The Stability of the Rare Earth Complexes with

The two methods are shown to supplement each other, one being more accurate for the ... containing a proton and the rare earth cation simultaneously, ...
0 downloads 0 Views 713KB Size
4196

E. J. WHEELWRIGHT, F. H. SPEDDING AND G. SCHWARZENBACH

VOl. 75

[CONTRIBUTION NO. 243 FROM THE INSTITUTE FOR ATOMICRESEARCH AND DEPARTMENT OF CHEMISTRY, IOWASTATE COLLEGE. PART OF THE W O R K W A S PERFORMED IN THE AMES LABORATORY OF THE ATOMIC ENERGY COMMISSION]

The Stability of the Rare Earth Complexes with Ethylenediaminetetraacetic Acid BY E. J. WHEELWRIGHT,F. H. SPEDDING AND G. SCHWARZENBACH~ RECEIVED OCTOBER30, 1952 The stability constants of the complexes formed between the rare earth metal ions and the anion of ethylenediaminetetraacetic acid have been measured a t a temperature of 20" and an ionic strength of g = 0.1 by two independent methods. The two methods are shown t o supplement each other, one being more accurate for the lighter rare earths and the second method more accurate for the heavy rare earths. The first, a potentiometric method, involves the formation constants of copper with the anion of ethylenediaminetetrcetic acid and of copper with the trihydrochloride of B.B',fl"-triaminotriethylamine. The second method involves the polarographic determination of the amount of free copper in the presence of both the copper and rare earth tetraacetate complexes.

Introduction The tervalent rare earth cations, all having the same charge and apparently the same electronic configuration of the outer orbitals, form a unique series for the study of the influence of the size of a cation upon its physical and chemical properties. Complex formation of these cations in homogeneous solution is of special interest in this connection, because of its independence of crystal structure and lattice requirements of a solid phase. Very few quantitative data of this nature have been obtained because of the small tendency of the rare earth cations to form complexes. I t is known, however, that the rare earths form stable complexes with the anion of ethylenediaminetetcetic acid. The symbol HdY is used in this paper to designate the acid. This substance has been used for an improved fractional separation of the rare earths one from another.2 Stability constants of the rare earth-ethylenediaminetetraacetate complexes have However, we do not been reported by V i ~ k e r y . ~ ~ consider the method used by this author to be applicable.ab The equilibrium between the copper complex CUY-~,uncomplexed cupric ions Cuf2and the rare earth complex MY- in the presence of uncomplexed rare earth cation M+3has been investigated in order to obtain the stability constants of these complexes. The rare earth coniplex is defined as

information as to the extent to which this reaction had proceeded, and would allow the calculation of its equilibrium constant. A simple exchange of the protons by the metal cation in the reaction M+*

+ H,Y

MY

I-4

-+ x H +

(1)

was used in order to obtain the stability constants and the other alkaof the calcium complex line earth complexes. This is not suitable for the purpose if M is a rare earth metal, since the equilibrium of (I) lies too far to the right hand side. A mixture equivalent in rare earth salt and the salt NazH2Y behaves very much like a strong acid. The hydrogen ion concentration of such a mixture is only slightly less than twice the total concentration, [MIt, of the metal or the complexing agent, [Ylt. Only a very small amount of uncomplexed M+3 is left over; the concentration of this is obtained as difference between two nearly equal quantities: [M]t - 2[H]. This example indicates that accurate measurements by means of pH determinations can best be made when the reaction in question exerts a buffer action in a suitable pH range, which has to be between pH 4 and pH 10. It has been found that reaction (11) is suitable for the purpose4

+

C U Y - ~ Ha tren+3

+ M+s

MY-

+ Cu tren+* + 3 H +

(11)

