Micro Laser Drilling: Tolerances, Materials and Use Cases
This guide is written for engineers deciding whether laser is the right process for a specific part, covering achievable tolerances, how different material classes behave, where laser beats mechanical and EDM, and where it does not.
How Micro Laser Drilling Works
Laser drilling relies on controlled ablation. A focused beam deposits energy into a small volume of material faster than that energy can conduct away, vaporizing the material rather than cutting it mechanically. No tool touches the part.
That single fact drives most of the process advantages. There is no tool to wear, deflect or snap, so hole 1 and hole 100,000 are dimensionally identical. There is no lateral cutting force, so thin films, unfired ceramic and brittle substrates survive the operation. And because the beam can be steered and shaped, hole geometry is not limited to what a round tool can produce: square, rectangular, and other non-round profiles are achievable.
Four techniques cover most production work:
- Percussion drilling. The beam stays fixed and fires repeated pulses until the hole is through. Fastest method, best suited to smaller diameters.
- Trepanning. The beam traces a circular path, cutting a cylindrical core out of the substrate. Slower, but gives far better control over taper and exit quality on larger diameters.
- Direct-write. The focused beam is positioned by a scanner or stage and writes each hole individually. Maximum flexibility for mixed geometries and low-to-mid volumes.
- Mask projection. The beam is shaped and imaged through a mask, producing many holes simultaneously. Higher throughput for repeated patterns, particularly in polymers.
As Sergey Broude documented in Laser Focus World, the choice between direct-write and mask projection is as consequential as the wavelength selection itself, especially in polymer medical devices where edge quality and repeatability are the governing specs.
Why Micro Laser Drilling Outperforms Mechanical Drilling
The case for laser over a physical tool is strongest at the small end, and it compounds as diameter drops.
Minimum diameter. Mechanical micro drills become impractical below roughly 100 microns, because the tool is too fragile to survive the cutting load. Laser processes reach 5 microns routinely and are not bounded by tool geometry.
Tool wear and dimensional drift. A mechanical drill changes dimension as it wears, so the first hole and the thousandth hole differ. A laser has no contact element. Across a 1,000-hole array, every hole is produced by the same beam conditions.
Part stress. Drilling forces load the workpiece. On thin films, unfired ceramic, sapphire, or thin-wall tubing, that loading causes cracking, delamination, or deflection. Ablation applies no mechanical force.
Burr and secondary operations. Mechanical drilling produces burrs that need deburring. Laser ablation removes material as vapor and ejecta, which generally eliminates the deburring step, a real cost factor on high-count parts.
Geometry freedom. Non-round holes, blind holes to a controlled depth, dense arrays at tight pitch, and holes drilled at an angle to the surface are all accessible to a steered beam and not to a rotating tool.
Micro Laser Drilling Compared to Other Micro-Hole Processes
| Micro laser drilling | Mechanical micro drilling | EDM | Punching | |
|---|---|---|---|---|
| Practical minimum diameter | ~5 µm | ~100 µm | ~50 µm | ~200 µm |
| Material restriction | Metals, ceramics, glass, polymers, silicon, diamond | Machinable solids | Conductive only | Ductile sheet |
| Contact force on part | None | High | None | Very high |
| Tool wear | None | Significant | Electrode wear | Die wear |
| Burr | Minimal | Yes | Minimal | Yes |
| Brittle material capability | Strong | Poor | Not applicable | Poor |
| Setup for a new geometry | Programming change | New tooling | New electrode | New die |
| Economics | Prototype through production | Mid volume | Low volume | Very high |
The pattern is consistent. Laser wins on small diameters, brittle and non-conductive materials, geometry flexibility, and short-run changeover. Punching wins on very high volume commodity parts in ductile sheet. EDM remains relevant for deep features in conductive metals where aspect ratio outweighs cycle time.
Materials and How They Behave
Material class determines wavelength, pulse width, and achievable quality more than any other input. Laserod drills stainless steel, aluminum, copper, glass, sapphire, ceramic, silicon, diamond, Kapton, plastics, carbon fiber and composites, and the behavior differs meaningfully across them.
Metals: stainless steel, aluminum, copper. Readily absorbing at common wavelengths and well characterized. The engineering concern is the heat-affected zone and recast layer, molten material that resolidifies on the hole wall. Shorter pulse widths suppress both. Copper and aluminum are more infrared-reflective and often process better at shorter wavelengths. Representative work includes 10 µm diameter holes through 500 µm stainless across a 1,000-hole array, and 11 µm holes at 100 µm pitch in 75 µm stainless.
Ceramics, fired and unfired. Brittle and intolerant of mechanical load, which makes them a natural fit for a non-contact process. Unfired ceramic in particular would not survive a mechanical drill at these sizes. Representative work: 10 µm to 100 µm holes in unfired ceramic, and 25 µm holes in 250 µm and 500 µm fired ceramic.
