Yes, the OH proton does show up on 1H NMR, but almost never as the sharp, predictable peak that CH protons give. It appears as a broad singlet, its chemical shift wanders with solvent, concentration, and temperature, and in protic solvents like D2O it disappears entirely through exchange. That variability is why most structure-elucidation workflows treat the OH signal as a diagnostic afterthought rather than a fixed reference point.
The proton on an alcohol or phenol oxygen is a real hydrogen with real magnetic resonance. What makes it difficult is not whether it resonates but how the surrounding chemistry controls when, where, and how sharply it appears. Understanding that control is the difference between reading an OH peak confidently and misassigning it. This explainer works through the mechanism, the numbers, the edge cases, and the practical steps that separate a clean assignment from a guess.
Exchange mechanism: why the OH proton broadens
The OH proton sits on an electronegative oxygen that pulls electron density away from hydrogen. That deshielding pushes the resonance downfield relative to most alkyl CH signals. But the dominant behavior comes from chemical exchange, not shielding. The hydroxyl proton hops between oxygen atoms through hydrogen bonding and trace acid or base catalysis, and the rate of that hopping determines the linewidth.
NMR sees an averaged environment only when exchange is fast on the NMR timescale. For a proton, that timescale is set by the reciprocal of the chemical-shift difference in hertz. If two exchanging environments differ by, say, 200 Hz at a given field, exchange faster than roughly 10³ per second collapses them into one averaged line. Slower exchange broadens the line, and intermediate rates can broaden it almost into the baseline. This is why OH peaks are notoriously variable in width: the molecule is not changing, but the exchange rate is.
The exchange rate itself depends on traces of water and acid. A few microliters of D2O or a speck of HCl can catalyze proton transfer and sharpen or collapse the OH signal within minutes. A rigorously dried sample in an aprotic solvent often shows a sharper, more reproducible OH peak precisely because exchange is slow. The same compound can therefore give a broad lump in one tube and a defined peak in another. The proton is present in both cases; only the dynamics differ.
Quantified context: where the OH proton lands
The chemical shift of an OH proton spans an unusually wide window, roughly 0.5 to 5.0 ppm in most cases, with phenols reaching 4 to 12 ppm and carboxylic acid protons often between 10 and 13 ppm. That range overlaps the aromatic region and the aldehyde region, which is the practical problem. A broad peak at 7.2 ppm could be an OH, an aromatic CH, or an NH, and shift alone cannot decide it.
Concentration shifts the equilibrium between free and hydrogen-bonded hydroxyls. Diluting a sample typically moves a hydrogen-bonded OH upfield toward its free value, sometimes by 0.5 to 2 ppm across a dilution series. Temperature has a similar effect: warming weakens hydrogen bonds and usually shifts the OH upfield while sharpening it. Running a variable-temperature series is a standard way to confirm an exchangeable proton, because the shift tracks temperature in a way CH signals do not.
| Signal type | Typical shift (ppm) | Linewidth | Exchange behavior |
|---|---|---|---|
| Alcohol OH | 0.5–5.0 | Broad, variable | Exchanges with D2O |
| Phenol OH | 4–12 | Broad to sharp | Exchanges, often slower |
| Carboxylic acid OH | 10–13 | Very broad | Exchanges readily |
| Amine NH | 0.5–5.0 | Broad | Exchanges with D2O |
| Amide NH | 5–9 | Broad | Slow exchange, often visible |
| Aromatic CH | 6.5–8.5 | Sharp | No exchange |
| Aldehyde CH | 9–10 | Sharp | No exchange |
The table captures why shift matching fails as a standalone test. Linewidth and exchange behavior do the discriminating work.
Edge cases: when the OH proton vanishes or sharpens
Deuterium exchange is the decisive edge case. Shaking a sample with D2O replaces the OH proton with deuterium, which is NMR-silent at the proton frequency. The OH peak disappears, and that disappearance is positive evidence for an exchangeable proton. The same treatment removes NH and SH signals, so D2O shake confirms exchangeability without specifying which heteroatom carries the proton.
