1H, 13C, and 15N NMR Stereochemical Studyof cis-Fused 7a(8a)-Methyl and 6-Phenyl

Octa(hexa)hydrocyclopenta[d][1,3]oxazinesand [3,1]Benzoxazines

PETRI TAHTINEN,1 JARI SINKKONEN,1 KAREL D. KLIKA,1 VILLE NIEMINEN,1 GEZA STAJER,2

ZSOLT SZAKONYI,2 FERENC FULOP,2 AND KALEVI PIHLAJA1�1Department of Chemistry, University of Turku, Turku, Finland

2Institute of Pharmaceutical Chemistry, University of Szeged, Szeged, Hungary

ABSTRACT Four 7a-methyl octa(or hexa)hydrocyclopenta[d][1,3]oxazines, five 8a-methyl octa(orhexa)hydro[3,1]benzoxazines, two 6-phenyl hexahydro[3,1]benzoxazinones, and 8a-methyl hexahy-dro[1,3]benzoxazinone, all cis-fused, were prepared and their stereostructures studied by various one-and two-dimensional 1H, 13C, and 15N NMR spectroscopic methods. In solution, the cyclopentane-fused2-oxo derivatives and the 1,3-benzoxazinone were found to attain exclusively the N-in/O-in conforma-tion, whereas the 6-phenyl 2-oxo/thioxo derivatives were found to be present predominantly in the N-outconformation. The C-2 unsubstituted and the 2-oxo/thioxo 7a/8a-methyl derivatives were all present insolution as a rapidly interconverting equilibrium of the N-in and N-out conformations. The C-2 methylderivatives were each found to be interconvertable mixtures of epimers (at C-2) with the N-in conformerpredominating for one epimer and the N-out conformer predominating for the other, with both pre-dominating conformers having the C-2 methyl group equatorially orientated. The substituent on thenitrogen (H or Me) was found to be always predominantly equatorial with respect to the heteroring,except for the epimeric 2-methyl derivatives with N-out conformations where steric constraints and thegeneralized anomeric effect resulted in the axial orientation of the C-2 methyl being favored. Chirality14:187�198, 2002. � 2002 Wiley-Liss, Inc.

KEY WORDS: saturated heterocycles; preparation; conformation; substituent effect; epimerization

A systematic study of the structure and stereochem-istry of ring-saturated heterocyclic compounds has beenundertaken for many years concomitant with improve-ments to their synthetic methodologies.1 For example,Eliel2,3 has thoroughly studied the reactions of 4,4,7a-trimethyl-trans-octahydro-1,3-benzoxazines substituted atposition 2 for use as chiral adjuvants, which were ob-tained optically pure by derivation from natural pule-gone. In particular, the amino alcohol fi 1,3-benzoxazinetransformation in the work described here was based onhis synthetic methodology. References 4�14 provide il-lustrative examples of the wide use of this synthon forenantioselective transformations and it has subsequentlybecame known as the Eliel-synthon. In addition to beinginherently interesting, many heterocyclic compoundsalso display marked physiological activity which providesfurther impetus for their study. Assessment of the sub-stituent effects on the stereochemistry can be accom-plished by systematically modifying the structure of aheterocyclic compound and utilizing the power of mod-ern NMR spectroscopic methods for structural analysis,as demonstrated in this report.

In this work, the preparation of a set of compoundscomposed of four 7a-methyl octa(or hexa)hydrocyclo-

penta[d][1,3]oxazines (1, 4, 7, 9), five 8a-methyl octa(or hexa)hydro[3,1]benzoxazines (2, 3, 5, 8, 10), two6-phenyl hexahydro[3,1]benzoxazinone derivatives (11,12), and 8a-methyl hexahydro[1,3]benzoxazinone (6),all cis-fused, are described (see Fig. 1). Theirstereostructures adopted in solution as determined byNMR spectroscopic methods are also reported. Similarcompounds possessing, or lacking, the 7a/8a-methylhave been studied earlier,15�20 whereby it was ascer-tained that the cis-fused ring system in 3,1-oxazine (1,3-oxazine) derivatives can attain either a biased N-in or N-out (O-in or O-out) conformation (see Fig. 2), or exist asan interconverting equilibrium mixture of the two con-formers, depending on their relative stabilities asdetermined by ring size, the ring substituents, and ring

Contract grant sponsors: the Academy of Finland, the National ScientificResearch Foundation (Hungary), and the Foundation for HungarianResearch and Higher Education; Contract grant numbers: 4284; OTKA,F 032828; and AMFK, respectively.�Correspondence to: Professor Kalevi Pihlaja, Department of Chemistry,University of Turku, Vatselankatu 2, FIN-20014 Turku, Finland. E-mail:kalevi.pihlaja@utu.fiReceived for publication 10 May 2001; Accepted 21 July 2001

� 2002 Wiley-Liss, Inc.DOI 10.1002/Chir.10056

CHIRALITY 14:187�198 (2002)

structure. The steric orientation of the labile ring sub-stituents in this series of compounds (i.e., attached to N)was also determined. Compounds 9 and 10 are distinctfrom the other members of the set in that they exhibitepimerization with respect to the configuration at C-2 inaddition to the conformational processes observed in theother compounds. (Compounds 7 and 8 can alsoundergo the same interconversions as 9 and 10, but thetransformation results in the same species.) The 15Nchemical shifts were also measured to ascertain corre-lations with regard to the structural differences betweenthe compounds of the set, but the scope of this workwas found to be too limited to discern the exact natureof the effects on the 15N chemical shifts.

EXPERIMENTAL

Melting points were determined on a Kofler apparatusand are uncorrected. The physical and analytical data forcompounds 1�12 are listed in Table 1. NMR data arelisted in Tables 2�4. Compounds 13a,b and 14a,bwere prepared by methods described previously.1,20

Synthetic Methodology (Scheme 1)

(4aR�,7aS�)-7a-Methyl-4,4a,5,6,7,7a-hexahydrocyclopen-ta[d][1,3]oxazin-2(1H)-one (1), (4aR�,8aS�)-8a-methyl-4H-4a,5,6,7,8,8a-hexahydro[3,1]benzoxazin-2(1H)-one (2), (4aR�,7aS�)-1,7a-dimethyl-4,4a,5,6,7,7a-hexahydrocyclopenta[d][1,3]oxazin-2-one (4), and (4aR�,8aS�)-1, 8a-dimethyl-4H-4a,5,6,7,8,8a-hexahydro[3,1]benzoxazin-2-one (5).Ethyl chloroformate (0.26 g, 2.4 mmol) was added drop-wise to a stirred solution of the appropriate amino alcohol(13a,b or 14a,b, 2 mmol) in a mixture of benzene (15 ml)and 5% aqueous sodium hydroxide (15 ml) at 10�C and thereaction monitored by TLC. After stirring for 2 h thephases were separated and the organic phase dried overNa2SO4. Removal of the solvent yielded an oily productwhich was then heated at 100�C for 30 min together with0.05 g of sodium methylate. After dissolution in ethyl ac-etate (30 ml), the organic phase was then washed withwater (2 · 15 ml), dried (Na2SO4), and the solvent re-moved by rotary evaporation.

