MAGNETIC RESONANCE IN CHEMISTRYMagn. Reson. Chem. 2002; 40: 659–665Published online in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/mrc.1074

Phenylselenenylmenthane derivatives and theirenantiomeric discrimination by 1H and 13C NMRspectroscopy in the presence of a chiral dirhodiumcomplex

Shahid Malik,1 Stefan Moeller,1 Helmut Duddeck1∗ and Muhammad Iqbal Choudhary2

1 Universitat Hannover, Institut fur Organische Chemie, Schneiderberg 1B, D-30167 Hannover, Germany2 H. E. J. Research Institute of Chemistry, International Center for Chemical Sciences, University of Karachi, Karachi-75270, Pakistan

Received 11 April 2002; Accepted 18 June 2002

The phenylselenenylmenthane and -menthene derivatives 1–4 were studied in terms of 1H and 13Csignal assignments, conformational analysis and also complexation shifts (1d) and dispersion effects(1n) observed when non-racemic (ca 2 : 1) mixtures of the chiral selenides were exposed to an equimolaramount of the chiral dirhodium complex Rh∗. The complexation site is the selenium atom exclusively.Whereas 1d values are moderate or small, dispersed signals (split into two owing to the existence ofdiastereomeric adducts) can be observed, many of which are large enough for a facile determination ofenantiomeric ratios of the selenides irregardless of the conformational behaviour of the selenides and theadduct composition. Thus, the ‘dirhodium method’ is simple and reliable for chiral recognition in the classof organoselenium compounds which is gaining increasing importance in developing new techniques ofasymmetric synthesis. Copyright 2002 John Wiley & Sons, Ltd.

KEYWORDS: NMR; 1H NMR; 13C NMR; phenylselenenylmenthanes; dirhodium complexes; chiral recognition; adductformation shifts

INTRODUCTION

In a series of papers we reported studies of the potential ofthe dirhodium complex Rh2(MTPA)4 [RhŁ, (R)-(C)-MTPA-H,methoxytrifluoromethylphenylacetic acid D Mosher’s acid;Scheme 1, top] as a solvating agent for the determinationof enantiomeric ratios of various chiral monofunctionalligands L.1 It has been shown that the ‘dirhodium method’is particularly suitable for soft-base functionalities where theclassical method of chiral lanthanide shift reagents (CLSR)2

fails. Typically, RhŁ and L form kinetically labile adductsso that in equilibria as depicted in Scheme 1, only averagedNMR signals can be observed for the L molecules (in analogywith the CLSR method).

Chiral organoselenium compounds have gained increas-ing importance in asymmetric synthesis during the lastdecade.3 Hence this class of compounds needs an inde-pendent method of chiral recognition, so we decided toreinvestigate it in more detail. During the course of ourearlier studies, we demonstrated that selenides can be dis-criminated by 1H NMR spectroscopy in the presence of RhŁ

if they are chiral.4 To our surprise, and in contrast to previ-ously studied ligands (olefins, epoxides, nitriles, sulfoxides,

ŁCorrespondence to: Helmut Duddeck, Universitat Hannover,Institut fur Organische Chemie, Schneiderberg 1B, D-30167Hannover, Germany. E-mail: duddeck@mbox.oci.uni-hannover.deContract/grant sponsor: Deutsche Forschungsgemeinschaft.

phosphines, sulfides, etc.),1,4 a recent re-investigation by 77SeNMR spectroscopy has shown that the exchange rate of theselenide ligands at the rhodium atoms is low enough tostudy the underlying equilibria by variable-temperature 1H,13C and 77Se NMR spectroscopy.5 Even more, 77Se signalsmay be invisible owing to coalescence effects even at roomtemperature. During the investigation of a primary selenide,5

it turned out that the equilibrium is strongly shifted towardsthe adduct RhŁ Ð Ð ÐL (L D selenium ligand). At room temper-ature and in a 1 : 1 molar ratio, chosen as standard conditionsfor the present study, a fast exchange exists between 1 : 1 and1 : 2 adducts (Se! Rh–Rh and Se! Rh–Rh Se, respec-tively). This situation allows a systematic investigation ofchiral discrimination of selenides (seleno ethers) which con-tain only hydrocarbon residues in order to make sure thatthe selenium atom is the exclusive complexation site. Asmodels, we chose the menthane derivatives 1–4 (Scheme 1)because the rings are conformationally rigid owing the transconfiguration of the two alkyl substituents but the seleniumatoms are in different stereochemical positions. Moreover,these compounds can be synthesized easily in pure enan-tiomeric form.

