Matrix-Assisted Laser Desorption/Ionization (MALDI)

A pulse of ultraviolet light strikes a crystalline spot the size of a pinhead. A moment later, an intact 66,000-dalton protein floats into the gas phase. It is charged just once, undamaged, and ready to be weighed. That is the everyday miracle of matrix-assisted laser desorption/ionization. This technique turned mass spectrometry from a tool for small volatile molecules into something more. It became an instrument that can weigh antibodies, whole viruses, and synthetic polymers with molar masses exceeding 100,000 Da. The problem MALDI solves is brutal in its simplicity. Large, fragile, nonvolatile biomolecules cannot be boiled into vacuum without shredding them. Any ionization method energetic enough to desorb them tends to fragment them into uselessness. Proteins, oligonucleotides, and synthetic polymers decompose thermally long before they volatilize. Their many polar and ionizable groups make them essentially involatile as neutral species. Electron ionization and chemical ionization, the workhorses of small-molecule GC-MS, are simply inapplicable. MALDI, developed in the late 1980s, solved this problem differently. It embeds the fragile analyte inside a vast molar excess of a small, UV-absorbing organic acid — the matrix. This matrix absorbs the laser energy and sacrifices itself into a plume of ablated material. In doing so, it gently carries the analyte along and hands it a single proton. The big picture is a story of energy management. A pulsed laser (typically nitrogen, 337 nm, or frequency-tripled Nd:YAG, 355 nm) dumps roughly 10⁶–10⁷ W cm⁻² onto the sample. This happens for just a few nanoseconds. Bare protein cannot absorb this efficiently and would simply char. The matrix, present at a 1000- to 10,000-fold molar excess, absorbs essentially all of the photons. It undergoes rapid photoexcitation and decomposition, then expands explosively into vacuum as a dense plume. The analyte is isolated within the matrix lattice and thermally insulated by it. It is swept along in this expansion without absorbing damaging energy directly. Somewhere in the resulting plasma of ions and neutrals, it picks up (or loses) a proton. The method shows an overwhelming preference for singly-charged ions. This contrasts with the highly multiply-charged ions typical of electrospray. That preference is itself a defining, almost paradoxical feature of the method. It is precisely what makes MALDI the natural ionization partner for the time-of-flight mass analyzer. A single, well-defined charge state produces one clean, easily interpreted m/z peak per analyte mass. This article works through the physical mechanism of matrix-assisted desorption. It covers the chemistry of proton transfer that generates the charge, and the reasons the method favors z = +1. It also presents the equations that connect measured flight times and matrix chemistry to molecular mass. Three fully worked problems ground the theory in numbers a bench chemist would actually compute. These include a peptide mass calculation, an isotope-resolution argument for a monoclonal antibody, and a mixed-matrix optimization problem.

Key takeaways

MALDI embeds analyte in a large molar excess of a UV-absorbing organic acid matrix, which absorbs the laser pulse and co-desorbs the analyte intact.
Ionization proceeds mainly by proton transfer in the dense, transiently formed gas-phase plume, not by direct photoionization of the analyte.
The method overwhelmingly produces singly-charged ions [M+H]⁺ (or [M−H]⁻), unlike electrospray's characteristic charge-state ladders.
Common matrices — sinapinic acid, α-cyano-4-hydroxycinnamic acid (CHCA), 2,5-dihydroxybenzoic acid (DHB) — are chosen by analyte class and mass range.
MALDI is almost always coupled to a time-of-flight analyzer, since $m/z = 2eV t^2/L^2$ turns one flight-time measurement into one mass.
Soft, tolerant of salts and mixtures, and capable of masses beyond 500 kDa, MALDI is the standard tool for peptide mass fingerprinting, intact protein and polymer analysis, and imaging mass spectrometry.
Key idea: a sacrificial, UV-absorbing matrix crystal absorbs the laser energy and supplies protons, letting a fragile macromolecule desorb intact and singly charged into the gas phase.

