The Acid Dissociation Constants of Diethylenetriaminepentaacetic

The Acid Dissociation Constants of Diethylenetriaminepentaacetic Acid and the Stability Constants of Some of its Metal Chelates1. Edward J. Durham, an...
0 downloads 0 Views 678KB Size
4512

EDWARD J. DURHAM AND DANIELP. RYSKIEWICI-I

tion and greater water structure-breaking) , unless the cation is so small and highly charged as to cause appreciable interaction with the iodide ion (through a water molecule) and so lower significantly thc activity coefficient of the iodide salt. Among the alkali metals this would be most important with LiI and least with CsI, and so, the activity coefficients of LiI and LiC104 should be most similar, that of NaI should be Significantly greater than that of NaC104 and as one descends the alkali family t o the cesium salts, where localized hydrolysis effects are negligible, the predicted order of activity coefficients should become more markedly that determined by hydration and water structurebreaking effects alone, namely, F > C1 > Br > I > c104. Experimental data are available only for the acids and the lithium and sodium salts, but they do show the predicted behavior (Table I ) . Since the dipositive alkaline earth ions as a group are more highly hydrated than the unipositive alkali metal cations, the rise in activity coefficients with increasing concentration should be more marked, and, experimentally, this is so with a larger spread in the values of the activity coefficients of a particular halide from M g to Ba than from Na to Cs. But the localized hydrolysis effect should also be enhanced due to the greater fields of the alkaline earth cations, and so the activity coefficient order I>Br>C1 should also be more marked; and i t is, see Table I. As with the alkalies, this effect should decrease as one descends the family to the larger and less polarizing cations, but with the highly polarizing dipositive ions, no reversal to the order expected of independent halide ions in solution occurs, even a t barium, although there is a trend in that direction. This trend toward the order F>C1>Br>I>C104 is more obvious if one compares the iodide and perchlorate salts. With magnesium, the localized hydrolysis effect is so important, even with the iodide, that the activity coefficients of MgIZ are below those of Mg(C104)2. With calcium the decrease in the activity coef-

Vol.

so

ficients of the iodide salt due to this effect is smaller and y + for CaIz and Ca(C10& are similar. With strontium and barium iodides the effect becomes unimportant, and the order I > CIOs becomes increasingly marked. The effect of localized hydrolysis should be still more important with small tripositive cations, and so, if no actual complex ion formation occurs, the order I>Br>Cl should be expected there, too, with the activity coefficients of the perchlorates becoming perhaps larger than those of the iodides, at least for small cations. There are few experimental data with which to compare, but in the case of the rare earth chlorides and bromides whose activity coefficients have been measured up to about O . O X W Z , the ~ ~ order Br>C1 is apparent. In summary it is seen that the idea of an interaction between oppositely charged ions through the intermediary of a water molecule is a reasonable one even for the anions of acids normally considercd strong, and this leads t o a decrease in the mean activity coefficient of the ions. This effect, which depends upon the polarizing power of the cation on the water molecule (proton repulsion) and up011 the proton accepting ability of the anion, can explain, then, the anomalous anion order of the activity coefficients of the alkaline earth and of most of the alkali metal halides, and the reversal of thiq order with the rubidium and cesium halides. The activity coefficients of the perchlorate salts (and presumably of other large, relatively unhydrated and non-associating ions) are in agreement with what would be expected from the effects discussed in this paper of water structure-breaking (a decrease in ? & with an increase in the size of the ion), and those ions where the charge is localized a t one part of a hydrophobic structure, as with the normal carboxylate anions, mill show the opposite behavior because of their water structure-tightening effect. (21) F IT Spedding, P E Porter and J RI Wright THE J o r R V ~ , 74, 2781 (lOSZ), F H Spedding and I S Yaffe, z h r d , 7 4 , 4 7 5 1 ( 1 c l i Z )

IIHACA, S. Y. -

[CONTRIBUTIOS FROM THE

XICIIOLS

LARORArORY,

S E W YORK UXIVERSITY ]

The Acid Dissociation Constants of Diethylenetriaminepentaacetic Acid and the Stability Constants of Some of its Metal Chelates1 BY EDWARD J. DURHAM AND DANIEL P. RYSKIEWICFI~ RECEIVEDMARCH26, 195S The acid dissociation constants of diethylenetriarninepentaacetic arid, H5Y, and the stability constants of its chelatcs C U + ~Zn+2 , and Cd+2 were determined a t 20" and an ionic strength with Mg+2, Ca+2,Sr+2,Ba+z, M n C z ,Fe+2, CoC2, of 0.10. The existence of a stable monohydrogen chelate species. MHY-2, and the usual chelate, MY-3, is illustrated. No bimetallic chelate is formed. The pH dependence of chelate formation is quantitatively described by means of pM, % MHY-2 and % MY-3 uersus pH plots.

