Stereochemistry of
`Organic Compounds
`
`ERNEST L. ELIEL
`Department of Chemistry
`The University of North Carolina at Chapel Hill
`Chapel Hill, North Carolina
`
`SAMUEL H. WILEN
`Department of Chemistry
`The City College of the City University of New York
`New York, New York
`
`With a Chapter on Stereoselective Synthesis by
`
`LEWIS N. MANDER
`Research School of Chemistry
`Australian National University
`Canberra, Australia
`
`A Wiley-lnterscience Publication
`JOHN WILEY & SONS, INC.
`New York
`• Chichester
`• Brisbane
`
`• Toronto
`
`• Singapore
`
`PARAGON - EXHIBIT 2022
`
`

`
`This text is printed on acid-free paper.
`Copyright © 1994 by John Wiley & Sons, Inc. ·
`All rights reserved. Published simultaneously in Canada.
`
`Reproduction or translation of any part of this work beyond
`that permitted by Section 107 or 108 of-the 1976 United
`States Copyright Act without the permission of the copyright
`owner is unlawful. Requests for permission or further
`information should be addressed to the Permissions Department,
`John Wiley & Sons, Inc., 605 Third Avenue, New York, NY
`10158-0012.
`
`Library of Congress Cataloging in Publication Data:
`Eliel, Ernest Ludwig, 1921-
`Stereochemistry of organic compounds I Ernest L. Eliel, Samuel H.
`Wilen, Lewis N. Mander.
`p. em.
`"A Wiley-Interscience publication."
`Includes index.
`ISBN 0-471-01670-5
`1. Stereochemistry. 2. Organic compounds. I. Wilen, Samuel H.
`II. Mander, Lewis N. III. Title.
`QD48l.E52115 1993
`547.1'223--dc20
`
`93-12476
`
`Printed in the United States of America
`
`10 9 8 7 6 5 4 3 2 1
`
`

`
`1071
`
`ascribed to conformational rigidity and to the presence of isotactic helical
`of a preferred helical sense in the copolymer. A similar example of such a
`cooperative phenomenon (with a resulting high optical rotation) involving a
`·
`consisting of mostly achira:l poly(n-hexyl isocyanate) incorporating as
`as 0.12mol% of a nonracemic chiral isocyanate [(S)-(-)-2,2-dimethyl-1,3-
`isocyanate] has been described (Green et al., 1989).
`
`Attachment of achirotopic chromophoric groups such as dehydrophenylalanine and
`azobenzene to polypeptides [e.g., poly(L-glutamic acid)] generates synthetic macro(cid:173)
`molecules whose CD spectra reflect the presence of inherently chiral chtomo(cid:173)
`phores. This spectral feature results because the pendant side chains serve as
`chirality reporters of the secondary structure of the peptides ( Ciardelli and Pieroni,
`1980).
`
`In subsequent studies on dimethyloctene-styrene and similar copolymers, the
`CD could be measured deeper into the UV spectral region. Other CD bands
`typical of isolated benzenoid electronic transitions CB and 1La) signaling the
`chirotopicity of the benzene ring were observed. In contrast to model compounds,
`such as the conformationally restricted 2-phenyl-3,3-dimethylbutane (Salvadori et
`. al., 1972), the copolymer exhibits, in addition, an exciton-like couplet centered at
`about 190 nm. The latter was attributed to a helical conformation in which
`benzene rings in the same chain are sufficiently close to couple. The enantiomer
`. exhibiting negative chirality incorporates a right handed helix ( Ciardelli et al.,
`1972). The CD of such helical copolymers can be calculated by means of a
`classical theory developed by DeVoe (DeVoe, 1969, 1971; see Hug et al., 1974 and
`Section 13-4.a).
`It is only since the 1980s that CD studies have permitted the observation of
`CEs in the vacuum UV region of hydrocarbon polymers devoid of aromatic
`groups. The CD spectra of films of poly-(S)-4-methyl-1-hexene and poly-(R)-3,7-
`dimethyl-1-octene exhibit a CD band at 158 nm that is ascribed to conformations
`containing helical segments having a common helix sense ( Ciardelli and Salva(cid:173)
`dori, 1985).