Where this lives today

MALDI-TOF is now the standard front-line tool for peptide mass fingerprinting: a protein digested by trypsin yields a characteristic set of peptide masses that, matched against a genome-derived database, identifies the protein with high confidence from a single spectrum — the backbone of routine bottom-up proteomics workflows in research and increasingly in clinical laboratories. In clinical microbiology, MALDI Biotyper and comparable platforms identify bacterial and fungal species directly from a colony smear in minutes by matching whole-cell protein mass fingerprints against reference libraries, replacing days-long biochemical identification panels in many hospital microbiology labs. In materials science, MALDI-TOF measures the molar mass and mass-distribution (dispersity, $\mathit{Đ}$ ) of synthetic polymers such as polyethylene glycols and polystyrenes, giving end-group and repeat-unit information unavailable from bulk techniques like gel permeation chromatography. MALDI imaging mass spectrometry rasters a laser across a thin tissue section, matrix-coated in situ, to build spatial maps of lipid, peptide, or drug distribution directly on histological sections — a technique now used in pharmaceutical drug-distribution studies and in tumor-margin research. In glycobiology, DHB-matrix MALDI-TOF remains a primary method for profiling released N-glycans from therapeutic antibodies and other glycoproteins, informing quality control in biopharmaceutical manufacturing. And because MALDI so naturally produces well-behaved, singly charged high-mass ions, it remains the textbook partner for the time-of-flight mass analyzer, whose theoretically unbounded m/z range keeps pace with whatever mass MALDI can desorb.

:::quiz[A MALDI-TOF instrument is calibrated using $[\text{M+H}]^+$ standards. Which variable should be plotted linearly against flight time $t$ for an accurate mass calibration?]

$m/z$ directly
$\sqrt{m/z}$ ✓
$\ln(m/z)$
$(m/z)^2$

Ask the tutor: How does the finite ion-formation time and initial kinetic-energy spread of MALDI ions limit resolving power in linear versus reflectron time-of-flight instruments?

The physical picture

Consider the sample preparation first, because it dictates everything downstream. A dilute solution of analyte (say, 1–10 pmol µL⁻¹ of a protein) is mixed with a large excess of matrix solution — typically a saturated solution of a small aromatic acid in a mixture of water and organic solvent (acetonitrile or ethanol), often with a trace of trifluoroacetic acid to promote protonation. A microliter is spotted onto a metal target plate and allowed to air-dry. As the solvent evaporates, the matrix crystallizes, and — critically — the analyte molecules become physically trapped within or on the surface of the growing matrix crystals, isolated from one another at the molecular level. This co-crystallization is not a minor detail; poor co-crystallization (wrong matrix, wrong pH, wrong solvent ratio) is the single most common cause of a failed MALDI spectrum. When the pulsed laser strikes this crystalline deposit, the matrix — chosen specifically because its absorption maximum overlaps the laser wavelength — absorbs nearly all of the incident photons within a shallow surface layer (typically sub-micrometer penetration depth for these strongly absorbing solids). Absorbed photon energy is redistributed extremely rapidly among the vibrational and lattice modes of the matrix crystal, driving an explosive, localized phase transition: a "hot" plume of matrix molecules, matrix clusters, matrix ions, and entrained analyte molecules is ejected from the surface into vacuum on a nanosecond timescale. This is desorption, not simple evaporation — the process is far too fast and too energetic to be ordinary thermal vaporization, and it carries solid directly into a dense, transiently confined gas/plasma above the target. Ionization occurs within this expanding plume, principally by gas-phase proton-transfer reactions. Matrix molecules M can be photoionized or photoexcited to a state from which they lose or gain a proton; the resulting matrix cation MH⁺ or matrix radical cation M•⁺ then reacts, within the dense early plume, with neutral analyte molecules A:

\text{MH}^+ + \text{A} \rightarrow \text{M} + \text{AH}^+

This is favorable whenever the proton affinity of the analyte exceeds that of the matrix, which is generally true for basic sites on peptides and proteins (the guanidinium of arginine, the ε-amino of lysine, N-terminal amines). Because the plume is dense only for the first few tens to hundreds of nanoseconds before it expands and cools past the point where further ion–molecule collisions are likely, the charge state that is thermodynamically and kinetically accessible in that narrow window is essentially always z = +1: a large multiply protonated ion would require several independent, closely spaced proton-transfer events on a molecule that is simultaneously being physically separated from the matrix by supersonic expansion, and Coulomb repulsion between the added protons discourages the buildup of high charge on a species that only has a few nanoseconds of "wet" chemical contact with the matrix plasma. This stands in sharp contrast to electrospray ionization, where the analyte is protonated in solution over many available basic sites before droplet evaporation ever removes the last solvent shell, favoring extensive multiple charging.

Figure

Matrix crystal absorbs the laser pulse and co-desorbs the trapped analyte as a singly protonated ion

Matrix choice is not incidental chemistry — it is tuned to the analyte class. Sinapinic acid (3,5-dimethoxy-4-hydroxycinnamic acid) is preferred for intact proteins above roughly 10 kDa, where its slightly "harsher," more energetic desorption tolerates larger, more robust structures and gives good ion yield despite broader peaks. α-Cyano-4-hydroxycinnamic acid (CHCA) is the standard for peptides and small proteins (roughly 500–10,000 Da), giving sharper peaks and higher resolution because of gentler, more homogeneous crystallization. 2,5-Dihydroxybenzoic acid (DHB) is favored for oligosaccharides and glycopeptides, whose glycosidic bonds are labile and benefit from DHB's comparatively soft desorption. 3-Hydroxypicolinic acid (3-HPA), often mixed with picolinic acid, is the standard for oligonucleotides and DNA, whose phosphodiester backbone is prone to in-source fragmentation with the acidic cinnamic-acid matrices. Choosing the wrong matrix for the analyte class is the classic beginner error and typically produces either no signal or a spectrum dominated by matrix clusters.

