CONFORMATIONAL RIGIDITYOF
I
= =
+ km2 (1
2 ) 2
THE
4227
AMIDEBOND
+ Io
+ 2 ():cos0
+ (:)'
+
while that of ketone (m = 51) is taken by fitting it to the experimental intensity data. 10
(9)
I is the absorption intensity of the substituted compound; 0 is the angle between and 2. For the case when > 121,I is proportional to m2. Figure 2 shows a straight line for the two series of both 6-one and 6-thione purine ribonucleosides. The same plot is shown for the charge transfer band of purine-6-thione ribonucleoside derivatives. The intensity of the charge transfer band is seen to be more sensitive toward the substituents. The spectroscopic moment of the substituents are obtained from Petruska for benzene,46
Acknowledgments. The authors are grateful to Dr. Dennis J. Caldwell, Dr. Morris J. Robins and Professor Pete D. Gardner for helpful discussions. This investigation has been supported by the National Institutes of Health Grant GM 12862-02; National Science Foundation Grant G P 6496; Army Ordinance Contract DA-31-124-G618; and the National Cancer Institute of the National Institutes of Health, CA08109-04, U. S. Public Health Service. (46) J. Petruska, J. Chem. Phys., 34,1120 (1961).
Conformational Rigidity of the Amide Bond.
A Variable-Temperature
Nuclear Magnetic Resonance Study of the System Ag+-N,N-Dimethylacetamide by P. A. Temussi, T. Tancredi, and F. Quadrifoglio Istituto Chimico dell' Universitd d i Nap&, Naples, ltaly
(Received May 6, 1969)
Nmr variable-temperature studies of the internal rotation of N,N-dimethylacetamide-da in water and of its complex with AgN03 in water by high resolution nmr spectroscopy gave, respectively, the activation parameters E,, = 21.0 kcal/mol, log A = 13.6, AG* = 19.3 kcal/mol, and E, = 19.0 kcal/mol, log A = 13.9, AG* = 17.9 kcal/mol. These results are discussed in terms of different possible structures of the complex. The data for dimethylacetamide in water are also discussed with respect to recent literature data for organic media.
Introduction Complexes formed by simple amides and inorganic cations have been studied by means of various physicochemical techniques.1-6 Many of these studies have been undertaken with the aim of understanding the influence of salts on conformational transitions of polyaminoacids. A direct interaction of lithium salts with the polymer a,mide groups has been invoked, for instance, to explain the helix-coil transition of poly-Lalanine and poly-L-methionine induced by various lithium salts.' Although considerable evidence has been collected pointing to complexation of cations a t the oxygen of amides in the solid state,Iv5it can not be excluded that complexes with cations linked to nitrogen may be of some importance in solution, at least as transition states. Complexation a t the nitrogen can greatly influence the rigidity of the amide N-C bond, reducing the barrier to rotation. As reported previously,* there are indications that a t least small amounts of a complex of Ag+ linked to the nitrogen of
N,N-dimethylacetamide (DMA) actually exist in aqueous solutions of the system AgN03-DMA. Here we report a complete line shape analysis of the nmr spectra (at various temperatures) of the system AgN03-DMA in aqueous solution which gives the activation parameters connected with rotation around the N-C bond of the amide. For sake of comparison, the isomerization of pure DMA in water has been also studied.
(1) W. E. Bull, 8. K. Madan, and J. E. Willis, Inorg. Chem., 2 , 303 (1963). (2) A.Fratiello andD. P. Miller, Mol. Phgs.,11,37(1966). (3) A. Fratiello, D. P. Miller, and R. Sohuster, ibid., 12, 111 (1967). (4) J. Bello and H. R. Bello, Nature, 190,440 (1961). (5) J. Bello and H. R. Bello, ibid., 194,681 (1962). (6) P.D. Crispin and R. L. Werner, Aust. J. Chem., 20, 2689 (1967). (7) J. S. Franzen, C. Bobik, and J. B. Harry, Biopolymers, 4, 637 (1966). (8) P. A. Temussi and F. Quadrifoglio, Chem. Commun., 844 (1968).