`
`13-5. APPLICATIONS OF OPTICAL ACTIVITY
`
`a. Polarimetry
`
`The actual measurement of optical activity may be carried out with either manual
`or photoelectric polarimeters. Manual polarimeters have changed relatively little
`, since the first instruments were developed some 140 years ago (Lowry, 1964, p.
`180). Photoelectric polarimeters, the type nowadays commonly found in research
`laboratories, have greatly reduced the tedium formerly associated with the
`measurement of optical rotation with manual instruments. Moreover, photoelec(cid:173)
`tric polarimeters are much more accurate and sensitive, permitting the rapid and
`meaningful recording of quite small absolute rotation values a to about ±0.002°
`and, consequently, the use of smaller samples. Polarimeters fitted with microcells
`may even serve advantageously as detectors in HPLC resolutions (Mannschreck,
`
`

`
`1072
`
`Chiroptical Propertie{· .... ·
`
`·.
`Eigelsperger, and Stuhler, 1982; Mannschreck, 1992; Pirkle, Salvadori, et al.;.
`1988; Lloyd and Goodall, 1989). A laser-based polarimetric HPLC detector h~s · ·
`been shown to be sensitive to as little as 12 ng of sample (Yeung et al., 1980). For· .
`the advantages of the use of CD detectors in HPLC, see Salvadori, Bertucci, and
`Rosini (1991); Mannschreck (1992).
`The laser polarimetric detector has been adapted to HPLC analysis not only
`of optically active samples but also, in a different way, as a universal detector for
`achiral, that is, optically inactive, substances. In this technique, termed "indirect .
`polarimetry," the mobile phase is optically active, containing, for example.
`(-)-2-methyl-1-butanol or ( + )-limonene, and the detector output due to the
`optically active solvent is zeroed. Under these conditions, any optically inactive ·
`fraction passing through the detector cell is sensed since the concentration of the'
`optically active solvent is thereby reduced. The response of the detector is
`universal, like that of a refractive index detector, but is more sensitive than the. · · · ·
`latter (Bobbitt and Yeung, 1984, 1985; Yeung, 1989). The simultaneous measure- .
`ment of absorbance and optical rotation during the liquid chromatographic . .
`resolution of chiral substances on enantioselective stationary phases make pbs: . ·
`sible the determination of the enantiomer composition in spite of extensive peak-- ~­
`overlap (Mannschreck et al., 1980; Mannschreck, Eigelsperger, and Stuhler,· ·. ·
`1982; compare Drake, Gould, and Mason, 1980).
`·:"·
`For a brief discussion of polarimetry and its instrumentation, see the review·.:· ·
`by Lyle and Lyle (1983); for a more extensive treatment, see Heller and Curme':·
`(1972).
`.
`The measurement of optical activity has traditionally been the method -of ; ·
`choice to establish the nonracemic character of a sample of a chiral compound·.·' ..
`and, when quantitatively expressed as a ratio [a] I [a] max, of its enantiomeric
`composition (optical purity). In contemporary practice, chiroptical measurements
`have to a large extent been replaced by NMR and by chromatographic analyses :.:
`for the purpose of determining enantiomeric compositions (cf. Chapter ,6).'~:
`Nevertheless, the use of [a] for this and other purposes continues. The reasons. ·
`are that the measurement is easy to carry out and one may wish to compar~.
`experimental values of [a] with those in the literature. While substantial coliec-: .
`tions of optical rotation data exist, for example those in various handbooks and< .
`chemical supplier catalogs, it should not be assumed that values of [a] provide·d'·:
`are those of enantiomerically pure compounds. A consistent set of spe<;tfi~­
`rotation data for amino acids including temperature coefficients has been com~.
`piled by ltoh (1974).
`.
`..
`Optical activity has been used (a) to determine if a given unknown substap:ce. ~-.
`is chiral or achiral; (b) to ascertain the enantiomeric composition of _chi~al.
`samples, either qualitatively or quantitatively; (c) to study equilibria; the m.uta~:
`rotation or change in rotation of equilibrating stereoisomers as a function ofJime0 . _·.
`is one such phenomenon (Eliel, 1962; for a recent example, see Arjona;: :
`Perez-Ossorio, et al., 1984); and (d) to study reaction mechanisms. Qt~~l': · ·
`chiroptical techniques, namely, ORD and CD, have increasingly replaced pol:ari-: ..
`metry in these applications, especially in the past 20 years. For review( of
`applications of polarimetry, see Lowry (1964), Eliel (1962), Legrand and Roumep. -·
`(1977), and Purdie and Swallows (1989).
