Temperature-Dependent Splitting Constants
1853
Temperature-Dependent Splitting Constants in the Electron Spin Resonance Spectra of Cation Radicals. IV.' The Ethoxy Group Paul D. Sullivan Department of Chemistry. Ohio University. Athens. Ohio 45701 (Received March 8. 7973) Publication costs assisted by the Petroleum Research Fund and the Ohio University Research institute
The temperature dependence of the proton splitting constants of a series of ethoxy substituted cation radicals has been measured. Particular attention has been paid to the magnitude and temperature coefficients of the splittings from the 6- and y-ethoxy protons. Qualitative conclusions regarding the molecular conformation of the ethoxy group have been drawn from the ðoxy proton splitting constants. INDO calculations on the ethoxymethyl radical and ethanol cation radical have been used to rationalize the 8and y-ethoxy proton splittings. The averaged p-alkoxy proton splittings are given by ugH = Qofi-klpoT + Q ~ f l - ~ where p ~ ~ ,Q o P - ~ = +33.4 G and Q c 6 - H = -4.99 G. Similarly the y-alkoxy proton splitting is " Q o T - ~ and Q c Y - ~ depend on the dihedral angle about the 0-C given by u y H = Q O ~ - H p o r Q ~ y - ~ p cwhere (alkyl) bond. Good agreement for the freely rotating y proton is found between the experimental and calculated splitting constants.
+
study are shown in Table I. The important points to note are that the methylene ( p ) protons8 of the ethoxy group have a splitting constant of 3.80 G (at -40") compared to 3.35 G for the methoxyl protons of p-dimethoxybenzene (DMB)7 and that their temperature coefficient is -3.72 mG/deg (cf. 0.09 mG/deg for DMB). Also, a splitting of 0.144 G is detected from the y-ethoxy protons. 1,4-Diethoxy-2,5-dimethylbenzene (DEDMB). The cation radical of DEDMB was formed in AlC13-CH3N02 and gave well-resolved spectra (line width ca. 0.040 G) over the temperature range from 0 to -50". The spectra are readily analyzed in terms of only one isomer, presumably the trans form, and the splitting constants (see Table I) account for all the protons in the molecule. Again a large Experimental Section temperature dependence is found for the @-ethoxyprotons p-Diethoxybenzene was a commercially available com(-3.46 f 0.3 mG/deg) and splittings from the y-ethoxy pound and was recrystallized before use; 1,4-diethoxyprotons are observed which are also temperature depen1,4-diethoxy-2,5-di-tert-butylben- dent (-0.29 f 0.03 mG/deg). The methyl splittings are 2,5-dimethylbenzene, zene, and 4,4'-diethoxybiphenyl were prepared by allownot measurably temperature dependent whereas the ring ing ethyl iodide to react with the appropriate phenol folproton splittings do show a significant dependence (-0.62 lowing a method outlined by R i ~ t o w 1,2,4,5-Tetraethoxy.~ 0.06 mG/deg). benzene is found as a -1% impurity in commercially 1,4-Diethoxy-2,5-di-tert-butylbenzene(DEDBB). The available 1,2,4-triethoxybenzene and can be prepared as esr spectrum of the cation radical of DEDBB produced in the cation radical from this source. 