A. R. GUPTAAND S. K. SARPAL
500
Nitrogen Isotope Effects in Nickel-Ammonia Complex and Ammonia System
by A. R. Gupta and S. K. Sarpal Atomic Energy Ertablishment Trombay, Chemistry Division, Bombay 88, India
'
(Receiued June 8, 1966)
The nitrogen isotope effect in nickel-ammonia complexes/ammonia equilibrium has been determined, using an ion-exchange breakthrough operation, as a function of resin crosslinking and ammonia concentration. The experimental (over-all) single-stage factors have been calculated from the data using the material balance equation as well as Glueckauf's theory of ion-exchange columns. The good agreement between the calculated and observed breakthrough curves provides an experimental proof of the validity of Glueckauf's theory. The effect of cross-linking on the height equivalent of a theoretical plate and over-all efficiency of the column has been discussed in terms of various parameters like the diffusion coefficients and the effective single-stage factors. The single-stage separaus. ammonia equilibria in the tion factors for Ni(NH3)42+,Ni(NHa)s2+, and Ni(NH3)~2+ aqueous phase have been obtained from an analysis of the data, taking into consideration the equilibrium concentratjons of various nickel-ammonia complexes at different ammonia concentrations; the values are 1.0062, 1.0079, and 1.010, respectively. The singlestage factor for the hexaammine nickel-ammonia system has been calculated from the available vibrational data for this complex. There is good agreement between the calculated and experimentally determined values for the single-stage factor of the Ni(NH&Z+NH3(aq) system.
comparable to the effect observed in NH4+-NHa system and on the other side, for very weak complexes, it will approach unity. For this effect to be experimentally observed, a further requirement is that the above exchange reaction should be rapid. Thus besides a strong interaction between metal cation and ammonia, the complex formed should also be [M(14NH3) 14NH3)]"(as) 15NH3(aq) labile. Many of the very stable transition metal amand Cr(NH3)63+, also happen mines, e.g., CO(NHI)B~+ [ilII(14NH3).-1(15n"3)]"(as) 14NH3(aq) (A) to be inert. Out of the labile ammonia complexes This reaction has an isotope separation factor defined of transition metal cations, nickel, copper, and zinc by2 complexes are quite stable. Nickel-ammonia comC ~ A= (N15/N14)comp~ex/(N16/N14) ammonia = ( ~ / % ) K A plexes have special features, e.g., more pronounced dependence of the equilibrium distribution of various where K A is the equilibrium constant of the reaction A species (complexes with different n values) on ammonia as written, i.e. concentration. For these reasons, we have first selected nickel-ammonia complex for a detailed investigation, {M(14KH3).-~(15NH3)]Z+(aq) 15NH3(aq) K A = so that we can have a better understanding of the { M('4NH3).-1(14NH3) 1I+(aq) 14NH3(aq) theoretical implications of these isotope effects as well The magnitude of this isotope effect varies with the nature of the transition metal ion which determines (1) T. Ishimori, Bull. Chem. SOC.Japan, 3 3 , 520 (1960). the strength of the metal-nitrogen bond formed. On (2) J. Bigeleisen in "Separation of Isotopes," H. London, Ed., the one hand, for very strong complexes, it may be George Newnes Ltd., London, 1961, pp 100-123.