EXPERIMENTALSynthesesThe synthetic procedure of RhŁ has been described before.1a

The phenyselenenylmenthane and -menthene derivatives

Copyright 2002 John Wiley & Sons, Ltd.

660 S. Malik et al.

ORhRh

O O

O

R

O OR

OO

R

R

ORhRh

O O

O

R

O OR

R

OO

RC

OMe

CF3

Ph

+ L L

Rh* R =

SePhSePh

SePh

SePh

2 3 4

SePh

1

7

89 10

123

456

S

(pro-R ) (pro-S )

Scheme 1. Structures of the phenylselenenyl compounds 1–4. Only (1S)-enantiomers depicted derived from the D-menthol/neomen-thol series; for the enantiomeric and diastereomeric composition of each selenide, see the text and footnotes to Tables 3 and 4. In allfour compounds, the methyl group 9 is pro-R and the methyl group 10 is pro-S.

1–4 were synthesized as described by us earlier.6 Startingmaterials were D-(�)- and L-(C)-neomenthol (for 1), D-(C)-and L-(�)-menthol (for 2) and D-(C)- and L-(�)-menthone (for3), which were converted to the selenides by SN2 reactionwith the respective tosylates.6 Phenyselenenylmenth-1(2)-ene (4) was a by-product in the synthesis of 3. Allselenides were purified by column chromatography usinglight petroleum as eluant and were isolated as yellowish oils.Yields refer to the respective tosylates.

1. Yield: 55%. IR (neat): 3072, 2944, 2920, 1579, 1475, 1436,1229, 1182, 1022, 730, 858, 691 cm�1. EI-MS (relativeintensity, %), m/z [80Se] D 296 (51, MC), 294 (27), 190(72), 176 (43), 158 (34), 142 (72), 139 (73), 128 (41), 105(34), 97 (49), 83 (100). (1R)-1 [L-1]: [˛]20

D D �128.4; (1S)-1(corresponding D-1): [˛]20

D D C128.4.2. Yield: 50%. IR (neat): 3073, 2955, 2917, 1579, 1475, 1436,

1177, 1022, 733, 931, 691 cm�1. EI-MS (relative intensity,%), m/z [80Se] D 296 (24, MC), 190 (82), 176 (41), 142 (80),139 (31), 128 (44), 105 (39), 97 (22), 83 (100). (1R)-2 [L-2]:[˛]20

D D C93.0; (1S)-2 (corresponding D-2): [˛]20D D �93.0.

3. Yield: 20%. IR (neat): 3055, 2955, 2923, 1576, 1475, 1436,1130, 1021, 733, 688 cm�1. EI-MS (relative intensity, %),m/z [80Se] D 452 (3, MC), 314 (97), 296 (64), 294 (98), 234(42), 199 (87), 176 (58), 157 (100), 142 (85), 137 (85), 128(58), 105 (68), 95 (85), 81 (85). (1R)-3 [corresponding L-3]:[˛]20

D D C50.0; (1S)-3 [D-3]: [˛]20D D �50.0.

4. Yield: 15%. IR (neat): 3055, 2919, 2851, 1577, 1435, 1370,1251, 1022, 732, 687 cm�1. EI-MS (relative intensity, %),m/z [80Se] D 294 (100, MC), 279 (12), 251 (8), 198 (15), 157(11), 137 (43), 121 (77), 107 (21), 95 (58), 81 (66), 77 (33).

NMR measurementsRoom-temperature 1H (400.1 MHz) and 13C (100.6 MHz)measurements of the free ligands 1–4 (Scheme 1) wereperformed on a Bruker DPX-400 spectrometer equipped with

a QNP probe head. The ligands were dissolved in 0.7 mlof CDCl3; concentrations were 0.094 mol l�1, in analogywith those in the chiral recognition experiments (see below).Chemical shift standards were internal CHCl3 (υ D 7.24) for1H and the central peak of CDCl3 (υ D 77.0) for 13C. Signalassignments are based on 1Hf1Hg NOE-difference, DEPT,HMQC and HMBC experiments (standard Bruker software).Digital resolutions were 0.24 Hz per point in the 1H and1.53 Hz per point in the 13C NMR spectra.

