K . MISLOW,E. BUNNENBERG, R . RECORDS, K. WELLMAN AND C. DJERASSI
1342
[JOINTCOKTRIBUTION FROM
THE
DEPARTMENT OF CHEMISTRY OF XEW YORKLSIVERSITY, NEW YORK53, K. Y , A N D UNIVERSITY, STASFORD, CALIF.]
Inherently Dissymmetric Chromophores and Circular Dichroism. B Y KURTMI SLOW,^^ E. BUNNENBERG, RUTHRECORDS, KEITHwELLMAN3b
AND
1701. OF
85
STANFORD
II1s
CARL DJERASSI
RECEIVEDKOVEMBER 28, 1962
In a continued investigation of the ultraviolet circular dichroism ( C . D . ) curves of optically active substances possessing inherently dissymmetric chromophores, the previously discussed1s2,8relationship between ultraviolet absorption spectra, optical rotatory dispersion (O.R.D.) and C.D. curves has now been extended to various representatives of bridged and unbridged biphenyls and binaphthyls I n addition to confirming the earlier O.R.D. results, this work serves to accent the relative advantages of C.D. and O.R.D. measurements in a class of complex chromophores.
In a preliminary survey, we reported that circular dichroism (C.D.) curves of twisted biphenyls are extremely useful in the identification of the electronic transitions which are responsible for the individual Cotton effects and which give rise to the associated optical rotatory dispersion (O.R.D.) curves. The intensities of the C.D. curves were seen to parallel both the high extinctions in the absorption spectra
The marked reduction of overlapping tails in the C.D., as compared to the O.R.D. curves, was found to be especially significant in that i t permitted identification of weakly optically active transitions whose Cotton effects had previously been unsuspected because they had been completely masked by the intense tailings of high amplitude Cotton effects centered a t shorter wave lengths. As a result of these preliminary ob-
TABLE I
R R DATAO N BIARYLSI-XIV
OF
FORMULA TYPE
R' R' N0
R
R'
Solvent
I
XO2
CH3
Dioxane
I1
SO2
CH2Br
Dioxane
I11
s0 2
COOH
Dioxane
IY
so2
COOCH:.
Dioxane
1.
KO2
CH2OH
Dioxane
VI
CH:!
Bt
Isooctane
1.11
CHa
CI
Isooctane
\.I11
CHa
I
Isooctane
IX X
CH3 CH;
COliH2 COOCHB
Dioxane Dioxane
XI XI1
c1 c1
XI11
SH?
COOH CONH9 CHI
Dioxane Dioxane Dioxane
x11.
SH3
CHa
0.1 N HCI
'
Circular dichroism d a t a
-2000, -2300, [el32o-1300, !elzQ8-9200, -6700, [e]2ij -3200, +i6,ooo, [eiZa9$8700 ( 6 0.50-0.01) +MO, [81360 +5i60, [elara + 6 a o , $2280, +8500, [O]z9o $9500, +i3,300, [eiZi60.00, [e1za6-30,400 ( c 0.7-0.02s) [ e l 4 3 0 0.00, [e13io f 5 2 0 0 , [elaao $10,700, [el31o+5700, i-4700, 0.00 ( C 0.9-0.018) [Olazo 0.00, [elaio -4900, [el33o-10,200, [el31o-7700, [elzQo-5500, [elnio0 . 