`· · ··
`Polarimetric methods remain useful for quality control in pharmacology
`
`

`
`Applications of Optical Activity
`
`1073
`
`food-related industries (Lowman, 1979; Chafetz, 1991); there are also numerous
`applications in forensic, clinical, pharmaceutical, and agricultural chemistry (Pur(cid:173)
`die and Swallows, 1989). The percentage of sucrose in commercial samples is still
`being determined by polarimetry (saccharimetry); in ·the trade this is called
`"direct polarization". The cost of raw sugar is based on the results of the
`polarimetric analysis; if the analyte solution is dark, the raw sugar is first clarified
`· by precipitation of the dark side products with basic lead acetate (Cohen, 1988).
`An example of application d (above) is the methanolysis of the tosylate of
`(R)-( + )-C6H 5CH2SCH2CH(CH3)CH20H that leads to a partially racemized
`methyl ether. The intervention of a cyclic (symmetrical, and hence achiral)
`intermediate, via neighboring group participation, was inferred (Eliel and Knox,
`1985).
`The magnitude of rotation a, in degrees, fundamentally depends on the
`number of molecules of the sample being traversed by the linearly polarized light
`as well as on their nature, hence optical activity is not a colligative property.
`Values of a are affected by many variables, among which are wavelength, solvent,
`concentration, temperature, and presence of soluble impurities. It must also be
`mentioned that large molecules, such as proteins, may spontaneously orient
`themselves in solution, and consequently no longer be isotropic. The measure(cid:173)
`ment of the rotatory power of such substances may then be complicated by the
`occurrence of linear dichroism (see Heller and Curme, 1972, p. 67).
`As already pointed out in Section 1-3 (q.v.), rotation magnitudes are usually
`normalized to a quantity called the specific rotation [a] that was introduced by
`Biotin 1835 (Biot, 1835; cf. Lowry, 1964, p. 22), Eq. 13.28,
`
`[a]= a/tp = altc
`(13.28)
`where e is the length of the cell in decimeters, p (for undiluted liquids) is the
`density in grams per milliliter (g mL - 1
`) and c is the concentration also in grams
`per milliliter. The units of [a] are 10- 1 deg cm2 g- 1 (see also Eq. 1.1 and Section
`1-3).
`Comparison of specific rotations of homologues, and of organic compounds
`generally, is more significant if a modified Biot equation is used in which the
`quantity called the molar rotation [<I>] depends on the number of moles of
`substance traversed by the linearly polarized light, Eq. 13.29.
`
`[<I>]= [a]M/100
`
`(13.29)
`
`where M is the molecular weight. The units of [<I>] are 10 deg cm2 mol- 1 (see also
`Eq. 1.2) (IUPAC, 1986).
`The cumulative effect of the above-mentioned variables on [a] or [<I>] is
`potentially very large. A practical consequence is that precise reproduction of
`published rotation values, from laboratory to laboratory, or even from day to
`day in the same laboratory, is difficult to achieve (Lyle and Lyle, 1983).
`This sensitivity to numerous variables (Schurig, 1985) and the absence of
`major tabulations of critically evaluated absolute rotation data is responsible
`for the decreasing reliance on optical activity as a measure of enantiomeric
`composition .
`
`.
`·~·.
`
`. ·
`. .
`
`

`
`1074
`
`Chiroptical Properties·
`
`Much preliminary work may be necessary to increase the low accuracy of routine
`optical rotation measurements. For an outstanding example of a study of some of
`the variables affecting the rotation of a-methylbenzylamine, mandelic acid, and
`ephedrine, see the dissertation by Zingg (1981).
`
`.
`
`The sign of rotation is often the only experimental criterion for the specifica- ·
`tion of configuration. It is important to stress how frequently and how easily this.
`property may change for a given substance, for example, (R)-2-hydroxy-1,1'~
`binaphthyl has [a]~0 +4.77 [c = 0.86, tetrahydrofuran (THF)], +13.0 (c = 1.12,
`.
`w
`THF), and [a] 0 -5.2 (c = 1.03, CH30H) (Kabuto et al., 1983). Even tartaric ·
`acid, one of our configurational standards, does not exhibit an invariant sense of .