The chelating properties of diethylenetriaminepentaacetic acid, DTPA, are of interest in view of its structural relationship to the well known chelating agent ethylenediaminetetraacetic acid, EDTA. (1) Abstracted from a dissertation submitted by Daniel P. Ryskiewicb to the Faculty of New York University in partial fulfillment of reqiiirements for the degree of Doctor of Philosophy. (2) Geigy Chemical Corp., Ardsley, li.Y.

CH?COOHOOCCHz+ );CH~CH~SCH?CII~S H CHZCOO1% -0OCCHz DTP.1 HOOC CH2 CHzCOOH >$CH~CH~X< H CH2COO-OOCCIT, EDTA L I

+

ACID n I S S O C I A T I O N COXSTANTS AND ALKALINEEARTJI CHELATE STABILITY COSSTANTS Cation

so

used for the calculation of K , from the experimental ‘‘a”and pH values.

and

me calculated, where K i y and rium constants of the reactions

voi.

cu Cd Cd cu CO

co 17e Fe Zri

M11

Ca

zn

MI1

Xi

C3.

21.63 24.44 19.94 15.60 23.87 39.56 25.54 21.17 21.32 17.01 23.22

log

rCMy

21.03 21.02 18.93 15.13b 21.09(21.05) 1S.87 19.04 16.55 1G.66 1 s . 17 15.35 (18.28) 20.21

The ratio M, t2/Rf + 2 was 1 in systems contailling Mn i - 2 This is the value of K\I,~Y and 10 in all the other systems. calculated using the value log K C ~ = Y 18.93. (1

The preceding heavy metal chelate stability constants have been calculated by assuming that the amount of heavy metal hydrogen chelate, MHY-2, present was negligible since the stability constants H3X+3 MY-3 + M,+2 K M H Yand KEY were not known. Also, for the same MX+* M.Y-3 3 H + (5) reason, the manganese hydrogen chelate had to be with an equilibrium constant neglected when manganese was used as the second metal ion. However, using the preliminary K M Y K , = (M.Y -3) (MX”) (H +) 3/(MY -9 (HsX+a)(Ma”) = values of Table I1 it was possible to calculate these KMXKM~YKR~X/KMY (6) constants from data for the equimolar DTPR-metal where KH*Xis the product, KIK2K3, of the acid dissociation constants of H3X+8and K M Xis the sta- ion titrations. Between a = 4 mid a = 5 bility constant of the heavy metal-amine chelate. The following material balance equations were c = (T)TPAIt = (MIIY - 2 ~-4- ( M Y 3j + a ( T G ) (loa)

+

+

+

ACIDDISSOCIATION CONSTANTS OF DIETHYLENETRIAMINEPENTAACETIC ACID 4815

Sept. 20, 1958 c = (M+')t

=

(MHY-')

+ ( M Y - 3 ) + (M")

( M H Y - ' ) -I- ( M Y - 9 [I g = (5

-a

-

) ~ (H')

=

+ K N Y ( y - 5])

= (MHY-')

+ y(Y-')

(lob)

greater the drop in PH. This generally observed property is used qualitatively to determine the relative tendencies of metal ions to combine with a

(10~)

TABLE IV

where

PREVIOUSLY REPORTED DTPA CONSTANTS

+ (H+)'/kikakdks + (H+)'/ksk&s + (H+)'/krks + ( H + ) / k s + 1 and y = 5(H+)'/kik2k&4ks + 4(H+)4/kzk&aks +

H+

From (loa), (lob) and Kh1Y

Mg+' Ca + 2

x = (H+)6/kikpkakdks

3(H '.) 3/k&aks

4- 2(H +) 'lk4k5 f

- (MHY-')

c

( M Y - 9 = X K M Y ( Y -=~ ) ~

$-

1

(H +)/ka

(11)

KanY(Y-7

Cation

M

+*

pkl

Pkz

fik3

Okr

pki

Ref.

