formal report

5353

The Molecular Structures and Thermodynamic Functions of 2-Methylbutane and 2,3-Dimethylbutane Richard H. Boyd Contribution from the Department of Chemical Engineering and the Department of Materials Science and Engineering, University of Utah, Salt Lake City, Utah 84112. Received September 16, 1974

Abstract: Previous values of the stabilities of the conformational isomers of 2-methylbutane and 2,3-dimethylbutane as in- ferred from the Raman spectra and the thermodynamic functions of these compounds have not been in accord with confor- mational concepts as expressed by the number of gauche (skew methyl) interactions. Recent Raman studies have removed some previous ambiguities and have resulted in improved values of the conformational isomer stabilities which show that they cannot be accounted for in terms of numbers of gauche interactions alone. Further, the redetermined stability of the 2- methylbutane conformers is not in accord with previous interpretation of the thermodynamic functions. In the present work, we show that the isomer stabilities, the thermodynamic functions, and the conformational energy minimization calculations are all in reasonable mutual accord. It is emphasized that valence angle distortion is important in reducing gauche strain and accounts for the lack of correlation with the number of gauche interactions.

Interest in the interpretation and prediction of the con- formational properties of complex organic molecules and polymers has focused a great deal of attention on the prop- erties of those relatively few simple molecules whose prop- erties have been studied thoroughly experimentally. Ob- viously, methods for property prediction must work well on these “test” molecules if we are to have confidence in pre- dictions on more complex molecules. Two examples of the apparent failure of current qualitative concepts of hydro- carbon structure have been the properties of 2-methylbu- tane and 2,3-dimethylbutane. The series «-butane, 2-meth- ylbutane, and 2,3-dimethylbutane each should have two conformational isomers. The conformers of each molecule differ by one gauche (skew methyl) interaction (see Table I and Figures 1, 2, and 3). Hence, the difference in energy between each isomer pair should, on this basis, be nearly the same. In the case of «-butane, it has been known for some time that both the intensity of the Raman vibrational bands1 and the thermodynamic functions2 (S° and Cp°) are in accord with the gauche isomer being ~800 cal more en- ergetic than the trans. This value along with values3 from «-pentane and «-hexane forms the basis of much of the cur- rent interpretation of hydrocarbon conformational proper- ties. However, the situation with respect to 2-methylbutane and 2,3-dimethylbutane has been perplexing. In earlier work, the Raman spectrum of 2-methylbutane indicated an energy difference of ~100 cal between conformers.4 How- ever, from analysis of the thermodynamic functions (S°, Cp°) Scott et al.5 concluded that the Cs isomer was of much higher energy (at least several kilocalories) than the C\ form. For 2,3-dimethylbutane Szasz and Sheppard4 found no temperature sensitive conformer bands from which it was concluded that either one isomer was of much higher energy or that both existed in equal population (AH = 0). Scott et al.5 concluded from the thermodynamic functions that both conformers have the same energy. Allinger et al.6 on the basis of conformational energy calculations predicted that the isomers should be of nearly equal energy. They pointed out the importance of valence angle distortion in determining conformer stabilities.

The advent of laser Raman spectroscopy has made a much more careful analysis of the spectrum possible. Verma, Murphy, and Bernstein7 have recently restudied the temperature dependent conformer bands in 2-methylbutane and have found such bands in 2,3-dimethylbutane allowing them to assign energy differences between conformational isomers (see Table I). They also have redetermined the en-

ergy difference in «-butane. They find a systematic drop in AH through the series with the isomers of 2,3-dimethylbu- tane being of nearly comparable energy. In confirmation of the latter, they find the ratio of intensities of the two forms to be ~2 to 1 in agreement with the statistical weights. The crystalline phase band corresponds to the less intense liquid band as is consistent with it being due to the more symmet- rical C2/1 form. In summary then, the situation seems to be that for 2-methylbutane the previous interpretation of the thermodynamic functions is not consistent with the new Raman results. For 2,3-dimethylbutane the energy differ- ence between conformers is now unambiguously settled in favor of nearly equally stable forms. This energy difference is anomalously low in the context of gauche interactions but is consistent with conformational energy calculations in which all internal degrees of freedom are allowed to partici- pate.

