Q: When a pharmaceutical quality-control or quality-assurance laboratory weighs a monogrammed substance for assay, the balance used to weigh the sample must conform to the requirements laid out in US Pharmacopeia <41>. How do you assign a minimum sample weight for a USP <41> balance application when the tested repeatability gives a standard deviation (SD) of zero?
A: The idea that 0 SD for a repeatability test means an excellent result is a misconception. In fact, 0 SD indicates a problem with the choice of instrument or methodology used for the test.
If a measuring instrument indicates a 0 SD during testing, it does not have enough sensitivity to measure the tested mass accurately. Because most balances used for USP <41> applications are analytical balances (i.e., those with 0.1 or 0.01mg readability), a result of 0 SD during a repeatability test should never be considered a pass criterion for minimum-weight verification. The same can be said for micro and ultramicro balances (i.e., those with 1 or 0.1µg readability). A 0 SD generally means that the balance does not have enough displayed digits (i.e., readability) for the tested minimum weight. USP <41> recommends a limit of 0.1% error in the measurement. If the balance cannot display 0.1% of the desired minimum weight, then testing for suitability is pointless because you cannot meet the USP requirements.
Course-reading top-loader balances sometimes give a 0 SD result during minimum-weight tests. The problem is that top-loader balances of low readability (e.g., 0.1 or 0.01 g) are inherently insensitive because of their course readability. The theoretical test mass needed to achieve suitable sensitivity levels with such an instrument may be way beyond the balance’s actual capacity. In compliance with current legislation, the correct solution is to employ a balance with the correct readability to display 0.1% of the net sample under investigation.
—Ian Ciesniewski, technical director at Mettler Toledo
If you have a problem with your equipment or process, an industry expert may have the solution. Please send your question to Erik Greb, editor of Equipment and Processing Report, and we may be able to provide an answer in a future issue. All questions will remain anonymous.