl 13C atoms are coupled only to ester-methylgroups. The coupling constant is similar to that(3.7 Hz) observed16 for the coupling between the 13Cof the carbonyl group and the OCH, of methyl benzoate.The other two 13C signals of carbonyl groups now ap-peared as doublets, J(13C,H) 5.5 and 3.5 Hz.The wholepicture is consistent with the coupled proton beingpresent at position 9. It interacts most strongly withthe 9-carbonyl l3C atom, less so with the 8-carbonyl13C atom, and not with the others. Wasylischen andSchaefer have shown l7 that 13C,H-coupling over threebonds depends on the dihedral angle between the atoms,and our results are consistent with this but incompatiblewith the proton being present at either position 7 (whenthree ester groups would have coupled with it) or position5a (when only one weak coupling would be anticipated).The same method has been used to establish l8 the struc-ture of the pyridoimidazole (24).l6 A.M.Ihrigand J . C. Marshall, J . Amev. Ciiem. SOC., 1972,94,17 R. Wasylischen and T. Schaefer, Canad. J . Chem., 1973, 51,18 F. Troxler, H. P. Weber, A. Jaunin, and H. R. Loosli, Helv.3268.961.Chim. Acta, 1974, 57, 7601976 1273The 13C spectrum for the red adduct from l-methyl-benzimidazole, for which an improved preparation isdescribed, showed that it must possess structure (16)and not the earlier suggested l4 isomeric formulation witha bridgehead hydrogen atom cf.(3). This new struc-ture is consistent with the failure of the compound toisomerise thermally to (20), and with its reaction withbromine in perchloric acid, which causes substitution ofone hydrogen atom.15 Oxidation to a correspondingpyridobenzimidazolium salt, which is formed from (20)Ytetrahydrofuran gave 15 mainly the corresponding aze-pine, along with an orange-red substance affording agood elemental analysis for a 1 : 2 molar adduct. Theproton n.m.r. spectrum suggested that the substance wasa mixture, and re-examination shows that without doubtit consists of the bridgehead methyl (7 8.57) cf. (26) andrearranged ( T ~ ~ 7.08) cf. (17) isomers in ca. 1 : 2 ratio.The corresponding product from e-ethyl-l-methyl-benzimidazole l5 is homogeneous, and the high-fieldposition of the CH,-CH, resonance and the similarity ofFIGURE I Stereoscopic projection for tetramethyl 3,8a-dimethylpyrido2.1-bthiazole-5,6,7,8-tetracarbosylate (4) down the z axisunder these conditions,15 is expected of a compoundanalogous to (3).1,2-Dirnethylbenzimidazole with the ester in tetra-hydrofuran gives l5 mainly tetramethyl 9,10-dihydro-5-methylazepino 1,2-abenzimidazole-7,8,9,10-tetra-carboxylate, along with 1 : 2 molar, isomeric, red andorange adducts, which we confirm are both formed andextremely difficult to separate.The red adduct, with ahigh-field C-methyl group (7 8.56), is undoubtedly thedirect cyclisation product (26), and the orange compoundcan now be identified from its much lower-field (7 7.04)C-methyl signal and 13C spectra as (17), formed by therearrangement of (26).The aromatic proton resonancepattern for (26), which shows one low-field proton signaland three others of similar chemical shift, differs markedlyfrom that for (17), where the signals are much morespread out.1-Ethyl-2-methylbenzimidazole with the ester inthe aromatic proton resonance pattern to that of (26)but not that of (17) confirm that the assigned structure iscorrect.The adduct of 1,2-dimethylimidazole and dimethylacetylenedicarboxylate was assigned l9 structure (25)in spite of the fact that eight ester-methyl resonanceswere seen in the lH n.m.r. spectrum. Re-examinationof this substance has shown that it indeed has a verycomplex