In the Laboratory

Spectrophotometric Determination of Iron(III)–Glycine

Formation Constant in Aqueous Medium
Using Competitive Ligand Binding
Rajendra Prasad* and Surendra Prasad
School of Chemical Sciences, Faculty of Science and Technology, The University of the South Pacific, Suva, Fiji;
*prasad_re@usp.ac.fj

In his seminal work Ramette described spectropho- Materials and Methods

tometric determination of the formation constant of
monothiocyanatoiron(III) (1), which, in presence of another Equipment
weakly coordinating ligand, L−, is likely to exist in equilibrium All absorbance measurements were made on a PerkinElmer
with Fe(L)2+. In such a mixed-ligand environment the color- Lambda 16 UV–vis spectrophotometer. The Fe(III)–SCN−
developing Fe(III)–SCN− reaction can be used as an indicator complex exhibits maximum absorption in the visible region
reaction to determine the formation constant of Fe(L)2+. This between 450 and 460 nm, hence inexpensive glass or PVC cu-
competitive binding forms the basis of this study. The formation vettes were used. pH of the solutions was measured with a glass
constants of metal–ligand interactions play decisive roles in electrode connected to an Orion pH meter. The molar absorp-
the speciation of metal ions in the environment as well as their tion coefficient and formation constant of the Fe(III)−SCN−
selective uptake, transport, and storage by the living organisms. complex were calculated using the Solver tool in Excel on a
Various complex-forming agents compete for binding with the Dell PC and were subsequently used to calculate the formation
soluble iron(III) present in extremely low concentration in constant of the Fe(III)–glycine complex.
the living cell as well as outside in the environment (2–4). The
degree of iron complexation with a particular ligand critically Chemicals
depends on its formation constant, which in turn depends on the The common chemicals anhydrous ferric chloride, potassium
oxidation state of iron, the pH of the medium, and the nature thiocyanate, glycine, potassium permanganate, nitric acid, sulfuric
of the ligand itself. acid, and granular zinc used were purchased from Ajax Finechem
Amino acids form complexes with Fe(III) that are known (Australia). Stock solutions of ferric chloride in 0.2 M HNO3 and
to exist in equilibrium with the free metal ions. Although glycine and potassium thiocyanate in distilled deionized water
quantitative information about the iron–amino acid forma- were prepared (see the online material for details).
tion constants is desired in many studies, the complexes are less
studied due to inherent complications arising from formation Experimental Procedure
of oxo-bridged polynuclear complexes (5–8) and also from hy-
drolysis of the metal ion. Thus even the formation constant of Standardization of Iron(III)
iron(III)–glycine, Fe(Gly)2+, the simplest α-amino acid, has not The Fe(III) concentration in the stock solution was deter-
been extensively studied (9–13). Polarography and pH-metric mined titrimetrically using the Byars and McCreary method
methods employed relatively high iron(III) concentrations and (14). The Fe(III) was initially reduced to Fe(II) with zinc gran-
highly acidic conditions to avoid hydrolysis and to suppress the nules in dil H2SO4 and was subsequently titrated with standard
formation of polynuclear complexes (10). Because of high proto- KMnO4 until a faint pink color persisted.
nation constants, glycine was often used in large excess (11, 12).
Recently, differential pulse cathodic voltammetry (DPCV) on Determination of Formation Constants
a static mercury drop electrode (SMDE) was used to evaluate Standard solutions were prepared by mixing 1.0–6.0 mL
overall formation constants, β2 and β3, of the iron(III)–glycine of Fe(III) and 1.0 mL of KSCN stock solutions and diluting
system that employed low iron(III) concentration (25 μM) and to 25 mL with 0.2 M HNO3. Similarly the test solution was
moderately alkaline pH (6.36–9.05) (13). prepared by adding 2.0 mL Fe(III), 1.0 mL KSCN, and 1.0 mL
Since these methods require sophisticated experimenta- glycine solutions and diluting the mixture with 0.2 M HNO3
tion and advanced computational methods that are difficult to to 25 mL. All solutions were found to have pH between 0.7 and
emulate in undergraduate laboratories, we herein report a simple 0.8. UV–vis spectra of the solutions were subsequently recorded
and convenient spectrophotometric method for the laboratory between 350 and 600 nm against a blank solution containing
determination of the iron(III)–glycine formation constant via 1.55 × 10‒3 M Fe(III) in 0.2 M HNO3. The absorbance values
competitive ligand binding. Solver, Excel’s suite of analysis tools, of the solutions were recorded at λmax of 457 nm. Five replicate
was used for the dynamic optimization of the molar absorption analyses were carried out independently by five different groups
coefficient and the equilibrium constant. The program finds a re- of students.
quested optimal value in the target cell, by adjusting the values in
the adjustable cells, as specified by the user, while subjecting the Hazards
model to a number of restrictions or constraints. This method
was tested in the class and was found to work satisfactorily. It can Potassium thiocyanate causes skin irritations, iron(III)
easily be extended to determination of formation constants of chloride, potassium permanganate, and acids (oxalic, sulfuric
other non-phenolic amino acids and organic bases with iron(III) and nitric) are corrosive and may cause burns to skin and other
and gives good quantitative comparisons. body tissues.