0 ~ ( c 1.0-0.1) [el410 0.00, [el316-5300, [8]3zo -2350, [e1300-zQo -8820, [elzi6 -4420 ( C 0.320.013) [ e b 8 - - 2 8 8 -425, [ B l z i ~+2000, [6]zij +1400, [6]2il +2420, [6]za-z45 1800, [ e l 2 3 6 -5600, [elm -4200 ( C 0.2-0.02) ~ e 0.00,~ [ B l z i ~+2200,~ ~ +i500, ~ ,2800, ~ [81266-263 ~ +2000, 0.00, -7700, [elzi7 -3100 ( C 0.2-0.04) [ e l 3 1 3 0.00, [8]?s8 -900, [QIZSI -4300, [ 0 ] 2 i ~-3000, [8]?iz -4000, [@I266 -2500, PI^,, -6000, +38,ooo, [8]236--234 +32,500 ( c 1.0-0.02) [8]300-290 0.00, [@I280 +1500, [ O ] Z L +31,000, [8]250+33,000 ( C 0.04-0.01) [~1320-310 - 1400, [elzQ5-2400, [e]290-2800.00, [el2&@ -7000, [el2520.00 ( c 0.0540.014) [e1320 0.00, [el3O3-:io, [ 8 ] 3 0 0 - 2 9 @ 0.00, [8]zi6 +2000, [e]251 -15,400 ( C 0.2-0.02) [eIaoo-zso0.00, [ e b -32,000, [8]243 -228,000, (C 0.04-0.005) [@I330 -1200, [~IB -4820, x [8]305 -12,000, $19,300, [0]zi6 7220, [ e l 2 6 0 +5800, [ e 1 2 5 5 +14,400 ( C 0.044-0.011) S o deflections were obsd. from 350 to 245 mp, The init. concn. mas diluted 1 : 5 Subsequent 1:2 dilut. were made as required. 181367
and the high amplitudes in the O.R.D. curves: all three phenomena are characteristic of molecules which contain inherently dissymmetric chromophores. The shapes of the C.D. curves are in excellent accord with those of the absorption curves, a correspondence which is predicted by theory. The signs of the C.D. curves were found to correspond to the signs of the related O.R.D. curves, which had previously been shown to reflect the chirality of the c h r ~ m o p h o r e . ~ ( 1 ) P a r t I, E. Bunnenberg, C. Djerassi, K . Mislow a n d A. Moscowitz, J . A m . Chem. SOL.,84, 2823, 5003 (1962). (2) T h e present paper represents part L X X X of the series on "Optical
Rotatory Dispersion Studies" from t h e Stanford Laboratories; f o r part L X X I X , see C. Djerassi, H. Wolf and E. Bunnenberg, J . A m Chem. SOL., 86, 324 (1963). T h e work a t Stanford has been supported by t h e National Science Foundation (Grant S o . G-19905). (3) (a) Fellow of the Alfred P. Sloan Foundation. (b) National Institutes of Health Postdoctoral Research Fellow, 1962-1963.
-
servations it became immediately obvious that the presence of weakly optically active transitions might also have gone undetected in our previous O.R.D. , ~ an investigation of study of other related b i a r y l ~and the C.D. curves of these substances seemed to be indicated on these grounds alone. In addition, the subject of dissymmetric chromophores is of sufficient theoretical interest so that a broadening of the experimental basis of our earlier conclusions' appears to be more than justified. Discussion of Results Chromophoric Groups.-Our findings are reported in Tables I-IV and Fig. 1-15. As in the preceding (4) K . Mislow, A m . N . Y . Acad. SC;., 93, 457 (1962), and references cited therein. ( 5 ) K . Mislow, h f . A. W, Glass, R. E, O'Brien, P. Rutkin, D . H. Steinberg, J . Weiss and C. Djerassi, J . A m . C h e m . Soc., 84, 1455 (1962).
DISSYMMETRIC CHROMOPHORES AND CIRCULAR DICHROISM
May 5 ) 1963
1343
TABLE 11
DATAON BIARYLSXV-XXV
OF
FORMULA TYPE
( & CH? ?HZ B
h'o.
R
Solvent
Circular dichroism d a t a
XV
NO2
Dioxane
[elaoo0.00 [el338$44,900, [elzs7-55,200, [e]2is-25s0.00, [el253-30,900
XVI
so2
Dioxane
XVII
NO2
Water
XT'III
so2
Dioxane
XIX
KHz
Dioxane
XX
NH3 I.