lH n.m.r.spectrum, not appreciably changed bycrystallisation from methanol or by chromatography.However, crystallisation from acetonitrile resulted in amaterial, the lH n.m.r. spectrum of which (see Experi-mental section) was entirely in accordance with struc-ture (25). The compound is not very stable and de-composes in solution, particularly in solvents such as1s 0. Diels, K. Alder, W. WinckIer, and E. Peterson, Annalen,1932, 498, 1; R. M. Acheson and G. A. Taylor, J. Chew. Soc.,1960,46001274 J.C.S. Perkin Ichloroform. Its 13C n.m.r. spectrum (solvent trideuterio-nitromethane) was complex owing to decomposition,but, although not all the signals could be assigned, theessential features were in accord with structure (25) andshowed the expected differences from the spectrum l8of (24).zx itwo lines of equal height, the biggest downfield shiftbeing observed for the 9-proton, which must thereforebe close to the europium atom.Stereoscopic projections for the similarly shapedthiazole adducts (4) and (11) are shown in Figures 1 and2, respectively. The nitrogen atoms and the carbontFIGURE 2 Stereoscopic projection for tetramethyl 5-methylpyrido2,l-bthiazole-6,7,8,8a-tetracarboxylate (1 1) down the y axisThe effect of adding the chiral shift reagent, tris-3-(2,2,2-trifluoro-l -hydroxye t hyl) -( +) -camphorat01 -europium(II1) 2o to solutions of the adducts (4), (ll),(15), (16), (18), and (20) in deuteriochloroform wasexamined in the hope of detecting the resulting diastereo-isomers.Although marked shifts of most of the protonresonances were observed in all cases, only for (18) andY IFIGURE 3 Atom numbering for compound (4)(20) were diastereoisomers detected. The resonances dueto all the methyl groups and the 9-proton were split into2o V. Shurig, Tetrahedron Letters, 1972, 3297; Inorg. Chem.,1972, 11, 736.atoms to which they are attached are effectively co-planar in both compounds (Table 2; plane 1) whichXFIGURE 4 Atom numbering for compound (11)means that the electronic hybridisations of the nitrogenatoms are essentially sp2. For an sp3 nitrogen this atomwould be placed 21 ca. 0.5 A away from the mean planeof the attached atoms. The bonds from the nitrogenatoms are unequal in length, the longest being to thebridgehead carbon atom and the shortest, which is evenshorter than the N-C bond of pyrrole (1.383 A),22 leading21 A.H.-J. Wang, I. C . Paul, E. R. Talaty, and A. E. Dupuy,J.C.S. Chern. Comm., 1972, 43.22 B. Bak, D. Christensen, L. Hansen, and J. Rastrup-Ander-sen, J. Chem. Phys., 1966, 24, 720; cf. C. W. N. Cumper, Trans.Faraday SOC., 1958, 54, 12661976 1275to the 5-carbon atom. The deviations of individualatoms from the mean planes of the 4(N), 5-, 6-, 7-, andcarbon atoms are small (Table 2). The orientations ofthe mean planes (Table 2) for all the ester groups ( i e .the C-CO, atoms for each) to this last mean plane werealso calculated. For the ester groups at position 6, forcompounds (4) and (1 1) , the divergences from coplanarityTABLE 2Deviatit a s (-amp;) from calculated least-squares best planesCompound (4) Compound (1 1)r p v - - - - - - lN(4) -0.004 N(4) -0.062Plant Atoms Deviation Atoms Deviation1 C(3) 0.001 C(3) 0.021C(6) 0.001 C(5) 0.023C(8a) 0.001 C(8a) 0.018C(6) -0.088 C(6) 0.0812 N(4) 0.021 N(4) -0.020C(5) 0.021 C(5) -0.020C(7) 0.111 C(7) -0.104C(8) --0.068 C(8) -0.064C(6) 0.008 C(6) -0.007C(10) -0.026 C(10) 0.024N(4) 0.113 N(4) 0.069O(11) 0.010 O(11) -0.009O(12) 0.008 O(12) -0.008C(5) -0.160 C(5) -0.028C(6) -0.009 C(6) -0.093C(10) -0.015 C(10) -0.029O(11) 0.013 O(11) 0.058O(12) 0.056 O(12) 0.022C(8) -0.003 C(8) 0.001C(18) 0.011 C(18) -0.005O(19) -0.004 