C. G.

Ramsay
University of Aberdeen
I
1
EDTA Titration of Cadmium and Mercury

Cadmium(11) and mercury(I1) can he determined by rear-

tiun with excess of ethvlenediaminetetra-accticwid (EDTAI
and hack-titration with standard zinc(I1) solution, followed
by masking of the mercury(II), e.g. with iodide ( I ) , and con-
tinued hack-titration of the released EDTA. The principle of
the method is illustrated in Figure 1. These reactions are of
interest in the undergraduatr nnnlytical lahoratorysince they
demunstratc the use of masking toobtain drlectivitv in cum-
.~
nlexometrv. and narticularlv since exnerimental results aeree
~ ~

most impresiveiy with theutheoretical predictions fromihe

"conditional" ("apparent," "effective") stability constants, J
a concept introduced by Schwarzenbach ( 2 ) and developed Figure 1. Stickdiagram representing the cadmium(li)and mercury(l1)titration.
by Ringhom (3,4). (1)Amition of excess of EDTA. (2)back-tivation with zinc(ll),(3)second back-
tiration after masking mercury(1l) to release EDTA from HgEDTA complex. Cd
The Conditional Constant = (1H2H3):Hg = (3).
The degree of completeness (i.e. the quantitative nature)
of a reaction can be evaluated from the conditional stahilitv
constant. The thermodynamic equilibrium constant for a l:i [M'] = [MI + [MA] + .. . + [MA,O
complexation reaction, as occurs with EDTA
= [MI (1 + f: [ A I V ~ )
I=,

MtLtML
= [MI UM(A1
is defined by the law of mass action as where aMcA) is the side reaction coefficient or "alpha coeffi-
cient" for the reactions of A with M, and ,3i is the overall for-
mation constant of the comolex MAG'
However, the equilibria in real aqueous systems are never so
simple as to involve only metal and complexant, since protons
and hydroxide ions are always present, and since competing . .. .
If these formation constants and the concentration of A are
ligands (e.g. buffer components and masking agents) and iu- known, a can be evaluated. Corresponding alpha coefficients
terferine" metals mav also "eive rise to side reactions. Conse- for side reactions involving the ligand and the metal complex
quently, the conditional constant which takes into account can he defined.
all competing equilibria is defined as Consequently

where [M'I
. .. .
. . represents
. the total concentration of all species
containing M nut complexed tu I., where [L'Irepresents the side reactions invdlving ML,M, h d ~respectively.
, Ringhom
total concentration of L not hound tu the metal ion M, and (3.4) has discussed these concents and calculations in neater
where lh(l.'l is a function of the total concentrution of all detail.
species'con<aining hoth M and L.
Clearlv it is this conditional constant which is of analvtical Completeness of Reaction
importance since it dcfinrs the ratio of determinand and dc- 'I'he minimum value of K'req~tiredfor reaction to proceed
t ~ m n i n a nwhich
~ have taken r m t in the analscical reaction to to various degrees of complt~wnesswith \,ariou nmcentrations
those which have not. of reactants-can he readily calculated. For example, in an
The nrecise definition of IML'I denends on the oarticular EDTA titration where the total concentration of metal at the
reaction. In a complexometr~ctitiation it is usually~rrelevant equivalence point is 0.001 M, and the reaction is 99.9% com-
to the titration result whether ML is formed as such, or as a plete, the concentrations of free EDTA and metal not com-
mixed complex (e.g. the protonated complex MLH). In this
case [ML'] represents the total concentration of all species
containing hoth M and L. In other systems where a property
peculiar to the species ML is measured, the formation of a
&xed complex constitutes an interfering side reaction.
The concentration of free M is related to the concentration
plexed by EDTA are both 0.001 X
mum required value of the stability constant is
K' = 0.000999
(0.001 x 1or3l2
_
M. Hence, the mini-

