Antaris Site Map

Our Answers to Your Questions

Q1: I want to do tablet API quantification by FT-NIR but I am not sure which technique is right for me, reflection or transmission?

Thermo Scientific: Depending on the tablet type, either would be suitable. If you are doing uncoated tablets, you can get reliable results using reflection analysis. If the tablets are coated or bi-layered, however, most of the light that comes off the sample is due to the first 300 microns or so of the tablet at the focus which would make reflection inappropriate for this application. In general, transmission analysis is the best experiment for API qualification as it tends to give a better correlation coefficient and lower RMSEC.

Q2: I want to take measurements of material in a fluid bed dryer as it is running. Would I put an analyzer up to the sight window and take data that way?

Thermo Scientific: The best way to do data analysis in a fluid bed dryer is by inserting a probe into the dryer. The material that sits at the window is actually a crust that forms during dryer operation that has little to do with the concentration of residual solvent (Closed cycle) or moisture (Open cycle) in the solid cycling inside the dryer. The best solution here is to use either a purgeable-tip probe or a retractable, self-cleaning probe that will actually penetrate into the dryer's sample stream. Then you can take NIR measurements normally as there is a good deal of powder cycling around throughout the drying process. These probes are available for purchase with Antaris FT-NIR analyzers and have been used successfully in many different fluid bed dryer applications.

Q3: I am trying to do Raw Material Identification and I know that FT-NIR can penetrate thru the sample bag. But don't you get a residual signal from the bag in the process? Doesn't that affect your calibration?

Thermo Scientific: Yes, you definitely see a signal from the plastic. However, this issue is easily solved with how you set up your calibration. There are 3 major strategies of factoring bags out of a NIR calibration.

• If the bags are all the same polymer, typically the variability in the bag thickness doesn’t hurt the performance of a Raw Material ID method. So the answer here is to do nothing and leave the bag resonances in your calibration. This is the easiest strategy is actually the easiest to implement. Difficulty: Easy
• Change the regions in your calibration to exclude the polymer resonances. This will perform as if there were no plastic at all. To do this, go into your TQ Analyst chemometric method and simply click and drag until your analysis regions exclude the polymer. This strategy is still easy but involves going into your method to make the change. Difficulty: Intermediate.
• Do an automated spectral subtraction to remove most of the polymer resonances. The tolerances on commercial plastic bag thicknesses is good enough that whatever resonance are left over from the bag after the subtraction will be minimal. You can also set RESULT to reject spectra where the residual is too high. Difficulty: Advanced.

Q4: I am using a process control system that runs on OPC and I want this sytem to tell the NIR that it is time to start or stop an analysis.

Thermo Scientific: Absoutely. RESULT software has an internal OPC server that can also take commands like Start and Stop at anytime during a worklow. All you need to do is figure out where in your analysis, you would like to have a Start/Stop and simply click "Add Start Stop."

Q5: I want to measure a thick, viscous sample (molasses) for moisture and sugar profile by NIR but trying to get this stuff into vial or cuvettes is really messy. Any ideas?

Thermo Scientific: There is a NIR accessory with Antaris NIR Analyzers that will allow either transmission or transflectance measurements to be made at varying sample pathlengths for viscous samples. The accessory uses disposable polyethylene bags so there is no cleaning involved. Samples can be easily placed into the bag and sealed avoiding the mess created trying to place a viscous sample in a cell.

Q6: I have a heterogeneous cattle feed sample with varying particle size that I would like to analyze "As Is" by NIR to avoid grinding and Dry Basis corrections.

Thermo Scientific: You will want to use a sample cup spinner accessory over the integrating sphere. Sample spectra are continuously collected as the sample is rotated through the NIR beam. The resulting diffuse reflectance spectrum is representative of the bulk sample. This type of sampling eliminates the need to take multiple spectra of the same sample and use the average of these individual results to generate an answer.

Q7: I have a GC in my lab to monitor the transesterification reaction for conversion of oil to biodiesel. It takes almost 30 minutes for the answer. How fast can NIR do this?

Thermo Scientific: NIR gives important biodiesel component concentrations such as glycerin and glycerides in real time, analysis time is less than 1 minute. An FT-NIR installed inline using fiber optic probes produces a continuous flow of results for trending the completion of the transesterification reaction. The FT-NIR analyzer can also give results for contaminants such as water, methanol and free fatty acids in biodiesel process streams.

Q8: Can you do analysis through amber vials for lyophilized materials?

Thermo Scientific: Lyophilized proteins and materials are a great application for NIR. NIR has a large penetration depth and can readily go through glass and plastic to analyze the material inside. Typically, lyo vials are clear and colorless but there are some enzymes or proteins that require amber vials. The amber vials typically have an absorption in the low energy region of the NIR so what this means is you will have the change the regions in your method to avoid these resonances.

Q9: We use are a lactose supply house and we have many different particle sizes of this material in house. Since it is all lactose, can NIR distinguish?

Thermo Scientific: NIR spectra can actually be successfully used to distinguish between different sized particles of the same chemical. While it is true the absorbance peaks are going to be similar, the spectral baselines of the different particles will likely be different. Larger particles will have different baselines and slopes than smaller particles. When analyzing these samples, it is best to use the raw spectra as opposed to the derivative spectra. Using derivative spectra tends to remove these baseline differences which will make it more difficult to discriminate between different-sized particles.