- Email: [email protected]
Spectrophotometric and analogue derivative spectrophotometric determination of ultramicro amounts of cadmium with cationic porphyrins
0039-9140/82/070545-06$03.00/O
TdUnrU,vol.29.pp.54sto 550.1982
Copyright 0 1982Pergamon Press Ltd
Printed in Great Britain. All rights reserved
SPECTROPHOTOMETRIC AND ANALOGUE DERIVATIVE SPECTROPHOTOMETRIC DETERMINATION OF ULTRAMICRO AMOUNTS OF CADMIUM WITH CATIONIC PORPHYRINS HAJIME ISHII,*KATSUHIKO
SATOH, YASUHIROSAT~H
and HIDEMASAKoHt
Chemical Research Institute of Non-Aqueous Solutions, Tohoku University, Katahira, Sendai-shi, 980 Japan (Received
13 November
1981. Accepted 29 January
1982)
Summary-a,~,y,6-Tetrakis(l-methylpyridinium-3-yl)porphine n(3-MPy)P] and z$,,y,&tetrakis( I-methylpyridinium-4-yl)porphine n(4-MPy)P] have been found to react rapidly with cadmium to give coloured complexes in weakly alkaline media at room temperature. Simple and practical methods for the determination of cadmium at ng/ml levels by conventional and analogue derivative spectrophotometry have been proposed. The analogue method gives higher sensitivity. T(3-MPy)P gives higher sensitivity than T(CMPy)P. The interference of various foreign cations and anions has also been examined and in many cases eliminated or reduced. Adsorption of the porphyrins and their cadmium complexes onto the glassware, which is usually observed under the conditions of reaction and causes’ significant errors in the determination, can be suppressed almost completely by addition of fairly large amounts of a salt such as
sodium chloride. Water-soluble meso-substituted porphyrins are very useful as highly sensitive colour-producing reagents for metal ions because they possess Soret bands which have extremely large molar absorptivity (1 x IO’-6 x 1051.mole-‘.cm-1) and are more easily handled than water-insoluble reagents. However, there are also a few disadvantages. One is that, in general, the complexation reaction of the porphyrin with the metal ion in aqueous medium is very slow at room temperature. Hence several attempts have been made to accelerate it, including (a) heating,lw3 (h) addition of an auxiliary complexing agent such as pyridine or imidazole4 or a reducing agent such as hydroxylamine or ascorbic acid,5 and (c) utilization of the metal-substitution reaction with a cadmium, lead or mercury(H) porphyrin complex,’ which is very rapid even at room temperature (although in the case of the cadmium complex, addition of pyridine or imidazole is usually necessary to accelerate the complexation4). These attempts were found to be quite effective for acceleration of the complexation of the parphyrin (a) with many metal ions, (b) with cadmium or copper(H), and (c) with copper(H). cobalt(II), mangi-
nese(I1) and zinc. Another disadvantage is that the cationic porphyrin and its metal complexes are readily adsorbed on glass, especially in neutral and weakly alkaline media where the complexation reaction is more rapid than in acid. This causes significant errors in the determination of traces of metals. Therefore, acceleration of the complexation reaction and * To whom correspondence should be addressed. t Present address: Research Laboratory, Asahi Glass Co., Ltd., Hazawa-cho, Kanagawa-ku, Yokohama-shi, 221 Japan.
suppression of the adsorption are important in determinations with porphyrins. On the other hand, during a series of studies concerning analytical application of porphyrins, it has been found that the cationic porphyrins r&y,&tetrakis(l-methylpyridinium-3-yl)porphine JT(3-MPy)P] and a$,y,d-tetrakis(l-methylpyridinium-4-$)porphine [T(CMPy)P], (Fig. 1) form complexes directly and rapidly with cadmium, copper( mercury(II), lead and zinc at room temperature in neutral and/or weakly alkaline media without addition of any auxiliary complexing agent or reducing agent. Adsorption of these porphyrins and their complexes on glass can be suppressed by addition of a salt such as sodium chloride, nitrate or sulphate. On the basis of these findings, the spectrophotometric determination of cadmium with T(3-MPy)P and T(CMPy)P has been studied. The analogue derivative technique’.’ has been introduced to make the determination more sensitive. Thus simple and practi-. cal methods for the determination of cadmium at ng/ml levels have been developed, and these are described in this paper. EXPERIMENTAL Reagents
T(3-MPy)P was synthesized by the method described in an earlier paper’ and T(CMPy)P by the method of Pasternack, et cd9 All other chemicals used were of analyticalreagent grade. All solutions were prepared with distilled. demineralized water unless otherwise described. Apparatus
For measurements of the absorbance and the absorption spectrum, a Hitachi 139 spectrophotometer and a Hitachi 556 dual wavelength spectrophotometer respectively. were
HAJIME ISHII et al.
