| Author | Message | ||
     Dr.K.Balasubramanian (drkbala) New member Username: drkbala Post Number: 2 Registered: 9-2006 |
Pl perform the Second derivative | ||
| thomas ricour (tricour) Junior Member Username: tricour Post Number: 7 Registered: 2-2006 |
Thanks Gab, it is 1 ml of alcohol in 1 liter of water. Waiting for your answers my direct email is [email protected] kind regards Thomas Ricour | ||
| Gavriel Levin (levin) Junior Member Username: levin Post Number: 9 Registered: 1-2006 |
Hi Thomas, Lets's get clear - is it 1 ml/L of water in alcohol, or 1 ml/L of alcohol in water? In either case, write to me so I am not commercial and I will provide you data on this. I will post what can be posted after knowing the exact issue. My e-mail: [email protected] Thanks, Gabi Levin | ||
| Howard Mark (hlmark) Senior Member Username: hlmark Post Number: 38 Registered: 9-2001 |
Thomas - before you worry about data pretreatments, I'd say first you should worry about getting good data. There are two big differences between the spectrum of water and the spectrum of alcohol: first, water has no -CH bands, so the -CH bands of alcohol are very distinctive. Secondly, there is a big shift in the wavelength of the -OH band between the two materials. Either of these characteristics should enable you to detect alcohol at much lower concentrations than "normal", if they're not overwhelmed by other effects. So the first question that comes to mind is how much extraneous variation in composition there might be. For pure alcohol in pure water I think it might be doable, but if there are lots of other things changing, or if there are other components in the solution, at higher concentrations than the alcohol, then it becomes very problematic. The next issue is temperature. The -OH band of alcohol is very distinctive, but is still close to the water -OH band, and temperature changes can easily wash out the difference. So your detection limit will depend, among other things, on how close temperature control you can maintain. I suggest that in order to learn how your system is going to behave, you start with high concentrations of alcohol and make up dilutions -not for calibration but to inspect the spectrum and see how low you can go and still detect the distinctive absorbance bands of the alcohol. For calibration, since the bands are relatively broad, you could probably apply a fair amount of smoothing, and then use a derivative to enhance the spectral distinctions. Let us know what happens. Good luck. Howard \o/ /_\ | ||
| thomas ricour (tricour) Junior Member Username: tricour Post Number: 6 Registered: 2-2006 |
hello everybody. I'm trying to use NIR AOTF technology to make a calibration on water samples with inside 1ml/L of alcohol component and i want to know if someone else did that before and if it isworking or if we are under the detection of the technology. If it could work, what kind of pretreatement do you suggest to succeed in this project: first derivative, second...? I would be very interested to know about your opinion. Kind regards Thomas Ricour |