THE IOKIZATIOX CONSTANTS FOR THE LIGAND %,%’,2”-TRIPYRIDISE BY PETERO’D. OFFEXHARTZ,PHILIP GEORGE, AND GILBERT P. HAIGHT,JR. Chemistry Departments, University of Pennsylvania, Philadelphia, Pennsylvania, and Swarthmoie College, Swarthmore, Pennsglvania Receired June 2, 1062 The ionization constants for the ligand 2,2’,2”-tripyridine have been redetermined, using a spectrophotometric method which does not require the direct observation of the extinction coefficients of the singly-protonated forms. The new pK values are significantly larger than those previously reported in the literature. Accurate values of the ionization constants are necessary to obtain the stoichiometry and binding constants of tripyridine with transition metal ions.
Introduction The stability constants of many nitrogenous bidentate ligands with transition metal ions have been determined, including 2,2‘-dipyridine,l 1,lO-phenanthroline,2 and several substituted phenanthrolines. 2,2’,2”-Tripyridine is one of the few tridentate nitrogenous ligands studied thus far. A recent report3 indicated that in moderately strong acid, the bis-(2,2’,2”tripyridine)-iron(I1) complex existed mainly in a doubly protonated form, despite a very high stability. This would seem very unlikely, since there is no evidence that any of the bidentate ligands complex with the ferrous ion in this modification except in concentrated perchloric acid. Since an apparent stoichiometry of the bis complex which includes two protons could be produced by incorrect values of the ionization constant of the ligand, it was decided to remeasure the first two ionization constants of tripyridine, using more accurate techniques and more rigorous methods to analyze the data. Theoretical Since there is good evidence that 2,2’-dipyridine accepts only one proton in up to 0.1 N hydrochloric acid,l it will be assumed that 2,2’,2”-tripyridine can accept two protons only when they are in non-adjacent positions. Thus, the asymmetric doubly-protonated form is assumed to have no appreciable formation in the pH region studied. The symmetric form, abbreviated tripyHz+z, dissociates to form the asymmetric singly-protonated base, abbreviated tripyAH+, which is in tautomeric equilibrium with the symmetric form, tripysH+. Both of the singly-protonated compounds dissociate to form the free base, tripy. Let
KIA =
[tripyI [H 1 [tripy~”] +
Kis
where D is the observed optical density and eo, E ~ A , etc., are the extinction coefficients of the various species. As is readily shown 1
[tripy] = T
If all experiments are carried out a t the same total concentration of tripyridine, then the quantities Do, D I ~Dls, , and Dz can be defined analogously to their respective extinction coefficients as Then
or
where 6
=
KIAKIS (DO- D) KIA Kis [H+l
+
This expression is useful a t low pH values where 6 is small. Also, defining
[tripy1 [H I [tripysH+] +
=
it is easily shown that where [H+]is operationally defined as [H+] = antilog (-pH), and the constants KIA, Kls, and KZare implicKlAKlS ( D - DO) itly concentration constants. Also let KIA f KlS [Ef] ( 2 ) D = ~~[tripyla[tripY~H+] Equation 2 is useful a t higher pH values where 6’ is elg [tripysH +] €2 [ t r i ~ y H z + ~ ] small. D, Do,Dzl and the pH are directly observable. 6 and 6’ can be calculated by re-iteration. The quantities and (KlS/KZ)/(KIA Kls) and (KIAK~s)/(KIA K d T = [tripy] [ t r i p y ~ H + ] [tripysH+] [ t r i p y H ~ + ~ ] can be obtained from the straight line plots of D 6 vs. [H+](Dz- D ) and D - 6’ us. (D - Do)/[H+], ( 1 ) J. H. Baxendale and P. George, Trans. Faraday Soc., 46, 55 (1950). (2) T. S. Lee, I. M. Kolthoff, and D. L. Leusaing, J . Am. Chem. Soc., 70, respectively. 2348 (1948). It is obviously impossible, by spectrophotometric (3) R. B. Martin and J. A. Lissfelt, zbzd., 78, 938 (1956). means, to solve for either KIA, Kls, or Kz without (4) E. 4. IIealy and R. K. Murrnann, zbzd., 79, 5827 (1937).