Figure

Matrix choice by analyte class, showing UV absorption tuned to common desorption lasers

The core ideas

The measured quantity in every MALDI experiment coupled to time-of-flight is a flight time $t$ , converted to mass through the accelerating voltage and flight-tube geometry of the analyzer. An ion of charge $ze$ accelerated through potential $V$ acquires kinetic energy

zeV = \frac{1}{2}mv^2

where:

$z$ — number of elementary charges on the ion (almost always $z=1$ in MALDI),
$e$ — elementary charge, $1.602\times10^{-19}\ \text{C}$ ,
$V$ — accelerating potential (V),
$m$ — ion mass (kg),
$v$ — ion velocity after acceleration (m s⁻¹).

Because the ion then drifts a field-free distance $L$ at this velocity, $t = L/v$ , and substituting gives the working equation of time-of-flight mass spectrometry:

\frac{m}{z} = \frac{2eV t^2}{L^2}

where:

$m/z$ — mass-to-charge ratio (using $z$ as a pure number, $m$ in the same units implied by the calibration),
$t$ — measured flight time (s),
$L$ — field-free drift length (m),
$V$ — accelerating voltage (V).

Reading: flight time scales as $\sqrt{m/z}$ , so heavier ions of the same charge simply arrive later, and — since MALDI ions are almost all $z=1$ — the flight-time axis is, to excellent approximation, a direct proxy for molecular mass. This is precisely why MALDI and time-of-flight are such a natural pair: TOF analyzers have, in principle, unlimited mass range (there is no upper m/z cutoff imposed by a quadrupole's stability diagram or an ion trap's storage limit), which matches MALDI's ability to generate ions of extremely high mass, while MALDI's single dominant charge state means one time-of-flight peak corresponds unambiguously to one mass, with no need to "charge-deconvolve" a ladder of peaks the way electrospray spectra require. The second governing relationship concerns the proton-transfer chemistry itself. In the gas-phase plume, the favorability of transferring a proton from matrix (M) to analyte (A) is set by the difference in gas-phase proton affinities:

\Delta G_{\text{rxn}} = \text{PA}(\text{M}) - \text{PA}(\text{A})

where:

$\text{PA}(\text{X})$ — proton affinity of species X, defined as $-\Delta H$ for $\text{XH}^+ \rightarrow \text{X} + \text{H}^+$ (kJ mol⁻¹),
a negative $\Delta G_{\text{rxn}}$ (i.e. $\text{PA(A)} > \text{PA(M)}$ ) favors net proton transfer to the analyte, forming $[\text{A+H}]^+$ .

Reading: matrices are chosen partly because their proton affinities sit below those of the basic residues on peptides and proteins (arginine side chains, in particular, have very high proton affinity), ensuring that once matrix and analyte ions are thrown together in the dense plume, the thermodynamic driving force pushes the proton onto the analyte rather than leaving it on the matrix. This is why MALDI of peptides overwhelmingly gives $[\text{M+H}]^+$ rather than radical cations, even though the initial photoexcitation event on the matrix may well proceed through a radical or excited-state pathway.

Derivation AFrom accelerating voltage to molecular mass: the TOF mass equation

Goal. Show explicitly how a measured flight time $t$ , together with the known instrument geometry ( $V$ , $L$ ), yields the mass-to-charge ratio of a MALDI-generated ion, and confirm the $\sqrt{m/z}$ scaling that underlies every MALDI-TOF calibration.

Step 1 — energy imparted by the accelerating field. Ions formed near the target surface sit at a high potential $V$ relative to the grounded flight tube. As the ion of charge $ze$ falls through this potential difference, it converts electrical potential energy entirely into kinetic energy (ignoring the small initial thermal/kinetic spread from the desorption process itself):

zeV = \frac{1}{2}mv^2

Step 2 — solve for velocity. Rearranging,

v = \sqrt{\frac{2zeV}{m}}

This says that at fixed $V$ , lighter ions (small $m$ ) or more highly charged ions (large $z$ ) are accelerated to higher velocity — the physical root of TOF mass discrimination.