`rotation: [<I>] 578 -12.9 (24°C), -0.9 (57°C), and +10.8 (94°C) (all c = 10; ··
`dioxane); +21.3 (24.7°C, H20), +6.6 (24°C, EtOH), +0.3 [25.3°C, N,N-
`·
`dimethylformamide (DMF)], -12.9 (24°C, dioxane), and -14 (25.2°C, Et20) (ll,ll
`c = 10) [both sets of data measured on (R,R)-tartaric acid] (Hargreaves and
`Richardson, 1957).
`When the sign of rotation shows a strong solvent, concentration, wavelength·,. · · ·
`or temperature dependence, the association of such sign with a given configura~ ..
`tional descriptor is arbitrary. In the case of 4-amino-1-( diethylamino )pentane·
`(Fig. 13.70), 37 [a] is less than 0 in ethanol (365-589 nm) but greater than 0 when
`measured in the absence of solvent (neat). The configuration of the sample
`derived from L-glutamic acid was referred to as (R)-(-) rather than (R)-( +}. · ·
`because of the much larger magnitude of the rotation of the neat sample over that.
`in ethanol (Craig et al., 1988). This, and the other examples cited, serve to
`emphasize the crucial importance of specifying and recording the precise ex-·
`perimental conditions of measurerp.ent of optical rotations and of chiroptical. ·.
`properties in general. In particular, confusion can arise when the sign of rotation
`is related to a given configuration and the solvent is not specified.
`Occasionally, specific rotations of samples are very small. When that situation · ··
`arises, especially in resolutions or in stereoselective syntheses in which strongly
`rotating reagents are used, exceptional care must be taken to insure that th~. ·.
`rotation of the product is not spurious. A small amount of impurity having a large'· ·
`[a] may overwhelm (or at least seriously falsify) the rotation of a sample having·.a · ·
`small [a] (for a problem case, see Baldwin et al., 1969; see also Goldberg et at.;
`1971). Achiral contaminants, particularly solvents, will also affect the opti¢ai ··
`activity of a sample (Lyle and Lyle, 1983; Schurig, 1985).
`. ..
`Traces of achiral compounds, including solvent residues, normally would be.
`expected to reduce [a] (by dilution of the sample) and hence to artificially lower . ·.
`the optical activity of nonracemic samples (but not the enantiomeric purity of
`chiral solute). However, the converse may also be observed, for example, the.:
`optical activity of 1-phenylethanol at 589 nm is increased when acetophenone, ·a
`possible contaminant, is present in the sample (Yamaguchi and Mosher, 197~)~.
`The enhanced optical activity of the alcohol comes about because
`·
`.
`properties are induced in the achiral ketone by the alcohol; in the example,
`induction is superimposed on and swamps the typical and opposite dilution e:ffe~t ·
`(cf. Section 13-4.e).
`The case of low rotation warrants further comment. There are two sltlJati.O!lS
`in which no optical rotation is observed with enantiomerically enriched ., .. ., .. ..,~,,~, ,,
`
`

`
`· ... Applications of Optical Activity
`
`1075
`
`(a) the experimental device used (by implication this includes the eye) is of
`insufficient sensitivity, and (b) the specific conditions of measurement are such
`·.·that a is, in fact, accidentally equal to zero.
`In the first situation, the measurement threshold is such that there is no clear
`signal (rotation) distinguishable from iJ1.Strumental noise. The condition is one of
`· operational null. Progressive dilution of a solution of an optically active com(cid:173)
`. pound eventually leads to a sample that is no longer palpably optically active
`when the operational null threshold is crossed. Such a sample no longer reveals its
`enantiomeric excess; the sample is said to be cryptochiral (Mislow and Bickart,
`1976/1977).
`
`The term cryptochiral is not to be confused with the analogous expression
`"stereochemically cryptic" that refers to a stereoselective chemical reaction whose
`stereochemical outcome is hidden (Hanson and Rose, 1975; cf. also Section 8-5).
`
`Notable examples of enantiomerically enriched compounds that are crypto(cid:173)
`chiral as a consequence of inherently low optical rotation magnitude are shown in
`Figure 13. 71. The cryptochirality condition may conceivably be lifted by measur(cid:173)
`ing a different chiroptical property, for example, vibrational circular dichroism
`(VCD) (Section 13-6).
`The second type of cryptochirality arises when the measurement of rotation
`accidentally takes place in the vicinity of a change in sign (see below and above).
