Understanding M² and Beam Quality: TEM₀₀ vs. Multimode

M² is the most misread number on a laser datasheet. It looks like a quality grade — a smaller number is better, so surely you want the smallest one you can afford. Buyers routinely specify M² < 1.1 the way they would specify a tolerance, and pay for it in the only currency that matters: output power.

M² is not a quality grade. It is a focusability spec. It predicts one thing with precision — how small a spot your beam can be focused to, and how fast it spreads afterwards. For applications that focus tightly or couple into single-mode fibre, it is the most important number on the page. For applications that do neither, it is close to irrelevant, and chasing it costs you power you could have had.

This guide explains what M² measures, what it predicts, why our multimode lasers do not quote it at all, and how to tell which camp your application falls into.

What M² Actually Measures

Diffraction sets a hard floor on how tightly light can be focused. A perfect Gaussian beam — the fundamental TEM₀₀ mode — sits exactly at that floor. It is the best any beam of a given wavelength can do.

M² measures how far your beam is from that floor. Formally it is the ratio of your beam's parameter product (waist radius × divergence half-angle) to the diffraction-limited value of λ/π:

M² = (w₀ · θ) / (λ / π)

A perfect Gaussian gives exactly 1.0. Real beams give more. M² < 1.1 means the beam is within 10% of the physical limit; M² < 1.5 means it is 50% off it. Values below 1.0 are physically impossible, so any datasheet quoting one is describing a measurement error rather than a laser.

Note that M² says nothing about power, wavelength stability, noise, or lifetime. A laser can have superb M² and be useless for your experiment on every other axis.

What M² Predicts

Three consequences follow from the number, and they are the entire practical content of the spec.

1. Focused spot size scales linearly with M²

Focus a beam of diameter D with a lens of focal length f, and the spot diameter you get is approximately:

d = 4λf M² / (π D)

Everything except M² is fixed by your optics and wavelength. So a beam at M² < 1.5 focuses to a spot 50% wider than one at M² < 1.0 through the same lens — and because area goes as the square of diameter, it spreads the same power over 2.25× the area.

2. Divergence scales with M²

For a given waist size, a higher-M² beam spreads faster. Over a long free-space path, or through a long optical train, that shows up as a larger beam at the far end and more light clipped by apertures along the way.

3. Brightness scales as 1 / (M²)²

This is the one that reframes the buying decision. What determines how much light you can deliver into a small spot is not power alone but power divided by the square of M².

The consequence is counter-intuitive and worth sitting with: a 1500 mW multimode laser can deliver less light to a diffraction-limited focus than a 30 mW TEM₀₀ one. Broad-area multimode diodes commonly run M² in the tens along their slow axis. Square that, divide it into the power, and the high-power laser loses — often by an order of magnitude. Raw wattage on a datasheet is not the same as usable intensity at your sample.

TEM₀₀, Near-TEM₀₀, and Multimode

The mode label on a datasheet describes the transverse structure of the beam, and it maps onto M² ranges directly.

Across our catalogue the split is 28 TEM₀₀ variants, 16 Near-TEM₀₀, and 34 multimode — and the reason multimode outnumbers everything else is simply that most applications do not need a diffraction-limited beam and would rather have the power.

Why Our Multimode Lasers Quote No M² At All

Look through our specification tables and you will notice something consistent: every one of our 28 TEM₀₀ variants carries an M² figure, and not one of our 34 multimode variants does. That is deliberate, and the reason is worth understanding — because it also tells you how to read anyone else's datasheet.

A multimode laser diode is a broad-area emitter. Its output aperture is a wide, thin stripe, and the beam it produces is astigmatic: it behaves completely differently along its two axes. Perpendicular to the junction — the fast axis — emission comes from a region only a micron or two across and is nearly diffraction-limited. Parallel to the junction — the slow axis — emission comes from a stripe tens or hundreds of microns wide and supports many transverse modes.

The result is a beam with two wildly different M² values depending on which way you measure. Quoting a single number would require either averaging them, which describes no real axis of the beam, or quoting the better one, which flatters the laser. Neither tells you what your optics will do.

So instead we specify what is actually useful for a multimode beam: its physical dimensions. A figure such as ~ 5 × 8 mm at aperture (1/e²) tells you the beam is rectangular, tells you both axes, and lets you size your optics honestly. The TEM₀₀ variants, whose beams are symmetric and near-Gaussian, get both a single meaningful M² and a round diameter — typically ~ 1 mm or below.