lo,
Correspondingly, the minimum required value of K' increases
of total M by the formation constants of the MA, species as the concentration of titrant and titrand decrease, and de-
formed by the interfering ligand A. creases as the desired degree of completeness of the reaction
~ ~

decreases, and vice versa (see Fig. 2). The general equation
1 Alpho wefficirnti are sornrtlrnrp dcfinecl ar the inberic of thr
function dcscr~bedhere. In t h p authm'i upinion, a z ~ d erenrtwn
coefficient (a) which increases as the side reaction becomes more
predominant is the more appropriate. Further complications are in- ,-~- - .,
troduced hv the occasianal desimation of the side reaction coefficient where C is the total metal concentration a t the equivalence
as 0. Care must he exercised when material from differentsources is point and P is the fraction of total metal species cimplexed
studied. by EDTA.
714 1 Journal of Chemical Education
Figure 2. Dependence of the degree of completeness of the reaction(%R) on
lkg K' lor sclutions of fixed total metal cancenbstion (C) at Uw equivalence paint.
C = 0.1 M(1); 0.01 M(2); 0.001 M(3): 0.0001 M(4).

Figure 5. Conditional stability constants for EDTA complexes wllh cadmlum(ll)

in absence of iodide (1). In 0.1 M iodide (2), and in 2.0 M iodide (3):and with
mwcuy(ll) in absence of iadide (4). and in 0.1 M iadide (5)(log IC in 2.0 Mlodide
is always less than zero).

lndlcator Error
Resides analytical errors arising from the incomplcre re-
actionsof the metals with EDTA, it isalso ~ossibletocalculnre
the error associated with the color changeof the indicator by
consideration of the difference in the pM' values at the
equivalence point and at the indicator traosition point 1 4 , s )
It issufficient tonote here that sharp,accurateendlx~inrscan
be predicted and are obtained withxylenol orange and erio-
chrome black T in the titrations described. However, traces
of metal ions such as Cu(II), Co(II), Ni(II), AI(III), and Fe(II1)
must be avoided since they interfere by blocking eriochrome
black T (6).
0 1 .
0
. .
L
,
4
. . . . . .
6 S 10 I2
. 1
14
Reaction With Zn, Cd, and Hg
PH The conditional stability constants for the reactions of
Figure 3. Conditional stability constams for EDTA complexes with mercury(1l) EDTA with zinc(II), cadmium(II), and mercury(I1) are shown
in absence of NH3(I), in 0.2 Mlotal NHI (2). and in 2.0 M NH3 (3). in Figures 3-7 as functions of p H and concentration of buffer
(ammonia) or masking agent (iodide, bromide, chloride).
Whereas the calculation of K' for fixed conditions is relatively
simple, the repetitive calculations needed to obtain the data
in Figures 3-7 are best carried out by computer. The required
formation constants were taken from the com~ilationsin
Ringborn's monograph ( 4 ) with the exception of ;he stability
constant for the mixed Hg(I1)-EDTA-NHrcomplcx 17,
HgEDTA + NH3 = HgEDTA.NH3 log K = 6.3.
Although a mixed Hg(I1)-EDTA-hexamine complex has
similarlv been r e ~ o r t e d(7).
. it is ..
eenerallv acceoted that hex-
amine ~he~ameth~lenetetraminei is a i e a k c;mplwant (81
and itseffect on the formation of the metal-WTA complexes
should be slight since K' is increased hy formation of the
mixed complex, but decreased by the formation of the
metal-hexamine complex; full quantitative data are not
available.
Unless the titration conditions are adjusted correctly, the
wrong results will be obtained. The most suitable titration
conditions can be oredicted loeicallv as follows.
First, the feasklity of cor;lplehng Zn(II), Cd(II), and
Hg(I1) in ammoniacal solution is assessed. It is predicted from
Figures 3 and 4 that 0.001 M solutions of these metals will
react auantitativelv (>99.9%) with EDTA at DH 10 in 0.2 M
Figure 4 Condmonal staD lily conslants for EDTA complexes solh 2 nc( ) tn ammonia (log K' >.9). Even if theconcentrat~onof ammonia
absence of NH, ( I ) m 0 2 Mtola NH3(21 and n 2 0 M hH, (31 and warn caa- is raised considerablv. the errors should remain nealicible. In
mwn(ll) In absence of NHl (41, m 0 2 MNh, (51, and m 2 0 M hH, (61 the case of mercuG,it is the added stability of the mixed
Volume 54. Number 11, November 1977 / 715
Figare 6. Cond tional olabilily canslants tar EOTA complexes u Ihcadmium(l1) Figure 7 Conditional stability constants fn EOTA complexes with cadmium(l1)
in aosence ol bromlde (1). in 0.1 Morom de (2). and in 2.0 Moromlae (31:and in absence of chloride (1). in 0.2 Mchloride (2). and in 2.0 Mchloride (3):and
w i n mercurylll) n absence ol orom de (4). in 0.1 M oromos (5).and in 2.0 M wilh mercury(l1)in absence of chloride (4). in 0.2 M chloride (5).and in 2.0 M
bromide (6). chloride (6).