1.90 1.79 1.86

2 70 2.56 2.79

4.42 4.42 4.29

10 62 8.76 8.61

12.59 10.42 10.48

3 4 5

log

KMIIY

5.71

Sr + 2 Baf2

3.64

log K M Y

9.34 9.98 9.48 8.78

6

5 5 6

given reagent. If the titration curves of Fig. 1 are considered, i t would seem, for example, that the order of stability of the heavy metal chelates of Substituting (12) into (11) and rearranging yields DTPA was Ni+2< Zn+2< F e f 2 < C O +< ~ C U +< ~ x K N Y ( Y - ~ ) ' ( x - y)(Y-') (g - C) = 0 (13) M n + 2 < Cd+2 which is, of course, not the case. which can be solved for (Y-6), permitting the calcu- This unusual behavior of DTPA readily can be lation of all other concentration terms and leading explained when one realizes that two chelates are lutimately to the determination of K M H Yand then being formed, one a monohydrogen species which is a relatively weak acid in most cases. I6 When both K M H Yand KLIY are high, as in the with the exception of log Kcay. This discrepancy is case of the heavy metal systems, chelation occurs to due to the fact that the monohydrogen chelate was such an extent that the pH is much lower than that not considered by Wanninen in that particular de- of DTPA alone, even a t "a" values below 3. The termination. The acid dissociation modes of relative position of each curve is seen to follow the DTPA have been described p r e v i ~ u s l y . ~ J order of decrease in p K t y in the range a = 4 to a = The unusual shapes and relative positions of the 5 where the conversion of h9HYV2to M Y + is takmetal-DTPA titration curves are of importance in ing place. a discussion of the metal chelate constants. It is a The quantitative relationship of the concentrawell known fact that, the greater the tendency for a tions of the two chelates present in a solution a t metal ion to combine with an amino acid, the various p H values serves to illustrate the points disFrorn (1Oc)

(MHY-')

+

= g

- y(Y-')

+

(12)

EDWARD J DURHAM AND DANIELP. RYSKIEWICH

4516

cussed above. Figures 2, 3 and 4 are plots of the percentages of MHY-2 and MY-3 simultaneously present a t various p H values in equimolar metal ion-DTPA systems a t a concentration of 1 X M . These values are determined readily from the experimentally determined values of the various constants. The instabilities of the hydrogen chelates of Sr + 2 and Ba+2are readily apparent since their maximum concentrations are only 9 and 3yo of the total metal ion concentration, respectively. h much greater percentage of the metal ion is seen to be present as the hydrogen chelate a t low pH values in the Ca+2and M g f 2 systems as was qualitatively deduced previously. The predominance of the hydrogen chelate a t low pH values and its subsequent conversion to the AMY-3chelate with increase in pH in the heavy metal ion systems is self-evident. ,inother way of describing the PH dependence of chelate formation is to plot PA1 versus PH curves whcre pM = -log The pill versus pH

Vol. 80

IO0

00

60

a-0 40

20

0

3

5

4

6

7

8

PH. 13ig 4 --tkr cciit X l I i \ i 2 . t i i d hIT fuiictioli of pH: -1 = CoHY-2, -1’ = B = ITCHY-~;B’ = C = ZnHY-2, C’ = ZiiY-l, (UTPA1)t= (M’L)t J

= 1

x

10-J

data of metal ioii-DTP,Z systeiiis having a total metal ion concentration of -121 in the presence of a 100yGmolar excess of DTPiZ were calculated. The results are presented in Table V. Results for some of the cations are presented in the form of pM E’S. pH curves in Fig. 3 with coni22

E D T A -----

20

/

18

16

14

P M. 12

10

0

6

4

2

2

4

6

8

IO

12

14

PH.

( M + * )= ~ 1

x

10-3.

Fig. S.-phf as a furictioii of pli: A atid -1’= Cui*; u and 13’ = Cd-2; C anti C’ = A l i i - 2 ; 1) atid 11’ = Cat*; E and E’ = lIg’.*; (Chelatiiig ageiitlt = 2 X ( M + Z ) t = 1 x 10-2.

THEINFRARED SPECTRA OF NITROAND NITRITOCOMPLEXES

Sept. 20, 1958

4817

TABLE V

pM versus pH DATA OH

Ca + z

Mg+2

Srt2

3.0 4.0 5.0 6.0 7.0 8.0 9.0 10.0 11.0 12.0 13.0 14.0

2.00 2.09 2.71 3.83 5.52 7.39 8.93 9.96 10.50 10.62 10.63 10.63

2.00 2.03 2.33 3.09 4.23 5.83 7.33 8.36 8.90 9.01 9.03 9.03

2.00 2.00 2.11 2.82 4.54 6.43 7.98 9.01 9.55 9.66 9.68 9.68

(M+2) = 1 X 10-1 M Ba+? Cu+2

2.00 2.00 2.01 2.23 3.31 5.38 6.92 7.96 8.50 8.62 8.63 8.63

(DTPA) = 2 X 1 0 - 2 M Ni+l c o +a

8.29 10.28 12.01 13.90 15.87 17.78 19.33 20.36 20.90 21.01 21.03 21.03

parable data for EDTA.15 It is immediately evident that DTPA is a more effective chelating than EDTA a t high pH values due to the greater stability of the MY-3 chelate. However, a t lower PH values, due to the influence of the less stable hydrogen chelate, DTPA is in some cases less effective, e.g., in the and Mg+2systems. Chelates of the type M2Y- are not formed to any extent since it has been shown that $K,H, is independent of the concentration of metal ion. If ligand-metal binding in the hydrogen chelates involved only one terminal iminodiacetic acid group (15) S. Chaberek. Jr., Arch. Biochem. Biophys., 55, 3 2 2 (1955).