(4aR�,8aS�)-8a-Methyl-4H-4a,5,6,7,8,8a-hexahydro[3,1]benzoxazine-2(1H)-thione (3). Thiophosgene (0.25g, 2.2 mmol) in dry benzene (5 ml) and triethylamine(0.45 g, 4.4 mmol) in dry benzene (10 ml) were added

Fig. 1. The structures of compounds 1�12 studied in this work.

Fig. 2. The N-in and N-out conformational equilibria of 7 and 8 and theepimerizable compounds 9 and 10.

TAHTINEN ET AL.188

dropwise together to a stirred solution of amino alcohol13b (0.32 g, 2.2 mmol) in dry benzene (15 ml) at 0�C.After the addition was complete, the solution was stirred atroom temperature for a further 2 h and then washed with5% HCl (20 ml) followed by water (20 ml). The organicphase was dried (Na2SO4) and the solvent removed byrotary evaporation.

(4aR�,8aR�)-8a-Methyl-4H-4a,5,6,7,8,8a-hexahydro[1,3]benzoxazin-2(3H)-one (6). Compound 6 was pre-pared according to the procedure described for com-pounds 1, 2, 4, and 5 starting from the corresponding cis-aminomethyl-1-methylcyclohexanol.20

(4aR�,7aS�)-1,7a-Dimethyl-1,2,4,4a,5,6,7,7a-octahy-drocyclopenta[d][1,3]oxazine (7), (4aR�,8aS�)-1,8a-dimethyl-4H-1,2,4a,5,6,7,8,8a-octahydro[3,1]benzoxazine(8), (2R�,4aS�,7aR�)- and (2S�,4aS�,7aR�)-1,2,7a-trimethyl-1,2,4,4a,5,6,7,7a-octahydrocyclopenta[d][1,3]oxazine (9a and 9b, respectively) and (2R�,4aS�,8aR*)-and (2S�,4aS�,8aR�)-1,2,8a-trimethyl-4H-1,2,4a,5,6,7,8,8a-octahydro[3,1]benzoxazine (10a and 10b, respec-tively). Amino alcohol 14a or 14b (3 mmol) was stirredwith 15 ml of either 33% aqueous formaldehyde (for com-pounds 7 and 8) or 20% aqueous acetaldehyde (for com-pounds 9 and 10) at room temperature for 1 h. The

aqueous solution was basified with 10% aqueous potassiumhydroxide and extracted with diethyl ether (3 · 30 ml).The combined organic phases were dried (Na2SO4) andthe ether removed by rotary evaporation to yield a nearcolorless oil which was purified as the hydrochloride salt.After liberation, the free base was a viscous oil which wasthen subjected to spectroscopic examination.

(4aR�,6S�,8aS�)-6-Phenyl-4H-4a,5,6,7,8,8a-hexahy-dro[3,1]benzoxazin-2(1H)-one (11). Amino alcohol15b (2.1 g, 10 mmol) and NaHCO3 (0.84 g, 0.1 mol) weredissolved in water (10 ml) followed by the dropwise addi-tion of ethyl chloroformate (1.1 g, 10 mmol) and the re-sulting mixture refluxed for 30 min. After cooling, thesolution was extracted with ether, dried (Na2SO4), and thesolvent removed by rotary evaporation. Dissolution of theresidue in ethyl acetate and standing at 5�C overnightprovided crystals of the carbamate. The carbamate (2.0 g,10 mmol) was heated with EtONa (50 mg) in an oil bath at200�C for 15 min. After cooling, the melt was extractedwith EtOAc and the product crystallized.

(4aR�,6S�,8aS�)-6-Phenyl-4H-4a,5,6,7,8,8a-hexahy-dro[3,1]benzoxazin-2(1H)-thione (12). CS2 (1.3 g) indioxane (8 ml) was added dropwise to amino alcohol 15b(3.4 g, 17 mmol) in an 11% aqueous KOH solution (10 ml)at 0�C with stirring. A further 5.5% aqueous solution of

TABLE 1. Physical and analytical data for compounds 1–12

Formula(M.W.)

Requires (%) Found (%)

COMPOUND Yield (%) Recrys.solvent M.p. (�C) C H N C H N

1 53 hexane-isopropylether

80-81 C8H13NO2

(155.20)61.91 8.44 9.02 62.35 8.89 9.31

2 59 hexane-isopropylether

97-98 C9H15NO2

(169.23)63.88 8.93 8.28 63.43 8.98 8.02

3 42 i-propyl ether 191-194 C9H15NOS(185.28)

58.34 8.16 7.56 58.46 8.31 7.41

4 54 hexane-isopropylether

69-71 C9H15NO2

(169.23)63.88 8.93 8.28 64.12 8.68 8.54

5 46 hexane-isopropylether

72-74 C10H17NO2

(183.25)65.54 9.35 7.64 65.91 9.52 7.54

6 49 isopropyl ether 128-130 C9H15NO2

(169.23)63.88 8.93 8.28 64.02 9.01 8.20

7 74 acetone-diethylether

188-191 C9H18ClNO(191.70)

56.39 9.46 7.31 56.11 9.57 6.96

8 79 acetone-diethylether

193-195 C10H20ClNO(205.73)

58.38 9.80 6.81 58.58 9.65 6.76

9 56 acetone-diethylether

135-144 C10H20ClNO(205.73)

58.38 9.80 6.81 58.72 9.99 7.15

10 61 acetone-diethylether

166-176 C11H22ClNO(219.76)

60.12 10.09 6.37 59.86 10.24 6.42

11 73 ethanol 179-181 C14H17NO2

(213.27)72.70 7.41 5.87 72.82 7.48 5.95

12 67 ethanol 239-242 C14H17NOS(247.34)

67.98 6.93 5.66 67.79 6.79 5.61

NMR STUDY OF CIS-FUSED OXAZINE DERIVATIVES 189

TABLE 2. 1H NMR Chemical shifts for compounds 1�12

Comp. NH/Me H2ax H2eqa H4ax H4eq H4a H5axb H5eqb H6axb H6eqb H7axb H7eqb H8ax H8eq Me-8/9