Glass and sapphire. Transparent at many wavelengths, which inverts the usual problem. The challenge is coupling energy in without inducing subsurface cracking. Ultrashort pulses, which deposit energy faster than the material can fracture in response, are the standard approach.
Polymers: Kapton, PET, plastics. Ablate cleanly at ultraviolet wavelengths, where photons break molecular bonds directly rather than heating the bulk. This is why UV is the default for polymer medical components, and why mask projection is viable here: the clean ablation threshold supports parallel drilling of repeated patterns.
Silicon and diamond. Silicon is central to semiconductor via work, including blind vias drilled to a controlled depth. Diamond is among the hardest substrates available and effectively unmachinable by conventional means. Laserod has produced a 175 µm hole in 500 µm synthetic diamond.
Multi-layer and dissimilar stacks. Blind holes that must stop on a buried layer are a distinct class of problem, requiring the process to terminate at an interface rather than a depth. A 200 µm blind hole in epoxy stopping on copper is a representative case.
Full material and process detail is available on the micro drilling capabilities page.
Hole Diameter, Tolerances and Geometry
This is where most specification decisions get made.
Diameter range. Down to 5 µm at the low end, through to large diameters at the upper end. Below roughly 25 µm, process selection narrows considerably, and pulse width becomes the dominant variable.
Positional tolerance. Feature registration of ±5 µm. For dense arrays, hole-to-hole positional accuracy usually matters more than absolute position on the part, which is worth specifying explicitly on the drawing.
Substrate envelope. Up to 700 mm by 700 mm, up to 3 mm thick. Parts outside that envelope need a different approach.
Aspect ratio. The depth-to-diameter ratio is the hardest constraint in micro drilling. Ratios around 10:1 are common production territory. Beyond that, the beam has to deliver useful energy to a point increasingly shadowed by the hole walls. Research is pushing the ceiling: Zhang and colleagues, publishing in the Journal of Materials Research and Technology in January 2024, produced approximately 100 µm holes with aspect ratios above 20:1 in Inconel 718 using a two-step femtosecond method. Higher aspect ratios are achievable, but cycle time and process development cost rise sharply.
Taper. Holes naturally taper because the beam converges to a focus and loses fluence with depth. Taper is controlled, not eliminated. Trepanning, focus position, beam shaping, and helical techniques all adjust it. Work published in the Chinese Journal of Lasers demonstrated taper engineered across a range from -2.6 degrees to +2.3 degrees through helical drilling, where negative taper means the entrance is smaller than the exit. The same 2024 Inconel study reduced taper below 0.05 degrees using a secondary laser spinning step. The practical takeaway for a drawing: state whether you need straight-wall, controlled positive taper, or a specific entrance-to-exit relationship, because the process is selected around that answer.
Quality metrics that belong on the print. Beyond diameter and position, specify what actually matters for function: roundness or circularity, entrance and exit diameter separately, maximum heat-affected zone, maximum recast layer thickness, acceptable micro-cracking, and edge condition. Hole quality is assessed against these, and a drawing that omits them leaves the process open to interpretation.
Pulse Width and Wavelength Selection
Two variables set the quality ceiling before anything else is decided.
Pulse width governs thermal damage. Nanosecond pulses are fast and economical and suit applications where some heat-affected zone is acceptable. Picosecond pulses narrow that zone considerably. Femtosecond pulses deposit energy faster than heat can diffuse into surrounding material, producing what the industry calls cold ablation: minimal HAZ, minimal recast, minimal micro-cracking. Femtosecond processing costs more per part, so it is specified where thermal damage is functionally unacceptable rather than as a default.
Wavelength governs absorption. Laserod operates across 1064 nm, 532 nm, 355 nm and 266 nm. Infrared at 1064 nm couples well into many metals. Ultraviolet at 355 nm and 266 nm is preferred for polymers, where direct bond-breaking produces cleaner edges, and for reflective metals where infrared absorption is poor. Shorter wavelengths also focus to a smaller spot, which matters directly at the small end of the diameter range.
The practical consequence: a vendor limited to one laser source is limited to the materials that source suits. Access to femtosecond, picosecond, and nanosecond sources across four wavelengths means the process gets matched to the material rather than the material being forced to accept the available process.
Industry Applications
Medical and biotech. The densest application area. Catheters requiring side-port holes, microfluidic devices, drug delivery components, biosensors, surgical instruments, and implants. Picosecond and femtosecond processing suits these parts because thermal damage and recast are functionally unacceptable in devices that contact tissue or meter fluid. Laserod’s picosecond and femtosecond processes are used in FDA-regulated device manufacturing, including components for cataract surgery systems and surgical stents. Representative work includes 50 µm holes in a composite medical device. More detail on medical and biotech laser services.
