Conformational Studies of Chelated Ethylenediamines by Nuclear Magnetic Resonance Paramagnetic Nickel( II) Complexes of N-AIkylethylenediamines Floyd F.-L. Ho1 and Charles N. Reilley Department of Chemistry, University of North Carolina, Chapel Hill, N . C. 27514 Large proton magnetic resonance contact shifts were observed for ethylenediamines coordinated with nickel(l1) ion in aqueous solution. The observed spectra are highly sensitive to the conformational states of the ligand molecule, and considerable structural information can be obtained by investigating the effects of alkyl substitution and of temperature on the spectra. The chelated ligand was found to exist in various gauche forms, with fast conformational exchange between corresponding k and k‘ forms and with slow exchange between configurations that require Ni-N bond rupture for interconversion. The alkyl substituents are shown to prefer the pseudo-equatorial position in the ring; the N-methyl group, for example, was found to be 0.43 kcal/mole more stable thermodynamically at this position than at the pseudo-axial position for the tetraaquo-N,N’-dimethylethylenediaminenickel(l1) ion.
CONSIDERABLE EVIDENCE, based on X-ray and infrared studies in the solid-state, has now been accumulated ( I , 2) to establish convincingly that the ethylenediamine molecule in coordination complexes exists as a puckered five-membered ring and in a gauche configuration; much earlier it had been tacitly assumed that a planar cis configuration was predominant. Conformational energy differences estimated on the basis of a theoretical model proposed by Corey and Bailar (3) also suggest the gauche form to be the preferred configuration. Several attempts to acquire experimental proof of this preference by the nuclear magnetic resonance (NMR) technique (1,4-6) have been reported. However, in the most thoroughly studied example, tris-ethylenediaminecobalt(II1) ion, the methylene protons do not exhibit well refined NMR splitting, which would theoretically be an AzBzsystem in gauche configuration and in the absence of additional splitting from N, Co, and NH protons. A broad, highly overlapping pattern was observed in HzO, D20, DzSOd, and in trifluoroacetic acid, even over a range of temperatures. This had been rationalized as arising either from too small differences in the chemical shifts between the magnetically different protons or too great mobility of the ring in the gauche form. In this report, conformational studies were carried out, for a number of reasons, on various N-alkylethylenediamines coordinated with nickel(I1) ion in D20. Because of the strong magnetic moment of the unpaired electron spins of the paral Present address, Research Center, Hercules, Inc., Wilmington, Del. 19899
(1) A. M. Sargeson in “Transition Metal Chemistry,” R. L. Carlin, Ed., Vol. 3, Marcel Dekker, Inc., New York, N. Y., 1966, p 303. (2) R.D. Gillard and H. M. Irving, Chem. Rec., 65, 603 (1965). (3) E. J. Corey and J. C. Bailar, Jr., J. Amer. Chem. Soc., 81,2620
(1959). (4) D. B. Powell and N. Sheppard, J. Chem. Soc., 1959,791. (5) H. Yoneda and Y. Morimoto, BUN. Chem. Soc. Japan, 39, 2180 (1966). (6) B. M. Fung, J . Amer. Chem. Soc., 59,5788 (1967).