546
tion of cationic porphyrins and their complexes on glassware is thought to be to the electrostatic interaction between positive charges on the substituents in meso-positions of the porphine ring and negative charges on the glass surface, it was thought that an increase of the cation concentration in the solution by addition of acids, bases or salts would be effective in suppressing the adsorption. In fact. adsorption is not observed in acidic or strongly alkaline medium. However, in acidic medium, the cadmium-porphyrin complex tends to decompose, and in strongly alkaline T(3-MPy) P : R= medium even such metal ions as silver and calcium --c --N+ interfere with the cadmium determination because ‘CH, their complexation reaction with the porphyrin is also ’ + accelerated. As salts such as sodium, potassium or T(4-MPy)P : RN -CH, -c ammonium chloride, nitrate or sulphate seemed to minimize the adsorption, in this study sodium chlorFig. 1. T(3-MPy)P and T(CMPy)P. ide was added to give a concentration of about 0.4M used, the latter being used as an ordinary double-beam in the final solution, which made the adsorption neglispectrophotometer throughout all the measurements. To gible. If perchlorate or thiocyanate is present, the caobtain the derivative spectra a modified Hitachi 200-0576 tionic porphyrins and their complexes generally tend derivative unit (composed of two analogue differentiation to precipitate because of formation of ion-pairs. Adcircuits) was connected between the spectrophotometer dition of fairly large amounts of these salts, therefore, output and a Hitachi 057 X-Y recorder input. The details should be avoided. of this apparatus and the principles and characteristics of the analogue derivative spectrophotometry are described in Chloride as an axial ligand. In general, the absorp earlier papers.‘*’ tion spectrum of the metal-porphyrin complex is known to shift gradually and slightly with increasing Procedures Specfrophotometrywith T(3-MPy)P. Place an aliquot of pH or the concentration of a salt such as sodium the sample solution containing less than 8 fig of cadmium chloride or thiocyanate. Figure 2 shows a set of the in a 25-ml standard flask. Add 1 ml of 0.3M sodium tarSoret bands for a series of solutions containing identtrate, 1 ml of 0.3M sodium citrate, and 1 ml of 10W3% ical concentrations of cadmium-T(3-MPy)P complex triethylenetetramine solution or 0.5 ml of 1% dimethyland sodium sulphate but of varied pH, from which it glyoxime solution in ethanol as the masking agent, if necessary. After addition of 4ml of 1 x 10m4M T(3-MPy)P, is seen that the Soret band of the complex gradually adjust the pH of the solution to 9.5-11 by adding a suf- shifts to longer wavelengths as the pH rises, and an R
ficient volume of sodium borate-sodium hydroxide buffer solution. Then add 2.5 ml of 4M sodium chloride and dilute to the mark with water. After allowing to stand for about 5 mitt, measure the absorbance at 441 nm against a reagent blank, using l-cm glass cells. Spectrophotometrywith T(4-MPy)P. Place an aliquot of the sample solution containing less than 12 pg of cadmium in a 25-ml standard flask and treat as described above, but with T(CMPy)P instead of T(3-MPy)P, and absorbance measurement at 450nm (this wavelength is that of the absorption maximum when the complex is measured against a reagent blank, and differs slightly from that for measurement against water). Second derivative spectrophototnetry with T(3-MPy)P. Place an aliquot of the sample solution containing less than 380na of cadmium in a 25-ml standard flask. Add masking agent as in the spectrophotometric procedure, if necessary. Add 1 ml of 1 x 10m6M T(3-MPy)P solution, 1 ml of O.lM sodium hydroxide and 2.5 ml of 4M sodium chloride, and dilute to the mark with water. After allowing the solution to stand for about 5 min, record the second derivative absorption spectrum of the resultant solution, in the Soret region, against a reagent blank, using l-cm glass cells and the following conditions; circuit No. 6, scan-speed 150nm/min, and recorder sensitivity x 1. Measure the second derivative value (vertical distance from a peak to a trough or the base-line to a trough) on the chart. RESULTS AND DISCUSSION
Effect of addition of salts Suppression of adsorption. As the reason for adsorp-
Wavelength,
nm
Fig. 2. Absorption spectra of Cd-T(3-MPy)P complex as a function of PH. [Cd-T(3-MPy)P] = 5.7 x IO-‘M; 1NarSOJ = 0.5M; reference, water; pH-a 7.6. b 11.0, c 11.3, d 11.8, e 13.1, f 13.6.
Determination
isosbestic point exists at 44Onm over pH range 7.6-13.1. (In this case, sodium sulphate is added as an adsorption inhibitor which differs from chloride or thiocyanate and does not affect the spectrum of the complex.) A similar result was obtained for the T(CMPy)P complex. These spectral variations are thought to correspond to dissociation of a proton from a water molecule co-ordinated axially to the central metal ion (i.e., cadmium) of the complex, corresponding to the equations Cd(H20)2P4+ z$ Cd(H20)(OH)P3+
+ H+
k, = CCdW,WOWf’3+lCH+l d CC4W%P4+1
Cd(H20)2P4+ + x- = Cd(H20)XP3+ + Hz0
Table I. pk, of Cd(H)-porphyrin complexes and log k, of Cd(H)-porphyrin-Cl_ complexes at 25°C. p = 1.5 (Na,SOJ
(2)
Complex
pk.,
Cd-T(3-MPy)P Cd-T(4-MPy)P Cd(H,O);+
11.6 11.6 9.0*
log k, 1.5 1.7 1.4*
* Data from Stability Constants, The Chemical Society, London, 1964: pk,: 25°C. p = 3(HCIO, + NaCIO,); log k,: 25’C, p = 2 (NaCIO,).
(1)
where P and k, represent the non-dissociative part of T(3-MPy)P or T(CMPy)P and the acid dissociation constant, respectively. Further shift of the Soret band at higher pH (Fig. 2) suggests the deprotonation of a further water molecule co-ordinated to the central metal ion. but this was not studied in detail. However, by analysis of the results shown in Fig. 2 and for the T(4-MPy)P complex, according to the method used by Pasternack et al., lo the dissociation of a proton from each complex was ascertained and the acid dissociation constants for both complexes were determined (Table 1). A spectrophotometric titration in the Soret region of the cadmium(U)-T(3-MPy)P complex at pH 9 with sodium chloride was carried out and the result is shown in Fig. 3. The Soret band of the complex shifts gradually and slightly to longer wavelengths as the sodium chloride concentration is increased and an isosbestic point exists at 436.5 nm for the chloride concentration range 0-0.4M. This spectral variation is thought to be due to axial co-ordination of chloride (in general, the salt anion, X-), in accordance with the equations
[Cd(H20)XP3 ‘1 k1 = [Cd(H,0),P4+][X-]
547
of cadmium with cationic porphyrins
nation because of their extremely large molar absorptivities) shift slightly to longer wavelengths than those found in the absence of sodium chloride. This suggests interaction between the complexes and chloride ions as described above. Injuence of sodium chloride concentration and selection of analysis wavelength
As already described addition of sufficient sodium chloride is required to suppress the adsorption, and A,,,,, for the Soret band of the cadmium complex depends on the sodium chloride concentration. The selection of wavelength for the cadmium determination, therefore, needs to take both effects into consideration. Fortunately, the absorbance is practically unaffected over the chloride concentration range 0.2-l.OM for measurement at 441 nm in the determination with T(3-MPy)P and the range O.lIO.SM for measurement at 450 nm in that with T(4-MPy)P.
