The 60-second version
When an ion fragments in a tandem mass spectrometer it breaks into a charged piece (which the instrument can see) and a neutral piece (which it cannot — no charge, no steering, no detection). The neutral piece is only revealed by the mass gap between precursor and fragment: the neutral loss, Δm.
A neutral-loss spectrum is a filtered spectrum: on a triple-quadrupole, Q1 and Q3 are scanned together, always Δm apart. A compound reaches the detector only if it loses exactly Δm in the collision cell. The output looks like an MS¹ spectrum (intensity vs. precursor m/z) but contains only members of one structural class — e.g. Δm = 98 Da shows only things that lose H₃PO₄ → phosphopeptides; Δm = 176 Da → glucuronide metabolites.
1 · Why “neutral” losses are invisible
Mass spectrometers manipulate and detect charged particles with electric and magnetic fields. A neutral fragment keeps the precursor's velocity, drifts out of the ion beam, and is pumped away — it never produces a peak. You only ever infer it from arithmetic:
[M+H]⁺ at m/z 305 → fragment⁺ at m/z 261 (detected) + CO₂ · 44 Da (neutral — invisible, inferred from 305 − 261)The gap of 44 Da is a fingerprint: something in that molecule was a carboxyl group. Different chemistries give characteristic gaps — water (18), ammonia (17), CO₂ (44), SO₃ (80), H₃PO₄ (98), glucuronic acid (176)… That turns a mass difference into a class test.
GC-EI note: in classical GC-MS the molecular ion M⁺• and its fragments are all born in the ion source, so one EI spectrum already contains these gaps (M−15, M−18, M−29…). Reading them is neutral-loss analysis — see the interactive explorer in section 5.
2 · The triple-quadrupole neutral-loss scan — simulator
Below is a virtual Q–q–Q. Q1 scans precursor mass M upward; q2 is a collision cell filled with N₂/Ar; Q3 scans in lock-step at M − Δm. Watch the readouts under Q1 and Q3 stay exactly Δm apart. Colored dots are ions; gray dots are neutral fragments (invisible to a real detector). Ions of a compound only enter q2 when the scan reaches their m/z — that's Q1 doing its filtering job.
What to watch for
- Q1 and Q3 move together, glued Δm apart. Nothing else about the hardware changes during the scan.
- In q2, each precursor fragments. Only the fragments whose mass equals the current Q3 setting (= precursor − Δm) survive Q3 and hit the detector (flash). Wrong-mass fragments are deflected into the rods.
- Analyte C has no labile neutral group — it is abundant in the full scan (top panel) but never appears in any neutral-loss spectrum. Absence of a peak ≠ absence of the compound.
- Analyte D and the phosphopeptide lose two different neutrals — they show up in two different NL spectra.
- Peak height ∝ abundance × fragmentation yield, not abundance alone. NL spectra are screening tools, not quantitative ones.
3 · How to read a neutral-loss spectrum
- X-axis = precursor m/z (same axis as an MS¹ scan). Y-axis = detector counts.
- A peak at m/z M asserts two things at once: a precursor of mass M exists and it ejects Δm under CID → the molecule contains the corresponding functional group.
- It is a class filter on a dirty matrix: e.g. a 176 Da NL scan of urine extract shows glucuronide metabolites and hides almost everything else.
- It tells you nothing about where in the molecule the loss came from, and gives no full fragment fingerprint — follow up hits with product-ion scans or HRMS.
- QqQ scans run at unit resolution: isobaric coincidences cause false positives. A peak at M could be an unrelated M′ that happens to satisfy M′ − fragment′ = Δm. Confirm.
- Intensity depends on collision energy and fragmentation yield — don't compare abundances across compound classes.
4 · Common neutral losses worth memorizing
| Δm (Da) | Neutral species | Typical compound class | Where you meet it |
|---|---|---|---|
| 17.03 | NH₃ | amines, amino acids, peptides | LC-ESI & CI |
| 18.01 | H₂O | alcohols, carboxylic acids, peptides | everywhere — GC-EI & LC-ESI |
| ~28 | CO (27.99) / C₂H₄ (28.03) | aldehydes, ketones / ethyl groups | GC-EI |
| 32.04 | CH₃OH | methyl esters, methoxy compounds | GC-EI |
| 44.00 | CO₂ | free carboxylic acids, carbamates | LC-ESI, GC-EI |
| 56.06 | C₄H₈ (isobutene) | Boc-protected amines | LC-ESI (med-chem) |
| 63.96 | SO₂ | sulfonamides, sulfones | LC-ESI |
| 79.96 | SO₃ | sulfate conjugates, sulfonates | LC-ESI (metabolite ID) |
| 97.98 | H₃PO₄ | phosphopeptides (pSer/pThr), phosphate metabolites | LC-ESI (phosphoproteomics) |
| 129.04 | C₅H₇NO₃ (pyroglutamic acid) | glutathione (GSH) adducts | LC-ESI (reactive-metabolite screening) |
| 162.05 | C₆H₁₀O₅ (hexose) | O-glycosides, flavonoid glycosides | LC-ESI (natural products) |
| 176.03 | C₆H₈O₆ (glucuronic acid) | glucuronide conjugates | LC-ESI (drug metabolism) |
5 · The same idea, two different worlds
GC-MS
- Most GC-MS systems are single-quadrupole EI — there is no collision cell and no Q3, so a hardware NL scan doesn't exist (you'd need a GC-QqQ).