If a mixture of the copper complex of ethylenediaminetetraacetic acid, NazCuY, the rare earth salt, MCb, and the trihydrochloride of PI@'$"The brackets denote concentrations and all data triaminotriethylamine, N(CHrCHrNH&, desigobtained are valid for an aqueous salt solution of nated as "tren," is titrated with NaOH, three formula weights of the base are used in a buffer ion strength p = 0.1 and a temperature of 20". The Potentiometric Method.-Because of the region between PH 4 and 5.5. During this process, great difficulty of measuring the concentration [MI reaction (11)is delivering the protons. The tri-(P-aminoethy1)-amine,"tren," is a strong of the uncomplexed rare earth cation potentiometrically, a process was sought wherein the triprotonic base with the pK values of 10.29 , ( = log KHatren) and 8.56 complex MY- would be formed or decomposed (= log K H * ~ ) 9.59 simultaneously with the production or consumption (= log KHltren) and therefore we have of an equivalent amount of hydrogen ions. The concentration of the hydrogen ion, [HI, would give (1) Dr. Schwarzenbach held the position of "Distinguished Visiting Professor" in the Institute during 1951 while on leave from the University of Zurich, Switzerland. (2) G . Beck, Mikrochcmic, 88, 344 (1948); Anal. Chim. Acta, 8 , 41 (1949); J. K.Marsh, J . Chcm. SOC.,1819 (1950); 3057 (1951). (3) (a) R. C. Vickery, ibid., 1817 (1951); 421 (1952); 1895 (1952). (b) Vickery defined the constants by the equation K Z [MHY]/ [M].[HY], and reported values which are approximately the same as those reported in this paper. Complexes of the composition MHY, containing a proton and the rare earth cation simultaneously, do not occur in the equilibrium mixtures.

The tri-(P-aminoethy1)-amine,"tren," also forms a very stable complex with bivalent coppers .

.-

.

(4) G . Schwarzenbach and E. Freitag, Hslv. Chim. Acta, 8 4 , 1503

(1951). (5) J. E. Prue and G. Schwarzenbach, Hrlv. Chim. Acta, 88, 963 (1950).

&pt. 5, 1953

STABILITY OF RARE

EARTHCOMPLEXES WITH ETHYLENEDIAMINETETR~ACETIC ACID 4197

ml. of solution B, 10 ml. of solution C and 9.00 ml. of 1.00 M KCl. Next, different amounts of 0.01 M NaOH were added, for instance 5 ml. t o the first flask, 10 ml. t o the second, 15 ml. to the third and so on. Each mixture was C U + ~ Hstren+S CutrenfZ 3 H t (111) brought up t o a volume of 100 ml., placed in a thermostat which exhibits a buffer action between pH 3.9 and set a t 20.00 f 0.02’ and allowed t o equilibrate for 24 hours. At the end of this period the PH was determined using a glass 4.8. electrode and a Beckman pH meter. Check determinations The stability constanta of the complex CUY-~, were allowing 48 hours equilibration, for praseodyof copper and ethylenediaminetetraacetate, is mium,made, neodymium, samarium and yttrium. Since the resomewhat smaller than K c u t r e n sults were the same as those obtained with 24-hour equilibration, the practice was discontinued.8 The pH meter was calibrated against a solution of 0.001 M HCl containing (3) 0.10 M KC1 which gave a p H of 3.00 at a constant ionic of p = 0.10. Note that the concentration and not Because of the greater proton aflinity of the base strength the activity of the hydrogen ion was measured. Several “tren” in comparison to the anion “Y,” the com- check solutions were independently prepared and all plex Cutren+2is less stable than the complex C U Y - ~ checked a t a pH of 3.00 i0.02 relative to each o t h y c ,, The experimental results are given in Table I, a dein a pH range below 8.5. If NaOH is added to an noting the number of equivalents of NaOH added per equivalent mixture of C U Y - ~and Httren+a, a equivalent of rare earth cation. K I I denotes the equilibbuffer region between pH 8.5 and PH 9.5 is ob- rium constant of reaction (11)

This value has been obtained from the equilibrium constant of reaction (111) : K m =

