| Author | Message | ||
     Donald J Dahm (djdahm) New member Username: djdahm Post Number: 1 Registered: 2-2007 |
Since a few of my friends have written me saying I should get involved in this topic, here goes. However, I really don't have much to add to what Andrew McGlone said. We discussed this in several NIR News issues (16.6, 16.8, 17.1, and 17.2). The most relevant article is in 16.8, especially Figure 2. The limiting case argument was made by Kortum in his book Reflectance Spectroscopy. As I recall, he used the formulas for radiation from an oscillating dipole to express the Mie scatter as a series, the first term of which was the Raleigh scatter, and suggested that as particles got larger more terms were necessary to get a reasonable approximation. His equations are not the ones that you will usually see to describe Mie scatter. There is a complicating factor. Whenever particles are closer together than the size of their diameters, the light waves do not see them as independent particles. Perhaps one thing worth mentioning is the confusion over the isotropic nature of Raleigh Scatter. When people do calculations for a mixture of particles (like the ones being contemplated here), they frequently assume scatter that is truly isotropic: the same in all directions. In dipole radiation, the power distribution is not isotropic, but is rather concentrated around the directions lying perpendicular to the dipole moment. The Raleigh Equation also does not yield a scattering that is equal in all directions. What it yields is a power distribution pattern that is as if a mirror (perpendicular to the incident beam) has been placed through the particle. The effect is that in the directions directly forward and backward, the intensity is twice that in the sideward direction. | ||
| Nico Heigl (nch) New member Username: nch Post Number: 3 Registered: 1-2007 |
Hi! Tony, sample amount varies from minimum 10 mg up to gram. It is a very time consuming procedure to get fractions with different particle size, took us nearly two days to get a 10 mg fraction of some pharamceutical powder sample cause it has a certain amount of humidity and therefore agglomerates quite strong. But it is the time worth to separate the frations as good as possible in order to establish precise calibration models. Most of the time i use CHCL3 suspensions for determination of physicochemical parameters of powders and powder-like samples. What i experienced is that it can be dangerous just to stick in the fibre optic into bulk samples because density and other parameterns change during measurements and you get not-wanted scatter effects that can lead to deceptive erronous predictions. But with that low amount of sample, < 10mg, it is nearly impossible to analyse suspensions, viscosity is too low. I�m really thankful for your suggestion to use sulphur and carbon black i will do some measurements to test, thanks! Once i did some measurements where i used the background material as a reference spectrum, and what i foud was that measurements lead to better results if you scan your sample and substract the spectrum of the backgound material afterwards. Andrew, thanks for the links - very nice!! I just keep on trying, thank you all for your suggestions and interest! Greetings Nico. | ||
| Andrew McGlone (mcglone) Junior Member Username: mcglone Post Number: 9 Registered: 2-2001 |
Mie vs Rayleigh? I learnt a 'rule of thumb' to only consider the simple rayleigh solution when the particle size is at least below 1/10th of the light wavelength. Nanoparticles generally range from 10's to 100's of nanometres, so whether you can successfully use the simple rayleigh solution depends on your sizes. I'd suggest you find the Mie solution, as the rayleigh solution is just the same solution in the small particle limit. That is, the Mie and rayleigh solutions agree at that limit! I'm not particularly experienced in Mie scattering, but have made the calculations when necessary. In these days of fast processors, the large sums take no time at all to calculate. Depending on exactly what you are doing, you may be quite satisfied with simple calculators, like Scott Phral has on his website http://omlc.ogi.edu/calc/mie_calc.html Otherwise there is lots of web based code, in various languages, as can be found at: http://www.iwt-bremen.de/vt/laser/wriedt/Mie_Type_Codes/body_mie_type_codes.html | ||
| Tony Davies (td) Moderator Username: td Post Number: 144 Registered: 1-2001 |
Hello Nico, Gabi, Howard (and all other readers), Gabi, I was using a Neotec 6350 grating spectrometer. (the work was done in �85-6) so I was working with 55 mm reflectance sample holders. Perhaps I should explain that I placed the sample on the cover glass and then filled the cell with powdered sulphur. A cell containing only sulphur was used as the reference. The idea is that you have two well matched samples so that all the differences are due to the small amount of sample. You could try the same approach with your seed. Sulphur was suggested as a reference material by Tkachuk and Law in 1978! When I worked for Oxford Instruments (�86-89) I made up a series of reflectance �standards� by adding carbon black into the ceramic standard on a set of sample drawers. This enabled me to take readings from instruments as they came out of production. It was messy (Howard is always right!) but it worked. I got reproducible results from the same instruments but worryingly different results from different instruments. I left the company before I was able to decide how to use the information. Nico, Question: how much material do you have? My small sample idea is a modification of Mark Meurens� DESIR technique. He evaporates dissolved samples on a glass fibre filter paper. From a theoretical point of view absorption measurements are at optimum at 0.434, which assumes that measurements are detector noise limited. In these experiments, while absorption levels are low I think this is a reasonable approximation so I would suggest that the ideal would be to mix sulphur and carbon black to give an absorption (when measured against a sulphur standard) of about 0.434 (i.e. 0.4 � 0.5) and then use this material as the reference and as the background for the sample. This could be a small amount of deposited material (Nico) or a seed (Gabi). This is a different approach to what you are both doing. I think as well as keeping the absorption levels similar you also need to keep the reflectance characteristics similar otherwise you could get some strange effects! Should this be described as Mie or Raleigh scatter? This is really a question for my friend Donald Dahm (where are you Don?) but Raleigh deals with small objects so my answer is: Raleigh is closest. Best wishes, Tony | ||