In view of these new data that have removed the previous experimental ambiguities and the crucial importance of these well-studied molecules as test cases for predictive methods, it now seems appropriate to undertake a unified critical comparison of the relationships among the confor- mer stabilities, thermodynamic functions, and results of conformational energy mimimization calculations. This is the purpose of the present paper. Calculations

Energy minimization calculations were carried out using previously developed algorithms.8,9 The parameters with one exception have been reported earlier.10 The exception is an adjustment to the intrinsic rotational barrier. Recent work with barriers11 had shown that our previous intrinsic barrier is a bit low and in the present work we have in- creased it by 20% from 2.1 to 2.5 kcal/mol. The latter gives a total barrier of 2.8 kcal/mol for ethane. Calculated ener- gy differences reported by Allinger and his coworkers6 using their parameters are also listed.

In order to evaluate the thermodynamic functions, the vi- brational frequencies are required. Calculated values of these are also available from the minimization algorithm.8 The calculated frequencies are listed in Table II along with observed frequencies. The latter are principally those re- ported by Snyder and Schachtschneider.12a'b A few com- ments concerning our calculated frequencies are in order. Snyder and Schachtschneider have shown123·6 that a proper set of transferable force constants leads to excellent agree- ment between calculated and observed frequencies for al-

Boyd / 2-Methylbutane and 2,3-Dimethylbutane

D ow

nl oa

de d

vi a

L A

W R

E N

C E

B E

R K

E L

E Y

N A

T L

L A

B O

R A

T O

R Y

o n

N ov

em be

r 6,

2 01

8 at

0 4:

27 :1

8 (U

T C

).

S ee

h tt

ps :/

/p ub

s. ac

s. or

g/ sh

ar in

gg ui

de li

ne s

fo r

op ti

on s

on h

ow t

o le

gi ti

m at

el y

sh ar

e pu

bl is

he d

ar ti

cl es

.

5354

Table I. Summary of Experimental and Calculated Conformer Energies No. of //(earlier)

gauche (skew Conforma- methyl) in- Thermodynamic AH(recent

tional isomers teractions Oja Raman functions Raman)/ Ah (conf caled.)# (1) (2) (3) (4) (5) (6) (7) (8)

«-Butane Trans 0 1/2 Hob 800d 966 ± 54 675 (730) 670

Gauche 1 2/2 cal/mol cal/mol cal/mol cal/mol 2-Methylbutane c, 1 2/1

Cs 2 1/1 10CK >2000-? 809 ± 50 588 (640)440 2,3-Dimethylbutane C2h 2 1/2

c2 3 2/2 0 or >100 0<? 0<? 54 ± 30 201 (250)80 a Statistical weight (number of stereo isomers divided by rotational symmetry number), b Reference 1. c Reference 4. d Reference 2.

e Reference 5. / Reference 7. S From conformational energy minimization calculations. The first value is from this work, the value in parentheses is the first value corrected for zero-point and vibrational energy (ref 10), the third value is from ref 6.

Figure 1. Calculated structures of conformers of n-butane. In Figures 1-3 torsional angles are underlined, are based on atoms 1, 2, 3, and 4, and are.based on eclipsed as = 0°. In gauche n-butane both torsional angle adjustment (from 60 to 66.0°) and valence angle adjustment (both 1, 2, 3 and 2, 3, 4) contribute to increasing the nonbonded dis- tance (1,4) and reducing methyl—methyl repulsion.

Figure 2. Calculated structures of conformers of 2-methylbutane. In the Ci form, torsional angle adjustment (from 180° to 186.6°) assists in increasing the methyl—methyl (1,5) distance but only one valence angle adjustment (1, 2, 3) can assist. Thus, the methyl—methyl dis- tance is less and the repulsion greater than in gauche-n-butane (see Figure 1). Although torsional adjustment is not possible in the Cs form, methyl-methyl distances (1, 5 and 1, 4) greater than in the Ci form re- sult from valence angle (1, 2, 3) adjustment to 115°.

kanes (about 1% overall). To achieve this agreement, they found it necessary to include interaction force constants, especially between bending and stretching. Our conforma- tional energy force field does not include (valence) interac- tion constants and therefore our overall agreement with the observed frequencies is not as good. However, the largest discrepancies involve principally various C-H bending mo- tions with frequencies above 1000 cm-1. The thermody- namic functions are relatively insensitive to these and our calculated values are quite satisfactory. The thermodynam- ic functions are most sensitive to the low frequency torsion- al motions.