2.6
(3) (4)
where kl represents the formation constant of the ternary complex. Further shift of the Soret band at a higher chloride concentration region (Fig. 3) suggests the axial co-ordination of another chloride ion. A similar result was obtained for the cadmium-T(4MPy)P complex. The formation constants were determined in the manner used by Pasternack et aI.” and are also given in Table 1.
I.C
Absorption spectra
Figure 4 shows the absorption spectra of T(3-MPy)P, T(CMPy)P and their cadmium complexes under the conditions of the determination. The Soret, /I and a bands of the T(3-MPy)P complex lie at 441, 573 and 613nm. and those of the T(4-MPy)P complex at 448, 580 and 624nm, respectively. The Soret bands (which are important for the determi-
I
430
440
Wavelength,
450
nm
Fig. 3. Absorption spectra at pH 9 of Cd-T(3-MPy)P of clconcentration. complex as a function [Na,SO,] = 0.5M: [Cd:T(3-MPy)P] = 5.7 x IO-‘M; reference. water; [Cl-]--a 0. b 0.01, c 0.03, d 0.1. e 0.4. f 2.0. g 3.OM.
548
HAJIME ISHIIet a/.
0.06
/ 0.c )_ 600
500
Wavelength,
nm
Fig. 4. Absorption spectra of T(3-MPy)P (a), T(4-MPy)P (b) and their cadmium complexes (c, d) in 0.4M 1.6 x 10e6M, b,d nil; sodium chloride medium: pH 10.5; reference, water; [T(3-MPy)P]-a,c [T+MPy)P]-b, d 2.6 x tOe6M, a, c nil; [Cdl-a, b nil; c, d 5.3 x 10-6M.
Injuence of pH
Figure 5 shows that both porphyrins form the cadmium complexes rapidly and give almost constant absorbances in the pH regions 9.3-11.7 and 8.7-11.1 for T(3-MPy)P and T(CMPy)P respectively, in the presence of 0.4M sodium chloride. In these pH regions,
cadmium exists as a monohydroxo, Cd(H20)&OH)+, or a monochloro complex, Cd(H20)Jl+, or both (depending on pH), before it reacts with the porphyrins, which is thought to be one of the reasons for the fast complexation. Further, the pH ranges where almost constant absorbances are obtained are remarkably broad, whereas in the absence of sodium chloride no such pH regions are found. The reason 0.6 r
for this seems to be that the predominant complex formed in such pH regions in the presence of 0.4M
sodium chloride is Cd(H20)C1P3+, as may be deduced from the values of pka and k, in Table 1; further, the deprotonation of Cd(H20)C1P3+ is negligible because it occurs only in more strongly alkaline media. Thus, the complex species scarcely change in the pH regions where almost constant absorbances are obtained, and the adsorption of the porphyrins and their complexes is considerably suppressed by the presence of the chloride as already described. Influence of porphyrin concentration The porphyrins were found (by the mole-ratio method) to form a 1:l complex with cadmium, but about 100’~ excess of porphyrin is required to complete the complexation rapidly. To prevent decolouration of the porphyrin complex and allow for consumption of porphyrin by other metal ions, however, it is preferable to add fairly large amounts of the porphyrin. Stability of the complexes to light
DH
Fig. 5. Influence of pH. [NaCl] 0.4M; standing time 5 min; reference, reagent blank. (a) Cd-T(3-MPy)P system: [Cd] 178ng/ml; [T(3-MPy)P] 1.6 x 10-‘M; wavelength 441 nm. (b) Cd-T(4-MPy)P system: [Cd] 284ng/ml; [T(CMPy)P] 1.6 x 10e5M; wavelength 450 nm.
One of the present authors has already reported that solutions of the cadmium complex of the anionic porphyrin, ~$,y,&tetrakis(4_sulphophenyl)porphine or a,/?,y&tetrakis(4_carboxyphenyl)porphine, are ap preciably decolourized by light.4 Figure 6 shows that a solution of the cadmium-T(3-MPy)P complex is stable to light for at least 5 hr if excess of T(fMPy)P is present, whether sodium chloride is also present or not. On the other hand, with excess of cadmium, the absorbance decreases by only about 3% in 5 hr in presence of 0.4M sodium chloride, but by about 12%
Determination of cadmium with cationic porphyrins 0.6
0.2
1I 0
2
Table 2. Optimum ranges and molar absorptivi-
d
V-r-
3
Standing
ties of the recommended procedures
4
time,
549
5
6
hr
Fig. 6. Influence of light on stability of Cd-T(3-MPy)P
complex at pH 10.5. [T(3-MPy)P]-a, b 1.6 x lo-‘M, c, d 1.6 x IO-“M; [Cd(H)]-a, b 1.3 x lo-%, c,d nil, b. d 0.4M; wave6.3 x 10m6M; [NaCl]-a,~. length-a, c 443 nm*, b, d 441 nm; reference-a, b reagent blank, c, d, water. *The wavelength of the absorption maximum in sodium chloride medium.
in absence of the sodium chloride. Similar results were obtained for the T(CMPy)P complex. Thus the decolouration of the solutions of the complexes, which is attributed to a redox reaction, was found not to be as serious as expected, and it was suppressed almost completely by addition of excess of .porphyrin and appreciably by addition of sodium chloride.
Reagent
Optimum range, w
Molar absorptivities, I.&e-‘.cm-’
T(3-MPy)P T(CMPy)P
0.48 0.6-13
3.61 x 10’ 2.20 x lo5
of sodium tartrate, sodium citrate, ttiethylenetetramine or dimethylglyoxime, and the tolerance limits thus obtained are those shown in Table 3. The interference of zinc can be reduced by modifying the procedure as follows. The same volume of sample solution is placed in each of two 25-ml standard flasks. One sample flask is treated by the recommended procedure, and gives absorbance A,. To the other sample flask 2m! of O.lM EDTA are added after the cadmium has reacted with T(3-MPy)P to give a full colour, and the mixture is let stand for about 10 min, for the cadmium complex has to decompose completely (the zinc complex remains stable). Then, 2.5 ml of 4M sodium chloride are added, the solution is diluted to the mark with water, and the absorbance (A& is measured. The cadmium content is obtained by subtracting A2 from Al. With this modified procedure, up to 1Opg of zinc will not interfere.