[CONTRIBUTION FROM

THE

8.26 10.20 11.68 13.19 15.07 16.97 18.51 19.54 20.08 20.20 20.21 20.21

6.32 8.31 10.04 11.91 13.89 15.80 17.34 18.37 18.91 19.03 19.04 19.04

Zn+Z

Cd+?

Fe+2

6.09 8.04 9.53 11.12 13.02 14.92 16.46 17.50 18.03 18.15 18.16 18.17

5.52 7.68 9.76 11.78 13.77 15.69 17.23 18.26 18.80 18.92 18.93 18.93

4.30 6.25 7.84 9.57 11.51 13.42 14.96 15.99 16.53 16.65 16.66 16.66

Mn

+

2.40 4.09 5.86 7.99 9.97 11.89 13.43 14.46 15.00 15.12 15.13 15.13

the stability would be of the same order of magnitude as the chelates of iminodiacetic acid or methyliminodiacetic acid. However, the chelates are very much more stable indicating a greater degree of ring formation. Several possibilities also exist for the binding in the MY-3 chelates; however, no definite choice can be made a s to the actual structures from these data alone. Acknowledgment.-The authors wish to express thanks to the Geigy Chemical Corporation for financial support of this work. NEWYORK53, N. Y.

DEPARTMENT O F CHEMISTRY, OSAKA UNIVERSITY RESEARCH INSTITUTE]

AND THE

OSAKA MUNICIPALTECHNICAL

Infrared Spectra of Metallic Complexes. V. The Infrared Spectra of Nitro and Nitrito Complexes’ BY KAZUO NAKAMOTO JUNNOSUKE ,~~ FUJITA AND HIROMU MTJRATA~~ RECEIVED FEBRUARY 21, 1958

-

The infrared spectra of nitro and nitrito complexes have been measured in the 5000 400 ern.-' region. The normal coordinate treatment of the [Pt(N02)4I2-ion has been carried out to give complete assignments. The following conclusions have been obtained: (1) in a series of nitro complexes of various central metals, the metal-nitrogen bond becomes stronger progressively in the order of [iKi(NO,)6]4- < [Co(N02)6]3-< [Pt(NOz)a]2 - - . (2) the spectra of nitroammine complexes can be correlated with the structure more easily in the KBr region than in the NaC1 region; (3) the structure of the nitro bridge O ,H .in NH3)aCo-OH-Co( NH& ion can be determined by the infrared study; (4) [Cr(NH&N02]2+ is spectroscopically

r(

“02’

1 3 +

shown to be a nitrito and not a nitro complex.

Introduction Recently Beattie and Tyrrel13 studied the infrared spectra of a series of nitroammine coniplexes, [Co(”3) 6- w ( NOz) .](3--n) + in the 5000650 cm.-l region, and attempted to correlate the stereoisomerism and the number of nitro groups with the infrared spectra. It was found, however, that the spectra were too complicated to allow band assignment because the absorptions due t o nitro and ammine groups overlap each other. Faust and Quagliano4 also compared the infrared spectra of (1) Presented before the annual meeting of the Japan Chemical Society, April, 1957, Tokyo University, Tokyo, Japan. (2) (a) Department of Chemistry, Clark University, Worcester 10, Massachusetts; (b) Osaka Municipal Technical Research Institute, Osaka, Japan. (3) I. R. Beattie and H. J. V. Tyrrell, J. Chem. Soc., 2849 (1956). (4) J. P. Faust and J. V. Quagliano, Tars JOURNAL, 76, 5346 (1954).

trans- and cis- [Co(NH,),(N02)z]Cl.

However, no substantial difference was observed between these two isomers, although the latter exhibits a more complicated spectrum than the former in the 7 4 3 p region. It is expected that a study of the infrared spectra of these nitroammine complexes below 650 cm.-1 will afford more information, since the Co-NH3 and Co-NO2 stretching modes as well as the wagging, rocking and twisting modes of the nitro group appear in this region, and the overlapping of the nitro and ammine bands may possibly be avoided. The infrared spectra of ammine complexes have already been studied extensively by Powell and S h e ~ p a r d ,and ~ Mizushima, et a1.6 In order to ( 5 ) D. B. Powell and N. Sheppard, J . Chem. Soc., 4495 (1956). ( 6 ) S. Mizushima, I. Nakagawa and D. hf. Sweeny, J . Chem. Phys., a6, IOOG

(ism).