D. W. ROGERS, D. A. AIKESS,R K D C. N. REILLEY
1582
and the stationary state relation Rt
=
R,
=
k,[Initiator]f~
where Io is the intensity of actinic radiation. We have kt[M-]"= ki[Initiator]Ioand a plot of log {dR,/d(d - X ) / ( a ) ) against logarithm of the photoinitiator concentration has a gradient 1,ja. With benzoin methyl ether as photoinitiator, the data are shown in Fig. 7 where the gradient is 0.52 and
n, =
lit[?vI.]'.B2
This experimental relation also has been explored by changing the intensity of photo exposure. The logarithmic plot of Fig. 8 suggests 1,'a = 0.49, or Rt
=
kt[M.I2.03
T'ol. 66
This demonstrates a general consistency among the quantitative aspects of the postulated mechanism. The broad assumption of a heterogeneous polymerization system in which most of the propagation reaction occurs in virgin pockets has yielded a relation between the limiting degree of conversion and the initiator concentration or the intensity of photo exposure. This is in close agreement with the experimental work with a polyfunctional monomer. Acknowledgment.-The author wishes to acknowledge helpful conversat.ons with many colleagues in the Photo Products Department of E. 1. du Pont de Nemours Co., and especially with nlr. W.R. Saner and Dr. D. R. White.
THE KISETICS OF EXCHANGE OF COPPER(I1) BETWEEN ETHYLENEDIAJIISETETRAACETIC ACID BKD ERIOCHROJIE BLUE BLACK R BY DOXALD I T . ROGERS, DlzVID A. AIKEXS, A S D CHARLES s.REILLEY Department of Chemistry, University of ,VOIth Carolina, Chapel Hill, 4 o r t h Carolina RseetLed A'ovemher IS, 1061
The exchange of Cu( 11)between EDTA and Eriochronie Blue Black R, I-( 2-hydroxy-l-naphthylazo)-2-naphthol-4-sulfonic acid, a typical o,o'-dihydroxyazo compoun'd,,occurs by acid-catalyzed dissociation of the chelates below pH 5 and by secondorder displacement reactions above pH 5. Bonding of the attacking ligand to the Cu chelate followed by competition between the two ligands is postulated for the second-order reartions. EDTA, in order to displace Erio R from Cu effectively, must possess at least one proton, which weakens the phenolic Erio R-Cu bond in an 832 step. The rate of attack of Erio R on Cu-EDTA depends on the fraction of the Erio R having one unprotonated phenolic group, suggesting an S N reaction ~ in which the activated complex involves one chelate ring with the attacking ligand. The formation of Cu-OH-EDTA enhances the rate of this reaction since the Cu-Erio R product is stabilized more by mixed hydroxide complex formation than is CuEDTA. Effective second-order rate constants for the combination of Cu(1I) with Erio R and with EDTA a t pH 3.3 calculated from the dissociation rates of the Cu chelates are of the order of 106 i1f-l sec. for both ligands.
Introduction The importance of ethylenediaminetetraacetic acid (EDTA) as a chelating agent has led naturally to study of the kinetic behavior of its metal chelates. The bulk of this work, which has been concerned with the exchange of aquated metal ions with metal chelates, has been reviewed by Rfargerum. The interesting related problem of the exchange of a metal ion between two multidentate ligands has, as yet, received little attention for ligands of the aminopolycarboxylate type, the work of Bosnich, Dwyer, and Sargeson on the exchange of metal ions between the d and 1 isomers of 1,2-propylenediaminetetraacetic acid being the only example. The exchange of a metal between EDTA and a second multidentate ligand of contrasting structure is of interest since it allows speculation on the relation between the observed rates and the structures of the entering and leaving ligands. Eriochromc Blue Black R, 1-(l2-hydroxy-l-naphthylazo)2-naphthol-4-sulfoniv acid (Erio R), \vas chosen as the second ligand as representative of the o,o'-dihydroxyazo type of ligand, the Cu chelates of which are of appropriate stability and of well defined structure. (1) D. UT.Marpeium, .I P h ~ s Ciiem., . 63,386 (1'259) (2) R. Bosnioh, F. P. Dnyel, and A. AI. Saigeson, S a t m e , 186, 966 (1960).