Aerospace and defense. Cooling holes, instrumentation orifices, precision pinholes, and sensor components, frequently in superalloys and composites where the geometry is demanding, and the documentation requirements are strict. Laserod has manufactured spacecraft parts and components for NASA and its subcontractors, and high-precision components for defense applications, and is ITAR registered. See aerospace and defense.
Semiconductor. Micro vias, blind vias, buried vias, and through-silicon features for 3D packaging. Minimum thermal damage and zero debris are non-negotiable here. Representative work includes 100 µm holes in 500 µm silicon, 75 µm holes in 250 µm silicon with minimal HAZ, and 1 mm blind vias in silicon. See semiconductor laser services.
Automotive. Injector and spray orifices, sensor components, flex circuits, and resistor sensor trimming. Hole geometry directly controls spray pattern in fuel systems, making taper and exit quality functional specifications rather than cosmetic ones.
Solar. Via drilling in cell architectures, on thin and fragile substrates where mechanical processing is not viable.
Universities and research. Prototype and single-run work for institutions including Stanford, MIT, and Lawrence Livermore National Laboratory, where the requirement is usually a geometry that has never been produced before rather than a repeat production run.
When Micro Laser Drilling Is Not the Right Process
Laser is not the correct answer for every hole. Five cases where another process is the better engineering and commercial choice:
Very high volume commodity parts in ductile sheet. Punching amortizes tooling across millions of parts and will beat laser on cost per hole. Laser wins when volumes are moderate, geometry changes, or the material cannot be punched.
Substrates thicker than 3 mm. Outside Laserod’s envelope. Thicker material generally calls for a different process entirely.
Holes above roughly 1 mm in diameter. Conventional machining is usually faster and cheaper at that size. Laser’s advantage is concentrated at the small end.
Aspect ratios beyond about 20:1. Achievable, as the published research shows, but cycle time and development cost rise steeply. Worth confirming the depth is a genuine functional requirement before specifying it.
Highly reflective metals without wavelength flexibility. Copper and aluminum absorb infrared poorly. A vendor without shorter-wavelength sources will struggle, and the result is slow cycles and inconsistent quality.
An honest process assessment at the quoting stage saves more money than an optimistic one.
Specifying a Micro Laser Drilling Job
What to include when requesting a quote, ordered by how often it gets left off:
- Material and exact thickness
- Hole diameter, with entrance and exit called out separately if they differ
- Through-hole or blind, and for blind, depth or the layer it must stop on
- Hole count per part and positional tolerance, including array pitch
- Acceptable taper, or straight-wall requirement
- Maximum HAZ and recast layer, if functionally relevant
- Roundness and edge quality requirements
- Annual volume and prototype versus production intent
- Any inspection or documentation requirement
Material behavior is specific enough that a sample part usually settles questions a drawing cannot. Laserod produces sample parts from supplied material and drawings for exactly this reason.
Frequently Asked Questions
What is the smallest hole micro laser drilling can produce?
Down to approximately 5 microns in diameter, with feature registration of ±5 microns. Below 25 microns, pulse width and wavelength selection become the limiting factors on quality rather than the positioning system.
What materials can be laser drilled?
Stainless steel, aluminum, copper, glass, sapphire, ceramic, silicon, diamond, Kapton, plastics, carbon fiber and composites, among others. The process is not restricted to conductive materials, which is its main advantage over EDM.
How does micro laser drilling compare to mechanical drilling on tolerance?
Mechanical drilling loses dimension as the tool wears, so hole dimensions drift across a production run. Laser drilling has no contact element and no wear, so dimensional consistency holds across large hole counts.
Can laser drilling produce holes that are not round?
Yes. Square, rectangular, and other non-round profiles are achievable because beam position is programmed rather than determined by tool geometry. Cross-sectional profile can also be controlled: cylindrical, tapered, or a specified entrance-to-exit relationship.
What is the maximum aspect ratio achievable?
Around 10:1 is common production territory. Higher ratios are achievable with ultrashort pulse techniques, and published research demonstrates above 20:1 in nickel superalloys, but cycle time and process development cost increase substantially.
Does micro laser drilling damage the surrounding material?
Every laser process produces some heat-affected zone, but pulse width controls its extent. Femtosecond pulses deposit energy faster than heat diffuses, minimizing HAZ, recast and micro-cracking. Where thermal damage is functionally unacceptable, that is the process specified.
Closing
Micro laser drilling is a process-selection decision before it is a vendor decision. The material, the diameter, the aspect ratio, and the tolerance together determine whether laser is correct, and which laser.
Laserod has spent over forty years on that question, across thousands of micro-hole and via projects in materials from thin film to 3 mm substrates. Send us the material and the drawing, and we will produce a sample. Request a quote or get in touch with our engineering team.