March 20, 1964
N.M.R.SPECTRA
T h e in-plane components of the CB,C( p-orbitals are then resolved along their respective internuclear lines with C,. The question now remains regarding the proper vectorial decomposition of the C? in-plane component with regard t o the two internuclear lines. One can project the in-plane component onto one of the internuclear lines, obtain the magnitude of this component, and then repeat this operation along the other internuclear line (method -4). Alternately, one can require t h a t the vector sum of the components along the two internuclear lines be equal to the original in-plane component (method B ) . These two procedures are compared in the figures below.
OF
Method A gives results virtually identical with the procedure described by Roberts,24 b u t method B gives different*values. Using Kopineck’s tables,26a t a Cz-Ci distance of 2.296 A , , S2, is 0.0792 and 0.0641 for methods A and B, respectively, which compare t o Spi = 0.0802 quoted by Roberts. In order t o bracket 6 2 7 the secular equation was solved for 6 2 7 = 0.3394 (ref. 24) and 6 2 7 = (0.0641/0.28)8 = 0.22898 (method B ) . The values of the carbonyl oxygen coulomb integral ( a 26) and C=O exchange integral (0) were those listed by S t r e i t w i e ~ e r . ~ ’ The energy, E”, of the nonbonding ( n ) orbital of oxygen is assumed to be independent of the environment of the carbonyl group. Since the energy of the n K * transition is En-.-* = E.* - E”, the difference in n K* transition energies is then dependent upon only the relative energy levels of the first antibonding orbital, E(l)n+l* E0, but < 1 , since w h for the entire multiplet was 10.6 i 0.1 c.p.s. Each olefinic proton (as H d ) is coupled directly to the adjacent bridgehead proton (as H,; J c d = 2.85 c.P.s.) and is virtually coupled through the adjacent olefinic hydrogen (as He) to the other bridgehead proton (as H f ) . The outer line separation S = 3.5 C.P.S. (Table I ) for the resulting olefinic “triplet” observed should equal the sum J c d Jce. Since J c d = 2.85 c.p.s, the value Jce= 0.63 C . P . S . is very likely. “Virtual coupling” once again is effective in averaging the actual coupling constants between each olefinic proton and both allylic, bridgehead hydrogens. The remaining two protons H a , H b are coupled to each other l J a b ’ = 9.1 + 0.1 c.p.s. (Table I ) , b u t the chemical shifts remain to be assigned. The upfield ( a ) resonances are split into “regular” triplets with 1: 2 : 1 intensities; this can be due only to interaction with the two equivalent bridgehead protons, J H ~ H , ( H ( ) = 2.0 f 0.1 C.P.S. The other bridge hydrogen ( p ) Fig. 2.-Suggested geometry for the association between comgives rise to lines broadened, but not split, presumably pound I and benzene, as deduced from the solvent shifts; the also due to coupling with the two bridgehead protons observed (&) and calculated (6,) chemical shifts are relative to Hc and Hf. Estimation of J H ~ Hand ~ ( an H~ explana) tion for the obscuration of the details of the @-proton CC1, solution. peaks, assignment of the Ha and H b chemical shifts, and the case for I ; solvents of higher dielectric constant dean explanation of the appearance of the 192 c.p.s. shield the protons in an approximately regular manner, bridgehead H,,Hf hydrogen resonance remain. Solvent shifts in benzene are quite different due to Use of Selective Solvent Shifts.-In order to assign “ring current.” anisotropy. Upfield shifts of relatively Ha and Hp resonances, the effect on the position of the large magnitude are observed with a polar solute such various absorptions with change in solvent was exas I. Dipole-induced dipole interactions between I and amined (Table 11). It is believed that polar solutes, such benzene result in a weak 1: 1 molecular complex with the solute positioned above the plane of the benzene TABLE I1 ring. 4 2 - 4 5 The well-known anisotropy of benzene reS O L V E X T SHIFTS FOR COMPOUSD I” sults in increased shielding for all of the protons of I ( € - l)P (Table II), since all are in the diamagnetic region above Solvent ( c + 1) A(Hd,H,) A(H,,Hh) A(H,,Hr) A& AH@ the ring.25.46 cc1, 0 38 0.0 0.0 0.0 0.0 0.0 A reasonable approximation of the possible geometry CeH I 2 .33 -2.9 +4.6 f6.0 . ... . . . . of the association of benzene with I is depicted in Fig. 2. cs2 .45 -2 8 +l.5 +3.0 -0.5 +4 5 It is observed experimentally that the olefinic hydrogen CHCla .66 -9.3 -6.0 -2.5 -2.0 $3.5 ( H d and He) resonances are shifted least in benzene; CHzC12 .79 -8.3 -i.0 -1.8 -1.5 +3 0 these protons should be farthest away from the aromatic (CHa)2CO .90 -6.7 -16 5 -1 0 ... ring.43.47Associations such as this are usually found to CHaCN .95 -7 8 -12.5 -1 0 . . . have the negative end of the solute dipole tending t u lie CsH6 .34 +3 2 +32 0 +29 8 + 3 1 . 