The methyl torsional frequencies are sensitive to differ- ences in nonbonded interactions in different conformations. The Snyder-Schachtschneider force field does not include nonbonded interactions and thus does not accurately reflect the effect of steric interactions on methyl torsional frequen-

Figure 3. Calculated structures of conformers of 2,3-dimethylbutane. In the Cih form, methyl—methyl distances (1,5 and 4, 6) are excep- tionally short since alleviation by torsional adjustment or by valence angle adjustment is not possible. In the C2 form, torsional angle adjust- ment increases distances (1,4) and (5, 6). The otherwise shortened 1, 5 distance is increased by adjustments of valence angles (1, 2, 3) and (2, 3,5).

cies. Since there was little experimental information on such frequencies available to them, this inadequacy was not apparent in the overall accuracy of their calculated frequen- cies. In Table II calculated frequencies for both the Snyder- Schachtschneider force field and ours are compared with the observed frequencies for propane. Experimental values of the methyl torsional frequencies are now available for the latter from neutron diffraction.13 The above mentioned points of the superiority of the (valence) interaction con- stant containing force field in the middle frequency region and the superiority of our nonbonded interaction containing force field for the methyl torsions are illustrated by this molecule.

The calculated frequencies together with the calculated moments of inertia were used to calculate the thermody- namic functions for each conformer. The functions for the torsional vibrations were corrected for anharmonicity using the tables of Pitzer.14 The required barrier heights were cal- culated from the harmonic frequencies and the effective moments of inertia. The conformer functions were then combined to obtain the thermodynamic functions of the equilibrium mixture of conformers by methods previously described.9 The equilibrium mixture calculation requires the enthalpy difference between conformers. The calcula- tion was carried out for both the observed (column 7) and calculated (column 8) AH values of Table I. Both sets of re- sulting thermodynamic functions are listed in Table III. In the case of 2-methylbutane, the value of AH = <=(>2000 cal) proposed by Scott et al.5 is also included. The observed values are shown also. The values for 2-methylbutane and 2,3-dimethylbutane are those tabulated by Scott et al.5 For 2-methylbutane they are based on the experimental values of Scott et al.5 for Cp°(gas) and AHvap° and the values of

Journal of the American Chemical Society / 97:19 / September 17, 1975

Table II. Calculated and Observed Vibrational Frequencies (cm l)a Propane

5355

trans-n-Butane

Caled. Obsd. Caled. Obsd. Caled. Obsd. Caled. Obsd.

A,

a;

2974 (2966)6 2872 (2882) 2855 (2856) 1493 (1471) 1421 (1445) 1404 (1378) 1064 (1151)

851 ( 870) 388 ( 382)

2973 (2964) 1441 (1459) 1200 (1279) 970 ( 903) 212 ( 200)

2965 2875

1473 1449 1385 1157 868 375

1278 899 217 ± 8C

B, 2971 2865 1561 1412 1370 1000

929

B2 2973 2925 1444 1053

774 259

(2963) (2882) (1465) (1367) (1342) (1046) ( 924)

(2965) (2921) (1464) (1185) ( 747) ( 220)