Step 3 — introduce the flight-time measurement. After acceleration, the ion travels a field-free drift region of length $L$ at this constant velocity, so the measured time of flight is

t = \frac{L}{v} = L\sqrt{\frac{m}{2zeV}}

Step 4 — invert to obtain $m/z$ . Squaring both sides and rearranging isolates the mass-to-charge ratio in terms of purely measured or instrument-fixed quantities:

t^2 = \frac{L^2 m}{2zeV}

\frac{m}{z} = \frac{2eVt^2}{L^2}

Step 5 — note the calibration in practice. Real instruments are calibrated with two or more standards of known mass to fix an effective $L$ and a small time offset $t_0$ (accounting for trigger delay and the finite acceleration region), giving a two-point linear calibration in $\sqrt{m/z}$ vs. $t$ rather than relying on nominal geometric $L$ and $V$ alone; this is standard practice precisely because stray field penetration and finite ion-formation volume make the idealized geometry only approximately correct.

Go deeper · Derivation B · 2 minWhy MALDI ions are overwhelmingly singly charged: a proton-transfer/plume-lifetime argument

Derivation BWhy MALDI ions are overwhelmingly singly charged: a proton-transfer/plume-lifetime argument

Goal. Give a quantitative feel for why MALDI plumes deliver essentially one proton per analyte molecule, using the competition between the number of available proton-transfer collisions and the density/lifetime of the reactive plume.

Step 1 — set up the plume as a transient, dense gas. Immediately after the laser pulse, the ablated plume has a very high initial number density of matrix neutrals and matrix ions — locally comparable to or somewhat below solid density before rapid adiabatic expansion — but this density falls essentially as the inverse cube of the expansion radius, $n(t) \propto n_0 (r_0/r(t))^3$ , and the plume becomes collisionless (mean free path exceeds plume size) within roughly $10^2$ – $10^3$ ns for typical MALDI expansion velocities of $5\times10^2$ – $10^3$ m s⁻¹ over a $\sim$ mm standoff distance to the extraction grid.

Step 2 — proton transfer requires an ion–molecule encounter within this window. The rate of a bimolecular ion–molecule reaction is $R = k[\text{MH}^+][\text{A}]$ , with a Langevin-type rate constant $k$ typically of order $10^{-9}\text{–}10^{-10}\ \text{cm}^3\ \text{molecule}^{-1}\ \text{s}^{-1}$ . Because the analyte is present at 1000- to 10,000-fold lower concentration than the matrix, and the whole reactive period lasts only nanoseconds to a few hundred nanoseconds while density falls precipitously, the expected number of successful proton-donation events experienced by a given analyte molecule before the plume becomes collisionless is a small number of order unity, not tens.

Step 3 — compare to electrospray for contrast. In electrospray, protonation of multiple basic sites occurs in the condensed phase, in a droplet that persists for microseconds to milliseconds while solvent evaporates and repeatedly re-equilibrates surface charge; every basic residue has ample time and ample proton supply (bulk solvent, not a rapidly rarefying vacuum plume) to become protonated, so proteins there commonly emerge from the +5 to +50 charge states depending on size. MALDI simply never grants the analyte that much residence time in a dense, wet, proton-rich environment.

Step 4 — Coulombic self-limitation reinforces the effect. Even where a second protonation is kinetically accessible early in the plume, the electrostatic penalty of placing a second positive charge on a molecule that has not yet had time to solvate or conformationally rearrange to spread the charges apart is substantial; for two point charges separated by a typical folded-protein dimension of order 3–5 nm, the Coulomb repulsion energy,

E_{\text{Coulomb}} = \frac{e^2}{4\pi\varepsilon_0 r}

is on the order of tens of kJ mol⁻¹ at these separations even in vacuum permittivity — large enough, in the still-condensing, low-dielectric plume environment, to strongly disfavor a second proton addition on the timescale available.

Step 5 — conclusion. The combination of (i) a short, rapidly rarefying reactive window, (ii) a large stoichiometric excess of matrix over analyte that nonetheless only grants each analyte molecule a handful of collision opportunities, and (iii) unfavorable Coulomb energetics for closely spaced multiple charges in a not-yet-solvated gas-phase ion, together explain why the dominant, often near-exclusive, MALDI charge state is $z=+1$ , occasionally accompanied by weak $z=+2$ or $z=+3$ satellite peaks for larger proteins.

