R. L. GUSTAFSON AND A. E. MARTELL
576
VOl. 67
HYDROLYTIC TENDENCIES OF FERRIC CHELATES' BY RICHARD L. GUSTAFSON AND ARTHURE. MARTELL Departments of Chemistry of Clark University, Worcester, Massachusetts, and the Illinois Instilute of Technology, Chicago, Ill. Received August 1, 1968 Potentiometric equilibrium measurements are reported for the hydrolysis and polymerization of Fe( 111) complexes of ethylenediaminetetraacetic acid (EDTA), cyclohexanediaminetetraaceticacid (CDTA), N-hydroxyethylethylenediaminetriacetic acid (HEDTA), and nitrilotriacetic acid (NTA), which have a I. :1 molar ratio of ligand to metal ion. Equilibrium constants are reported for the formation of monohydroxo mononuclear complexes, and for their conversion to p-dihydroxo binuclear chelates. Heats and entropies of reaction for the formation of EDTA and CDTA mono- and binuclear dihydroxo species were calculated from temperature coefficient data. The results are interpreted on the basis of the probable structures of the metal complexes formed. cation of the method outlined by Sweetser and Bricker.'6 DeterIntroduction mination of the amount of free acid in the stock solution was The purpose of the work described in this paper mas carried out on the basis of potentiometric titrations of solutions to extend to the Fe(II1) chelates the studies previously containing equimolar amounts of ferric ion and the disodium salt of EDTA. The difference between the quantity of base recarried out in these Laboratories on the hydrolytic bequired to reach the midpoint of the first inflection corresponding havior of chelate compounds of copper (11),2.3 oxourato formation of the ferric EDTA chelate, and the quantity calcun i ~ r n ( V I ) , ~ Jt h ~ r i u m ( I V ) , ~ ,and ~ J zirc~iiiurn(IV).~~~ lated to neutralize the two moles of acid liberated from the Schwarzenbach and co-workers have reported the first ligand, was a measure of the amount of free acid present. Diand second hydrolysis constants of ferric chelates of sodium EDTA was obtained from Distillation Products Industries, Rochester 3, N. Y. A Sample of CDTA was kindly doethylenediaminetetraacetic acid,lo N,N-ethylenediaminediacetic acid, l1 N-hydroxyethyliminodiacetic nated by the Geigy Chemical Co., Ardsley, N. Y., and samples of HEDTA, NTA, HIMDA, and HXG were supplied through acid,]-l and nitrilotriacetic acid.12 Skochdopole and the courtesy of the Dow Chemical Co., Midland, Mich. EsChaberek13 have studied the hydrolysis of N-hydroxytablishment of purity of the various ligands was carried out on the basis of potentiometric titration of dried samples with standard eth ylethyleiiediaminetriacetato-iron(II1). I n these NaOH, with and without the addition of excess calcium ion. studies, the possibility of formation of polynudear Standard carbonate-free sodium hydroxide was prepared by the ferric chelates was not investigated. Chaberek, et u Z . , ~ ~ usual procedure from a saturated NaOH solution. suggested the formation of an unhydrolyzed binuclear Potentiometric Measurements.-Calculation of the various Fe(II1) chelate of N,N-dihydroxyethylglycine but did equilibrium constants was carried out from data obtained from potentiometric measurements of 1:1ferric chelates over a concennot calculate equilibrium constants for the reactions intration range 8 X 10-8 to 8 X 10-2 M. The hydrogen ion convolved. Recently, Richard, et u Z . , ~ demonstrated the centration waa recorded with a Beckman Model G pH meter existence of a binuclear diolated species containing two fitted with extension glass and calomel electrodes. Measuremoles of 8-hydroxyquinoline-&sulfonate per mole of ments were carried out under a nitrogen atmosphere in a medium 1.00 M in KC1. The details of calibration of the electrode sysferric ion. tem have been described in a previous publication.8 In the The ferric chelates studied iii this iizvestigation are cases of ferric chelates of NTA, HIMDA, and HXG, equilibthose of ethylenediaminetetraacetic acid (EDTA), rium was reached slowly and it was not possible to complete a trans-l,2-diaminocyclohexanetetraaceticacid (CDTA), series of measurements within a single day. In these cases E-hydroxyethyliiinodiacetic acid (HIMDA), dihysolutions containing appropriate amounts of metal, ligand, KCl and NaOH were allowed to equilibrate in sealed glass containers droxyethylglycine (HXG), N-hydroxyethylethylenea t constant temperature for extended periods of time. The pH diaminetriacetic acid (HEDTA), and nitrilotriacetic values of the various solutions were measured at appropriate acid (NTA). intervals varying from one day to several months.