1 5.9 — — 4.09 4.25 2.02 1.95c 1.77t 1.68 1.75 1.73C 1.79t — — 1.302 6.3 — — 4.47 4.06 1.65 1.59 1.62 1.30 1.66 1.49 1.49 1.46 1.71 1.323 8.4 — — 4.53 4.17 1.77 1.49 1.66 1.32 1.68 1.46 1.54 1.48 1.88 1.374 2.87 — — 4.17 3.96 2.11 1.96 1.69 1.64 1.66 1.67c 1.97t — — 1.355 2.90 — — 4.43 3.97 1.71 1.66 1.64 1.37 1.67 1.30 1.52 1.41 1.96 1.326 6.3 — — 3.63 2.99 1.65 1.57 1.52 1.29 1.75 1.63 1.53 1.40 1.94 1.387d 2.20 4.05 4.15 3.76 3.56 1.69 1.75 1.75 1.62t 1.80c 1.29c 2.04t — — 1.128 (N-in)e 1.91 4.03 3.96 3.88 3.35 1.09 2.03 1.25 1.17 1.65 1.41 1.23 0.95 1.88 0.938 (N-out)e 2.27 4.41 4.02 3.69 3.51 1.87 1.18 1.57 1.32 1.14 1.36 1.57 2.02 1.12 1.159a 2.09 3.97 1.24 3.81 3.75 1.53 2.15 1.69 1.56 1.84 1.32 2.00 — — 1.079b 2.22 4.10 1.30 3.21 3.74 1.94 179c 1.13t 1.66 1.71 1.27c 2.06t — — 1.2210a 2.01 4.15 1.26 4.05 3.42 1.15 2.31 1.33 1.27 1.77 1.63 1.31 1.00 1.99 1.0310b 2.24 4.43 1.28 3.84 3.63 1.94 1.70 1.33 1.43 1.23 1.43 1.69 2.05 1.31 1.2911 6.8 — — 4.54 4.16 2.50 1.91 1.92 2.61 — 1.55 1.93 1.73 2.04 —12 8.1 — — 4.54 4.32 2.57 1.95 1.95 2.63 — 1.58 1.96 1.76 2.08 —

aH2eq = Me-2 in 9 and 10.bFor cyclopentane-fused compounds 1, 4, 7, and 9, H-5, H-6, and H-7 protons are denoted either byc(= cis) or t(= trans), indicating their stereochemistry relative to the C-8 methyl, where such notation is not provided, the stereochemistry of the proton is unknowndAverage chemical shift value at +25�C for 7.eChemical shifts for 8 in CD2Cl2 at )50�C.

TA

HT

INE

NE

TA

L.

190

TABLE 3. 1H, 1H coupling constants for compounds 1�12

Comp. 2ax,2eqa

4ax,4eq

4ax,4a

4eq,4a

4a,5axb

4a,5eqb

5ax,5eqb

5ax,6axb

5ax,6eqb

5eq,6axb

5eq,6eqb

5eq,7eqb

6ax,6eqb

6ax,7axb

6ax,7eqb

6eq,7axb

6eq,7eqb

6eq,8eqb

7ax,7eqb

7ax,8ax

7ax,8eq

7eq,8ax

7eq,8eq

8ax,8eq

1 — )11.3 3.9 3.2 7.1 9.7 )13.1 2.7 8.5 8.7 4.2 — )11.1 10.6 4.5 6.8 6.9 — )9.9 — — — — —2 — )11.2 3.3 3.6 10.8 2.6 )14.4 13.6 3.5 7.0 4.7 2.8 )12.9 10.5 3.8 2.9 2.9 1.7 )14.6 12.1 3.5 2.8 3.6 )12.13 — )11.5 3.4 3.5 10.7 4.8 )14.1 11.5 4.0 3.9 5.1 1.3 )13.4 11.2 3.5 3.3 5.1 1.3 )13.8 11.6 3.6 4.0 4.8 )14.44 — )11.1 3.6 5.7 6.7 9.5 )13.6 9.2 5.7 9.8 4.7 — )10.7 7.6 7.4 7.2 7.5 — )15.2 — — — — —5 — )11.1 3.0 3.7 10.9 4.4 )14.1 11.3 4.1 4.1 5.4 0.5 )13.4 11.0 3.8 3.4 5.5 1.4 )13.5 11.4 3.1 3.4 5.6 )14.76 — )11.8 5.3 4.2 12.8 3.6 )13.6 12.9 4.0 3.1 3.7 n.r. )13.3 12.8 4.2 3.7 3.4 1.7 )13.9 12.8 3.9 4.3 3.1 )14.27 )8.5 )11.6 4.3 5.5 8.3 7.0 )13.1 10.1 5.3 5.4 10.7 — )13.1 9.7 5.0 7.0 9.3 — )13.2 — — — — —8(N-in) )8.1 )11.4 2.4 <0.5 12.5 n.r. )13 13 3 n.r. n.r. n.r. )13 13.2 n.r. 3.0 n.r. n.r. )13.2 13.6 3.0 3.3 n.r. )13.98(N-out) )9.7 )11.4 12.1 5.1 n.r. n.r. n.r. n.r. n.r. n.r. n.r. n.r. n.r. n.r. n.r. n.r. n.r. n.r. n.r. 13.1 n.r. 4.3 n.r. )13.19a 5.5 )11.5 3.1 1.5 n.r. n.r. n.r. n.r. n.r. n.r. n.r. n.r. n.r. n.r. n.r. n.r. n.r. — )13 — — — — —9b 5.5 )11.5 12.2 6.4 n.r. n.r. n.r. n.r. n.r. n.r. n.r. n.r. n.r. n.r. n.r. n.r. n.r. — )12.2 — — — — —10a 5.3 )11.3 2.4 1.6 12.3 3.9 )13.2 13.2 4.2 4.0 n.r. n.r. n.r. n.r. 4.0 n.r. n.r. n.r. n.r. 13.6 n.r. 3.2 n.r. )14.310b 5.6 )11.5 12.5 5.3 5.3 1.5 n.r. n.r. n.r. 4.0 n.r. n.r. n.r. n.r. 4.0 3.9 n.r. n.r. n.r. 13.2 3.9 4.5 n.r. )13.211 — )10.9 11.8 4.4 5.7 2.7 )14.3 12.3 — 4.0 — n.r. — 11.6 3.5 — — — )13.4 12.8 3.3 3.3 4.6 )13.712 — )11.1 11.8 4.4 4.8 3.9 )12.0 8.5 — 8.3 — n.r. — 11.6 3.4 — — — )13.8 12.6 3.2 3.4 4.2 )13.8

aH2, Me-2 in 9 and 10.bIn the cyclopentane-fused compounds 1,4,7, and 9, the H-5, H-6, and H-7 protons cannot strictly be assigned as axial or equatorial ; n.r.=not resolved.

