The following apers were ven in the &her Award
Fisher Award Symposium Chelometric Titrations Fisher Award Address CHARLES N. REILLEY Department o f Chemistry, University of North Carolina, Chapel Hill, N. C.
S
the advent of the classical papers of Schwarzenbach (2, IS), the role of EDTA in volumetric analysis of. metal ions and for masking purposes has been rapid and impressive, particularly during the last decade. Table I summarizes in periodic table form the elements determinable by EDTA titration. The solid lines indicate the metal ions which form stable complexes with EDTA and have been determined by direct or back titration procedures. The dotted lines encircle those species which are determined by an indirect procedure-e.g., precipitation of sulfate ion with excess barium ion and subsequent back titration of residual barium ion. Other chelons, such as the polyamines and other aminopolycarboxylic acids, have found specific applications where greater selectivity is needed. I n the present paper, it is not possible to review the entire scope of chelometric INCE
2751 5
titrations, and emphasis will, therefore, be given to selected areas, attempting to acknowledge the contributions of my former students and associates. INDICATOR END POINTS
The most popular method of end point detection is certainly by way of use of chemical indicators, usually dyestuffs which have complexing ability toward metal ions. Indicators should have certain general properties such as appropriate stability toward the metal ion in question, suitable color differences in the metallized and free forms, and sui& able rates of reaction. Sharpness. One characteristic of a suitable indicator is the sharpness of the end point transition. By this is meant the extent of conversion from metallized to free indicator form as titrant is added, the system being assumed to be in equilibrium a t all points. Clearly, this is determined
Table I. Elements Determinable by EDTA Titrations 7------
ILi
Be
_ _ _ A
I
I
Fr Ra ** *Lanthanides [La Ce Pr NdlI Pm I[Sm Eu Gd Tb Dy Ho Er L
Tm Yb Lul
**Actinides Ac
1298
Pa
Np
ANALYTICAL CHEMISTRY
Ipu/ Am
Cm Bk Cf
Es Fm Md
No
103
by the thermodynamics of all interacting species in question. A large number of factors need to be considered, the particular metal ions present, the chelon titrant employed, the indicator used, the auxiliary complexing agents (used to control the pH and, a t times, to solubilize the metal ion), pH, ionic strength, and temperature. In attempting to simplify this situation, the ionic strength and temperature will be assumed to be 20" or 25" C. and the ionic strength 0.1; many of the stability constants in the literature (16) refer to these conditions, and this assumption, therefore, permits use of these reported constants provided no unsuspected mixed complexes exist. In considering the net effect of these various parameters, Schmid (7) has proposed a useful indicator sharpness index, symbolized by AI and A*, which may be readily read from a pM-pH or pY-pH diagram. A typical pM-pH diagram is illustrated in Figure 1. Such diagrams summarize the various competitive effects in an easily understood manner. Line A, for example, indicates the concentration of free metal ion as a function of pH, taking into account the effect of pH and auxiliary complexing agents. Line C represents the concentration of free metal ion after an equimolar excess of titrant has been added-i.e., a situation beyond the end point. This curve is influenced by the (Y factor, and mixed chelate formation sometimes must be considered. Line B represents the concentration of free metal ion present in solution when the indicator exists in equal amounts in free and metallized forms; the shape of this curve is influenced by the (Y factor for the indicator and possible mixed complex formation with buffer constituents. Line D represents the concentration of free metal ion at the equivalence point in the titration and is
exactly midway between lines A and C. The end point indices, Al and Az, have the meanings indicated in the figure. Schmid found that the extent of indicator change upon adding increments of titrant is governed entirely by the values of A1 and Az. For suitable indicator titrations, A1 Az should generally be 6 or larger, and the symmetry of the color change is governed by the ratio A1/Az with the symmetrical color change occurring when Al = Az. A recent paper by Still and Ringbom (16) illustrates the use of the A indices in photometric titrations. Color Quality of End Point. After the competitive equilibrium has been elucidated and optimum conditions (pH, buffer, etc.) established-Le., by the AI, Az indices-a remaining problem concerns the ability of the eye to detect the conversion of the indicator from metallized to free form. For example, one may have a sharp indicator end point (where the transition from metallized to free form of the indicator occurs within a fraction of the drop of the titrant) yet the overall quality of the titration is hopeless if the indicator color transition is simply from one shade of red to another shade of red! Thus, some objective method of characterizing the color quality of indicator transitions was desired. For this purpose, Flaschka, Laurent, and Laurent (6) studied the possible application of tristimulus colorimetry. Included in this investigation was consideration of the number of chromaticity steps crossed in the indicator transition, the ability of a standard observer to remember previous colors, the relative greyness (the dirtiness) of the indicator color as well as the possibility of systematically screening an indicator so as to yield a perfectly grey color at the end point. An objective scheme using visible absorbance spectra was found for elucidating these quantities. For dilute color indicators, the calculations are fairly simple, absorbance data being used in place of the transmittance data in the common tristimulus calculations. The resulting points were termed complementary color points since these points fell on the tristimulus colorimetry diagram in areas of color complementary to the actual color of the indicator. These complementary color points are invariant with concentration, and mixtures of two dyestuffs, for example, will have a complementary color point falling on a straight line connecting the complementary color points for the individual dyestuffs. This leads to a great simplification in treating mixtures of dilute colorants. The ability of the eye to detect a n indicator end point is enhanced by adding to the mixture a screening color of suitable characteristics.
