1593
Communications to t h e Editor found, as would be expected from earlier work on organic iodides,l5-17 that the f3C chemical shifts of ICHzCHzI is strongly dependent upon concentration and solvent. The value included in Table I was obtained on a 0.3 M solution in dioxane; a saturated solution and 0.3 and 0.12 M solutions in cyclohexane gave values of 1.58, 0.64, and 0.53 ppm, respectively, with respect to TMS. In contrast, similar experiments on BrCHZCHzBr and ClCHzCHzCl did not reveal analogously large sensitivities to solvent and concentration. With only two exceptions, the 24 shifts predicted by eq 3 and 4 for the acetates and the ammonium compounds are within about 1.5 ppm of the experimentally determined values. Within the list of 22 successfully predicted shifts for these two classes of compounds, and belonging to both of these classes, is the biologically important case, acetylcholine, CH&OzCHzCH2N+(CH3)3. The fact that the 13C shifts in acetylcholine are so "regular" in this case suggests that any large deviations that might be observed in biological studies could be interpreted in terms of significant alterations of structural detail, e.g., complexation or conformation effects. Table I1 sumimarizes the substituent parameters of eq 3 and 4 for the cy and /3 positions which, according to the
linear least-squares regression, give the best fit to the experimental data. Also included in Table I1 are the individual standard deviations of the fit for the compounds corresponding to each substituent. The overall standard deviation for all of the 13C chemical shifts included 'in this study is 1.62 ppm. The I and C=CH groups remain as the structural units associated with the largest deviations from simple additivity predictions. I t is interesting that neither of these substituents is associated with structural characteristics such as conformational variability about the C-X bond, hydrogen bonding, or ion-dipole interactions, any of which might be expected to manifest themselves much differently from compound to compound. However, these two are probably the most polarizable substituents among the compact structural groups considered. Additivity predictions for compounds containing the other groups covered in this study are likely to be successful within about 1.8 ppm.
Acknowledgment. The authors are grateful to the National Science Foundation for equipment grants for purchase of the spectrometer and data system, and to Dr. Lee M. Huber of Dow Chemical, Midland Division, for pertinent suggestions.
COMMUNICATIONS TO THE EDITOR
Nuclear Magnetic Relaxation of Sodium-23 in Polyphosphate Solutions Publication costs assisted by The University of Leiden
Sir: Nuclear magnetic relaxation of counterions in polyelectrolyte solutions should in principle yield interesting information on polyion-counterion interactions. For example, 23Na relaxation, occurring by way of a quadrupolar mechanism, is determined by the electric field gradient a t the site of the nucleus and the correlation time for this gradient. A study of the relaxation rate of 23Na ions in the presence of negatively charged macroions should therefore contribute to our knowledge of the details of the behavior of counterions in these systems. We wish to report some results obtained on aqueous solutions of sodium polyphosphates (NaPP) with samples of different degree of polymerization (DP). Some sodium phosphates were prepared by heating sodium dihydroorthophosphate in a platinum crucible for 40 hr a t 900". The rlesulting ( N ~ P O Sglass ) ~ is very soluble in water. The polyphosphates of different degree of polymer-
ization were obtained by solubility fractionation of the aqueous solution with acetone by the method of van Wazer.1 Some other polyphosphates were made by the same fractionation method of 10% aqueous solutions of commercially obtained sodium metaphosphate (E. Merck, Darmstadt). All the fractions were freeze dried. The samples contained about 11% water. The DP was found by viscosity measurements in 0.035 N NaBr solutions with an Ostwald viscosimeter .2 The nuclear magnetic relaxation rates were measured at 26" in 15-mm diameter tubes at a frequency of 16 MHz with a Bruker B-KR 302s 16/60 MHz pulsed nmr spectrometer. All the nmr measurements were carried out within a few days after making the solutions of the sodium polyphosphates to exclude the influence of hydrolysis of the polyphosphates.3 The measurements of the longitudinal relaxation time (TI) were performed with a 180"-~90" pulse sequence a t different 7's. For a number of cases (1) J. R. van Wazer, J. Amer. Chem. SOC.,72,647 (1950). (2) U. P. Strauss, E. H. Smith, and P. L. Wineman, J. Amer. Chem. SOC.,75, 3935 (1953). (3)J. B. Gill and S.A. Riaz, J. Chem. SOC.A , 183 (1969). The Journal of Physical Chemistry, Vol. 77, No. 12, 1973
Communicationsto the Editor
1594
800.