In the chemical equilibrium between transition metal ammines and ammonia in aqueous solutions, a considerable nitrogen isotope effect has been observed' because of the very different chemical environment of the nitrogen atom in the two compounds. The isotope-exchangereaction involved can be written as
+
+
I
The Journal of Physical Chemistry
NITROGEN ISOTOPE EFFECTS IN NICKEL-AMMONIA COMPLEX
as of the special features one encounters in the experimental determination of the same. In principle, the isotope effect for reaction A can be calculated if the normal vibrational frequencies of the various species involved are known. From a survey of the literature on the vibrational spectra of transition metal ammines, it appeared that frequencies corresponding to some of the vibrational modes either have not been observed at all or have not been assigned unequivocally. The existing data, therefore, are not sufficient to calculate these isotope effects exactly. With the help of the available data, however, one can calculate these effects, though not t o a very high degree of accuracy. These results for the hexaammine nickelammonia system are reported here. Whereas Ishimoril depended upon a salting-out technique for the experimental determination of the equilibrium isotope effects in this complex, an ion-exchange column technique has been successfully used here. The ionexchange method is based on the property of transition metal ions that they form stable complexes in the exchanger phase.3 In general, it has been observed that, if the situation is not complicated by any specific interactions between the ligand and the resin matrix or between the metal ion and the resin, the stability of the variety of such complexes formed by the transition metal ions is almost the same as in the solution As similar data on the nickel-ammonia complexes are not available, the stability of these complexes in different ion exchangers vis a vis their stability in aqueous solutions was first investigated. Having confirmed the nature of the nickel-ammonia complex in the resin phase, the isotope effect was determined by a breakthrough operation on ion-exchange columns. The use of the column technique, besides permitting a clean separation between the complex and ammonia, also multiplies the isotope effect. The latter can be of great advantage in those cases where the separation factor is small. The isotope effects have been calculated from the data in two independent ways: (i) the usual application of material balance equation, and (ii) Glueckauf’s’ theory of ion-exchange columns. The effective separation factors have been measured using ion exchangers of 4, 8, and 12% DVB content. In the presence of the ion exchanger, instead of reaction A, the following isotopic exchange reaction takes place 1
~~
[Ki(14NH3)n-1(14NH3) ]RZ+ 4- 15NH3(aq)E
[Ni(14nTH3)n_1(15NH3)]RZ+ + 14NH3(aq) (B) This reaction is additional equilibrium
to the reaction A through the
501
[Ni(14NH3)n-1(14NH3) ]Rz+
+
[Xi(l4NH3)n--1(15NH3) l2 +(as> [Ni(14NH3)7t-l(15NH3) ]RZ+ [Ni(14NH3)%-I(
+ l2
+
(as) (C)
The relation between the three equilibria is K.A = KB/Kc and the separation factor for reaction A is given by aB/ffC
=
CYA
There is no way of directly observing a ~ as , the nickel-ammonia complexes are stable only in an excess of ammonia. The importance of equilibrium C is shown by the varying values of CYB obtained from experiments on ion exchangers of different cross linking. In the absence of directly measured values of ac they have been estimated from a consideration of similar equilibria on these resins. The single-stage factor for reaction A thus obtained has been discussed in the light of the theoretically calculated value and the previously observed value for the same reaction.
Experimental Section Chemicals. Resins used were of sulfonic acid type, viz., Dowex 50 W of different cross-linking (4, 8, and 12% DVB) and different mesh size (20-50 and 100-200) supplied by J. T. Baker Chemicals Go., Phillipsburg, N. J. Nickel chloride was BDH(LR) grade and was purified by filtering its -1 M solution and then passing it through an ion-exchange column. Ammonia used was E. Merck GR grade. Stability of Nickel-Ammonia Complexes in the Ion Exchanger. The stability of nickel-ammonia complexes in the ion-exchanger phase was determined as follows. A stock of Ni form of the resin (20-50 mesh) was prepared in the usual manner in a column and airdried at room temperature (-30”). Its capacity was determined by volumetric estimation of Ni by the KCN method.8 The value was verified by converting the resin into H form and then estimating the H + ions. The values of the capacity obtained by the two methods were in very good agreement. The capacities and the moisture contents of the different resins were as follows: Dowex 50 W-X4, 2.224 mequiv, 40.97%; Dowex (3) R. Nelson and H. F. Walton, J. Phys. Chem., 48,406 (1944). (4) R. H.Stoke and H. F. Walton. J . Am. Chem. Soc... 76,. 3327 (1954).