2965 2875 1464 1370 1332 1049 921

2965 2915 1459 1187 748 265 ± 8c

Ag 2973 2872 2859 1593 1465 1412 1403 1080 968 864 394

Au 2973 2922 1442 1216 999 727 224 119

2965 2872 2853 1462 1455

1148 1053 835 427

2965 2920 1455 1257 944 733

Bg

Bu

2973 2926 1444 1203 1038 846 249

2972 2867 2853 1510 1421 1406 1331 991 954 297

2965 2912 1460 1300

2965 2875 2861 1468 1459 1375 1293 1010 965

gauche-n-Butane Cj 2-methylbutane Cs 2-methylbutane

Caled. Obsd. Caled. Obsd. Caled. Obsd. Caled. Obsd. Caled.d Calcd.d

A 2974 B 2973 2974 1400 1351 A’ 2975 298 2972 2971 2973 1345 1337 2971 230 2925 2923 2973 1304 1298 2971 A" 2974 2872 2868 2972 1220 1268 2897 2972 2860 2854 2972 1113 1176 2870 2971 1576 1529 2970 1073 2864 2923 1473 1440 2923 1066 1149 2858 2865 1450 1417 2897 1014 1037 1570 1549 1410 1404 2869 1000 1011 1516 1452 1398 1368 2865 970 969 1462 1438 1210 1282 1207 1233 2864 967 952 1445 1410 1063 1167 1027 1133 2857 941 917 1420 1341 1023 1077 994 1584 904 910 1415 1209 983 981 937 956 1544 827 796 1397 1066 854 835 752 747 1503 758 764 1345 1014 777 789 449 1464 443 459 1110 972 329 325 213 1448 419 1069 940 274 1444 361 368 1010 770 110 1442 292 961 375

1417 1384 255 896 275 1414 1377 227 760 207 1410 1366 212 531 87

93 388

C2n 2,3-dimethylbutane C2 2,3-dimethylbutane Caled.d Caled. Obsd. Caled ß Caled. Obsd. Caled. Obsd. Calcd.d Caled. Obsd.

2974 Au 2972 Bg 2971 Bu 2973 A 2975 931 1410 2973 2970 2968 2972 2973 746 1352 2899 2865 2863 2892 2972 454 1284 1297 2866 1533 1561 2865 2970 343 1082 1168 1571 1456 1439 1496 2897 308 1063 1103 1471 1409 1368 1412 1442 2866 274 1021 1038 1449 1306 1304 1350 1417 1377 1569 234 970 1403 1053 1067 1079 1292 1278 1549 67 940 910 1137 969 956 972 1089 1155 1457 877 835 1069 941 918 938 1019 989 1447 B 2974 537 932 314 401 870 871 1445 2972 412 799 219 207 421 1416 2971 298 481 65 360 1409 2970 242 372 223 1341 1297 2894 209 248 1139 1199 2866

1088 1161 2863 1045 1029 1501

976 954 1548 942 1455

1437 1419

a Observed frequencies are from ref. 12a and 12b except where noted. 6 Calculated values from force field using interaction constants (ref 12b). c Methyl torsional frequency from neutron diffraction (ref 13). d No observed values.

Boyd / 2-Methylbutane and 2,3-Dimethylbutane

5356

Guthrie and Huffman15 for S°(liquid). For 2,3-dimethylbu- tane they are based on the results of Waddington et al.16 for Cp°(gas) and AHnp° and those of Douslin and Huffman17 for S°(liquid). For comparison the calculated and experi- mental18 functions for -butane are also shown.

Discussion From comparison of columns 7 and 8 of Table I, it is ap-

parent that although there is not exact agreement, the con- formational energy minimization calculations reproduce reasonably well the features of conformer stabilities. It is of interest to emphasize why the concept of the number of gauche interactions determining stability fails. Since, in en- ergy minimization calculations, simultaneous adjustments of all of the internal coordinates of the molecule are made and the total energy is the sum of a large number of individ- ual energy functions, it is difficult to ascribe the overall re- sult to any given structural feature. However, simple quali- tative rationales can often be extracted from the details of the calculation. In the present examples, we point out the following. If the skeletal geometries all involved the same bond lengths and valence angles, and the torsional angles were at the exact 60, 180, 300° gauche, trans, gauche' values, the energy function method would be essentially equivalent to counting gauche methyl interactions. The lack of correlation with the latter is the result of relatively mod- est adjustments in the valence and torsional angles. In «- butane (see Figure 1), the methyl-methyl nonbonded inter- actions19 in the gauche conformation result in the minimum energy position of the skeletal torsional angle being dis- placed (in our calculation) from 60 to 66.0° and the valence angles (1, 2, 3 and 2, 3, 4) being increased over the trans to 113.3 from 111.9°. In 2-methylbutane this adjustment is possible for the torsional angle and one of the valence an- gles (1, 2, 3) (valence angle 2, 3, 5 does not increase due to hindrance from methyl group 4) in the C\ form.