.1 (1)
+
Final
The screening color to be added is that which will cause the color at the end point to be perfectly grey. When this is done, the.color of the solution prior to the end point (Color I) will be perfectly complementary to the color present beyond the end Doint (Color 11). The problem securing the proper amount and type of screening color($ remains. After an extensive study of various inert dyestuffs, a set of, four suitable ones were given. By knowing the complementary color point present a t the end point (secured from compleOx
$0.3
Hg
-0-3 Cu
Bi
(8) found that sophomore students using such screened acid-base indicators, were able to adjust solutions to a grey point very reproducibly, with the pH of solutions being within *0.03 pH units of the designated value. AMALGAM EXCHANGE
I n order to extend the applications of EDTA titrations, Scribner (14) made use of oxidation-reduction principles involving metal amalgams as indicated in the following scheme:
+ M(Hg)
-0.9
-0.6 ’
3
PbCd
M+”
+ Red -1.2
V V S S.C.E. .
(2)
Zn
mentary tristimulus colorimetry of the metal indicator and the free indicator in conjunction with AI and A2), the selection and amount of screening color($ could be calculated. Indicator color changes, which had been considered very poor when unscreened, were found to be useful when appropriately screened by this procedure. The screening principle can be utilized for acid-base titrations, and this permits achieving grey at a desired pH value. In a trial experiment, Smith
Here any species, designated, by Ox, which is reduced by a suitable metal amalgam to Red, can in principle be analyzed via EDTA titrations. In this procedure, the sample containing 0%is shaken with an appropriate amalgam, and an equivalent quantity of metal ion is formed which is then titrated with a suitable chelon. As any survey of organic polarography will indicate, a number of organic species are reducible within the potential range of metal amalgams. Selectivity may sometimes
PH Figure 1.
A typical pM-pH diagram VOL 37, NO. 11, OCTOBER 1965
1299
be obtained by appropriate choice of the potential. This is achieved in a coarse way through the choice of a metal for the amalgam. Fine control of the potential where needed is then obtained by the use of auxiliary complexing agents and pH. In principle, this system is somewhat analogous to controlled potential electrolysis in that the amalgam and conditions chosen establish the required controlled potential, and the number of equivalents of the species 0 2 reduced is determined via titration. A distinct advantage of the amalgam method is that of time. By shaking the amalgam, it is broken up into many fine droplets (with subsequent large surface area) and since these travel rapidly through solution the “convection currents” are very large. As a result, the half life of reduction for a solution species can be as little as 30 seconds. Thus, in 300 seconds (10 half lives), the reduction may be assumed quantitative. The method has been applied for reduction of nitroorganic compounds, quinones, and those metal ions which are readily reducible but not easily titrated with EDTA [thallium(I)1. The method has also been applied to certain mixtures of metal ions. For example, in the titration of a lead and manganese mixture, it was known that cyanide would not mask either one of the two. However, by shaking the mixture with a zinc amalgam, the lead is completely reduced, and zinc ions are formed in its place. The zinc ions can subsequently be masked by cyanide, and the manganese ion, which is not reduced by the zinc amalgam, can then be titrated with EDTA. By titrating a separate aliquot of the mixture, the total amount of lead and manganese is determined. Several schemes of this type were developed. THERMODYNAMICS