Calibration graphs
Linear calibration plots were obtained by the recommended procedures, the weight of cadmium in the sample taken being given by Cd = 7.75A~g Cd = 12.7sA /.q
(6)
where A is the absorbance, and equations (5) and (6)
refer to T(3-MPy)P and T(4-MPy)P respectively. The optimum ranges for the cadmium determination and the molar absorptivities are summarized in Table 2. The relative standard deviation for 3.55 pg of cadmium determined by the T(ZMPy)P procedure was 0.4% (16 variates). interferences
The possible interference in the determination of 3.55 pg of cadmium by the procedure with T(3-MPy)P, which gives rather high sensitivity, was examined. Cations were added in the form of chlorides, nitrates or sulphates; anions were added as sodium or potassium salts. The limiting value of the concentration of foreign ions was taken as that which caused an error of not more than 3%. The results ate summarized in Table 3. Most anions do not interfere even when present in fairly large amounts, but most cations interfere because of their hydrolysis or their complexation with T(3-MPy)P when no masking agent is used. Interferences caused by cations except zinc, however, can be removed or reduced by addition
Sensitization by analogue derivative spectrophotometry
We have already reported that derivative spectrophotometry by use of an analogue differentiatibn circuit is extremely effective for increasing the sensitivity.‘.’ As an example, the second derivative spectrophotometric determination of cadmium with T(3-MPy)P is described here. Selection of conditions for measurement of the skond derivative spectrum. The second derivative spectrum of
the analyte is recorded and the vertical distance from a peak to a trough or from the base-line to a trough of the spectrum is measured. Since this distance (D) depends on both the time constant of the analogue differentiation curcuit (our apparatus has 6 circuits, with different time constants) and the scanning speed used, these need to be selected to give a well-resolved large peak (to give good selectivity and higher sensitivity in the determination). This is done on the basis of the breadth of the bands in the ordinary absorption spectrum. In general, a large time constant and/or a fast scanning speed should be used for ? broad band in the absorption spectrum. In Fig. 7 the second derivative spectra of the Soret band of a cadmium-T(3MPy)P complex solution measured with varying circuit number or scanning speed are shown; circuit No. 6 and a scanning speed of 150 nm/min are seen to be preferable for the cadmium determination (the circuit numbers increase with increasing time constant).
550
HAJIMEISHIIet al. Table 3. Influence of foreign ions Ions added
Tolerance limit
AI(IH)’ Ca(II)*, Mg(II)‘, V(IV)rU’*.V(V)@‘* Cr(VJ)@‘,Fe(II)t”‘. Ni(II)‘“’ Fe(II1)‘“’ Cu(II)‘d’ Ag(I), Co(II)““, Cr(III)u”, Hg(II)“’ Mn(II)fd’, Pb(II)‘“’ Zn(II) Br-*, Cl-*, NO;*, SO:-*, tartrate*. citrate* c10; I-, SCN-
1~Pcg 500 pg 1OOlcg 50 /rcg 1Opg 4pg <4flg 1OOmg 50 mg 1Omg
Cd(I1) taken 3.55 pg. * Maximum tested. (a) 1 ml of 0.3M sodium tartrate added. (b) 1 ml of 0.3M sodium citrate added. (c) 1 ml of 0.001% triethylenetetramine solution added. (d) 0.5 ml of 1% dimethylglyoxime solution (in ethanol) added.
nm
nm
0
Fig. 7. Influence of (A) circuit number and (B) scan-speed on the second derivative spectrum of the Cd-T(3-MPy)P complex. [Cd(II)] 236 ng/ml; [T(3-MPy)P] 1.2 x 10e6M, [NaCI] 0.4M, pH 10.5; cells 1 cm, reference water. Numerical values indicate 1st and 2nd differentiation circuit numbers in (A) and scan-speed (nm/min) in (B), respectively. The recorder sensitivity was set so that a derivative value of 1.0 corresponded to l.Ocm on the chart. In (A) the scan-speed was lSOnm/min; in (B) both circuits were No. 6. Calibration graphs. Linear plots of D vs. weight of cadmium were obtained up to 0.4 pg of cadmium, the equations being
Cd=4OODng for the peak-to-trough
(7)
measurements, and
Cd = 6OODng
Acknowledgement-This study was supported partly by a scientific research grant from the Ministry of Education, Japan to which our thanks are due.
REFERENCES
(8)
for the baseline-to-trough measurements, where D was the second derivative value obtained with the recorder sensitivity set so that a derivative value of 1.O gave a signal of 20 cm (i.e., full scale deflection) on the chart. Equation (8), although the sensitivity is lower, is preferable for the analysis of “real” samples because it is less affected by concomitant ions which give T-(3-MPy)P complexes with a Soret band near to that of the cadmium-T(3-MPy)P complex. As a second derivative value of 0.1 corresponds to a signal of 2cm on the chart, cadmium at the ng/ml level can be readily determined by the proposed method.
1. J. Itoh. T. Yotsuyanagi and K. Aomura, Anal. Chim. Acta, 1975, 74, 53. 2. H. Ishii and H. Koh, Talanta. 1977, 24. 417. 3. Idem, Nippon Kagaku Kaishi, 1978, 390. 4. H. Koh, K. Kawamura and H. Ishii, ibid., 1979, 591. 5. H. Ishii and H. Koh, Bunseki Kagaku, 1979, t8, 473. 6. H. Ishii, H. Koh and K. S&oh, Nippon Kagaku Kaishi, 1980, 1919. 7. H. Ishii and H. Koh, ibid., 1980, 203. 8. H. Ishii and K. Satoh, 2. Anal. Chem., submitted. 9. R. F. Pasternack, P. R. Huber, P. Boyd, G. Engasser, L. Francesconi, E. Gibbs, P. Fasella. G. C. Ventura and L. deC. Hinds, J. Am. Chem. Sot., 1972,94,4511. 10. R. F. Pasternack and M. A. Cobb, J. Inorg. Nucl. Chem., 1973, 35, 4327.
