Carn-06_Book_Final

Carn-06 - Validation Report FOR ERP USE ONLY DO NOT DISTRIBUTE

nitine esters to form free carnitine than an acidic hydrolysis thereby giving higher results this is seen as ok because we desire to measure the total content of car- nitine. Sample: NIST SRM 1845A Measuring range: The measuring range in this validation covers the interval: LOQ – 1000 ng/g. Samples containing concentration higher than the highest standard will be reana- lysed with a more fitting sample size and/or dilution leading to a concentration in- side the standard range. LOD has been determined by treating a standard at low level (9 ng/g) as a sam- ple, the reason this approach has been used is that it has been impossible to find a natural sample containing carnitine at a level near the expected limit of detec- tion. The sample was analysed as a 10 replicates measurement on the same day, the standard deviation (S r ) has been calculated from these results. See appendix 3. LOD is calculated as 3 x S r and LOQ as (10 x S r ). The limit of detection and quantification for carnitine is calculated to: Limit of detection / quantification:

LOD: 0,9 ng/g LOQ: 3,0 ng/g

LOQ in samples as mg/kg is calculated from an aliquot of 20 g sample (maxi- mum) extracted into 110 g extraction solution. 10 g extraction solution is diluted to 250 g, (this sums up to a total dilution factor of 137,5). Carnitine in samples: LOQ: 0,4 mg/kg

Robustness:

The robustness of the method has been evaluated on the following critical pa- rameters; extraction time and saponification time. Extraction time: The robustness test has been performed on a milk powder and a pet food con-