The carbonyl peak is at 1689 cm⁻¹ with the two peaks for the NO₂ group at 1530 and 1347 cm⁻¹.

The hydroxyl group has a broad peak centred around 3334 cm⁻¹ with the two peaks for the NO₂ group at 1521 and 1345 cm⁻¹.

With modern instruments it tends to be very difficult to record a spectrum without a background to subtract, and the example below has been extracted from the datafile. A spectrum where no background has been recorded will be the sum of the peaks observed in the background, in this case absorbances from the diamond crystal, along with absorbances from the air (particularly water vapour and CO₂), along with absorbances from the compound. It is likely low transmission values would be observed for the y-axis (values omitted here).

The example below is for a sample of 3-nitrobenzaldehyde where the background has not been subtracted.

With solid samples, it is necessary for force to be applied to the sample (using the clamp) in order to achieve sufficient contact with the crystal surface for Evanescent waves to be absorbed by the compound. In the sample below, insufficient pressure was applied to the clamp, which resulted in low levels of signal. This can be seen from the limited transmission values on the y-axis (strongest bands are only seeing about a 3% drop in transmittance), and results in a very noisy baseline. The same effect can also be observed if sufficient pressure is observed, but there is very little compound on the crystal, e.g. the compound gets knocked off the crystal as pressure is applied.

The sample below is a spectrum recorded for a sample of 3-nitrobenzaldehyde where no pressure has been applied to push the solid sample into contact with the diamond.

If the spectrometer is not clear of sample from the ATR crystal when the background is recorded, then the absorbances from the compound will be included within the background data. When this contaminated background is subtracted from the acquired data to produce the spectrum, no peaks will be present. The data is effectively the same for both the acquired sample and the background measurement.

The data below is a spectrum recorded for a sample of 3-nitrobenzyl alcohol where the background had been contaminated with the same compound. Note that whilst peaks do seem to be present, the y-axis changes are tiny, and we are essentially viewing a zoomed in baseline, which is why the baseline appears so noisy.

Where the background measurement has been contaminated by another compound, this will result in absorbances due to the contaminant being subtracted from the measurement for the compound of interest and will likely lead to transmission values which are greater than 100%, with an appearance of peaks going in the wrong direction. This can easily be avoided by ensuring that the ATR crystal is cleaned prior to the background measurement being recorded.

The data below is a spectrum recorded for a sample of 3-nitrobenzylalcohol where the background measurement had been recorded with the ATR crystal contaminated with 3-nitrobenzaldehyde. Note the y-axis has values which are greater than 100%, with the carbonyl peak of the aldehyde particularly prominent with a negative absorbance.

The example above shows the effects of a background being recorded which is contaminated with another compound. This example shows the same process, where the contaminant is the solvent used to clean the ATR crystal (usually isopropanol). This arises if the solvent used to clean the crystal isn't cleaned off or allowed to evaporate before the background measurement is acquired.

The data below is a spectrum recorded for a sample of 3-nitrobenzyl alcohol where the background was recorded before the isopropanol used to clean the ATR crystal had been allowed to dry. Below this is a spectrum for isopropanol. The peaks in the isopropanol at 2970 and 950 cm⁻¹ are particularly prominent as negative peaks in the contaminated recording.

This data collection issue doesn't relate to instrument use problems, but arises from the sample. When recording an IR spectrum for a compound, it is desirable for the compound to be free of any contaminants. The most commonly encountered contaminations arise from incomplete removal of solvents prior to data acquisition. Simple steps such as allowing the compound to dry for longer e.g. by spreading the compound out on a watch glass and allowing to air dry, can really aid the quality of the data obtained.

In this recording, a sample of the 3-nitrobenzaldehyde has been recorded which also contains water from the recrystallisation step prior to data acquisition. Allowing the compound to have dried for longer, prior to recording the IR spectrum would have minimised or eliminated the absorbance observed here from the water present in the sample. The broad peak around 3300 cm⁻¹ is due to the water in the compound.

Below is a spectrum of water, and by comparing the peaks in this spectrum with the contaminated sample above and the good spectrum for 3-nitrobenzaldehyde, the contamination of the 3-nitrobenzaldehyde spectrum with water is fairly clear, with the OH absorbance centred around 3300 cm⁻¹ is evident. Similar effects can be observed if samples are wet with other solvents, e.g. ethanol or ethyl acetate. Usually the easiest solution is to allow the compound to dry (e.g. by spreading out on a watch glass) and repeating the measurement when the compound has had more time to dry. The sample data folders contain spectra for a range of common solvents.