and intermediate precision, accurac

and intermediate precision, accuracy, specificity, detection (LOD)
and quantitation (LOQ) limits, and measurement uncertainty.
2. Materials and methods
2.1. Chemicals and standards preparation
All chemicals were of analytical reagent grade: ethylenediaminetetraacetic
acid disodium salt di-hydrate, ferric chloride
hexa-hydrate, tetrabutylammonium hydroxide solution 1 M, glacial
acetic acid, ammonium acetate, and methanol were purchased
from Sigma–Aldrich. All aqueous solutions were prepared with
high-purity water (resistivity 18.2 M cm1
) obtained from a MilliQ
system (Millipore, Milford, MA, USA). Ferric chloride solution
was prepared adding 4 g of ferric chloride hexa-hydrate to
100 mL of a 30% (v/v) glacial acetic acid solution, and pH 3 was
achieved by adding potassium hydroxide pellets. An EDTA standard
stock solution at 2000 mg L1 was prepared by dissolving
the stoichiometric quantity of EDTA disodium salt dihydrate
(254.2 mg) in 100 mL of water. Working standard solutions, ranging
from 0.25 to 20 mg L1
, were daily prepared by dilution of
the stock solution with water, previously adding 100 lL of the ferric
chloride solution.
2.2. EDTA extraction procedure
Analyzed complete feed aliquots (fat 1%, protein 2.8%, ashes 3%,
fiber 30%, non-nitrogen compounds 53.2%) and pre-mixtures (containing
as main ingredients zinc – around 7% – and copper –
around 0.05%) were collected during official controls from Italian
suppliers. Complete feed aliquots of approximately 5.00 g were
added with 2 mL of the ferric chloride solution; then, a dilution
with 40 mL of water was settled. As regards to premix formulations,
35 mL of the ferric chloride solution and 7 mL of water were
added to approximately 2.50 g of sample. After shaking for 5 min,
sample solutions were heated in oven, being careful that they were
kept at 70 C±2 C for 30 min; then, centrifugation was performed
at 1300 g for 15 min, and the supernatant was filtered with PVDF
filters (polyvinylidene-difluoride, 0.45 lm).
2.3. HPLC determination
Data were acquired with an Ultimate 3000Thermo Fisher HPLC
device, equipped with a multi-solvent delivery system, an autosampler
with a 10 lL sample loop, a thermostatic column oven,
and a diode array detector. The mobile phase consisted in: (A)
methanol and (B) acetate buffer/TBAOH 0.025 M, running isocratically
at 5% and 95%, respectively. Solution B was prepared by dissolving
4.1 g of ammonium acetate in 800 mL of water, then
25 mL of TBAOH 1 M were added, and pH was adjusted to 4.5 with
glacial acetic acid; the final volume was made up to 1 L with water,
checking again pH. Flow rate was set to 0.8 mL/min; runtime was
10 min for EDTA standard solutions, and 40 min for sample solutions.
Column temperature was maintained at 50 C; UV detection
was performed at 254 nm (absorption maximum), and the injection
volume was 10 lL.
The analytical column was a Phenomenex Kinetex C18 (Phenomenex,
Torrance, USA), 250  4.6 mm, particle size 5 lm (example
chromatograms are reported in Fig. 1a,b).
2.4. Method development, optimization and validation
The evaluation of the quantity of ferric chloride needed for a
quantitative conversion of EDTA species into the Fe(III)-complex,
either in complete feed or premix formulations, was of utmost
F. Ch