This is the same principle as stating the bandwidth alongside a noise figure, which we cover in what “< 1% RMS” really means. A number quoted without the context that makes it meaningful is worse than no number, because it invites a comparison that does not hold. If a vendor quotes a single tidy M² for a high-power multimode diode, ask which axis it refers to.

The Trade-Off You Are Actually Making

Beam quality and output power pull against each other, and our own catalogue shows the cost plainly. Within a single wavelength:

The 488 nm case is the clearest: choosing M² < 1.1 takes you from 1500 mW to 30 mW at the same wavelength. That is not a premium tier being withheld — it is the physics of extracting a single transverse mode from a semiconductor. We discuss where that trade lands for blue in 488 nm vs. 473 nm.

DPSS lasers escape much of this. Because the beam is defined by a resonator cavity rather than a semiconductor stripe, TEM₀₀ comes naturally and at useful power: our 532 nm line holds M² < 1.2 across all three tiers up to 1500 mW. That structural advantage is a large part of why DPSS still exists, and it is covered in diode laser vs. DPSS laser.

There is one notable exception in our diode range. The 1030 nm family is TEM₀₀ in both variants — DLCE-1030 and DLCZ-1030, M² < 1.1 with a ~ 1.0 mm beam, up to 500 mW. No multimode option exists at that wavelength because the line is built for spectroscopy, telecommunication, and precision measurement, where a multimode version would have no buyers.

Which Applications Actually Need Low M²

The single-mode fibre cliff

One case deserves separating from the rest, because it is not a gradual trade-off but a threshold. A single-mode fibre supports exactly one spatial mode. Light that does not match that mode does not couple in at reduced efficiency — it does not couple in at all, and ends up as heat at the connector.

Coupling efficiency therefore collapses as M² rises above about 1.1. If your system requires single-mode fibre delivery, the beam quality specification is not a preference to be balanced against power. It is a gate: below the threshold the laser works, above it the laser does not, and no amount of extra output compensates.

Reading the Number in Context

Two habits will keep you out of trouble.

Check the tier, not just the family. M² is a per-variant specification, not a property of a wavelength. Our 1064 nm standard and low-noise models are TEM₀₀ at M² < 1.5, while the SLCP-1064 Pro reaches M² < 1.1. Our 561 nm line runs M² < 1.5 in its first two tiers and M² < 1.2 in Pro. Specifying “1064 nm, TEM₀₀” does not pin down the beam you will receive.

Do not assume a mode label implies a number. Near-TEM₀₀ covers a real range. Our 633 nm standard, low-noise, and narrow-linewidth variants are all Near-TEM₀₀ at M² < 1.5, while the DLCZ-633 Pro is fully TEM₀₀ at M² < 1.1 — a substantial difference within one wavelength and one mode family.

How to Decide

1. Are you coupling into single-mode fibre? M² < 1.1. This is a gate, not a preference — decide it first and let the rest follow.
2. Are you focusing to a diffraction-limited spot — confocal, trapping, Raman, precision machining? TEM₀₀ at M² < 1.2. Remember that brightness goes as 1/(M²)², so beam quality buys you more here than extra watts do.
3. Are you illuminating an area, pumping, or coupling into multimode fibre? Buy multimode and spend the budget on power. A tight beam delivers nothing extra.
4. Working out the power you need? Compute it at the sample, after focusing losses — not at the aperture. This is where a low-M² beam quietly wins, because more of its output survives the optical train.
5. Comparing vendors? For any multimode laser, ask which axis a quoted M² describes. For any laser, confirm the figure applies to the specific variant and power level you are buying.

Aimpico Beam Quality Options

Our catalogue spans three beam-quality tiers. Multimode variants (DMCE / DMCZ) deliver maximum power — up to 2500 mW at 785, 808, and 980 nm — with beam dimensions specified rather than a misleading single M². Near-TEM₀₀ variants sit at M² < 1.5. TEM₀₀ variants reach M² < 1.2 or < 1.1, with beam diameters at or below 1 mm.

The DPSS lines carry TEM₀₀ throughout: 473 nm and 532 nm hold M² < 1.2 across every tier. Among the diodes, Pro variants (DLCZ) provide M² < 1.1 at 405, 488, 520, 633, 635, 660, 785, and 808 nm, and the 1030 nm line is TEM₀₀ in both variants to 500 mW. A full breakdown by technology is on our diode and DPSS pages.

Tell us your focusing optics, your target spot size, and whether fibre is involved, and we will tell you the minimum beam quality that does the job — so you can put the rest of the budget into power.