1,4-Diethoxy-2,5-di- AlC13-CHaN02 was obtained over the temperature range methylthiobenzene and 1,4-diethoxy-2,5-diethylthioben- from -40 to +20°. The spectrum is again analyzed in zene were a gift from Dr. Z. I. Ariyan. The cation radicals terms of only one species. The major splittings are a quinwere prepared by treating the neutral compounds with tet from the @-ethoxyprotons which shows a large temperaluminum chloride or concentrated sulfuric acid in nitroature dependence (-5.2 mG/deg) and a triplet from the methane or nitroethane.6 The esr spectra were measured ring protons which has a small but significant temperaon a Varian E-15 spectrometer in a dual cavity using a ture coefficient (-0.76 mG/deg). Each of these 15 major sample of the perylene radical anion as a secondary stanlines is split further by small splittings from the y-tertdard. The least-squares analyses of the experimental specbutyl protons and the y-ethoxy protons. The overall pattra were carried out as previously described.3 tern of this additional splitting changes markedly with temperature. By using expanded field sweeps these small Results splittings were investigated and computer simulations inp-Diethoxybenzene (DEB). The cation radical derived dicated that the splittings were made up of two compofrom this compound in AlC13-CH3N02 has been previousnents. First, a temperature-independent splitting of 0.100 ly investigated rather extensively.7 The radical was shown G which could be attributed to the y-tert-butyl protons. to exist as cis and trans isomers a t room temperature and This assignment was made on the basis of expected intenbelow. The averaged values of the splitting constants and sity ratios for 18 equivalent protons and by analogy to the the temperature coefficients obtained from the previous same splitting in 1,4-dimethoxy-2,5-di-tert-butylbenzene
Introduction Previous papers in this series have studied the temperature dependence of the methoxyl and hydroxyl proton splitting constantsl-3 in a number of cation radicals in an effort to obtain potential barriers to rotation. It is the purpose of this paper to extend these studies to the temperature dependence of the ethoxy proton splitting constants. Our object in this paper is not, however, to obtain quantitative information on the potential barriers to rotation but rather to give an explanation of the magnitude and temperature dependence of the p- and y-ethoxy proton splitting constants4 in terms of specific molecular motions and conformations and with reference to INDO calculations.
*
The Journal of Physical Chemistry. Vol. 77, No. 75. 7973
Paul D. Sullivan
1854 TABLE I: Summary of Esr Results for Substituted p-Diethoxybenzenes
OCH,CH, Substituents R2
= R5
R3
Splitting constants, G, and temperature coefficients,e mG/"C
= Rs
a2 = a5
a3 = a s
aoCHzH
2.272
3.802 (-3.72) 3.636 f 0.002 (-3.46 f 0.3) 4.0000f 0.0008 (-5.21 f 0.24) 2.6556 f 0.0008 (-2.90f 0.03) 1.967f 0.001 (-2.56.f 0.06) 1.964f 0.006 (-2.52 f 0.16) 2.0796 f 0.0008 (-2.30 f 0.10)
~ O C H Z C H ~ ' ~ ~
____ g factoru
a#/a, _ I
H
H
2.272
CH3
H
4.113 f 0.001
C C
(CH3)3C CH3CH20 CH3S
H
0.1002 f 0.0011
H
C 2.6556 f 0.0008
H
(-2.90f 0.03) 3.457 f 0.001 C
CH3CH2S
H
3.194 f 0.004 (-0.75 f 0.10)
4,4'-Diethoxybiphenyl
d
C
0.6032f 0.0007 (-0.62 0.06)f 1.0536 f 0.0009 (-0.76 f 0.11) 0.8579 f 0.0008 (+0.27f 0.03) 0.789 f 0.001 (4-0.37f 0.05) 0.770 f 0.006 (0.52f 0.11) d
*
0.144
2.00371
26.40
2.00363
26.63
2.00366
25.77
2.00401
23.33
2.00690
25.35
C
0.1365f 0.0005 (-0.29 f 0.03) 0.1552f 0.0011 ( - 0.25) 0.1138 f 0.0008 (-0.16 f 0.02) 0.0775f 0.0013 (-0.25 f 0.02) C
2.00690
0.0819f 0.0006 (-0.12 f 0.02)
2.00323
25.24
Estimated errors f0.00002. The ratio of the splittings of the fl and y protons of the ethoxy group. Not measured. Not applicable. e Given in parentheses immediately below the approximate splitting constant. f The signs of the temperature coefficients for the ring protons are related to the sign of thelr coupling constant. It is planned to discuss this point in more detail in a future publication.
