IR Spectroscopy: Where the Peaks Actually Show Up
Most people memorize a chart and think they understand IR. They don't. The chart is a starting point, not a rulebook. Here is how the Ir Range For Functional Groups actually works when you are sitting in front of a spectrometer trying to figure out what you have. The carbonyl stretch is the most useful signal in the whole spectrum. It sits between 1650 and 1850 cm^-1 and it is sharp, strong, and impossible to miss if you have a good sample. Ketones land around 1715. Amides drop lower to about 1650-1690 because of resonance. Acid chlorides push up to 1800. Anhydrides show two bands split by roughly 40 cm^-1 apart. If you see two carbonyl peaks close together, think anhydride or carboxylate before anything else. The O-H region is messy and I mean properly messy. Free hydroxyl groups appear as a narrow spike near 3600 cm^-1. Hydrogen-bonded O-H stretches smear across 3200 to 3550 and often swallow neighboring C-H signals. If your sample is wet, the broad trough around 3300 might be water and not your compound. I learned this the hard way trying to characterize a phenolic product and spending three hours convinced I had a carboxylic acid dimer until I dried the KBr pellet over P2O5 overnight and the broad hump shrank to a manageable shoulder.
C-H stretching happens just below 3000 cm^-1 for sp3 carbons and just above 3000 for sp2 and sp. A terminal alkyne C-H shows up as a sharp spike near 3300 and it is easy to confuse with an O-H if you are not paying attention. The triple bond CC itself sits around 2100 to 2260 but it is often weak or invisible unless the molecule is unsymmetric. Nitriles give a cleaner signal around 2250. Below 1500 cm^-1 is the fingerprint region and everyone tells you it is useless. That is wrong. This is where you confirm identity against a reference spectrum. The patterns here are compound specific. Benzene rings show multiple sharp peaks between 1450 and 1600, plus overtones in the 1600 to 2000 range that follow substitution patterns. I use that region to check ortho versus para disubstitution when the substituent patterns are ambiguous from the main bands alone. Fingerprint peaks alone are not enough to identify an unknown. You need the functional group region above 1500 to build the first hypothesis and then the fingerprint to test it against known spectra.
Practical Problems With Real Samples
Water vapor and CO2 from the air will insert their own peaks into your spectrum if your instrument is not properly purged. Water adds a broad band near 1640 and a weaker one around 2130. CO2 produces a sharp doublet near 2350. If you see these and you did not add water or carbon dioxide to your sample, run a background scan with the sample chamber open and subtract it. This usually fixes the problem in under two minutes. KBr pellets can absorb moisture from the air if you leave them sitting out. I once spent forty-five minutes chasing a phantom O-H peak on a product that was supposed to be a clean ketone. The KBr had absorbed atmospheric water. Drying the pellet under vacuum before mounting solved it immediately. Conjugation shifts carbonyl stretches down by about 20 to 30 cm^-1. A conjugated ketone appears near 1685 instead of 1715. Ring strain does the opposite and pushes cyclic ketones higher. Cyclobutanone sits near 1780. Cyclohexanone is closer to 1715. If your observed carbonyl is significantly higher than the standard value for that class, check ring size before declaring an error in your synthesis.
Get the Full Details
Hydrogen bonding also affects N-H stretches. Primary amines show two N-H bands near 3300 and 3400. Secondary amines show only one. Tertiary amines show none. Amides follow a similar pattern but shifted lower, around 3180 to 3350 for the N-H stretch, plus the characteristic amide I and amide II bands near 1650 and 1550 respectively. Halides do not produce useful IR signals in most cases. C-Cl, C-Br, and C-I stretches fall below 800 cm^-1 where instrument sensitivity drops and peaks broaden. Don't rely on IR to confirm a halogen is present. Use NMR or mass spectrometry for that.
What IR Cannot Do Well
IR struggles with symmetric molecules and nonpolar bonds. A symmetrical alkene C=C stretch may be absent or extremely weak because there is no dipole change during vibration. Similarly, O2, N2, and other homonuclear diatomics are completely invisible in IR. This is why you should never use IR as your only characterization technique for a new compound. Quantification is also unreliable without careful calibration. Peak height varies with path length, concentration, and instrument alignment. Peak area is more consistent but still requires a standards curve for accurate work. If you need concentration data, use UV-Vis or HPLC instead. IR is for functional group identification, not quantitation. The technique also fails for aqueous solutions because water absorbs strongly across almost the entire mid-IR range. You can use ATR accessories with thin samples to minimize water interference, but pure aqueous samples are essentially opaque. Use D2O exchange for NMR or switch to Raman spectroscopy if your compound is water-soluble and you need vibrational data.
Quick Reference For Common Functional Groups
O-H stretch (alcohol): 3200-3550 cm^-1 broad
O-H stretch (carboxylic acid): 2500-3300 cm^-1 very broad, often overlaps C-H
N-H stretch: 3100-3500 cm^-1
C-H stretch (sp3): 2850-2960 cm^-1
C-H stretch (sp2): 3010-3100 cm^-1
C-H stretch (sp): ~3300 cm^-1 sharp
C=O stretch: 1650-1850 cm^-1 strong
CN stretch: ~2250 cm^-1 medium
CC stretch: 2100-2260 cm^-1 weak to medium
C-O stretch: 1000-1300 cm^-1 strong
NO2 asymmetric stretch: ~1550 cm^-1
NO2 symmetric stretch: ~1350 cm^-1 Keep this list on your desk. Reference it while you interpret. Do not memorize it and expect perfect recall under pressure. The numbers shift with environment, substitution, and instrument type. The ranges above are guidelines, not laws. When a peak falls outside the expected window, check your sample purity, your solvent, your cell path length, and your calibration before you change your structural assignment. IR spectroscopy is fast and it is cheap. A routine scan takes three minutes on a modern FTIR with ATR. The trade-off is that speed comes with limitations. Learn those limitations before you trust the spectrum to make a decision that affects the rest of your project.