O(19) 0.002O(20) -0.003 O(20) 0.001C(7) -0.000 C(7) -0.004O(15) -0.001 O(l5) -0.005O(16) -0.001 O(16) -0.004C(14) 0.002 C(14) 0.013C(5) 0.003O(23) 0.004O(24) 0.003C(22) -0.009Angles between the planesPlanes Compound (4) Compound (1 1)2 and 3 16.46" 10.23"2 and 5 47.60" 43.53"2 and 6 69.93" 69.36"2 and 7 63.75"are onIy 16.5 and 10.2", respectively, those for the S-estergroups are much greater, and those for the other estergroups greater still.This suggests that of the chargedresonance structures including a positively chargednitrogen atom which can be written, those involving the6-ester group 1e.g. (27) are the most important. Thelengths of the bonds joining the various ester groups tothe ring q-stern are in agreement with this concept, theshortest and next shortest being to the 6- and 8-estergroups , respectively. The carbon-oxygen bond lengthsfor the amp;ester group are also significantly greater thanthe others for compound (a), but not for (11).The bondlength situation for compound (15) is ~imi1ar.l~ Ageneral implication of these results is that quinolizinessuch as (21) and (23) will be expected to possess similarsp2-hybridised nitrogen atoms. The suggestion 23 that23 R. M. Acheson and R. S. Feinberg, J . Chem. SOC. (C), 1968,350.the large difference in U.V. spectra between the quino-lizines (28) and (29) is due to the importance of chargedresonance structures involving the 5-ester group isconfirmed by OUT new results, which also fit in with themuch smaller spectral change observedB when the 3-ester group is removed from compound (29).The low-field position of the amp;methyl proton resonance in then.m.r. spectrum of the pyridobenzimidazole (20) hasbeen attributed l5 to significant resonance contributionfrom charged structures, corresponding to (27) andinvolving the 6ester group. Bond-length data are notavailable for this compound, but those l3 for the ana-logous pyridobenzothiazole (18) show a steady increasein the lengths of the bonds joining the 6-, 8-, 7-, and 9-ester groups, respectively, to the ring. This indicatesthe importance of such charged contributors in com-pound (lS), and their importance in (201, sulphur havingbeen replaced by nitrogen, should be greater.EXPERIMENTALInstruments and chromatographic procedures used havebeen described in earlier papers in the series.2 , 3-Dihydro-2,3-dimethylbenzothiazole was prepared asdescribed in ref.24; T (CD,NO,) 8.51 (6 H, d, J 6 Hz,2-Me), 7.31 (3 H, s, 3-Me), 5.02 (2 H, q, J 6 Hz, 2-H), and2.9-3.8 (4 H, m, ArH) ; very similar in CDC1,.Reaction of Dimethyl A cety Zenedicarboxylate with Benzo-thiazoZe.-The ester (1.08 g) in methanol (2.5 nil) mixed withbenzothiazole (0.50 g) in methanol (2.5 ml) was left over-night at room temperature. The methanol was removed invucuo, toluene was added and evaporated off in vacuo withgentle warming (twice), and the residue was triturated withether. The solid (0.54 g) yielded fluorescent yellow crystals(from acetonitrile) of tetramethyl gH-pyridoS, l-blbenzo-thiazoZe-6,7,8,9-tetracarboxyZate (18), m.p.249-249.5'(Found: C, 54.2; H, 4.3; N, 3.4; S, 7.4. C,,H,,NO,S re-quiresc, 54.4; H, 4.1; N, 3.3; S, 7.6); T (CDC1,) 2.3-2.9 (4 H, m, ArH), 3.60 (s, 9-H), and 6.10, 6.19, 6.20, and6.37 (s, 4 x OMe), Amx. (MeOH) 229 (E 22 800), 256 (13 300),295infl (32 900), 309 (18 900), and 418 nm (19 400). Theinitial filtrate from (18) contained compound (15).Tetramethyl 2,5-Di.~.zeth~~lpyrido2,l-btkiazoZe-6,7,8,8a-tetracarboxylute (12) .