Experimental Reagents.-A 0.3 M solution of Baker's Analyzed Reagent grade Fe(NO& was prepared such that the solution was also 0.1 M in HC1. Standardization was carried out by titration with standard 0.200 M disodium salt of EDTA in the presence of 0.2 M acetate buffer with salicylic acid as an indicator, in a modifi(1) This work was supported by the U. S. Atomic Energy Commission under Contraots AT-(30-1)-1823 (Clark University), and AT-(ll-1)-1020 (Illinois Institute of Technology). (2) R. C. Courtney, R. L. Gustafson, 6 . Chaberek, and A. E. RIartell, J . A m . Chsm. SOP.,81, 519 (1959). (3) R. L. Gustafson and A . E. Martell, ibid., 81, 525 (1959). (4) C. F. Richard, R. L. Gustafson, and A. E. hlartell, zbid., 81, 1033 (1959). (5) R. L. Gustafson, C. I?. Richard, and A. E. Martell, ibid., 82, 1526 (1960). (6) R. F. Bogucki and A. E. RIartell, %bid., 80, 4170 (1958). (7) R. L. Gustafson and A. E. Martell, ibid., 82, 5610 (1960). (8) B. J. Intorre and A. E. Martell, ibzd., 88, 358 (1960). (9) B. J. Intorre and A. E. Martell, zbid., 83, 3618 (1961). (10) G. Schwarzenbach and Heller, Helv. Chim. Acta, 84, 576 (1951). (11) G. Anderegg and G. Schwarzenbach, ibid.. 88, 1940 (1955). (12) G. Schwarzenbaoh and J. Beller. ibid., 84, 183Y (1951). (13) R. Skochdopole and S. Chaberek, J . Iaorg. Nucl. Chsm., 11, 222 (1959). (14) S. Chaberek, R . C . Courtney, and A. E . Martell, J . A m . Chem. Soc., 76, 2185 (1953).
Mathematical Treatment of Data.-The solution equilibria may be expressed in terms of the equations
K,
FeLa - " =+=
Fe [OH]L2-
K,
=
+ H+ $e [OH]L2- "1 [H+] (1) [FeLa - "3
KD 2FeL3 - It =+= (Fe [0H]L)z4 - 2"
KD =
+ 2H+
[(Fe[OH]L)24- z n ] [H+I2 (2) [FeLS-
.r,
Kd (15) P.
=
[(Fe[OH]L)24- z n ] (3) [Fe [OH]L2- "I2
E. Sweetser and C. E. Bricker, Anal. Cham., 26, 195 (1954).
March, 1963
HYDROLYTIC [PENDENCIES
Kb
Fe [OH]L2-
=+:
Fe [OH]2L1- la
Kb
=
2Fe [OHIzL1- " $. 2H+
OF
FERRIC CHELATES
577
+ H+
[Fe [OH]zL1- "1 [H+] (4) [Fe [OH]L2- "1
+(Fe [OH]L)24-
K O=
[(Fe [OH]L)24- 2 + ] (5) [Fe[OHI2L1- n]2[H+]2
Here H,L represents the organic ligand. Thus FeL3 - is an unhydrolyzed metal chelate, Fe[OH]L2 - and Fe[OH]2L1 - " would represent mono- and dihydroxo chelates, respectively, and (Fe[OH]L)e4 - zn indicates a binuclear diolated chelate. Combination of the above equations with the usual material balance and electroneutrality relations gives
a
n
+ I U
(9
0 -I
I
+
where [FeLS - "1 = T M - TOH- [H+] [OH-], TOHis the total concentration of NaOH added beyond the formation of the normal ferric chelate, FeL3 A plot of [ H + ] ( ~ o H 4- [H+l - [OH-])/FeL3 -"I as ordinate us. 2[FeLa -"]/[H+] as absicssa will yield a straight line of slope KO and intercept K, if the binuclear diolated species is the only polynuclear chelate present in appreciable concentration. If the concentration of FeLa -- " is sufficiently small, K , rad K b may be expressed by the relationship
".