494 Journal of Chemical Education • Vol. 86 No. 4 April 2009 • www.JCE.DivCHED.org • © Division of Chemical Education
 In the Laboratory

0.8
Results and Discussion f
e
For simplicity sake, it was assumed that in the mixed-ligand d
0.6 c
systems only these two equilibria existed: b

Absorbance
a
Fe3 SCN  Fe(SCN)2 (1) 0.4

Fe3 Gly  Fe(Gly)2 (2)

0.2

Hence, the molar absorption coefficient, ε, and formation con-

stant, KFe-SCN, for the Fe(III)–SCN− complex were calculated 0.0
from the absorbance data, which were subsequently used to 400 450 500 550 600

calculate the formation constant for the Fe(III)–glycine com- Wavelength / nm

plex, KFe-Gly. Figure 1. Absorbance spectra of Fe(SCN)2+ solutions containing
3.02 × 10−4 M SCN− and increasing concentrations of Fe(III):
Formation Constant of Fe(SCN)2+
(a) 7.75 × 10−4 M, (b) 1.55 × 10−3 M, (c) 2.32 × 10−3 M, (d)
In very dilute solutions of low pH (<1.0) the 1:1 complex 3.10 × 10−3 M, (e) 3.87 × 10−3 M, and (f) 4.65 × 10−3 M in
Fe(SCN)2+ is likely to be present as the dominant species presence of 0.2 M HNO3.
as shown in eq 1 (1). However, at higher concentrations of
SCN− additional complexes such as Fe(SCN)2+, Fe(SCN)3,
and Fe(SCN)6 3− are also likely to be present (15). They all are
blood red colored and have intense absorption bands between From the Beer–Lambert law
400­and 500 nm. The UV–vis spectra of Fe(SCN)2+ at different
Fe(III) concentrations are shown in Figure 1. A  F l Fe(SCN )2 (5)
Although the Fe(III)–SCN − formation constant,
KFe-SCN, is a function of ionic strength of the solution (16), it where A is the absorbance at λmax, ε is molar absorption coef-
can be written in terms of concentrations, as shown in eq 3, if ficient at λmax, and l = 1 cm. Equation 4 can be rewritten as
ionic strength is kept constant:
A
Fe(SCN )2 Fl
KFe-SCN  (6)
KFe-SCN  (3) A A
Fe 3
SCN  CM  CL 
Fl Fl

From the ICE table for the equilibrium reaction, the formation Equation 6 can be solved in Excel using the Solver optimization
constant for Fe(III)–SCN− can thus be written tool or it can be expanded and rearranged (15) to

KFe-SCN  CM C L CM C L 1
 (7)
A Fl F l KFe-SCN
Fe(SCN )2 (4)
2 2 The left-hand side of eq 7 contains only experimental data that
C M  Fe(SCN ) C L  Fe(SCN )
can be plotted against (CM + CL) to obtain a straight line. Values
of ε and KFe-SCN are thus obtained from the slope and intercept
where CM and CL are the total Fe(III) and the total SCN− con- of the straight line, respectively. The experimental values of CM,
centrations, respectively. Since in the dilute solutions, aquated CL, and absorbances for the different solutions along with the
Fe(III) and SCN− ions show negligible absorption between 400 calculated parameters obtained from the graphical method and
and 500 nm, only Fe(SCN)2+ is present as an absorbing species. by using Solver optimization are listed in Table 1.

Table 1. Experimental Values and Calculated Data for the Fe(III)–SCN– Complex
ε/(L mol–1 cm–1) KFe-SCN/( L mol–1)
CM / CL / (CM + CL )/ (CMCL /A)/
Absorbance
(10–3 mol L–1) (10–4 mol L–1) (10–3 mol L–1) (10–6 mol2 L–2) Graph Solver Graph Solver

0.775 3.02 0.1617 1.08 1.45

1.55 3.02 0.3076 1.85 1.52
2.32 3.02 0.4252 2.63 1.65
4830 4690 179 183
3.10 3.02 0.5053 3.40 1.85
3.87 3.02 0.5880 4.18 1.99
4.65 3.02 0.6242 4.95 2.25
Note: The shaded area designates the calculated data.