+

+

served during which the copper changes over from the “Y” to the “tren” CuY-2 H a t r e n + 3 z C u t r e n + 2+ HY-3 2 H + (IV) If a cation capable of binding the complexing agent “Y” is added during this neutralization, not only two, but three hydrogen ions can be neutralized in a buffer region which lies below the buffer region of (IV). In the presence of calcium ions the reaction takes place around PH 7, in the presence of manganous ions around pH 6 and in the presence of lead ions around PH 4. Reaction (II), with a rare earth metal as additional cation, has, as mentioned above, a buffer action between pH 4 and PH 5.5, the heavy rare earths being similar to lead and the first member of the series similar to zinc or manganous ion.

+

+

Solutions. A. 0.01 M NazCuY.-A solution of exactly 0.1 M CuCls was prepared. Ten ml. was titrated with approximately 0.1 M Na2H;Y after the addition of 5 ml. of ammonia and diluting t o 300 ml., using murexide as indicator. For this titration 10.20 ml. was needed. One hundred ml. of the original copper chloride solution was therefore mixed with 102.0 ml. of Na2HZY. This mixture is strongly acid because of reaction (V) CU+* HzY-2 +CuY-2 2H+ (V) Therefore, the solution was carefully adjusted t o PH 6 with NaOH and diluted to 1000 ml. Solution A then contains [Na] = 0.04, [CuY] =. 0.01, [Cl] = 0.02 B. 0.01 M “tren.”-One-hundredth molecular weight of the pure hydrochloride of the tri-( 8-aminoethy1)-amine base: tren.3HC1 was dissolved in water and diluted to 1000 ml. Solution B contains [Hltren] = 0.01, [Cl] = 0.03 C. 0.01 M MCl8.-Sufficient pure rare earth oxide’ was dissolved in a slight excess of N HCl to give a solution of 0.001 molar weight in MC&. A sample of this solution was titrated with NaOH using a glass electrode as indicator, and from a graph A pH/A ml. against ml. of the base added, the PH of a neutral solution was obtained. These p H values ranged from 5.2 for lanthanum down t o 4.70 for lutecium. Then the two solutions were recombined, adjusted t o this pH value and diluted to 100 ml. Solution C contains [MI = 0.01, [Na] Ei 0.01, [CI] Gi 0.04 Procedure.-Five solutions were made up in 100-ml. volumetric flasks, each containing 10 ml. of solution A, 10

+

.

+

(6) G. Schwarzenbach and E. Freitag, Hclo. Chim. Acta, 84, 1503 (1951). (7) The rare earths used in this research were separated on ion-exchange columns according to the method of Spedding, et at. (THIS JOURNAL, 69, 2786 (1947); 1%.2354 (19501, etc.), and each rare earth was spectrographically pure.

Calculation of K~l.-For the calculation of the equilibrium constant of reaction I1 from “a” and PH, the following equations have been used

+ a[Cutren] [Y], = 10-8 = [MY] [M]t [CUY]+ + [MI [MY] [trenlt = = [Cutren] + [Hztren] [Hlt = (3 - a ) X = [HI + 3[H*tren] [Cult = 10-3 = [CuY]

!7,

(dl (e) Equation (5a) is valid if no complexes of copper containing both “tren” and “Y” are present. Since “tren” is tetradentate and “Y” is hexadentate, each is capable of satisfying the coiirdination number of 4 which is typical for copper. Therefore, it seems unlikely that mixed complexes would form. The last term of (5a) takes account of the complex Cutren +2 as well as of the uncomplexed copper

+

a[Cutren] = [Cutren] [CUI The latter can be obtained from the concentration of H r tren+3 and the pH value [HIa x RHatren [CUI = [Cutren] X [Hstren] X Kcutran Therefore we obtain