Dstermination of Cadmium and Mercury (0.2 mmole in totall

in a Mixed Solution Relatwe to the Determmation of Each
EDTA-ammine complex which permita quantitative reaction. in t h e Absence of the Other
If formation of this mixed complex is ignored, a value of log
Cd: Hg Masking Recovery Recovery
K' = 2 a t p H 10 (i.e. only 8% reaction) is obtained in 2 M (molar rafio) Agent of Cd 1%) of Hg 1%)
ammonia; the falseness of this prediction is readily checked
experimentally. Quantitatiw reaction with all three metals 1:3 iodide 100.5 (0.1) 100.1 (0.1)
is ~ l s nossiblr
o at DH5.5 in hexamine buffer. assumine that 1:l iodide 100.4 (0.1) 100.0 (0.1)
3:l iodide 100.1 (0.1) 100.1 (0.1)
the effects of the lhffer are negligible. 1:3 bromide 99.1 (0.4) 100.2 (0.1)
The predicted influence of iodide, bromide, and chloride 1:l bromide 99.6 (0.21 100.3 (0.2)
on the EDTA complexation reactions is now examined (Figs. 3:l bromide 99.7 (0.11 100.3 (0.11
5-71, Since zinc forms verv weak halide comulexes, the re- Standard deviations are given in parentherer.
duc;ion in the conditional constant of the zinc-~DTAcomplex
can safelv he nealtvted. even in 2 hi halide. A \,slue of lop. K'
= 0 correipondsto 0.1% reaction for a metal concentration of
are in complete agreement with the data in Figure 7. The ti-
0.001 M a t the equivalence point. At pH 10, mercury(I1) is tration conditions are, therefore, too critical for chloride to
quantitatively masked by 0.1 M iodide (log K' = -3.6) be recommended as a selective masking agent.
whereas EDTA still reacts quantitatively with cadmium(II), Experimental
even in the presence of 2 M iodide. Consequently, i t should L)isrillcd water and analytical grade reagents (apart from the in.
be possible to determine cadmium in the presence of mercury dicators) should be used throughout. Prepare a mixture of cadmi-
by masking the latter as tetraiodomercurate, [HgI4]2-: there um(lll and mercuryill, tn dissolving the metal nitrates in 0.05 M
is considerable latitude in the choice of pH and iodide con- nitric acid: the total metai concentration should be approximately
centration. With bromide or chloride as masking agent, very 0.01 M. Dissolve pure zinc metal in the minimum volume of8Mnitric
high concentrations (>2 M ) would be necessary to mask acid, and dilute to volume to give a standard 0.01 M zinc(I1)solution.
mercury quantitatively a t p H 10. Standardize a 0.01 M solution of the disodium salt of EDTA against
At pH 5.5, however, mercury should be quantitatively the zinc solution. Dissolve NH&l (70 g) in concentrated ammonia
masked by bromide a t concentrations greater than 0.1 M , solution (600 ml, sp.gr. 0.88), and dilute to 1I. Prepare a 1%dispersion
of eriochrome black T in powdered KC1, and a 0.5%aqueous solution
while the complexation of cadmium by EDTA remains of xylenol orange.
ouantitative a t bromide concentrations less than 0.5 M. Add35ml ofEDTAsolution to25mlofCdlHgsolution.After5min
Complete masking of mercury at pH 5.5 requires a chloride (the reaction between HefII) and EDTA is relativelv slow) add am-
concentration of at least 2 M (Fig. 7) while this concentration donia buffer (about 2 mfi'kd eriochrome black T iddicator. Titrate
permits very nearly quantitative reaction (99.7%) with cad- with zinc to the first perceptible permanent red-purple, then add K1
mium. but the DH and the chlorideconcentration arecritical. (2 g) and titrate to the second endpoint.
At such high ionic strengths, the data upon which the calcu- Instead of ammonia buffer, solid hexamine (1g) may be added to
lations of K' are hased are not strictly applicable. Moreover, raise the pH to 5.5, in which case xylenolorange isused as indicator.
the indicator color change may be adversely affected, as is the The solution is titrated with zinc to the first permanent pink, then
KBr (1.2 g) is added and the titration continued as before.
case with xslenol orange where the maximum p H value for a To prevent discharge of heavy metals into the environment, the
clear color change decreases as the ionic strength increases (9). residues from the titration should he collected and concentrated (ex..
Nevertheless, by back-titration of EDTA with zinc at p H 5.5, by evaporation to dryness) prior to appropriate disposal.
it is ~ o s s i b l invractice
e to determine cadmium and mercury
with'less than i% error when mercury is masked by 2 M Conclusions
chloride. If the concentration of chloride is decreased to 0.2 Both bromide and iodide yield satisfactory results (the
masking of mercury is incomplete and the endpoint is table), but iodide is the preferred masking agent in analytical
nrotracted and inaccurate, while at D H 5.0 and in 2 M chloride applications owing to the greater permissible latitude in the
medium, some of the cadmium is aiso masked. These findings reaction conditions.