Experimental Reagents.-Solutions of EDTA were standardized against copper metal. Erio R (Color Index No. 202, supplied bv American Cyanamid Company, Kew York, S . Y.) was freed of metallic impurities by three preci itations from aqueous solution by the addition of hydrochoric acid. The precipitated dye was dried at 60' under vacuum. The purified dye contained less than 27, metal impurity transferable to EDTA, as judged from the increase in absorbance at 635 mp of a dye solution at pH 9.3 on addition of a 100-fold excess of EDT.4. Solutions of Erio R were prepared daily and standardized against copper by photometric titration. Stock solutions of copper-Erio R, about 2 X M, were prepared in 0.001 M x e t a t e buffer, pI-1 5 , ttnd used within 8 hr. to avoid deterioration. Stock EDTA solutions were prepared and converted to Cu-EDTA by addition of the appropriate amount of Cu(I1). I n both Cu-EDT.4 and Cu-Erio R, a 1% excess of the ligand was present to prevent intwference by free Cu(I1). Demineralized water was used in all solutions. All kinetic and equilibrium mrasurements were made at f = 0.1, maintained with sodium nitrate. Chloroacetatc,, aretatc, phosphate, carbonate, and borate huffws were uscd tit 0.01 .If concentration, giving pH control diiring thc rourw of a rcac*tionto within 0.02 pH unit, thc limit of drtertion. A properly cdibrated Leeds snd Sorthrup Motlcl 7664 pII meter equipped with a Leeds and Northrup STD. 1 I B R - X I glars electrode was used for all pH determinstions. Beraustl of the smsitivity of Erio R and Cu-Erio R solutions to slow aggregation at low pH in the presence of electrolytes, and to oxidation of the dye at high pEI, the buffer and sodium riitrate were incorporated in the EDTA and Cu-EDTA solutions. Procedure. Rate Data.-All sprctrophotometric measure-
ments were made with a Cary Model 14 recording spectrophotometer with the cell compartment thermostated at 25.0 & 0.2". Rate constants were obtained from the limiting rate of change of absorbance a t 545 m p extrapolated to zero time. The reaction was initiated by mixing the reactants directly in a stoppered spectrophotometer cell, permitting the absorbance to be monitored within 10 sec. of the start of the reaction. I n a typical experiment 2.00 ml. of 2.0 X N Cti-EDTA (or EDTA) was placed in the cell and 1.00 ml. of 3.00 X M Erio R (or Cu-Eric) R ) waa blown from a calibrated pipet into the cell, which was capped and inverted several times. The concentration of Cu-EDTA (or EDTA) aolution wa8 varied from about 1.0 X 10-6 M when the reaction rate constant was of the order of 102 M-1 see.-' to about 1.O X 10-3 M when the rate constant was of the order of 1 M-1 set.-'. The Cu-Erio R (or Erio R) concentration was of the order of 1 X 10-6 M . The rate of reaction is related to the rate of change of absorbance by the expression
a 2
oy
%
0
O
I
r
0
Y
a0
53
I
-2
3
where dA/dt is the rate of change of absorbance extrapolated to zero time, az and acuz are the absorptivities of Erio R and Cu-Erio R , respectively, d(Z)/dt is the rate of appearance (or disappearance) of Erio R a t zero time, and b is the cell length. The contribution of Cu-EDTA to the absorbance change can be neglected as its absor tivity is less than that of the Erio R species by afactor of atfeast a hundred. The reproducibility of the rate data from day to day was of the order of 30y0in spite of precaut,ions to standardize the experimental techniques. This variation may be caused by irre roducibility in the mixing technique and by variations ine!t Erio R and Cu-Erio R solutions, such as undetected aggregation. In addition, the inherent difficulty in determining the slope of the absorbance us. time plot is undoubtedly a factor. The problem of reproducibility was accentuated below p H 5 where the reaction rate was independent of the concentration of the attacking ligand. Above pH 5, the second order of the reaction allowed ready adjustment of the reaction rate to a convenient value. Equilibrium Data.