0.1 N HC1
XXI
CH3
Dioxane
XXII XXIII
CH3
c1
Isooctane Isooctane
XXIV
c1
Water
XXP
CH3
Isooctane
e +IOOO, ~ ~ [elaaa ~ ~ +48,ooo, ~ e + ~ O ~O O , [ ~ e -38,000, ~~ ~ ~ [elzao ~ -10,000 ( C 0.1-0.05) [elazo0.00, [e1368 -19,000, [e]327-56,000, [8]300 - 15,000, -9000, [ O ] Z ~+82,000, ~ [e]22s- 16,500 ( C 0.13-0.0054) [0I4oz 0.00, [ e h -15,000, -57,000, [e]3oo -8000, [elzQo -16,000, [ 8 ] m +107,000 ( C 0.05-0.01) 0.00, [elaao-2380, - 11,800, [elaoz+i9,800, -11,900, i10,ooo ( c 0.057-0.011) No deflections were obsd. from 400 to 245 mp. The init. concn., 1.15 g./1., was diluted 1: 5 . Subsequent 1:2 diln. were made as required. [ e l 2 9 4 0.00, [ e l 2 8 4 $2700, [8]279-276 -1800, [ e l 2 7 1 -4600, [9]268 0.00, +i28,000 (c 0.1-0.005) [Bl3o0 0.00, [@I286 -8500, [elsn $4100, [elgdz- 11,800 (C 0.2-0.002) [el3o0 0.00, [el2g1+10,500, [el275- 17,200, [elza5+%,loo, [elza3 +70,600, [ e l 2 2 4 +121,000 ( C 0.25-0.005) [81300 0.00, [elzg3+3680, -9200, [elzas+90,20o, [ O l z r ~+ ~ I , O O O , (01226 $184,000, [ @ I 2 2 2 +153,000 (C 0.0b0.005) +56,300, [e]2aa0.00, [Olzas -48,300 (C 0.0016) [eizs00.00,
( c 0.05-0.005)
[
TABLE I11
R R R
KO.
XXT'I
Config.
Solvent
Circular dichroism d a t a
(R)
Dioxane
[el313 +4500, [ e l 2 8 4 +41,300, [el26s+9000, [ 8 I z 3 ~ -38,000, [81220 -241,000 0.02-0.004) [ 8 I 3 ~ o0.00, [ d ] z s 5 - 15,700, [e]26~ +23,500, [0]250 +33,000, [e]z4s +78,500 (c 0.0044) [elsa20.00, [el32z-11,000, [elal3 -8500, [el295-35,500, [elzgo-45,600, [elzs5 -53,300, 0.00 ( C 0.04-0.004) [8Iasa f6000, [elsz6- 11,800, [8]305 5000, [ O l Z s ~ -70,700, [e1245 +524,000, [ ~ I z ~+460,000 o . . ( c 0.1-0.004)~ [8lal00.00, [e]2s, -42,400, [0Izi2 -30,000, [B]260 -22,500, [ B ] 2 s s 0.00 (c 0.0074-0.0019) [e]35o0.00, [ e l 3 0 6 -5750, [e]28g -46,000, [8]ZSQ 0.00 (c 0.47-0.0076)
CHOH
[BI:isz +4000, (C
XXT'II XXVIII
CH3 COOH
(5')
Dioxane Dioxane
XXIX
COOCHI
(S)
Dioxane
XXX XXXI
COKH2 CHZBr
(S)
Dioxane Dioxane
(S)
(S)
-
TABLE IV
DATAO N BIARYLS XXXII-XXXV
OF
@
FORMULA TYPE
CHz CHz
\R/
No.
K
Config.
Solvent
XXXII
co
(R)
Isooctane
XXXIII
CHOH
(R)
Dioxane
XXXIT'
C(C02GH:)?
(S)
Dioxane
NXX\'
0
(S)
1)ioxane
Circular dichroism d a t a
[81333 +lO,OOO, [elaz7+95,000, +60,000, +100,000, [el315 +95,000, [e1300 $-155,000, [8]z~a +105,000, [8]287 +85,000, [ e l 2 6 5 -125,000, -45,000, -42,000, - 1,210,000, [ o -640,ooo ~ ~ ~(c 0.022-o.ooo9) ~ 101340 0.00, [01330 +7000, [8Iso5 +54,000, [ B ] ~0.00, ~ ~ [e1267 -128,000, [el25z0.00 (C 0.006-0.0012) 18Iato0.00, [el326 -41,000, [elao6-69,000, [e]286-35,000, [e]265 +103,000, [ ~ I ~ O O +96,000 (C 0.016-0.079) [el8,(, +3300, [e1364 + M O O , [elaos - 51,000, je]268 +144,000, [8]zja-z51 $133,000, - 145,000, [eiZzs0.00 ( C 0.02-0.0004)
1elato 0.00,
O.R.D. study,j chromophorically related substances were found to exhibit very similar rotatory properties which in the present case are reflected in the very similar C.D. curves for each such group. This behavior is expected as a matter of course from the similarity of the relevant electronic transitions.