Worked examples

Example 1 — Foundational: mass of a tryptic peptide from its flight time. A MALDI-TOF instrument operated in linear, positive-ion mode is calibrated with two standards: angiotensin II, monoisotopic $[\text{M+H}]^+ = 1046.54$ Da, flight time $t_1 = 24.981\ \mu\text{s}$ , and insulin (bovine), $[\text{M+H}]^+ = 5734.56$ Da, flight time $t_2 = 58.626\ \mu\text{s}$ . An unknown tryptic peptide peak is observed at $t_x = 31.407\ \mu\text{s}$ . Find its monoisotopic mass. Because $m/z \propto t^2$ (with $z=1$ throughout, since these are all singly protonated peptide ions), the two-point calibration is done in the linear variable $\sqrt{m/z}$ vs. $t$ :

\sqrt{m_1} = \sqrt{1046.54} = 32.351\ \text{Da}^{1/2}, \qquad \sqrt{m_2} = \sqrt{5734.56} = 75.727\ \text{Da}^{1/2}

A linear fit $\sqrt{m} = a\,t + b$ through these two points gives the slope

a = \frac{75.727 - 32.351}{58.626 - 24.981} = \frac{43.376}{33.645}\ \mu\text{s}^{-1}\ \text{Da}^{1/2} = 1.2892\ \mu\text{s}^{-1}\ \text{Da}^{1/2}

and the intercept

b = \sqrt{m_1} - a\,t_1 = 32.351 - (1.2892)(24.981) = 32.351 - 32.204 = 0.147\ \text{Da}^{1/2}

For the unknown at $t_x = 31.407\ \mu\text{s}$ :

\sqrt{m_x} = (1.2892)(31.407) + 0.147 = 40.484 + 0.147 = 40.631\ \text{Da}^{1/2}

m_x = (40.631)^2 = 1650.9\ \text{Da}

So the observed $[\text{M+H}]^+$ corresponds to a monoisotopic neutral mass of $1650.9 - 1.0 = 1649.9$ Da (subtracting the mass of the added proton, 1.0073 Da, to obtain the neutral peptide mass) — consistent, for instance, with a tryptic peptide of 14–15 residues. The key numerical habit worth internalizing: TOF calibration is always linear in $\sqrt{m/z}$ versus $t$ , never in $m/z$ versus $t$ directly, precisely because of the $t \propto \sqrt{m/z}$ relationship derived above.

Example 2 — Applied: can MALDI-TOF resolve the glycoforms of a monoclonal antibody? An IgG1 monoclonal antibody has an intact monoisotopic mass near 148,000 Da when carrying the common core-fucosylated biantennary glycan G0F, and a mass roughly 162 Da higher when carrying one additional galactose residue (glycoform G1F, since one galactose adds $\text{C}_6\text{H}_{10}\text{O}_5$ , monoisotopic mass 162.053 Da). A linear MALDI-TOF instrument, appropriate for such a large intact protein, typically delivers a mass resolving power $R = m/\Delta m$ of only about 500–1000 for species in this mass range (far lower than the tens of thousands achievable for small peptides in reflectron mode, because of the fundamentally broader initial kinetic-energy and metastable-decay distributions of very large ions). Using $R \approx 700$ as a representative value for a well-tuned linear instrument at 150 kDa, the peak width (FWHM) expected at this mass is

\Delta m = \frac{m}{R} = \frac{148{,}000}{700} = 211\ \text{Da}

Since the glycoform mass difference, 162 Da, is smaller than this expected FWHM of 211 Da, the two glycoforms G0F and G1F would not be baseline-resolved as separate peaks in standard linear MALDI-TOF of the intact antibody; instead they would appear as a single broadened, possibly slightly asymmetric envelope, or at best a poorly resolved shoulder. This is a genuine, practically important limitation: resolving antibody glycoforms by mass in practice requires either enzymatic release of the glycans (analyzed separately, as much smaller, better-resolved species) or a higher-resolution technique such as native electrospray with Orbitrap or FT-ICR detection, where the multiply charged ions compress the observed m/z scale and higher intrinsic resolving power is available. This calculation illustrates a limiting case worth stating explicitly: as the analyte mass grows while instrumental $R$ stays roughly fixed, the absolute mass window $\Delta m = m/R$ grows in direct proportion, so resolving small compositional differences becomes intrinsically harder for larger intact biomolecules — a genuine physical limitation, not merely a matter of poor tuning.