-The acid dissociation constants of Erio R were determined spectrophotometrically as described by Irildebrand and Rcilleya and the formation constant of the hydroxide derivative of Cu-EDTL4 spectrophotometrically as debcribed by Bennett and Schmidt.* The effective equilibrium constant for the exchange reaction was determined from the absorbance of mixtures containing Erio R and Cu-Erio R in approximately equal concentrations and at least a fivefold excess of both EDTA and Cu-EDTA. The absolute stability constant of Cu-Erio R was estimated from spectral measurement of the dissociation of the chelate in acid solution rontaining a tenfold excess of Cu(I1) as described by Rogers.5
Results Equilibrium data for the exchange reaction are presented in terms of effective constants based on total concentrations of ligand and chelate species without regard to such factors as chelate derivative formation or the degree of protonation of ligands. Such constants are useful as the numerical magnitudes reflect directly the composite effects of all the competing equilibria involved. Effective exchange constants are readily derived from the absolute stability constants of the chelates by consideration oE all the competitive equilibria. Equilibrium Data.-The sucocssivc pKa's of li;rio R UTV 1,7.3, aiicl 13.6. The first value, that for the sulfonic acid proton, is somewhat uiiccrtain hecause of the small spectral diffcrciicc betwecu (3) G. P. Hildebrand and C . N. Reilloy, Anal. Chem., 29, 339 (1057). (4) hl. C . Bennett and N. 0. Schmidt, T m n s . Faraday Soc., 61, 1412 (1955). ( 5 ) D. W. Rogers, P h . l ) . Thesis, University of North Carolina,
1960.
4
5
6
7
8
9
IO
II
I2
PH.
Fig. 1.-Effective equilibrium constant for the exchange reaction CuZ Y -P. CuY Z as a function of pH: solid line, calculated from forination constants of the chelates, acid-base constants of the ligands, and formation constants of mixed ligand chelates with hydroxide: dashed line, as above, except formation of Cu-OH-Erio R neglerted; filled circles, ratio k&,; open circles, measured values of keu.
+
+
the acid and base forms. The latter values, for the phenolic prot,ons, are much more reliable and are in good agreement with those of previous workers.3 The effective equilibrium constant, K,tf, for the exchange reaction (1) where Y is EDTA and 2 is Erio CUZ
+ Y +CUY 4-Z
(1)
R, is given from pH 3 to pH 12 in Fig. 1. Experimental values of K,ff determined by direct measurement (open circles) agree within 0.2 pK unit with values calculated from kfefr/Jcrerf, the ratio of the forward and reverse effective rate constants (filled circles) above pH 5. The relation of the rate data to the equilibrium constant below pH 5 , where the reaction rates are independent of the concentration of the attacking ligand, is more complex and will be discussed separately. Agreement between the two experimental sets of Keffand values calculated from the effective stability constants of the two chelates (solid line) is between 0.1 and 0.2 pK unit, if the formation of hydroxide derivatives of Cu-EDTA and Cu-Erio R is considered. The directly measured value of the equilibrium constant for the formation of the former from Cu-EDTA and hydroxide is 2.1 log units, in good agreement with the value found by Bennett and S ~ h m i d t . The ~ formation constant of Cu-OH-Erio R cannot be determined directly by spectral methods because of the similar spectra of the chelate and its hydroxide derivative, but a value can be inferred readily from the pH dependence of mcasurcd values of Kerf. Neglect of this species results in large discrepancies between measured and calculated values of Keff above pH 9 (dashed line), while a value of' 4.0 log units for its formation constant gives excellent agreement8to pH 12. Calculated values of Keffare based on data quoted and the (6) G. Sohwarzenbach and W. Uiedermann, Helu. Chrm. Acta, 31, 678 (1948).