For example, the C.D. curves of the unbridged 2,2'dinitrobiphenyls display common features which are in an obvious way related to their absorption characteristics. Figure 1 shows C.D. and absorption curves of one such compound, (S)-6,6'-dinitro-2, 2'-dimethylbiphenyl (I). The short-wave length posi-
~
~
1344
K . MISLOW,E. BUNNENBERG, R.RECORDS,K . WELLMAN AND C. DJERASSI
VOl. 85
10
t
20
9
I \
0
5l I '\ \ \
15
x
2
0 x
q
6
W
\
'0
-5L
4
5
I
240 9
3
1 : : 250
220
300A(rn,u)
350
L
2
0
Fig. 1 Fig. 3.
2 0
300 X(rnA). 350
400
Fig. 2.
tive C.D. maximum6 a t 251 mp corresponds to the absorption maximum a t 260 mp, and the negative C . D . maximum6 a t 298 mp corresponds to the inflection a t ca. 310 mp. The long-wave length negative C.D. maximum near 350 mp is associated with a transition which is not apparent in the ultraviolet spectrum. These signed features are characteristic of the (S)configuration and are shared by the chromophorically related IT-V which have in common the signed (nega(6) T h e maxima of circular dichroism curves are signed, and it becomes necessary t o speak of positive and negative maxima. T h e close relationship of circular dichroism maxima t o necessarily unsigned transitions in t h e absorption spectra makes i t misleading t o speak of negative maxima as "minima."
-I0
t I
.
240 '
:
.
.
:
:
:
300
h(rny). 350 '
400'
I
Fig. 4.
tive for the (S)-configuration and positive for the (R)-configuration) long-wave length C.D. maxima near 300 and 340 mp (Fig. 2). I t is instructive to recall the previously discussed O.R.D. results in the same series.; I t had then been found that the 2,2'-dinitrobiphenyls exhibit two Cotton effects of opposite sign centered near 260 and 330 mp, respectively, and i t had been concluded that the positive 330 mp Cotton effect corresponded to the (R)-configuration. The present C.D. results completely support the earlier conclusions. I t is particularly noteworthy (a) that the masked nitrobenzene
May 5 , 19G3
DISSYMMETRIC CHROMOPHORES AND CIRCULAR DICHROISM
Fig. 7.
Fig. 5 .
1
1345
-6 -7
I
.
240 250
:
:
:
.
h(m,u).360
.
:
:
.
:
:
350
I
230
:
:
:
'
:
:
250 h(m,u).
'
' 300
"
:
30
Fig. 6.
Fig. 8.
transition near 340 mp has now been clearly revealed
tive C.D. maximum a t 267 mp corresponds to one of the absorption maxima below 280 mp, and one of the negative C.D. maxima near 290 and 330 mp corresponds to the inflection near 310 mp. These features are characteristic of the (R)-configuration and are shared by the chromophorically related compounds XVI and XVII (Fig. 4). Our earlier conclusions6 are therefore supported : for the bridged 2,2'-dinitrobiphenyls, two oppositely signed Cotton effects are centered near 260 and 330 mp, and a negative 330 mp Cotton effect signalizes the (R)-configuration. In agreement with the earlier conclusions, the intensity
both by C.D. and by O.R.D., and (b) that the transition near 300 mp has been shown to be optically active by C.D. measurement, a fact which had eluded us in the earlier experiments where the tails of neighboring Cotton effects had obscured the small Cotton effect associated with this transition. In much the same way, the C.D. curves of bridged 2,2'-dinitrobiphenyls, while confirmatory of earlier O.R.D. r e s ~ l t s ,disclose ~ heretofore unsuspected features. Figure 3 shows C.D. and absorption curves of the (R)-diester XVIII. The short-wave length posi-
1346
K. MISLOW,E. BUNNENBERG, R. RECORDS, K. WELLMAN AND C. DJERASSI
I
Vol. 85
1
I
I I
XXII R = C:OH H
I
Fig. 9. Fig. 11
ox IX
RCK, X="2
X
R=CY X=OCH, XI RCI, X= OH R=CI, X=NH,
I
I
\
-20 220
\
250 X(rn&u).
0 3 0
Fig. 10.