TN - [Fe[OH]zL'-"] -- [H+][Fe[OH]zL1--"I
5.0 I
1
I
1
0.4
0
m.
I
I
I .2
0.8
Fig. 1.-Potentiometric titration of 1:1 Fe(II1)-EDTA che, 8.5 X 10-2A!I; . . . .., 5.6 X late. Concentraticns: 10-2J.f; * -,3.0 x 10-2M; . - -, 1.6 X 10-2&f; - - - -, 8.2 X M. m = moles of base added per gram ion of Fe(III), m = 0 corresponds to complete formation of normal Fe(II1)EDTA chelate, t = 25.0°, p = 1.0 (KCl).
- .-
.
--.
3 . 2 1 where Tu is the total concentration of Fe(II1) in all forms. Here a plot similar to that described above will give a slope equal to Kc and an intercept equal to l/Kb. Values of AH0 and AS0 for the various reactions involved were calculated with the usual thermodynamic relationships.
Results
Fe(II1)-EDTA.---Potentiometricmeasurement of a solution containing equimolar quantities of ferric isalt and the disodium salt of EDTA (HzL2-) produces an inflection a t m = 2) followed by a buffer region terminating in a second inflection a t m = 3. Here m is equal to the number of moles of standard NaOH added per mole of metal ion. The first buffer region corresponds to the reaction Fe*+
+ HzL2- =e+ FeLl- + 2H+
whereas the second buffer region corresponds to the reaction FeLl-
=+= Fe [OHILZ-
+ H+
and possibly also to an olation reaction. Potentiometric data obtained over a tenfold concentration range a t 25" are shown in Fig. 1. Here m = 0 corresponda to the unhydrolyzed chelate, FeLl-. Beyond m = 1 precipitation was observed in all solutions in the concentration range studied (8.5 X to 8.2 X 10-2M). A plot of eq. 6 demonstrating the presence of a binuclear diolated ferric EDTA species is shown in Fig. 2 for data obtained a t 25'. Data from experiments carried out a t other temperatures did not exhibit such
0
I .o
20
30
4.0
2CMAI/CH+l xlo-'. Fig. 2.-Plot of data of Fig. 1 illustrating presence of the binuclear olated species FeEDTA [OH]2FeEDTA. Points calculated from data obtained a t the following concentrations: 0 , 8 . 5 X lo-' M; 0,5.6 X lo-' M; 6, 3.0 X lo-' Af; 8 , 1.6 x 10-2 M ; a, 8.2 x 10-3 M.
small deviations as those shown in Fig. 2, although the average deviations of the experimental points from the best straight line corresponded to an error of only =k0.02pH unit. Equilibrium constants obtained a t several temperatures are shown in Table I. Since the value for log Kd is determined by the relationship
R. L. GUSTAFSON AND A. E. MARTELL
578
log Kd = 2pK, - PKD where there is a relatively large uncertainty in the intercept K,, values of log Kdsomet,imesshowdiscrepancies
f
Vol. 67
TABLE I HYDROLYSIS AND OLATION OF Fe(1II)-EDTA oc. 0 4 13.7 25.0 42.4 t,
P&
PKD
log Ed
7.97 7.80 7.58 7.11
12.71 12 37 12.21 12.04
3.23 3.24 2.95 2.18
n
y6
such as that for the value obtained a t 0.4" in Table I. Data obtained a t 13.7, 25.0, and 42.4' gave nearly linear plots of log K us. 1/T and data at these three temperatures were employed in calculating AHo for the various reactions involved. A summary of thermodynamic constants pertaining to the hydrolysis and olation reactioiis of Fe(II1)-EDTA are shown in Table 11. Here it may be seen that the reaction of two moles of monohydroxo chelate to form one mole of a binuclear species proceeds because of a favorable enthalpy change.