Certain chelating agents, such as the polyamines (9) and EGTA (10, l a ) , permit selective titrations primarily because of their relative stability constants toward the metal ions in question. For example, the stability constants, log K(Ca-EGTA) = 10.7, log K(MgEGTA) = 5.4, log K(Ca-EDTA) = 10.7, log K(Mg-EDTA) = 8.8, suggest the utility of EGTA for the selective titrations of ,calcium in the presence of magnesium. This differentiating b e havior is of some interest, and Wright and Holloway (SO)investigated the heats and entropy changes for most of the common metal ions and chelons in order to understand in more depth the relationship between the structure of these chelons and their complexing ability. The heats of hydration of metal ions vary from approximately 300 to 500 kcal. per mole whereas the heats of metal-complex formation in 1300
0
ANALYTICAL CHEMISTRY
aqueous solution are seldom greater than 10 kcal. per mole. Thus, the replacement of waters of hydration by ligand atoms of the chelating agent represents a relatively small but important energy effect. The fine structural features of the chelons which cause or remove ring strain, alter configurational entropy, etc., are, therefore, of importance. Interestingly, the heat of formation of magnesium-EDTA is endothermic (AH = 3.1 kcal. per mole); the stability of this complex, therefore, results not from the greater energy of the magnesium chelate bonds over those of the hydrating water molecules, but rather from the entropy increase accompanying chelate reaction (AS = 40 entropy units). The entropy increase arises primarily from an increase in translational entropy accompanying the liberation of the numerous waters of hydration as indicated by the reaction: Mg(H20),+2 Y(H20),-4 -, MgY(H20),-2 (n m - z)HzO. In fact, for all of the various metal complexes with aminopolycarboxylate chelons, the contribution of the heat of reaction is generally less than that of the entropy change (standard state in moles per liter). Thus, entropy can be considered as the main driving force in the metal chelon reactions, especially in dilute solutions. In the case of cyclohexanedinitrilotetraacetate (CyDTA), the entropy increase is appreciably larger than for EDTA and is attributed in part to the lower configurational entropy difference between the free chelon and its metal complexes. The heat of complexation is also larger for CyDTA than for EDTA; hence, both thermodynamic properties contribute to the greater stability of the CyDTA complexes.
+
+ +
KINETICS
The titration of dilute metal ions with chelons using indicators for end point detection places great demands on rates of reaction. For example, consider the end point reaction MZn Y MY In (3) where MZn corresponds to the metal indicator and Y corresponds to the chelon titrant. Assuming the titrant to be 10-3M1 a solution volume of 50 ml., an indicator concentration of lO+M, an end point color change with one drop of excess titrant (5 X 10-2 ml.), then a t the end point the concentration of MZn is 1 x 10-6M, and the concentration of Y is 1 X 10-6M. For a color change rate equivalent to 10% color change in one second, the value of the secondsrder rate constant must be equal to 1 x 104M-’ second-’. This is a rather high rate constant for multidentate chelate exchange reactions. Aikens and Rogers (1) and Crus (4), therefore, investigated a number of
+
-
+
factors influencing the rates of typical reactions pertinent to chelometric titrations. The demands for direct titrations and for back titrations are different, and they are summarized accordingly. Direct Titrations. The end point reaction, Equation 3, must be fast for suitable direct titration. Titration a t elevated temperatures is sometimes necessary. Additional factors that govern this rate include the structure of the indicator and chelon, solution conditions, and the nature and oxidation state of metal ion. For the indicator, generally the larger the number of ligands the slower will be the rate of reaction. Also, if these ligand groups are tied to a rigid structure, the rate of reaction will be slower. The type of ligand atom is also important; carboxylate ligand-metal bonds are made and broken more rapidly than metal-nitrogen ligand bonds. For the chelon, the influence of its effective stability with the metal is often unimportant. The number, nature, state of protonation, and steric availability of ligand atoms in the chelon molecule are controlling. Sometimes a change in solvent can enhance the rate of reaction. The rate of the exchange reaction is generally higher a t low pH and a t high pH, the reaction being slowest a t an intermediate pH. At low pH the mechanism of exchange generally involves an acid-catalyzed SN, reaction (where the proton attack liberates the free metal ion which then reacts with the chelon). At intermediate and high pH, the reaction is SN2; the rate of reaction is more rapid a t high pH because the ligand atoms of the attacking chelon are not protonated. Sometimes the addition of auxiliary complexing agents will speed up the overall exchange rate by providing a new and more rapid pathway somewhat analogous to the chelate chain reactions of Margerum. The type and oxidation state of the metal ion itself, of course, is extremely important. This is exemplified in the rates of H 2 0 exchange determined by NMR (3,19) :
+ HzO* * ki
Jfw(H~0)
M,,(HzO*)
Metal ion Mn +2 c u +2
+ HzO
H 2 0 exchange (kl) 3.1 X 10’ set.-' 2 X 108 (axial) 1 X lo4 (equatoria1)a c o +2 1.13 X loe Ni +Z 2 . 7 x 104 Fe +3 2 . 4 x 104 Fe +Z 3 . 2 X loe a More recent work by Connick indicates absence of appreciable equatorial exchange. From these data, it is seen that the HzO exchange reaction of Fe+a is approxi-
mately 100 times slower than the corresponding reaction with Fe+2. This tendency carries over qualitatively to many, if not most, other exchange reactions. On this basis the reactions with Mn+2 are expected to be more rapid than those of the other transition metal ions listed. Also, the smaller the ionic radius and the larger the oxidation state, the slower will be the rate of reaction. These features are certainly in accord with experience in chelometric titrations. Some evidence for the influence of pH was obtained by Cruz (4) from the rate of exchange of Pb-EDTA with an optically pure lJ2-propylenedinitri1otetraacetic acid (PDTA, Y*). ki
PbY
+ Y* k-
PbY*
+Y
Erio T
Blocked by : Ti+4, A+*, Fe+* Mn+3, Co+2 ! Ni+2, Cu+2,'Pt metals
Not blocked by : R.E.+3, Pb+2, Cd+2, Zn+2, Mn+*, Ca+*, Mg+2
Fe +3, Ni +2
Th+*,Bi+2, Sc+3, c u f2
&OH
Metalphthalein
Glycinethymol Blue
1
k-1 M-1 ki sec.-l 6 7.43 3.40 7 1.22 0.55 1.04 0.47 8 9 6.33 2.88 10.3 21 .o 9.57 11 25.5 11.6 11.7 11.25 25.7 13.2 11.5 29.0 These data indicate that the rate of the forward and backward processes are greatly influenced by pH, with the slowest rate of reaction occurring near pH 8. At low pH's (below pH 7), the reaction becomes first-order in respect to PbY (or PbY*) and zero-order in respect to attacking ligand Y* (or Y ) , and the rate of the reaction increases as the pH is lowered. I n the intermediate region, from 7 to 9, the reaction rate is second order-first order in respect to the lead complex and first-order in respect to the attacking ligand. The same is true a t pH values higher than 10.3, the enhanced rate constant at higher pH being attributed to the presence of nonprotonated form of the attacking chelon. The data cited were obtained in solutions containing sodium ion; hence, the transition region (which appears to be somewhat above pH 9) is lower than the pk, of EDTA or PDTA because of formation of the sodium complex of the attacking ligand. The structure of the indicator itself is of importance, some being more readily blocked by certain metal ions (see above right). It is seen that the rigid structure of Erio T causes a number of metal ions to block the indicator transition. Even the reaction of Mg-Erio T with EDTA is visibly slow. The situation is somewhat improved in the case of Metalphthalein by removing one of the glycinate arms, forming Glycinethymol Blue. Thus, Glycinethymol Blue may be utilized for the titration of copper whereas Metalphthalein is blocked by its presence. It is important, in con-