`For an example involving a change in concentration of dimethyl a-methylsucci(cid:173)
`nate, CH30 2CCH(CH3)CH2C0 2CH3 , see Berner and Leonardsen (1939). At a
`certain concentration, the measured rotation is necessarily zero (crossover point)
`and the sample is then accidentally cryptochiral. Note that a distinction between
`stochastic achirality (cf. Section 13-2.a) and cryptochirality cannot be made unless
`
`38
`
`CH3 H
`I
`I
`H3c-c-c-oH
`I
`I
`CH3 D
`
`40
`
`C2Ha
`
`I
`n-CeH13-C-n-C4He
`I
`
`n-C3H7
`
`41
`
`Sanderson and
`Mosher (1966)
`
`Wynberg, et al.
`(1965)
`
`Figure 13.71. Compounds illustrating cryptochirality.
`
`

`
`1076
`
`Chiroptical Properties
`
`the former be lifted by a change in measuring device or the latter by a change in
`conditions, the latter being easier to achieve. Other examples of the second type
`of cryptochirality have been given above and in Section 13-4.
`The dependence of optical rotation on the wavelength of the light, ORD, has
`been discussed in Sections 13-2.b and 13-4.
`
`Effect of Temperature
`
`The effect of temperature on chiroptical properties may be ascribed to the
`following phenomena (Legrand and Rougier, 1977): (a) changes in density of the
`solute and/or the solvent that alter the number of molecules being observed; (b)
`changes in the population of vibrational and rotational energy levels of the chiral
`solute; (c) displacement of solute-solvent equilibria; (d) displacement of con(cid:173)
`formational equilibria; and (e) aggregation and microcrystallization of the chiral
`solute (cf. enantiomer discrimination, Section 6-2).
`In general, [a] changes 1-2% per degree Celsius, but larger changes (up to
`10% per degree Celsius) are not unknown, for example, [aJn of aspartic acid,
`H02CCH(NH2)CH2C02H, in water (c=0.5%) is 4.4 at 20°C, 0 at 75°C, and
`-1.86 at 90°C. The change in sign at 75°C (temperature of cryptochirality, see
`above) is noteworthy (Greenstein and Winitz, 1961, p. 78).
`An early example of an increase in specific rotation with increasing tempera(cid:173)
`ture that was ascribed to a shift in a conformational equilibrium is that of
`2-butanol (Horsman and Emeis, 1965). Other examples are discussed in Section
`13-4.e).
`Strong dependence of the optical rotation on temperature may be found even
`among hydrocarbons, for example, 3-phenyl-1-butene whose neat rotation,
`[a]~ -5.91, for the enantiomerically pureR enantiomer increases linearly 0.18°/
`oc from 16 to 29°C. Here, it is likely that the temperature exerts a strong
`conformational bias (Cross and Kellogg, 1987).
`
`Effect of Solvent
`
`The "nonspecific" influence of solvent on the specific rotation may be corrected
`by calculation of a quantity called the specific rotivity fi' that includes the
`refractive index of the solvent ns (see Heller and Curme, 1972):
`
`fi' = [3a]l(n; + 2)
`
`(13.30)
`
`Several examples of dramatic changes in the angle of rotation as a function of
`solvent have been given above. Many instances of changes in sign of [a] have also
`been recorded for amphoteric substances, such as the amino acids, as the pH is
`changed (Greenstein and Winitz, 1961, p. 1727). An exceptional example of the
`effect of solvent on [a] is given in Figure 13. 72. Given examples such as these, it
`is disconcerting how frequently the mention of the solvent is omitted from
`experimental descriptions of the optical rotation.
`Care in choosing the solvent to be used in the measurement of [a] is
`necessary in view of the several specific types of interaction that are possible
`between solute and solvent. In general, one recognizes the intervention of .·
`hydrogen bonds when oxygen-containing solutes, such as carboxylic acids, aide-
`
`

`
`1077
`
`Forma midi
`
`~
`lNJ
`bH3
`
`Ethylene bromide
`
`-6
`-8-
`
`-10
`-12
`-14
`
`-16
`
`-18
`- 20100
`
`eo
`
`60
`
`40
`p
`Figure 13.72. Specific rotation of nicotine in various solvents ( p =concentration of solute in grams
`· · per 100 grams of solution). At p = 100, the "bulk" rotation [a] 100 should be a constant, as observed,
`· · and at p = 0, [a] should tend to the intrinsic rotation {a} (p. 1079). [Adapted with permission from
`• Winther, C. (1907), Z. Phys. Chern., 60, 621.]
`
`and ketones, and alcohols, are dissolved in hydroxylic solvents; in some
`'cases reactions, such as hemiacetal formation, may occur. In addition, dipole(cid:173)
`. dipole interactions and changes in conformer populations are important sources of
`·.solvent-induced variations in rotation magnitude (Lyle and Lyle, 1983).
`The effect of intermolecular solute association of polar solutes on [a] in
`· nonpolar solvents has already been.,pointed out (see above). Solute-solute
`' association effects may be leveled out or suppressed in polar solvents by competi~
`·. · tion with (concentration-independent) solute-solvent association. Polar solvents,
`, such as ethanol, may break up solute-solute association leading to a smaller
`concentration dependence of [a], as is found with nicotine (Fig. 13.72). Such
`· , findings illustrate the desirability of using methanol or ethanol as a solvent in
`polarimetry.
`In some instances, hydrogen bonding is known to be responsible for changes
`•. in [a] with concentration and/or solvent. Compounds 42 and 43 (Fig. 13.73)
`. exhibit a remarkable solvent dependence of the sign of [a] 0 for the RRISS (syn)
`, 42 diastereomer that is not found in the case of the RS I SR (anti) diastereomer 43 .
`. The sign of [a]~ of (4R,5R)-42 is(+) in methanol and(-) in chloroform. This
`
`

`
`1078
`
`Chiroptical Properties·
`
`OR
`HsC&~~1oH21
`
`OCH2C&Hs
`
`Rx;
`
`OH
`
`43 (R= H) S,R shown
`
`44
`
`42 (R = H) R,R shown
`42a (R=Ac)
`
`Figure 13.73. Structures 42-44.
`
`difference has been ascribed to a conformational change: the predominant
`methanol-solvated (OH/OCH2C6H 5 ) anti conformer gives way to a (OH/
`OCH2C6H 5 ) gauche intramolecularly hydrogen-bonded conformation in chloro~
`form. Such sign reversal is not seen in the benzyl ether-acetate derivative 42a, in
`the corresponding diol (the latter appears to prefer the gauche conformation ·
`regardless of solvent) or in the diol acetonide. Reversal of the sign of [a] 0 would.
`seem to be precluded in the predominant zigzag (all-anti) conformation of the
`molecular skeleton. A similar sign reversal was observed in a series of 2-alkoXy.
`alcohols 44 (Fig. 13.73) presumably for the reasons advanced above. The free dibl.
`(S)-1,2-dodecanediol exhibits [a]~ -10.1 (EtOH) but +0.9 (CHC13 ). This
`suggests that here, too, the intramolecularly hydrogen-bonded conformer prevails c
`in CHC13 (Ko and Eliel, 1986).
`Another example of optical rotation sign reversal is found with compound 45
`(Fig. 13.74): [a] 0 +14.0 ± 0.6 (CHC13 ) and -2.7 ± 0.6 (CH30H) (Suga et al.;
`1985). The molar rotation [<I>] 0 was found to be independent of concentration .irt
`12 solvents of varying polarities. The principal factor responsible for the s,ign·
`reversal was intramolecular hydrogen bonding between the carbonyl and hydroxyl· :
`groups in nonpolar solvents (confirmed by IR and 13C NMR measurements)and ·.
`its absence in the presence of strongly solvating media (alcohols, CH3CN,··'or
`acetone) as a result of competing intermolecular hydrogen bonding between·t~e ..
`solvent and the solute (Suga et al., 1985). A sign reversal of aCE was also noted·
`in the ORD of 46 (Fig. 13.74) in CHC13 and CH30H but this was not reflected in' .
`the sign of [a] 0 . The latter fact points up once again the desirability of carrying···
`out studies of chiroptical properties over a range of wavelengths and preferably.·
`into regions that reveal the responsible CEs.
`,.,. : ·. ·
`A particularly clear-cut example of a conformational equilibrium that ·is.·
`responsible for changes in chiroptical properties over the range of 210-350nm.is
`shown by ketone ( + )-47 (Fig. 13.75). As the solvent is changed from cyclohexane.· ,.