In typical chiral recognition experiments, 45.2 mg ofRhŁ (0.04 mmol) and a molar equivalent of the respectiveselenide in 0.7 ml of CDCl3, and 7 µl (one drop) of acetone-d6

were added to increase the solubility of RhŁ.1d,e All selenidesamples were prepared by mixing the pure enantiomers in2 : 1 molar ratios; the respective major components were (1R)-1, (1R)-2, (1S)-3 and (1S)-4. Thereby, it was ascertained thatboth diastereomeric adducts experience identical mediumeffects and the signals of the individual enantiomers maybe identified by their signal intensities. In order to verifyoverlapping 1H signals, the experiments in the presence ofRhŁ were repeated with the pure enantiomers in some cases.

The 77Se NMR signals of 1–4 showed significant linebroadenings due to coalescence effects. This phenomenonmay provide information about the adduct formationprocesses5 and is currently under investigation in thecontext of other chiral and achiral phenylselenenylalkanesand -cyclohexanes (H. Duddeck, S. Malik, S. Moeller, T. Gati,G. Toth and M. I. Choudhary, in preparation).

RESULTS AND DISCUSSIONSignal assignments of thephenylselenenylmenthanes 1–41H and 13C signals (see Tables 1 and 2, respectively) wereassigned by using standard correlation techniques (DEPT,COSY, HMQC and HMBC). For 1H resonances inspection

Copyright 2002 John Wiley & Sons, Ltd. Magn. Reson. Chem. 2002; 40: 659–665

Enantiomeric discrimination of phenylselenenylmenthanes 661

Table 1. 1H chemical shifts of 1–4 in CDCl3a

1 2 3 4

H-1 1.29 m 1.99 M 2.01 m 1.68 mH-2 (eq) 2.07 dm 1.99 dm 1.79 dt 2.34 dmH-2 (ax) 1.25 m 1.33 ddd 1.23 dd 1.94 ddmH-3 (eq) — 3.71 m — —H-3 (ax) 3.07 td — —H-4 1.29 tm 1.04 ddt 1.38 dd —H-5 (eq) 1.70 dm 1.82 dm 1.72 dm 2.23 dmH-5 (ax) 1.06 td 1.15 qm 1.49 dddd 2.14 tmH-6 (eq) 1.70 dm 1.77 dm 1.73 dm 1.77 dmH-6 (ax) 0.87 qm 0.91 m 0.69 qd 1.22 mH-7 0.81 d 0.87 d 0.61 d 0.88 dH-8 2.42 sept d 1.69 m 3.02 sept 3.40 septH-9 0.77 d 0.89 d 0.97 d 1.00 dH-10 0.92 d 0.93 d 1.00 d 0.98 dH-ortho (eq) 7.55 dm 7.57 dm

7.37 dmH-ortho (ax) 7.55 dm 7.82 dmH-meta (eq) 7.27 tm 7.26 tm

7.24 tmH-meta (ax) 7.24 tm 7.34 tmH-para (eq) 7.27 tm 7.25 tm

7.21 tmH-para (ax) 7.24 tm 7.33 tm

Samples with RhŁ contained a small amount of acetone-d6 (for details see Experimental).a Spectra recorded at 400.1 MHz.b Vicinal coupling constants involving methyl groups are ca 7 Hz.c Abbreviations d, t, q, sept and m indicate doublet, triplet, quartet, septet and (unresolved)multiplet, respectively.

Table 2. 13C chemical shifts of 1–4 in CDCl3a

1 2 3 4

C-1 34.3 27.7 29.9 30.6C-2 44.9 41.8 48.0 42.6C-3 48.3 50.4 66.8 120.7C-4 47.7 49.5 51.7 148.3C-5 25.0 27.2 23.1 24.8C-6 34.7 35.2 34.8 30.9C-7 22.1 22.1 21.1 21.2b

C-8 29.1 31.3 28.9 34.5C-9 15.1 20.6 24.2 20.5C-10 21.4 21.0 19.0 21.1b

C-ipso (eq) 129.0 129.5 131.6C-ipso (ax) 130.8 126.6C-ortho (eq) 135.5 138.1

131.2C-ortho (ax) 134.1 138.3C-meta (eq) 128.8 128.5

128.9C-meta (ax) 128.8 128.5C-para (eq) 127.3 128.6

126.0C-para (ax) 126.9 128.9

Samples with RhŁ contained a small amount of acetone-d6

(for details see Experimental).a Spectra recorded at 100.6 MHz.b Not resolvable from HMQC; may be interchanged.

of the signal multiplicities in terms of 1H,1H couplings andNOE contacts were most helpful. If not obscured by overlap,

signals of axial hydrogens can be identified by large vicinalcouplings in antiperiplanar configurations (J D 10–12 Hz)whereas 3J values for equatorial hydrogens are much smallerin the gauche orientations (2–5 Hz) (Fig. 1). Except forfew cases, unambiguous spectral interpretations includingstereochemical hydrogen assignments could be achieved.