Equation (5b) is true only if M+’ does not form a complex with the base “tren.” This assumption is justified because of the low PH ranges of the measurements. Denoting the stability constant of the complex as K M ~we~obtain ~ ~ for , the acid decomposition 3 H + Mtren+*I _ M+3 H3tren+* (VI)

+

+

the equilibrium constant KVI = Ka,trdKmtren (7) This shows that, with a maximum concentration of M+* and Hltren+s of 10-8, KMW- would have to be larger than 1012 in order to make the concentration of Mtren+a exceed 10-5, which is 1%of [MIt. Since very little is known about the ammonia complexes of the rare earth metals, it seems very improbable that the rare earth tren complexes should be more stable than Mntrene or F e t r e n e which have stability constants of 105.8 and 108.8 and certainly would not be nearly as stable as the cobalt( 11) complex (1019.*) and the zinc complex ( l O 1 4 ~ ) . 9 (8) In preliminary runs on La and Y ,equilibration rates of a few minutes to a few hours were tried. In the former it was evident that equilibrium while approached had not been reached and in the latter case the approach to equilibrium was 80 close that it seemed evident that in 24 hours equilibrium would be attained. (9) J. E. Prue and G. Schwarzenbach, H e k . Chim. Acto, 88, 963 (1950).

E. J. WHEELWRIGHT, F. H.

4188

$’If

G. SCHWARZENBACII

Vol. 75

TABLE I NazCUY, [H3treri]Cla,MCIa, EACH I N A CONCENTRATION O F 0.001i \ f A N D NaOIi CONCENTRATION OF “a” X 10 - 3 Jf. SUPPORTING ELECTROLYTE IS 0.09 M KCI

VALUES O F SOLUTIONS CONTAINING

Metal

M

SPEDUING AND

”a”

La+3 0.50 0.85 1.20 1.50 1.75 Ce+3 0.50 0.85 1.20 1.50 1.75 0.50 0.85 1.20 1.55 1.75 Xd+’ 0.51 0.84 1.20 1.50 1.75 Smt3 0.50 0.85 1.20 1.55 1.75 E u + ~ 0.50 0.85 1.20 1.50 1.75 Gd+3 0.50 0.85 1.20 1.55 1.75 Tb+3 0.50 0.85 1.20

9H

4.92 5.12 5.29 5.44 5.56 4.71 4.92 5.08 5.21 5.36 4.60 4.80 4.92 5.15 5.22 4.51 4.70 4.86 5.00 5.10 4.36 4.54 4.72 4.85 4.97 4.33 4.50 4.65 4.81 4.93 4.32 4.50 4.63 4.83 4.94 4.21 4.36 4.52 1.50 4.64 1 . 7 5 4.74

-log

-4~. -log KII

KII

13.05 13.00 13.00 13.02 13.01 12.29 12.39 12.35 12.33 12.40 12.04 12.02 11.87 12.05 11.98 11.70 11.72 11.68 11.68 11.62 11.17 11.18 11.23 11.14 11.21 11.04 11.03 11.01 11.09 11.09 11.00 11.03 10.93 11.07 11.12 10.46 10.46 10.53 10.51 10.47

13.02f0.06

log

KMY

14.72i0.06

12.35=k00.06 15.39fO.OG

11.9!3f0.Uti

15.75i0.06

11.68 f 0 . 0 6

16.06 f 0 . 0 6

11.19 2 ~ 0 . 0 7 l f i . 5 5 f 0 . 0 7

11.05i0.08 16.69f0.08

11.04fO.08

16.70f0.08

10.49 f O . l l

17.25i0.11

Equation (5c) is true if there is no uncomplexed “Y,” which would have to be present mainly in the form of the ion HzY-2. It can be shown that the concentration of this ion will be below byequations(l),(2) md(3)if weintroduce [Hatrenl = [Cutren] = [CuY] = 5 X Equation (5d) follows from the assumption that there will be no M t r e x ~ + ~The . complexing agent “tren” therefore can be present only in the form of the copper complex or uncomplexed total “tren” = [Cutren] [Hdren] [Hztren] [Htren] [tren] The last three terms of the equation are so small that they can be neglected in the p H ranges concerned. The total amount of acid hydrogen available is summarized in equation ( 5 e ) . With the equation (5), the calculation proceeds as follows: (e) gives immediately [Hjtren] from the experimental values of [HI and “a ” “> KC,,^), the complex MYwould be formed to 100Toduring the niixing process, and the neutralization of the solution with NaOH would not be itlflueri(c ( l IIV Ilir f w r u a t i o i i < J ( t i e r a e earth (omplex.

