New Year’s Resolutions and Infrared Resolutions

As we get older, the grand schemes to force ourselves into the best we can be (thinner, healthier, nicer…) become simplified into more attainable New Year’s resolutions such as flossing every day.

IR spectroscopy also has the deluxe instrumentation schemes that offer everything under the sun and the simpler versions that can make a measurement goal easier and more economical to attain.  In spectroscopy, resolution is often used as a yardstick of instrument performance.  This is especially true when the measurements are energy limited as they often are in the mid infrared region.

For qualitative analysis, resolving individual peaks can be important in distinguishing one molecule from another.  With quantitative measurements, on the other hand, a lower resolution could mean higher performance and better accuracy.

Using the example of biodiesel in diesel, the spectra below show 5% and 20% samples at 4 cm-1 and 16 cm-1 resolution with a 10 bounce ATR (Attenuated Total Reflection) sample stage.  Biodiesel is typically measured at the carbonyl band (5.7 µm or 1745 cm-1).

Note that the wider and less intense peaks are the 16 cm-1 resolution scans.  While the spectra would make it appear that the higher resolution (4 cm-1) would give better results, the opposite was found to be true.

Infrared spectra comparing IR resolution for biodiesel in diesel

 A calibration was generated using a measurement area of half the maximum peak intensity on either side of the carbonyl peak.  A baseline correction was applied at either end of the peak.

The calibrations were analyzed using leave-one-out cross-validation, which allows a prediction to be made as to how robust the calibration is since it forces a calibration to predict samples not seen in the calibration set.  Both the cross-validation results and the Standard Error of Cross Validation (SECV) are in Figure 2 below.

IR calibrations at different resolutions

Figure 2: Cross Validation Results

Figure 2 shows that the low resolution calibration performed better on most samples than the high resolution samples.  The SECV of the low resolution sample was calculated to be half the SECV of the high resolution.

While the high resolution spectra had a more visibly intense and clearly defined peak, there is a trade off in the signal-to-noise ratio.  Typically, instruments operating at a lower resolution will have a higher signal to noise ratio as shown in Figure 3.   This lower noise level with lower resolution translates to a better performing calibration.

IR noise at different resolutions

Figure 3: Noise Levels of the Nicolet FTIR at 4 & 16 cm-1

Portable handheld Mid IR Spectrometer

InfraSpec VFA-IR Spectrometer

Since lower resolution infrared instruments can be made with no or few moving parts, they are easier to make more rugged and portable for field use.  The InfraSpec VFA (Variable Filter Array)-IR Spectrometer, manufactured by Wilks Enterprise, Inc., is a typical example of an instrument that provides very good signal to noise in a compact low resolution design capable of accurate, on-site quantitative analysis for a wide variety of measurements.

Lowering infrared resolution, as well as our New Year’s resolution expectations, may actually lead to better results!

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s

%d bloggers like this: