ultraviolet-visible-near-infrared spectrophotometer

A new high-accuracy spectrophotometer has been developed at the National Research Council of Canada to measure regular transmittance factors over the spectral range from 200 to 2500 nm. The most significant feature of this automated single-beam instrument is a highly collimated normal-incidence beam geometry, which eliminates the need for polarization corrections or for an averaging sphere for the calibration of regular-transmittance reference materials. The instrument also possesses a large uniform measurement beam that minimizes errors caused by sample nonuniformity. We describe the instrument's design and the testing, optimization, and verification procedures that have been carried out for measurements in the visible and near-infrared regions. Systematic errors that have been determined and corrected for include wavelength shifts, stray light, system drift, and nonlinearity. In the visible and near-infrared regions, the overall photometric accuracy is estimated to be 2.5 and 4.0 parts in 104, respectively. The wavelength scale is accurate to within ±0.1 nm with a reproducibility of ±0.03 nm over its entire design range from 200 to 2500 nm. I. Introduction Spectrophotometry involves the measurement of the ratio of two radiometric quantities. The accuracy of the measurement is not based on a fundamental physical standard but derives from the design of the measuring instrument and the identification and characterization of all sources of error. Most commercial spectrophotometers are designed more for ease of use and reliability than for the highest accuracy. Frequently they employ pulsed sources, double-beam operation and restricted source, sample, and detector compartments to maximize stability and signal-tonoise ratio performance. Unfortunately, these design features interfere with and complicate an analysis of the instrument's systematic errors. Consequently, commercial spectrophotometers are limited in their measurement accuracies typically to a few parts in 103.