+

+

Sept. 5, 1953

STABILITY O F

RAREEARTHCOMPLEXES

WITH

ETHYLENEDIAMINETETRAACETIC ACID 4199

The equality of the total concentrations [MI$, [Cult, and [Y], leads to the expression (8) KVII= (%Cu)2/(100 - % C U ) ~ (8) From the equilibrium constant of reaction VLI, K M Y ,the formation constant of the complex MY- is KvlI KMY/KC~Y The error made in the estimation of the percentage of uncomplexed copper amounts to about f 2 % and will be roughly the same for all the metals investigated. By introduction of these limits in equations (8) and (9), the unranges for log K ~ I and I log KMY given in Table The Polarographic Method.-This method is certainty I1 are obtained. I t is seen that the polarographic method based on the fact that it is possible to measure the gives better values for the heavy rare earths and uncertain concentration of free uncomplexed copper ion values for the light metals. The potentiometric procedure Cufe polarographically in the presence of its and the polarographic method supplement each other. The uncertainty ranges of the values obtained by the two complex C U Y - ~ . This determination can be independent methods overlap up t o the element dysprosium carried out much more easily and more accurately (see Fig. 1). The heavy rare earth polarographic values in a solution which does not contain any chloride are definitely higher than the potentiometric values. It was ion. The salt KNOBhas therefore been used this felt that this difference was real, in spite of the fact that the potentiometric data for the heavy rare earths involve large time as supporting electrolyte. I n such a solution errors. If this is so, the difference must be due t o the change ~ C U Y - ~produces two sepa- in the supporting electrolyte, L e . , resulting from substitution a mixture of C U + and rated waves. The first of these has the shape of a of KNOI for KCl. These differences can be explained if we two-electron wave with El/, = -0.04 v. p s . S C E . assume that the rare earth cations form chloro complexes a slight extent in 0.1 M KC1, thus reducing the apparent (versus saturated calomel electrode). It is due to to formation constant of the complex MY-. If only the most the reduction of C U + to ~ the metal a t the surface simple of such chloro complexes with the composition of the mercury drop. The second is about as MC1+* are considered the stabilitv constant can be calcusteep as a one-electron wave and is due to the lated from the difference of the two values found for KMYin 0.1 M KCl and 0.1 M KN08. The value of 28 for the stareduction of the complex CUY-~,its Et/, being bility constant of LuC1+2 exdains auantitativelv the differ-0.32 v. Between the two waves the diffusion ence found for K M Y namely, , 101g.07-fromthe poientiometric current i d remains constant and is proportional to method and lOl9.66 from the polarographic method. The the concentration of free uncomplexed Cuf2 chloro complexes would be less stable for the other metals. It seems that the formation of chloro complexes to such a present. slight extent is not unlikely for the triple-charged cations Mixtures were prepared containing: copper complex of the rare earths. NazCuY (0.001), rare earth nitrate M(NOa)s (0.001), soIn order to prove this assumption, the polarographic dium acetate (0.01), acetic acid (0.01) and KNOs (0.091). method was applied to solutions containing chloride. The In these solutions equilibrium VI1 is set up. The polaro- same equivalent mixtures of CuY -a and M +3 were prepared gram was taken after the addition of two drops of basic with 0.1 M KC1 instead of