`to acetonitrile or methanol, the effect of increasing solvent polarity on. the. • ·
`dipole-dipole repulsion (as well as solvation effects) between the adjac~nt .
`permanent dipoles (C=O and C-Br) causes a conformational change: the broniine , ·
`changes from axial to equatorial and significant changes in the CD ·.
`· ·
`
`45
`Figure 13.74. Structures 45-46.
`
`46
`
`

`
`Applications of Optical Activity
`
`1079
`
`s=b-··
`
`H
`
`0
`
`47
`48
`Figure 13.75. Structures 47-48.
`
`(Kuriyama et al., 1967). Striking changes in [a] 0 of propylene oxide, including
`sign reversal, are observed as the solvent is changed from benzene to water. In
`this instance, we are cautioned against ascribing the effect of solvent directly to
`conformational changes (Kumata et al., 1970).
`
`Effect of Concentration
`
`Equation 13.28 suggests that the specific rotation should be independent of
`concentration. It is not hard to find evidence that this constancy holds only over
`very narrow concentration ranges and, in some solvents, not at all (the example
`of nicotine is found in Fig. 13.72; other examples may be found in the book by
`Lowry, 1964, Chapter VII).
`As early as 1838, Biot suggested that the specific rotation followed a linear
`relationship, such as that of Eq. 13.31, where a and b are constants
`
`[a]= a+ be
`
`(13.31)
`
`and cis the concentration (Lowry, 1964, p. 90). Constant a has been equated with
`a new quantity called the "intrinsic rotation" {a}, a true constant corresponding
`to the specific rotation in a given solvent at infinite dilution; [a Jc__,0 = {a} (Heller
`and Curme, 1972, p. 163). Obviously, {a} can only be calculated since ex(cid:173)
`perimentally, as the concentration is reduced the rotation must vanish.
`The intrinsic rotation is the specific rotation for a system free of solute-solute
`interactions. However, solute-solvent interactions are maximized in {a}, which
`can differ greatly from solvent to solvent, for example, for nicotine (Fig. 13.72).
`Since, obviously a = oo at 0% solute, the values at very low concentrations must
`be extrapolated. Conversely, as the concentration of solute increases, solute(cid:173)
`solute interactions become dominant and the effect of the solvent eventually
`vanishes: [a Jc__, 100 = [a lneat tends to a constant value that is identical for all
`solvents.
`rotation of (S)-2-phenylpropanal,
`the specific
`A
`recent report on
`CH3CH(C6H 5)CH=O, makes it clear that even at relatively low concentrations in
`benzene (c = 1-4), changes of the order of 1-2% are found in [a] 0 as the
`concentration is doubled (see Table 13.3; Consiglio et al.; 1983). The accurate
`determination of optical purities is thus seen to be dependent on the careful
`measurement of rotations as well as on comparison of the resulting [a] values
`with reference [a] values measured in the same solvent, at the same temperature,
`and at the same concentration (Consiglio et al., 1983). The reader is also
`
`

`
`1080
`
`Chiroptical Properties
`
`Influence of Dilution on [a]~ for {S)-2-Phenylpropanal•.b
`TABLE 13.3.
`Concentration (g/100 mL- 1
`
`)"
`
`[a]~t
`
`[a]~
`166.6
`161.8
`Neat
`182.2
`177.9
`46.43
`195.4
`190.5
`18.57
`201.9
`196.6
`9.29
`207.9
`202.7
`7.43
`205.8
`211.3
`3.72
`214.7
`209.1
`1.49
`• Reprinted with permission from Consiglio, G., Pino, P., Flowers, L. 1., and
`Pittman, C. U., Jr. (1983), J. Chern. Soc. Chern. Cornrnun., 612. Copyright©
`Royal Society of Chemistry, Science Park, Milton Road, Cambridge CB4 4WF,
`UK.
`b Optical purity 68%.
`c Benzene solution.
`
`reminded that [a ]0
`reflects, but is not necessarily linearly related to, the . .
`enantiomeric composition (Horeau effect: Horeau and Guette, 1974; see Section
`· ..
`6-S.c).
`Some very subtle effects are manifested by the rotatory power. For example·:.: ·
`[a ] 436 of menthol and of menthol-0-d diverge as the solute concentration i~ .