Cyclohexane ring inversion is excluded in the menthanederivatives 1–3 because the 1-methyl and the 4-isopropylgroups retain their equatorial position. Therefore, the PhSesubstituent is equatorial in 1 and axial in 2, whereas bothpositions are occupied by the geminal selenium atoms in3. However, the conformational preference of the isopropylgroups varies.

Figure 1. Section of the 400 MHz 1H NMR spectrum of 2, inCDCl3.

Copyright 2002 John Wiley & Sons, Ltd. Magn. Reson. Chem. 2002; 40: 659–665

662 S. Malik et al.

The equatorial selenium atom in 1 forces the isopropylgroup into a position with one methyl (9-CH3) to be antiperi-planar with respect to H-4 (Scheme 2) in order to avoida strongly disfavoured skew-pentane-type arrangement (C-9—C-8—C-4—C-3—Se). This can be proved by the 13Cchemical shifts of C-9 and C-10; the carbons see a differentnumber of �-gauche oriented groups suffering from differentextents of diamagnetic signal shifts: C-9 two �-gauche inter-actions with υ�13C� D 15.1; C-10 only one �-gauche interactionwith υ�13C� D 21.4 (Scheme 2, Table 2).7 Moreover, the axiallypositioned H-3 shows a strong NOE contact with H-9 but noNOE with H-10. On the other hand, the same conformation-energy argument (skew-pentane) stabilizes the position ofboth methyl groups C-9 and C-10 gauche with respect toH-4 in the neomenthane derivative 2 with only one �-gaucheinteraction for each and υ�13C� D 20.6 and 21.0, respectively(Scheme 2, Table 2). The vicinal H-4/H-8 coupling constantis 10.0 Hz in accordance with the antiperiplanar orientationof these atoms. Interestingly, the C-9 and C-10 chemical shiftsare υ D 24.2 and 19.0, respectively, in the bis-seleno acetal3. This leads to the conclusion that the isopropyl group in3 prefers a conformation similar to that of 1, but somewhatdistorted (Scheme 2). The distortion is ca 30° as suggestedby AM1 calculations; any other rotamer is at least 15–20 kJmol�1 less stable. In 3, the atom C-9 (υ D 24.2) experiences adeshielding skew-pentane-type υ-effect8 from the axial sele-nium atom instead of the �-gauche-effect from C-3/H-3 in 1.The chemical shift of C-10 is υ D 19.0. Here, a �-gauche effectsmaller than those on C-10 of 1 or C-9 of 2 is effective inaccordance with the larger spatial distance by the distortion.A further confirmation comes from the 3J(H-4,H-8) couplingconstant, which is very small (0–1 Hz) because the torsionangle is close to 90°.

1

9CH3

H8H3C

5

810

H2C C3

H4

SePh

2

H8

CH3H3C

5

9 108

H2C C3

H4

SePhH3

H3

3

H3C9

H8

H3C

5

108

H2C C3

H4

SePh

SePh

CCH3

SePhH8

H3C

H3C9

H

H5(qa)

34

8

10

17

4

pro-R

pro-S

5(qe)

Scheme 2. Preferred conformations of the isopropyl groups in1–4.

Finally, the menthene derivative 4 with only one chiralcenter adopts a half-chair conformation with a quasi-equatorial methyl at C-1, and the preferred conformationof the isopropyl group is as shown in Scheme 2, i.e. H-8points towards the selenium atom and is slightly out-of-plane (15–20° according to AM1 calculations). This allowedus to assign the two diastereotopic methyl groups 9 and 10:whereas irradiation of the methyl signal at υ D 1.00 produces

NOE responses for both H-5 protons, that at υ D 0.98enhanced the intensity of the quasi-equatorial H-5(qe) only.Hence the former signal was assigned to H-9 and the latterto H-10.

