NIR Laser Wavelengths for Deep-Tissue Imaging: 730 to 1550nm

Visible light does not travel far in tissue. Send 532 nm into skin and it is scattered and absorbed within a millimetre or two. Move into the near-infrared and the same tissue becomes translucent enough to image centimetres deep. This is why almost every technique that looks into a living subject — rather than at a thin slice of it on a slide — operates somewhere between 700 and 1400 nm.

But the rule is not simply “longer is deeper.” The NIR region contains water absorption bands that cut penetration sharply at specific wavelengths, and one of them sits directly on a wavelength many buyers assume is a safe deep-tissue choice. This guide maps the region properly and shows which of our twelve NIR wavelengths belongs where.

The Optical Window, and What Closes It

The usable NIR range is bounded at both ends by different absorbers, and understanding both boundaries explains the whole region.

Below about 650 nm, haemoglobin dominates. Both oxy- and deoxyhaemoglobin absorb strongly through the visible, which is what makes blood red and tissue opaque. That absorption falls steeply through the deep red, and by 700 nm it has dropped by more than an order of magnitude.

Above about 1350 nm, water dominates. Tissue is mostly water, and water absorption climbs steeply through the short-wave infrared, with a strong band near 1450 nm and a very strong one near 1900 nm.

Between those two limits lies the optical window — roughly 650 to 1350 nm — where tissue is at its most transparent. Everything in NIR biomedical optics happens inside it.

Why Longer Wavelengths Reach Deeper

Within the window, absorption is low, so what limits you is scattering. Photons are not absorbed so much as randomised — deflected repeatedly until all spatial information is lost.

Scattering in tissue falls off with wavelength, roughly as λ⁻¹ to λ⁻¹·⁵ depending on tissue type. Longer wavelengths are deflected less per unit distance, so they travel further before losing directionality. That produces two benefits at once: greater penetration depth, and sharper images at a given depth, because less of the light arriving at your detector has taken a scrambled path.

This is the entire argument for pushing to longer wavelengths, and it is why the field has moved steadily from 800 nm toward 1300 nm over the last two decades.

The Water Bands — and Why 980 nm Is Not What You Think

Here is the part that catches people out, because the scattering argument above suggests a smooth improvement with wavelength and the reality is bumpy.

Water absorption does not rise monotonically. It has local peaks inside the window, and the practically important one sits at around 980 nm. Water absorbs noticeably more at 980 nm than at either 940 nm or 1064 nm.

The consequence is counter-intuitive: 980 nm penetrates tissue less well than 1064 nm, despite being the shorter wavelength. A buyer reasoning purely from “longer is deeper” will pick 980 nm over 940 nm expecting more depth and get less.

That same absorption makes 980 nm genuinely useful for the opposite purpose. Light absorbed by water becomes heat, so 980 nm deposits energy efficiently in wet tissue — which is precisely what therapeutic and surgical applications want. Our 980 nm line reaching 2500 mW is aimed at that use, not at imaging depth.

The same logic sets the far edge of the window. 1550 nm sits closer to the 1450 nm water band than 1310 nm does, so it penetrates wet tissue less well despite being longer. Two other minor absorbers are worth knowing if your target is fatty tissue: lipids absorb around 1210 nm and again near 1720 nm.

NIR-I and NIR-II

The literature splits the region into two windows, and the terms appear constantly in instrument specifications.

The detector column is the practical catch. Silicon sensitivity ends near 1100 nm, so everything in NIR-II requires InGaAs detection — substantially more expensive than the silicon cameras and photodiodes that serve NIR-I. Moving from 808 nm to 1310 nm is not only a change of laser; it is a change of detector, filters, and often budget.

Mapping the Wavelengths

The three wavelengths that carry the most weight

808 nm is the haemoglobin isosbestic point — the wavelength where oxy- and deoxyhaemoglobin absorb equally. That makes it the reference channel for tissue oximetry and photoacoustic imaging: measured signal reflects total blood volume independent of oxygenation state, so pairing 808 nm with a second wavelength where the two forms differ lets you separate volume from saturation. No other wavelength in the region does this.

785 nm and 808 nm both excite indocyanine green. ICG is one of very few NIR fluorophores approved for human use, and it anchors a great deal of surgical guidance, lymphatic mapping, and perfusion imaging work. Its excitation sits around 780–805 nm with emission near 820–830 nm, so both of these lines drive it well.

1310 nm is the depth champion. Scattering is low, and it sits in the gap between the 1200 nm and 1450 nm water features. For imaging as deep as scattering tissue allows, this is the wavelength — which is why our 1310 nm line lists medical diagnostics among its applications.

Why Multimode Is the Right Choice Here

Ten of our twelve NIR wavelengths are multimode, and for tissue work that is a feature rather than a compromise.

Beam quality buys you a tight focus. But tissue scatters light within a millimetre or so regardless of how well collimated it arrived — the medium destroys your careful beam profile almost immediately. For diffuse illumination of a volume, what matters is how many photons you can get in, not how precisely they were aimed.

So the 2500 mW multimode module is genuinely the better instrument here than a 30 mW diffraction-limited one, even though the latter costs more. This is the clearest practical case of the argument in understanding M² and beam quality: M² is a focusability spec, and when you are not focusing, paying for it is waste.

The exceptions are the two wavelengths whose applications are not tissue illumination at all — 1030 nm and 1064 nm, which are TEM₀₀ because they serve spectroscopy, interferometry, and metrology.

Choosing a Wavelength

1. Start from your label, if you have one. ICG points at 785 or 808 nm. Other NIR dyes have their own peaks — match the excitation maximum first, since nothing else recovers lost excitation efficiency.
2. Label-free and want maximum depth? Go as long as your detector budget allows. 1310 nm is the best of the region; 1064 nm is strong and works with cheaper silicon detection below 1100 nm.
3. Check your detector before your laser. Anything past about 1100 nm requires InGaAs. This constraint often decides the wavelength on its own.
4. Avoid 980 nm for imaging depth — and choose it deliberately when you want tissue heating.
5. Measuring oxygenation? Anchor at 808 nm and pair it with a wavelength where oxy- and deoxyhaemoglobin diverge.
6. Specify power at the tissue, not at the aperture. Losses through optics, fibre, and the first millimetre of tissue are substantial, which is why the high-power multimode variants exist.

Aimpico NIR Lasers

Our NIR range spans twelve wavelengths from 730 to 1550 nm. Standard (DMCE) and low-noise (DMCZ) multimode diode variants reach 2500 mW at 785, 808, and 980 nm. The 1064 nm line is DPSS rather than diode — TEM₀₀ throughout and reaching 3000 mW, the highest power we offer in the region. Every family specifies power stability of < 1% RMS, with low-noise variants adding a guaranteed amplitude noise figure across 20 Hz – 20 MHz — which matters for quantitative work, as covered in what “< 1% RMS” really means.

All twelve families carry TTL and analog modulation from DC to 30 kHz, suitable for gating and lock-in detection but not for nanosecond timing. The one exception is the narrow linewidth DMCN-785, which is CW only — see TTL vs. analog modulation for where that boundary falls. For Raman specifically, 785 nm has its own guide on why 785 nm is the standard.

These are laser sources intended for research and OEM integration; any clinical instrument built around them carries its own regulatory path at the system level.

Tell us your target depth, your label if you have one, and what detector you are working with, and we will tell you which wavelength fits — including when a shorter one serves you better than a longer one.