Skip to main content

Ultrafast lasers: why femtosecond pulses change the physics of material removal

Two identical holes in polyimide, one clean and one melted, at the same pulse energy. The difference is pulse duration. What happens when energy arrives faster than heat can move, and which applications that opens up.

Put two holes side by side in a sheet of polyimide, made with the same pulse energy and the same focused spot. One has a clean vertical wall and no discolouration. The other has a raised melted lip, a brown halo and stringy debris. The only variable changed was pulse duration. That single comparison is why ultrafast laser processing exists as a separate field rather than as a faster version of the pulsed lasers already on the shop floor.

Energy in before heat can get out

Laser light is absorbed by electrons almost instantly. Those hot electrons then pass their energy to the atomic lattice through electron and phonon coupling, and that transfer takes a few picoseconds in most metals, longer in dielectrics. Everything follows from the relationship between the pulse duration and that coupling time.

With a nanosecond pulse, the lattice heats up while the pulse is still arriving. A melt pool forms, material is expelled partly as liquid, and the surrounding solid conducts heat away for the rest of the pulse and beyond. The recast layer, the dross, the burr and the oxide band are all consequences of that. With a pulse of a few hundred femtoseconds, the pulse has ended before the lattice has warmed. The energy is already sitting in the electrons of a very small volume, which then leaves as vapour and plasma, carrying its heat away with it.

The scaling argument is simple enough to write in one line. Thermal diffusion length grows with the square root of the interaction time, so shortening the pulse by six orders of magnitude reduces the depth of thermally affected material by roughly three. That is the whole case for ultrafast processing, and it is a physical argument rather than a marketing one.

What cold ablation does and does not mean

The phrase is convenient shorthand and slightly misleading. Nothing about the process is cold. A very hot, very small plasma is created. The useful property is that the heat leaves with the removed material instead of conducting into the part that remains.

This matters because heat accumulation is real and catches people out. Each pulse leaves a small residue of heat behind. At high repetition rates with a slow scan, the next pulse arrives before that residue has dissipated, and after a few thousand pulses you have a melt pool again. Repetition rate, scan speed, hatch spacing and the order in which the geometry is traversed all become part of the thermal design. Stated bluntly: buying an ultrafast source does not buy you a cold process, it buys you the possibility of one, and the process development that follows is where the real work sits.

Nonlinear absorption, and machining things that are transparent

Compressing a modest pulse energy into a few hundred femtoseconds produces enormous peak intensity at the focus. At those intensities, materials that are perfectly transparent at the laser wavelength start absorbing through multiphoton and avalanche processes. That absorption happens only where the intensity exceeds a threshold, which is inside the focal volume and nowhere else. The beam passes through the surface and the bulk without effect and deposits its energy at a point you choose in three dimensions.

A sharp threshold gives features smaller than the spot

For ultrashort pulses the ablation threshold is unusually well defined and reproducible, because the removal mechanism does not depend on messy heat flow. With a Gaussian beam profile, only the central part of the spot exceeds threshold when you work just above it. That means you can produce features smaller than the focused spot diameter, at the cost of removal rate and with tight demands on pulse-to-pulse energy stability. Together with the near absence of recast and burr, this is why the post-processing chain shortens: fewer deburring, polishing and cleaning steps, and in some medical device flows one fewer qualification headache.

Where the extra cost is justified, and where it is not

ApplicationIs a nanosecond source enough?What ultrafast adds
General metal marking and engravingYes, comfortablyNothing that justifies the price difference
Sheet and plate cuttingYesNothing
Selective removal of one thin film without damaging the layer beneathSometimes, with careA genuine process window instead of a lucky setting
Polymer and biodegradable medical device cuttingMarginalNo melt, no rolled edge, no thermal degradation of the polymer chain
Stent struts and fine medical device edgesYes, with heavy post-processingFar less recast, so less electropolishing and less risk
Glass, sapphire, display glass, semiconductor dicingNoAccess to the material at all, plus internal modification
Waveguide writing and integrated photonicsNoThe only practical route
Ophthalmic and other tissue proceduresNoCutting with negligible thermal damage to adjacent tissue
Nanoparticle generation by ablation in liquidPartlyCleaner, surfactant-free particles with better size control

The honest summary for an industrial buyer in Bangladesh: if the work is marking, engraving and cutting ordinary metals, a nanosecond system does the job and the money saved should go into the motion system, the extraction and the training. Ultrafast earns its price on transparent and brittle materials, on polymers, on layered structures and on anything where the customer specifies a heat affected zone limit that a thermal process cannot meet.

Research programmes a university can build around one

For a university group, the calculation is different, because the instrument is a platform rather than a production tool. A single ultrafast source with a good optical bench and a decent stage supports several independent lines of work at once.

What owning one actually demands

Questions worth asking before the purchase order

  1. What is the pulse duration at the workpiece, after the full delivery path, and how is that verified at commissioning?
  2. Can repetition rate be varied independently of pulse energy, and over what range? This determines whether you can control heat accumulation.
  3. What burst or pulse-shaping modes are available, and are they included or licensed separately?
  4. What is the expected service life of the pump diodes, what does replacement cost, and who fits them?
  5. What does the system need in cooling, purge gas and electrical supply, stated as a site requirement document we can hand to the works department?
  6. Will you process a sample of our material before award, and may we inspect the result ourselves?

Back to all Insights