Nicolet™ Summit FTIR 光谱仪

<p>DRIFTS is used in both mid-IR and near-IR. In the mid-IR, DRIFTS requires the sample be blended with diluents like KBr, with 3-10% sample. This is typically undesirable as the sample is now mixed. However, DRIFTS is heavily used in catalysis research where powdered material is exposed to high temperature, elevated pressures, and mixtures of reactant gases. Several accessory suppliers make devices specific for this. In the near-IR, DRIFTS is used without dilution through direct measurement - many hand-held probes exist allowing analysis through a container wall (like plastic bags) meaning the sample can be analyzed without touching or contaminating it. <br/><br/> ATR involves making contact with the sample by forcing it into contact with a crystal. ATR generally does not require dilution and works well with solids like credit cards or car bumpers which would be tough in DRIFTS. ATR has, for the most part, displaced DRIFTS in the mid-IR except in special cases, while DRIFTS remains a method of choice in the near-IR world.</p>

<p>Fourier Transform Infrared Spectroscopy (FTIR) responds to a change in dipole moment, regardless of whether it is organic or inorganic. Metal oxides, carbonates, and carbonyls are good examples. The basic equation states that the wavenumber is proportional to the square root of the spring constant (bond strength) and one over the square root of the reduced mass. Simply put, as mass of the atoms involved in the bond goes up, the wavenumber goes down. Many inorganics have peaks below 400cm-1, such as ferrocene, acetylferrocene and cadmium oxide. This necessitates the use of “far-IR” optics. Many forensics users have found far-IR useful in identifying paint chips, due to their inorganic content. There are several ATR accessories that now permit far-IR ATR (mostly monolithic diamond devices). The Thermo Scientific Nicolet iS50 FTIR Spectrometer was designed to make far-IR performance trivial with a built-in ATR as well. Ultimately, if you have further interest in this area, you need to speak with an FTIR sales person to understand the capabilities and limitations.</p>

<p>Quantitative work requires one of two approaches. First, you may know or can calculate the absorptivitiy (the epsilon in Beer's Law). This is exceedingly rare. Much more commonly, you can develop a set of training / calibration standards and record spectra. You can then either use a chemometrics package like Thermo Scientific TQ Analyst software to automate the analysis using Beer's Law or more complex modeling, or you can record basic information (peak height or peak area) in a spreadsheet and then use linear (or non-linear) regression. It is generally the same idea as done for chromatography or atomic spectroscopy.</p>

<p>Apodization can increase from light (like Happ-Genzel) to heavy (like Blackman-Harris). The heavier the function, the more effect you will see on lineshape. For most normal uses, when the resolution is 4 wavenumbers or more, even heavy apodization does not distort spectra badly. However, as the lines narrow, as in gas-phase spectroscopy, the effect of apodization can be profound. Strong apodization in broad peaks improves the signal to noise ratio over Boxcar (essentially no apodization) by a lot with minimal effect on linewidth. Happ-Genzel was the original OMNIC default (and still often is) due to its “gentle” effect on linewidth coupled with decent signal to noise ratio. Generally, signal to noise ratio improves in the order Happ-Genzel, Norton-Beer-Weak, Norton-Beer-Medium, Norton-Beer-Strong, Blackman-Harris, while lineshapes are also affected more in the same order.</p>

<p>Apodization works on the interferogram. If you have that raw data, which Thermo Scientific OMNIC software can do by setting a single check box, then post-processing by changing apodization is trivial. The interferogram typically requires a lot more memory than the Fourier transformed spectra, so older software packages used to throw away the interferogram to save memory. In cases where the interferogram is not present, you can't reprocess as apodization is a time domain function.</p>

<p>Presenting spectra as transmittance tends to emphasize the smaller peaks, so you can sometimes visually assess your sample better. Absorbance spectra are used for any quantitative analysis, spectral subtractions or other manipulations as the spectra are linear with concentration whereas transmittance spectra are not. For searching or more general use, it is really more of a personal preference. Older literature tended to use transmittance most often, while detailed peak analysis always moved to absorbance, again because of the linearity feature.</p>