-2,5-Ditnethylthazole 25 (1.43 g) indimethylformamide (5 ml) a t 0 "C was added dropwise withstirring to dimethyl acetylenedicarboxylate (3.55 g) indimethylformamide (5 ml). After 40 min at 0 "C and 10days a t room temperature, most of the solvent was removedin zlacuo, water (450 ml) was added, and the methylenechloride-soluble material was collected.Chromatographyon alumina and elution with methylene chloride gave thethiazole (12), orange crystals (2.85 g), m.p. 183-185"(from MeOH) (Found: C, 51.5; H, 4.9; N, 5.4. C,,H,,NO,requires C, 51.4; H, 4.8; N, 3.5); +T (CDCl,) 3.55 (9,J ca. 1 Hz, 3-H), 6.23, 6.30, 6.37, and 6.37 (s, 4 x OMe),7.43 (s, 5-Me) ; and 8.03 (d, J ca. 1 Hz, 2-Me) ; Amax. (MeOH)229 (E 17 700), 287 (28 900), and 442 nm (6 400)' unchangedby addition of a few drops of 72 HC10,.Tetramethyl 2-Methylpyrido2,l-bthiazole-6,7,8,8a-tetra-carboxylate (9) .-This was prepared similarly to ( 12)a4 J. Metzger, H. Larive, E. J. Vincent, and R. Dennilauler,J . Chim. phys., 1963, 60, 944 (Chem. Abs., 1963, 59, 9,763b).25 M.Poite and J. Metzger, Bull. SOC. chim. France, 1962, 20781276but from 5-methylthiazole 25 (0.56 g), and obtained asorange crystah (0.39 g), m.p. 185-186" (from MeOH)(Found: C, 50.2; H, 4.6; N, 3.6. C16Hl,N0,S requires(9, J ca. 1.5 Hz, 3-H), 6.17, 6.28, 6.35, and 6.35 (s, 4 XTABLE 3Bond lengths (A) for compounds (4) (Figure 3) and (11)(Figure 4) ; estimated standard deviations in paren-theses. The atom numbers for (1 1) are only given whenthey differ from those for (a), but exactly comparablebond lengths are shown on the same lineC, 50.1; H, 4.5; N, 4.7) ; z (CDCI,) 1.99 (s, 5-H), 3.77Compound (4)w-Compound (1 1)AI C (22)-0 (24)0 (24)--C(26)C( 2)-H (1 02)C (9)-H (1 09)C (9)-H (209)C(9)-H (309)C(13)-H(113)C( 13)-H (2 13)C( 13)-H(313)C(17)-H( 117)C( 17)-H( 21 7)C( 17)-H (31 7)C(21)-H( 121)C(21)-H(221)C(2 1)-H (32 1)C(25)-H(125)C(25)-H(225)C(25)-H(325)C(26)-H (1 26)C(26)-H(226)C (26)-H (326)1.730(6)1.832(6)1.320(8)1.41 1 (6)1.483(8)1.359( 6)1.496 (6)1.385(6)1.466 (7)1.338( 7)1.521 (6)1.439( 7)1.207 (6)1.342( 7)1.429( 8)1.504(6)1.193( 7)1.326(7)1.436( 7)1.464(7)1.187( 7)1.335(6)1.466(8)1.5 18 (8)1.503( 7)1.191 (7)1.322(7)1.446(9)0.95(9)1.03(6)0.93(6)0.96 (6)0.95(10)1.06( 11)0.92 (1 0)1.09(8)0.7 9 (8)1.02(9)0.97 (9)0.93(9)0.82( 11)0.93 (9)0.97(8)0.92 (9)1.07(6)0.99 (6)0.98(6)C( 8a)-C( 22)C(22)-0(23)C(22)-0(24)0(24)-C(25)C(5)-C(9)C( 2)-H( 102)C( 3)-H ( 103)C(9)-H(109)C(9)-H(209)C(9)-H (309)C( 26)-H ( 1 26)C(25)-H(225)C (2 5)-H (3 2 5)1.744( 6)1.838(4)1.321(9)1.405(6)1.350(6)1.468( 6)1.392 (6)1.461(6)1.350(6)1.507 (5)1.470(6)1.206( 7)1.309( 6)1.450( 7)1.513 (6)1.386(9)1.330(7)1.439( 6)1.489 (6)1.209(7)1.316( 6)1.444(6)1.564(7)1.203(6)1.302( 7)1.441 (8)1.496 (7)1.03(7)0.97(7)0.92( 12)0.82(11)1.20( 12)0.84(9)1.05( 10)0.89(9)0.77( 11)0.98(7)0.87(7)0.95(6)0.79(11)1.09(10)1.01 (1 1)1.09( 11)1.05 (1 1)OCH,), and 8.02 (d, J ca.1.5 Hz, 2-Me); (MeOH) 228(E 21 300), 286 (30 900), and 443 nm (6 500); no change onacidification.Tetramethyl Pi-Methy@yridoS , l-blhiazole-6,7,8,8a-tetva-J.C.S.Perkin Icarboxylate ( 11) .-This compound, prepared in the sameway as (12), had m.p. 163-164" (lit.,3 159.5-161').TABLE 4Bond angles (") for compounds (4) (Figure 3) and (11)(Figure 4) ; estimated standard deviations in paren-theses. The atom numbers for (1 1) are only given whenthey differ from those for (4), but exactly comparablebond angles are shown on the same lineCompound (4) Compound (1 1) . - c'c(2)-S(l)-C(8a)C( 2)-C (3)-N (4)C (2)-C (3)-C( 26)N (4)-C( 3)-C (26)C(3)-N(4)(8a)C( 6)-N(4)-C(8a)N (4)-C (6)-C (6)S(1)--C(2)-C(3)C ( 3 W (4)-c(5)N (4)-C (5)-C (22)C( 6)-C (6)-C (22)C(S)-C(6)-C( 10)C(7)-C(6)-C(10)C(5)-C(6)-C(7)C(6)-C(7)-C(8)C(6)-C(7)-C(14)C(7)-C(@-wa)C (8)-C ( 7)-C ( 1 4)C(7)-C(8)-C(18)C(8a)-C(8)-C(l8)S ( 1) -C (8a)-N (4)S (1 )-C( 8a)-C( 8)S( 1)-C( 8a)-C (9)N (4)-C( 8a)-C (9)C( 8)-C( 8a)-C (9)C(6)-euro;(10)-0( 11)C (6)-C (1 0)-0 ( 1 2)C(10)-0 (1 2)-C( 13)C(7)-C( 14)-O(15)C( 7 ) 4 ( 14)-0 ( 16)O( 15)-C( 14)-O( 16)C(l4)-0(16)-C(17)C (8)-C (1 8)-0 (1 9)C( 8)-C ( 1 8)-0 (20)0 (1 9)-C( 1 8)-0 (20)C (1 8)-0 (20)-C( 2 1)C(6)-C(22)-0(23)C(5)-C(22)-0(24)O( 23)-C( 22)-O(24)C (22)-0 (24)-C (25)N (4)-C(84-C@)O( 1 l)-c( lo)*( 12)Tetramethyl..90.9(7)11 6.9 (1 .O)112.9(9)12 3.6 (9)1 23.1 ( 1.0)128.7 (7)113.8 (8)1 1 7.4( 8)119.6(8)11 7.8(8)122.4 (8)11 7.0( 9)123.3( 8)119.7(9)121 .O( 8)118.6( 8 )120.4( 8)117 .O (9)1 22.2 ( 8)120.7 (8)1 04.4 (9)1 1 1.1 (1.0)1 1 1-1 (9)107.7 (9)108.6( 1.1)1 13.4( 1.0)1 24.7 (9)1 13.7( 1 .O)12 1.4 (9)1 17.9 (1 .O)124.9( 1 .O)110.4(1.1)124.7 (9)1 15.7 (1.0)12 5.4 (9)111.6(1)123.0 (9)11 6.0( 1)122.7( 1.0)1 26.2( 9)11 6.7 (1.2)11 1.2(1.1)90.1(9)114.1(1.1)114.4( 1.3)127.5 (7)1 12.2 (9)1 19.2 (8)11 8.1 (8)1 16.2 (8)1 2 5.6 ( 8 )N (4)-C (5)-C( 8)C (6)-C (5)-C(9)11 7.5(8)126.1(7)1 16.3 (8)121.3( 7)11 6.6( 7)12 1.2( 7)1 16.7 (8)1 2 5.6 (6)11 8.6( 7)105.0( 8)1 12.9( 7)107 .O (9)109.4( 8)S( 1)-C (8a)-C( 22)N(4)-C(8a)-C(22) 107.1(1.0)C (8)-C (8a)-C(22) 1 14.7( 9)122.6(1 .O)11 5.6(9)121.6(9)1 16.6( 1 .O)112.4(1.1)126.6( 8)1 15.0 (9)122.2 (9)113.5(9)124.2 (8)1 1 6.4 (8)122.0 (1.2)C( 8a)-C( 22)-O( 23) 122.8 (9)C (8a)-C (22)- (24) 1 1 1.8 (9)O( 23)--C( 22)- (24) 1 23.4 (9)C(2!2)-0 (24)- (26) 1 1 6.3 (1.0)3, 8a-DimetlzyZpyrido2,l-bthiazoZe-5,6,7,8-tetracarboxylate (4) .-This was prepared as for compound(12); m.p.149-149.5' (lit.,3 147.5-149'). The n.m.r.spectrum was essentially as described in ref. 3, and de-coupling of the 2-H and 3-Me (7 8.02) caused marked shar-pening of the 3-Me and 2-H resonances, respectively.The 2-deuterio-3-trideuteriomethyl analogue (5) wasobtained similarly from bromopentadeuterioacetone (ob-tained by brominating perdeuterioacetone 2s) and was thenconverted into the deuteriated thiazole, 7 (CDC1,) 7.63(2-Me), by Poite and Metzger's proced~re.~~ The proton26 B. A. Levene, Org. Synth., Coll.Vol. 11, 1943, p. 881976 1277n.ni.r. spectrum (CDCl,) was as described in ref. 3 exceptthat the resonances at 't 4.36 and 8.02 were missing.Tetramethyl 5-Methylpyrido 1,2-a benzimidazoZe-5a, 6,7,8-tetracarboxylute (I 6) .--lm$voved preparation. Dimethylacetylenedicarboxylate (5.38 g) in toluene (25 ml) at 0 "Cwas added to 1-methylbenzimidazole (2.5 g) in tolueneTABLE 5-Atomic co-ordinates for compound (4) (Figure 3) ;estimated standard deviations -x- 104la Y lb0.213 24(7) 0.177 2(1)0.134 7(3)0.126 4(3)0.182 8(2)0.195 6(2)0.265 l(3)0.324 2(2)0.311 l(2)0.233 5(2)0.216 2(3)0.280 7(3)0.341 2(2)0.230 9(5)0.400 4(3)0.449 l(2)0.405 8(2)0.475 6(4)0.368 6(3)0.421 4(2)0.356 2(2)0.402 1(4)0.131 6(3)0.081 7(2)0.138 O(2)0.081 2(6)0.069 4(4)0.102(4)0.264(3)0.168(3)0.226(3)0.189(5)0.2 37 (6)0.260(5)0.51 9 (4)0.485(4)0.471 (4)0.454( 5)0.3 73( 5)0.391(5)0.094(4)0.032(5)0.070(4)0.034( 3)0.042(3)0.094 (3)0.220 9(2)0.047 7(6j-0.099 9(4)-0.221 O(5)-0.087 6(6)-0.211 O(5)- 0.076 4(5)0.046 3(5)0.041 4(6)0.054 l(6)-0.326 3(6)- 0.324 9(5)- 0.434 3(4)-0.550 8(8)- 0.078 6(6)-0.023 6(5)-0.151 l(4)- 0.165( 1)0.181 2(6)0.190 l(5)0.296 O(4)0.436 4(8)-0.360 3(5)-0.379 l(4)-0.453 l(4)- 0.593 9(9)-0.213 7(8)0.072(8)0.147 (6)0.048(6)- 0.012(7)- 0.62(1)-0.53(1)-0.59(1) - 0.061 (8) - 0.230(8) - 0.196 (8)0.452 (9)0.49(1)0.43(1)- 0.654(9)- 0.603(8) - 0.611 (9)-0.277(6) - 0.183 (6)- 0.2 67 (7)ZIC0.152 3(2)0.092 5(6)0.120 O(5)0.200 6(4)0.237 2(5)0.276 O(5)0.247 l(5)0.228 5(5)0.251 9(6)0.396 2(6)0.340 8(6)0.366 7(5)0.380 9(5)0.452(1)0.233 6(6)0.308 9(5)0.126 5(4)0.105(1)0.190 4(6)0.128 O(6)0.238 5(5)0.186(1)0.242 