i + is
TABLE I1 VALUESOF AFo, AHo, ASD AS0 FOR HYDROLYSIS AND OLATION REACTIONS OF Fe(II1)-EDTA
9'
-
0
27 4
2 U \
A
n
E
0 U
In
AsO (25'),
Y
+
AFO
35 Reaction
+ +
FeL1-* Fe[OH]LZ- H + 2FeL S (Fe[OH]L)z*2H* 2Fe[OH]L2- (Fe[OH]L)z4-
4
I
I
1
1
I
4 8 2CMAI/CH+l x
0
12
Fig. 3.-Plot of data illustrating the formation of the binuclear diolated Fe( 111)-CDTA chelate. Points calculated from data obtained a t the following Concentrations: 0 7.3 X 10+ M ; 0 , 3.8 X 10-2M; c)! 2.0 X 10-2M; @, 8 X 10-3M. t = 25.0", p = 1.0 (KCl).
(25'). ked./ mole
cal./ mole deg.
AHO,
kcal./ mole
-
1-10.3 +10 f 1 2 t16.7 +4.7 f 0.4 -40 - 4 . 0 -15 i3 -36
Fe(I1I)-CDTA.-Potentiometric measurements of equimolar mixtures of ferric ion and CDTA over a tenfold concentration range resulted in a family of curves similar to that shown in Fig. 1. A plot of data according to eq. 8 gave the straight line shown in Fig. 3, indicating that a binuclear diolated species is the predominant polynuclear chelate. The equilibrium constants obtained at the three temperatures studied are given in Table 111. TABLE I11 HYDROLYSIS ASD OLATION OF Fe( 111)-CDTA t , QC.
1.o 25 0 42 3
PK.
PKD
log Kd
9 95 9 32 8 90
18 58 17 62 1 G 92
1.31 1.01 0.89
Thermodynamic constants for reactions 1-3 for Fe(111)-CDTA are presented in Table IV. Bs in the case of the analogous EDTA compound, the driving (Fe[OH]L)24-is a force in the reaction 2Fe[OH]L2favorable enthalpy change whereas the entropy change is -9 e.u. a t 25'.
+
TABLEIV THERMODYKAMIC CONST.4KTS FOR HYDROLYSIS A S D OLATIOX REACTIONS OF Fe(111)-CDTA
Reaction
0.d
3.0
I
I
3.4
I
I
38
I
I
4.2
-LOG CH+l. Fig. 4.-PIot of average number of hydroxo groups bound per mole of metal ion us. -log [H + I for 1:1 Fe(II1)-HEDTA chelates at various concentrations: A, 9.0 X M ; B. 3.6 X M; M; D, 1.0 X 10+ AT. t = 25.0", p = 1.0 (HC1). C, 1.9 X
+
AFQ (25'), kcal./ mole
ASo,
AHo. kcel./ mole
+12 7 $10 0 f 0 2 FeL1- + Fe[OH]Lz- H + +24 0 +16.1 f 0 9 2FeL1-* (Fe[OH]L)s42H 2Fe[OH]L2- e (Fe[OH]L)z4- - 1 . 4 - 3 9 f 0 6
+
Cd./
mole deg.