5 + 5 1 . 5 off the ring and to be parallel to the plane of the Hydrogens H, and H h may be sufficiently activated by a 5C;; a . / v solutions with TMS internal reference. The variations in chemical shifts with changes in solvent ( A , in c . P . s . ) the adjacent chlorine atoms48 to act as weak proton are taken as positive when the resonances move upfield, negadonors in hydrogen bonding to the benzene ring. Soltive when they move downfield, with CCla positions a s the refervent shifts of H, and H h in proton-accepting solvents ence. Dielectric constant function; see text. (Table 11),while small in absolute magnitude, are larger than those of other hydrogens in the molecule. On the as I , polarize nonaromatic solvents in such a way as other hand, a downfield shift of only 3 C.P.S. from CC1, to create a “reaction field” which affects the magnetic was observed for H, and H h in ether solution.45 Weak shielding of solute protons in an approximately linear way; solvent shifts are roughly in proportion to a func( 1 2 ) L. W Reeves and W G Schneider, C a n . J . Chem . 36, 2 5 1 (15571 (43) J V. H a t t o n and R E . Richards, Moi P h y s . , 3, 253 (1560). 6 , tion of the dielectric constant of the solvent, ( e - l ) / 135, I53 (1962). (e 1).40,4’Inspection of Table I1 reveals t h a t this is (44) J 1’.H a t t o n and W G Schneider, C a n J . Chem , 40, 1283 (1962)
+
/m
+
(40) A D . Buckingham, C a n . J C h e m . , 38, 300 (1560); A. D. Bucking. h a m , T . Schaefer, and W G . Schneider, J . C h e m . P h y s . , 32, 1227 (19601, R . J . Abraham, ibid., 34, 1062 (1961); A . D Buckingham, T . Scbaefer, and W . G Schneider. ibid. 34, 1064 (1961); J . I. Musher, ibid.,37, 34 (1962) (41) P. Diehl and R Freeman, M o l . Phys., 4, 39 (1961).
(45) W. G . Schneider, J . P h y s C h e m . , 66, 2653 (1962). (46) C E. Johnson, Jr , and F A Bovey, J Chem P h y s , 2 9 , 1012 (19.58) (47) H. M . McConnell, ibid., 27, 226 (1557). ( 4 8 ) A Allerhand and P von R . Schleyer, J (1963).
A m Chem S o c . 8 6 , 171.5
1174
PIERRE
LASZLO AND
P A U L VOX
RAGCBS C I I L E Y E R
VOl. 86
T A B L E111 DESCRIPTIOV O F S P E C T RAA\ D CHEMICAL SHIFTS Hb
HB
85
101 d.t 107 d.q. 106 d.d.t 92 d.d.t 65'
80'
?
?
117
117
d 120 d 120 d 80-90
H C
Hi
H,
Ht
H,
Hh
HI
192 VI1 179
376
376
192 \ I1
261
264
..
186
255 ABX
I11
376
\
1' 208 1-11 181 br
111
89 br. a wh = 7 . 5 . = 5.0. solution. Kef. 21d.
Z'h
'
89 br . = 8.0.
A
w,,
=
399
(96)
(96)
ti7 dd ?
111
363
reference solvent, CCl,, is known, furthermore, to give rise to erratic shifts.l' ,4 critical test for the coupling constants resulting from our analysis is the appearance of the bridgehead hydrogen (H,,Hf) resonance, seen from Fig. 1 to be an "irregular" septet (5'= 10.0 c.P.s.). The multiplicity is accounted for by coupling or virtual coupling of each bridgehead proton to six other hydrogens-every other hydrogen in the molecule except its isochronous counterpart. The sum of the coupling constants alJ f g J i h = 2.0 -k ready determined is: Jar J d f J,i 0.65 2.85 3.2 0.0 = 8.7 c.p.s. I n consequence, the upper limit of the one coupling constant remaining, J b f , is 10.0 - 8 . i = 1.3 c.P.s., a value consistent with the observed width a t half-height (3.0 c.P.s.) of the Hb ( p ) resonance. X splitting of this magnitude should be resolvable as a triplet, yet only a broad line is observed. A likely explanation is broadening due to additional coupling of Hb to both olefinic protons Hd and He, perhaps via overlap of the a-orbitals of the double bond with the rearward lobe of the Hb sp3-hybrid orbital. Well documented examples of such specific longrange coupling between olefinic hydrogens of i-substituted norbornene and norbornadiene derivatives and the anti- but not the syn-C, proton have now been reported ( J E 0 . S ~ . p . s . ) . ~ I n~the ~ ?case ~ ~ of compound I , this coupling (Jbd) must be less than 0 . 5 c.p.s. judging from the appearance of the olefinic side band pattern. The evidence of long-range Hb-Hd(He) Coupling reinforces the assignments, CY = Ha, /3 = Hb.