TABLE II: Least-Squares Analysis of the Splitting Constants for the 1,2,4,5-Tetraethoxybenzene Cation Radical
-30 -20 -10 0 4-10 +20
2.6821 f 0.0012 2.6556 f 0.0008 2.6250 f 0.0009 2.5951 f 0.0009 2.5684 f 0.0008 2.5373 f 0.0018
0.8569 zk 0.0012 0.8579 f 0.0008 0.8623f 0.0010 0.8650f 0.0009 0.8685 f 0.0008 0.8691 f 0.0018
0.1141 f 0.0012 0.1138 f 0.0008 0.1109f 0.0010 0.1096 f 0.0009 0.1072 zk 0.0008 0.1069 f 0.0017
cation radical ( u H t - ~=~ 0.103 G).9 Second, a temperature-dependent splitting of L'U. 0.150 G with a temperature coefficient of cu. -0.25 mG/deg was assigned to the yethoxy protons. I , 2,4,5Tetruethoxy benzene (TEB).The cation radical of TEB can be produced by the reaction of commercial samples of 1,2,4-triethoxybenzene with either HzS04 or AlC13 in nitromethane or nitroethane. The TEB is believed to be present as an impurity to about l% and the 1,2,4-triethoxybenzene is apparently not oxidized under the conditions employed. The esr spectrum of TEB is readily interpreted in terms of a nonet of triplets from the @-ethoxyand ring protons with additional hyperfine splittings from y-ethoxy protons. In order to indicate the precision of our results the least-squares analysis of the splitting constants for TEB over the range from -30 to +20° is given in Table 11. From these results one finds temperature coefficients of -2.90 f 0.03, +0.27 f 0.03, and -0.16 f 0.02 mG/deg for the @-ethoxy,ring, and y-ethoxy protons, respectively. 1,4-Diethoxy-2,5-dimethylthiobenzene (DEDMTB).The cation radical of DEDMTB was obtained in AlCl3The Journal of Physical Chemistry. Vol. 77. No. 75. 7973
CH3N02 and was analyzed in the temperature range from -40 to +lo". The esr spectrum consists of a septet from the 8-methylthio protons (3.457 G), a quintet from the @ethoxy protons (1.967 G), a triplet from the ring protons (0.789 G), plus a small splitting from the y-ethoxy protons (0.0775 G). The g factor is significantly larger than the radicals containing only oxygen atoms and the line widths are also significantly increased. Temperature coefficients were obtained for the @-ethoxy,the ring, and the y-ethoxy 0.06, +0.37 f 0.05, and -0.25 f 0.02 protons of -2.56 mG/deg, respectively. l,4-Diethoxycy-2,5-diethylthiobenzene (DEDETB). The esr spectrum of the DEDETB cation radical in AlC13CH3N02 is entirely analogous to the previous compound DEDMTB. The major difference is the substitution of a major quintet splitting of 3.19 G from the @-ethylthioprotons. Another difference between the two compounds was the failure to resolve any splittings from either the y-ethoxy or the y-ethylthio protons in the DEDETB spectrum. Temperature coefficients were obtained for the @-ethylthio, @-ethoxy,and ring protons of -0.75 f 0.10, -2.52 0.16, and +0.52 f 0.11 mG/deg, respectively. 4,4'-Diethoxybiphenyl(DEBP).The cation radical was produced in AlC13-CH3N02 over the temperature range from -30 to +lo". The analysis of the esr spectrum was complicated by the fact that rotational motions about the C(ary1)-0 bond are occurring. This results in a line width alternation of the lines associated with the ring protons. The ethoxy group splitting constants are unaffected by this cis-trans isomerism and since these are our main concern in this paper only the splitting constants for the @and y-ethoxy protons are shown in Table I. Temperature coefficients were measured as -2.30 f 0.10 and -0.12 f 0.02 mG/deg for the @ andy-ethoxy protons.