(5) L. Cockerel1 and H. F. Walton, J . Phys. Chem., 66, 75 (1962). (6) M.G. Suryaraman and H. F. Walton, ibid., 66,78 (1962). (7) E. Glueckauf in “SeDaration of IsotoDes,” H. London, Ed., George Newnes Ltd., London, 1961,pp 2091248. (8) A, 1, Vogel, ‘ D@Y0 DVB) > D(12Oj, DVB). As isotopic equilibrium in the column is achieved by the diffusion of ammonia molecules into and out of the resin particles, one would expect the largest number of plates (or the smallest HETP) in 4% DVB resin column and the least number (or the largest HETP) in 12% DVB resin column, the particle size and flow rate being the same. The data in Table I11 confirm the above expectation. The separation, i.e., the depletion in the front of the band, however, depends on both the number of plates in a column and the effective separation factor. As these two factors vary in opposite directions with an increase in cross linking of the resin, the frontal depletion does not follow a dkfinite pattern. The largest separation has been achieved on resin of 8% DVB content, which shows the importance of considering both the factors, i.e., HETP and Eeff, simultaneously if the aim is to obtain maximum separation. E. E$ect of Ammonia Concentration on Separation The interpretation of the experimental Factors. values of the separation factors is complicated by the presence of many different species in the system. In fact, all the nickel complexes having one to six ammonia molecules coordinated to the nickel ion are present in equilibrium with one another at a particular ammonia concentration. Thus cyspp or aeXprepresents the weighted average of the six independent separation
NITROGEN ISOTOPE EFFECTSIN NICKEL-AMMONIA COMPLEX
507
factors of the equilibria represented by reaction A Table VI: Separation Factors for Different having n values from one to six. The equilibrium conNickel-Ammonia Complexes centrations of the various species change with ammonia concentration and thus their contribution to aexp No. of ammonia also varies. The data in Table IV clearly show that molecules a aeXp increases with ammonia concentration. Ishimori’s 4 I . 00615 value of 1.0076 for CY of this system at -0.3 N ammonia 5 1.0079 concentration is intermediate between the values at 6 1.010 0.15 and 0.40 N reported here and confirms the above trend. This dependence of the separation factor on ammonia concentration, Le., concentration of various three different values of cu,,, for different x4’, x5, and complex species present in the system, implies that the 2 6 , a set of three simultaneous equations is obtained separation factors for equilibria involving individual which can be solved for the three unknowns, e4, e5, complexes are different. These will be denoted by and €6. The results are given in Table VI. all cy2, etc., and (a1 - 1) and (a2- 1) by el, e,etc. The theoretically calculated value of the separation To evaluate the individual contributions of the equifactor is for the hexaammine nickel-ammonia system and libria with the nickel complexes having specific values compares very favorably with the experimentally obof n, one needs to know the equilibrium distribution of served value for the isotopic exchange reaction involving the various species at different ammonia concentrahexaammine nickel ion. tions. Using the stability constants for the nickelammonia complexes given by Bjerrum, et aZ.,22the equiConclusions librium distribution of the various stepwise complexes The above agreement shows that the various apwas calculated for the three different ammonia concenproximations made in the theoretical calculations of trations and is given in Table V. It is obvious that the isotope eff ects-particularly the one proposed by the major contribution to aexpcomes from equilibria Kresge, et aZ.,12 that only those atoms which are where complexes with three, four, five and six ammonia directly bonded to the isotopic atom need be conmolecules are involved. sidered in such calculations-are quite valid. The success of the simple ion-exchange breakthrough technique in the isotope effect determination demonstrates Table V : Equilibrium Composition“ of Nickel-Ammonia the power of this technique in solving such problems. Complexes at Various Concentrations A much needed experimental proof of the validity of Concn Glueckauf’s theory of ion-exchange columns, especially N o . of ammonia molecules of amfor the breakthrough operations, is provided for the monia 1 2 3 4 5 6 first time by the good agreement between the values 33.13 5.5 17.27 40.3 2.16 0.15 0.83 of the single-stage factor obtained from this theory and 48.7 15.7 6.2 28.9 0.39 0.3 0.07 the material balance equation. A simple way of cal23.7 53.0 22.7 0.18 0.32 0.40 0,025 culating single-stage factors and the number of theoa I n mole per rent,. retical plates from the experimental data based on this theory has been found and successfully applied. A more convincing proof of Glueckauf’s theory is proThe concentration of the complex having three amvided by the good agreement between the calculated monia molecules is significant only at 0.15 N ammonia. and observed breakthrough curves. This new appliAs a first approximation, a3 is taken to be equal to cation of Glueckauf’s theory can be of great importance can be written as a4. Then aexp for the industrial chemist interested in large-scale separations on ion-exchange columns. However, the ( a e x p - 1) = ( 2 3 2 4 ) ( a 4 - 1) most significant result which emerges out of this in25(a5 - 1) %(a6 - 1) vestigation is that the isotope effects, in a series of _
_
_
~ ~
+
+
+
or Eexp
=
24‘64
+
25E5
+
where z3,2 4 , etc., represent the mole fractions of the corresponding nickel-ammonia complexes. Using the
stepwise complexes of a metal cation with the same ligand, are different. This implies that the funda(22) J. Bjerrum, G. Schwarzenbach, and L. G. Sillen, “Stability Constants,” Vol. 11, The Chemical Society, London, 1958, p 47.