- Instead, neutral losses live inside every EI spectrum: read the gaps below the molecular ion. M−15 (•CH₃), M−18 (H₂O), M−29 (•C₂H₅/CHO), M−31 (•OCH₃), M−43, M−60 (acetic acid from acetates), M−90 (TMS-OH from silylated analytes).
- In chemical ionization, [M+H]⁺ often loses H₂O, NH₃, etc. — same arithmetic.
- On high-res GC-Orbitrap/Q-TOF systems you can compute constant-neutral-loss views in software.
LC-MS
- ESI + triple quadrupole is the classic home of the NL scan: 98 (or 49 for z = 2) for phosphopeptides, 80 for sulfates, 176 for glucuronides, 129 for GSH adducts, 44 for acids.
- Use it to screen a class out of a filthy matrix, then characterize hits with product-ion scans.
- Q-TOFs and Orbitraps can't link-scan; you acquire DDA/DIA and mine constant neutral losses in software (MS-DIAL, mzMine, Compound Discoverer, MassHunter…).
- Mostly qualitative — quantitation is done by MRM instead.
GC-EI explorer — the “other” neutral-loss spectrum
Hover the peaks of this stylized EI spectrum (molecular ion M⁺• = 158). Then flip the axis:
re-plotting every peak at M − m/z converts an ordinary spectrum into a neutral-loss spectrum —
the software meaning of the term you'll meet in data-processing tools.
6 · Check yourself
Score: 0/5
7 · Going deeper
Charge states: why Δm/z = Δmass ÷ z matters
Quadrupoles select by m/z, not mass. A neutral loss of 98 Da appears as an offset of 98 for z = 1, 49 for z = 2, 32.7 for z = 3. Phosphopeptide NL methods on QqQ instruments therefore run two linked scans (98 and 49) to catch singly and doubly charged precursors. In EI-GC-MS everything is z = 1, so the gap equals the mass directly.
Where NL scan sits among the four QqQ scan modes
| Mode | Q1 | q2 | Q3 | What you get |
|---|---|---|---|---|
| Product ion | fixed on M | CID | scans | full fragment fingerprint of one precursor |
| Precursor ion | scans | CID | fixed on f | all precursors that produce fragment f |
| Neutral loss | scans | CID | scans, locked at Q1 − Δm | all precursors that lose Δm — a structural class |
| MRM / SRM | fixed on M | CID | fixed on f | one specific transition — the quantitation workhorse |
Practical tuning tips (LC-QqQ)
- Collision energy is class-specific: too low → no fragmentation, no signal; too high → secondary losses (e.g. phosphopeptides losing H₃PO₄ and H₂O) and weaker signal. Optimize on a standard.
- Scan speed trades against duty cycle and sensitivity; don't outrun your chromatographic peak (≥ 10 points across it).
- Unit resolution → confirm hits by product-ion scan or accurate mass. Isobaric coincidences are the main artifact.
- Common paired strategy: NL survey scan → automatically trigger product-ion scans on detected peaks (information-dependent acquisition).
Common GC-EI neutral losses (read as M⁺• minus fragment)
| Gap (Da) | Lost | Suggests |
|---|---|---|
| 15 | •CH₃ | methyl branch; ubiquitous in TMS derivatives |
| 17 | •OH | alcohols, acids |
| 18 | H₂O | alcohols (often with M−15 companion) |
| 29 | •C₂H₅ / •CHO | ethyl chains / aldehydes |
| 31 | •OCH₃ | methyl esters, methoxy |
| 35/36 | •Cl / HCl | chlorinated compounds (check 3:1 isotope) |
| 43 | CH₃CO• / C₃H₇• | methyl ketones / propyl chains |
| 60 | CH₃COOH | acetate esters, alditol acetates (sugar analysis) |
| 90 | TMS-OH | silylated –OH groups (derivatized GC-MS) |