Complexation shifts (1d) in the presence of Rh∗

The complexation site in the ligand molecules 1–4 is theselenium atom; phenyl rings are not able to bind at therhodium atom(s)1c,4 (it should be noted that chiral recognitionof aromatic substrates by 31P NMR spectroscopy has beenclaimed for mononuclear rhodium complexes containingphosphine ligands9). As mentioned above,5 phenylselenidesform 1 : 1 and 2 : 1 complexes with RhŁ which are in a fast-exchange equilibrium with free RhŁ because both partnersexist in equimolar ratios. Therefore, this ratio was chosen forthe present study.

Unambiguous signal assignment is mandatory for areliable evaluation of complexations shifts and dispersioneffects (see below) by adduct formation of the selenides 1–4with RhŁ. In these experiments, signal assignments are basedon the same 1D and 2D experiments as for the free selenidesand were secured by comparative experiments with pureenantiomers under identical conditions. Nevertheless, somesignals could not be identified beyond doubt, in particularthose of the aromatic hydrogens which are obscured by themuch larger 1H signals of RhŁ.

1H and 13C chemical shifts in the ligand molecules may bechanged by forming adducts between RhŁ and the selenideligands. By definition, complex or adduct formation shifts(υ) are the differences between chemical shifts of ligandatoms in the adducts relative to the respective ones in thefree ligands. In agreement with all previous observationsfor selenides and other monofunctional ligands,1,4,5 υ aremoderately positive (deshielding effect) for atoms close tothe complexed atom in terms of intervening bonds. This canindeed be observed for the selenides 1–4; typical positiveυ values exist only for the H-3 and H-4 atoms (C0.09 toC0.70) and for the C-3 atoms (C2.1 toC4.0 ppm) (see Tables 3and 4).

However, negative υ values (shielding) of 1H signalsmay be observed in certain regions of the molecules,particularly for atoms further away from the complexationsite. This is the case for H-6(eq) and H-7, the atomswith the largest distances to the complexation site (Se). Areasonable explanation is as follows: The nuclei experiencea diamagnetic ring current effect if they come into closecontact with the phenyl groups within the Mosher acidresidues; electric field effects from the methoxy and/or thetrifluoromethyl groups may also be contributing.

Alternatively, significant υ values may arise fromconformational distortions of parts of the ligand moleculesduring adduct formation, compared with the free ligandmolecule. However, it is evident from the 1H and 13C signalsthat the isopropyl groups in 1–4 do not experience significantcomplexation shifts. Hence it can be concluded that theconformational preferences of this group are hardly affectedby complexation with RhŁ. This is different for H-1 and C-1in 2 and also for H-2(eq) and C-2 in most compounds 1–4. Atentative explanation is the flexibility of the phenyl group at

Copyright 2002 John Wiley & Sons, Ltd. Magn. Reson. Chem. 2002; 40: 659–665

Enantiomeric discrimination of phenylselenenylmenthanes 663

Table 3. Complexation effects (υ, in ppm) on the 1H chemical shifts of 1–4 in CDCl3 containing a small amount of acetone-d6 (fordetails see Experimental)a

1 2 3 4

1R(major)

1S(minor)

1R(major)

1S(minor)

1R(minor)

1S(major)

1R(minor)

1S(major)

H-1 ¾C0.4 ¾C0.4 �0.68 �0.68 n.d. n.d. ¾�0.1 ¾�0.1H-2 (eq) C0.45 C0.40 C0.71 C0.63 C0.17 C0.17 C0.24 C0.17H-2 (ax) C0.18 C0.25 C0.16 n.d. n.d. n.d. ¾C0.1 ¾C0.1H-3 (eq) — — C0.70 C0.68 — — — —H-3 (ax) C0.55 C0.57 — — — —H-4 ¾C0.4 n.d. C0.09 C0.11 n.d. n.d. — —H-5 (eq) ¾0 ¾0 �0.03 �0.03 C0.11 C0.07 C0.02 �0.02H-5 (ax) C0.02 n.d. C0.22 C0.22 C0.02 C0.02 ¾�0.1 ¾�0.1H-6 (eq) �0.11 n.d. �0.20 �0.16 n.d. n.d. �0.13 �0.13H-7 (ax) ¾�0.1 ¾�0.1 �0.09 �0.09 C0.08 C0.07 ¾�0.5 ¾�0.5H-7 �0.16 �0.25 �0.30 �0.30 �0.17 �0.08 �0.24 �0.27H-8 C0.51 C0.51 C0.25 C0.25 n.d. n.d. C0.38 C0.39H-9 C0.08 C0.04 �0.11 �0.03 �0.08 �0.04 0.00 �0.02H-10 C0.05 C0.07 0.00 0.00 �0.02 �0.02 C0.07 C0.07H-ortho(eq) C0.34 C0.36 �0.04 C0.02 C0.43 C0.41H-ortho (ax) C0.43 C0.38 n.d. n.d.H-meta (eq) n.d. n.d. n.d. n.d. �0.05 �0.05H-meta (ax) n.d. n.d. n.d. n.d.H-para (eq) n.d. n.d. n.d. n.d. �0.08 �0.08H-para (ax) n.d. n.d. n.d. n.d.