Example 3 — Exam-level: optimizing matrix-to-analyte ratio and estimating proton-transfer efficiency. A researcher prepares a MALDI sample by mixing $2.0\ \mu\text{L}$ of a $50\ \mu\text{mol L}^{-1}$ peptide solution with $18.0\ \mu\text{L}$ of a saturated CHCA matrix solution whose concentration is $8.0\times10^{-3}\ \text{mol L}^{-1}$ (a typical saturated CHCA concentration in 50:50 acetonitrile/0.1% TFA/water). Compute the matrix-to-analyte molar ratio in the mixed spotting solution, and comment on whether this falls within the generally recommended range of $10^3$ – $10^4$ -fold molar excess. First find moles of each component. For the peptide,

n_{\text{peptide}} = (50\times10^{-6}\ \text{mol L}^{-1})(2.0\times10^{-6}\ \text{L}) = 1.00\times10^{-10}\ \text{mol}

For the matrix,

n_{\text{matrix}} = (8.0\times10^{-3}\ \text{mol L}^{-1})(18.0\times10^{-6}\ \text{L}) = 1.44\times10^{-7}\ \text{mol}

The molar ratio of matrix to analyte is therefore

\frac{n_{\text{matrix}}}{n_{\text{peptide}}} = \frac{1.44\times10^{-7}\ \text{mol}}{1.00\times10^{-10}\ \text{mol}} = 1.44\times10^{3}

This ratio, about 1400-fold, sits at the low end of — but still within — the commonly cited working range of $10^3$ – $10^4$ -fold molar excess for good co-crystallization and efficient proton transfer; it would be reasonable to expect usable signal, though increasing the dilution of peptide relative to matrix (or reducing peptide concentration modestly) toward a ratio closer to $5\times10^3$ would typically improve spot homogeneity and shot-to-shot reproducibility, based on standard MALDI sample-preparation practice. Now suppose that of the $1.00\times10^{-10}$ mol of peptide deposited, mass spectral peak integration combined with an external calibration curve indicates that $6.0\times10^{-13}$ mol worth of ion signal (i.e., ions actually detected as $[\text{M+H}]^+$ , corrected for detector and transmission efficiency) was recorded from a single laser shot ablating an estimated 2% of the total spot material. The overall ionization efficiency per ablated analyte molecule is then

\text{ablated analyte} = 0.02 \times 1.00\times10^{-10}\ \text{mol} = 2.0\times10^{-12}\ \text{mol}

\text{efficiency} = \frac{6.0\times10^{-13}\ \text{mol detected}}{2.0\times10^{-12}\ \text{mol ablated}} = 0.30, \text{ i.e. } 30\%

A per-shot ionization efficiency of order tens of percent for the ablated fraction is broadly consistent with reported values for well-optimized MALDI spots (bearing in mind that overall analyte utilization across an entire sample spot, summed over the many laser shots typically fired during data acquisition, is much lower — often well under 1% — because most of the deposited material away from any given laser impact site is never ablated at all). This distinction between per-shot, per-ablated-material efficiency and whole-spot utilization efficiency is one that is frequently confused, and it is worth keeping straight when reasoning about detection limits.

Limiting cases, scope & common traps

Regime / question	Behavior	Why
Very small analytes (< 500 Da)	Matrix cluster ions and matrix fragments dominate the low-mass region, often obscuring or overlapping the analyte peak	Matrix itself has $M \sim 150$ – $220$ Da and forms abundant $[nM+H]^+$ clusters in the same mass range
Very large analytes (> 300 kDa, e.g. IgM, large protein complexes)	Signal becomes weak and broad; resolving power collapses	Fewer intact, well-desorbed ions per shot; larger metastable decay and kinetic-energy spread
Highly acidic analytes (e.g., oligonucleotides, phosphopeptides) with cinnamic-acid matrices	In-source fragmentation, loss of phosphate/backbone cleavage	Acidic matrix protons can catalyze gas-phase elimination reactions at labile linkages
Salt- or buffer-contaminated samples	Adduct peaks ( $[\text{M+Na}]^+$ , $[\text{M+K}]^+$ ) broaden or split the main peak	MALDI tolerates salts far better than electrospray, but high mM salt still competes for charge and disrupts co-crystallization
As matrix:analyte ratio → very low (< ~100-fold)	Poor or no signal; large crystals with analyte excluded from lattice	Insufficient matrix to isolate individual analyte molecules and absorb the laser pulse efficiently
As matrix:analyte ratio → very high (> ~10⁵-fold)	Signal suppression, matrix ion saturation of detector	Analyte becomes too dilute within crystal to be sampled by a typical laser shot footprint
$z$ in MALDI vs. electrospray	MALDI: essentially always $z=1$ (occasionally 2–3 for large proteins); ESI: broad charge-state ladder	Gas-phase, nanosecond plume chemistry (MALDI) vs. condensed-phase, microsecond-to-millisecond droplet chemistry (ESI)

Common trapCommon misconception: "the matrix just protects the sample from the laser like a sunscreen"

The matrix does far more than shield the analyte — it is the primary UV chromophore, the dominant proton donor/acceptor, and the medium whose explosive phase transition physically launches material into vacuum. Describing it merely as protective obscures the central chemistry: without the matrix's photoexcitation and subsequent proton-transfer chemistry, there would be no efficient ionization at all, even if the analyte somehow survived the pulse unfragmented. It is equally wrong to say the analyte is "photoionized directly" — for the great majority of biomolecular analytes, direct photoabsorption at 337–355 nm is negligible, and essentially all of the observed charge arises secondarily, via proton transfer from matrix-derived ions in the plume.