The torsional angle adjustment is not possible in the sym- metrical Cs form (see Figure 2). Thus, it might be expected that the gauche-trans difference would be greater than in «-butane rather than less as is observed. However, in the Cs form an exceptionally large adjustment of the valence angle (1, 2, 3) results in an increase in the nonbonded distances and a reduction of the methyl—methyl repulsions between centers 1, 4 and 1, 5 to below that in the C¡ form. The large valence angle change appears to be possible because it alle- viates two simultaneous gauche interactions.

Turning to 2,3-dimethylbutane, we see that in the C2h form no adjustment of the skeletal torsional angle as in gauche «-butane or Cj 2-methylbutane is possible. Further, any adjustment of a valence angle would be hindered by the other methyl substituent on the center carbon (i.e., presence of methyl group 6 hinders changes in angle 1, 2, 3, etc.). Thus, the Czh form is a relatively high-energy conformation for the number of gauche interactions it possesses. In con- trast, in the C2 form adjustment of the skeletal torsional angle reduces two methyl—methyl repulsions (1,4 and 5, 6) and increases one (1, 5). However, simultaneous adjust- ments of the valence angles 1, 2, 3 and 2, 3, 5 are effective in reducing the 1,5 repulsion. Just as in Cs 2-methylbutane, these valence angle adjustments are effective because each alleviates two gauche interactions. Thus, the overall effect of the lack of strain-relieving possibilities in the C2/, form and the presence of them in the C2 form results in the two forms being of nearly equal energy in spite of the greater number of gauche interactions in the latter.

In discussing the thermodynamic functions, we will focus our attention on the entropy, S°, and the heat capacity, Cp°, as they are more or less independently measured quantities.

Table III. Calculated and Observed Thermodynamic Functions0 T - (G° - G0°)/r (H° - H0°)/T S° C 0LP

298.15 58.68 «-Butane6

15.41 74.10 23.36 58.44 15.42 73.86 23.76 58.54 15.58 74.12 23.29

400.0 63.57 18.26 81.84 29.59 63.34 18.35 81.70 29.86 63.51 18.35 81.86 29.60

500.0 67.97 21.13 89.10 35.49 67.76 21.24 89.00 35.67 67.91 21.19 89.10 35.34

600.0 72.07 23.90 96.03 40.58 71.88 24.08 95.97 40.71 72.01 23.98 95.99 40.30

298.15 64.95 2-Methylbutanec

17.62 82.58 28.70 64.90 17.42 82.52 28.56 64.70 17.35 82.05 28.45 64.36 17.75 82.12 28.39

400.0 70.66 21.46 92.12 36.58 70.61 21.47 92.08 36.43 70.33 21.20 91.53 36.41 70.07 21.49 91.56 36.49

500.0 75.88 25.21 101.10 43.77 75.84 25.02 101.07 43.81 75.50 24.97 100.47 43.64 75.28 25.24 100.51 43.71

600.0 80.81 28.81 109.63 49.90 80.77 28.83 109.61 49.93 80.39 28.59 108.98 49.79 80.21 28.83 109.05 49.89

298.15 67.13 2,3-Dimethylbutanetf

20.56 87.70 34.11 67.48 20.25 87.74 34.01 67.58 19.84 87.42 33.59

400.0 73.80 25.25 99.05 43.55 74.08 25.00 99.08 43.51 74.06 24.58 98.64 43.30

500.0 79.93 29.79 109.72 52.01 80.15 29.58 109.54 51.99 80.04 29.22 109.26 51.94

600.0 85.74 34.12 119.87 59.26 85.93 33.95 119.88 59.05 85.77 33.61 119.38 59.23

0 All units are cal/°K/mol. At each temperature the values are calculated for the experimental and the calculated energy differences between conformers (see Table I columns 7 and 8). For 2-methyl- butane, the value AH = °° is also included. The values of AH used are indicated at the 298°K entries. 6 At each temperature row one is at AH = 730, row two is at Ah = 966, and row three is the ob- served values. c At each temperature row one is at AH = 640, row two is at AH = 809, row three is at AH= ==, and row four is the observed values. d At each temperature row one is at AH = 250, row two is at AH = 54, and row three is the observed values.