NM

RS

TU

DY

OF

CIS

-FU

SE

DO

XA

ZIN

ED

ER

IVA

TIV

ES

191

TABLE 4. 13C and 15N NMR chemical shifts for compounds 1�12

Comp. NMe C2 Me-2 C4 C4a C5 C6 C7 C8 C7a/8a Me-7a/8a i o m p N

1 — 155.3 — 66.5 41.6 26.8 22.4 41.9 — 61.8 27.6 — — — — )276.42 — 154.3 — 68.6 36.6 24.8 24.1 21.5 37.8 52.2 30.1 — — — — )278.83 — 185.7 — 70.4 35.7 24.9 23.8 21.5 37.1 53.9 29.3 — — — — )235.94 30.4 155.2 — 65.3 43.8 26.4 22.4 38.1 — 65.7 25.1 — — — — )288.75 28.7 154.0 — 67.1 38.6 25.1 24.3 21.7 35.6 56.1 26.2 — — — — )290.36 — 154.5 — 42.8 35.1 26.5 24.7 21.2 37.5 79.0 26.2 — — — — )305.47 33.0 81.3 — 67.1 43.7 26.0 20.5 33.8 — 62.4 18.1 — — — — )337.78(N-in) 30.4 81.2 — 68.6 41.3 25.5 26.2 20.4 36.5 53.4 15.0 — — — — )342.18(N-out) 33.7 80.8 — 69.2 35.5 24.2 20.8 22.9 25.8 54.3 23.5 — — — — )325.29a 31.8 85.3 21.3 65.1 46.4 27.6 21.5 39.6 — 62.9 15.3 — — — — )329.69b 32.0 84.9 20.9 68.8 41.5 24.3 19.7 26.2 — 65.0 26.7 — — — — )322.810a 29.8 85.2 21.6 68.0 42.5 26.2 26.7 20.5 38.0 54.3 16.8 — — — — )333.010b 29.3 83.6 21.1 67.5 36.9 24.7 21.0 23.1 25.1 55.8 24.7 — — — — )316.811 — 154.2 — 67.0 30.7 32.7 38.1 30.8 30.8 50.0 — 144.9 126.6 128.5 126.4 )286.612 — 186.3 — 68.6 29.7 32.5 38.0 30.6 30.1 51.4 — 144.4 126.6 128.6 126.7 )244.2

TA

HT

INE

NE

TA

L.

192

KOH was added (10 ml) followed by Pb(NO3)2 (5.5 g in 30ml of water) and the mixture left to stir at 60�C for 10 min.Precipitated PbS was removed by filtration and evapora-tion of the solvent from the filtrate yielded the solidproduct

(1R�,2R�,4S�)-2-Hydroxymethyl-4-phenyl1-cyclohexy-ylamine (15b). (1R�,2S�,4S�) - 4 - Phenylcyclohexane-1,2-dicarboxylic anhydride21 (10.0 g) was added portionwiseto concentrated ammonium hydroxide (30 ml) at 10�C.The mixture was allowed to stand for 30 min, followed byremoval of the solvent under reduced pressure. The resi-due was cooled to 10�C and neutralized with 10 M HCl.The resulting solid was then filtered off, washed with wa-ter, and dried. The monoamide obtained (7.0 g) was thenadded portionwise to a hypobromite solution generated bythe dropwise addition of bromine (1.7 ml) at 0�C to NaOH(6 g) in water (25 ml). A further portion of 30% aqueousNaOH (15 ml) was then added and the resulting solutionstirred at 75�C for 2 min. After filtration, concentrated HCl(14 ml) and glacial acetic acid (6 ml) were added to thecooled mixture and the separated solid amino acid wascollected by filtration, washed with water, and then dried.

The amino acid (9.0 g unpurified, ca. 40 mmol) was thencarefully added to a suspension of LiAlH4 (4.2 g, 0.11 mol)in dry tetrahydrofuran (200 ml) with cooling and stirringfollowed by refluxing for 20 h. After cooling to 0�C, excessLiAlH4 was decomposed by the dropwise addition of water(7 ml) and the mixture stirred until a white suspension hadformed which was removed by filtration. The solvent wasevaporated and the residue recrystallized from benzene-petroleum ether (b.p. 40�60�C) yielding beige crystals of15b, m.p. 102�105�C, yield 3.3 g (32%).

NMR Measurements

NMR spectra were acquired on either a JEOL JNM-LA400 (operating at 399.78 MHz for 1H and 100.54 MHzfor 13C) or a JEOL JNM-A500 (1H, 500.16 MHz; 13C, 125.78MHz; and 15N, 50.69 MHz) spectrometer equipped witheither 5 mm normal configuration or 5 mm inverse probes.Depending on sample availability and solubility, NMRsamples comprised of 2�40 mg of the material dissolved inca. 0.6 ml of either CDCl3 or CD2Cl2. Samples were mea-sured within the temperature range of )95 to +57�C; the1H spectra of 1, 4, and 7 were also measured in a mixtureof CDCl3 and (CD3)2CO (1:3, eutectic point of ca.)115�C22) down to a temperature of )110�C. 1H spectrawere referenced internally to TMS (0.00 ppm); 13C spectrato the solvent signal (CDCl3, 77.00 ppm; CD2Cl2, 53.80ppm); while 15N spectra were referenced externally toCH3NO2 (0.00 ppm) containing ca. 10% of CD3NO2 forlocking purposes.

1H and 13C spectra were acquired with normal single-pulse excitation using a 45� flip angle and broad-band 1Hdecoupling for the 13C measurements. The 15N chemicalshifts were measured using single-pulse experiments withinverse-gated decoupling, or refocused INEPT+ measure-ments (optimized on J = 90 Hz) in the cases where thenitrogen possessed a directly bound proton, or by selectiveINEPT (INAPT) measurements optimized on a two-bondcoupling of 2 Hz to the N-methyl protons. 1H chemicalshifts are reported in Table 2, 1H,1H coupling constantsTable 3, and 13C and 15N chemical shifts in Table 4.

The assignment of the chemical shifts was based on theconcerted application of PDQF-COSY, PDQF-COSY withlong-range delay (200 ms), f1-decoupled CHSHF, HMQC,DEPT, and PHSQC-TOCSY experiments. PHSQC-TOCSYexperiments were invaluable for the assignment of thecarbocycle nuclei, while DEPT spectra were useful for theassignment of 13C signals in equilibrium cases. 1H spectrawere analyzed in selected cases using PERCH simulationsoftware for the extraction of chemical shifts and couplingconstants in complex spectra.23

The stereochemistry of the compounds was based onthe assigned 1H and 13C chemical shifts, the 1H,1H cou-pling constants, and the nuclear Overhauser effect(NOE). NOE difference experiments were acquired usingsaturation times of 6�8 sec and reported enhancementsare expressed as a percentage, integrated with respect tothe irradiated spin (set to )100%). Prior to NOE mea-surements, samples were deoxygenated by nitrogenbubbling.

RESULTS AND DISCUSSION

1H NMR Results for Compounds 1�6

Compounds 1, 4, and 6 were determined to essentiallyadopt predominantly only one conformation in solutionsince there was scant evidence for the presence of anyother conformation upon lowering the temperature. (Al-though the signals in the 1H NMR spectra of the com-

Scheme 1. Synthetic routes for compounds 1–5 and 7–12.

NMR STUDY OF CIS-FUSED OXAZINE DERIVATIVES 193

pounds in CDCl3, CD2Cl2, or a 1:3 mixture of CDCl3 and(CD3)2CO solutions did broaden slightly upon loweringthe temperature, this could well be attributed to the in-creased inhomogeneity of the system.) Saturation of thebridgehead methyl resulted in an NOE enhancement to H-4ax (for 1, 0.9%, and for 4, 0.6%), conversely saturation ofH-4ax resulted in an NOE enhancement to the bridgeheadmethyl (for 1, 2.6%, and for 6, 3.4%). These NOEs are onlypossible in the N-in (O-in for 6) conformation and not thecorresponding N-out (O-out) conformation. Unequivocally,the vicinal 1H,1H coupling constants between H-4a and thetwo H-4 protons were both small (see Table 3). For the N-out (O-out) conformation, one sizeable diaxial coupling of11�13 Hz should be present between these protons, es-pecially in the benzoxazine derivatives. Thus, it is the N-inconformation (O-in) which is the predominant conformerfor these compounds in solution.