-
9
-27
+
-
9
Fe(II1)-HEDTA.-Potentiometric measurements of equimolar amounts of Fe(II1) and N-hydroxyethylethylenediaminetriacetic acid (HEDTA), H3L, results in a curve of -log[H+] us. m which has a steep inflection
HYDROLYTIC TENDENCIES OF FERRIC CHELATES
IMarch, 1963
after the addition of four moles of base per mole of metal chelate. This corresponds to the formation of a monohydroxo chelate and corresponding polymerillation products. Silldn,16-.x8in his treatment of polynuclear complexes, has shown that in many cases a plot of 2 = TOH [H+] - [011-] as ordinate us, -log [H+] 8s abscissa produces a family of parallel aurves for potentiometric titration data obtained a t various metal ion concentrations. A plot of -log M'!! us. -log [H+]for data obtained a t constant 2 values then yields a straight line plot, the slope of which is equal tr, t in the general "core plus links" type complex, N(P1[OH]&, As may be seen in Fig. 4, a plot of Z us. -log [H+]for Fe(II1)--HEDTA a t 25' in the region m = 3-4 produces a series of curves which are essentially parallel. The plots of Fig. 5 show that an average value of (3 log TM/b log = 2.14 is obtained, suggesting that a polymer of the general type ML(M [OH]zL)ngn-is obtained. The only polymer which is consistent with the fact that one mole of hydroxide ion is bound per mole of metal chelate is the dimer where n = 1. It should be pointed out that the presence of significant amounts of monohydroxo species in equilibrium with the dimer would tend to give values of ( b log Ti& log [H+])in excess of 2.0. The fact that a value of 2.14 is obtained indicates that the dimer is by far the most predominant hydrolyzed species in solutione in the concentration range studied In Fig. 6, plots of [H+](!!'oR [H+]))[ML] us. 2. [ML]/ [H+]obtained from the same data as those used in Fig. 4 and 5 show a definite drift toward higher intercept values as the concentration of metal chelate increases. I n calculalhg the best straight line through the data, values obtained at the highest concentration (8.9 X 10-2Jl) were not used since these data also did not conform with the straight lines in Fig. 5. E o apparent explanation can be offered for this inconsistency on the basis of the information now available. It had earlier been noted that stock solutions of Fe(II1)HEDTA decomposed on standing, with the result that plots similar to those of Fig. 6 yielded curves which nearly doubled back on themselves a t high [ML]/ [H+] values. This difficulty wa$ eliminated by carrying out each titration with a freshly prepared chelate solution. The equilibrium constants obtained! for the hydrolysis and olation of the Fe(II1)-HEDTA chelate compound are listed in Table V. TABLB V
579
+
+
3.0
34 3.8 -LOG CH+l
4:
Fig. 5.-Plot of -log T X ZJS. -log [ H + ]a t constant 2 based on data shown in Fig. 4 for Fe(II1)-HEDTA chelates.
4.41*
1
3.6
EQUILIBIZIUM CONSTANTS FOR HYDROLYSIS A N D OLATIQX OF Fer: [II)-HEDTA AT 25" IN 1 M KC1 REACTIONS K,
[Fe[OH]L1-] [H+] I FeL]
= ___-
= 10- 4.11
f 0.07
0
0.4
0.8
1.2
I .6
2CMA3~CH+l~10'2. Data obtained in $he buffer region from m = 4-5 were plotted according to eq. 7. Because of the con( 1 6 ) L. G. Sill&
Acta Chem. Seand., 8, 299 (1954), (17) L. G. S i l l h . ibid., 8, 318 (1954). (18) S. Hietanen and L. G SillBn, ibid., 8, 1607 (1954).
Fig. 6.-Plot of data of Fig. 4 illustrating presence of the binuclear Fe(II1)-HEDTA chelate. Concentrations: 0, 3.6 x 10-2M; a,1.9 X 10-2M; 0, 1.0 X 10-2M.
siderable scattw af points, it 'was difficult to determine the intercept, which is equal to 1/Kb. Hence an algcbraio determinatiola af Kb was employed. Combination of eq. 3 , 4 , and 7 leads to the equation
R. L. GTJSTAFSON AND A. E. MARTELL
580 Kb =
~ K ~ [ H + ] ( T o-HTM4- [ H f ] - [OH-]) -1 S= 41 8Kd(22'~ - TOH- [H+] [OH-])
+
+
(8) The average value of pKb obtained for 20 experimental points is included in Table V. The value of K , was obtained from previously calculated constants by the relationship
KO= K D / K , ~ K ~ ~
Vol. 67
Fe(II1)-HIMDA and Fe(II1)-HXG.-Solutions of ferric chelates of N-hydroxyethyliminodiacetic acid (HIMDA) and dihydroxyethylglycine (HXG), containing up to two moles of hydroxide ion per mole of chelate, were allowed to stand €or several months. The recorded p H values showed a continuous drift, resulting in eventual precipitation in the Fe(lI1)-HXG solutions. Mathematical treatment of data obtained within the first few days of equilibration indicated the formation of polynuclear chelates in both cases, although calculation of even approximate values of the equilibrium constants was not possible.