vo.hme 79,Number 18 December 1969
4228
Experimental Section Materials. AgKOI (reagent grade) was purchased from C. Erba and used without further purification. DMA was purchased from C. Erba and distilled under Nz before use. (CH&N(C=O)CDI (DMA-d3) was prepared by reacting acetyl chloride-d3 and dimethylamine.$ The obtained DMA-da was purified by filtration through a basic alumina column. The solutions of the system A~NOI-DMA-D~Owere 4.36 M in and 1.07 M in DMA. The solutions of the system AgN03-DMA-d3-D20 were 4.36 M in AgN03 and 1.04 M in DMA-da. The solutions of the system DMA-d3-D20 were 1.04 M in DMA-d3. D20 was employed instead of H2O only to avoid strong side bands of the solvent. Apparatus and Procedures. The high-resolution nmr spectra were recorded on a Varian A-60-A spectrometer equipped with a V-6040 variable-temperature accessory. Temperatures were determined using methanol and ethylene glycol samples. Temperature readings were taken before and after each run. The samples were allowed at least 10 min to attain thermal equilibrium before spectra were recorded. Nmr Results As mentioned above, evidence has been givens of the existence of an Ag+-DMA complex both in a solution of DMA itself and in aqueous solution. The spectral changes observed at probe temperatures (ca. 40") are consistent with the presence of varying amounts (for increasing Ag+ concentrations) of two types of DMA molecules, bulk and complexed, in rapid equilibrium, characterized by different N-CHI doublet separations. The variable temperature study reported in the present paper has been restricted to aqueous solutions with a ratio of Ag+ to DMA concentrations equal to ca. 4.1. The N-CH, resonance is temperature dependent and can be used to extract exchange rates at different temperatures. It has been amply demonstrated9Jothat the only method capable of furnishing reliable exchange rates from steady-state, high-resolution nmr spectra is the one based on complete line-shape analysis of the spectra. In the case of DMA, a calculation of the N-CH, resonance without any simplifying approximation amounts to the treatment of an eight-site problem, that is if one wants to take explicitly into account the coupling between the C-CH:, and each of the N-CHI groups. Such a calculation, although not difficult to program on a digital computer, is certainly rather time consuming. This is why all couplings have generally been neglected in calculating the exchange rates of most amides. However, as shown by a recent paper,ll the influence on the activation parameters of altogether neglecting these couplings can be quite large. In a preliminary study we have tried to take into account the effect of couplings with the C-CHa group of DMA in an empirical way, by assuming fictitious Tz"values The Journal of Physical Chemistry
P. A. TEMUSSI, T. TANCREDI, AND F. QUADRIFOOLIO for the two N-CH3 lines. For each calculation, the line width in absence of exchange ( ~ / T z " was ) first taken equal to the apparent line width of the envelope of the quartet a t the lowest studied temperature (ie., ignoring that this line width depends mostly on the coupling constant JN-cH~,c-cH~) and it was subsequently allowed to vary by small amounts in order to optimize the fit between calculated and experimental resonance lines. The spectra were calculated by using the GMS formulation, l2 with individual relaxation times for each of the two resonances of the N-CHa groups. Comparison with digitized experimental spectra was performed with the aid of an IBM 1620 computer, as described in previous papers.'**14 The calculated exchange rates are reported in Table I, along with the corresponding AG*'s calculated by
Table I : Temperature Dependence of the Isomerization Rate of Agf-DMA" K, T,OK
sec -1
310 315 326 333 338 349 371 383
1.0 1.8 6.0 17.0 18.0 80.0 1000.0 2000.0
AG*A
~
kael/mol
kcal/mol
18.2 i 0.5 18.1 =!= 0 . 5 17.9 =t 0 . 5 17.8 =!= 0 . 5 17.7 f 0 . 5 17.4 f 0 . 6 17.0 i 0.6 16.7 f 0 . 6
~
*
~
~
18.2 f 0 . 5 18.1 i 0 . 5 17.9 i 0.5 17.7 i 0 . 5 17.9 i 0.5 0.6 17.5 i 16.8 It: 0 . 6 16.8 i 0.6
a Data include corresponding free energies of activation determined by complete line-shape analysis. AG f s l l ~are calculated from points of the Arrhenius straight line; AG*:exptl values are calculated from experimental rates at the same temperature.