TdUnrU,vol.29.pp.54sto 550.1982
Copyright 0 1982Pergamon Press Ltd
Printed in Great Britain. All rights reserved
SPECTROPHOTOMETRIC AND ANALOGUE DERIVATIVE SPECTROPHOTOMETRIC DETERMINATION OF ULTRAMICRO AMOUNTS OF CADMIUM WITH CATIONIC PORPHYRINS HAJIME ISHII,*KATSUHIKO
SATOH, YASUHIROSAT~H
and HIDEMASAKoHt
Chemical Research Institute of Non-Aqueous Solutions, Tohoku University, Katahira, Sendai-shi, 980 Japan (Received
13 November
1981. Accepted 29 January
1982)
Summary-a,~,y,6-Tetrakis(l-methylpyridinium-3-yl)porphine n(3-MPy)P] and z$,,y,&tetrakis( I-methylpyridinium-4-yl)porphine n(4-MPy)P] have been found to react rapidly with cadmium to give coloured complexes in weakly alkaline media at room temperature. Simple and practical methods for the determination of cadmium at ng/ml levels by conventional and analogue derivative spectrophotometry have been proposed. The analogue method gives higher sensitivity. T(3-MPy)P gives higher sensitivity than T(CMPy)P. The interference of various foreign cations and anions has also been examined and in many cases eliminated or reduced. Adsorption of the porphyrins and their cadmium complexes onto the glassware, which is usually observed under the conditions of reaction and causes’ significant errors in the determination, can be suppressed almost completely by addition of fairly large amounts of a salt such as
sodium chloride. Water-soluble meso-substituted porphyrins are very useful as highly sensitive colour-producing reagents for metal ions because they possess Soret bands which have extremely large molar absorptivity (1 x IO’-6 x 1051.mole-‘.cm-1) and are more easily handled than water-insoluble reagents. However, there are also a few disadvantages. One is that, in general, the complexation reaction of the porphyrin with the metal ion in aqueous medium is very slow at room temperature. Hence several attempts have been made to accelerate it, including (a) heating,lw3 (h) addition of an auxiliary complexing agent such as pyridine or imidazole4 or a reducing agent such as hydroxylamine or ascorbic acid,5 and (c) utilization of the metal-substitution reaction with a cadmium, lead or mercury(H) porphyrin complex,’ which is very rapid even at room temperature (although in the case of the cadmium complex, addition of pyridine or imidazole is usually necessary to accelerate the complexation4). These attempts were found to be quite effective for acceleration of the complexation of the parphyrin (a) with many metal ions, (b) with cadmium or copper(H), and (c) with copper(H). cobalt(II), mangi-
nese(I1) and zinc. Another disadvantage is that the cationic porphyrin and its metal complexes are readily adsorbed on glass, especially in neutral and weakly alkaline media where the complexation reaction is more rapid than in acid. This causes significant errors in the determination of traces of metals. Therefore, acceleration of the complexation reaction and * To whom correspondence should be addressed. t Present address: Research Laboratory, Asahi Glass Co., Ltd., Hazawa-cho, Kanagawa-ku, Yokohama-shi, 221 Japan.
suppression of the adsorption are important in determinations with porphyrins. On the other hand, during a series of studies concerning analytical application of porphyrins, it has been found that the cationic porphyrins r&y,&tetrakis(l-methylpyridinium-3-yl)porphine JT(3-MPy)P] and a$,y,d-tetrakis(l-methylpyridinium-4-$)porphine [T(CMPy)P], (Fig. 1) form complexes directly and rapidly with cadmium, copper( mercury(II), lead and zinc at room temperature in neutral and/or weakly alkaline media without addition of any auxiliary complexing agent or reducing agent. Adsorption of these porphyrins and their complexes on glass can be suppressed by addition of a salt such as sodium chloride, nitrate or sulphate. On the basis of these findings, the spectrophotometric determination of cadmium with T(3-MPy)P and T(CMPy)P has been studied. The analogue derivative technique’.’ has been introduced to make the determination more sensitive. Thus simple and practi-. cal methods for the determination of cadmium at ng/ml levels have been developed, and these are described in this paper. EXPERIMENTAL Reagents
T(3-MPy)P was synthesized by the method described in an earlier paper’ and T(CMPy)P by the method of Pasternack, et cd9 All other chemicals used were of analyticalreagent grade. All solutions were prepared with distilled. demineralized water unless otherwise described. Apparatus
For measurements of the absorbance and the absorption spectrum, a Hitachi 139 spectrophotometer and a Hitachi 556 dual wavelength spectrophotometer respectively. were
HAJIME ISHII et al.