Discussion
+
0 0
3
I
2 2 CM A 11 CH 1x I O-2. +
Fig. 7.-Plot of data illustrating presence of the binuclear diolated Fe( 111)-NTA chelate. Points calculated from data obtained at the following concentrations: 0, 5.3 X M ; 0 , 1.8 X 10-2M; @, 6 X 10-aM. t = 25.0°, = 1.0 (KC1).
-
Fe(II1)-NTA.-Potentiometric measurement of log [H+]us. m for the 1:1 Fe(II1)-NTA chelate system results in a steep inflection, followed by precipitation, after the addition of one mole of hydroxide ion per mole of chelate, indicating the formation of monohydroxo chelate and its corresponding binuclear form. Because equilibrium was reached slowly, the titration procedure was modified by using a number of sealed glass vessels containing equal aliquots of metal chelate and supporting electrolyte to which varying amounts of standard hydroxide were added. Plots of titration data based on the assumption of formation of a monohydroxo chelate and its corresponding dimer are shown in Fig. 7 for samples which were allowed to equilibrate for 1, 2, and 5 days, respectively. The results indicate the probable formation of a binuclear chelate although the following equilibrium constants may be considered to be only approximate 5.0 ~ K= D 6.0 log Kd = 4.0 pKa
=
Upon standing for 2 to 3 months, precipitation, presumably of ferric hydroxide, was observed in all of the solutions employed in the measurements which are summarized in Fig. 7.
The data in Table I provide a new, more precise, concept of the aqueous chemistry of Fe(II1)-EDTA chelate compounds. Interpolation of values of pKa gives a value of 7.68 a t 20' in 1 M KC1, whereas Schwarzenbach and HellerlOobtained avalue of pK, = 7.49 a t 20' in 0.1 M KC1 under conditions where the metal chelate M . Part concentration was approximately 1.5 X of the discrepancy is accounted for by the differences in the supporting electrolyte concentrations and the corresponding changes in activity coefficients. However, the higher value of also was caused by the fact that the presence of polynuclear hydroxo species was overlooked and that all of the reacting hydroxide ion was assumed to be utilized in forming Fe[OH]EDTA. M Fe(II1)Titrations carried out using 1.6 X EDTA in 0.1 M KNOI yielded values of pK, having average deviations of less than 0.01 p H unit over a wide pII range, even though contributions by binuclear species were neglected. Such small deviations would usually indicate that the reaction assumed in the calculatioii procedure was in fact the only one taking place in the experimental solution. Only by increasing the ferric-EDTA concentration markedly was it possible to detect the existence of the binuclear chelate species. It is noteworthy that the tendency of Fe(II1)-CDTA to hydrolyze and polymerize is much less than that of Fe(II1)-EDTA as may be seen by comparison of results in Tables I and 111. This may be due to the fact that the more stable Fe(II1)-CDTA chelate has a weaker affinity for an additional donor group such as a hydroxyl ion than does its EDTA analog; ie., the ferric ion in the CDTA chelate is less acidic than that in the EDTA chelate. This may be illustrated by comparison of the dissociation constants of the two ligands (EDTA: pK1 = 1.99; pKz = 2.67; pK3 = 6.16; pK4 = 10.26in 0.1 M KCl a t 25'. CDTA: pK1 = 2.43; pK2 = 3.52; pKa = 6.12; pK4 = 11.70 in 0.1 in 0.1 M KC1 at 25') where it may be seen that the donor groups of CDTA are considerably more basic than those of EDTA. The difference in dimerization constants also correlates with the difference in hydrolytic tendencies of the Fe(II1)-EDTA and CDTA chelates. Thus the much higher pK of the CDTA chelate (9.32 us. 7.58) indicates a lower affinity of the Fe(II1) ion in this complex for the hydroxyl ion. Since this indicates greater coordination saturation for the Fe(II1) ion when combined to CDTA, one would therefore expect that the hydroxyl ion would also be less effective in forming hydroxo bridges in the presence of this ligand. These two metal chelates therefore provide good examples of the relationships between basicity of the ligand, stability of
HYDROLYTIC TEXDENCIES OF FERRIC CHELATES
March, 1963
the chelate, tendency to form hydroxo mononucleax derivatives of the chelate, and the tendency to form binuclear chelates through hydroxo bridges (olation) . It is of interest t o compare the thermodynamics of dimerization of Fe(II1)-EDTA and Fe(II1)-CDTA chelates with the results obtained for dimerization of a number of Cu(I1) alielates of substituted diamii~es.~ The summary in Table VI indicates that in the cases of the Cu(I1) chelates of DMEN, TMEN. and HEN, the reaction
58 1
i"
/CH2-cH2
/N\
2
y z
CHz
\"
/cu\oH 'CH2-CH2 I
/ CHzCHzOH
/
H
0
/\ \
2Cu[OH]L(H20) =-I LCU
/cuL+
2Hz0
I1
0 H
TABLEVI THERYODYNAWC CONSTANTS FOR REACTION
2M[OH]L S (M[OH]L]z AHO,
Metal
Ligand
kcal./mole
AS0 ( 2 5 ' ) , oal./mole deg.
