| Author | Message | ||
     lawrence |
Dear Members: There is a data set related to Protein measurment originally worked out by Fearn. It is applied by many professionel to make research project. May I know where i can download the data set? Thank you for your information. Regards, Lawrence | ||
| Tony Davies (Td) |
Lawrence, I have asked Tom Fearn. Could you be more specific? There are at least two data sets that fit your description one is possible the other is more difficult. Best wishes, Tony | ||
| Solomon |
Hi every one, I am trying to build a calibration model which can predict a particle size distribution in crystallisation process. I want a comment about my procedure in building the calibration model. 1st, am going to take a spectra from the solution of the CH compound/ Ethanol. Since the CH compound do not dissolve in Ethanol. By taking the pure Ethanol as a background i presume i would get the CH pure spectra at predetermined size (reference method Master sizer). I will have a library of spectra for different size and concentration. Once i build the calibration model using PLS chemo. technique, i will run a crystallisation process using the same CH compound and distilled water. Taking pure water as background prior taking the kinetic spectra during crystallisation process. The later spectra will be used to test the calibration model. Hope fully we can determine the particle size distribution during crystallisation period. The problem may be i did not take the temperature effect in building the calibration model. Can you comment on the procedure please? | ||
| David W. Hopkins (Dhopkins) |
Solomon, While you are probably right that it may be possible to construct a calibration for the crystal size, I doubt that the procedure you outline will be successful. When you use a EtOH solution as 'background', it is not likely to be a good reference for EtOH with a suspension of particles, because the optical path in the suspension will be longer than in the pure solvent, due to scattering. Ditto for the water system. The problem will be even more severe in the ethanol system, because the bands in the CH target molecule may overlap significantly with the EtOH, so that particle size effects and scattering effects will be difficult to separate. EtOH will also show temperature effects. The primary rule for effective calibrations is that the calibration samples should be representative of the samples you wish to measure. I suspect you would be better off to put your effort into obtaining crystal size data on the media you will be using, and vary anything other components in the media too, so that the calibration will not be sensitive to changing levels of your CH compound and other constituents. Also be sure to vary the temperature, as that will also be a factor, as you suggest. Good luck, Dave | ||
| DJDahm |
Is what you hope to determine the change in average particle size with time during the crystallization process, or do you really mean particle size distribution? If it is the later, what evidence do you have that the spectra of two spectra of samples of the same �average� particle size, but different particle size distributions will give distinguishable spectra? | ||
| Gabi Levin |
Hi Solomon, Is the old traditional way of creating calibrations by collecting spectra from real crystallization process and pulling samples at the same time spectra is collected, performing size analysis on each sample not feasible for some reason? If it isfeasible I would not think of any other scheme. Unfortunatley, NIR is influenced by a variety of parameters, and usually simple sumation or deduction of one from another does not yield the spectrum for one or the other in another matrix. If the "simple" case of powders which are as non-interacting as you could think of is not lending itself to the simple "summation" of spectra in order to create the spectrum of a mixture, then the situation in an aqueous suspension, is even worse. Theoretically, the spectrum of crystallized material shall be identical to the spectrum of the crystals once they have been "dried". However, there are interactions between the liquid and the solid, which may affect both the spectrum of the crystals and the liquid. The magnitude of the effect depends greatly on the ratio between the surface area and the bulk volume of the crystals. This changes as the crystals keep growing during crystallization. How can all these effects be taken into consideration in a simplified approach? I hope this is of any use. Gabi Levin Brimrose Corp. [email protected] |