magnetic metal ion, NMR signals of different protons are spread out over large range (150 ppm). In addition, the paramagnetic shift (7, 8) is very sensitive and informative as to the metal-ligand bonding, the geometry of the chelates, and configurational and conformational changes of the ligand molecule. EXPERIMENTAL
Equipment. The NMR spectra were taken at 100 MHz frequency on a Val ;an HA-100 spectrometer, operated in HR mode. Because the contact shifts of the paramagnetic complexes are large and spread out over a few 100-ppm range, the audio-frequency modulation (at 2 KHz) from the integrator unit is undesirable, for it generates sidebands at 1 2 n KHz around the central signal of each kind of proton. Therefore, a Sargent model SR recorder was employed as external recorder and was connected directly to the output on the R-F unit V-43 11. A battery bias of variable potential was incorporated with the Sargent recorder to achieve a large range of recorder zero adjustment. The ac sweep was disconnected in these experiments, and great care was exercised to obtain proper probe tuning and balancing and to compensate for field drift. Spectra were recorded at a sweep rate of about 60 ppm/min. The instrument is equipped with a V-4333 variable temperature probe and a V-6040 temperature controller. The probe temperature was calibrated with a copper-constantan thermocouple. The sample was contained in the usual 5-mm 0.d. sample tube and was equilibrated at each temperature for 15 min before spectra were taken. Generally, for each spectrum, 4 scans, both downfield and upfield sweeps, were recorded at a given temperature. The chemical shifts were measured against the internal reference of the methyl resonance of sodium 3-(trimethylsilyl)-l-propane-sulfonate, abbreviated as TMS*. Precision of chemical-shift measurement is better than 1 %. Spectral calibrations were obtained by measuring the separation of the peaks between chloroform and TMS of the standard sample, which stand 726 Hz apart at the magnetic field strength of the HA100 spectrometer. Chemicals. Ethylenediamine and its N- and C-alkyl substituted derivatives were analytical reagent grade of Aldrich Chemical Company. Ethylenediamine was further redistilled over sodium metal and its derivatives over barium oxide. Anhydrous salt of nickel chloride was purchased from Alfa Inorganics, Inc., and deuterium oxide from Columbia Organic Chemicals. Solutions of ethylenediamines and anhydrous NE12 in 1 :1 ratio were prepared by direct weighing of the stoichiometric amounts into D20, followed by thorough stirring. The acidity of the solution was then measured on a Corning research pH meter equipped with a Fisher microprobe combination electrode. Concentration was 0.400 molar of
(7) D. R. Eaton and W. D. Phillips, A h . Magnetic Resoname, 1, 103 (1965). (8) E. De Boer and H. Van Willigen in “Progress in NMR Spectroscopy,” Vol. 2, J. W. Emsley, J. Feeney, and L. H. Sutcliffe, Eds., Pergamon Press, New York, N. Y., 1967, p 111.
ANALYTICAL CHEMISTRY, VOL. 41, NO. 13, NOVEMBER 1 9 6 9
1835
-90
“
-
8
” -85
-
-60
-
Field 1 n c . d
Figure 2. N M R spectrum of tetraaquoethylenediaminenickel(I1) ion in D2O at 60 “C Ax.
;6.
nickel chloride. The solution thus obtained was nearly neutral, with pD values around 7.6, which was obtained from the meter readings by the equation (9): pD = meter reading
+ 0.40
Signals from right to left are, respectively, the internal TMS* reference, HDO, and the methylene protons of ethylene-diamine. The temperature dependence of the latter is shown in the inset
two electrons, unpaired, one in each of the two higher ep orbitals. In a magnetic field, these electrons will tend to be oriented with their spins pointing in the field direction, giving the ion a measurable magnetic moment. In the bond between the metal and the ligand nitrogen atom, the nitrogen “lone pair” orbital, with two paired electrons in it, points toward one of the lobes of an eo orbital of the central nickel ion, which have the proper symmetry for a-bonding. Thus, a mechanism is provided for the unpaired electron spins to be delocalized onto the ligand skeleton, mainly through the ethylenediamine a-bonding system. Large proton chemical shifts are observed because of the finite spin density reached at the Is orbital of the hydrogens. This is the Fermi contact interaction and can be described by
RESULTS AND DISCUSSION
NMR spectra of the 1 :1 complexes of various substituted ethylenediamines with nickel(I1) chloride in D 2 0 are shown diagramatically in Figure 1. As compared to the diamagnetic free ligands, whose shifts lie in a narrow region between the solvent water and the reference TMS*, the paramagnetic shift of the protons of the complexed ligand is very large. The tetraaquoethylenediaminenickel(I1) ion probably takes a pseudo-octahedral configuration (10, ZI), in which the 8 delectrons on nickel ion are distributed among the 3d orbitals with six electrons paired in the three lower f2g states and with
(9) K. Mikkelsen and S. 0. Nielsen, J . Phys. Chem., 64, 632 (1960). (10) R. S. Milner and L. Pratt, Discussions Faraday Sac., 34, 88 (1962). (11) L. Sacconi in “Transition Metal Chemistry,” R. L. Carlin, Ed., Vol. 4, Marcel Dekker, Inc., New York, N. Y . , 1968, p 199. 1836
where & is the observed contact shift (in ppm) of the ith group of protons and A i is the corresponding nuclear spin-electron spin hyperfine coupling constant (in gauss) and S is the total spin of the paramagnetic complex. Other symbols carry their usual significance(Z2,13). The actual amount of the unpaired electron spin transmitted onto a given proton need not be large and has been successfully correlated with the observed contact shift (7). It is readily seen from Equation 1 that the observed chemical (12) H. M. McConnell and D. B. Chesnut, J . Chem. Phys., 28,107 (1958). (13) R. H. Holm, A. Chakravorty, and G. 0. Dudek, J . Amer. Chem. SOC., 86, 379 (1964).