DMEN"." +1.0 $21 Cu(I1) CujlI) TMEN"' +3 4-28 Cu(I1) HENC,' +4 23 2-HENd*' -1 + 3 Cu(I1) Fe( 111) EDTA/ - 15 - 36 Fe(111) CDTL4! -3.9 - 9 a N,N'-Dimethylethyleiiediamine; N,N,N',N'-TetramethylN,N'ethylenediamine; N-€1ydroxyethylethylenediamine; Dihydroxyethylethylenediamine; e I n 0.1 M "08; In 0.1 M KC1.
+
111 /CHz-CHz,
'
proceeds because of a, large favorable entropy change produced by the incre,ase in translational entropy upon release of the coordinated water molecules when dimerization occurs. In the cases of Fe(II1)-EDTA and Fe(II1)-CDTA chelates, where there probably are .no coordinated water molecules involved in the reactions, negative entropy changes are observed. It might be expected that if the dimerization reaction occurs between I and I1 rather than between I11 and. IV, a negative value of AXo would be observed. The fact that a slightly positive value of AXo is obtained for dimerization of the hydrolyzed Cu(II)-2-HEN chelate suggests that there may be a contribution by both reactions and that the equilibrium between I and I11 favors the former structure. No explanation is alpparent for the fact that the entropy change associated with the dimerization of Fe(111)-CDTA is more favorable than that of the analogous EDTA chelate. The hydrolysis and dimerization constants shown in Table V for the Fe(II1)-HEDTA chelate indicate that although the tendency toward hydrolysis is considerably greater than in the case of the analogous Fe(II1)-EDTA chelate, the tendency to polymerize is somewhat less. The low value of pK, suggests that binding probably takes place through the hydroxyethyl group and that the hydrolysis reaction is the conversion of V to VI. The binuclear complex may also involve bridging by alkoxide groups of the ligand, as is indicated by VI1 and VIII. On the basis of evidence now available, it is, of course, impossible to distinguish between the alterna-
IV
tive arraiigemeiits in I and 111,V and VI, and VI1 and VIII. The reactiom of Fe(II1)-HEDTA are similar to those studied previously in the cases of Cu(I1)-diamine syst e m where ~ ~ it has been shown that further hydrolysis of binuclear olated chelates results in depolymerization and the formation of monodentate dihydroxo complexes at elevated pH. The degree of polymerization of the hydroxo derivative of the 1: 1Fe(II1)-NTA chelate is greater than that of any of the ferric chelates described above. This is undoubtedly due to the fact that only four coordination positions are occupied by donor groups of NTA, leaving two aquated coordination positions available fol hydrolysis and olation.
Po\,
C,HZ
I
582
YUKITO
hIURAK.4MI
AND
ARTHURE. MARTELL
'Vol. 67
0
CHZ
\
\ co/o VI11
VI1
PH2
co
CATALYTIC HYDROLYSIS OF SALICYL PHOSPHATE IS THE PRESENCE OF COPPER(I1) CHELA4TES1 BY YUKITOMURAKAMI~" AKD ARTHURE. MART ELL^^ Departments of Chemistry, Clark University, Worcester, Massachusetts, and Illinois, Institute of Technology, Chicago, Illinois Received August 1, 196?2 Rates of hydrolysis of salicyl phosphate (SP) in the present investigation generally followed first-order kinetics. Studies of the catalytic hydrolysis of SP a t 30.0" were carried out a t an ionic strength of 0.100 M . The 1 : 1 hydroxyethylethylenediamine-Cu(I1) (HEN-Cu( 11)) catalyzed hydrolysis of SP was carried out in the middle pH range (;log[H*] = 4.00, 5.50, 6.00, 6.50, and 7.00). The increased hydrolysis rate is attributed to interaction of the diionic and the triionic species of the substrate with Cu(I1) ion. The Cu(I1)-dipyridyl system was found to be inactive as a catalyst in the hydrolysis of SP. The interactions of salicyl phosphate with Cu(I1) ion an6 the dipyridyl-Cu(I1) chelate were studied by means of potentiometric measurements, and the formation constants of the Cu(I1)-SP chelate and the Cu(I1)-dipyridyl-SP chelate are reported.