D. W. ROGERS,D. A. AIICENS,AND C. N. REILLXY
1584
Vol. 66
a t pH 3.15 are limited to Erio R concentrations greater than 1.0 X &I and less than 1.0 X M , the first because of the unfavorable equilibrium constant of the reaction, and the second because of the high absorptivity of Erio R. Over this range of Erio R concentrations no deviation from first-order kinetics is observable. The reactions of EDTA with Cu-Erio R and of 1 . G X 10’(CuZ)(H) 7 X 101(CuZ)(H2Y) Erio R with Cu-EDTA both follow pseudo-first 1.5 X 10z(CuZ)(Hk’) order kinetics above pH 5 over 95% of the reaction 1 . 5 X 102(CuZOH)(HY) when carried out with a large excess of attacking from pH 3 to 13 and the rate of the reaction of Erio ligand. If more than one Cu chelate species is present, equilibrium among them is rapid. Good R with Cu-EDTA in the same pH range by agreement was found between the pseudo-first order rate constants and those calculated from the 8 . 4 X 10-4(CuHY) 2 X 10-1(C~~Y)(H2Z) limiting rate of change of absorbance. 2 1 X 101(CuYj(HZ) Discussion 3 . 1 X 102(CuY0H)(€IZ) Chelate Stability Constants.-The absolute In each case the reaction rate is independent of the concentration of the attacking ligand below stability constant of Cu-Erio R is estimated as pH 5 and first order in the attacking ligand above 21.2 log units, somewhat higher than would be pH 5 . The reaction pairs chosen for the formula- expected from published stability constants of metal tion of the rate law are not necessarily unique and ions with EDTA and with Erio R. The absolute other combinations of the same over-all stoichiom- stability constants of Ca-EDTA and Ca-Erio R etry will fit the data as well. However, except are 10.7 and 5.2 log units, respectively, while that for the last term of the rate law for the reaction of of Cu-EDTA is 18.8 log unit^,^,^ so that the order EDTA with Cu-Erio R, a reasonable and mechanis- of stability of the copper chelates is opposite to that t>ically consistent choice can be made simply by of the calcium chelates. This increase in the relative stability of Cu-Erio formulating the rate lam in terms of the predomiR is due to the large ligand field of the aromatic nant species of each reactant. This criterion cannot be applied to the term in question as the pH of the ligand, and may be caused in part by the effect of CuZ/CuZOH transformation (pH 10.0) is nearly n-bonding between the d electrons of Cu(I1) and equal to the pH of the HY/Y transformatioii (pH the conjugated aromatic ring system of the dye. 10.3) and appreciable concentrations of all four Qualitatively, the stability of the complex is ensperies exist near pH 10. Although this suggests hanced by delocalization of the electrons in the t, that both the path (CuZ)(Y) and the path (Cu- orbitals of the central ion over the vacant n-orbitals ZOH)(HY) may be important in the reaction above of the ligand as discussed by Orgel.g As EDTA has pH 10, indirect evidence presented in the Discussion no delocalized n-orbitals, its chelates cannot benefit section strongly favors the latter pair, which has from n-bonding. Second-Order Displacement Reactions.-The obbeen chosen for the formulation of the rate law. servation of second-order kinetics above pH 5 in The second-order rate constants for both exthe reaction of EDTA with Cu-Erio R and in the change reactions are valid over a 100-fold range of reaction of Erio R with Cu-EDTA does not in itreactant concentration product and from reactant self shed much light on the actual reaction mecheoncentration ratios of 0.2 to 5.0. The rate conanism. Because of the large number of bonds instant for the reaction of EDTA with Cu-Erio R fluctuated between 1.1 X 102 and 1.4 X 1022W-l volved, a one-step reaction is very unlikely. Rather it is quite probable that a stepwise reaction occurs, S P C . - ~ a t pH 7.0 as the concentrations of EDTA and of Cu-Erio R were increased simultaneously by up the first step of which is the bonding of the attackto tenfold from their lower limits of 1.0 X 10-6 and ing ligand to a loosely chelated or unchelated site (or sites) of the Cu chelate to form a mixed ligand 2.1 x X , rcspectively. The rate constant for chelate in a second-order step. This is followed by the reaction of Erio R with Cu-EDTA a t pH 8.5 varied randomly between 20 and 24 J1-l sec.-l stepwise replacement of the ligand coordination sites of the original chelate by those of the attackwith the same variation in the EDTA and Erio R ing ligand. The activated complex is associated species. The pH value for each reaction was chosen with the formation of a certain critical number of to minimize variation of reaction rate with pH bebonds between Cu(1I) and the attacking ligand. tween experiments. Whether these steps proceed primarily by disThe reaction of EDTA with Cu-Erio R at pH 3.26 gives consistent first-order rate constants over sociation of the individual metal-chelate bonds to 3.3 (Sisl) or by displacement of the original chelate coan EDTA concentration range of 2.0 X X 10-5 M . At EDTA concentrations above 3.3 ordination links by the attacking ligand ( S N ~ )or by some other path must be deduced from detailed X 10-6 M , the effect of the concurrent second-order examination of the rate data. This second phase of reaction becomcs apparent. Data for the firstthe reaction is undoubtedly complex, but consideraorder path of the reaction of Erio R with Cu-EDTA tion of the ratc data together with the acid-base (7) G. Schwarzenbach and H. Ackermann, HeEv Chzm Acta, 30, constants of the ligands and the chelate formation 1798 (1947). acid-base constants of EDTA given by Schwarzenbach and Ackermann,’ and the stability constant of Cu-EDTA as determined by Schwarzenbach, Gut, and Anderegg.8 Kinetic Data.-The rate of reaction of EDTA with Cu-Erio R in J4-l sec.-l is given by the expression
+
+
+
+
+
+
( 8 ) G . Sohwvareenbach, R. Gut, and H. Anderegg, %bud., 37, 937 (1954).