For compounds 2, 3, and 5, distinct spectra for twointerconverting conformers were observed in CD2Cl2 so-lution at )95�C. The two conformers were readily identi-fied by the coupling constants of the H-4 and H-4a protons.Integration of their respective signals yielded their con-former ratio and, for compounds 2 and 3, the conformerratio was also available from the integration of the NHprotons. The minor conformer (2 and 3, 10%, 5, 12%)present in each compound was the N-out conformer, rec-ognizable by the two large couplings (ca. 11.5 Hz) clearlypresent for one of the H-4 protons, one of which, in addi-tion to the geminal coupling, is due to a vicinal diaxialcoupling between this proton and H-4a (see Table 3). Thisis only possible in the N-out conformation (see Fig. 2). Themajor conformer in each case possessed only small cou-plings between H-4a and both H-4 protons (<2.1 Hz),consistent with the N-in conformation.

Complete PERCH23 analyses were made for the protonspectra recorded at +25�C for 1�6 (except for 2, wherethe quality of the spectrum precluded a satisfactory result)and the resulting spectral parameters are presented inTables 2 and 3. In compound 6 the coupling between H-4aand H-5ax is large because of the diaxial configuration ofthe protons in the O-in conformation. In the benzoxazinederivatives 2, 3, and 5 the observed coupling between H-4a and H-5ax is not as large because of the contribution ofthe N-out conformation to the equilibrium in addition tothe preferred N-in conformation. The extracted couplingsfor the cyclohexyl moiety of 2, 3, 5, and 6 all indicate achair conformation for this moiety, particularly the long-range w-type couplings between H-5eq and H-7eq, andbetween H-6eq and H-8eq, which are very typical for achair conformation. Additionally in 6, a 1.67 Hz couplingwas observed between H-4eq and the NH proton. Thechemical shifts are mostly invariant and only small sub-stitution effects are perceivable.

1H NMR Results for Compounds 7 and 8

Although the signals in the spectra of 7 in either CDCl3or CD2Cl2 solution broadened significantly upon loweringthe temperature, only an average spectrum was observable

at the lower limit of the temperature range. In a 1:3 mix-ture of CDCl3 and (CD3)2CO subspectra of two conform-ers started to decoalesce at temperatures below )100�C,but totally sharp spectra could not be obtained. A roughestimate for the ratio of the two conformers at )110�C isca. 85:15. In the spectrum of the minor conformer, twolarge couplings (ca. 11.5 Hz) were clearly visible for one ofthe H-4 protons, which again was an indication of the N-outconformation (see Fig. 2). The predominant conformerwas therefore the N-in conformer.

It is also possible to estimate the position of the con-formation equilibrium for 7 at +25�C by using model val-ues for the vicinal coupling constants obtained from the N-in and N-out conformations of 9a and 9b (vide infra).Compounds 9a and 9b differ from 7 only by the methylsubstitution at C-2. Couplings J4a,4ax and J4a,4eq were se-lected for estimating the ratio of the N-in and N-out con-formations in 7 by using Equations 1 and 2:

J4a;4axð7Þ ¼ xðN-inÞJ4a;4axð9a;N-inÞþð1� xðN-inÞÞJ 4a;4eqð9b; N-outÞ ½1�

J4a;4eqð7Þ ¼ xðN-inÞJ4a;4eqð9a; N-inÞþð1� xðN-inÞÞJ4a;4axð9b; N-outÞ ½2�

yielding in both cases x(N-in) = 0.63. The approximateratio for the N-in:N-out conformers is thus about 3:2 at+25�C, which means that the equilibrium shifts towardsthe N-out conformation at elevated temperatures. Similar-ly, it is possible to estimate the conformer ratio for com-pounds 2, 3, and 5 by using as model values thecorresponding couplings obtained for the epimers 10aand 10b (vide infra; these values are also consistent withvalues obtained from a review1). Using this method, theapproximate ratios for the N-in:N-out conformers forcompounds 2 and 3 are 3:1 and for compound 5 4:1 at+25�C, and again the equilibrium is considerably shiftedtowards the N-out conformation at elevated temperatures.

The oxo/thioxo substitution at C-2 together with amethyl at the bridgehead carbon (C-7a/C-8a) increasesthe predominance of the N-in conformation, which is inagreement with the observations made on analogousquinazolinone derivatives.24 However, when position 2 isunsubstituted the relative stability difference between theN-in and N-out conformers decreases (vide supra 7),which was an observation also made in earlier studiesfor the case where C-8a is unsubstituted.19 For 8 it waspossible to spectroscopically freeze out the equilibriumand the ratio of the two conformers (see Fig. 2) wasdetermined to be 3:4 at )60�C (N-in:N-out) in CDCl3 and3:2 (N-in:N-out) in CD2Cl2. The shifting of the equilibri-um by this extent when changing from CDCl3 to CD2Cl2is quite exceptional and reflects an extensive degree ofsolvation. The conformers were once again readilyidentifiable by H-4ax and the two large couplings presentin the N-out conformation to H-4a and H-4eq, whereas

TAHTINEN ET AL.194

the large diaxial coupling between H-4a and H-4ax islacking in the N-in conformation. In the N-in conforma-tion of 8, the coupling between H-4a and H-5ax wasagain large because of the diaxial configuration of theprotons. Further confirmation of the N-out conformationwas also obtained from NOE enhancements, whereby a3.8% NOE to H-8ax upon irradiation of H-2ax was ob-served. The 1H NMR spectrum measured in CD2Cl2 at)50�C was analyzed in more detail using PERCH23 (seeTables 2, 3). The analysis of the carbocycle protons wasdifficult due to the breadth of the signals at this tem-perature; however, this problem was not alleviated bylowering the temperature to )90�C due to loss of ho-mogeneity.

1H NMR Results for Compounds 11 and 12

Both compounds 11 and 12 were found to adopt pre-dominantly the N-out conformation in solution at +25�C.No other conformations were directly observed uponlowering the temperature to )60�C but the signals didbroaden slightly; however, this might be attributable toconformers arising from the hindered rotation of thephenyl ring rather than N-in:N-out conformers. The protonspectra measured at +25�C were fully analyzed usingPERCH.23 In addition to two large couplings for H-4ax, theN-out conformation is also indicated by a large diaxialcoupling between H-8a and H-8ax. For compound 11, thiswas found to be 11.05 Hz and for compound 12 11.09 Hz.By using the couplings between the H-4a and H-4 protonsin the same manner as earlier for compounds 2, 3, 5, and7 for estimating the N-in:N-out conformer ratio using Eqs.1 and 2, an approximation of 5% for the presence of the N-in conformation was calculated—a result which is withinthe limits of experimental error.