€00.
400
-
0
1
2
+POI
E
5
4
3
cow (eq
?
r')
Figure 1. Concentration dependence of 23Na relaxation rate in aqueous NaPP solutions of NaPP sample with DP = 34.
-
RP
Figure 2. Variation of 23Na relaxation rate with DP at several polymer concentrations in aqueous solutions: 0 = 0.089 equiv I.-'; 0 = 0.89equivI.-'; A = 3.50equivI.-'. TABLE I: 23NaRelaxation Rate in Aqueous Solutions of Sodium
Dihydroorthophosphate Concn, equiv i,-1
1-1,
sec-
3.00
51
2.01
39
Concn, equiv I.-1
1 .oo 0.0
(4) P. A. Speight and R. L. Armstrong, Can. J . Phys., 45, 2493 (1967). ( 5 ) M. Eisenstadt and H. L. Friedman, J. Chem. Phys., 44, 1407 (1966); 46, 2182 (1967). (6) C. Hall, R. E. Richards. G. N. Shulz, and R. R. Sharp, Moi. Phys., 16, 529 (1969). (7) H. G. Hertz, G .Stalidis, and H. Versrnold, J. Chim. Phys. Physicochim. Bioi., 177 (1969). (8) G. S. Manning, J. Chem. Phys., 47, 2010 (1967); 51, 924 (1969). (9) F. Oosawa, "Polyelectrolytes," Marcel Dekker, New York, N. Y., 1970. (IO) A. Katchalsky, J. Pure Appi. Chem., 26, 327 (1971). (11) U. Schindewoif and K. F. Bonhoeffer, Z. Eiektrochem., 57, 216 (1953). (12) U. Schindewolf, 2. Phys. Chem. (frankfurt am Main), 1, 134 (1954). D e p a r t m e n t of Physical Chemistry 111 University of Leiden Leiden, The Netherlands
H. S. Kielman J. C. Leyte*
Received December 27, 7972 T-. 1 sec-
28 17.5
(extrap01ated)~ the transverse relaxation time (T2) was measured by the standard spin-echo,technique, and was always found equal to T I within experimental error. The accuracy of all Ti measurements is estimated to be 5%. As a reference point for the investigation of the macromolecular systems, we first determined the 23Na relaxaThe Journal of Physical Chemistry, Vol. 77, No. 12, 1973
tion rate in aqueous NaH2P04 solutions. The results are given in Table I and it is concluded that the longitudinal relaxation time is of the same order as has been found in other sodium salt solutions.4-6 In solutions of NaPP (degree of polymerization: 34) the relaxation rate is increased by about one order of magnitude as may be seen on comparing the results shown in Figure 1 with the experimental values for the NaH2P04 solutions in Table I. Now, according to Hertz, et al.,7 the field gradient a t the nucleus contains contributions from the surrounding water molecules and all other ions. Therefore, if ion condensations-10 occurs an increase in the relaxation rate is to be expected. From the fact that T I and T2 are equal in these solutions it is seen that the correlation time for the field gradient is very short ( < l O - Q sec) and this directly confirms to the conclusion of Schindewolf,llyl2 based on potentiometric and transport results, that the sodium ions do not reside for any appreciable time on given sites on the polyelectrolyte, even though ion condensation occurs. As, in theoretical work on polyelectrolytes, the concept of infinitely long charged rods is used, it is of interest to investigate what degree of polymerization is effectively infinite from the point of view of the counterions or, in other words, a t what chain length do the properties of the counterions become independent of this length. In Figure 2 it is shown that a t a degree of polymerization of about 60 a steep initial increase of TI-1 as a function of the D P levels off sharply. This is observed for concentrations up to 3.50 equiv 1.-1. For solutions of NaPP with a degree of polymerization of 338 we still found the same values for TI-1 as for samples with a D P = 198. Again, these results are in surprisingly good agreement with the work of Schindewolf in view of the difference in the time scales of the methods used.
Effect of 2-Butanol on the Activity of Sodium Sulfate in Aqueous Solutions. Implications for Electrosorption Studies Publication costs assisted by the Air Force Office of Scientific Research
Sir: In studies of the electrosorption of organic compounds on electrodes from aqueous solutions it is customary1 to