© Division of Chemical Education • www.JCE.DivCHED.org • Vol. 86 No. 4 April 2009 • Journal of Chemical Education 495
 In the Laboratory

2.4
Calculation of Formation Constant of Fe(Gly)2+

(10ź6 mol 2 Lź2)

Using the Graphical Method 2.2

The plot of (CM + CL) versus CMCL/A for the Fe(III)−SCN−

2.0
complex (Figure 2) gave a slope of 2.069 × 10‒4 mol L‒1 and in-
tercept of 1.161 × 10‒6 mol2 L‒2. These values when substituted 1.8
in eq 7 gave ε = 4830 L mol‒1 cm‒1 and KFe-SCN = 179 L mol‒1.
Then, assuming that only two equilibria, eqs 1 and 2, existed 1.6

CM CL
in the mixed-ligand solution, concentrations of uncoordinated

A
1.4
Fe(III), Fe(Gly)2+, and uncoordinated glycine were calcu-
lated. The mixed Fe(III)–glycine–SCN − solution, having 0 1 2 3 4 5 6
initial [Fe(III)] of 1.55 × 10‒3 M, [SCN−] of 3.02 × 10‒4 M, (CM á CL) / (10ź3 mol Lź1)
and [glycine] of 5.05 × 10‒4 M in 0.2 M HNO3, exhibited
Figure 2. (CM + CL ) versus CMCL /A for the Fe(III)–SCN– complex.
an absorbance of 0.2969 at λmax 457 nm. This corresponded
to [Fe(SCN)2+] of 6.15 × 10‒5 M. Thus the concentration of
uncoordinated SCN− was calculated to be 2.41 × 10‒4 M. Sub-
sequently using the equilibrium constant KFe-SCN of 179 L mol‒1,
eq 4 gave free [Fe(III)] of 1.43 × 10‒3 M. By subtracting the
combined concentration of free Fe(III) and Fe(SCN)2+ from the deviation values equal and to <5% of their respective calculated
total Fe(III) concentration, the concentration of Fe(Gly)2+ was average ε and KFe-SCN values. Again both standard deviations
calculated to be 6.08 × 10‒5 M and that of the uncoordinated were sequentially minimized. After two iterations average values
glycine to be 4.44 × 10‒4 M. Substituting these concentrations of ε = 4690 ± 80 L mol‒1 cm‒1 and KFe-SCN = 183 ± 3 L mol‒1
into eq 8 gave the formation constant of Fe(III)–glycine, were obtained. As seen in Table 2 both standard deviations are
KFe-Gly, 95.9 L mol‒1. <1.7% of the calculated mean values of ε and KFe-SCN.
Using the optimized average value of ε = 4690 L mol‒1 cm‒1
Fe(Gly)2 for the Fe(III)–SCN− complex in the mixed-ligand Fe(III)–gly-
KFe-Gly  cine–SCN− solution and eq 5, the concentration of Fe(SCN)2+
C M  Fe(SCN )2  Fe(Gly)2 t (8) was calculated to be 6.32 × 10‒5 M. Thus with initial concen-
2
trations Fe(III) of 1.55 × 10‒3 M, SCN− of 3.02 × 10‒4 M, and
C Gly  Fe(Gly) glycine of 5.05 × 10‒4 M in 0.2 M HNO3, the concentrations
of uncoordinated SCN− and Fe(III) were calculated to be
2.39 × 10‒4 M and 1.44 × 10‒3 M, respectively. From the differ-
Using Solver in Excel ence of combined concentrations of uncoordinated Fe(III) and
The Solver tool in Excel facilitates direct calculation of ε Fe(SCN)2+ from total [Fe(III)], the Fe(Gly)2+ and uncoordi-
and KFe-SCN from the absorbance data of the solutions. Initially nated glycine concentrations were calculated to be 4.24 × 10‒5 M
an arbitrary value of ε was first assigned to the data and the and 4.63 × 10‒4 M, respectively. Substituting these concentration
corresponding formation constant values, KFe-SCN, for different values in eq 8 the formation constant for Fe(Gly)2+, KFe-Gly, of
sets were calculated from their respective absorbance data using 63.3 L mol‒1 was obtained.
eq 6. Four other cells were designated to calculate average and
standard deviation values of ε and KFe-SCN. At first the target Conclusion
cell in Solver was set to minimize the standard deviation in the
calculated KFe-SCN values, while the ε values in the column (see The results presented in Table 1 show a fairly good agree-
the online material) were set for changing. Later on the target ment between the calculated values of εFe-SCN and KFe-SCN by the
cell was changed to minimize the standard deviation in ε, while two methods and yield a consistent value of KFe-Gly. Depending
retaining the ε values in the column for changing. After two such upon the expertise of the students in use of Excel software ei-
iterations constraints were introduced to keep both standard ther of these methods is equally suitable for the undergraduate