The couplings between H-8a and H-8eq were deter-mined to be 4.64 Hz for compound 11 and 4.93 Hz for 12.The w-type couplings present over four bonds from H-4eqto H-8a (in compound 11, 1.36 Hz, and in 12, 1.42 Hz) andfrom H-4a to H-8eq (in 11, 0.81 Hz, and in 12, 0.91 Hz)are also consistent and indicative of an N-out conforma-tion. Neither of these couplings are possible in an N-inconformation. Furthermore, strong NOE enhancementsfor both H-6ax (in 11, 8.5%, and in 12, 7.3%) and H-8ax(11: 3.6%, and 12: 3.1%) upon irradiation of H-4ax are alsostrongly supportive of the N-out conformation. The strongNOE observed from H-4ax to H-6 indicates that H-6 isaxially orientated and that therefore the phenyl ring isequatorial; this is also consistent with the large diaxialcouplings from H-6ax to both H-5ax and H-7ax (see Table3). It is actually the preference for the phenyl ring to beequatorial which is responsible for both 11 and 12 toprefer the N-out conformation. In the benzoxazine deriv-ative lacking a phenyl group at C-6, the N-in conformerwas found to dominate.15 Additional couplings between H-4a and H-8a (in 11, 5.05 Hz, and in 12, 5.07 Hz) and H-8aand NH (in 11, 4.13 Hz, and in 12, 4.12 Hz) were ob-served but have not been listed in Table 3.

1H NMR Results for the EpimerizableCompounds 9 and 10

Both compounds 9 and 10 are able to epimerize at C-2in solution, with the result that the C-2 methyl can beorientated equatorially in both of the predominate N-in andN-out conformations adopted by each of the epimers (seeFig. 2). Initially, solid samples of 9 and 10 contained bothepimers and this was apparent after immediate dissolutionin CDCl3 (ratio of the epimers, 5:3 and 10:3, respectively,for 9 and 10), but within several hours the epimericequilibrium settled, with a shift towards the epimer withthe predominant N-out conformation (3 h, 4:3 and 8 h, 1:1,respectively, for 9 and 10). The configuration at C-2 mostprobably inverts by a ring-opening mechanism as postu-lated by Szakonyi et al.20 There was no evidence for thepresence of a minor amount of the other conformer ineither epimer.

The coupling constants between the H-4a and H-4 pro-tons (see Table 3) again provided the essential informationregarding the conformations adopted by the two epimerspresent in 9 and 10. In the epimers 9a and 10a whichattain a predominantly N-in conformation, H-4a has onlysmall couplings to the two H-4 protons. By contrast, in theepimers 9b and 10b which attain a predominantly N-outconformation, the coupling between proton H-4a and H-4ax is large (ca. 12.5 Hz). These coupling constants areconsistent with the couplings proffered as model values forpure conformers.1 Additionally, the couplings from H-4a tothe H-5 protons are both small in the N-out conformationof the benzoxazine derivative 10b, whereas in the N-inepimer 10a a large diaxial coupling was measured be-tween H-4a and H-5ax (13.2 Hz).

The epimeric configurations were proven by NOEmeasurements. Upon irradiation of the bridgehead meth-yls, an NOE enhancement at H-2 was observed (1.6% in 9aand 2.9% in 10a) for the predominantly N-in conforma-tions, which indicates that H-2 is axial and thus the C-2methyl is equatorial. Irradiation of the bridgehead methylsalso resulted in an NOE to H-4ax (0.5% in 9a and 1.7% in10a), which is not possible in the N-out conformation. Inthe epimers with a predominant N-out conformation, irra-diation of H-4ax provided an NOE to H-2 of 3.2% in 9b and7.3% in 10b, which thus has to be axial and therefore theC-2 methyl is again in an equatorial orientation. Irradiationof H-2 resulted in an NOE to H-4ax (1.7% in 9b and 3.8% in10b) and additionally to H-8ax in 10b (5.6%), furtherconfirming the C-2 configuration.

Orientation of the N-substituent in Compounds 1–12

The NH or the N-methyl was found to be predominantlyequatorially orientated with respect to the heteroring incompounds 1�5, and 7, which are primarily in the N-inconformations and in the N-in conformation of 8, as in-ferred by an NOE (0.8�1.7%) from the bridgehead methyl.Conversely, the bridgehead methyl was enhanced(0.9�3.2%) when the proton or the methyl on the nitrogenwas irradiated. In 3, a w-type long-range coupling (1.43

NMR STUDY OF CIS-FUSED OXAZINE DERIVATIVES 195

Hz) between H-4a and the NH proton was discernable,which is consistent with an equatorial proton on the ni-trogen in the N-in conformation. Additionally for 7, irra-diation of H-2ax resulted in 2.8% enhancement of theN-methyl signal and irradiation of the N-methyl resulted in0.9% NOE of H-2ax, implying that the N-methyl is orien-tated equatorially, and this is probably true for both the N-in and N-out conformations which comprise the equilibriumat +25�C. In the N-out conformation of compound 8, irra-diation of H-2ax resulted in 2.1% NOE in the N-methyl andirradiation of the N-methyl resulted in 0.8% NOE in H-2ax,again indicating the equatorial disposition of the N-methyl.For 6, irradiation of the NH proton resulted in NOEenhancements at both H-4ax (2.3%) and H-4eq (4.5%) butnot at any of the carbocyclic protons, which suggests thatit too is equatorial. The NH proton in 11 and 12 is alsoequatorially orientated based on a 1.7% NOE at H-8eq uponirradiation, and vice versa when H-8eq is irradiated a 1.7%NOE is observed at the NH proton.

The N-methyl N-in conformations of 9a and 10a werealso found to be predominantly equatorial with respect tothe heteroring, as irradiation of the N-methyl resulted inan NOE to the bridgehead methyl (for 9a, 1.1%, and for10a, 2.5%). Additionally, in 9a irradiation of the bridge-head methyl resulted in a 1.2% enhancement at the N-methyl and in 10a the H-2ax signal was enhanced by 0.9%upon irradiation of the N-methyl. In the N-out conforma-tion of the epimer 10b, irradiation of the N-methyl yieldeda 1.0% enhancement at H-4a, which suggests that the N-methyl is axially orientated with respect to the heteroring,although at the same time H-2ax was also enhanced by0.7%. For 9b, irradiation of the N-methyl yielded only aweak NOE at H-4a. Irradiation of H-2ax in both 9b and10b did not result in an NOE at the N-methyl, implyingthat the N-methyls are indeed axial. For these two epimerswith an N-out conformation, equatorial orientation of theN-methyl is sterically more hindered than an axial orien-tation. Furthermore, axial orientation is also preferred bythe generalized anomeric effect.25

13C NMR Results for Compounds 1–12

The chemical shifts for the carbon nuclei of compounds1�12 are presented in Table 4. The 13C chemical shifts inthe carbocyclic parts of 1�10 agree with the adoptedconformation inferred from the 1H NMR analysis. Modelvalues for the 13C shifts were obtained from the litera-ture,15 from compounds 9a,b and 10a,b, which weremixtures of epimers predominate in the N-in (a) and N-outconformers (b), and from the conformers of compound 8.In the benzoxazines 2, 3, 5, and 6, which are mainly inthe N-in (O-in) conformation, and the N-in conformationsof 8 and 10a, C-7 is more shielded than when it is in the N-out conformations of 8 and 10b. Conversely, C-6 and C-8are less shielded in 2, 3, 5, and 6 and the N-in confor-mations of 8 and 10a in comparison to the N-out confor-mations of 8 and 10b. Similarly, C-6 and C-7 are lessshielded in the cyclopentaoxazines 1, 4, and 9a with the

N-in conformation in comparison to the N-out conforma-tion of 9b.

