bulk electrolyte results in the deactivation of the electrode surface. In the above experiment, platinum was deposited at a rate greater than that of impurity adsorption. The identity of the impurity is not known. Bruckenstein et al. (21) found that oxygen reduction at platinum electrodes in 0.2M HzS04is inhibited by deposition of copper. Their electrodes were reactivated by anodic polarization at f0.6 V at which potential the copper is anodically stripped from the electrode surface. In our case, electrode activation was not
achieved unless anodic potentials greater than 1.2 V were used in the pretreatment procedure. The results reported above support the conclusion of Damjanovich et ul. (22) that the deactivation of a platinum electrode will occur as a result of the adsorption of impurities from the solution. RECEIVED for review March 15, 1971. Accepted April 29, 1971.
Proton Magnetic Resonance Identification of Nonconjugated Cis-Unsaturated Fatty Acids and Esters David J. Frost and John Barzilay Unileuer Research, Vlaardingen, The Netherlands PROTONMAGNETICRESONANCE (PMR) has already been extensively used in the structural analysis of unsaturated fatty acids (1-4)> although the two important factors of the stereochemistry and position of double bonds have hitherto been determined for only the simpler of these compounds. Use of high-field nuclear magnetic resonance (NMR) spectrometers amplifies chemical shift differences and has enabled these to be correlated with more subtle structural differences, The effect of the stereochemistry of double bonds upon the chemical shift of the protons upon adjacent carbon atoms has already been reported (5). In the present work, 8 saturated and 39 nonconjugated cis-unsaturated fatty acids and esters have been studied at 220 MHz, and the long-range deshielding effects (6) of the relevant functional groups were found to be additive. These parameters enable the &values of the absorptions in other such fatty acids to be calculated to a sufficient accuracy to provide a sensitive method of determining the positions of unsaturation in a long chain. EXPERIMENTAL
The unsaturated fatty acids and esters were synthesised in our laboratory, while the commercial saturated compounds were of pro-analyse standard. Measurements were made upon 10-20 vol solutions in CClr at 220 MHz and about 13 “C (Varian HR 220). 6Values measured were accurate to +0.005 ppm (i.e.,about 1 Hz at 220 MHz) and were relative to internal tetramethylsilane (TMS). RESULTS
The measurements upon the short chain (C2-C5) saturated fatty acids were found to be concentration dependent, pre(I) J. M. Purcell, s. G. Morris, and H. Susi, ANAL.CHEM.,38, 588 (1966). (2) . , F. D. Gunstone and I. A. Ismail. Chem. Phvs. LiDids, . . 1.. 337 (1967). (3) F. D. Gunstone, M. Lie Ken Jie, and R. T. Wall, ibid.,3, 297 (1969). (4) M. van Gorkom and G . E. Hall, Specrrochim. Acta, 22, 990 (1966). ( 5 ) D. J. Frost and J. Barzilay, Rec. Trac. Chim. Pays-Bas, in press. (6) R. F. Ziircher, “Progress in NMR Spectroscopy,” Volume 2, Pergamon Press, Oxford, 1967, p 205. ~
1316
sumably because of the considerable hydrogen-bonding effects present. As the chain length increases, concentration changes produce smaller effects, and measurements from acids of Cg and longer were included in the correlations. As was expected, the results from esters were less affected by concentration, and ethyl butyrate was the smallest ester measured. The 220 MHz spectra revealed that the cis double bonds and carboxylic groups produced measurable deshielding effects upon methylene groups up to five or six carbon atoms removed from the functional group. Consequently, all unambiguously assignable absorptions were written in terms of a basic value for a CH2 group in the middle of an alkyl chain &, plus a- to {-substituent effects of the functional groups concerned. A similar procedure was applied to the methyl absorptions, and it was found that there was no significant difference between the substituent effect derived from both sets of data. The substituent effects calculated are given in Table I. The parameters given are essentially the result produced by replacing an alkyl group with the functional group concerned, which explains the apparent anomaly of an a-deshielding effect produced by a methyl group. From the 215 assignable absorptions in the aliphatic spectral region of the 47 fatty acids and esters measured, all but four of the calculated &values agreed to better than 0.015 ppm with the measured value. The only two large deviations, 0.041 and 0.043 ppm, are included in Table 11. These two groups are readily recognized, however, and if treated as exceptions to the additivity rule (6), present no problems of interpretation. Characteristic Groups. Use of the substituent effects in Table I readily identifies most characteristic groups in the unsaturated fatty acids. The most prominent of these are presented in Table 11. The 6-values given concern the situation when no other groups are sufficiently close to be of influence (Le., more than five carbon atoms distant). The presence of unsaturation in the neighborhood of the terminal methyl group is also readily detected by variation in the 6-value of this absorption (Table 111). The pattern of the absorption itself of course also varies in this series, being a doublet (J = 5.2) when n = 0, a sharp triplet (J = 7.5) when n = 1, and a less well defined triplet in the other cases.
ANALYTICAL CHEMISTRY, VOL. 43, NO. 10,AUGUST 1971
Table I. Deshielding Effects of Functional Groups along an Alkyl Chain (in ppm); Basic &Values; Methylene Methyl & = 0.883
an = 1.253;
Abb.