of both Cotton effects is significantly enhanced by bridging but there appears to be no marked red-shift. It is remarkable how sharply the weak transition near 290 mp, hidden in both absorption and O.R.D. curves, has now been brought out in the C.D. curves. In the 1,1'-binaphthyls, the correspondence between earlier O.R.D. data5 and the present C.D. results is also satisfactory. The unbridged binaphthyls , ~ have absorption maxima a t 285 and ca. 230 m ~ and, as exemplified by compound XXVI (Fig. 5), the C.D. curves are signed reflections of these spectral characteristics. A positive 285 mp Cotton effect corresponds to the (R)-configuration in this and related compounds
io
XXVII-XXXI (Fig. 6), as had been concluded earlier on the basis of O.R.D. data alone. The 2,2'-bridged 1,l'-binaphthyls XXXII-XXXV have in common oppositely signed C.D. maxima a t ca. 270 and 310 mp (Fig. 7 and 8). These correspond to the oppositely signed O.R.D. curves centered near 265 mp and near 300 mp, respectively, as had been observed b e f ~ r e . ~ Furthermore, in harmony with the earlier conclusion,6 a positive 306 mp Cotton effect corresponds to the ( R ) -
May 5 , 1963
DISSYMMETRIC CHROMOPHORES AND CIRCULAR DICHROISM
1347
.u
32
220
250
h(rnrJ).
300
Fig. 13.
configuration. In the case of compound XXXV, it was even possible to ascertain that the 230 mp transition is also strongly optically active. Although the C.D. curve of the binaphthyl ketone X X X I I bears a general resemblance to the curves of the other 2,2’-bridged 1,l’-binaphthyls, the greater complexity in the region above 320 mp, as well as the greater intensity of the long-wave length maximum, reveal the operation of a second effect. We believe that the intensified carbonyl n -+ r*-transition which characterizes5 P,y-benzo-ketones having the geometry of X X X I I may be held to account for these departures from the norm. The bridged 2,2’-dimethyl- and 2,2‘-dihalobiphenyls XXI-XXII and XXIII-XXIV, respectively, show a C.D. pattern (Fig. 9 and 10) which is characteristic of this group of compounds and which, as expected, resembles the C.D. behavior of the previously discussed bridged dimethylbiphenyls. In the wave length region between 270 and 300 mp, two relatively lowintensity, oppositely signed C.D. maxima are displayed by all four compounds. A positive maximum a t longer wave lengths signalizes the (R)-configuration ; in this respect the present C.D. data merely serve to confirm the conclusions drawn from the earlier O.R.D. data.5 However, only through C.D. has the presence of the second of the two long-wave length maxima (negative for the (R)-configuration) been detected : as in the case of the related bridged 2,2’-dimethylbiphenyls discussed in part I,’ the O.R.D. tailings of the conjugation band Cotton effect had in effect swallowed the small Cotton effect associated with this transition. The conjugation band Cotton effect in the biphenyls has been demonstrated by C.D.’ as well as by 0 . R . D . 5 measurements. Although we were not able to extend measurements below 260 mp in the case of XXI-XXII, the conjugation band Cotton effect is clearly displayed in the 2,2’-dichlorobiphenyls XXIII-XXIV (Fig. 10) in the form of two adjacent high-intensity C.D. maxima
240 250
‘
3oo h(mA).
350
400
Fig. 14.