means of the Eyring formulation. The AG*'s calculated from points of the straight line determined by least-squares treatment of the Arrhenius equation (Figure 1) are also given for comparison. A 10% confidence limit was taken for the exchange rates on the basis of previous experience.14 An accuracy of *2"K was assumed for all temperatures. All the activation parameters derivable from the Arrhenius plot are reported in Table 11. The unusually'high value of AS*, if real, would point to a mechanism of isomerization qualitatively different from that of the pure amide either in organic solutionsg or in aqueous solution (vide (9) R.C.Neuman, Jr., and V. Jonas, J . Amer. Chem. Soc., 90, 1970 (1968). (10) A. Allerhand, H. 8. Gutowsky, J. Jonas, and R. A. Meinzer, J . Amer. Chem. Soc., 88,3185(1966). (11) A. Pines and M. Rabinovitl;, Tetrahedron Letters, 3529 (1968). (12) G. 8. Johnson, Jr., in "Advances in Magnetic Resonance," J. S. Waugh, Ed., Academic Press, New York, N. Y., 1965,pp 33-101. (13) J. Jonas, A. Allerhand, and H. 8. Gutowsky, J . Chem. Phys., 42, 3396 (1965). (14) P. A. Temussi and T. Tancredi, J . Phys. Chem., 72,3581 (1968).
~
t
l
,
CONFORMATIONAL RIGIDITYOF
THE
4229
AMIDEBOND
Table I1 : Activation Parameters' Tog OK
338 a
A,
5
880-1
x 1017
Ea,kcal/mol
25.2 f 0.5
AH *,koal/mol
AQ
*,kcal/mol
AS*^ eu
17.7 k 0.5
24.6 f 0.5
+20.4
Determined from the temperature dependence of isomerization rate of Ag+-DMA.
Table I11 : Temperature Dependence of the Isomerization Rate of Ag+-DMA-da and of DMA-ds"
IogK..\l. 3.00
T,OK
323 328 331 337 343 349 354 360 366 345 350 355 36 1 362 366 372 378 382 383
K ,880 -1
AG *,l, kcal/mol
A 0 fsxptl,
kcal/mol
Ag +-DMA-da 17.9 f 0.7 17.9 f 0.7 8.0 17.9 f 0.7 10.0 17.9 f 0.7 18.0 17.9 f 0.7 29.0 17.9 f 0.8 45.0 70.0 17.9 f 0.8 100* 0 17.8 f 0.8 170.0 17.8 f 0.8
17.9 f 0.7 17.9 f 0.7 17.9 rrt 0.7 17.9 f 0.7 17.9 f 0.7 17.9 rrt 0.8 17.8 f 0.8 17.9 f 0.8 17.8 f 0.8
DMA-ds 19.4 f 0.9 19.3 =k 0.9 19.3 f 0.9 19.3 f 0.9 19.3 f 0.9 19.3 f 0.9 19.3 f 1 . 0 19.3 f 1.0 19.3 f 1.0 19.3 It 1.0
19.3 f 0.9 19.4 rrt 0.9 19.4 f 0.9 19.4 f 0.9 19.2 f 0.9 19.3 f 0.9 19.4 f 1 . 0 19.3 f 1 . 0 19.2 f 1 . 0 19.3 f 1.0
5.0
4.0 6.0 9.0 14.0 18.0 23.0 30.3 54.5 85.0
78.0
' Includes corresponding free energies of activation, determined by complete line-shape analysis. AG fal's are calculated from points of the Arrhenius straight line; AG*sxptl'~ are calculated
Figure 1. An Arrhenius plot of the isomerization rate of Ag+-DMA as obtained from an nmr complete line-shape analysis.
infra), The documented15 influence of incorrect Tz" values on the activation parameters led us to suspect the AS* value rather to be an artifact, due to the empirical way of accounting for long range couplings. This was actually shown to be the case by the study of the system AgNO3-(CH3)zN(C=O)CD8 in which the influence of long range couplings is almost negligible. In Figure 2 and Figure 3, the spectra of DMA, DMA-d3, Ag+-DMA and Ag+DMA-dS are shown for comparison. The exchange rates for the system Ag+DMA-da are reported in Table 111; the Arrhenius plot and the activation parameters are reported respectively in Figure 4 and Table IV. Since no detailed variable temperature nmr study of DMA itself in aqueous solution existed in the literature, the system DMA-dS in water was also studied in order to be able to judge the relative influence of Ag+ on the rigidity of the N-C bond. The relevant data of the chemical exchange study of DMA-d3 are reported in Tables I11 and IV and in
from experimental rates a t the same temperature.
Figure 4. It must be noted that the somewhat high values of the standard errors in all activation parameters reported in this paper are attributable to the limited number of experimental points. Nevertheless, the uncertainties in the quoted parameters are not so high as to prevent a useful comparison between DRiIA and the complex Ag+-DMA.