*
*
1855
Temperature-Dependent Splitting Constants
Discussion T h e Ethoxy Group. In order to explain the temperature dependence of the ethoxy group protons one should consider the various torsional oscillations which can occur in the alkoxy group and also the mechanism by which the hyperfine splitting is produced. Previous studies on methoxyl-substituted compounds have led us to believe that the primary mechanism responsible for the hyperfine splitting of the @-alkoxyprotons is via a hyperconjugative interaction with the spin density on the oxygen atom. In addition, there is likely to be a small long-range contribution to the splitting from spin density on the adjacent carbon atom; this contribution should be small and has been ignored for the moment (see INDO Calculations). Two torsional oscillations can effect the magnitude of the hyperfine interaction. First, oscillations about the C(ary1)-0 bond would change the conjugation and hence the spin density on the oxygen atom (Le., a change in 8, see Figure 1). Second, oscillations about the 0-C(alky1) bond (change in a) can alter the hyperconjugative interaction between the @ protons and the spin density on the oxygen pz orbital. Both of these oscillations can in principle lead to a temperature dependence of the hyperfine splitting constant. Our studies on methoxyl compounds have indicated that the oscillation about the C(ary1)-0 bond has a large potential barrier to rotation associated with it (ca. 16 kcal/mol for p-dimethoxybenzene), when the methoxyl groups are not sterically hindered, and hence leads to only a very small temperature dependence of the methoxyl splitting constants. The oscillation about the 0-C( alkyl) bond is thought to have a very small barrier to rotation associated with1 it and for the methoxyl compounds essentially free rotation occurs around this bond. In the light of these results it is proposed that torsional oscillations about the C(ary1)-0 bond are of little importance to the ethoxy group and contribute only slightly to the temperature dependence of the @ protons. Restricted rotation about the 0-C(alky1) bond is considered to be the major origin of the temperature dependence. Since this is so, the results can be interpreted by analogy with previous work on @-alkyl protons.lOJ1 Assuming that the @-ethoxyproton coupling is given by eq 1, where @ is the dihedral
angle between the C-H bond and the oxygen pz orbital and (cos2 a) represents the time-averaged value of cos2 (a. For a freely rotating methoxyl group (cos2 (a) = l/2. If the other alkoxy groups were freely rotating one would expect no change in the @-alkoxy proton coupling constant. Deviations from the methoxy proton splitting constant can, however, be regarded as evidence for a preferred orientation of the alkyl group. These deviations are most easily compared by calculating the ratio of the @-alkoxyproton splitting to the methoxyl splitting in a similar compound ( R = u ~ ~ - o c H ~ ~ / ~ oThis c H ratio ~ ~ ) .is given in Table 111 for the compounds studied in this paper and for two neutral radicals containing alkoxy groups studied by other workers.12J3 As can be seen, the ratio for the P-ethoxy protons is >1.0. This can be compared to the same ratio for @-ethyl protons in a variety of ethyl-substituted aromatic compounds, which is consistently 1.0 implies that the @ pro-
A I 1 I
c Figure 1.
,1
-B
Definition of angles H, @, and $.
TABLE I l l : Calculated R Values for Some Dialkoxy Compounds
Cation Radicals .1.13 1.15
3.242a
3.802 3.636 4.000
2.208b 1.78c
2.655 2.079
1.20 1.17
1.85d 1.82e
1.17 1.20
1,4-Dialkoxybenzene
3.36
1,4-DiaIkoxy-2,5-dimethyl benzene
3.153a
1,4-Dialkoxy-2,5-di-tert-butylbenzene
1,2,4,5-Tetraalkoxybenzene 4,4'-Dial koxybiphenyl
Neutral Radicals 2,6-Dimethyl-4-alkoxyphenoxy 1 .58d 2,6-Di-tert-butyl-4-aikoxyphenoxy
e
a Reference 27. Reference 13.
Reference 9.