ANALYTICAL CHEMISTRY, VOL. 41, NO. 13, NOVEMBER 1969
shift should be inversely proportional to temperature with the experimental slope proportional to the hyperfine interaction. Deviations from this dependency have been observed for a few paramagnetic tetrahedral nickel(I1) complexes and extensive studies have revealed that configurational equilibrium existed between tetrahedral and square planar forms, the latter usually being diamagnetic (11). Ethylenediamine. X-ray studies indicate that the ethylenediamine molecule in a coordination complex could exist in two conformations, which were designated to be k and k’ by Corey and Bailar (3). If one views the model from the ligand molecule toward the metal ion along the N-Ni-N plane, these two conformations can be represented by Structure I. d
b
C
k
k’
(I) The temperature dependence of the chemical shift for the coordinated ethylenediamine is presented in Figure 2 , together with an actual spectrum taken at 60 “C. The straight line suggests that a single type of chemical entity is, indeed, under observation and that it is very unlikely that there exists any equilibrium between different configurations (other than k e k’) or different spin states (14). It is also noted from Figure l a that a single signal was observed for all four methylene protons of the coordinated ligand. This could indicate that configuration of the five-membered complex ring is either in the planar cis form or in gauche forms with rapid conformational exchange over equally populated different conformations. The latter seems to be the situation in light of the observed spectra of complexes of the substituted ethylenediamines. N-Methyl and N-Ethyl Ethylenediamines. In the case of unsymmetrical N-alkyl derivatives, the two conformational states, k and k ’ , should not be equally populated. For example, if 1 = alkyl, and 2 , 3, and 4 = H the k form is preferred because the bulky alkyl group prefers the pseudo-equatorial position offered in this form. As the conformational exchange between k and k’ forms is expected to be fast on the NMR scale, only an averaged signal would be observed for a given proton at room temperature. However, the time-averaged chemical shifts for the various individual methylene protons are each different because the weighed average is dominated by the structurally favored conformer, in which the four ring protons are all magnetically different. A spectrum containing four equal area peaks is, therefore, expected. See Figure Ib, c, and f . The spectral assignment for the various protons was reached from direct area comparison, coupled with the fact that the d and b protons are in equatorial positions and, therefore, trans to the metal-nitrogen bond in the preferred conformation. It has been shown recently that the delocalization of the unpaired electron spin on nickel ion to the nucleus of the methylene protons depends on the dihedral angle of the Ni -N and the C--H bonds in the fragment of Ni-N-C-H (1.5-1 7) with a dihedral angle dependence somewhat analogous to that for the coupling constant in the diamagnetic fragments, (14) G. N. LaMar and L. Sacconi, J . Amer. Chem. SOC.,90, 7216 (1968). (15) K. I. Zamaraev, Yu. N. Molin, and G. I. Skubnevskaya, Z h . Srrukt. Khim.,7, 798 (1966); J . Struct. Chem., 7, 740 (1966). (16) R. J. Fitzgerald and R. S. Drago, J. Amer. Chem. SOC.,90, 2523 (1968). (17) L. Pratt and B. B. Smith, Trans. Faraday SOC.,65, 915 (1969).