a For each compound: left, (1R)-enantiomer and right, (1S)-enantiomer; ‘major’, major constituent; ‘minor’, minor constituent in thenon-racemic mixture; n.d., not detectable safely. Spectra recorded at 400.1 MHz.

Table 4. Complexation effects (υ, ppm) on the 13C chemical shifts of 1–4 in CDCl3 containing a small amount ofacetone-d6 (for details see Experimental).a

1 2 3 4

1R 1S 1R 1S 1R 1S 1R 1S(major) (minor) (major) (minor) (major) (minor) (major) (minor)

C-1 �0.5 �0.6 �0.7 �0.9 C0.1 C0.1 0.0 0.0C-2 �4.0 �3.7 �1.2 �1.6 C0.1 C0.1 �2.6 �2.8C-3 C4.0 C3.8 C2.7 C2.4 C3.2 C2.9 C2.1 C2.2C-4 �2.3 �2.3 C0.8 C0.8 C0.2 C0.3 C1.5 C1.4C-5 �0.1 �0.1 �0.5 �0.5 C0.6 C0.6 C0.9 C1.3C-6 �0.5 �0.3 C0.1 C0.1 �0.5 �0.7 0.0 0.0C-7 �0.5 �0.5 �0.3 �0.3 �0.2 �0.1 C0.1b �0.2b

C-8 0.0 0.0 �0.2 �0.2 C0.3 C0.2 0.0 0.0C-9 �0.3 �0.3 C0.3 C0.2 C0.1 0.0 C0.3 C0.3C-10 �0.2 �0.2 0.0 0.0 �0.1 �0.1 C0.6b C0.8b

C-ipso (eq) n.d. n.d. �2.5 �2.5 C1.2 C1.3C-ipso (ax) �1.9 �1.7 C0.1 C0.2

C-ortho (eq) C0.8 C0.9 �0.1 �0.2 C2.0 C2.1C-ortho (ax) C2.1 C2.2 �0.1 �0.1C-meta (eq) n.d. n.d. C0.2 C0.2 C0.6 C0.4C-meta (ax) �0.2 �0.2 C0.2 C0.3C-para (eq) n.d. n.d. �0.3 �0.3

n.d. n.d.C-para (ax) C1.7 C1.7 �0.4 �0.4

a For each compound: left, (1R)-enantiomer and right, (1S)-enantiomer; ‘major,’ major constituent; ‘minor,’: minor constituentin the non-racemic mixture; n.d., not detectable safely. Spectra recorded at 100.6 MHz.b Values for C-7 and C-10 may be interchanged pairwise for each enantiomer.

Copyright 2002 John Wiley & Sons, Ltd. Magn. Reson. Chem. 2002; 40: 659–665

664 S. Malik et al.

the selenium atom leading to different average orientationswith respect to the C-1/C-2 fragments of 1–4. In particular,the C-2/H-2(eq) bond being close in space with respectto the selenium atom(s) appears to suffer from a distinctbond polarization from the hydrogens (υ > 0) towards thecarbons (υ < 0).

Among the phenyl protons, only those in ortho-positionsare significantly affected; they are deshielded in the adductwith respect to the free ligands. The ipso-carbons, how-ever, are somewhat shielded whereas slight deshielding isobserved for most of the ortho- and para-carbons but not forthe meta-carbons. This suggests a change of the inductiveand mesomeric properties of the selenium atoms by bindingto rhodium.

Diastereomeric dispersion effects (1n):determination of absolute configuration?If non-racemic mixtures of the ligand molecules 1–4 areadded to equimolar amounts of RhŁ, diastereomeric adductsare formed and can be discriminated by their different 1Hand 13C signals. In principle, each signal of the free ligandis split into two. The distance within such signal pairs iscalled diastereomeric dispersion (�; in Hz). Note that �

values are field dependent; in this study they are referredto 400.1 MHz (1H) and 100.6 MHz (13C). Owing to differentquantities of the enantiomers in each sample, it is possibleto identify most of the signals of each by its intensity.All identified � values are compiled in Table 5. They aredefined as

�n�i� D �n[�1R�� i]� �n[�1S�� i][in Hz]

i.e. the chemical shift of a given hydrogen or carbon atom n ofthe 1S-enantiomer (in Hz) is subtracted from the respectivevalue in the 1R-enantiomer regardless of which is the majorand which the minor constituent; i is the selenide 1, 2, 3, or4. Thereby, stereochemical influences of the chiral Mosheracid residues in RhŁ can be compared within the series ofselenides via the signs of the � values.

Even a brief overview of the data in Table 5 shows thatmost hydrogen and carbon atoms reflect the stereochemicaldifference in the diastereomeric adducts; some hydrogensdisplay signal splittings up to 35 Hz. As an example, Fig. 2shows signal dispersions in the 1H NMR spectrum of 1. Itcan be seen that it is easy to determine the enantiomeric

Table 5. Dispersion effects (�, Hz) at the 1H and 13C chemical shifts of 1–4 in CDCl3 containing asmall amount of acetone-d6 (for details see Experimental)a

1H 13C

Atom 1 2 3 4 1 2 3 4

1 n.d. ¾0 n.d. 0.0 C14 C21 ¾0 �292 (eq) C17.8 ¾C33 n.d. �21.2 �23 C26 �5 C132 (ax) ¾C45 n.d. n.d. n.d.3 (eq) — C10.8 — — C24 C28 C34 �153 (ax) ¾�10b — — —

4 ¾C17b ¾�6 n.d. — 0 C6 —9 C85 (eq) <2 ¾0 n.d. n.d. �5 �5 C4 �225 (ax) <2 ¾0 n.d. n.d.6 (eq) n.d. ¾�15 n.d. 0.0 �15 0 �18 06 (ax) n.d. ¾ 0 ¾C3 0.0

7 C35.4 �3.1 �35.6 �7.3 0 0 �8 �128 ¾0 ¾0 n.d. �2.4 ¾�1 C4 C7 0.09 C15.4 �31.2 �15.0 �6.4 C3 C2 C9 C810 �7.8 0 ¾�1 ¾�1 0 0 ¾0 �12

ipso (eq) — — ——

n.d. ¾C1-12

ipso (ax) — — ¾C16 �6Ortho(eq) �10.0 �26.8 ¾�7

�14 C7c

�12Ortho (ax) C14 n.d. �3 ¾0c

meta (eq) n.d. n.d.n.d.

n.d. �7d

C15meta (ax) n.d. n.d. �2 �10d

para (eq) n.d. n.d.n.d.

n.d. �6n.d.

para (ax) n.d. n.d. 0 �2

a The � values (Hz, at 400.1 and 100.6 MHz, respectively) are defined as follows: the chemical shift (inHz) of the (1S)-enantiomer is subtracted from the respective value for the (1R)-enantiomer, regardlesswhich is the major and which the minor constituent. Note that � values are field dependent; in thisstudy they are referred to 400.1 MHz (1H) and 100.6 MHz (13C); n.d., not detectable safely.b Signal of the 1S-enantiomer is not detectable in the 1D spectrum. However, from the HMQC experimentand/or by signal overall-width comparisons, we expect a dispersion as indicated.c,d May be interchanged pairwise.

Copyright 2002 John Wiley & Sons, Ltd. Magn. Reson. Chem. 2002; 40: 659–665

Enantiomeric discrimination of phenylselenenylmenthanes 665

Figure 2. Section of the 400 MHz 1H NMR spectrum of 1 inthe absence (bottom) and presence (top) of an equimolaramount of RhŁ.

purity of selenides by inspecting the intensity ratios of 1HNMR signals.

