A pH-mediated stacking method in capillary electrophoresis as an assay for low concentrations of melamine in milk products was established. Real samples were treated with acetone and sodium acetate and injected directly after centrifugation and filtration. Several experimental factors, such as buffer pH, buffer concentration, sample matrix, injection/sweeping ratio, sweeping time/voltages, separation voltages, as well as sample pretreatment, which affected stacking and separation, were investigated and optimized. Under the selected condition, a low LOD of 0.01 μmol/L (S/N = 5) and a wide range of linearity of 0.01∼1.0 μmol/L could be easily achieved with a good reproducibility (RSDs < 5.8% for both migration time and peak area) and an acceptable recovery of 94.0∼103.2% (for milk, infant formula, yogurt, and milk products). The proposed method was suitable for routine assay of melamine in real milk samples.

Melamine was measured in real milk products with capillary electrophoresis (CE) based on acetonitrile-salt stacking (ASS) method. Real milk samples were deproteinized with acetonitrile at a final concentration of 60% (v/v) and then injected hydrodynamically at 50 mBar for 40.0 s. The optimized buffer contains 80.0 mmol/L pH 2.8 phosphates. Melamine could be detected within 20.0 min at +10 kV with a low limit of detection (LOD) of 0.03 μmol/L. Satisfactory reproducibility (inter- and intraday RSD% both for migration time and peak area was lower than 5.0%) and a wide linearity range of 0.05 μmol/L ~ 10.0 μmol/L were achieved. The proposed method was suitable for routine assay of MEL in real milk samples that was subjected to a simple treatment step.

A capillary zone electrophoresis method was developed for determination of melamine in food samples, such as quail egg and milk products. The CE procedure was performed on fused silica capillary (41 cm × 75 μm I.D.) at 17 kV using pH3.1 60mmol/L phosphate buffer as run buffer and detecting at 200 nm. The proposed method showed good linearity (0.5 -10.0 μmol/L) and low LOD (0.5 μmol/L) with good reproducibility (RSD% was 2.4 and 3.2 for migration time and peak area respectively), which made it suitable for quantity control of the related food product.