-10
I
1
I
1
I
H-C-C-H (18), 1g5Pt-N-C-H ( I 9 ) , or I99Hg(20). As a result, the equatorial protons d and b C-C-H are expected to come into resonance at lower fields. Furthermore, the distinction between d and b is made on the fact that N-alkyl substitution would weaken the metal-nitrogen bond (21) and, therefore, reduce the coupling with the adjacent methylene proton. This also receives support from the finding that the 19bPt-H coupling in Pt-N-C-H decreases with increasing N-alkyl substitutions (22). In the diamagnetic cobalt(II1) ( I ) or in the cyclohexane system ( 2 3 , there is a well established empirical axial-equatorial rule, which states that an axial group absorbs at higher field than its equatorial counterpart. The chemical shift difference in such diamagnetic systems is small, ranging only from 0.13 to 0.68 ppm in cyclohexane derivatives and often unresolvable in cobalt(II1) cases. In contrast, the observed chemical-shift difference between the axial and equatorial protons for the paramagnetic complexes of ethylenediamines is large, ranging from 10-150 ppm, as shown in the spectra in Figure 1. Because the spectral splitting for the N-alkyl derivatives of ethylenediamine comes from the conformational preference among different possible conformers, studies of the variation of these chemical shifts over a range of temperatures should (18) A. A. Bothner-By, Adc. A4agneric Resonance, 1, 195 (1965). (19) L. E. Erickson, J. W. McDonald, J. K. Howie, and R. P. Clow, J . Amer. Chem. Soc., 90, 6371 (1968). (20) M. M. Kreevoy and J. F. Schaefer, J . Organometal. Chem., 6, 589 (1966). (21) F. Basolo and R. K. Murmann, J. Amer. Chem. SOC.,74, 2373, 5243 (1952). (22) L. E. Erickson, University of North Carolina, 1969, private
communication. (23) L. Jackman, “Nuclear Magnetic Resonance Spectroscopy,” Pergamon Press, New York, N. Y . , 1959, p 116.
ANALYTICAL CHEMISTRY, VOL. 41, NO. 13, NOVEMBER 1969
1837
-40
cz 7 2 P X k 3
-30 7
T
t
-20
-10
1
300
I
I
1
I
320
340
T
I
L 360
OK
Figure 4. Plot of product of (6 X T ) 6s. T for N,"-dimethylethylenediaminecomplex Lines 1, 2, 3, 4, 5, and 6 are signals from left to right in Figure Id. Line 7 is mean value of lines 1 and 6
methyl of the ethyl group prefers to protrude in a position such that the C-C bond is trans to the metal-nitrogen bond, leaving two methylene protons of the ethyl almost gauche to the metal-nitrogen bond and with slightly different dihedral angles, As temperature is increased, again two opposing effects are operating in determining the chemical shift of these protons, namely an upfield shift according to the Curie behavior expressed in Equation l and a downfield shift because of the more equal chance for the methylene protons of the ethyl group to take up a position trans to the metal-nitrogen bond. The result is the lagging of the observed chemical shift with temperature as observed in lines 3 and 4 in Figure 3 . It is interesting to note that the multiline spectral pattern of these N-alkyl ethylenediamine complexes prevails over the entire range of observed temperatures (32 to 92 "C). This indicates that the Ni-N bonds in these complexes are relatively inert. Had the Ni-N bonds been much more labile, the resulting uncoordinated nitrogen would be free to invert its tetrahedral structure, and the frequent interchanges of the magnetic environments of the ligand protons would produce coalescence of the two lines assigned to a and b, the two lines assigned to c and d , and the two lines assigned to the two -CH2protons of the N-ethyl substituent. From the observed chemical-shift difference, one can estimate the upper limit of the rate of nitrogen inversion at room temperature to be less than 6.2 x lo3 sec-l, calculated from the observed 10 ppm difference for the methylene protons in the ethyl substituent (Figure IC). N,N'-Dimethylethylenedhmine. For this symmetrically disubstituted ligand, two geometric isomers are possible, the optically active dl and the inactive meso forms, as shown in Structures I1 and 111, respectively. c1
provide a useful route to determination of conformational equilibria. It is expected from Equation 1 that the product of contact shift and temperature should be independent of temperature if no conformational exchange occurs or if exchange occurs only among equally populated states. The data for the Ni(I1) complex of ethylenediamine (line 6 in Figure 3) is an example of the latter where the spatial interaction with the coordinated water molecules in the complex is the same for the ethylenediamine in either of its conformations, k or k'. Analogous plots for the N-ethyl ethylenediamine complex are given in Figure 3. For the N-alkyl derivatives the methylene protons at low field are assigned to the protons d and b, which are