5(6)0.316 l(6)0.153 9(6)0.1-55 ( 1 )0.0597 (8)0.035(9)0.435( 6)0.401(6)0.448 (6)0.42 (1)0.65( 1)0.40(1)0.1 2 1 (8)0.1 33 ( 8)0.009(9)0.19(1)0.20(1)0.11(1)0.097( 9)0.130(9)0.24(1)0.126(7)0.009(7)0.009(7)(25 ml) over 1 h at 0 "C with stirring.After 3 h at 0 "Cand 5 days a t room temperature the precipitate was col-lected and chromatographed on deactivated alumina. Thefirst fraction, eluted with methylene chloride, on recrystal-lization from methanol, gave the adduct (0.9 g), m.p.173-174" (lit.,l6 173-174"), which showed one spot ont.1.c.silica gel; ethyl acetate-petroleum (b.p. 60-80")Reaction of 1,2-Dimethylbenzimidazole with DimethylA cetyZenedicarb0xyZate.-The reaction was carried out asdescribed in ref. 15, the azepine was removed and the residuewas chromatographed. Elution with methylene chloridegave first the red adduct (26). The mother liquor from thecrystallization of this was combined with further methylenechloride eluates. Rechromatography and elution withtoluene-light petroleum (b.p. 80-100") (7 : 3 v/v) gave a(1 : 1 v/v)J.series of fractions. A slow-moving orange material crystal-lized with methanol to give the orange adduct (17), m.p.161-162' (lit.,15 158"); 't (CDCl,) 2.80(d) and 3.35(d),(1- and 4-H), 2.97(t) and 3.23(t) ( J ca.8 Hz, 2- and 3-H),6.92 ($Me), 7.04 (5a-Me), and 6.20-6.28 (4 x OMe).Tetramethyl 1,8a-DimethyZpyrido 1,2-airnidazole-5,6,7,8-tetrucarboxylate (25) (with G. PROCTER) .-This was preparedas described in ref. 19 but the product was chromatographedover alumina with methylene chloride as solvent, and aftercrystallisation from acetonitrile had m.p. 163-1 64" ;T (CDCl,) 7.00 (1-Me), 4.03(d) and 4.12(d) ( J 4 Hz, 2- and3-H), 8.72 (8a-Me), and 6.06, 6.15, 6.24, and 6.29 (4 x OMe).X-Ray Structure Determinations.-Approximate cell di-mensions were obtained from oscillation and Weissenbergphotographs. The intensity data for compound (4)were collected on a Hilger and Watts linear diffractometerwith Mo radiation and balanced filters.The structure of the adduct (4) (Figures 1 and 3) wasdetermined by Patterson and Fourier methods.All theatoms were located after two Fourier maps. Oneround of least-squares refinement with isotropic tempera-ture factors, followed by three rounds with anisotropictemperature factors gave an R value of 0.078 6. A differenceFourier synthesis located 13 of the hydrogen atoms; theremainder were Iocated by calcdation.The hydrogen atom positional parameters and isotropictemperature factors were refined individually and thenrefined together. Refinement of positional parametersand temperature factors (anisotropic except for H atoms)continued until convergence root mean square (shift/e.s.d.) G0.131.The root mean square (shift/e.s.d.) for allatoms except hydrogen was 0.07. The temperature factorsTABLE 6Anisotropic temperature factors * and estimated standarddeviations for compound (4) (Figure 3)Atom U,, ua a Us, Upara;3 Ul, Ull0.036 4(7)0.035(3)0.024(2)0.025(2)0.023(2)0.01 7(2)0.020(2)0.037(3)0.042(3)0.038(2)0.044(2)0.092(6)0.026(3)0.033(2)0.037(2)0.052 (4)0.032(3)O.O44( 3)0.034(2)0.049(4)0.027(3)0.034(2)0.038(2)0.053(4)O.OSS(3)0.021(2)0.021 (2)0.034 O(7)0.049(3)0.044( 3)0.036(2)0 034(30:032(3)0.036(3)0.028(2)0.023(2)0.031(3)0.031(3)0.059(3)0.030(2)0.035(3)0.039(3)0.081 3)0.051{2)0.0 75 (5)0.037(3)0.056(3)0.026(2)0.035(3)0.028(3)0.040(2)0 043 2)0:059{5)0.055(4)0.048 2(9)0.043(3)0.03 7(3)0.030(2)0.037(3)0.037(3)0.033 (3)0.035 (3)0.038(3)0.036(3)0.041 (3)0.086 (4)0.077(3)O.OSl(6)0.040(3)0.070(3)0.048(3)O.OSS(6)0.043(3)0.104(4)0.070(3)0.110(7)0.053(4)0.074(3)0.069(3)0.129 (9)0.064(5)0.005 S(6)-0.002(3) - 0.003 (2)-0.001(2)0.000(2)O.OOl(2)-0.004(2) - 0.001( 2) - 0.003 (2)O.OOO(2) - 0.005(2)0.01 0.021( 5(2) 3)0.020(4)0.005(2)0.005( 4)0.007(3)0.015(4)-0.022(3)-0.004(2)-0.000(2)-0.001(2)- 0.007(2) - 0.001 (2)-0.023(2) - 0.057( 5)0.001( 3)-0.005 3(6)-0.004(2)-0.011(2)-0.000(2)O.OOO(2)0.001 (2)0.002(2) - 0.001 0.002(2) (2)-0.001(2) 0.002(2)0.003(2)0.004(4)0.005(2)0.007(2)0 026(40:005(2{0.037( 3)0.00 3 (2)0.008(4) - 0.006 (2)O.OlS(2)- 0.005( 2)-0.011(2)-0.000(2)-0.001(5) - 0.020(3)0.016 7(6)0.021(3)0.009(2)O.OOS(2)0.014(2)0.007(2bsol;0.016(2)0.015(2)0.027(2)0.026(2)0.013(0.024(2)0.043(4)0.014(2)0.007(0.006(3)0 008 2)0.012(2)O.OlS(2)0.005(0.032(4) O.OlO(2)O.OlO(2)0:ooelz)0.011(2)-0.003(4)0.005(3)Atom UiSO t Atom UISO t Atom Uiao iH(102) 0.03(2) H(117) 0.04(1) H(125) 0.05(1)H(109) 0.020(9) H(217) 0.04(1) H 225) 0.05(1)H(317) 0.04(1)Ht325) H(126) 0.023(9) 0.05(1)H(209) 0.020(9)H(309) 0.020(9) H(121) 0.07(2)H(113) 0.07(2) H(221) 0.07(2) H(226) 0.023(9)H(213) 0.07(2) H(321) 0.07(2) H(326) 0.023(9)H(313) 0.07(2)Temperature factor T = exp (-2n*U,amp;a*)' j- .. . +2U1,(ha*kb*)}.-f Temperature factor T = exp ( - S ~ x U i 8 ~ ) a whereS = sin O/h.for H atoms on the same carbon were equivalenced to thesame least-squares parameter in the refinement.Final cell dimensions and intensity data for compoun1278(11) were measured on a Hilger and Watts four-circlediffractometer with A10 radiation.TABLE 7Atomic co-ordinates for compound (1 1) (Figure 4) ;estimated standard deviations x lo4xla0.347 O(1)0.212 4(6)0.126 7(5)0.161 O(4)0.094 9(4)0.151 7(4)0.281 5(4)0.344 7(4)0.273 4(4)-0.033 4(5)0.093 4(4)0.146 7(4)-0.022 2(3)-0.081 5(7)0.340 6(4)0.329 8(4)0.407 3(3)0.475 5(7)0.476 6(4)0.530 7(3)0.526 9(3)0.655 l(5)0.241 3(4)0.142 4(4)0.333 8(4)0.317 4(6)0.2 13 (6)0.046(6)- 0.07( 1)-0.071(9)-0.04(1)- 0.051 (8)- 0.154(8)-0.125(8)0.509(8)0.434( 8)0.490( 9)0.695(6)0.653( 6)0.685(6)0.2 7 7 (9)0.26(1)0.397 (9)Ylb-0.000 00.072( 1)0.093( 1)0.086 7(8)0.075 3(9)0.060 4(8)0.063 3(7)0.022 7(7)0.120(2)0.021(2)-0.042 3(7)0.053 9(9)0.074( 1)0.065(2)0.143 2(9)0.296 l(7)0.025 6(6)0.028 6(7)0.144 2(7)0.139( 1)0.098(1)-0.066 5(8)-0.245 3(7)-0.300 2(7)-0.343 7(7)- 0.534(1)O.lO(1)0.13(1)0.03(2)0.14(2)0.18(1)0.27(2)0.09(1)-0.02(1)0.20(1)-0.06(1)0.07(1)0.13(1)0.06(1)0.23(1)- 0.62(2)- 0.54(2)- 0.61 (2)ZlC0.044 2( 1)0.107 5(5)0.019 2(5)-0.109 O(4)-0.219 O(4)-0.337 8(4)-0.336 l(4)-0.113 8(4)-0.227 4(4)-0.201 4(7)-0.466 9(5)-0.468 O(4)-0.594 l(8)- 0.4539(4)-0.481 5 ( 5 )-0.516 4(3)-0.621 4(7)-0.213 7(4)- 0.564 9(4)-0.138 5(4)-0.291 9(4)-0.297 l(6)-0.118 l(4)-0.200 l(5)-0.142 2(5)- 0.1325(7)0.205 (7)0.032 (6)- 0.16(1)- 0.15( 1)-0.649(9)- 0.586(9)- 0.57819)- 0.653(9)- 0.694(9)- 0.62 1 (9)- 0.215(6)-0.374(7)- 0.342(6)-0.06(1)-0.18( 1)- 0.2711)- 0.13(1)The same procedure as used for (a) was employed todetermine the structure of (11). Refinement by usingpositional parameters and anisotropic temperature factors(isotropic for the hydrogen atoms) continued until con-vergence root mean square (shift/e.s.d.) GO, 171, The rootmean square (shift/e.s.d.) for all atoms except hydrogen was0.07.The temperature factors for hydrogen atoms on thesame carbon atom were equivalenced to the same least-squares parameter in the refinement.The carbon-hydrogen bond lengths were constrained to1.0 + 0.01 i as there were large variations in bond lengthsand angles. The constraints were removed near the end ofthe refinement.The structures for compounds (4) and (11) were refinedby use of weights calculated from expression (i) 27 wherex = Fo/Fo(,,1 / ~ = {AOTO'x -+ AjlTl'x . . . + ANP - 1T-NP - 1'x} (i)* For details of Supplementary Publications see Notice toAuthors No. 7 in J.C.S. Perkin I, 1975, Index issue.J.C.S. Perkin IStructure factor tables are available as SupplementaryPublication No. SUP 21717 (35 pp., 1 microfiche).