KN03. The polarograms of fuchsin (0.2%) to suppress the maxima. The i d value for these chloride-containing solutions have a different appeareach rare earth nitrate solution was evaluated from these ance. Instead of the original two-electron wave with El/, polarograms at E = -0.17 v. A reference i d value for 100% = -0.04, we now have two one-electron waves due t o the uncomplexed copper was obtained from a polarogram of a stepwise reduction of uncomplexed copper Cu(11)-Cu( I), solution containing CuSOl instead of NazCuY. The per- and Cu(1)-Cu(0). The first step is situated to the left and centage of uncomplexed copper was calculated from the & the second to the right of the original two-electron wave value for each rare earth and the i d value for 100% uncom- observed in KNOJ. The first lies in the positive potential plexed copper. The results are summarized in Table 11. range (us. S.C.E.) and overlaps with the anodic oxidation of mercury: Hg(0)-Hg(I), so that it is impossible to determine TABLE I1 its height. The height of the second step is proportional to the concentration of uncomplexed Cu+a; it is impossible POLAROGRAPHIC DETERMINATION O F UNCOMPLEXED COPPER t o measure this height exactly because of partial overlapping IN AN EQUIVALENT MIXTUREOF NaaCuY AND RAREEARTH of the first step and some interference from the wave repreOF 0.001 M NITRATEM(NO&, BOTHAT A CONCENTRATION senting the reduction of the complex CuY-2. The error involved in the determination of the uncomplexed copper is I N KNOa, /k 0.1 therefore much larger in KC1 solution than in KNOa soluRare Uncomplexed tion. We estimate this error to be f 4 % . This is so large earth Cu,% log KVII 1% KYY that the method cannot be applied t o the lighter members La Uncertain .. . .. . .... .......... of the rare earth series. The results for the heavier memCe 4.1 f 2 -2.75 1 0 . 4 15.6 f 0 . 4 Pr 4.6 f2 -2.63 f .34 15.8 f .34 TABLE I11 Nd 5.9 1 2 - 2 . 4 f .25 16.0 f .25 POLAROGRAPHIC DETERMINATION 4p UNCOMPLEXED COPPER Sm 8.7 f 2 -2.1 f . 2 16.3 f . 2 I N AN EQUIVALENT MIXTUREO F NalCuY AND RAREEARTH EU 10.4 1 2 -1.91 f .15 16.5 .15 OF 0.001 M CHLORIDEMCl,, BOTH AT A CONCENTRATION Gd 11.9 1 2 -1.74 f .15 16.6 1 .15 IN KCl, p = 0.1 UnTb 24.1 f 2 -1.00 f .10 1 7 . 3 8 f .10 Rare complexed 17.75 f .OS 32.5 f 2 -0.63 f .OS Dy earth Cu, yo log KVII 1% K M Y HO 48.1 f 2 -0.07 f .07 18.31 i .07 Gd 16 f 4 -1.4 f 0 . 3 17.0 f 0 . 3 Er 55.0 f 2 +0.17 f .07 18.55 f .07 Tb 24 f 4 -0.98 f . 2 17.4 f . 2 Tm 68.8 f 2 +0.69 f .08 19.07 =k .OS Dy 33 f 4 - .60 f .16 17.78 f .16 Yb 76.3 f 2 +1.01 f 1.10 19.39 f .10 HO 40 f 4 - .33 f .15 18.05 f .15 Lu 81.3 f 2 +1.27 f 1.12 19.65 f .12 Er 50 f 4 .OO f .14 18 3 8 i .14 Y 2 8 . 1 zh 2 -0.82 f .OS 17.56 f .08 Tm 57 f 4 .24 =t 14 18 A2 f .I4 .50 f .16 18.88f . 1 R Yb 64 f 4 The equilibrium constant of reaction VI1 is calculated from the percentage of uncomplexed copper in the mixtures. LU 69 f 4 .69 i . I 7 19 07 :+= . 1 7

This neutralization would simply be reaction (111) and it would be impossible t o obtain any information as to the stability of the complex MY- from the PH measurements of such a titration. Increased stability of MY- will shift the equilibrium (VII) more and more to the right, increasing the uncertainty of the information obtained from the neutralization. The figures in Table I show that the results, L e . , the constants K M Y ,are fairly accurate up t o the element terbium. However, only very approximate values can be obtained for the last six elements. This was the reason a second method for the determination of the K M Yvalues was adopted.