CY
P
Y
6
6
s
//O C 'OH
AH
1.037
0.362
0.094
0.055
0.029
0.005
//O C
AM
0.959
0.322
0.061
0.036
0.021
0.005
AE
0.938
0.320
0.059"
c=c
CD
0.733
0.063
0.019
0.015
0.007
O.Oo0
H H H H C=C-CHz-C=C
CDS
0.768
0.073
0.025
0.024
0.009
O.Oo0
Me
M
0.030
0.000
0.000
O.Oo0
O.Oo0
O.Oo0
Functional group
'OMe C
yo
0,024"
'OEt H H
a
One measurement only. Table 11. PMR Features of Characteristic Groups in Fatty Acids and Esters
&value Group H H CHzC=C H H H H CHzC=CCHzC=C H H H H C=CCHzC=C H H H H H H C=CCHzC=CCHzC=C
Effects"
Found
Calcd
CYCD
1.973-1.995
1.986
CYCDS
2.011-2.030
2.021
2CYCD
2.112-2.125
2.719
2.741-2.764
2.754
2.779 2.754
2.189 2.754
2.036-2.041
2.049
1.995 1.368-1.375
2.005 1.379
CYAE
2.289-2.295 2.205-2.218 2.182-2.198
2.290 2.212 2.191
PAH
1.611-1,620
1.568-1.582 1.559-1.582
1.615 1.515 1.513
3,025
2.980
W D
H H H H H H H H C=CCHzC=CCHzC=CCHzC=C b a b
H H H H C=CCH2CHzC=C
H H H H C=CCHzCHzCHzC=C
a
b
a
CHzCOzH CHzCOzCHz CHaCOzCHzCH3 CHzCHzCOzH CHzCHzCOzCH3 CHzCHzCOzCHzCH3
f
W D
ffCD
f
aCDS
f
WDS
PCD
+
a ~ C D j b 2PCD
YCD
ffAH ffAM
PAX
PAE
Pattern
A
A
h
A
A A
Deviations H H H H C=CCHzC=CCHzC02CH, H H C=CCHzCHzCOzCH 3
b a
a
(YAM
a b
~
+ A
CYCD
ffCDS
+ PCD + PAY
J4
2.275
Y
ca. 2.265
2.308
For abbreviations see Table I.
ANALYTICAL CHEMISTRY, VOL. 43, NO. 10, AUGUST 1971
e
1317
Table 111. Influence of Cis Double Bonds upon the Terminal Methyl Absorption 0 1 2 3 4
n
H H CHa(CH&C=C H H H H CHa(CHz),C=CCHzC=C n Only one measurement.
>4
1.616
0.946
0.902
0.898
0.890
0.883
1.651
0.956
0.908
0.907"
0.892
0.883
Table IV. Calculation of &Values for (CH2), Protons in AS, A9, and A10 Octadecenoic Acids Cornpound
R
A8 CHI
=
CHa CHz CHZ CHz CHZ CHZ C j i h g f
C
C
CHz CHz C e d 1.28 1.25 1.25 1.26 1.27 1.27 1.32
C
C
C
C = C
C
I
CHz C e 1.28 1.25 1.26 1.27 1.28 1 . 3 2
C = C
1.31-1.34 C CHz d 1.32 1.30 1.32 1 . 3 5
d L
A
2
1.25-1.28
1.30-1.35
C
CH1 CHz e d 1.28 1.26 1.27 1.27 1.32 1.32 1.28 1.29 1.30 1 . 3 1 all absorptions within 1.26-1.32
a
C;10:5
2
A
1.34 1 . 3 1 1 . 3 3
1.25-1 28
A10 R = CH3
220 M H pattern ~
Y
L
A9 R = H
CHZ CHZ CHQ CHz CHz C //O c b a \OR
-
C = C
C
,A,
1.40
1.a 1.20
c a b O' H
1
f C16:7
-
e c a d b O' H
Figure 1. 220 MHz spectra of C18A5 (a), c18A 6 (b), and cl8A 7 (c) acids
Application. The sensitivity of the results from the 220 MHz spectrometer is such that almost all cis-unsaturated nonconjugated fatty acids and esters (chain lengths up to about 22 carbon atoms) should be identifiable by PMR. The application of this to the series of mono cis-unsaturated c 1 8 acids now means that positive identification may be made of all but two isomers (it is not possible to distinguish between the AlO, A l l compounds). This is demonstrated in Figure 1 and Table IV. It is particularly significant that the difference between the A8, A9, and A10 isomers, which is primarily apparent in the splitting of the large (CHJ, absorption (6 1.25-1.35), may be qualitatively and quantitatively explained (see Table IV). Application of the parameters in Table I further gives a very good correlation with the measurements of Gunstone and coworkers (2, 3). 1318
CONCLUSIONS
The results show that high resolution PMR at high magnetic field has a unique potential for the identification of unsaturated fatty acids. Further work is in progress to extend the results to trans-unsaturated, and acetylenic fatty acids. ACKNOWLEDGMENT
We thank G . J. N. Egmond, J. B. A. Stroink, G . L. van der Schee, L. van der Wolf, and A. Steenhoek for making samples available, and J. A. C. M . van der Ven, L. Hoogendoorn, and Miss A. Groenewegen for some experimental assistance. RECEIVED for review January 4, 1971. Accepted April 26, 1971.
ANALYTICAL CHEMISTRY, VOL. 43, NO. 10,AUGUST 1971