【正文】 here coating,with BaSO4 being more popular after that time. For many decades theonly option for performing spectral diffuse reflectance measurements in the midinfrared was a conic mirror reflectometer,due to the absence of suitable midinfrared integrating sphere coatings. In 1966, Morris fabricated a glass sphere with a roughened, aluminized surface for use beyond 181。m. About 10 years later, Egan and Hilgeman tested a sphere coated with sulfur flowers for use in the – 12 181。m region. A major advance in instrumentation occurred in 1976 with the introduction of the Willey 318 infrared (IR) diffuse spectrophotometer. Willey’s innovative instrument, shown in Figure 2, utilized a Michelsoninterferometer, an integrating sphere with a diffuse textured gold surface, and a doublebeam geometry. In the 1980s Labsphere mercialized diffuse gold integrating spheres. IR sphere accessories are now available for most mercial Fourier transform infrared (FTIR) spectrophotometers. The National Physical Laboratory (NPL) in the UK provides a flamesprayed aluminum diffuse reflectance standard National Institute of Standards and Technology (NIST) in the USA is also developing an IR diffuse reflectance standard based on roughened significant development in the integrating sphere field was the introduction of nonimaging concentrators for coupling detectors and sources to spheres. Most absolute sphere designs depend on either a restricted field of view (FOV) detector or a hemispherical FOV bination of nonimaging concentrators and baffles can be used to eliminate measurement errors associated with unequal scattering of radiation from the sample and reference into the sphere wall region viewed by the detector. Nonimaging concentrators also have the advantage of returning rejected light into the sphere, which results in significant enhancements in throughput pared with baffling or collimating a detector’s FOV.Integrating sphere accessories are the instruments of choice for accurate quantitative spectroscopy of scattering materials. Their design allows measurement directionalhemispherical reflectance, hemisphericaldirectional reflectance factor and/or hemisphericalhemispherical features of integrating spheres are critical to their ability to accurately measure reflected fluxes. The first is that the interior surface has (ideally) a Lambertian coating of high reflectance. This means that the radiance of the coating is constant in all directions. The second important feature is the spherical shape of the interior. For a Lambertian scattering surface, light incident on it results in a constant irradiance over any spherical shell that contains the light incident anywhere on an integrating sphere is immediately uniformly distributed over its entire detector located on the第 48 頁 spherical surface or viewing a part of it will always receive flux that is directly proportional to the light incident on the sphere, independent of location. Even if the sphere surface is not perfectly Lambertian, the high reflectance of the surface leads to a large number of reflections around the sphere prior to absorption of the incident light. This cavity effect further contributes to the uniform flux distribution over the sphere , when light is incident on a sample with arbitrary BRDF, the sphere can be used to accurately collect all the reflected flux.An ideal diffuse reflectometer can be defined as a reflectometer with a throughput that is independent of the angle of reflected radiation, as measured at the sample. Effects due to ports, baffles, the sample’ s BRDF, the detector FOV,the sphere coating ’ s BRDF and the curvature of samples can all introduce systematic measurement errors. The purpose of this section is to review the sphere design features and principles that will minimize these sources of assume that the sphere considered has its sample and reference simultaneously mounted on the wall and is operated in the parison mode in a d/h geometry. The error sources and effects described here also apply to “ substitution ” mode spheres (with a single port for both sample and reference). However,spheres have increased levels of measurement error and will not be addressed specifically here. Detailed descriptions of the substitution error can be found in Hardy and Pineo and others.After the reflection of the input beam from the sample and reference, the reflected light will be distributed on the sphere wall according to the sample and reference accurate measurement results, an equal fraction of the reflected sample and reference radiation must illuminate the sphere wall region viewed by the the radiation exchange factors from the sample and reference to the sphere wall viewed by the detector are not equal, a systematic measurement error will be introduced. After the first reflection from the sphere wall, the residual radiation that was reflected from the sample and reference will have equal radiation exchange factors to the sphere wall viewed by the detector if the sphere wall coating is are only three detector FOV sphere geometries that guarantee (irrespective of BRDF) that equal fractions of sample and referencereflected radiation is scattered into the region of the sphere wall viewed by the detector. The three geometries have detectors with a hemispherical FOV,a FOV that includes exactly half of the sphere wall, and a FOV that is restricted to a narrow angular region.Baffles are frequently required to prevent radiation exchange from one region of the sphere wall to another. For this reason they can be found in most integrating sphere systems, including several absolute methods. For example,for sphere designs that employ a restricted FOV detector,one or two baffles are required to prevent direct radiant interchange between the sample or reference and the region of the sphere wall viewed by the detector. An illustration of the design principle of placing a baffle in an integrating sphere is shown