716 / Journal of Chemical Education

 The close correlation found in these systems between ex- (2) Schwareenbseh.G.. "Die komplerometrisehe Turofian:2nd Ed., F.Enke, Stutfgart,
,.XL
perimental results and the predicted reaction feasibility is an (3) Ringbom. A., J. CHEM. EDUC.. 35.282 (1958).
excellent demonstration of the usefulness of conditional sta- (41 Ringbom, A,. "Complexation in Analytical Chemistry: Interaeienee, New York.
1963.
bility constants. (5) Ringbom, A., in "Indicators: (Editor: Bishop, E.), Pergamon Preas, Oxford, 1972, p.
9,w
Acknowledgment (6) Rpilley, C. N., Schmid, R. W., and Sadek, F. S., J. CHEM. EDUC.. 36, 555, 619,
(1959).
The author is indebted to Drs. R. A. Chalmers, I. L. Marr, (71 van dor Linden, W. E..and Beers, C., Talonto, 22.89 (1975).
and M. R. Masson for helpful discussion. (8) Grassino, S. L..and Hume, D. N., J.Inorg. Nuci Chem., 32,3112 i1970l.
(9) Flaschka, H., and Butcher,J., Microehem. J.,7.407 (1963): Chomiat-Anolyat, 54.36
Literature Cited (1965).
(11 "KampleromelrischeBe~limmungsmelhw'on
mif nlripkr:3.Aunage,E. Merck AG,
Darmstadt, p. 73.

Volume 54. Number 11. November 1977 / 717