’
Fig. 15.
centered a t 240 mp which are positive for the (R)configuration. This is precisely what might have been predicted on the basis of the O.R.D. b e h a v i ~ r . ~ Conformation of Biphenyls and C.D.-It has previously been established5 that the high amplitude Cotton effect associated with the biphenyl conjugation band suffers a blue-shift as the bridge length is increased, in parallel with the blue-shift of the absorption
1348
K. ~ I I S L OE. W BUNNENBERG, , R . RECORDS, K. XELLMAN AND C. DJERASSI
Vol. 85
band itself which is known to accompany an increase In certain ways the C.D. information has merely been in the angle of torsion around the pivotal l,l’-bond. of a confirmatory nature. First, all of the conclusions I n the present work these conclusions have been conwhich had previously been drawn, on the basis of firmed in a study of (R)-9,10-dihydro-4,5-dimethyl-O.R.D. information alone, concerning the relation bephenanthrene (XXV): in contrast to the seventween absolute configuration of various chromorphores membered ring 2,2’-bridged biphenyls discussed in and sign of the relevant Cotton effects, have been conthis and in the preceding paper.’ the conjugation band firmed without exception by the present and by the Cotton effect has its C.D. maximum a t 262 mp, in preceding’ C.D. studies. Second, the presence, sign excellent correspondence with the absorption maxiand intensity of the various Cotton effects discussed mum a t 261 mp (Fig. 11). I n the previous O.R.D. in the O.R.D. study5 have been verified in this and the study of this substance’ only the long-wave length preceding‘ investigation by the observation of C.D. low-intensity features had been instrumentally accurves (a) which are centered near the point of infleccessible. Nevertheless, it had then been predicted that tion of the corresponding O.R.D. curves, (b) whose (R)-XXV would display a high-intensity positive sign is the same as that of the corresponding O.R.D. Cotton effect centered near 261 m ~ The . ~ present curves, and (c) whose intensity (expressed in terms of finding of [ e l m a x $56,300 (isooctane) at 262 mk is in ellipticity) reflects qualitatively the amplitude of the complete accord with this prediction, particularly corresponding O.R.D. curves. when i t is noted that the sample is not likely to be On occasion, C.D. measurements have been of distinct optically pure‘ and that the cited ellipticity therefore measurements. Thus, while all advantage over O.R.D. represents a minimum value. major Cotton effects had been detected by 0 . R . D . , 5 Interpretation of the C.D. curves of the unbridged the presence of some low-intensity Cotton effects had biphenyls offers, in general, difficulties similar to the previously escaped scrutiny. The corresponding lowones encountered in the preceding O.R.D. study5 of amplitude O.R.D. curves had either been entirely these compounds. For example, the diphenic acid hidden by the overlapping tails of Cotton effects, derivatives IX-XI1 have relatively featureless C.D. associated with nearby electronic transitions, or their curves in the instrumentally accessible region (Fig. 12). presence had merely been suggested by barely disThis is particularly unfortunate because a study of the cernible inflections in the over-all O.R.D. spectrum. O.R.D. or C.D. characteristics of unbridged comExamples of the first kind were amply provided by the pounds and a comparison of these characteristics with present study and have been discussed above. An those of closely related bridged biphenyls of the same example of the second kind is given by unbridged and absolute configuration and (necessarily) fixed cis-conbridged amines XI11 and X I X , for which O.R.D. formation could, in principle, lead to important condata5 had been of very limited value because of the formational conclusions, as in the case of the 2,2‘obvious extreme overlapping of curves. The then dinitrobiphenylsj and (below) the 2,2’-diaminobistated reservations5 have now been resolved by the phenyls. The subgroup of 6,6’-dihalo-2,2’-dimethyl-anticipated5 identification of the optically active transibiphenyls VI-VI11 is of special interest. In the tions. As shown in Fig. 15, the amines of the same O.R.D. study it had been concluded5 that a positive absolute (R)-configuration show similar long-wave Cotton effect centered a t 260-270 mp corresponds to length features of the same sign: the red-shift of the the (R) -configuration : the identical conclusion is now curve for X I X corresponds to the red-shift of the arrived a t through comparison of the appropriate aniline band maximum from 286 mp (XIII) to 302 mp C.D. curves (Fig. 13). The problem of the conforma(XIX).5 By contrast, the C.D. curves of the acid tion of amines will be discussed separately below. solutions XIV and X X are of such low intensity (to ca. 300 mp under the same conditions of measurement) Summation of Chromophores.