Loading...
Please wait, while we are loading the content...
Similar Documents
Obtaining Accurate Reflectance and Transmittance Measurements for Optical Coatings
| Content Provider | Semantic Scholar |
|---|---|
| Author | Willey, Ronald R. |
| Copyright Year | 2012 |
| Abstract | The accuracy of the derivation of index of refraction values, n and k, for optical thin film coating materials depends heavily on the accuracy of the spectrophotometric measurement data supporting the derivation. Any reflectance and/or transmit tance errors will lead to errors in the resulting n- and k-values, which will in turn cause errors in the associated designs, and the actual coatings will have errors and depart from that which was intended. These derivations are an important part of the production of optical thin films, and need to be fully understood and carefully exercised. Possible causes of measurement errors, calibration procedures, and working standards are discussed. Tests for photometric accuracy and linearity are suggested. Techniques for calibrating spectrophotometers and obtaining accurate reflectance and transmittance data are reviewed. MEASURING TRANSMITTANCE There are occasional papers in the journals and sections in texts on the fine points of the measurement of optical coat ings such as Arndt et al.[1], but there is seldom a review of “common” practice for a person new to the field. This work attempts to help with that problem of “how to measure” before one embarks on what to do with those measurements. Different spectrophotometers have different f-number beams in their sample compartments and different minimum sample sizes in their beams. Older instruments may have been f/10 while some newer ones may be f/4. This beam convergence/ divergence may need to be taken into account where coatings are angle sensitive. For example, a 100 GHz fiber optics communication (DWDM) filter will shift a significant por tion of its bandwidth (0.4 nm) with a one degree tilt. This means that such a filter cannot be properly measured with an ordinary spectrophotometer because the beam divergence will distort the band shape significantly. A typical beam in a grating instrument is also a millimeter or more wide and about a centimeter high at its smallest point. Again, a DWDM filter of about 1.5 mm square could not be measured properly in such a beam. A new family of instruments has appeared for the DWDM field such as Optical Spectrum Analyzers (OSAs). These are built around the tunable lasers, Gradient Index Lenses (GRINs), and the fiber optics of the field that they serve. They usually measure in Decibels (dB) rather than %T. Although an OSA is very different in detail, in the final result, it is a spectrophotometer. Infrared instruments are subject to effects of the absorption of the water vapor and CO 2 in the atmosphere. These are most noticeable at about 2.7, 3.2, and 6.2 µm for H 2 O, and at 2.7, 4.3, and 15 µm for CO 2 . If the atmosphere within the instrument changes from when the 100%T calibration is run to when the sample is run, there may be spectral artifacts at some of these wavelengths. If this is a problem, the instrument can be purged with dry nitrogen to eliminate these effects. Similar things can be said for regions of the ultraviolet (UV, 200-400 nm). The region from 40 to 200 nm in the UV is referred to as the vacuum ultraviolet (VUV) because the atmosphere absorbs at those short wavelengths and instruments must be evacuated to be effective. It is understood that a dry nitrogen purge might be usable for some cases in the UV. The typical spectrophotometer is designed to measure transmittance (%T). Ways to measure reflectance are discussed below, but this section deals with transmittance only. Figure 1a illustrates a typical sample beam in the sample compartment of a grating spectrophotometer as seen from above, and Figure 1b shows the beam as seen from the side. The slit of the monochromator is typically imaged in the sample compartment; often near the center as shown, but sometimes at one side of the compartment. The reference beam is usually identical and toward the back of the compartment. The slit image is typically much higher than it is wide. It is assumed here that the light travels from left to right in these figures. A normal practice would be to run a 100%T calibration over the spectral range of interest with no sample in the beams, and then run a 0%T with an opaque sample in the sample beam, while being careful not to interfere with the reference beam in the same sample compartment. The sample of interest is then placed in the sample beam and its transmittance spectrum scanned. Barring errors and instrumental non-linearity, the displayed result is the actual transmittance of the sample. |
| File Format | PDF HTM / HTML |
| Alternate Webpage(s) | http://www.willeyoptical.com/pdfs/SVC2012Measurements.pdf |
| Alternate Webpage(s) | https://ftgsoftware.com/docs/augmeasure.pdf |
| Language | English |
| Access Restriction | Open |
| Content Type | Text |
| Resource Type | Article |