`increased 100-fold (in cyclohexane). This differential concentration effect reveals
`that intermolecular hydrogen bonding is subject to a thermodynamic isotope :
`'
`effect (Kolbe and Kolbe, 1982).
`In the absence of experimental information, calculations, or obvious structui:~ ,
`al features, the detailed aspects of chiroptical data including sign reversal of CBs .
`cannot be explained. A detailed study of the chiroptical properties of alkyL:·
`substituted succinic anhydrides 48 (R' = H, R =alkyl; Fig. 13.75) revealed a great'..
`sensitivity of the CD (CEs at ca. 220 and 240 nm) to solvent polarity; to'
`temperature, and to the size of the alkyl substituent. The results could generally··
`be accommodated by a sector rule (for anhydrides with local C2v symmetry of th.e- • ·
`chromophore, that is, a planar anhydride group) (Gross, Snatzke, and Wessling;:· .•
`1979), but the absence of information about conformational equilibria' and . ·
`solvation effects made it impossible to offer explanations for the detailed features:
`of the CD (Sjoberg and Obenius, 1982).
`··.r
`
`b. Empirical Rules and Correlations. Calculation of Optical Rotation
`
`Ever since simple curiosity about optical activity gave way to its ap]pU<;attoR; ,
`efforts have been made to calculate the magnitude and sign of the optical ro!a#qn.
`in relation to structure and configuration.
`.
`One of the very oldest correlations between structure and rotatory pow~r
`that of Walden who observed that the molar rotations of diastereomeric salt~ ·
`dilute solution are additive properties of the constituent ions (Walden, . ·
`Jacques et al., 1981, p. 317). Arithmetic manipulation of the molar rotattcms.o!'.:.
`diastereomeric salts, such as those obtained in a resolution, may thus permit
`to estimate the enantiomeric purity achieved during a resolution mediated,
`these salts.
`
`

`
`The molar rotations [<1>] 0 of two of the nonenantiomeric diastereomeric salts of
`a-methylbenzylamine mandelate are -182.3 and +169.7 (in H 20). From these
`data, the molar rotations of the constituent ions are calculated to be [<1>] 0 = ~
`([-182.3] + [+169.7]) = -6.3, that is, ±6.3 and [<1>] 0 = ~ ([-182.3]- [+169.7]) =
`-176, that is, ±176. The latter value is assigned to the mandelate ion [literature
`values of [<1>] 0 for sodium and potassium mandelates are 182 and 178, respectively
`(both in H 20; Ross, et al., 1937)]. Combination of this value with that for
`ephedrinium hydrochloride, [<1>] 0 69.8 (H 20) (Overby and Ingersoll, 1960) yields
`the molar rotations of the two ephedrine mandelate diastereomers, [<1>] 0 = 176 +
`69.8 =246 and 176-69.8 = 106. The experimental values of these rotations ob(cid:173)
`tained by Zingg, Arnett, et al. (1988) are ±250 and ±107, respectively (both in
`H 20).
`
`Analogously, molar rotations of inclusion compounds would be expected to
`additive properties of the host and guest molar rotations. In both cases,
`Ji,U~~·····vity of rotations would not necessarily obtain when strong intermolecular
`takes place.
`In the case of covalent compounds, early correlation attempts also made use
`of the concept that the rotations of compounds containing several chiral centers
`be calculated by adding rotation contributions from each of these centers.
`concept, incorporated in van't Hoff's empirical "Principle of Optical Super(cid:173)
`!i?PIOSJLtlcm." that individual chiral centers in a chiral compound make independent
`... vu•u .. •uu•vu" to the molar rotation (van't Hoff, 1893; Kuhn, 1933, pp. 394, 423) is
`successfully being applied in very limited contexts.
`The relative configurations of the diastereomeric (R)-0-methylmandelate
`esters of 49 were assigned by application of the van't Hoff principle. Contribu(cid:173)
`. tions to the specific rotations from the octalin portions of the ester molecules were
`· estimated from the rotations of ( + )-dihydromevinolin 50 and lactone 51 (Fig.
`13.77) to be approximately 100 (148.6- 48.8) while that of the (R)-mandelate
`was independently known to be strong and positive [(S)-(-)-methyl
`uuc:.uu•v•ate] has [a]~ -124 (Barth, Mosher, Djerassi, et al., 1970).
`, the configurational assignments were ±75 f