I

.

*

+ +-

+

4200

E. J. WHEELWRIGHT, F. H. SPEDDING AND G. SCHWARZENBACH

Vol. 75

one metal to the next, of This difference is very large in comparison with the difference calculated from literature information, Crouthamel and Martinlo give the following values for the stability of the monooxalate complexes M(ox) + -__ [M(Ox)l = in the case of Ce lOEJ, Nd 107.2,

[MI 10x1

Yb 107.3

A rather large difference has been found between Ce and Nd and a very small difference exists between Nd and Yb, the ionic radii of which are considerably different from each other. From the data in this paper, i.e., from the PH values of the dilute (approximately 0.01 M ) rare earth chloride solutions,approximate values of the stability constants of the hydroxy complexes MOH+2can be calculated = concn. of rare earth salt [HI = hydrogen ion concn. of this sol. Kw = ion product of water

c

Equation (10) gives the stability constant of the hydroxy complex La(OH)+2 as 106.6and for LU(OH)+~ as 108.6, the difference being a factor POTENTIOMETRIC WLUE WITH of 10. UNCERTAINTY RANGE. Larger differences between individual rare earths must be expected in POLAROGRAFHIC VALUE IN KN03 cases of chelating agents in compariWITH UNCERTAINTY RANGE son to simple ligands such as OHPOLAROGRAPHC VALUE IN KCI groups. A polydentate complexing WITH UNCERTAINTY RANGE agent possesses a number of ligand atoms which are donated to the metal cation during complex formation. Each of these will be bound more I firmly by the cation of a heavier member, in comparison to the cation Lo Ce R Nd 61 Sm ELI Gd Tb Dy Ho Er Tm Yb Lu of a lighter member. All of these differences are cumulative in arriving ATOMIC NUMBER. Fig. 1.-Logarithms of complex formation constants versus atomic number. a t the stability constants of the two chelate complexes, the logarithms of bers, giving rise t o uncomplexed copper of more than 10% which are proportional to the free enGgies to be in the mixtures, are summarized in Table 111. These re- gained during the addition of the whole polydentate sults are now consistent with the potentiometric measurements except in two ci~ses,H o and Er, where the errors group to the two cations to be compared. In the case of a simple complex such as M(OH)+2 one possibly have been underestimated. In order t o recheck the assumption of the formation of water molecule is replaced by OH- during complex rare earth chloro complexes, the Yb value was redetermined formation. In the case of MY- probably all the with perchlorate as the supporting electrolyte. This value was identical with the value reported with the nitrate elec- water molecules of the hydrated cation are expelled trolyte. and replaced by the atoms of the hexadentate Discussion of the Results agent Y-'. This explains the comparatively large I n Fig. 1, a smooth curve has been drawn through differences between the individual rare earth the points representing the better data, using the cations. 2. There is a large increase of KMYfrom La+8to potentiometric values for the light and the polarographic values for the heavier rare earths. The Ce+a. This large jump is paralleled by a similarly rapid decrease of the ionic radius from La to Ce. following points are of interest. 1. The difference in KMYbetween the individual 3. The curve of log KMY shows an irregularity metals is astonishingiy large. This difference (10) C. E. Crouthamel and D. S. Martin, THISJOURNAL, 72, 1382 corresponds to a factor of or, as a mean from (1950); 78. 669 (1951).

