Credit: Jennifer Lynch, NIST
Infrared (IR) spectroscopy, a form of vibrational spectroscopy, offers useful information for a wide range of both organic and inorganic samples. Used in numerous applications ranging from environmental to pharmaceutical to forensic, IR spectroscopy can provide both qualitative and quantitative information about the compounds present in a liquid, solid or gaseous sample. The transmittance or absorbance spectra of a sample can reveal which functional groups are present, as well as a unique “fingerprint” for each substance, but accurate and efficient identifications rely on both a high quality spectrogram and the skills and resources to interpret it. Here are a few tips to speed and simplify your analysis when investigating IR spectra:
1. Identify and avoid common spectral interferences
An ideal, easy-to-read IR spectrogram would have low noise, a flat baseline and no artifacts from atmospheric interferences like water and carbon dioxide.1 Obtaining a flawless spectrogram for every run would be a miracle, but knowing how to identify interferences in the spectra and how to mitigate them are important precursors for successful qualitative analyses.
Noise in the spectrogram is easy to spot from the “fuzzy” appearance of the baseline. Noise around 4000 to 3400 cm-1 and 2000 to 1300 cm-1 typically comes from water vapor, while noise near 2350 cm-1 and 670 cm-1 comes from CO2, which is easy to see in a background run. Depending on the severity of the interference, ambient water and CO2 can leave quite large artifact peaks in the spectrogram – locating and identifying these interferences should be among the first steps you take when examining the spectral output.
Noise and atmospheric peaks can be eliminated either by taking steps to prevent them during the run or by using digital methods to correct them. First, it’s important to perform the measurement run as soon as you can after running the background, as water and CO2 levels in the instrument can tend to fluctuate.2 When you notice these artifacts appearing in the spectrogram, some options include trying again with a fresh background, purging the instrument with dry nitrogen or changing your desiccant to better reign in water activity. Some IR software also include atmospheric correction functions that can clean up artifacts after the fact. Overall, it’s important to address interferences before you attempt to assign other peaks, and learning to recognize where the noise is coming from will make you more efficient in reducing it.
2. When assigning peaks, start with the “low-hanging fruit”
Trying to read an IR spectrogram from end-to-end, examining each and every peak in wavenumber order, can make the whole process much more cumbersome and time consuming than it needs to be. You can better prioritize your time by first assigning the most obvious and intense spectral bands – in other words, start with the “low-hanging fruit.”1,3 The largest peaks in the functional group range (4000 cm-1 to 1500 cm-1) are often not only the easiest to assign, but also among the most useful for narrowing down what is in an unknown sample.
Two of the most useful peaks to look for when dealing with an unknown sample are a broad, rounded peak around 3400-3200 cm-1 and a sharp pointed peak around 1850-1630 cm-1, representing hydroxyl groups (OH) and carbonyl groups (C=O), respectively.3 From just these two peaks, if they are present, you can glean more information based on their specific position and shape; for example, a hydroxyl peak right around 3300 cm-1 is typical of alcohol, while much broader hydroxyl peaks extending below 3000 cm-1 can indicate a carboxylic acid. The specific location of a carbonyl peak can help you determine whether it is part of an aldehyde, ketone, ester, carboxylic acid, amide or anhydride. Furthermore, identifying these obvious peaks help you assign less intense secondary bands and look for “companion” peaks to further confirm and narrow down your analysis.1,3
Of course, the usefulness of certain prominent peaks in a spectrogram will vary depending on your specific application; for example, if you are routinely analyzing alcohols, the presence of an OH group is far from a smoking gun. Whatever “low-hanging fruit” looks like to you, peaks that stand out the most can often be the best starting point to launch your investigation.
3. Remember that what you DON’T see may be as important as what you do
The peaks in a spectrogram allow you to see which functional groups are present in the sample, which can lead you down the path to identification based on this combination of groups. However, knowing which functional groups are NOT present in the sample can be just as useful for narrowing down the possibilities, and empty regions of the spectrogram should be considered along with the peaks. This is when scanning from left to right and taking note of what you see in each region is helpful – an empty area from 3500-3200 cm-1 lets you rule out O-H and N-H bonds, and a lack of peaks in the C-H stretch region eliminates a large number of organic compounds. Once you’ve identified your most prominent peaks and noted obvious absences, you could be well on your way to solving the molecular puzzle with little more than a glance.
4. For database searching, consider building your own custom library
Modern technology has made interpreting spectra far easier than it used to be – gone are the days of eyeballing comparisons between sample spectra and reference spectra printed out on paper. With reference libraries at our fingertips and algorithms available to search thousands of spectrograms for a possible match, demystifying the sticky “fingerprint” region and reaching an accurate identification could be just a few clicks away. Both public libraries like the NIST database and commercial libraries available for purchase are valuable assets for a wide range of applications, but in some cases the best resource can be the spectra collected at your own lab.4 Many IR software packages make it easy to compile your own custom library, and there are several benefits to building a custom database to search from.
One major benefit of constructing a custom database of known substances from your own lab is ensuring that the samples you typically see are fully represented in the library.4,5 Another benefit is having spectra from your own instruments, as spectra can appear different depending on factors like instrument resolution and sample handling methods. Having references measured under the same conditions in the same laboratory as the sample can increase the likelihood of finding a close match, and building a custom library can be cost-saving as it lessens the need to purchase libraries to account for specific or uncommon samples you might see in your work. When building a custom library, it’s important to include only high-quality spectrograms and that you are completely certain of the identification of the sample being used as a reference.4
References
- B.C. Smith, Spectroscopy 31(1), 14–21 (2016). https://www.spectroscopyonline.com/view/process-successful-infrared-spectral-interpretation
- "FTIR Analysis Q&A," Shimadzu. https://www.shimadzu.com/an/service-support/faq/ftir/4/
- "Infrared Spectroscopy: A Quick Primer On Interpreting Spectra," James Ashenhurst, Master Organic Chemistry (2020). https://www.masterorganicchemistry.com/2016/11/23/quick_analysis_of_ir_spectra/
- B.C. Smith, Spectroscopy 36(3), 24-27 (2021). https://www.spectroscopyonline.com/view/library-searching
- "Techniques for Obtaining Infrared Spectra," Webinar Presented by Andrea B. Champagne, The McCrone Group (2015). https://www.mccrone.com/mm/techniques-obtaining-infrared-spectra/