546
tion of cationic porphyrins and their complexes on glassware is thought to be to the electrostatic interaction between positive charges on the substituents in meso-positions of the porphine ring and negative charges on the glass surface, it was thought that an increase of the cation concentration in the solution by addition of acids, bases or salts would be effective in suppressing the adsorption. In fact. adsorption is not observed in acidic or strongly alkaline medium. However, in acidic medium, the cadmium-porphyrin complex tends to decompose, and in strongly alkaline T(3-MPy) P : R= medium even such metal ions as silver and calcium --c --N+ interfere with the cadmium determination because ‘CH, their complexation reaction with the porphyrin is also ’ + accelerated. As salts such as sodium, potassium or T(4-MPy)P : RN -CH, -c ammonium chloride, nitrate or sulphate seemed to minimize the adsorption, in this study sodium chlorFig. 1. T(3-MPy)P and T(CMPy)P. ide was added to give a concentration of about 0.4M used, the latter being used as an ordinary double-beam in the final solution, which made the adsorption neglispectrophotometer throughout all the measurements. To gible. If perchlorate or thiocyanate is present, the caobtain the derivative spectra a modified Hitachi 200-0576 tionic porphyrins and their complexes generally tend derivative unit (composed of two analogue differentiation to precipitate because of formation of ion-pairs. Adcircuits) was connected between the spectrophotometer dition of fairly large amounts of these salts, therefore, output and a Hitachi 057 X-Y recorder input. The details should be avoided. of this apparatus and the principles and characteristics of the analogue derivative spectrophotometry are described in Chloride as an axial ligand. In general, the absorp earlier papers.‘*’ tion spectrum of the metal-porphyrin complex is known to shift gradually and slightly with increasing Procedures Specfrophotometrywith T(3-MPy)P. Place an aliquot of pH or the concentration of a salt such as sodium the sample solution containing less than 8 fig of cadmium chloride or thiocyanate. Figure 2 shows a set of the in a 25-ml standard flask. Add 1 ml of 0.3M sodium tarSoret bands for a series of solutions containing identtrate, 1 ml of 0.3M sodium citrate, and 1 ml of 10W3% ical concentrations of cadmium-T(3-MPy)P complex triethylenetetramine solution or 0.5 ml of 1% dimethyland sodium sulphate but of varied pH, from which it glyoxime solution in ethanol as the masking agent, if necessary. After addition of 4ml of 1 x 10m4M T(3-MPy)P, is seen that the Soret band of the complex gradually adjust the pH of the solution to 9.5-11 by adding a suf- shifts to longer wavelengths as the pH rises, and an R
ficient volume of sodium borate-sodium hydroxide buffer solution. Then add 2.5 ml of 4M sodium chloride and dilute to the mark with water. After allowing to stand for about 5 mitt, measure the absorbance at 441 nm against a reagent blank, using l-cm glass cells. Spectrophotometrywith T(4-MPy)P. Place an aliquot of the sample solution containing less than 12 pg of cadmium in a 25-ml standard flask and treat as described above, but with T(CMPy)P instead of T(3-MPy)P, and absorbance measurement at 450nm (this wavelength is that of the absorption maximum when the complex is measured against a reagent blank, and differs slightly from that for measurement against water). Second derivative spectrophototnetry with T(3-MPy)P. Place an aliquot of the sample solution containing less than 380na of cadmium in a 25-ml standard flask. Add masking agent as in the spectrophotometric procedure, if necessary. Add 1 ml of 1 x 10m6M T(3-MPy)P solution, 1 ml of O.lM sodium hydroxide and 2.5 ml of 4M sodium chloride, and dilute to the mark with water. After allowing the solution to stand for about 5 min, record the second derivative absorption spectrum of the resultant solution, in the Soret region, against a reagent blank, using l-cm glass cells and the following conditions; circuit No. 6, scan-speed 150nm/min, and recorder sensitivity x 1. Measure the second derivative value (vertical distance from a peak to a trough or the base-line to a trough) on the chart. RESULTS AND DISCUSSION
Effect of addition of salts Suppression of adsorption. As the reason for adsorp-
Wavelength,
nm
Fig. 2. Absorption spectra of Cd-T(3-MPy)P complex as a function of PH. [Cd-T(3-MPy)P] = 5.7 x IO-‘M; 1NarSOJ = 0.5M; reference, water; pH-a 7.6. b 11.0, c 11.3, d 11.8, e 13.1, f 13.6.
Determination
isosbestic point exists at 44Onm over pH range 7.6-13.1. (In this case, sodium sulphate is added as an adsorption inhibitor which differs from chloride or thiocyanate and does not affect the spectrum of the complex.) A similar result was obtained for the T(CMPy)P complex. These spectral variations are thought to correspond to dissociation of a proton from a water molecule co-ordinated axially to the central metal ion (i.e., cadmium) of the complex, corresponding to the equations Cd(H20)2P4+ z$ Cd(H20)(OH)P3+
+ H+
k, = CCdW,WOWf’3+lCH+l d CC4W%P4+1
Cd(H20)2P4+ + x- = Cd(H20)XP3+ + Hz0
Table I. pk, of Cd(H)-porphyrin complexes and log k, of Cd(H)-porphyrin-Cl_ complexes at 25°C. p = 1.5 (Na,SOJ
(2)
Complex
pk.,
Cd-T(3-MPy)P Cd-T(4-MPy)P Cd(H,O);+
11.6 11.6 9.0*
log k, 1.5 1.7 1.4*
* Data from Stability Constants, The Chemical Society, London, 1964: pk,: 25°C. p = 3(HCIO, + NaCIO,); log k,: 25’C, p = 2 (NaCIO,).
(1)
where P and k, represent the non-dissociative part of T(3-MPy)P or T(CMPy)P and the acid dissociation constant, respectively. Further shift of the Soret band at higher pH (Fig. 2) suggests the deprotonation of a further water molecule co-ordinated to the central metal ion. but this was not studied in detail. However, by analysis of the results shown in Fig. 2 and for the T(4-MPy)P complex, according to the method used by Pasternack et al., lo the dissociation of a proton from each complex was ascertained and the acid dissociation constants for both complexes were determined (Table 1). A spectrophotometric titration in the Soret region of the cadmium(U)-T(3-MPy)P complex at pH 9 with sodium chloride was carried out and the result is shown in Fig. 3. The Soret band of the complex shifts gradually and slightly to longer wavelengths as the sodium chloride concentration is increased and an isosbestic point exists at 436.5 nm for the chloride concentration range 0-0.4M. This spectral variation is thought to be due to axial co-ordination of chloride (in general, the salt anion, X-), in accordance with the equations
[Cd(H20)XP3 ‘1 k1 = [Cd(H,0),P4+][X-]
547
of cadmium with cationic porphyrins
nation because of their extremely large molar absorptivities) shift slightly to longer wavelengths than those found in the absence of sodium chloride. This suggests interaction between the complexes and chloride ions as described above. Injuence of sodium chloride concentration and selection of analysis wavelength
As already described addition of sufficient sodium chloride is required to suppress the adsorption, and A,,,,, for the Soret band of the cadmium complex depends on the sodium chloride concentration. The selection of wavelength for the cadmium determination, therefore, needs to take both effects into consideration. Fortunately, the absorbance is practically unaffected over the chloride concentration range 0.2-l.OM for measurement at 441 nm in the determination with T(3-MPy)P and the range O.lIO.SM for measurement at 450 nm in that with T(4-MPy)P.