-Inspection of the that they are indistinguishable from zero on the same C.D. curve of the bridged dinitroketone XV (Fig. 14) plot. It had been pointed out before that protonation shows that there is a marked deviation from the pattern quenches the aniline band in the absorption s p e ~ t r u m . ~ set by the other bridged 2,2’-dinitrobiphenyls (Fig. 3 It can therefore be unequivocally stated that the longand 4). This deviation had also been noted in the wave length aniline band is optically active and that it preceding O.R.D. study5and it had been been remarked is of the same sign for the bridged and unbridged that the various chromophores appear to contribute amines of the same absolute configuration. Furtherindividually to the resultant O.R.D. curve, as revealed more, it is not unreasonable to suggest that the unby algebraic analysis. These conclusions have now bridged amine XI11 is conformationally related to the been confirmed. The algebraic summation of the bridged amine X I X which has both the same absolute C.D. curves (in dioxane) of bridged (S)-2,2’-dinitrobiphenyl XVI and of (S)-dimethyldibenzsuberone (a configuration and the identically signed C.D. pattern. Since the bridged amine is necessarily cis (or syn), it bridged biphenyl ketone) bears a fair resemblance to the experimental curve for (3)-XV (Fig. 14). may be concluded that the unbridged amine is also cis. This result is of considerable interest since it had earlier Comparison of C.D. and O.R.D. Results.-The been proposedg that amine XI11 may be trans (or anti) phenomena of C.D. and O.R.D. are complementary and that protonation, which inverts the sign of rotation manifestations of the Cotton effect which is associated in the visible region, may cause a change in quadrant with a particular electronic transition.‘ The recent from trans to cis.’O That the situation might be more introduction of automatic recording instrumentation complex had already been suggested5by the preliminary suitable for measuring C.D. curves has permitted a O.R.D. data. close examination of the relative merits of the two types of measurement in stereochemical studies. I n connecTo this example of the usefulness of C.D. in the contion with another problem, a preliminary judgment of formational analysis of complex chromophores should the relative advantages of C.D. and O.R.D. had been be added the further advantage, noted repeatedly in arrived at.* The present investigation has yielded the present work, that C.D. measurements may reveal further information on this point. ( 7 ) K . M i s l o w and H. B. Hopps, J . A m . Chem. Soc., 84, 3018 fIiJli’2). ( 8 ) C . Iljerassi, H. Q‘olf and E. Bunnenberg, ibid.,84, 4532 (19fi2).
(9) D. D . Fitts, M , Siege1 and K . Mislow, i b i d . , 8 0 , 480 (1958): d. also H. hlusso and J. Sandrock, A n p e w C h e m . , 72, 322 (1960). (10) E. Bergmann and L. Engel, Z . p h y s i k . Chem., 16B, 85 (19311, had previously assigned t h e cis conformation t o XI11 on the baris of dipole moment measureme-.
May 5 , 1963
COMMUNICATIONS TO THE EDITOR
transitions-necessarily optically active-which hidden even in the absorption spectrum itself.
are
On other occasions, O.R.D. measurements have aforded valuable information which could not have readily been culled f r o m present C.D. data. The very tailing of the O.R.D. curves which has been cited as a disadvantage in the preceding discussion can be turned to excellent advantage7S8: the presence of high-intensity Cotton effects lying in wave length regions which are at present spectropolarimetrically and spectrodichrometrically inaccessible (“invisible giants”) can often be easily and unequivocally inferred by an examination of the longwave length O.R.D. region. For example, it had previously been n ~ t e d that ~ . ~ background rotations from a curve centered a t short wave lengths is superimposed on all but one of the O.R.D. curves of simple six- and seven-membered ring bridged 1,1’-binaphthyls and 2,2’-dimethyl- and 2,2’-di~hlorobiphenyIs.~~ This curve dominates the whole visible region and is opposite in sign to that of the interposed conjugation band Cotton effect a t 240-245 m p and the binaphthyl band Cotton effects in the 250-330 mp region. The position (Le., below 220 mp for the biphenyls and below 240 mp for the binaphthyls) and the sign of this high-amplitude Cotton effect had been clearly discerned in the O.R.D. study. In fact, so dominant is the background rotation in the visible due to this Cotton effect that a fairly (11) I t is not unlikely t h a t t h e background rotation in the case of t h e one exception (compound XXI in t h e present study) also makes a weak contribution t o t h e O.R.D. in t h e visible region, but i t has not thus far been possible to demonstrate such a n effect experimentally 6
1349
successful correlation of sign of rotation a t the sodium D-line and absolute configuration in the biphenyl series (the Bridge rule) had a t one time been e ~ t a b l i s h e d . ~ ~ ~ ~ By contrast, in the C.D. study only compound XXXV clearly displays a short-wave length maximum, while compounds XXIV and XXV show bare intimations of this effect. A judicious use of both C.D. and O.R.D. information is therefore clearly indicated. Present instrumentation dictates that “invisible giants” are best detected by O.R.D. measurements since their Cotton effects are felt as background curves and frequently even as the dominant effect of optical rotatory power in the region under investigation. On the other hand, C.D. is most useful in identifying relatively weak optically active transitions, especially in the longer wave length region where their presence may be obscured in both absorption and O.R.D. spectra by broad, overlapping neighboring bands.