Discussion DMA-d3. Before comparing the data of the pure amide and those of the complex, a few words of comment on the parameters of DMA-d3 are in order. The electronic structure of the amide bond is generally describedI6 as a resonance hybrid between the following structures 0-
\ No \ / N-C ++ N+=C / \ / \ A
B Volume 73, Number 1.9
December 1960
4230
P. A. TEMUSSI, T. TANCREDI, AND F. QUADRIFOGLIO
Table IV : Activation Parametersa System
TO,
Ag +-DMA-da DMA-ds a
337 366
A , see-1
Ea, kcal/mol
4 x 1013 8 X 10la
19.0 1 0 . 7 21.0 f 0.9
*,kcal/mol
AQ*, kcal/mol
AS*, eu
18.4 f 0 . 7 20.3 f0.9
17.9 f 0 . 7 19.3 f 0 . 9
11.4 12.7
AH
Data determined from the temperature dependence of isomerization rate of Ag+-DMA-d3 and of DMA-ds.
1'1
DMA
I'* at 3 1 2 O K
10
Hr
-1
I
I
at
DMA-d,
312OK
Figure 2. The experimentally observed N-CHa resonance signals for solutions of DMA and of DMA-ds in DzO.
I n other words a partial delocalization of the C=O double bond on the N-C bond is assumed to be the The Journal of Physical Chemistry
Figure 3. The experimentally observed N-CHa resonance signals for solutions of Ag+-DMA and of Ag+-DMA-da in DSO.
main cause of the rigidity of the amide bond. The relative importance of B should increase when the amide is dissolved in highly polar liquids. The nmr spectral changes in going from organic to aqueous solutions of amides (compare, for instance, Figure 2 of this paper and Figure 1 of ref 9) are accordingly interpreted as an indication of a pronounced increase of the polar character of the amide bond in water. An enhanced conformational rigidity has been generally associated to this increase in polarization. A comparison (Table V) of the activation parameters of DMA-d3 in water with those of neat DMA-da and in dimethyl sulfoxide (16) Z. M. Holubec and J. Jonas, J. Amer. Chem. Soc., 90, 6986 (1968). (16) L. Pauling, "The Nature of the Chemical Bond," Cornel1 University Press, Ithaca, N. Y . ,1960.
423 1
CONFORMATIONAL RIGIDITY OF THE AMIDEBOND Table V: Comparison of Activation Parameters of Various Solutions of DMA-ds sec-1
Ea, koal/mol
6 X lo1*
19.6 20.6 21.0
A,
Solvent
DMA-ds DMSO-da DzO
2 x 1014 8 X 1013
AH
*,kaal/mol
AQ
*,koal/mol
19.0 20.0 20.3
18.2 18.6 19.3
AX*, eu
Ref
$2.7 $4.7 $2.7
9 9 This work
(including that formed by Ag+) in the solid state. A complex of formula
logh
CH3 2.0(
0-Ag
\
/
/
\
"+=e
CH3
NOSCHI
I does not account for the lowering of the barrier. Only complexation of the nitrogen18 or binding of the anion18 1.0c
Ag
CH8\l 2.90
2.70
Nf-C
yo
NOR-
3.10
T-lx103
Arrhenius plots of the isomerization rate of Ag+-DMA-& and of UMA& as obtained from an n m r Ag+-DMA-ds; 0 1 DMA-d3. complete line-shape analysis:
I1
Figure 4.
+,
(DMSO) shows that the increase of rigidity of the amide bond is not very high. By this we do not imply that DMA is not more polarized in water than in DMSO or in DMA itself. The increase in the long range coupling constants between each of the N-CH3 groups and the C-CH3 group clearly speaks in favor of an increased double bond character of the N-C bond. 'It should be concluded that the increase in double bond character is not paralleled by a large increase in the energy difference between the ground state form (the resonance hybrid) and the transition form (very likely structure A). A similar behavior has been recently observed14 also in methyl nitrite whose activation barrier seems to be almost insensitive to the value of dielectric constant of the medium. Ag+-DMA-ds. All relevant activation parameters of the amide are lowered" by the presence of silver ions, as shown in Table IV. The differences between corresponding activation parameters, being almost comparable with the errors, can not be discussed in any quantitative fashion. However they certainly indicate a reduced rigidity of the amide bond which can eventually be interpreted in terms of different possible complexes. Infrared evidence1 points to complexation of the cation at the oxygen for many complexes of DMA
CHs
0-Ag
\
/
N-C-NO3
/
CHI
I
CHa I11
" \I-
0-
/
N+-C-NOa
/
(3%
I
CH3
IV
*
can account for the observed lowering of E,, AH and AG*. It has been shown8 that by increasing the ratio Ag+/DMA, the separation of the N-CH3 doublet is reduced but the chemical exchange rate remains essentially unaltered. That is, increased amounts of Ag+DMA a t 40" do not alter appreciably the rigidity of the amide bond. On the other hand, the temperature dependence of the spectra is greatly affected by relatively small amounts of Ag+. Such a behavior can be interpreted, in our opinion, by means of mechanism similar to that-suggested by Berger, et aZ.,19to explain the hydrogen catalyzed lowering of the barrier of DMA. The most important equilibrium in solution should be (17) The lowering of AQ* is not so marked as reported in ref 8. The discrepancy is due to the fact that in the quoted paper the i\G f ' s had only been estimated on the basis of coalescence temperatures. This once again emphasizes the need of using complete line-shape analysis in extracting exchange rates from steady-state, high-resolution nmr spectra. (18) See W. E. Stewart, Ph.D. Thesis, Florida State University, 1966. (19) A. Berger, A. Loewenstein, and S. Meiboom, J. Amer. Chem. Soc., 81,62 (1959).
