| Author | Message | ||
     Howard Mark (hlmark) Senior Member Username: hlmark Post Number: 99 Registered: 9-2001 |
Don - I think Tony is on the mark, but there are some other considerations, too, that you might have to think about. One problem with most glasses (including the BK-7) in the NIR (and even with fused quartz, unless made specifically for NIR usage) is the water content of the glass. Not only does this extract useful energy from the beam (which decreases the signal and increases the amount of heating of both the window and the sample) but water is notorious for changing its spectrum when the temperature changes. This will introduce errors into both calibration data and measurement data. For this reason, I think you're better off to stay with NIR-quality quartz (or perhaps sapphire or other NIR-transparent material). Incidentally, I don't think you're delivering 550 watts of energy to the sample or to the cell. Presumably the 550 watts refers to the input electrical energy to the lamp; the total output optical energy will probably be somewhat less than a watt; this is enough, if focussed on your finger, to be felt as sensible heat, but is still far from 550 watts of energy on the finger. \o/ /_\ | ||
| Tony Davies (td) Moderator Username: td Post Number: 156 Registered: 1-2001 |
Hi David! Your question goes back to 1800; the NIR dawn when William Herschel was trying to determine which colour was associated with the greatest heating effect and accidentally discovered radiation he could not see! I suggest that much of 550 watts comes from the visible and the problem is not so much the glass as the sample. The glass has low absorption so will not be heated directly by the radiation, it will be at about the environment temperature. On the other hand the sample will absorb radiation and will get hotter than the environment. We know that sample temperature plays an important role in the measured spectrum. I have never been too keen on post-dispersive systems because of what I regard as an experimental flaw but I admit that this does not seem to stopping them from working. At NIR-2001 I suggested that we should make more use of the temperature effect -�thermometrical spectra�. (The name suggested by Herschel for his discovery). It just occurred to me that you could do this in the a post dispersive system by varying the number of sample scans before storage of the n+1 scan. I realise that this has nothing to do with your question but I think you need to worry about the sample temperature. However your question is interesting and perhaps Don (Dahm) would like to provide the answer! Best wishes, Tony | ||
| David Donald (ddonald) New member Username: ddonald Post Number: 2 Registered: 11-2006 |
Hi All, My question pertains to the thermal variation of the optical glass windows, BK7 (borosilicate crown glass). BK7 windows are commonly used to encase photometric standards, which are used for calibration transfer or standardisation. In most applications, the instrument using the BK7 windows are pre-dispersive instruments which transmit a low energy beam (less than a watt or so) through the window and onto the sample, etc. so little heating of the window occurs and the density of the BK7 alters very little (and hence the refractive index remains kinda constant). However,... I am in the situation where a post-dispersive instrument is in use and there is approximately 550 watts of energy being passed through the unfortunate window. After a literature search on the topic, I found that dn/dT (rate of change of refractive index Vs temperature) for borosilicate crown glass is approx. 4x10^-6 per K. (in the operational range 20 - 60deg C). Now, in the normal operation of these windows, I will be in an uncontrollable environment where the temperature can vary by up to 40 deg C (from 10 deg to 50deg) inside of a stinky mill - unavoidable. So the typical changes of the refractive index (at 633nm) would be 2x10^-3 at the most. I was unable to obtain any information on the NIR spectrum variation with respect to temperature for BK7 windows. The BK7 standards are to be used for standardisation type work and naturally enough I would like to have control of the sources of variation. So now the question: Is there a noticeable difference in the transmission NIR spectrum (from 100nm to 2400nm) of BK7 windows (3mm) with variations in temperature of 10-50 deg C? A simple solution to this question is: use fused silica instead, but my curiosity has been invoked and of course would like to know if the cheaper windows will suffice. I�ve read Phil Williams article: �The near infrared window�glass or quartz? A study in precision� which basically states that I shouldn�t be too tight when it comes to getting windows � you get what you pay for. Best Regards, David. |