PH
Reactivity of Metal-Indicators with EDTA
particular pH employed, the number of protons attached to the chelating agent and their sites of attachment must also be borne in mind. Sometimes alteration of the oxidation state of the metal ion will lead to sui& able end points. For example, consider
siderations of this type, to separate thermodynamic blocking from kinetic blocking (1). The rate of attack of different chelons on the metal indicator complex, Cu-Erio R, illustrates the influence of the chelon structure :
Rate of Attack of Chelons (Y) on Cu-Erio R (CuZ)
Y
k
+ CuZ
Chelon, Y
2
+
+ CuY, pH 8.8
Structure 0
0
\ / N-N / \
EDTA
0 0
\
DTPA
log Ketf (CU Y)
120
17.4
45
19.5
40
17.2
0
0
I
0
N-N-N
/
/
0 0
\ N-0 /
EGTA
k, M-l sec.-l
\
0 0
-0-N
0
/ \
0
0.3
From these data, it is seen that the number of ligand atoms and the geometry of complexing agent is of greater importance than the corresponding effective stability of the resulting metal complex product. Because of the Enhancement of Mln
18.6
the following titration principle proposed by Pribil: (Table below). The rate of attack of EDTA on Fe(II1)XO is so slow that a suitable end point is not achieved. However, simply by adding a small quantity of Fe(I1) to the
+
Y Rate by Redox Reaction Xylenol Orange, Fe(lll), EDTA Fe(I1) Absent Fe(I1) Present 1. Fe(II1) XO + Fe(II1)XO 1. Fe(II1) XO + Fe(II1)XO 2. Fe(II1) Y + Fe(1II)Y 2. Fe(II1) Y + Fe(II1)Y 3. Fe(II1)XO Y + Fe(II1)Y XO 3. Fe(I1) Y + Fe(I1)Y 81OW 4. Fe(I1) Y Fe(III)XO-cFe(III) Y
+ +
+
+
+ + + + Fe(I1) + XO
fast
VOL 37, NO. 1 1, OCTOBER 1965
+
1301
solution being titrated, a sharp end point is obtained. The addition of ferrous iron leads to the formation of Fe(I1)Y near the end point. Fe(II)Y, which is a rather strong reducing agent (ff), reduces Fe(I1I)XO to Fe(II)XO, which then rapidly dissociates, Back Titration. When suitable kinetic conditions cannot be secured for direct titrations, advantage can often be taken of these slow kinetics in performing a back titration. Back Titration
M
M' In
++ Y(X8) Y + M'
MY
+In
--P
slow
MY M'Y
I
?+
+Y
Q-cH~I-c:
NUCLEAR MAGNETIC RESONANCE STUDIES
Nuclear magnetic resonance can provide information concerning the properties of the chelons and their metal complexes which is difficult to secure with other physical methods. Our studies have been far from exhaustive and much remains to be learned. Sites of Rotonation. Consider the structure of diethylenetriaminepentaacetic acid (DTPA):
I
-0zC-CHz I n this compound, there are eight basic sites, three nitrogens and five carboxylates. If one equivalent of protons is added to this species, the question arises: to which of the eight basic sites does this proton become attached? An approximate answer will be the central nitrogen group [2]. After a second equivalent of protons is added, where are the two acid protons? An approximate' answer is that the one proton is on one end nitrogen groupi.e., [I]-and the other is on the opposite nitrogen group [3]; thus, the addition of the second proton causes the proton attached to the central nitrogen group to vacate this position and move to the remaining unoccupied end nitrogen site, leaving the central nitrogen site vacant. While it may appear unusual for the addition of acid to cause deprotonation of a basic site, in this case it is reasonable ANALYTICAL CHEMISTRY
H,
0
M'ln MIn+ Y
I n back titrations, it is desired that the last reaction given be slow. Here again, the influence of the indicator and chelon structure, solution conditions, and metal ion and its oxidation state need to be considered but from the viewpoint of achieving a slow reaction between MY and I n . CyDTA is often a useful choice for this purpose.
1302
in C\COOH
Xf
In
CF,COOO
Figure 2.