The entropy, S°, is derived from integrated condensed phase heat capacities down to low temperatures and heats of vaporization. The vapor heat capacity Cp° is indepen- dently measured. The entropy tends to be sensitive to.both the vibrational frequencies and the population of confor- mers while Cp° is sensitive to vibrational frequencies but somewhat less sensitive to conformer population.

Scott et al.5 estimate the uncertainty interval of their vapor-phase Cp° values for 2-methylbutane at ~0.3%, so an overall reliability of 0.1-0.2 cal/°K/mol is probably rea- sonable. For the liquid entropies, the uncertainties for 2- methylbutane and 2,3-dimethylbutane were estimated at ±0.1 and ±0.14 cal/°K/mol, respectively, by the investiga- tors.15·17 The overall uncertainty of the ideal gas entropies is likely to be of the order of 0.2-0.3 cal/°K/mol. Thus, it is seen in Table III that good agreement is obtained between calculated and observed values of S'0 and Cp° for «-butane for both values of AH used. For 2-methylbutane the agree-

Journal of the American Chemical Society / 97:19 j September 17, 1975

ment is good for Cp° for all three of the AH values calculat- ed. For S° the agreement is certainly better for AH = =° in accord with Scott et al.5 However, conversely we are not prepared to say that the discrepancy of 0.5 cal/°K/mol for AH = 600-800 cal/mol is significant in the light of experi- mental uncertainties and uncertainties in the calculated values. For 2,3-dimethylbutane the situation is similar, the Cp° values are in good agreement, and the calculated en- tropies are a bit high but probably not significantly so in view of the uncertainties.

In summary, it appears that the stabilities of the confor- mers of 2-methylbutane and 2,3-dimethylbutane are consis- tent with conformational energy calculations and have a simple qualitative explanation in terms of valence and tor- sional angle adjustments. Further, the stabilities are in rea- sonable accord with the thermodynamic functions.

Acknowledgment. The author is indebted to the U.S. Army Research Office (Durham) for financial support of this work.

References and Notes

(1) G. J. Szasz, N. Sheppard, and D. H. Rank, J. Chem. Phys., 16, 704 (1948).

(2) K. S. Pitzer, J. Chem. Phys., 6, 711 (1940). (3) N. Sheppard and G. J. Szasz, J. Chem. Phys., 17, 86 (1949). (4) G. J. Szasz and N. Sheppard, J. Chem. Phys., 17, 93 (1949).

5357

(5) D. W. Scott, J. P. McCullough, K. D. Williamson, and G. Waddlngton, J. Am. Chem. Soc., 73, 1707 (1951).

(6) N. L. Allinger, J. A. Hlrsch, . A. Miller, I. Tyminski, and F. A. Van- Catledge, J. Am. Chem. Soc., 90, 1199 (1968).

(7) A. L. Verma, W. F. Murphy, and H. J. Bernstein, J. Chem. Phys., 60, 1540(1974).

(8) R. H. Boyd, J. Chem. Phys., 49, 2574 (1968). (9) R. H. Boyd, S. M. Breitling, and M. Mansfield, AlChE J., 19, 1016 (1973).

(10) S. J. Chang, D. McNally, S. Shary-Tehrany, M. J. Hickey, and R. H. Boyd, J. Am. Chem. Soc., 92, 3109 (1970).

(11) K. B. Wiberg and R. H. Boyd, J. Am. Chem. Soc., 94, 8426 (1972). (12) (a) R. G. Snyder and J. H. Schachtschneider, Spectrochim. Acta, 21,

169 (1965); (b) J. H. Schachtschneider and R. G. Snyder, ibid., 19, 117 (1963).

(13) D. M. Grant, R. J. Pugmire, R. C. Livingston, K. A. Strong, H. L. McMurry, and R. M. Brugger, J. Chem. Phys., 52, 4424 (1970).

(14) K. S. Pitzer, "Quantum Chemistry", Prentice-Hall, New York, N.Y., 1952, p 492.

(15) G. B. Guthrie and . M. Huffman, J. Am, Chem. Soc., 65, 1139 (1943). (16) G. Waddlngton, J. C. Smith, D. W. Scott, and . M. Huffman, J. Am.