equatorial in the preferred k conformation (the alkyl group is assumed to be in position 1 in Structure I). These protons show an upfield shift with increasing temperature but at a rate greater than expected solely from Equation 1 ; see Figure 3, lines 1 and 2. Also, the contact shifts of the a and c protons, which are those at the axial position in the preferred form, show an upfield shift with increasing temperature but at a rate slower than expected solely from Equation 1 ; see Figure 3, line 5 . This behavior can be attributed to a shift in the distribution of k and k' conformers with temperature. As the temperature increases, the k' conformation is increasingly populated; thus the d and b protons can more frequently populate the upfield axial positions while the u and c protons can more frequently occupy the downfield equatorial positions. The two methylene protons of the ethyl substituent (lines 3 and 4 in Figure 3 ) also show a behavior similar to a and c protons. From a molecular model, it is easily seen that the 1838
d
A total of six peaks were observed for the ligand molecule in DzOsolution. The spectrum in Figure Id shows the presence of a mixture of dl and meso isomers in equilibrium. The most downfield and the most upfield resonances are attributed to the equatorial b,d protons and the axial a,c protons of the dl isomer, respectively. The dl form is expected to favor the k conformation where both methyl groups occupy the pseudoequatorial positions, and this results in a large chemical-shift difference (114 ppm) between the equatorial and the axial protons of the dl isomer in the coordinated ring. The close doublet at -85.0 ppm is attributed to the four methylene protons of the ring in the meso form. No conformational preference exists because one equatorial and one axial methyl are present in both the k and the k' conformations. The observed small shift arises from the residual differences in the averaged axial and equatorial pairs.
ANALYTICAL CHEMISTRY, VOL. 41, NO. 13, NOVEMBER 1969
The area of the dl signal is 1.4 times more intense than that for the meso protons, at the ambient probe temperature of 32 "C. The same ratio is also found for the methyl peaks at -153.6and -131.7 ppm, which originate from thedlandmeso isomers, respectively. The dl form is favored over the meso isomer because in the former less severe interaction is present between the N-methyl substituents and the coordinated water molecules at the axial apex of the octahedral complex. The free energy difference calculated on this basis is 0.21 =t0.03 kcal/mole (32 "C). An NMR study (24) of the same ligand in the square planar complex, N,N'-dimethylethylenediamineplatinum(II), showed that the dl and meso isomers exist in equilibrium with each other and are present in a nearly 1 :1 ratio; however, for N,N-'ethylenediaminebipyridylplatinum (11), the ratio is about 2 :1 while a 1 :1 ratio was reported for Pt(NH&(EDDA) (25). These data illustrate well the fact that the very small energy differences between configurations (and conformations) in these complexes are not readily calculable and require experimental evaluation. The temperature dependence of the chemical shift of each line was measured, and these data were used in constructing the plot shown in Figure 4. The signals from the meso form (lines 4 and 5) do not show appreciable deviation from Curie behavior as one would expect for its nonpreference for k over k' conformations (Structure 111). The low field peak (line l), attributed to the equatorial protons in the k conformer of the dl isomer (Structure 11) shows a positive deviation from Curie behavior while the upfield axial protons (line 6) show a corresponding negative deviation. The mean value of the upfield and downfield signals (lines 1 and 6) shows normal Curie behavior (line 7). The deviations of lines 1 and 6 reflect the temperature dependence of the conformational equilibrium between k and k' forms of Structure I1 and, from the following analysis, permit evaluation of thermodynamic properties of the system. The observed contact shift for the low field proton (b in Structure 11) can be expressed:
= cfb%
f
fb'Kb')
(1/T)
(2)
where f b andfbt are the mole fractions of the b proton found in the k and k' conformations, respectively, and 6 b and 6b' are the corresponding contact shifts for the b proton in the two conformers. K is the composite constant for the constant quantities given in Equation 1. A similar expression can be written for the upfield signal (proton a in Structure 11): aaa'
= fasa
+
fai6al
(3) After making use of the relationships
and
one obtains
(24) P. Haake and P. C . Turley, J. Amer. Chem. SOC.,90, 2293 (1968). (25) L. E. Erickson, H. L. Fritz, R. J. May, and D. A. Wright, ibid., 91, 2513 (1969).