NMR spectrum of benzylphenylmethylclmine
since the minimal electrostatic action between the positively protonated sites will be achieved by placing the two protons on the terminal nitrogen groups. Actually, the questions should be rephrased. For example, the question should be stated, ('After one equivalent of proton is added to DTPA, what percentage of the time does the proton spend on each of the eight basic sites?" The more basic sites will, obviously, be protonated for a longer percentage of
I
CH2-COzthe total time. We have computed that, after the addition of a single equivalent of proton, the terminal acetate groups are not protonated to any detectable extent, the terminal nitrogen groups [l] and [3] are protonated 26y0 of the time each, the central nitrogen [2] is protonated 41% of the time, and the central acetate group [4]is protonated 7% of the time (17). These values were determined through NMR measurements. In the case of DTPA, four different types of nonlabile protons exist, denoted as a, b, c , and d. Consider the two protons on the central acetate group (labeled d). The chemical-shift value of these protons will be most sensitive to protonation of the central nitrogen and the adjacent carboxylate group. It is found that the chemical shift of the d proton
moves downfield upon the addition of the first equivalent of protons and then moves back upfield as the second equivalent of protons is added. By studying a large number of model compounds containing a single basic site, Sudmeier (17) was able to establish empirically a substitutent constant table relating the structure of any chelon having any assigned protonation scheme with chemical-shift values for the nonlabile hydrogens. From these substituent constants, he calculated the relative number of possible protonation schemes which satisfied best the observed spectra. Certain chelons, such as trans-1,% cyclohexanedinitrilotetraacetic acid (CyDTA), l12-propylenedinitrilotetraacetic acid (PDTA), and meso-2,s butanedinitrilotetraacetic acid (BDTA) yield A B patterns for the glycinate protons (18). Also, robust metal complexes of EDTA which have octahedral or even highly distorted octahedral structures also exhibit AB patterns for the glycinate protons. In the latter case, the asymmetry is caused by the long lifetime of the tetrahedral nitrogen group, in which case the glycinate portion of the structure visualizes three different groups attached to the nitrogen. June Singer, in this laboratory, illustrated this principle with the relatively simple structure given in Figure 2. When benzylphenylmethylamine is placed in acid solution, the nitrogen is protonated but has a fairly short lifetime due to the weak basic strength of the nitrogen. The NMR in this case simply indicates two singlets, one for the methylene group of
area two and one for the methyl group of area three. I n order for the hydrogen to have a long lifetime on the nitrogen, we then chose a solvent, trifluoroacetic acid, which would be very weakly basic compared to water. After adding a small amount of perchloric acid, the spectrum for benzylphenylmethylamine was obtained and is shown in Figure 2. Here the methyl group, split by the nitrogen proton, is a doublet, and the methylene group is an ABX spectrum. To illustrate more clearly the nonequivalence of the methylene protons, deuterium was substituted for the acid hydrogen. The resulting spectrum for the methylene group is clearly an A B pat tern. Metal Complex. N M R measurements are often useful in elucidating the lability of the individual members of a multidentate complexing agent attached to a metal ion. Day ( 5 ) , in a study of metal-EDTA complexes, has shown that several situations may arise. First, the lifetime of both metal-oxygen and metal-nitrogen bonds may be short. In this case, the nonlabile methylene protons on the acetate group and the methylene protons on the ethylene portion of the molecule are each able to average. Since any possible multiplet caused by proton-proton splitting of the methylene hydrogens is collapsed, a simple spectrum is obtained. Such spectra indicate a high rate of internal exchange accompanied by rapid umbrella inversion of the nitrogen. If the metal ion has appreciable abundance of isotopes with spin one-half and a high degree of covalent bonding, metalproton splitting occurs and persists despite the short lifetime of the metalnitrogen bond and rapid inversion of the nitrogen because the nitrogen-metal bond always reforms to the same metal ion. Because the nuclei are still in their same spin states, the coupling is not relaxed. An example of this case is Pb-EDTA (PbzD7, I = l/z, occurs in a natural abundance of 21%). In a second situation where the metal-nitrogen bond has a long lifetime while the metal-oxygen bond lifetime is short, the acetate methylene protons exhibit A B splitting patterns. The two
ethylenic conformations (D and L ) can rapidly interconvert and average out any differences of the protons on a given ethylenic carbon. The two protons on a given acetate, even when the metaloxygen bond is broken, are not equivalent for the same reasons given above in discussion of benzylphenylmethylammonium ion. With a fixed metalnitrogen bond, the free acetate groups may rotate about the acetate-carbon nitrogen bond and may be depicted by various rotational-staggered configurations. These positions are not equivalent and are not necessarily equally abundant. An example of a metalEDTA complex where the long lifetime of the metal-nitrogen bond causes an A B splitting pattern of the acetate is cadmium-EDTA. Recently, Day has found that this situation pertains even for barium complexes when a complexing agent of fairly rigid geometry (CyDTA) is utilized. Where both nitrogen-metal bonds and oxygen-metal bonds have relatively long lifetimes, relatively complex spectra are obtained because the acetate groups themselves are not equivalent one to the other. In the case of EDTA, two different A B patterns are obtained, indicating the existence of two different types of acetate groups in Co-EDTA. Also, the rigidity of the structure causes the ethylenic protons to no longer average out. When one or two methyl groups are substituted for the ethylenic protons, each of the four acetate groups can be different [Co(III)-BDTA]. The Co (III)-1 ,2 - propylenedinitrilotetraacetic acid complex may be expected to occur in two forms, one with an equatorial methyl group (I) and one with an axial methyl group (11) (Figure 3). From the NMR spectra, it has been concluded that the structure with an equatorial methyl group is the predominant one. NMR also promises to have great utility in studying the rate of interchelate exchange. In favorable cases, the magnitude of the rates of exchange can be obtained. The exchange rate must fall within the NMR kinetic "window," and this rate can often be suitably adjusted and controlled by conditions such as pH, relative concentration of
I Figure 3.