A qualitative investigation on the catalytic hydrolysis of salicyl phosphate (SP) has been carried out recently3 with a series of copper chelate and vanadyl chelate compounds. I n order to explore further the mechanism for the catalytic hydrolysis of SP, more extensive study on the hydrolysis Catalyzed by Cu(I1)hydroxyethylethylenediamine (HEX) and dipyridyl (DIPY) has been performed. Experimental Salicyl phosphate (0-carboxyphenylphosphate, SP) was purchased from the California Foundation for Biochemical Research, Los Angeles, California. The purity of the compound, which was established by phosphorus analysis and potentiometric titration, was sufficiently high so that further purification was not necessary. N-Hydroxyethylthylenediamine N), purchased from as the dihydrochloride the Eastman Kodak Co., was is01 after two recrystallizations from methanol solution. ot,ot'-Dipyridyl, obtained from the S.A.F. Hoffman-La Roche and Co., Ltd., Basel, Switzerland, was used without further purification. Copper(I1) solution was prepared from its nitrate salt and was standardized by titration with standard disodium salt of ethylenediaminetetraacetic acid, with murexide as an i n d i ~ a t o r . ~ The procedures employed for the potentiometric measurements, kinetic measurements, and phosphorus analysis were described in the previous paper.8 The potentiometric measurements were carried out at 10.1 f 0.05" in aqueous media of an ionic strength of 0.100 M with KNOa as supporting electrolyte. All the kinetic studies were carried out at 30.0 f 0.05" by maintaining an ionic strength at 0.100 M withXN0,. (1) This investigation was supported by a grant from the Esso Education Foundation, Linden, New Jersey. (2) (a) Department of Organic Synthesis, Faculty of Engineering, Kyushu University, Japan: (b) Department of Chemistry, Illinois lnstitute of Technology, Chicago 16, Illinois. (3) R. Hofstetter, Y. Murakami, G. .Mont, and A. E. Martell, J . Am. Chem. SOC.,84, 3041 (1962). (4) G. Schwarzenbach, "Die Komplexometrische Titration," Ferdinand Enke. Stuttgart, 1.965 p. 68.
Results Interaction between SP and Cu(II).-To obtain information about the chelating tendency of salicyl phosphate with cupric ion, potentiometric titration of the SP-Cu(I1) system was carried out over an eightfold concentration range, as is shown in Fig. 1. These curves, together with the comparison in Fig. 2 of the titration curves of salicyl phosphate in the presence and in the absence of an equivalent concentration of Cu(I1) ion, give a clear picture of the nature of the interaction between SP and Cu(I1). Since SP undergoes hydrolysis a t or above room temperature, the measurements were carried out quickly a t 10.1 =t0.05'. Under these conditions the degree of hydrolysis of SP was observed to be negligible even after a titration had been completed. Since a precipitate was observed beyond m = 3.0, calculation of the formation constant, carried out algebraically in the usual way from data obtained below m = 2.60, indicated appreciable interaction between metal ion and ligand. Under these conditions the reactions are expressed by the following equilibrium constants: Dissociation of SP H2AHA2-
HA2A3-
+ H+
+ H+
Kz = 10-3*63 & = 10-6.37
where HIA represents the acid form of salicyl phosphate. Formation Constant Cu*+
+ A3-
CuA-
=
103*64
Interaction between SP and Cu(11)-D1PY.-Potentiometric titrations of solutions containing a 1: 1:1 molar