Table 2. Solver Calculated Parameters from Concentration and Absorbance Data or the Fe(III)–SCN– Complex
ε/(L mol–1 cm–1) KFe-SCN/(L mol–1)
CM / CL / ε/ KFe-SCN/
Absorbance
(10–3 mol L–1) (10–4 mol L–1) (L mol–1 cm–1) (L mol–1) Average SD Average SD

0.775 3.02 0.1617 4590 179

1.55 3.02 0.3076 4730 185
2.32 3.02 0.4252 4780 187
4690 80 183 3
3.10 3.02 0.5053 4710 184
3.87 3.02 0.5880 4750 185
4.65 3.02 0.6242 4600 181
Note: The shaded area designates the calculated data.

496 Journal of Chemical Education • Vol. 86 No. 4 April 2009 • www.JCE.DivCHED.org • © Division of Chemical Education
 In the Laboratory

laboratory demonstration. This method was tested in two classes 6. Zheng, Y.-Z.; Tong, M.-L.; Xue, W.; Zhang, W.-X.; Chen, X.-M.;
of 22 students each and was found to work satisfactorily across Grandjean, F.; Long, G. J. Angew. Chem., Int Ed. 2007, 46,
different groups of students. The calculated values of KFe-Gly 6076.
were mostly between 1.0 × 102 and 3.0 × 102 L mol‒1. It is eas- 7. Wu, L.; Pressprich, M.; Coppens, P.; DeMarco, M. J. Acta Cryst.
ily extendable to determination of formation constants of other 1993, C49, 1255.
non-phenolic amino acids and organic bases with iron(III) and 8. Anderegg, G. Inorg. Chim. Acta 1986, 121, 229.
gives reproducible quantitative comparisons. 9. Perrin, D. D. J. Chem. Soc. 1958, 3125.
10. Biruš, M.; Kujundźić, N.; Pribanić, M. Prog. React. Kinetics 1993,
Acknowledgments 18, 171.
11. Smith, R. M.; Martell, A. E. Critical Stability Constants; Plenum
We thank the students in CH306, Special Topics in Chem- Press: Oxford, 1976.
istry, for their hard work. This work was supported by a grant 12. Kiss, T.; Sovago, I.; Gergely, A. Pure Appl. Chem. 1991, 63, 597.
from the Faculty of Science and Technology, the University of 13. Cuculić, V.; Piźeta, I.; Branica, M. J. Electroanal. Chem. 2005,
the South Pacific, Suva, Fiji (Vote code 6C141-1321). 583, 140.
14. Byars, J.; McCreary, T. W. J. Chem. Educ. 1992, 69, 935.
Literature Cited 15. Frank, H. S.; Oswalt, R. L. J. Am. Chem. Soc. 1947, 69, 1321.
16. Cobb, C. L.; Love, G. A. J. Chem. Educ. 1998, 75, 90.
1. Ramette, R. W. J. Chem. Educ. 1963, 40, 71.
2. Dhungana, S.; Taboy, C. H.; Zak, O.; Larvie, M.; Crumbliss, A. Supporting JCE Online Material
L.; Aisen, P. Biochemistry 2004, 43, 205. http://www.jce.divched.org/Journal/Issues/2009/Apr/abs494.html
3. Martin, J. H.; Fitzwater, S. E. Nature 1988, 331, 341.
4. Timmermans, K. R.; Gledhill, M.; Nolting, R. F.; Veldhuis, M. J. Abstract and keywords
W.; de Baar, H. J. W.; van den Berg, C. M. G. Mar. Chem. 1998, Full text (PDF) with links to cited JCE articles
61, 229.
5. Zarta, M. K.; Powella, D.; Borovik, A. S. Inorg. Chim. Acta 2007, Supplement
360, 2397. Student handouts

© Division of Chemical Education • www.JCE.DivCHED.org • Vol. 86 No. 4 April 2009 • Journal of Chemical Education 497