and (T6bb')
- (TGaa')
= (Kb
- Ka) (2fb - 1)
(5)
An assumption of K, = kbt and Kb = Kat has also been made in writing Equations 4 and 5. This implies that the hyperfine coupling constants for the a and b' protons are identical; both protons occupy axial positions and, therefore, retain nearly the same dihedral angle with the metal-nitrogen bond. An analogous assumption is made for the b and a' protons, which occupy equatorial positions. From Equation 4, it can be seen that the sum of line 1 and 6 in Figure 4 should be independent of temperature, and this is, indeed, verified when one examines at their mean value represented as line 7. The difference of lines 1 and 6 is, however, related to the conformational fraction and, thus, to the equilibrium constant :
The constant-quantity (KO - Ka) in Equation 5 can be expressed from Equation 1 :
Kb - K,
=
T(6, - 6,)
(7)
which corresponds to the situation wherefbl and f a r in Equations 2 and 3 are zero. 6 b and 6, are the contact shifts of the equatorial and axial protons of a given "frozen" conformation-e.g., the k conformer in Structure I1 at the appropriate temperature. In some cases one could perhaps obtain this quantity directly by observing the chemical shifts of 6bb' and 6,,! at temperatures low enough to achieve the condition where only the preferred k conformation exists or where the k,k' conversion rate is sufficiently slow. In this work, the fast rate of conformational exchange and the low temperature limit imposed by the use of aqueous medium prevented the use of this direct approach. On the other hand, we can obtain a useful estimate of the quantity T(& - 6,) by a series of interwoven arguments. The contact shifts, 6 b b ' and 6,,!, for the dl isomer at room temperature are, respectively, - 150.98 and - 36.99 ppm relative to the diamagnetic free ligand. If we deliberately take the difference of these two quantities, - 114 ppm, as ( 6 8 - 6,) at room temperature and use Equations 5 and 6 to calculate the kT's, a very curved line results for a plot of log kT us. (l/T), This results certainly from ignoring the k' contribution present at room temperature. An increase in the k' contribution would lead to a decrease in the observed difference ( 6 b b ' - 6,,~) and vice versa. Thus, we expect that ( 6 , - 6,) is, in fact, larger than the observed ( 6 b b ' - S,,~). In the recorded spectra of 1,2-propylenediamine complex (Figure lh), a larger separation (- 130 ppm) for ring ethylene protons is found. For this complex, only one ethylene proton, 6, is at the downfield equatorial position and two, a and c, are at the upfield axial position in the k conformation of Structure I (where the methyl is at position d). The observed resonance for the a and c protons is at -20.0 pprn from the diamagnetic environment. This is to be expected because the C-methyl substituent of this ligand exists predominantly at the equatorial position in the coordinated ring according to the findings of Sargeson et al. ( I ) . By using this observed separation of - 130 ppm as ( 6 b - 6,) at room temperature and then calculating kT at the other temperatures, a curved line is still obtained for the plot of log k p os. (l/T); the curvature is in the same direction as previously but of less magnitude. This led us to believe that the true value of (6, - 6 0 ) should be substantially larger, and, consequently, the shift for the
ANALYTICAL CHEMISTRY, VOL. 41, NO. 13, NOVEMBER 1969
1839
0.64
A r -0" I
0.60
-- -:-'-I-- (d')ob.