Volume 73, Number 18 December 1060
4232
JOHN 6.SMITHAND E. L. PACE
*CHI
CH3
0
0-Ag
\
CH3
/ I
Only small amounts of either 11, 111, or IV are present a t any temperature but can act as transition forms in the isomerization
Ag+N03-
+ *CHa
0
\ / / N-C / \
CH3
Acknowledgment. It is a pleasure to acknowledge the skillful technical help of Mrs. A. Hermann.
The Thermodynamic Properties of Carbon Tetrafluoride from 1 2 ° K to Its Boiling Point.
The Significance of the Parameter v
by John H.Smith' and E. L. Pace Department of Chemistry, Case Western Reserve University, Cleveland, Ohio
(Received M a y 7, 1960)
Low-temperature heat capacity data for carbon tetrafluoride of 99.985 mol % ' purity in the temperature range from 12 to 145.12"K are presented. Values were also derived experimentally for the entropies and temperatures of the solid transition, fusion, and vaporization (1 atm) as follows: A&, = 5.387 cal/(mol O K ) , Tt, = 76.23"K; A& = 1.901 cal/(mol OK), Tr = 89.56"K; A& = 19.457 cal/(mol OK), T, = 145.12"K. The calorimetric entropy for carbon tetrafluoride is found to be 54.04 f 0.13 cal/(mol OK), a value in excellent agreement with the statistical value of 54.037 cal/(mol OK). The results are discussed in terms of the theory of Pople and Karasa. The parameter Y is calculated for carbon tetrafluoride and used to account for the temperatures, entropies, and volume increments for the fusion and transition anomalies.
Introduction '
Carbon tetrafluoride in the solid form is commonly classified2 as a plastic crystal which undergoes an orientationally disordering transition below the melting point due to a globular or spherically symmetric shape of its molecular envelope. Previous low-temperature heat capacity data are given by Eucken, et U Z . , ~ and Kostryukov, et ul.' I n the former work an entropy discrepancy of 0.38 cal/(mol OK) between the calorimetric and spectroscopic entropy is in evidence a t 298°K. This discrepancy is caused to a large extent by use of an incorrect value of the moment of inertia of the molecule. The latter work presents heat capacity data from 12°K to the melting point and offers no comparison between the statistical and calorimetric entropy involved. The purity of the samples used was in one case not stated and in the other slightly over 99%. Therefore, the The Journal of Physical Chemistry
present investigation was undertaken in part to obtain reasonably definitive thermodynamic data on this classic compound. A possible interpretation of the transition and fusion anomalies in carbon tetrafluoride is offered by a theory due to Pople and Karasz5 as an extension of the original works of Lennard-Jones and D e ~ o n s h i r e . ~This ,~ (1) Department of Chemistry, Butler University, Indianapolis, Ind. (2) J. Timmermans, J. Phys. Chem. Solids, 18, 1 (1961). (3) A. Eucken and E. Schroeder, Z.Phys. Chem., B41, 307 (1938). (4) U. N. Kostryukov, 0. P. Samorukov, and P. G. Strelkov, Zh. M z . Khim., 32, 1354 (1958). (5) (a) J. A. Pople and F. E. Karasz;, J. Phys. Chem. Solids, 18, 28 (1961); (b) ibid., 20, 294 (1961). (6) J. E. Lennard-Jones and A. F. Devonshire, Proc. Roy. Soc., A169, 317 (1939). (7) J. E. Lennard-Jones and A. F. Devonshire, ibid., A170, 464 (1939).