The chemical shift of the bridgehead methyl at C-7a/C-8a shows a clear trend in the conformer pair of 8 and theepimeric pairs of 9a,b and 10a,b. The 13C shift in the N-inconformation is 7.8�11.4 ppm towards higher frequency incomparison to the N-out conformation. Unfortunately,these bridgehead methyl shifts are not comparable to theother bridgehead methyl shifts in the series due to thevery different substitution at C-2 (oxo or thioxo).

The 13C NMR chemical shifts for 11 and 12 are es-sentially not comparable to the other compounds in thisstudy because of the phenyl substitution at C-6 and thelack of the bridgehead methyl at C-8a render them dis-tinct. Nevertheless, the chemical shifts of C-2 and C-4 in11 and 12 bear similarity to the corresponding shifts in 2,3, 5, 8, 10a, and 10b. Unfortunately, the chemical shiftof C-4 shift does not appear to be a reliable indicator forthe identification of the preferred conformation. The dif-ferences in the chemical shift of C-2 in 1�6, 11 and 12between oxo and thioxo functionalized compounds arequite pronounced (oxygen, 154.1�155.3 ppm; sulfur, ca.186 ppm). C-2 aside, the 13C chemical shifts of 11 and 12do not otherwise vary significantly from one another andonly the small differences can be seen in the heteroringcarbons.

15N NMR Results for Compounds 1�1215N chemical shifts are very sensitive to all manner of

structural changes,26 rendering the 15N nucleus a verygood probe that is highly responsive to systematic altera-tions in the structures of compounds. A difference inoverall conformation can also change the configuration ata nitrogen atom and appropriate substitution next to ni-trogen can alter its hybridization state. For example,within the compounds in this study the sp2 hybridization ofC-2 to oxygen or sulfur (cf. 7 and 4, and 8 and 5) resultsin the C-N bond having partial double bond character,which is reflected by the change in the chemical shift of49.0 ppm and 51.8 ppm, respectively. In both cases astrong shift towards higher frequency occurs as sp2 hy-bridized oxygen is incorporated, which is even more pro-nounced with the incorporation of the thioxo functionality.On this basis, the 15N chemical shifts were measured forthis series of compounds and the results are presented inTable 4. The observed 15N chemical shifts were all withinranges typical for amine- and amide-type nitrogens.26

Changing the size of the fused-carbocycle from cyclo-pentane to cyclohexane with the same substituents movesthe 15N chemical shifts slightly towards lower frequency incompounds with a predominate N-in conformation (cf. 2and 1, )2.4 ppm, 5 and 4, )1.6 ppm, 8a and 7, )4.4 ppm,and 10a and 9a, )3.4 ppm). However, for the N-out con-formation the shift is to the opposite direction (cf. 10b and9b, +6.0 ppm). Furthermore, comparison of the differ-ences in the 15N chemical shifts when changing from N-into N-out conformation results in a smaller shift to higherfrequency in the cyclopentane-fused case (cf. 9, +6.8 ppm)

TAHTINEN ET AL.196

than in the cyclohexane fused-cases (cf. 8, +16.9 ppm, and10, +16.2 ppm). In the case of benzoxazine derivatives,substituting the oxo group by a thioxo group does notdramatically affect the conformation of the compoundsand, thus, the chemical shift of the nitrogen moves to thehigher frequency by a rather constant amount in both theN-in and N-out conformations (cf. 2 and 3, +42.9 ppm, and11 and 12, +42.4 ppm). Similarly, methylation of C-2seems to have a rather constant effect on the 15N chemicalshift (cf. 7 (mainly N-in) and 9a, +8.1 ppm, 8 (N-in) and10a, +9.1 ppm, and 8 (N-out) and 10b, +8.4 ppm).Methylation of the nitrogen moves the 15N chemical shiftto lower frequency (cf. 1 and 4, )12.3 ppm and 2 and 5,)11.5 ppm). Also, the effect of methylation of the bridge-head C-8a carbon can be estimated by comparing com-pounds 2 and 11 or 3 and 12 and taking into account thatan average change in the 15N chemical shift from N-out toN-in conformation is ca. )16.6 ppm. This results in anapproximately +24.7 ppm effect for methylation of thebridgehead C-8a.

All these factors indicate that certain determinablesubstituent effects for the 15N chemical shifts do exist.However, the fact that in most of the cases within this setthe compounds are rapidly interconverting mixtures ofconformers which, together with technical limitations,preclude a precise determination of the trends in the 15Nchemical shifts.

CONCLUSIONS

Four 7a-methyl octa(or hexa)hydrocyclopenta[d][1,3]oxazines (1, 4, 7, 9), five 8a-methyl octa(or hexa)hy-dro[3,1]benzoxazines (2, 3, 5, 8, 10), two 6-phenylhexahydro[3,1]benzoxazinones (11, 12), and 8a-methylhexahydro[1,3]benzoxazinone (6), all cis-fused, were pre-pared and their stereostructures principally ascertained by1H, 1H coupling constants and by NOE enhancements to-gether with support from 1H and 13C chemical shifts. Insolution, the cyclopentane-fused 2-oxo derivatives (1 and4) and the 1,3-benzoxazinone (6) were found to attainexclusively the N-in/O-in conformation, whereas the 6-phenyl 2-oxo/thioxo derivatives were found to be presentpredominantly in the N-out conformation (11 and 12).The C-2 unsubstituted (7 and 8) and the 2-oxo/thioxo 7a/8a-methyl derivatives (2, 3, and 5) were all present insolution as a rapidly interconverting equilibrium of the N-in and N-out conformations. The C-2 methyl derivatives (9and 10) were each found to be mixtures of interconver-table epimers (at C-2) with the N-in conformer predomi-nating for one epimer (9a and 10a) and the N-outconformer predominating for the other (9b and 10b),with both predominating conformers having the C-2methyl group equatorially orientated. The substituent onthe nitrogen (H or Me) was found to be always predomi-nantly equatorial with respect to the heteroring, except forthe epimeric 2-methyl derivatives with N-out conforma-tions (9b and 10b), in which the substituent was orien-

tated axially due to both steric hindrance and thegeneralized anomeric effect.

Quite clearly there are many factors which contribute tothe final energy differences between the N-in and N-outconformers and hence the determination of the actualpresence of more than one conformer is really a questionof the limit of experimentation. There is always, at somelevel, a contribution from energetically unfavored con-formers and their quantification is bound by the con-straints of the particular approach applied. Slowing therate of interconversion by lowering the temperature toenable the visualization of separate spectra for each con-former can become a self-defeating exercise, as the con-tribution of the minor component is reduced in tandemwith the reduction in temperature; calculating the contri-bution of each conformer to an observed averaged spec-trum for a particular parameter (in this case, 1H,1Hcoupling constants) is heavily dependent on the availabilityand validity of correct model values, which is highlyquestionable in many instances.

