| Author | Message | ||
     Suresh kumar. B (Sureshkumar) |
Hello I tried to quantify Iodine Value , FFA in Groundnut, Sunflower Oils i am getting repeatable results for Iodine value but not for FFA can u suggest me the which region i have to scan i used from 1800 to 2400nm and 1100 nm to 2500 nm both. Iam not getting Repeatable results for FFA i had used PLS for prediction and i had devoloped calibration file from 50 samples will it suffiencent. i will be happy if you can help me in this regard thanking you | ||
| Weyer |
I have a few suggestions based on my experience with fatty acid in crude tall oil. 1) The acid hydroxyl band at 1450 nm is very sensitive to temperature. You'll probably need to control the temperature if you use this band. Also, there is interference from CH combination bands there. 2) There is another, small but useful peak near 1880 nm (a combination C=O and OH). A C=O/CH combination doublet near 2200 nm is less distinct, but may help you. 3) If you cannot control temperature and need to calibrate for different temperatures, you need to know that the hydroxyl band becomes very, very broad as temperature decreases and so you probably cannot use derivative data treatments or you will lose the information you need to quantify the OH. 4) In addition to temperature, solvent effects can broaden or sharpen the OH peaks. 5) Examine your referee method carefully. Is FFA being calculated as one compound, but actually measuring many? If so, you may never get a good calibration. Try recalculating the FFA as only the COOH group. 6) Look at what PLS is using for the equation (regression coefficients, beta coefficients). If it does not involve the hydroxyl and carbonyl peaks, it is only looking at circumstantial correlations which may not hold for all samples. You can either try to cut down the spectral regions to the specific peaks or add more variation to the sample set. Good Luck. I concluded that mid IR was better for my situation. | ||
| Suresh kumar. B (Sureshkumar) |
Thank you verymuch sir here my problem is i have a chance to scan for desired range of nm.but my instrument can not calibrate at specific nm it is pattern recgn. software naa? i tried in 1500-2300 nm range for nm to quantify cleary specify me how temp. plays here or can we degrade the oil by any means i tried by heating at elevated temp. to degrade , to see the FFA clear spectrum but cannot suggest me any publications or books reg. this thanking you bye suresh kumar | ||
| weyer |
Suresh: Please give me your direct e-mail address and I'll send you more information. | ||
| sureshkumar |
Dear Weyer This is Suresh herewith giving my direct E-mail I.D [email protected] thanking you suresh |