1.51e Reference 21
1.23
Reference 12.
tons spend more of their time in configurations for which @ < 45"; for a ratio of 45". The reasons for the different configurations of the ethoxy and ethyl groups are probably steric in origin. Thus, in order to mimimize steric interactions between the methyl group and the ortho ring protons the ethoxy group would be expected to have a minimum energy conformation as shown in Figure 1B ((a = 30"). In the same way for an ethyl group substituted on a benzene ring the minimum energy conformation should be one in which the methyl group is out of the benzene plane, (a = 60" for the @-CH2protons (Figure IC). In both cases temperature-dependent torsional oscillations about the equilibrium position will result in temperature-dependent splitting constants. It should be noted that the temperature coefficients are predicted to be of opposite sign for the two cases. For the ethoxy group with an equilibrium configuration of @ = 30" an increasing splitting with decreasing temperature is predicted, whereas for an ethyl group with an equilibrium configuration of CP = 60" a decreased splitting with decreasing temperature is predicted. Experimentally Vincow, et al.14 715 have found temperature coefficients of +1.80 and +2.10 mG/"C, for the 9-ethylxanthyl radical and the hexaethylbenzene cation radical, respectively, as compared to our results for the ethoxy protons which show negative temperature coefficients. Previous workers have used the temperature dependencies of the @-alkylprotons to evaluate the potential barThe Journal of Physical Chemistry. Voi. 77. No. 15. 7973
1856
Paul D. Sullivan
riers to rotation by fitting the experimental splitting congroups have also been noted and those of the ethyl group stants to calculated values using various models for the (ref 14, 15, 25, 29-32) are of particular relevance to this torsional oscillations.10~14-19All of these calculations have paper. used a symmetric periodic potential function in their calThe hyperfine splitting patterns of the y-ethoxy protons culations. Unfortunately in the case of the ethoxy group lead us to conclude that the CH3 group is freely rotating, the periodic potential is expected to be far from symmetsince no line width effects are observed and the relative ric and it was decided not to pursue further calculations intensities are consistent with the numbers of y-ethoxy since the significance of the barriers obtained would be, a t protons. The magnitude of the y-ethoxy proton splittings best, doubtful. It can be noted, for comparison purposes, are indicated in Table I. The ratio of the p- to y-proton that another series of radicals which are believed to have splittings a g H / a , H is found to be approximately constant the same Q, = 30" equilibrium conformation as the ethoxy a t 24.9 f 1.6. The only other documented example of a group are the n-alkyl radicals (propyl, butyl, etc).20 For splitting from the y protons of an ethoxy group occurs in example, the n-propyl radical has a @-proton splitting the neutral 2,6-dimethyl-4-ethoxyphenoxyradical,l2 where constant of 33.2 G a t -180", an R value of 1.23, and a the ag/a, ratio is found to be 20.5. This almost constant temperature coefficient of - 24 mG/deg, representing a ag/a, ratio is to be contrasted with the same ratio found 0.78%/deg change. This compared with changes of ca. for ethyl groups where considerable variation is found 1.2%/deg for the ethoxyl compounds studied. This result (e.g., for hexaethylbenzenel5 cation radical agla, = 7.16, certainly suggests that the barriers to rotation in the two 'for tetraethylbenzene cation radical32 aB/a, = 34.0, and systems are of the same order of magnitude. for 9-ethylxanthyl14radical agla, = 46.8). Another aspect of our results concerning @-protonsplitThe mechanism leading to the hyperfine splitting of the tings which is worth comment is related to the P-alkylthio y-ethoxy protons is undoubtedly complex. Qualitative protons. The R value for the @-alkylthio protons of considerations of this splitting usually consider spin polarDEDMTB and DEDETB is found to be 3.194/3.457 = ization, hyperconjugation, and homohyperconjugation as 0.92. This compares with R values of 0.83, 0.74, and 0.72 possible mechanisms. In order to obtain quantitative for 1,4.-bis(a1kylthio)benzene~~~ and tetrakis(alky1thio)ethagreement it appears that the best method presently ylene22 radical cations and 4-alkylthio-2,6-di-tert- available is a molecular orbital calculation in the INDO butylphenoxy23 radicals, respectively. Assuming, as has approximation. These calculations were therefore carried been suggested,21,22 that the mechanism of the 0-alkylout in order to further understand the splittings of the pthio splittings is similar to the @-alkoxy splittings (ie., and y-ethoxy protons. hyperconjugative coupling to the spin density on sulfur) INDO Calculations. Space and time limitations of our then, these R values
