| Author | Message | ||
     Howard Mark (hlmark) Senior Member Username: hlmark Post Number: 44 Registered: 9-2001 |
Tatjana - As Jim said, manufacturers generally trust the laser referencing to insure wavelength accuracy. That said, there are still some possible causes of inaccuracy, some will affect precision, some accuracy only. But only the most precise measurements will be noticeably affected by them: 1) If the optical beam does not enter the interferometer exactly along the optical axis of the interferometer, but is tilted compared to the optical axis, then all wavelengths will appear to be somewhat shorter than their true values. If the laser beam is also tilted and by the same amount, then that would compensate for the tilt. This would affect wavelength accuracy only, and not precision. 2) Changes in temperature, pressure or humidity of the air will change the index of refraction of the air and thus the wavelength of the beam. Peter Griffiths ran into this when using his FTIR to check the calibration of a NIST-type NIR wavelength standard; he got different values than NIST obtained. It turned out that NIST reported the values for wavelengths in vacuum and Peter reported the wavelengths in air (at his ambient conditions). 3) Under some conditions, there can be an apparent (but not real) wavelength inaccuracy, if the wavelength is being checked by using a standard based on absorbance bands. If the bands are non-symmetrical, then the measured wavelength will depend on the instrumental resolution and also on whether there are non-linearities in the instrumental measurement system. When NIST released their CRM-1920 wavelength reference standard for NIR wavelengths, they had to specify the wavelengths at different spectral resolutions, and also how the measurement should be made and how the wavelength should be calculated, because of these effects Howard \o/ /_\ | ||
| James Malone (jmalone) New member Username: jmalone Post Number: 1 Registered: 8-2006 |
Dear Tatjana: 1. FT instruments generally have very good wavelength accuracy - it's in the nature of the laser clock in the interferometer. Having said that, the manufacturer should still be able to give you the S.D. of the wavelength - although this would likely be an average of many instruments and may not relate to yours. I think is is more useful to simply make the measurements as you have described and determine "your" wavelength S.D. It will reflect the instrument's performance as well as your ability to re-position the wavelength standard in the sample compartment. In any case, the value in periodic checking will be to learn if any changes have occurred. That will almost certainly indicate a problem with the system. 2. If an instrument does not respond in a linear fashion to changes in the intensity of light reaching the detector, the differences between high transmission peaks and low transmission peaks may not be representative of the actual characteristics of the molecule. This may not be as important in qualitative work, but I would still do the test. Any changes in linearity could indicate a new source of stray light (new sampling accessory???) or some electronic problems. As with wavelength accuracy, any changes over time may indicate subtle problems that should be investigated. | ||
| tatjana tomazin (tomazin) New member Username: tomazin Post Number: 2 Registered: 5-2006 |
Does anybody have any experience regarding the FT-NIR instrument qualification? The instrument is used for identification tests � measuring diffuse reflectance, using fibre optics probe. I'm interested in two things: 1.Verification of wavelength scale repeatability using Extended range near-infrared diffuse reflectance secondary wavelength standard KTP-1920X . However I can not get the information about the standard deviation from the instrument supplier. Do you have any suggestion about it? I suppose I will make 10 measurements of the reference material one after another and determine the wavelengths, where the three characteristic peaks appears (1261 nm, 1681 nm and 1935 nm.). But I could not verify repeatability if I do not have the allowed standard deviation. 2.Checking linearity: I'm wondering is it really necessarily to check the linearity for the instrument, which is used only for qualitative applications? Thanks for your suggestions, Tatjana |