Metai-l,2-propylenedinitrilotetraacetic
acid complexes
ligand and metal ions, and temperature, hence permitting production of sharp and broadened splitting patterns required for evaluation of the rates. From an analytical point of view, it is obvious that NMR measurements may permit quantitative measurements of certain metal ion mixtures where their NMR spectra are different. This same principle can be utilized for determining the isotope content-Le., Pb2*. The use of paramagnetic ions is frequently interesting. For example, if small incremental quantities of Cu(I1) ion are added to a 0.5M solution of ethylamine at pH 10, a t approximately 10-4M copper ion, the-CH2-peak will be selectively broadened and a t higher concentrations will eventually disappear, illustrating the influence of the paramagnetism of the nearby copper ion and rapid exchange rate. If the pH is lowered to 5, the broadening diminishes, and the splitting pattern reappears. This occurs because the concentration of unprotonated ethylamine, the only form which takes part in the exchange, is now quite small, and the rate of exchange is low. Similarly, if EDTA were added to the solution at pH 10, the ethylamine peaks reappear. showing that the copper ion is now tied up in a slow exchanging EDTA complex. LITERATURE CITED
(1) Aikens, D. A., Rogers, D. W., Reilley, C. N., J. Phys. Chem. 66, 1582 (1962). (2) . . Biedermann, W., Schwarzenbach. , G.., Chimia 2, 1 (1962). (3) Connick, R. E., Fiat, D. N., J. Chem. Phys. 39, 1349 (1963). (4) Cruz, C. J., Reilley, C. N., ANAL.
CHEM.,in press.
(5) Day, R. J., Reilley, C. N., ANAL. CHEM.36,1073 (1964). (6) Reilley, C. N., Flaschka, H. A., Laurent, S., Laurent, B., Zbid., 32, 1218 (19601. (7) Reilley, C. N., Schmid, R. W., Zbid., 31,887 (1959). (8) Reilley, C. N., Smith, E. M., Zbid., 32,1233 (1960). (9) Reilley, C. N., Vavoulis, A., Zbid., 31,243 (1959). (10) Sadek, F. S., Schmid, R. W., Reilley, C. N., Talanta 2 , 3 8 (1959). (11) Schmid, R. W., Reilley, C. N., ANAL.CHEM.28,520 (1956). (12) Schmid, R. W., Reilley, C. N., Zbid., 29.264 (1957). (13) 'Schwarzenbach, G., Biedermann, W., Bangerter, F., Helv. Chim. Acta 29, 811 (1946): (14) Scribner, W. G., Reilley, C. N., ANAL.CHEM.30,1452 (1958). (15) SillBn, L. G., Martell, A. E., "Stability Constants of Metal-Ion Com lexes," Special Publication 17, The Ciemical Society, London, 1964. (16) Still, E., Ringbom, A., Anal. Chem. Acta 33,50 (1965). (17) Sudmeier, J. L., Reilley, C. N., ANAL.CHEM.36,1698 (1964). (18) Sudmeier, J. L., Reilley, C. N., Zbid., 36,1707 (1964). (19) Swift, T. J., Connick, R. E., J . Chem. Phys. 37, 307 (1962). (20) Wright, D. L., Holloway, J. H., Reilley, C. N., ANAL. CHEM. 37, 884 (1965). \ - - - - , -
'
1. Equatorial methyl group II. Axial methyl group
VOL 37, NO. 1 1 , OCTOBER 1965
1303