+
+
+
+
+
+
+
+
Jan., 1963
IOKIZATION CONSTANTS FOR
LIGAND2,2',2"-'~RIPYRI~INB
117
additional information or assumptions. Therefore, it is iiseful to define the operational constants ICl, IC2, and D1
D,,the intercept in the plots of either eq. 1 or 2, is in general dependent, upon temperature and ionic strength. Thus [tripy] = T
1
E H 1
tH +I
k1
klk2
1+-C-
+
so the concentration of free base can always be calculated when the total tripyridine concentration and the pH are known. Experimental A sample of 2,2',2"-tripyridine was obtained from the G. Frederick Smith Chemical Co., and was recrystallized from a 40-60" cut of petroleum ether, removing an insoluble brown oil. The material was further recrystallized from an aqueous methanol solution, in which it was soluble at room temperature but only moderately soluble a t 5'. The resulting crystals were very pale yellow in color. Calcd. for CiSHIiNa: C, 77.23; H,4.75; N, 18.02. Found: G76.96; H,4.85; N, 18.12. Stock solutions of tripyridine were prepared by dissolving several milligrams of the base in hydrochloric acid and diluting with conductivity water; the p H of such solutions was about 3. All experiments were carried out at a total concentration of tripyridine less than molar to prevent precipitation of the free base. Measurements of the pH were made on a Beckman Model GS p H meter, using the expanded scale. The electrodes were standardized with 0.05 M potassium acid phthalate, according to the recommendations of the National Bureau of Standards.5 Buffers were prepared from potassium acid phthalate-sodium hydroxide mixtures in the region around p H 4.5, and from hydrochloric acid-sodium chloride mixtures in the region around pH 3.2. The measured pH values probably are precise t o 0.01 or better. Measurements of the optical density were made on a Beckman Model DU spectrophotometer. Care was taken to ensure that all measurements in a particular run were made a t exactly the same wave length and slit width. Small corrections were made for the absorption of the phthalate buffer. Measurements of the optical density probably are precise to 0.005 or better. The spectra of tripyHz+2and the free base tripy are shown in Fig. 1, along with the calculated values of D1/T a t 25.0' and ionic strength 0.01. The observed extinctions vary with slit width, and so are only valid a t the experimental value of 0.26 mm.
Results pkl was measured a t 25.0 f 0.1' and ionic strengths of 0.01 and 0.05, and, a t the lower ionic strength, a t 13.2 0.2'. pkz was measured a t 25.0 f 0.1' a t ionic strengths 0.001 and 0.01, and, at the lower ionic strength, at 36.6 f 0.1' and 13.2 i 0.2'. All data were subjected to least squares analysis, and the mean errors were computed. The possible inaccuracy in the results is considered to be about 0.02 greater than the mean error in the case of pk2, due to an uncertainty in the absolute value of the pH caused by unknown liquid junction potentials, and about 0.04 greater in the
*
( 5 ) W. J. Hamer, G. D. Pinohing, and S. F. 36, 47 (1916),RP 1690.
Aaree, J . Rea. Nutl.
Bur. Std.,
Fig. 1.-Absorption spectra of 2,2',2"-tripyridine and its acid derivatives. Key: open circles, tripyHz+2; solid circles, tripy ; solid triangles, D1 (see text).
case of pkl, due to both the uncertainty in the pH and uncertainty in 6'. However, the sum of these errors never exceeded 40.07, so both the pk values can be considered accurate within this amount. The data are summarized in Table I, and the results in Table 11. All measurements of pL1 were made at 322 mp, slit width 0.20 mm., and all measurements of pkz were made at 288.5 mp, slit width 0.26 mm. According to Fig. 1, these are the two most favorable wave lengths. TABLE I Pkl PH
D 0.392 .340 ,310 ,267 .240 .218
PH
D
D
PH
4.644 0.494 4.304 0.550 4.29,5 4.817 .424 4.504 ,460 4.496 4.939 .368 4.666 .410 4.656 5.033 .292 4.876 .323 4.86,5 5.125 ,263 4.962 .290 4.952 5.209 I = 0.05 I = 0.01 I = 0.01 t = 25.0 f 0.1' t = 25.0 f 0.1" t = 13.2 f 0.2' X = 322 nip, slit width = 0.20 mrn., DO = 0.066, D2= 0.833 T 5 x 10-6 M
-
Pk2 D
pH
D
pH
0.394 ,370 .349 .326 ,314
D
pH
D
pH'
3.064 0.357 3.064 0.421 3.064 0.409 3.045 3.169 .333 3.169 .393 3.169 .393 3.142 3.284 .314 3.284 .370 3.402 .373 3.252 3.402 .285 3.538 .343 3.538 .355 3.366 3.538 .263 3.677 .341 3.470 .327 3.575 z = 0.002 I = 0.001 I = 0.001 I = 0.01 t = 25.0 i t = 36.6 f t = 13.2 f t = 25.0 f 0.1" 0.2' 0.2' 0.l0 X = 288.5 mp, slit width = 0.26 mm., Do= 0.390, D2= 0.515 T 2.5 X 10" M
-
DANIELCUBICCIOTTI
118
Vol. 67