+
,.--.d~-------
,
1
-7T
,I"
Fig 3 -Spectrum of compound I1 in CCla solution (bottom trace) box A , appearance of the olefinic multiplet in CsH6solution, box B, 13Cpattern relative t o t h e olefinic protons
Fig 4 -Spectrum
268 d t
161 5 dd 135 ddd
216 t 124 dd
167 147 49 A BXY br dd d d 6.0.e Insoluble in carbon tetrachloride; this spectrum was examined in acetonitrile
11
lm
'
111
363
C-H. . . K hydrogen bonding, if present, would be facilitated by the geometry indicated in Fig. 2. The two 7-protons, Ha and Hb, are shielded by different amounts in benzene solution. Examination of Fig. 2 shows t h a t Hb ( s y n to H, and Hh, strongly shifted) is closer to the benzene ring and should be more strongly shifted than Ha(syn to Hd and H e , weakly shifted). On this basis we assign resonance lines CY (benzene shift S 0 . 5 p.p.m.) to Ha and lines p (benzene
_- -t;--
\ I 5 380 20.3 .1BST; br 5 376 5 185 5 AHST; br 5 367 5 171 111 hr = He - Hr = 1-43? I11 1 399 208 111
178 5 368 br 173 5 363 br e 171 357 br H,i - H,
>
376
I11
1
I
+
r--
+
+ + + +
-
of compound 111 in CCla solution
shift E 0.9 p.p.m.) to H b . Using the Johnson-Bovey tables,?6which are known to give an accurate estimate of the anisotropy effects in the neighborhood of the benzene ring, we have calculated the benzene shifts expected from the model (Fig. 2). Agreement with experimental shifts are reasonably good, especially since all other types of solvent effects and the possibility of hydrogen bonding were ignored in making the calculations. The
5.0
Fig. 5.-Spectrum
L.0
of compound \*I in CH3CN solution.
Analysis of the n.m.r. spectra of the remaining compounds studied, 11-VJII, followed similar procedures. Data are summarized in Tables 111-V and the spectra of compounds not available in the l i t e r a t ~ r e ~ ,are ~~.'~,~~ reproduced in Fig. 1, 3-5,
COGPLINGBETWEEN Compd.
1175
N.M.R.SPECTRA OF NORBORNENES
hlarch 20, 1964
Jcd
Jdc
Jc f
JC,
TABLE IV OLEFINICASD ALLYLICPROTONS
THE
Jd i
Jbd
Jcf
Jbe
Appearance of t h e olefinic multiplet
J"C-H,j
I11 s = 3 . 5 I11 S = 3 8h . . . -4BXY W ] , = 2 0 ... L~BX\- ?Llh = 2 0 1-21" 0 55" 0.55" 166 f 1 111 s = 3 5 a 111 s = 4 0 1.I ./,,I J,,. = 4 0 a 111 S = 3 . 6 172 f 1 VI17 l i r 2 71) 0 95 5 05 0.95 2 70 J,, = J,ir = 0 . 9 0 " 1.11115 :3 0 ? 0 5 55 ? 3 0 ? . . .dBs\' W h = 2 0 Use of the selective benzene shift resulted in a striking modification in the appearance of 0 Calculated as the difference .C - Jcg,. this multiplet, which was thus transformed from an "irregular" triplet (111) to the regular 8-line pattern of a n A B X Y system, the accidental equivalence of H,] and H,, having been removed by their unequal shielding effects from the aromatic ring anisotropy (see Fig 3 ) Esaniination of this pattern yielded: JdP = 5.2 j= 0 . 2 ; J c d = Jet = 3.0 f 0.2. I I1 111 I 1.
2.85 2 90 3.0 2 80 2 93
-0 -0 -0 -0 -0
0.65" 0.70"
+
5.55 5.55 5 60 5 60 5 80
2 2 2 3 2
0.65" 0 70"
164 f 1 166 f 1
1 5 4 5
J,,
= J,i
2 2 1 1
1
Jal
0 0 8 95
+ JIx +
2 0 3 0 3 1
Jch
Jaf
+
Jbi
J fg
3 2 3 9 3 6
Jtx
= Jbf