Chem. Soc., 71, 3902 (1949). (17) D. R. Douslin and . M. Huffman, J. Am. Chem. Soc., 68, 1704 (1946). (18) F. D. Rossini, “Selected Values of Physical and Thermodynamic Proper-

ties of Hydrocarbons and Related Compounds", Carnegie Press, Pitts- burgh, Pa., 1953.

(19) D. H. Wertz and N. L. Allinger have recently (Tetrahedron, 30, 1579 (1974)) proposed that gauche H—H nonbonded Interactions play a domi- nant role In the structure of conformational Isomers. In our parameteri- zation, at least, they play a relatively minor role. In gauche- vs. trans-n- butane for example, we find methyl—methyl nonbonded Interactions contribute 0.50 kcal/mol to the gauche-trans energy difference with other contributions to the total difference of 0.66 kcal/mol from the fol- lowing sources of Me— —0.30, H—H 0.16, torsional angle distortion 0.11, valence angle distortion 0.21, bond length distortion 0.03, and non-bonded Interaction differences on the same side of the center C-C bond —0.05. It also seems clear that most of the torsional and valence angle distortional energies are due to the methyl—methyl repulsions.

A Study on the Interaction of Eu2+(aq) with Pyridinecarboxylic Acids

E. Vrachnou-Astra1 a and D. Katakis*lb Contribution from N.R.C. “Demokritos", Aghia Paraskevi Attikis, Athens, Greece, and Laboratory of Inorganic Chemistry, University of Athens, Athens, Greece. Received September 9, 1974

Abstract: Europous ion forms with isonicotinic, /V-methylisonicotinic, nicotinic, and picolinic acids one to one complexes hav- ing several features, which are rather unusual for a lanthanide ion. They are formed in strongly acidic aqueous solutions and have absorption maxima around 420 nm. The formation constants are 0.15 1. mol-1 for nicotinic acid, 0.2 1. mol-1 for picol- inic acid, 1.9 1. mol-1 for isonicotinic acid, and 0.4 1. mol-1 for /V-methylisonicotinic acid, respectively. Evidence is presented that the complexes involve charge transfer from the metal ion to the ligand. The complexes of nicotinic and picolinic acids are stable toward further redox reaction. The complexes of isonicotinic acid and its /V-methyl derivative, however, undergo further reduction leading in the first case to isonicotinaldehyde and in the second very likely to the dihydro derivative. In the presence of Eu3+(aq) the kinetics of the redox reaction of isonicotinic acid and its /V-methyl derivative are second order in europous ion, first order in the organic acid, first order in hydrogen ion, and inverse first order in Eu3+(aq). A unified mech- anism is proposed to explain the results for both of these acids, which is also consistent with the results obtained on complex formation and with the postulate of a charge transfer from europous ion to the ligand.

The mechanism of electron transfer through reducible organic ligands is related to the mechanism of transfer to such ligands. It must be recalled that even when these lig- ands are bound, the electron is very likely first transferred to them, before finding its way to the central ion.2

If the ligands are bound, the presence of the central metal ion makes it impossible to detect and study some im- portant details of the electron transfer process. In the reac- tions of free ligands3 with low valent metal ions some of these “missing aspects” become more pronounced and can be studied by conventional techniques. Focusing attention on substituted pyridine ligands, it is worth mentioning the following two such aspects, (i) In studies of electron trans-

fer through substituted pyridine ligands4 the electron may “reside” for a while on the ligand. The intermediate radi- cal-complex is, however, difficult to detect. In the corre- sponding reactions of free ligands5 complex formation and subsequent reaction are time-resolved, and the course of the reaction from the precursor complex to the products can be explored more effectively, (ii) Kinetically, the reactions be- tween substituted pyridine complexes and reducing metal ions are generally quite simple. The ligand, whether bound4 or free6 essentially acts as a catalyst. The differences in the overall chemistry caused by changes of the substituents on pyridine are rather trivial.

In the corresponding free ligand reactions the products

Vrachnou-Astra, Katakis / Interaction of Eu2+(aq) with Pyridinecarboxylic Acids