Accuracy and precision of a scanning and integrating microinterferometer

Abstract

SummaryInstrumental factors affecting the precision and accuracy of measurements with the Vickers M86 scanning and integrating microinterferometer were investigated. ‘Spot’ measurements of OPD (optical path difference) could be made to a precision better than 0.001 Λ (wavelengths); the coefficient of variation of scanning measurements of IOPD (integrated OPD) depended on the object, typical values being approximately ± 5% for a human erythrocyte in water, and ± 10% for a mouse sperm head in water. Repeated measurements enabled the dry mass of a specimen to be estimated to any desired precision.Instrumental calibration to give results in absolute units is described. Early prototypes required frequent calibration, but with production instruments the calibration varies little from day to day, and very similar results are obtained if a specimen is measured in different parts of the scanned field.A slight, approximately sinusoidal deviation from linearity was sometimes observed when the apparent OPD or IOPD was plotted against expected values. The error, which probably occurs with all instruments employing a Jamin double‐refracting interference system, was maximally about ± 0.02 Λ for a specimen of true OPD 0.25 Λ or 0.75 Λ, but was zero for a specimen of true OPD 0.5 Λ. The error could be minimized or eliminated by the correct choice of an interference fringe for ‘spreading’, careful adjustment of the condenser and background OPD, and especially by reduction of the condenser aperture.Scanning measurements of IOPD were shown experimentally and theoretically to be insensitive to changes in the measuring‐spot diameter or errors in focus. Valid measurements of dry mass can therefore be made even if the three‐dimensional specimen (e.g. a living cell or isolated nucleus) is considerably thicker than the depth of focus of the objective used.