2.6
(3) (4)
where kl represents the formation constant of the ternary complex. Further shift of the Soret band at a higher chloride concentration region (Fig. 3) suggests the axial co-ordination of another chloride ion. A similar result was obtained for the cadmium-T(4MPy)P complex. The formation constants were determined in the manner used by Pasternack et aI.” and are also given in Table 1.
I.C
Absorption spectra
Figure 4 shows the absorption spectra of T(3-MPy)P, T(CMPy)P and their cadmium complexes under the conditions of the determination. The Soret, /I and a bands of the T(3-MPy)P complex lie at 441, 573 and 613nm. and those of the T(4-MPy)P complex at 448, 580 and 624nm, respectively. The Soret bands (which are important for the determi-
I
430
440
Wavelength,
450
nm
Fig. 3. Absorption spectra at pH 9 of Cd-T(3-MPy)P of clconcentration. complex as a function [Na,SO,] = 0.5M: [Cd:T(3-MPy)P] = 5.7 x IO-‘M; reference. water; [Cl-]--a 0. b 0.01, c 0.03, d 0.1. e 0.4. f 2.0. g 3.OM.
548
HAJIME ISHIIet a/.
0.06
/ 0.c )_ 600
500
Wavelength,
nm
Fig. 4. Absorption spectra of T(3-MPy)P (a), T(4-MPy)P (b) and their cadmium complexes (c, d) in 0.4M 1.6 x 10e6M, b,d nil; sodium chloride medium: pH 10.5; reference, water; [T(3-MPy)P]-a,c [T+MPy)P]-b, d 2.6 x tOe6M, a, c nil; [Cdl-a, b nil; c, d 5.3 x 10-6M.
Injuence of pH
Figure 5 shows that both porphyrins form the cadmium complexes rapidly and give almost constant absorbances in the pH regions 9.3-11.7 and 8.7-11.1 for T(3-MPy)P and T(CMPy)P respectively, in the presence of 0.4M sodium chloride. In these pH regions,
cadmium exists as a monohydroxo, Cd(H20)&OH)+, or a monochloro complex, Cd(H20)Jl+, or both (depending on pH), before it reacts with the porphyrins, which is thought to be one of the reasons for the fast complexation. Further, the pH ranges where almost constant absorbances are obtained are remarkably broad, whereas in the absence of sodium chloride no such pH regions are found. The reason 0.6 r
for this seems to be that the predominant complex formed in such pH regions in the presence of 0.4M
sodium chloride is Cd(H20)C1P3+, as may be deduced from the values of pka and k, in Table 1; further, the deprotonation of Cd(H20)C1P3+ is negligible because it occurs only in more strongly alkaline media. Thus, the complex species scarcely change in the pH regions where almost constant absorbances are obtained, and the adsorption of the porphyrins and their complexes is considerably suppressed by the presence of the chloride as already described. Influence of porphyrin concentration The porphyrins were found (by the mole-ratio method) to form a 1:l complex with cadmium, but about 100’~ excess of porphyrin is required to complete the complexation rapidly. To prevent decolouration of the porphyrin complex and allow for consumption of porphyrin by other metal ions, however, it is preferable to add fairly large amounts of the porphyrin. Stability of the complexes to light
DH
Fig. 5. Influence of pH. [NaCl] 0.4M; standing time 5 min; reference, reagent blank. (a) Cd-T(3-MPy)P system: [Cd] 178ng/ml; [T(3-MPy)P] 1.6 x 10-‘M; wavelength 441 nm. (b) Cd-T(4-MPy)P system: [Cd] 284ng/ml; [T(CMPy)P] 1.6 x 10e5M; wavelength 450 nm.
One of the present authors has already reported that solutions of the cadmium complex of the anionic porphyrin, ~$,y,&tetrakis(4_sulphophenyl)porphine or a,/?,y&tetrakis(4_carboxyphenyl)porphine, are ap preciably decolourized by light.4 Figure 6 shows that a solution of the cadmium-T(3-MPy)P complex is stable to light for at least 5 hr if excess of T(fMPy)P is present, whether sodium chloride is also present or not. On the other hand, with excess of cadmium, the absorbance decreases by only about 3% in 5 hr in presence of 0.4M sodium chloride, but by about 12%
Determination of cadmium with cationic porphyrins 0.6
0.2
1I 0
2
Table 2. Optimum ranges and molar absorptivi-
d
V-r-
3
Standing
ties of the recommended procedures
4
time,
549
5
6
hr
Fig. 6. Influence of light on stability of Cd-T(3-MPy)P
complex at pH 10.5. [T(3-MPy)P]-a, b 1.6 x lo-‘M, c, d 1.6 x IO-“M; [Cd(H)]-a, b 1.3 x lo-%, c,d nil, b. d 0.4M; wave6.3 x 10m6M; [NaCl]-a,~. length-a, c 443 nm*, b, d 441 nm; reference-a, b reagent blank, c, d, water. *The wavelength of the absorption maximum in sodium chloride medium.
in absence of the sodium chloride. Similar results were obtained for the T(CMPy)P complex. Thus the decolouration of the solutions of the complexes, which is attributed to a redox reaction, was found not to be as serious as expected, and it was suppressed almost completely by addition of excess of .porphyrin and appreciably by addition of sodium chloride.
Reagent
Optimum range, w
Molar absorptivities, I.&e-‘.cm-’
T(3-MPy)P T(CMPy)P
0.48 0.6-13
3.61 x 10’ 2.20 x lo5
of sodium tartrate, sodium citrate, ttiethylenetetramine or dimethylglyoxime, and the tolerance limits thus obtained are those shown in Table 3. The interference of zinc can be reduced by modifying the procedure as follows. The same volume of sample solution is placed in each of two 25-ml standard flasks. One sample flask is treated by the recommended procedure, and gives absorbance A,. To the other sample flask 2m! of O.lM EDTA are added after the cadmium has reacted with T(3-MPy)P to give a full colour, and the mixture is let stand for about 10 min, for the cadmium complex has to decompose completely (the zinc complex remains stable). Then, 2.5 ml of 4M sodium chloride are added, the solution is diluted to the mark with water, and the absorbance (A& is measured. The cadmium content is obtained by subtracting A2 from Al. With this modified procedure, up to 1Opg of zinc will not interfere.