*Crystal data.(i) Tetramethyl 3,8a-dimethylpyrido2, l-bthiazole-5,6,7,8-tetracarboxylate (4) Cl,H1,N08S, M =397.45, monoclinic, a = 19.56(4), b = 9.89(2), c = 10.15(2)A, cc = 90.00 p = 90.00 y = 110.50 amp; 0.08", U = 1839.2A3, D, = 1.44 g cm-3, 2 = 4, F(000) = 832. Space groupP2,/a. Mo-K, radiation, 1, = 0.710 7 A, p = 2.25 cm-1.5 315 Reflections were observed up to 0 = 34", yielding 4 203independent reflections of which 2 220 were 30 (I).Merging X was 6.7 over 2 220 reflections; terminal R =0.063. In the weighting scheme above, AO = 8.5 andA1 = 6.5.(ii) Tetramethyl 5-meth~-lpyrido~2,1-bthiazole-6,7,8,8a-tetracarboxylate( 11) Cl,Hl,NO,S, f11 = 383.40, mono-clinic, n == 11.249(2), b = 7.498(2), c = 10.272(2) A, a =D, = 1.47 g ~ m - ~ , Z = 2, F(000) = 400.Space group P2,.hfo-I, radiation, 1, = 0.710 7 A, p = 2.35 cm-l. 3 091reflections were observed up to 0 = 2S0, yielding 2 246independent reflections of which 2 056 were 30 (I).Orange crystals 0.07 x 0.02 x 0.02 cni.90.00, p = 90.91 Ji 0.02, y = 90.00", U = 866.2 A3,TABLE 8Anistropic temperature factors * and estimated standardAtom li,, UZZ U S J u 2 4 Ul, L'l *deviations for compound (1 1) (Figure 4)S ( l ) 0.035 7(6)C(2) 0.055(3)0.042 (2)0.0637(!)) 0.0217() -0.0046(3)O.OBi(4) 0.029(2) -0.006(3)0.063(3) 0.028(2) -0.010(2)C(5) 0.026(2)C(7) 0.028(2) 0.036(2) 0.022(2) O.OOO(2)c(6) 0.029(2) 0.046(2) 0.023(2) o.ooi(f)0.027(2) 0.032(2) O.053(2) -0.000(2)C(0) 0.029(2) 0.16(1) 0.042(3) 0.010(4)C(l0) 0.033(2) 0.0;4(3) 0.030(2) 0.004(2):((::/ 0.031(3) 0.104(7) 0.053(4) -0.003(4)0.033(2) 0.047(3) 0.024(2) 0.005(2) :bsol;bsol; 0.058(2) 0.033(3) 0.048(2) O.OlS(2)O(1C;) 0.046(2) 0.047(2) 0.032(2) 0.003(2)C(1 r ) 0.069(4) 0.098(4) 0.039(3) 0.005(3) SK)) 0.031(2) 0.072(:1) 0.045(2) 0.015(2)0.020(1) 0.058(2) O.044(2) 0.006(2)O(11) 0.048(2) 0.185(9) 0.029(2) -0.011(4)0.032(2) 0.121(5) 0.039(2) -0.008(3)0.028(2) 0.041(3) 0.027(2) -0.003(2)amp;" o.ozs(2) n.o72(4) o.oSo(n) -n.o02(3)C(22) 0.032(2) 0.040(3) 0.022(Y) -0.000(2)O(23) 0.037(2) 0.059(3) 0.069(3) -0.001(2) :it 0.038(2) 0.038(2) 0.063(!) O.OOO(2)0.030(3) 0.048(3) O.O56(J) O.OOl(3)Atom CJisot ~4t0111 riao 1-0.ooiq4) o.o024(q0.011(2) O.OOS(3)0.015(2) 0.009(3)0.007(1) O.OOS(2)0.007(2) 0.00312)0.004jij0.008( 1)0.005(1)- 0.001(2)0.002(2)-0.009(1)0.008( 1 )O.OOS(2)- 0.022(3)0.007(2)0.020(2) 0.019(1)0.011(2) 0.005 (1)0.030(3)O.OOS(1)-0.003(1)0.008 (1)O.OlS(2)O.OlS(2)0.006( 3)Atoin0.001(2 jO.OOl(Z)0.002(2)0.004( 0.023(4)0.007(4)0.005( 3)-0.004(4) - 0.004(2)0.007(2)-0.007(3)O.OOS(2)Ooo(-0.002(2)-0.001(2)- 0.005(2)-0.008(3)-0.005(2) - 0.011 (2)-0.003(2)-0.002(3)Gieo t.H(102) 0.02(1) H(21::) 0.04(1) H(221) 0.011(8)H(103) 0.02(1) H(313) 0.04(1) H(321) 0.011(8)H(1OY) 0.06(2) H(1lF) 0.04(1) Hf125) 0.05(1)H(209) 0.06(3) H(217) 0.04(1) H(225) 0.05(1)H(309) 0.06(2) H(317) 0.04(1) H(3?6) 0.03(1)H(113) 0.04(1) H(121) 0.011(8)* t See Table 6Merging X was 3.8 over 2 056 reflections; terminal X =0.081. In the weighting scheme above, AO = 765,A1 = 1021, and A2 = 260. Golden-yellow crystals0.06 x 0.03 x 0.02 cm.We thank Mrs. E. E. Richards for the 13C n.m.r. spectra,with the exception of the data for compound (20), whichwere kindly obtained by Mr. H. R. Loosli, at the Pharma-ceutical Division of Sandoz Ltd., Basel, and the Directorand Staff of the Oxford University Computing Laboratoryfor computing facilities.5/2003 Received, 9th Octobev, 197812i ' Crystals ' User Manual, J. R. Carruthers, Oxford UniversityComputing Laboratory, 1975