the calculated values shown in Fig. 1 and Table 11. The disagreement in the pK values probably can be I = 0.05 4.81 0.02 0.584 0.004 traced to the difference in extinction coefficients used. t = 25.0" il preliminary value for plcl of 5.04 a t 25' and I = I = 0.01 4.66 .01 ,654 .004 0.1 was reported previously by the present author^.^ t = 25.0' No correction for the effect of pkz on the data was I = 0.01 4.68 .03 ,694 ,008 made in this earlier value, and the inaccuracy in the t = 13.2" pH determination was much greater. pka It is of interest to note that if 2 K 1 ~= K1s, the statisI = 0.01 3.28 * 02 ,224 ,005 tical ratio of the constants, then, from the expression t = 25.0' ICI = KIAKIS/(KIA Kls), it follows that pKla= I = 0.001 3.27 .05 .192 ,015 4.48 and pKls = 4.18 a t 25' and I = 0.01. The pK t = 25.0" for 2,2'-dipyridine a t this temperature and ionic I = 0.001 3.08 .04 ,188 ,005 strength is 4.30,' so that, as expected, the first ionit = 36.6" zation constants of dipyridine and tripyridine are not I = 0.001 3.46 .05 (0.196 used) much different. A second possible assumption is that t = 13.2' p&A = P K d i p y = 4.30; this gives p x l s = 4.41. Although LI and k2 are not true thermodynamic Discussion constants, it is possible to define AF = -RT In IC = Three other measurements of the ionization conAH - TA8,where AH = -RT2 (d In k/dT). Values stants of 2,2',2''-tripyridine have been made.3~~'' of the pK's, free energies, heats, and entropies of ioniOne of these measurements6 apparently was based on zation for pyridine,8 2,2'-dipyridine,' and 2,2',2"the extrapolation of potentiometric data in partially tripyridine are summarized in Table 111. The more aqueous solutions to pure water, and so cannot be positive entropy change for the ionization of the doublycompared rigorously with the present results. The protonated tripyridine is in keeping with the different pKz was 7.1, compared to the value value of pK1 charge type; that is, in this reaction, an ion with a (for pkl pk2) of about 7.9 a t 25O, I = 0.01, obtained +2 charge dissociates to two +1 ions, whereas each above. of the other ionizations in Table I11 is the dissociation Martin and L i ~ s f e l tusing ,~ an experimental method of a 1 ion to a proton and a neutral compound. similar to the present technique, found pK1 = 4.33, pK2 = 2.64 (both values corrected to I = 0). Their TABLE I11 values for the extinction coefficients of the singlyPK AF AH AS protonated base apparently were assumed rather than (I = 0.01) (kcal.) (kcal.) (e.u.) measured by extrapolation as in the present paper. Pyridine 3.86 -10.8 5.24 7.07 These assumed values are in serious disagreement with 2,2 '-Dipyridine 4.30 5.80 2.0 -13 TABLE I1
ph
R.m.s. error
D1
R.m.s. error
+
+
+
+
( 6 ) W. W. Brandt and J. P. Wright, J . Am. Chern. Soc., 76, 3082 (1954).
(7) P. Offenhartz, G. P. Haight, and P. George, Paper No. 55, Division of Physical Chemistry, American Chemical Society National Meeting, New York, N. Y., 1960.
2,2',2"-Tripyridine, plcl pkn
4.66 3.28
6.36 4.47
0.7 6.8
- 19 4- 8
(8) F. L. Hahn and R. Klockman, 2. physik. Chem., 146, 373 (1930).
THERMODYNAMICS OF LIQUID SOLUTIONS OF BISRIIUTH AND SULFUR' BY DANIELCUBICCIOTTI Stanford Research Institute, Menlo Park, California Received June I S , 1962 The pressure of Szin equilibrium with bismuth-sulfur meltB has been determined a t 705, 740, and 800' over the concentration range from pure bismuth to the composition a t which a second phase formed-solid Bi& a t 705 and 740°, and liquid sulfur a t 800". A transpiration method was used in the lower pressure range and a dewpoint method for the higher pressures. The thermodynamic quantities for solution of sulfur and bismuth in the melt were derived; these are discussed with regard to the possible nature of their binding in the melt. Evidence was found for the gaseous molecule BiS and its heat of dissociation was obtained.
Introduction We have been studying the chemistry of metals dissolved in their molten halides in this Laboratory for some time.2 It seemed of interest to investigate the bismuth-sulfur system since the phase diagram reported in the literature3 indicated that liquid Biz& was completely miscible with liquid bismuth. Our re-investigation of the phase diagram already has (1) This work was made possible by the financial support of the Research Division of the United States Atomic Energy Commission. (2) Consult for earlier references, D . Cubiociotti and F. J. Keneshea, Jr., ,I. Phys. Chem., 6S, 295 (1959). (3) M. Hansen, "Constitution of Binary Alloys," 2nd Ed., McGraw-Hi11 Book Co.. New York, N. Y . , 1958.
been r e p ~ r t e d . ~ The present article describes the chemistry of the system as derived from measurement of the sulfur pressure. Experimental The thermodynamic properties of these bismuth-sulfur melts were characterized by measuring the pressure of sulfur and its temperature coefficient over the liquid composition range. For most of the range, from zero to about 50 atom % sulfur, the partial pressure of sulfur was determined by a transpiration method. At greater sulfur concentrations, the pressure became too large for that method, so a dew-point method was used. Transpiration Method.-A stream of nitrogen was passed through a bulb containing the melt and then through R col(4) D. Cubicoiotti, J . Phys. Chem., 66, 1205 (1062).