Calibration graphs
Linear calibration plots were obtained by the recommended procedures, the weight of cadmium in the sample taken being given by Cd = 7.75A~g Cd = 12.7sA /.q
(6)
where A is the absorbance, and equations (5) and (6)
refer to T(3-MPy)P and T(4-MPy)P respectively. The optimum ranges for the cadmium determination and the molar absorptivities are summarized in Table 2. The relative standard deviation for 3.55 pg of cadmium determined by the T(ZMPy)P procedure was 0.4% (16 variates). interferences
The possible interference in the determination of 3.55 pg of cadmium by the procedure with T(3-MPy)P, which gives rather high sensitivity, was examined. Cations were added in the form of chlorides, nitrates or sulphates; anions were added as sodium or potassium salts. The limiting value of the concentration of foreign ions was taken as that which caused an error of not more than 3%. The results ate summarized in Table 3. Most anions do not interfere even when present in fairly large amounts, but most cations interfere because of their hydrolysis or their complexation with T(3-MPy)P when no masking agent is used. Interferences caused by cations except zinc, however, can be removed or reduced by addition
Sensitization by analogue derivative spectrophotometry
We have already reported that derivative spectrophotometry by use of an analogue differentiatibn circuit is extremely effective for increasing the sensitivity.‘.’ As an example, the second derivative spectrophotometric determination of cadmium with T(3-MPy)P is described here. Selection of conditions for measurement of the skond derivative spectrum. The second derivative spectrum of
the analyte is recorded and the vertical distance from a peak to a trough or from the base-line to a trough of the spectrum is measured. Since this distance (D) depends on both the time constant of the analogue differentiation curcuit (our apparatus has 6 circuits, with different time constants) and the scanning speed used, these need to be selected to give a well-resolved large peak (to give good selectivity and higher sensitivity in the determination). This is done on the basis of the breadth of the bands in the ordinary absorption spectrum. In general, a large time constant and/or a fast scanning speed should be used for ? broad band in the absorption spectrum. In Fig. 7 the second derivative spectra of the Soret band of a cadmium-T(3MPy)P complex solution measured with varying circuit number or scanning speed are shown; circuit No. 6 and a scanning speed of 150 nm/min are seen to be preferable for the cadmium determination (the circuit numbers increase with increasing time constant).
550
HAJIMEISHIIet al. Table 3. Influence of foreign ions Ions added
Tolerance limit
AI(IH)’ Ca(II)*, Mg(II)‘, V(IV)rU’*.V(V)@‘* Cr(VJ)@‘,Fe(II)t”‘. Ni(II)‘“’ Fe(II1)‘“’ Cu(II)‘d’ Ag(I), Co(II)““, Cr(III)u”, Hg(II)“’ Mn(II)fd’, Pb(II)‘“’ Zn(II) Br-*, Cl-*, NO;*, SO:-*, tartrate*. citrate* c10; I-, SCN-
1~Pcg 500 pg 1OOlcg 50 /rcg 1Opg 4pg <4flg 1OOmg 50 mg 1Omg
Cd(I1) taken 3.55 pg. * Maximum tested. (a) 1 ml of 0.3M sodium tartrate added. (b) 1 ml of 0.3M sodium citrate added. (c) 1 ml of 0.001% triethylenetetramine solution added. (d) 0.5 ml of 1% dimethylglyoxime solution (in ethanol) added.
nm
nm
0
Fig. 7. Influence of (A) circuit number and (B) scan-speed on the second derivative spectrum of the Cd-T(3-MPy)P complex. [Cd(II)] 236 ng/ml; [T(3-MPy)P] 1.2 x 10e6M, [NaCI] 0.4M, pH 10.5; cells 1 cm, reference water. Numerical values indicate 1st and 2nd differentiation circuit numbers in (A) and scan-speed (nm/min) in (B), respectively. The recorder sensitivity was set so that a derivative value of 1.0 corresponded to l.Ocm on the chart. In (A) the scan-speed was lSOnm/min; in (B) both circuits were No. 6. Calibration graphs. Linear plots of D vs. weight of cadmium were obtained up to 0.4 pg of cadmium, the equations being
Cd=4OODng for the peak-to-trough
(7)
measurements, and
Cd = 6OODng
Acknowledgement-This study was supported partly by a scientific research grant from the Ministry of Education, Japan to which our thanks are due.
REFERENCES
(8)
for the baseline-to-trough measurements, where D was the second derivative value obtained with the recorder sensitivity set so that a derivative value of 1.O gave a signal of 20 cm (i.e., full scale deflection) on the chart. Equation (8), although the sensitivity is lower, is preferable for the analysis of “real” samples because it is less affected by concomitant ions which give T-(3-MPy)P complexes with a Soret band near to that of the cadmium-T(3-MPy)P complex. As a second derivative value of 0.1 corresponds to a signal of 2cm on the chart, cadmium at the ng/ml level can be readily determined by the proposed method.
1. J. Itoh. T. Yotsuyanagi and K. Aomura, Anal. Chim. Acta, 1975, 74, 53. 2. H. Ishii and H. Koh, Talanta. 1977, 24. 417. 3. Idem, Nippon Kagaku Kaishi, 1978, 390. 4. H. Koh, K. Kawamura and H. Ishii, ibid., 1979, 591. 5. H. Ishii and H. Koh, Bunseki Kagaku, 1979, t8, 473. 6. H. Ishii, H. Koh and K. S&oh, Nippon Kagaku Kaishi, 1980, 1919. 7. H. Ishii and H. Koh, ibid., 1980, 203. 8. H. Ishii and K. Satoh, 2. Anal. Chem., submitted. 9. R. F. Pasternack, P. R. Huber, P. Boyd, G. Engasser, L. Francesconi, E. Gibbs, P. Fasella. G. C. Ventura and L. deC. Hinds, J. Am. Chem. Sot., 1972,94,4511. 10. R. F. Pasternack and M. A. Cobb, J. Inorg. Nucl. Chem., 1973, 35, 4327.