Common trapCommon misconception: "MALDI charge states behave like electrospray charge states, just with fewer charges"

It is tempting to think of MALDI's $z=1$ behavior as simply "electrospray with less charging," but the underlying mechanism is qualitatively different, not merely quantitatively suppressed. Electrospray charge states reflect equilibrium protonation of surface-accessible basic sites in a persisting, evaporating charged droplet; MALDI charge states reflect a kinetically frozen snapshot of whatever proton-transfer chemistry could occur during a few hundred nanoseconds of dense-plume existence. This is why, for example, adding a strong acid to an electrospray solution reliably shifts the charge-state envelope higher, while comparable manipulations of MALDI matrix acidity have far more muted and less predictable effects on the (already narrow) charge-state distribution.

The chemist behind it

MALDI's development is credited jointly to Franz Hillenkamp and Michael Karas, working at the University of Münster and later the Institute of Medical Physics and Biophysics, alongside foundational parallel work by Koichi Tanaka in Japan. Hillenkamp, a physicist by training who devoted much of his career to biophysical instrumentation, had spent years developing laser microprobe mass analysis for elemental and small-molecule analysis of biological tissue sections before turning his attention, with Karas, to the much harder problem of intact large-molecule desorption. Karas and Hillenkamp's key 1985–1987 experiments demonstrated that mixing a small, strongly UV-absorbing organic additive — initially amino acids and small aromatics, eventually the cinnamic- and benzoic-acid derivatives still used today — with a peptide or protein analyte, then irradiating the co-crystallized mixture with a pulsed UV laser, produced abundant, largely undamaged, singly charged molecular ions of species previously considered far too fragile and nonvolatile for any mass spectrometric approach. Tanaka, working independently at Shimadzu Corporation, achieved comparable large-protein desorption using a different matrix system (a suspension of ultrafine cobalt particles in glycerol) and shared the 2002 Nobel Prize in Chemistry for this contribution, alongside John Fenn for electrospray ionization — the two soft-ionization techniques that, together, opened biomolecular mass spectrometry to the routine analysis of proteins, a development the Nobel committee explicitly recognized as transformative for the life sciences.

Origins · 1 minFranz Hillenkamp and how this was discovered

Franz Hillenkamp

1936–2014 · German biophysicist, University of Münster; co-developer, with Michael Karas, of matrix-assisted laser desorption/ionization for intact biomolecule mass spectrometry

History

- 1985 — Karas and Hillenkamp report enhanced laser desorption ionization of amino acids and small peptides using an amino-acid additive matrix.

- 1987 — Karas and Hillenkamp demonstrate MALDI mass spectra of intact proteins beyond 10,000 Da, establishing the general matrix-assisted approach.

- 1988 — Tanaka, independently, reports desorption/ionization of proteins up to ~34,000 Da using an ultrafine-metal/glycerol matrix system.

- 1990s — Systematic development of sinapinic acid, CHCA, DHB, and 3-HPA matrices for distinct analyte classes; MALDI-TOF becomes commercially widespread.

- 2002 — Tanaka and Fenn share the Nobel Prize in Chemistry "for the development of soft desorption ionisation methods for mass spectrometric analyses of biological macromolecules."

Why can't proteins simply be ionized the way small volatile molecules are, without a matrix?

Electron ionization and chemical ionization both require the analyte to already exist as a free gas-phase neutral before ionization occurs, which in turn requires heating the sample enough to vaporize it. Proteins and other large biomolecules decompose — losing water, cleaving peptide bonds, denaturing irreversibly — at temperatures far below what would be needed to generate any appreciable vapor pressure, because their intermolecular and intramolecular interactions (hydrogen bonding, ionic bridges, extensive conformational entropy loss on unfolding) make purely thermal vaporization thermodynamically and kinetically impossible before decomposition wins. MALDI sidesteps the need for thermal vaporization entirely: the matrix absorbs the destabilizing energy, undergoes an extremely fast, localized, essentially explosive phase change, and physically carries the still-folded (or nearly so) analyte along for the ride, without ever requiring the analyte itself to boil.

Why is sinapinic acid preferred over CHCA for very large intact proteins, if CHCA gives sharper peaks?