(9) L E. Orgel, “An Introduction t o Transition Metal Chemistry,” Methuen, London, 1960, pp. 106-108.
Sept., 1962
EXCHANGE OF COPPER(II)BETWEEK EDTA AHID ERIOCHROME BLUEBLACK R
constants of the ligands allows some interesting mechanistic speculation. For convenience, the formation of a mixed complex will be discussed first, followed by consideration of the replacement phase of the reaction. The mixed ligand complex EDTA-Cu-Erio R formed in the exchange reactions, as inferred from the second-order kinetics, exists in lorn concentration relative to the reactants. First, attempts to detect this intermediate spectrophotometrically were unsuccessful, The absorbance of the reaction mixture, extrapolated to zero time, was equal within experimental error to that expected from the sum of the reactants. Second, the reaction of EDTA with Cu-Erio €1remained second order even in the presence of a fivefold excess of EDTA. Formation of a significant concentration of mixed ligand intermediate would result in a lowering of the apparent order of reaction in the presence of excess attacking ligand. The inference of a mixed ligand complex involving inultidentate ligands is supported by a considerable amount of indirect evidence. The existence of derivatives of Cu-EDTA is well known. Mixed bideiitate ligand complexes of Cu(I1) have been demonstrated by Watters.lo Mixed ligand Cu(I1) chelates involving o,o‘-dihydroxyazobenzeie were studied by Jonassen and Oliver,ll who concluded that 8-hydroxyquinoline can occupy two coordination sites in such chelates. In the case of substitution-inert Ki(I1) species, mixed ligand intermediates have been detected in kinetic experiments. In the reaction of Xi-EDT-4 with cyanide, Margerum, Bydalek, and Bishop12identified Xi-EDTAC S and Ni-EDT,4-(CX)2 as intermediates in the forward reaction and found evidence for Si-EDTA(CK), as an intermediate in the reverse reaction. Although cyanide is a monodentate ligand, .the analogy of the mixed complexes to those proposed in the present study is clear. Another mixed complex of Ni(I1) with tetraethylenepen tamine and Erio T, an analog of Erio R, has been reported by Reilley and Schmid.ld Some insight into the mechanism of the exchaiige reactions a t individual coordination sites is given by consideration of the magnitudes of the terms in the rate laws for the exchange reactions. In addition, estimates of the proton affinities of the coordination sites of the coordinald ligands and of the relative stabilities of individual metal-ligand bonds in the two chclates are necessary. The equilibrium constant for the protonation of Cu-EDTA (formation of a binuclear H-EDTA-Cu species), presumably at one of the carboxylate groups, is 3.0 log units.’ Data for the amine sites is not available, nor is there any data for the protonation of CuErio R. As a rough approximation, the proton affinities of the coordination sites of the chelated ligands have been equated with the proton affinities of the free ligands. The errors introduced tend to cancel to a large extent since the diflerence in proton affinities of the two ligands is important, and both (IO) J. I. Watters, J. A m . Chem. Soc., 81, 1560 (1959). (11) H. Jonassen and J . Oliver, zbzd., 80, 2347 (1958). (12) D. W. Margeruin, J. T. Bydalek, and John J. Bishop, tbzd., 83, 1791 (1961). (13) C. N. Reilley and R. W. Schmid, Anal. Chem., 31, 887 (1959).