| Nico Heigl (nch) New member Username: nch Post Number: 2 Registered: 1-2007 |
Hi, Thanks a lot for the literature hint Tony everything that considers this topic really helps me! Gabi, I�m using a FT-NIR coulped with an sample desk for diffuse reflection measurements. But my aim is a litte bit different from yours. What i try to do is to find an apporpriate background material that amplifies the signal in a special wavenumber range. What i tried is to use on one hand an solid absorbing material, diamond, so test whether i get a good reflectance signal of the sample. In this case the sample consisted of dendrimers (nanoscaled particles) dissolved in methanol. Therefore a solid background material is well suited to directly dry the sample on the material. The second appoach was to find a material that absorbs at specific wavenumber, e.g. 4590, 6370 (NH str. first overtone) in order to detect a combined signal of the NH backgound absorption and for example the absoption of a dendrimer with NH surface groups. The aim is to increase sensitivity and specificity. Another approach would be to bind the analyte of interest onto a material to increase the ability of the molecule to vibrate. But i�m not sure about that, just a try  So now the question is wheter to use a background material (BM) that has a strong absorption and to detect the scattering of the analyte of interest, or to use a BM that enhaces absorption in a wavenumber range of interest. I think in your case it makes sense to use a material that absorbs radiation in order to detect reflectance of the sample ? hmm.... Another problem is i�m not sure which theory to use to describe the scattering for particles at nano scale (Mie, Reyleigh). Thanks for sharing ideas, it really helps to get an insight of others NIR based problems! greetings, Nico. | ||
| Gavriel Levin (levin) Senior Member Username: levin Post Number: 33 Registered: 1-2006 |
Hello Mark, You are right, and thanks for your help, I tried graphite thin plate in the past, worked quite well, but still was a mess because when thin enough to useful to work with small seeds it breaks easily. I tried once developed X-ray film, it was so so, but easier to work with being thin and unbreakable. I will not go as far as powdered carbon black, too messy. Thanks a lot, I haven't seen you at IFPAC, have you been there, lots of NIR around. Gabi Levin Brimrose | ||
| Howard Mark (hlmark) Senior Member Username: hlmark Post Number: 72 Registered: 9-2001 |
Gabi - as far as I know, the "blackest" material available is carbon black in it's various forms; some may be "blacker" than others. The problem with it is that it's almost completely non-cohesive, so that you'd have to have a bed of the powder to use it in its pure form. Labsphere supplies a very dark standard by mixing in as little spectralon as they can manage and still get a cohesive mix; near as I can tell, this is what they sell as their 2% reflector. That's pretty dark, but if you want to get the absolute minimum amount of light from the background material, I don't see any way of avoiding the powder. And if you work with it, don't forget that this stuff is, essentially, "essence of black": it's the sort of stuff that paint, ink, and other purveyers of colors use for black - expect to get dirty, it doesn't take much! \o/ /_\ | ||
| Gavriel Levin (levin) Senior Member Username: levin Post Number: 32 Registered: 1-2006 |
I would like to know if possible the illumination set-up of the measurements. Were you using white light to illuminate and separated the wavelengths of the reflected light? If so, did you illuminate a large area and the actual sample covering only a small portion of the overall illuminated area, or were you illuminating only the small area of the sample? I am interested in this because sometimes I need to look at a small seed, and the illumination area is twice as large as the seed, or more (I am using preselected wavelenghts with AOTF and I scan the spectrum by changing the illuminating wavelength continuously) and I am using "spectralon" as background. This is not an ideal situation, because there is large reflected signal from the spectralon, and the signal coming from the seed is "weak" on a relative scale, so it reduces the sensitivity to the seed information. I then prefer to use a "shield", i.e., flat NIR black film in which I punch a small hole and place the seed on top, so I can collect the spectra of the seed only. But then I need to keep part of the seed "blocked" because it has to "sit" on the hole, some of it outside the perimeter of the hole, so my signal is reduced because of it. If there is a way to use a "background" that does not produce diffuse reflectance signal, it could simplify the set-up, I could just place it on the "background" and get the net signal. Any suggestions? Gabi Levin Brimrose | ||
| Tony Davies (td) Moderator Username: td Post Number: 143 Registered: 1-2001 |
Hello Nico, I'm sorry I had intended to reply earlier. I'm not sure about the nano scale but I have reported [Appl. Spec. 44, 1417-18 (1990)] spectra from as little as 10mg using sulphur as a background material. I agree with your proposed requirements. Best wishes, Tony | ||
| Nico Heigl (nch) New member Username: nch Post Number: 1 Registered: 1-2007 |
Hi there! I�m just doing measurements of materials, even at the nano scale with the help of "background materials" such as diamond and silica. My questions: Is it an appropriate way to analyse samples with the help of such background materials? I think cristalline materials and materials that consist of small particles, at best non porous, are quite well suited for that purpose. What do you think ? Thank�s a lot! Nico Heigl. |