Intramolecular hydrogen bonding is the counterintuitive case. A phenolic OH locked into a strong intramolecular hydrogen bond, as in salicylaldehyde or certain ortho-substituted phenols, can appear as a relatively sharp, concentration-independent peak far downfield. That behavior mimics a nonexchangeable proton and misleads analysts who expect all OH signals to be broad. The sharp downfield singlet is not a sign that exchange is absent; it is a sign that the proton is sequestered from the bulk solvent.
Very slow exchange produces another trap. In rigorously dry, aprotic solvents at low temperature, an OH proton can appear as a sharp line or even resolve into coupling with neighboring CH protons. That coupling is a gift when it appears, because it proves the connectivity directly. But it is fragile and disappears on warming or on adding a trace of water.
Misconceptions and professional caveats
The most common error is expecting a fixed OH chemical shift and rejecting a structure when the peak does not match a table value. Shift tables are ranges, not constants, and the OH range is the widest of any common proton. A second error is treating a missing OH peak as evidence the proton is absent. In protic solvents, at low concentration, or after inadvertent water exposure, the peak can broaden into the baseline and be missed entirely.
A subtler caveat concerns integration. A broad OH peak can integrate poorly because it overlaps the baseline and because exchange broadens it beyond the integration window. Analysts who total the integrals and find them short by one proton often have a broad exchangeable signal hiding under the noise. Reproducible quantification of an OH proton requires suppressing exchange, which means dry, aprotic solvent and careful sample handling.
Red flag: a broad peak that shifts position between repeat scans suggests exchange, not a new compound. Red flag: an OH signal that disappears after the sample sits overnight often reflects absorbed atmospheric water. Red flag: a sharp downfield singlet near 10–12 ppm may be a strongly hydrogen-bonded phenol, not a carboxylic acid. Red flag: overlapping OH and aromatic signals can inflate aromatic integrals and hide the exchangeable proton. Red flag: a "missing" OH in D2O is expected, not a failed synthesis.
Practical actions: confirming the OH proton
Confirming an OH proton is a workflow, not a single measurement. The first step is to record the spectrum in a dry, aprotic solvent such as CDCl3 or DMSO-d6, where exchange is slow enough to preserve a visible signal. The second is to note the linewidth and shift, then run a D2O shake and re-record. Disappearance of the candidate peak confirms exchangeability.
For ambiguous cases, a variable-temperature series settles the question. An exchangeable OH shifts upfield and usually sharpens on warming, while a CH signal shifts only slightly. A concentration-dilution series provides a second confirmation, since a hydrogen-bonded OH moves upfield on dilution while a CH does not. When connectivity matters, a 2D experiment such as COSY or HMBC can link the OH proton to its neighboring carbon framework and remove the ambiguity entirely, though the OH cross-peaks are often weak because of exchange.
| Confirmation step | What it tests | Expected result for OH |
|---|---|---|
| D2O shake | Exchangeability | Peak disappears |
| Variable temperature | Exchange dynamics | Shift and width change with T |
| Dilution series | Hydrogen bonding | Shift moves upfield on dilution |
| 2D COSY/HMBC | Connectivity | Cross-peaks to adjacent atoms |
Reconciling the thresholds: a candidate OH is credible when it is broad, shifts with concentration or temperature, and vanishes on D2O addition. Any two of those three are strong support; all three are conclusive in routine practice.
Limitations and what NMR cannot tell you
NMR identifies an exchangeable proton and, with 2D methods, its connectivity, but it does not by itself distinguish an alcohol OH from an amine NH or a thiol SH. All three exchange with D2O and all three can appear broad in the same shift window. Distinguishing them requires supporting evidence such as infrared stretching frequencies, heteronuclear correlation to the attached atom, or chemical derivatization.
Quantification of OH protons is inherently unreliable because exchange broadens the signal and undermines integration. Reporting an OH integral to two decimal places overstates precision the experiment cannot deliver. The honest conclusion from a routine 1H spectrum is qualitative: an exchangeable proton is present, its approximate environment is consistent with the proposed structure, and its exact shift is a property of the sample conditions rather than the molecule alone.