1585
are coordinated to some extent to Cu in the activated complex. The average strength of the bonds between Cu and Erio R is obviously much higher than the average strength of the bonds in Cu-EDTA. Erio R, with its o,o’-dihydroxyazo structure, can occupy three coordination sites around C U ( I I ) , yet ~ ~ the stability constant of Cu-Erio R is 21.2 log units, while the stability of Cu-EDTA with five or six metal-ligand bonds is only 18.8 log units. Although rather crude, this evaluation of the relative strength of the bonds in the two chelates is sufficiently clear cut for the present purposes. The rate of attack of monoprotonated Erio R 011 Cu-EDTA (21) is of the order of 100-fold faster than the rate of attack of the diprotonated form of Erio R (0.2). Although the diprotonated form has one free coordination site available for bonding to Cu-EDTA, its ability to participate in further bonding is limited. The high proton affinity of the Erio R (pK1 = 7.3) and the low proton affinities of the EDTA sites, (pK1 = 2.0 (COO-), pKz = 2.8 (COO-), pK, = 6.2 (X)) preclude the operation of a concerted SE2 reaction in which the proton on the attacking Erio R assists in breaking the CuEDTA bonds, Further, the high proton affinity of the Erio R impedes the formation of a strong bond between the Cu and the phenolic group of Erio R. Removal of the first phenolic proton allows the Erio R phenolic site to act as an effective displacing agent for the EDTA carboxylate group in an Sx2 step by virtue of the higher stability of the bonds between Cu and Erio R. A carboxylate seems more likely to be displaced than an amine because the peripheral nature of the former requires the breaking of only one metal-ligand bond. Removal of the second proton from Erio R has much less, if any, effect on the rate. At pH 13, where 20% of the Erio R is unprotonated, no increase in rate of attack on Cu-EDTA over the monoproiconateti species is observed. Apparently the formation of the first chelate ring b e k e e n Cu :irid Erio R is rate-determining. The rate of attack of EDTA on Cu-Erio R is nearly constant from pH 5 to 10, the range in which EDTA exists as a diprotonated or monoprotonated species, with a sharp decrease above pH 10. A choice between the alternative terms 150 (CuZOH) (HY) and 3OO(CuZ)(Y) applicable above pH 10 can be made by comparison of the reactivities of both forms of each species. as reflected in rate law coefficients. The coefficient for the term (CuZ)(HY) is 150, while in contrast that for the term (CuZOH)(Y) is of the order of 0.3, even if the total rate of the reaction is attributed to this path at pH 13, where these are the major species. The critical factor in determining the rate is therefore the presence of a proton on the attacking EDTA, and on this basis, the rate law term (CuZOH) (HY) has been chosen. The role of the proton is very likely to at least weaken the Cu-Erio R phenolic ~ followed by a displacement of bond in an S E step, the weakened bond by the EDTA. It is extremely unlikely that an EDTA coordination group could (14) H. D. K. Drew and J. 41. Landquist,