CHCA's gentler desorption and more homogeneous crystal formation give excellent resolution for peptides and small proteins, but this same "softness" becomes a limitation for very large, more thermally and mechanically robust protein assemblies, which may not desorb efficiently or intact under CHCA's milder plume conditions. Sinapinic acid's structure (with two additional methoxy substituents relative to simpler cinnamic acids) changes its photophysics and crystallization behavior in ways that empirically favor efficient desorption of larger, bulkier analytes, even though the resulting peaks are broader — a trade consciously made because getting any usable ion signal from a 60 kDa protein matters more than shaving resolution that was never going to be very high at that mass anyway (see the antibody glycoform example, where $\Delta m = m/R$ already dominates over subtle mass differences at these sizes).

Does MALDI ever produce multiply charged ions, and if so, when?

Yes, particularly for larger proteins (roughly above 30–50 kDa), where weak $[\text{M+2H}]^{2+}$ and occasionally $[\text{M+3H}]^{3+}$ satellite peaks are commonly observed alongside the dominant singly charged peak. These arise because larger proteins present more basic sites and a larger surface over which two protons can be accommodated with somewhat reduced mutual Coulomb penalty (greater achievable charge separation), and because larger analytes may spend marginally longer in a still-reactive plume environment due to their larger physical size and different desorption dynamics. Even so, the singly charged peak remains dominant in essentially all standard MALDI spectra, which is the qualitative signature that distinguishes MALDI spectra from electrospray spectra of the same protein at a glance.

How does the choice of laser wavelength relate to matrix selection?

Nearly all standard MALDI instruments use either a nitrogen laser (337 nm) or a frequency-tripled Nd:YAG laser (355 nm), and matrices are selected specifically because their solid-state UV absorption spectra have strong absorption bands overlapping one of these two wavelengths — this overlap is what allows the matrix to intercept essentially all incident photons in a thin surface layer rather than transmitting them uselessly through the crystal. Infrared MALDI variants exist (typically using an Er:YAG laser near 2.94 μm, matched to O–H stretching absorption in matrices like glycerol, urea, or succinic acid), and are sometimes used for very labile analytes such as intact oligonucleotides or noncovalent complexes, because IR-MALDI's gentler, more distributed absorption produces less analyte fragmentation, at some cost in ionization efficiency and typical mass resolution compared to UV-MALDI.

Why is a two-point (or multi-point) external calibration in

\sqrt{m/z}

necessary rather than trusting the nominal

V

and

L

values directly?

The idealized TOF equation, $m/z = 2eVt^2/L^2$ , assumes instantaneous ion formation exactly at the accelerating grid and perfectly uniform, ideal field geometry, but real instruments have a finite ion-formation region (ions are desorbed from a small area of the target surface, not a mathematical point), stray field penetration near apertures and grids, and small electronic delays between the laser trigger and the start of timing. All of these introduce a small, roughly constant time offset and a small effective correction to $L$ , both of which are absorbed automatically into an empirical linear fit of $\sqrt{m/z}$ against measured $t$ using standards of accurately known mass, as in the worked peptide-mass example. Skipping calibration and using nominal geometric values typically gives mass errors of many daltons, whereas a proper external (or, better, internal) calibration routinely achieves sub-dalton to few-ppm accuracy for well-resolved peaks.

Is MALDI compatible with mixtures, or does it require a pure, single-component sample?

MALDI tolerates mixtures reasonably well compared to many other ionization methods, and peptide mass fingerprinting explicitly relies on this — a tryptic digest is, by design, a mixture of dozens of peptides analyzed together in a single spectrum. However, MALDI is subject to strong ion-suppression effects: highly surface-active or more easily ionized components in a mixture can dominate the crystal surface and the plume chemistry, suppressing the signal of co-present but less readily ionized species, particularly when concentrations differ by orders of magnitude. This is why sample cleanup (desalting, removal of detergents and very high-abundance contaminating proteins) before spotting is standard practice, even though MALDI is comparatively forgiving of moderate salt contamination relative to electrospray.

Why is MALDI almost always paired with time-of-flight rather than a quadrupole or ion trap analyzer?

Time-of-flight analyzers have, in principle, no upper mass limit — every ion, however massive, simply arrives later, limited only by detector bandwidth and acquisition timing rather than any stability-diagram cutoff — which matches MALDI's ability to desorb intact ions across an enormous mass range, from a few hundred daltons to hundreds of kilodaltons. MALDI is also inherently a pulsed technique (each laser shot produces one packet of ions), which maps naturally onto TOF's requirement for a well-defined ion packet start time; by contrast, quadrupoles and ion traps are best suited to continuous or nearly continuous ion beams (as electrospray naturally provides) and have practical m/z ceilings set by rf stability limits that are far below what MALDI-generated intact-protein ions require.