Volume 71, Number S February 1967
508
KARLHEINZ
mental vibrational frequencies or the force constants are different in these stepwise complexes. It will be interesting to see whether one can experimentally observe such differences in the vibrational spectra of the different nickel-ammonia complexes.
K. BRANDES AND R. J. GERDES
Acknowledgment. The authors wish to express their sincere thanks to Dr. J. Shankar for his interest and encouragement during the course of this investigation. They also wish to acknowledge the help of Mr. K. N. Bhide in the isotopic analysis of the samples.
The Influence of 1,P-Dioxane and Tetrahydrofuran as Solvents upon the Stability and the Solvation of Negative Ions of the Potassium Compounds of Naphthalene and Anthracene
by Karlheinz K. Brandes and R. J. Gerdesl Contribution from the Department of Chemistry, Newberry College, Newberry, South Carolina, and the Georgia Institute of Technology, Atlanta, Georgia 30833 (Received June 10, 1966)
Potassium compounds of naphthalene and anthracene were prepared under high-vacuum conditions (lo-' to lom8torr). Electronic spectra of the solutions of the compounds indicate that dipotassium anthracene is stable in tetrahydrofuran but partly splits into monopotassium anthracene and potassium metal if dissolved in 1,4-dioxane. Dipotassium naphthalene is not soluble in 1,4-dioxane. In tetrahydrofuran as solvent it is completely decomposed into monopotassium naphthalene. The electric conductivities of the potassium compounds of naphthalene and anthracene dissolved in tetrahydrofuran and 1,4dioxane were determined to be about lo6times higher than the corresponding conductivities of the solutions in 1,4-dioxane as solvent. The dissociation energies of the potassium compounds of naphthalene and anthracene in the two solvents used were found to be 2 to 3 kcal/mole in 1,4-dioxaneand - 5 to -7 kcal/mole in tetrahydrofuran, a t room temperature.
Introduction Compounds of aromatic hydrocarbons and alkali metals are now generally agreed to consist of ion-pair complexes with the hydrocarbon as the negative ion. This assumption is mainly based on electronic spectra2-'0 and on esr spectra"-14 of the dissolved compounds. In some cases, however, the published results differed. This was, for instance, the case for a wide absorption band at 23,000 cm-' in the spectrum of potassium naphthalene and several maxima in the spectrum of dipotassium anthra~ene.'~There is alThe Journal of Physical Chemistry
ready some evidence from conductivity measurements16~16 that these differences may be due to the (1) To whom all correspondence s h d d be addressed at Georgia Institute of Technology, Atlanta, Ga. (2) G. J. Hoijtink and J. van Schooten, Rec. Trau. Chim., 71, 1089
(1958). (3) G. J. Hoijtink and J. van Schooten, ibid., 73, 355 (1954). (4) P. Balk, G. J. Hoijtink, and J. W. H. Schreuers, ibid., 7 6 , 813 (1957), ( 5 ) P. Balk, S. de Bruijn; and G. J. Hoijtink, ibid., 7 6 , 907 (1957). (6) G. J. Hoijtink and H. van de Meij, Z . Physik. Chem. (Frankfurt), 20, 1 (1959),