J. Chem. Soc., 292 (1938).
1586
t>. W. ROGERS,D. A. AIKENS,AND C. N. REILLEY
Vol. G6
effectively displace Erio R from its Cu bond, based 16(CuZ)(H) and that of Cu-EDTA with Erio R on the prior estimates of bond stabilities. by 8.4 X (Cu-H-EDTA), the formation conA twofold increase in rate does occur as the pH is stant of Cu-H-EDTA being taken as 3.0 log increased above 6, coincident with the tfansforma- units.7 50 evidence for the formation of a protion of H2Y into HY. The coefficient for the term tonated chelate derivative such as Cu-H-Erio R (H2Yj(CuZ) is 7 X lo1, while that for the term was found in pH titrations of Cu-Erio R. (HY)(CuZ) is 1.5 X lo2. In H2Y both amine Further evidence for a dissociation mechanism groups are protonated, while in HY the proton prob- is the inhibiting effect of free Erio R on the initial ably is localized between the two nitrogens as sug- rate of reaction of EDTA and Cu-Erio R. A gested by Olson and P l I a r g e r ~ m . ~The ~ amine mechanism consisting of a slow dissociation of Cunitrogens in HY are probably in a cis orientation Erio R, followed by competition between Erio R and both could readily attack the Cu-Erio R chel- and EDTA for the free Cu(II), yields the following ate. In H2Y, however, it is likely that the nitro- expression on assuming a steady state concentragens are better described as in a trans orientation, tion of Cu(I1). The rate of dissociation of Cudue to the electrostatic repulsion of the two pro- EDTA is negligible and is not considered. tons. Thus if one of the nitrogens is oriented to attack Cu-Erio R , an internal rotation about one of the C-C or C-Tu' bonds in the ethylenediamine nucleus is necessary before the second nitrogen is in a position to attack the Cu chelate structure. where An alternative explanation is that the protonation of both nitrogens in H2Yreduces the ability of these R is the observed initial rate of reaction groups to bond to Cu, simply because the nitro- kdcua is the effective first-order dissociation rate constant of Cu-Erio R (rate/(CuZ)) gens already are coordinately saturated. Because k, and k , are effective rate constants for the comEDTA has a t least 4 free coordination sites above bination of Cu(I1) with EDTA and with Erio R pH 5 , it is not possible to speculate as to the number of points of attachment of EDTA to Cu-Erio A test of the rate law at pH 3.30 gives general agreein the activated complex except to note that 3 ment between observed and predicted rates. The sites on Cu-Erio are unchelated and could be in- initial rate decreases from 8 X lo-* to 1.6 X volved. hi?-' sec.-l as the concentration of free Erio R is Superimposed on the basic reaction mechanism is increased from zero to 6 X 10-5 174. The ooncenthe effect of formation of hydroxy derivatives of the tratioiis of EDTA and Cu-Erio R are both 1 X 10-5 chelates. The formation of Cu-OH-EDTA is as- dl. sociated with an increase in the rate of attack of In addition to aiding in the elucidation of the Erio R on CWEDTA above pH 11. The rate in- reaction mechanism, these data allow an estimation crease reflects the relatively greater stabilization of of both the ratio k y / k , and of the magnitude of IC,. Cu-Erio R (compared to Cu-EDTA) by hydroxide Although the scatter in the data precludes an exact derivative formation, the formation constants of determination, 12, is roughly a factor of five greater the two derivatives being 4.0 and 2.1 log units. than k,. Even though the effective stability conThe reaction of CU-OH-EDTA with Erio R brings stant of Cu-EDTA is about a hundredfold greater together in one step all the reactants needed for the than that of Cu-Erio R, the effective rates of formation of Cu-OH-Erio R , with a resultant formation differ only by a factor of five. The effecstabilization of the activated complex. tive value of k , estimated from the product of kdcuz Dissociation Reactions.-Below pH 5, the rates (8 X and the effective stability constant of of both exchange reactions become independent CuZ a t pH 3.30 (8 X lo6) is 6 X lo4. of the concentration of the attacking ligand, sugAcknowledgment.-D. W. R. wishes to acknowlgesting dissociation of the chelates as the rate conedge the financial assistance of the American Vistrolling process. Further, the rates increase with cose Corporation. Part of this research was supacidity analogous to the observations of Cook and ported by the United States Air Force through Air Long16 and Jones and Long17 on the dissociation Force Office of Scientific Research, Air Research of the Ni and Fe chelates of EDTA. The rate of and Development Command, contract No. AF the reaction of Cu-Erio R with EDTA is given by 49(638)-333. Presented at the Combined Meeting, ( 1 5 ) D. C. Olson and D. W. Margerum, J . Am. Chem. Soc., 82, 8602 Southeastern and Southwestern Sections of the (1960). American Chemical Society, New Orleans, La., (16) C. hl Cook, Jr., and F A. Long, t h d . , 80, 33 (1958) Dec. 7-9,1961. (171 S, S. Jones and F. A. Lone, J. P k y s . Chem., 56, 26 (1962).
