Ferrozine binds to both Fe(II) and Fe(III) ions in solution and turns into a magenta-colored complex (note: Fe(II)-induced complex has ~60x stronger absorbance than Fe(III), thus constituting a larger part of the signal). Maximum absorbance is at λ=562 nm at pH 4-9. To determine the concentrations of Fe(II) and Fe(III) ions, two spectrophotometric absorbance measurements are made, termed A1 and A2, where A1 is an initial measurement and A2 after a reduction step to reduce all Fe(III) to Fe(II) ions.
After measuring A1 and A2, the following system of two equations is solved to find CFeII and CFeIII, the molar concentrations of Fe(II) and Fe(III) ions, respectively:
A1=βFeIICFeII+βFeIIICFeIIIA2=βFeII(CFeII+CFeIII)α
where βx=ϵxl , x is the species to be measured, ϵx is the molar absorbance coefficient of the specific complex, l the path length. β can be interpreted as the instrument-specific absorbance coefficient and α is the dilution factor (0 < α < 1) due to the addition of reductant reagent to obtain A2.
In order to solve for CFeII and CFeIII, all we need to determine are bFeII, bFeIII, and a (and estimate their uncertainties). We can obtain these values by analyzing a dilution series of standard solutions where the concentrations of Fe(II) and Fe(III) ions are known. A realistic α value (along with its uncertainty estimates) is obtained by doing the reduction step a second time to the standards only, obtaining A3 (even though all Fe(III) has already been reduced). Hence α = A3/A2. All analytical uncertainty estimates can be obtained by regression analysis after this step (usually linear regression, estimating error of the slope), using the same dataset already obtained.
After obtaining βFeII, βFeIII, and α, we may derive CFeII and CFeIII as the following:
CFeII=α(βFeII−βFeIII)A2−αA1
CFeIII=α(βFeII−βFeIII)βFeIIαβFeIIA1−βFeIIIA2
The range of measurement of this assay is between 0 - 80 µM Fe(II/III), thus for samples containing a higher amount of dissolved Fe, especially anoxic water samples, dilution is required. For unknown samples, it is useful to add a 10x dilution sample. For samples that may contain elevated Fe, an additional 100x dilution is recommended. An accurate way to perform dilution is by measuring mass instead of volume. A pipette (which measures volume) has about 1-3% accuracy, whereas an analytical balance (0.1-1 mg accuracy) is vastly better. Do not switch pipettes during colorimetric assay procedures and pay careful attention.