Experimental Circular dichroism measurements were conducted in the solvents presented in Table I using a Baird-Atomic/ Jouan Dichrograph (model JO-1). The concentrations employed (Table I ) were such as to maintain the slit width below 1.3 mm. in the region of the circular dichroism maximum. The molecular ellipticity [e] is expressed1 as [e] = 2.303(4500/~)( E L - E R ) with ( E L C R ) = d X sensitivity X lo-* X mol. wt./c X 1, where d is the recorder deflection (in m m . ) a t the particular sensitivity setting (1.5, 2 or 3 ) of the instrument, c is the concentration in g./l. and 1 is the cell path length in cm. Pertinent absorption spectra were obtained using a n Applied Physics Corpration model 14 spectrophotometer.
COMMUNICATIONS T O THE EDITOR THE METAL-METAL BONDED, POLYNUCLEAR COMPLEX ANION IN CsReC1, Sir :
Lye present here a brief preliminary report of a recently determined structure which we believe provides a key to understanding various aspects of the chemistry of rhenium in its + 3 oxidation state and probably has important implicatioris regarding the chemistry of some other heavy transition elements. Despite its rather innocent looking formula, CsReCld was found to have a large orthorhombic unit cell (a = 14.06 A.,b = 14.00 fi., c = 10.69 f i . ) containing twelve of the above formula units. The possibility of an unexpectedly complicated structure thus was suggested a t a very early stage. From systematic absences, the possible space group could be limited to Cmcm, Cmc2’ and Ama2. By conventional statistical treatment’ of the distribution of intensities in the principal zones, the last space group (a non-centric one) was unambiguously indicated. A positive piezoelectric test and the complete success of the ultimate refinement of the structure leave no doubt that this choice is correct. lJsing some 580 reflections, a three-dimensional Patterson synthesis was computed and rhenium and cesium positions were eventually established by semisystematic use of minimum functions of several ranks.? Using only these heavy atoms for phasing, a structure factor calculation was made giving a discrepancy index, It, of only 35%, thus indicating the probable correct(1) E. R . Howells, 11. C. Phillips and D. Rogers, A c f a C y y s I . , 3, 210 1030). ( 2 ) h l . J. Buerger, “Vector Space,” John Q’iley and Sons, I n c . , Xew York, K. Y., 19lil. (
ness of the heavy atom coordinates. One cycle of least squares refinements then was run varying all parameters and a Geller correlation matrix constructed. This showed several strong parameter interactions and refinement thus was continued allowing only those parameters which did not interact strongly to vary simultaneously. Ten such cycles reduced R (for Re and Cs only) to 18%. A difference Fourier map, in which the heavy atoms were subtracted out, revealed the chlorine positions quite clearly. After a number of cycles of full-matrix, least squares refinement, the discrepancy index had dropped to 7.3%, and a difference Fourier synthesis showed no anomalies. The structure of the complex anion is shown in perspecJive in Fig. 1. Some of the important bond lengths (in Angstroms) and bond angles (in degrees), with estimated standard deviations in parentheses, are ReI-Re11 = ReII-Re‘II = 2.47 (0.01) Re-Cl (bridging) = 2.39 (0.03) Re-C1 (out of plane) = 2.36 (0.03) Re-C1 (in plane) = 2.52 (0.03) CI?-Clj = Cl7‘-Cls = C17-Cl7’ = C16-C18, etc. = 3.28 (0.06)