Short-lived adducts such as those of RhŁ with selenidesand other ligands1,4 appear to be very flexible so that anyeffect on NMR signals (υ or �) is time-averaged. Generally,within the series 1–4 the magnitudes and particularly thesigns of the � values vary strongly for most of the 1H and13C atoms. Since the carbon skeletons are very similar, weare left with the assumption that the four ligands 1–4 adoptvery divergent average orientations relative to RhŁ owing tothe different stereochemical positions of the selenium atomsacting as binding sites. Therefore, we have to state that it isnot very promising to read absolute configurations10 fromthe signs of � values of the NMR signals.

CONCLUSIONS

Phenylselenenylmenthanes and -menthenes easily formshort-lived adducts with RhŁ. Thereby, deshielding effectsof nearby hydrogens and carbons (in terms of the number ofintervening bonds) are observed (υ), whereas some moreremote nuclei are shielded. The phenyl orientation is changedwhereas the isopropyl conformations are retained.

Many NMR signals are dispersed owing to the existenceof diastereomeric adducts (�) when the chiral phenylse-lenenylmenthanes and -menthenes are bound to RhŁ. Adetermination of enantiomeric ratios is facile regardless ofthe stereochemical position of the selenium atom as thecomplexation site. This constitutes a simple and valuablemethod for chiral recognition in this class of compounds,which is gaining increasing importance in developing newmethodologies of asymmetric synthesis.

AcknowledgementsThis work was performed within the project ‘Biologically ActiveNatural Products: Synthetic Diversity’ (Department of Chem-istry, Hannover University). It was supported by the DeutscheForschungsgemeinschaft and by a fellowship granted to S.M. by theGerman Academic Exchange Service (DAAD).

REFERENCES1. (a) Wypchlo K, Duddeck H. Tetrahedron: Asymmetry 1994; 5:

27; (b) Wypchlo K, Duddeck H. Chirality 1997; 9: 601; (c)Hameed S, Ahmad R, Duddeck H. Magn. Reson. Chem. 1998; 36:S47; (d) Meyer C, Duddeck H. Magn. Reson. Chem. 2000; 38: 29;Meyer C. PhD Thesis, Hannover University, 1999; (e) Rockitt S,Duddeck H, Omelanczuk J. Chirality 2001; 13: 214; (f) Malik S,Duddeck H, Omelanczuk J, Choudhary MI. Chirality 2002; 14:407.

2. (a) Sullivan GR. Top. Stereochem. 1978; 10: 287; (b) Rinaldi PL.Progr. Nucl. Magn. Reson. Spectrosc. 1983; 15: 291; (c) Parker D.Chem. Rev. 1991; 91: 1441; (d) Rothchild R. Enantiomer 2000; 5:457.

3. (a) Wirth T. Tetrahedron 1999; 55: 1; (b) Wirth T (ed.).Organoselenium Chemistry, Modern Developments in OrganicSynthesis. Topics in Current Chemistry, vol. 208. Springer: Berlin,2000.

4. Hameed S, Ahmad R, Duddeck H. Heteroat. Chem. 1998; 9: 471.5. Duddeck H, Malik S, Gati T, Toth G, Choudhary MI. Magn.

Reson. Chem. 2002; 40: 153.6. (a) Duddeck H, Wagner P, Gegner S. Tetrahedron Lett. 1985; 1205;

(b) Duddeck H, Wagner P, Biallass A. Magn. Reson. Chem. 1991;29: 248; (c) Duddeck H, Wagner P, Rys B. Magn. Reson. Chem.1993; 31: 736.

7. (a) Senda Y, Imaizumi S. Tetrahedron 1975; 31: 2905; (b) Firl J,Kresze G, Bosch T, Arndt V. Liebigs Ann. Chem. 1978; 87, andreferences cited therein.

8. Grover SH, Stothers JB. Can. J. Chem. 1974; 52: 870.9. Buriak JM, Osborn JA. J. Chem. Soc., Chem. Commun. 1995; 689.

10. (a) Uray G. In Houben–Weyl: Methods in Organic Chemistry,vol. 1, Helmchen G, Hoffmann RW, Mulzer J, Schaumann E,(eds). Georg Thieme: Stuttgart, 1996; 253; (b) Seco JM, Quinoa E,Riguera R. Tetrahedron: Asymmetry 2001; 12: 2915.

Copyright 2002 John Wiley & Sons, Ltd. Magn. Reson. Chem. 2002; 40: 659–665