N
0.56
L J 2.7
I
I
I
I
3.1
2.9
I/T
x 103
I
(
d Ilk
I
3.3
Figure 6. Relative free energies of different species of coordinated N,N'dimethylethylenediamine at 32 "C
(OK-+
Figure 5. Log kT us. (1/T) plot for conformational equilibrium of nickel(I1) complex of N,N'-dimethylethylenediamine in dl form
the geometric isomers mem and dl; the conformations k and k' of the latter
AF&c, = 0.86 kcal/mole; A H = 0.71 kcal/mole; A S = -0.56 e.u
comparable to the free energy difference of 0.53 kcal/mole found for trans (axial-equatorial) and cis (diequatorial) isomers of 1,3-dimethylcyclopentane (27). The free energy difference in the five-membered rings is much smaller than in the cyclohexane system. For example, in methylcyclohexane, the conformer with an equatorial methyl group is more stable by 1.7 kcal/mole than that with an axial methyl group. This difference is thought to be caused by a smaller extent of staggering in five-membered rings. In the case of the nickel(I1) complex, the N-Ni-N angle prohibits the ring from assuming a large degree of staggering. The NMR paramagnetic data further permit a qualitative estimate of the degree of staggering for the coordinated fivemembered ethylenediamine ring in solution. If one views the complex along the C-N bond, from the carbon to the nitrogen atom, the picture shown in Structure IV is obtained for the fragment (H2C-N-Ni).
axial protons should appear further upfield and exhibit, therefore, very little contact shift. This also suggested that the Cmethyl group in the 1 :1 propylenediamine complex does not appear exclusively in the equatorial position at room temperature, in contrast to the findings of Sargeson et al. ( I ) . Using the numbers arrived at later, the ratio of axial to equatorial forms of the C-methyl is approximately 0.1 5 at room temperature. In a study of nickel(I1) complex of ethylenediaminetetraacetate (EDTA) (26) where conformation on the coordinated ethylenediamine ring is frozen on the NMR time scale, a signal at -9 ppm (from TMS*) is observed. This signal did not disappear upon deuteration at the elevated temperature with base catalyst and is assigned to the ethylenic protons present in the axial position. If the axial protons of the dlform (Structure 11) are assigned a contact shift value of -6 i 2 ppm (from the 4 H 2 - of the diamagnetic free ligand, which is - 3 ppm from TMS*) the equatorial protons would then be assigned a contact shift value of -188 i 2 ppm. A value of -182 i- 4 ppm is, therefore, assigned to (&, - 6), at room temperature. Using this value, the calculations of kT were repeated, and a good straight line was obtained for log kT us. (1/T) plot (Figure 5). The free energy difference between the k and k' conformers of the dl isomer on this basis is 0.86 i 0.03 kcal/mole at 32 "C, and the enthalpy change is 0.71 + 0.03 kcal/mole. Taking into account the 0.21 f 0.03 kcal/mole free energy difference between the meso and the dl forms at 32 "C found above from the direct area comparison, the complete free energy profile for the coordinated N,N'-dimethylethylenediamine can be constructed (Figure 6). The mole fraction of different conformers of this coordinated ligand are summarized under each conformational structure in Structures I1 and 111. The free energy profile indicates that the free energy change necessary to bring a single N-methyl from the pseudo-equa0.03 torial position to the pseudo-axial position is 0.42 kcal/mole, which must, of course, be doubled to bring both methyls to the pseudo-axial position, as in the k to k' conformation change of the df form. This energy difference is quite
This assumes that the carbon and nitrogen atoms retain normal sp3 configurations. A value of 120" for the angle 0 would indicate the cis planar ring for the coordinated ethylenediamine. The decrease of 0 from 120" represents the increase of degree of ring puckering, with a 60' angle representing a gauche form. As it was pointed out earlier, the contact shift of a given C-H proton depends strongly on the dihedral angle, 0, of the C-H and N-Ni bonds. It is not uncommon to assume (15) that the angular relation of the hyperfine interaction constant of the sp3-electron of the nitrogen with
(26) F. F.-L. Ho, S. Watkins, L. E. Erickson, D. C. Young, J. B. Terrill, and C. N. Reilley, to be submitted for publication.
(27) E. L. Eliel, N. L. Allinger, S. J. Augyal, and G. A. Morrison, "Conformation Analysis," Interscience Division, John Wiley & Sons, Inc., New York, N. Y . , 1965.
*
1840
ANALYTICAL CHEMISTRY, VOL. 41, NO. 13, NOVEMBER 1969
~~
the nucleus of the C-H equation of the form:
proton can be described by an
A* = B,
+ B~cos2 et
(8)
This equation was first proposed by Heller and McConnell(28) for the ?r-electron alkyl radicals. Further experimental evidence for radicals with rigid geometry has led to the conclusion that Bo