LITERATURE CITED

1. Fulop F, Bernath G, Pihlaja K. Synthesis, stereochemistry andtransformations of cyclopentane-, cyclohexane-, cycloheptane-,and cyclooctane-fused 1,3-oxazines, 1,3-thiazines, and pyrimi-dines. Adv Heterocycl Chem 1998;69:349�477.

2. He X-Ch, Eliel EL. Highly enantioselective syntheses of a-hy-droxyacids using N-benzyl-4,4,7a-trimethyl-trans-octahydro-1,3-benzoxazine as a chiral adjuvant. Tetrahedron 1987;41:4979�4987.

3. Eliel EL, He X-Ch. Highly stereoselective syntheses involving N-alkyl-4,4,7a-trimethyl-trans-octahydro-1,3-benzoxazine intermedi-ates. J Org Chem 1990;55:2114�2119.

4. Pedrosa R, Andres C, Duque-Soladana JP, Mendiguchıa P. A novelcase of diastereoselection in 5-exo radical cyclization promoted byhydrogen bonding. Eur J Org Chem 2000;3727�3730.

5. Pedrosa R, Andres C, Nieto J. A short diastereoselective synthesisof enantiopure highly substituted tetrahydroepoxyisoindolines. JOrg Chem 2000;65:831�839.

6. Andres C, Duque-Soladana JP, Pedrosa R. Regio- and stereo-selective 5-exo radical cyclizations on a chiral perhydro-1,3-ben-zoxazine moiety. An access to enantiopure 3-alkylpyrrolidines. JOrg Chem 1999;64:4273�4281.

7. Andres C, Duque-Soladana JP, Pedrosa R. A novel approach tochiral, nonracemic pyrrolidines by 5-exo-trig diastereoselectiveradical cyclization on acrylamides derived from (-)-8-aminometh-anol. J Org Chem 1999;64:4282�4288.

8. Andres C, Duque-Soladana JP, Pedrosa R. Tributyltin radical-in-duced addition-carbocyclization on chiral perhydro-1,3-benzoxa-zines: a facile entry to enantiopure tin-containing auxiliaries.Chem Commun 1999;31�32.

9. Andres C, Garcıa-Valverde M, Nieto J, Pedrosa R. Thermal andLewis acid catalyzed diatereoselective intramolecular Diels-AlderReaction on a,b-unsaturated amides derived from (-)-8-amino-menthol. J Org Chem 1999;64:5230�5236.

10. Lacoste J-E, Soucy Ch, Rochon FD, Breau L. 2-Vinyl-trans-octa-hydro-1,3-benzoxazine: cyclization and 1,3-dipolar cycloaddition ofnitrile oxides. Tetrahedron Lett 1998;39:9121�9124.

11. Rae A, Castro JL, Tabor AB. Synthesis of homochiral propargylamines from N-(Boc)-tetrahydro-2H-1,3-oxazines. J Chem SocPerkin Trans 1 1999;1943�1948.

12. Soucy Ch, Lacoste J-E, Breau L. Synthesis of 2-isoxazolines fromolefins derived from norephedrine and pulegone. TetrahedronLett 1998;39:9117�9120.

NMR STUDY OF CIS-FUSED OXAZINE DERIVATIVES 197

13. Nakano H, Okuyama Y, Hongo H. New chiral phosphinooxathianeligands from palladium-catalyzed asymmetrical allylic substitutionreactions. Tetrahedron Lett 2000;41:4615�4618.

14. Andres C, Duque-Soladana JP, Iglesias JM, Pedrosa R. Synthesisof enantiopure 3-alkyl-perhydroazepines by diastereoselective 7-endo-radical cyclisation on a chiral 1,3-perhydrobenzoxazine de-rivative. Tetrahedron Lett 1999;40:2421�2424.

15. Stajer G, Szabo AE, Fulop F, Bernath G, Sohar P. Synthesis andconformational studies of 2-oxo- and 2-thiotetrahydro-1,3-oxazineswith condensed skeletons. Heterocycles 1982;19:1191�1195.

16. Stajer G, Szabo EA, Fulop F, Bernath G, Kalman A, Argay G,Sohar P. Synthesis and conformational analysis of stereoisomeric2-oxo- and 2-thioxo-cis- and trans-5,6-methylene-3,4,5,6-tetrahydro-1,3-oxazines. Tetrahedron 1983;39:1829�1836.

17. Bernath G, Fulop F, Kalman A, Argay G, Sohar P, Pelczer I.Connection between the diastereoselectivity and the dominantconformation in the formation of condensed-skeleton 1,3-oxazines,first X-ray diffraction evidence of N-outside conformation. Tetra-hedron 1984;40:3587�3593.

18. Pihlaja K, Mattinen J, Bernath G, Fulop F. A 1H and 13C NMRconformational study of cis- and trans-annelated 2-p-nitrophenyl-4,5- and -5,6-tetramethyleneperhydro- and 2-p-nitrophenyl-4,5- and-5,6-tetramethylenedihydro-1,3-oxazines. Magn Reson Chem1986;24:145�149.

19. Pihlaja K, Mattinen J, Fulop F. A 1H and 13C NMR stereochem-ical study on N-methyl-substituted cis- and trans-fused octahydro-

2H-1,3- and -3,1-benzoxazines. Magn Reson Chem 1996;34:998�1002.

20. Szakonyi Z, Fulop F, Bernath G, Evanics F, Riddell FG. Synthe-sis and ring-chain tautomerism of angularly substituted cyclo-alkane-fused tetrahydro-1,3-oxazines. Tetrahedron 1998;54:1013�1020.

21. Sugita K, Tamura S. Stereochemical studies in Friedel-Crafts re-actions. I. The reactions of cis- and trans-4-tetrahydrophtalic acidand its dimethyl ester with benzene. Bull Chem Soc Japan1971;44:3383�3387.

22. Prausnitz JM, Lichtenthaler RN, Azevedo EG. Molecular ther-modynamics of fluid-phase equilibria, 3rd edition. Upper SaddleRiver, NJ: Prentice-Hall; 1999.

23. Laatikainen R, Niemitz M, Weber U, Sundelin J, Hassinen T,Vepsalainen J. General strategies for total-line-shape type spectralanalysis of NMR spectra using integral transform iterator. J MagnReson 1996;A120:1�10.

24. Armarego WLF. Quinazolines. A stereospecific cis-addition of theelements of nitromethane across a tetrasubstituted ethylenicdouble bond. J Chem Soc C 1971;1812�1817.

25. Crabb TA. Conformational equilibria in azabicyclic systems. In:Lambert JB, Takeuchi Y, editors. Cyclic organonitrogen stereo-dynamics. New York: VCH; 1992. p 253�287.

26. Berger S, Braun S, Kalinowski HO. 15N NMR spectroscopy. In:NMR spectroscopy of the non-metallic elements. Chichester, UK:John Wiley & Sons; 1997. p 111�318.

TAHTINEN ET AL.198