Sept. 5, 1953

METALLATED DYECOMPLEXES

around gadolinium which is not paralleled by the curve of the radii. It is true that there is an irregularity in the radii curve a t erbium if the published radii are plotted against atomic number, but this may be accidental and due to errors connected with measuring the radii to 10.01 A. unit. The break in the curve of the log KMYobserved a t the point of gadolinium (substantiated by the change in slope when the “a” values as shown in Table I are plotted against PH for the individual rare earths) might be explained as follows: the carboxylate groups of the anion Y-4 are bulky groups and will have difficulty in finding enough room around the rare earth cation during COordination. This difficulty will increase as we go up the series because of the decrease in size of the

(CONTRIBUTION FROM THE

420 1

cation. There should be a critical point after which no longer all four, but only three, of the -COO- groups can be coordinated. This must cause a break in the curve log KMY versus atomic number. With gadolinium and the lighter members below it, the anion Y-4 will probably function as a hexadentate group. Going up the series from lanthanum to lutecium, the growing steric hindrance mentioned will prevent a large increase in stability, until a t the point of gadolinium the complexing agent can no longer function as a hexadentate group, but only as a pentadentate group. From this point upward, the stability increases again steadily as the ionic radius becomes smaller. A new steric hindrance does not seem to take place in the range gadolinium to lutecium. AMES,IOWA

RICHARDSON CHEMISTRY LABORATORY O F TULANE UNIVERSITY]

Studies of Metallated Dye Complexes. 11. Copper@) Complexes with 2-Carboxyphenyl-azo-j3-naphtholand 2-Carboxyphenyl-azo-6-naphthylamine BY HANSB. JONASSEN AND JOHN S. WILSON RECEIVED MARCH2, 1953 The azo dyes, 2-~arboxyphenyl-azo-@-naphthol and 2-~arboxyphenylazo-@-naphthyIamine were treated with copper( 11) ion. A one t o one reaction ratio was established in both cases by the method of continuous variation and by conductometric titrations. Evidence was advanced supporting the postulated structure involving the displacement of a hydrogen ion from each ortho substituted group of the dyes. On the basis of spectrophotometric measurements a monohydroxy addition complex was postulated for the copper( II)-2-carboxyphenyl-azo-@-naphtholcompound and a dihydroxy complex for the copper( 11)-2-carboxyphenyl-azo-,%naphthylamine.

Introduction photometric and conductometric methods that the Many metal lakes of the azo dyes have been pre- unoccupied positions in the copper(I1) coordination pared but few studies of these lakes are reported sphere of copper(I1) azo dye complexes may be in the literature. The lakes are formed between occupied by free hydroxy groups. A previous coordinating metal ions and azo dyes substituted in the ortho position by groups 4.0 such as the carboxy, hydroxy, or amino. The metal ion is postulated to react with the dye by replacing x3.0 hydrogen ions in the substituent groups and coordinating to the azo group.‘ h: It is well known that 2.0 the introduction of certain .$ metal ions to an azo dye, 2 used for dyeing cotton, produces better colors and w greater resistance to fading. This increased stability may 0.0 be explained on the basis of coordination. The dye may 200 250 300 350 400 450 500 550 600 not occupy all the positions Wave length, mp. in the coordination ‘phere Fig. 1.-Absorption spectra of 2-carboxyphenyl-azo-~-naphthol, 0 and Cu(11)-2Of the ion‘ These un- carboxyphenyl-azo-@-naphthol,0. Temperature = 30”; 1-cm. matched corex cells; occupied ~ositionsmay be 95% ethanol solutions. occupied by hydroxy groups of the cellulose fiber thus creating a chemical bond study2of the copper(I1) complex of o,o’-dihydroxybetween the fiber and the mordant dye. azobenzene has been reported, verifying the postuThis study has attempted to show by ‘pectro(2) H. B. Jonassen, Mae M. Cook and J. S. Wilson, THISJOURNAL, +

8

(1) M. Elkins and L. Hunter, J . Chcm. Soc., 1598 (1985).

18, 4688 (1961).