Description
Thorlabs PF20-03-M03-10 Ø50.8 mm (2 in) Unprotected Gold Mirror, 10-Piece Pack
Product Overview
This is the 10-piece pack of the PF20-03-M03; the optic is identical to the single-unit version. Each mirror in the box is a 50.8 mm (2 in) round fused silica substrate, 12.0 mm thick, carrying a bare gold coating with an average reflectance of 97 percent from 800 nm to 20 µm, a surface figure of λ/10 at 633 nm, a surface quality of 40-20 scratch-dig, a clear aperture of 45.7 mm and a laser-engraved part number on the fine-ground back face.
The pack exists because of how a particular kind of laboratory actually consumes this optic. In an ultrafast beamline, folding mirrors are not permanent fixtures. Every time the chirp is retuned, every time an amplifier stage is inserted or removed, every time a compressor grating pair is swapped for a different line density, the routing has to be rebuilt, and rebuilding means new fold angles at new positions. The mirrors that were in the old layout do not migrate cleanly into the new one, and the ones that do get handled during the transition.
Handling is the problem. Bare gold has no protective layer, so it cannot be cleaned, and a mirror that cannot be cleaned is a mirror that is consumed rather than maintained. A single fingerprint, a settling aerosol droplet, a chip lifted at the edge by a retaining ring: any one of these ends the optic’s useful life, and there is no procedure that brings it back. In that sense the PF20-03-M03 belongs in the same consumables category as a cuvette or a filter, not in the capital-equipment category where a protected mirror sits.
Buying ten at a time changes the tempo of the work rather than the price of the optic. It means that a layout change is limited by how long it takes to align, not by how long it takes a replacement to arrive; that a spare is always within arm’s reach at the moment a mirror is dropped; and that the decision to retire a marginal optic is made on optical grounds instead of on the basis of what is left in the drawer. A ten-piece box is, in practice, the smallest unit that supports a beamline that is under active development.
Key Features
- Ten identical optics, one part number. Every mirror in the pack is the same PF20-03-M03, drawn from the same production coating run family, so a replacement is a drop-in substitution rather than a re-characterisation exercise. There is no need to record which mirror went where for performance reasons, only for traceability.
- Bare gold, so no overcoat in the pulse path. The dielectric layer that protects a gold film also adds a small, wavelength-dependent phase and a polarization transformation that is difficult to quantify. In a system where the compressor is tuned to cancel the stretcher, removing that term from every fold keeps the dispersion budget attributable to the gratings.
- Average reflectance of 97 percent from 800 nm to 20 µm. A single coating spans the Ti:sapphire fundamental, its second and third harmonics’ long-wavelength side, the short-wave infrared from an optical parametric amplifier, and the mid-infrared from a difference-frequency stage. One spare box covers all of them.
- Pulse damage threshold of 4 J/cm² at 10.6 µm. Quoted for 100 ns pulses at 1 Hz over a 0.435 mm spot; the highest pulsed figure among the metallic coatings on this substrate, because there is no dielectric layer to be ablated ahead of the metal.
- CW damage threshold of 1000 W/cm at 10.6 µm. Quoted over a 1.18 mm spot. Relevant when a fold also carries a continuous alignment beam or a high-average-power idler that was not part of the original design.
- λ/10 at 633 nm on a 12.0 mm fused silica body. The thick body holds its figure in a kinematic mount, and the fused silica substrate has a thermal expansion of 0.55 × 10⁻⁶/°C, so a modest temperature drift in the room does not tilt the wavefront.
- Engraved identification that survives mounting. The part number is laser-engraved on the fine-ground rear face, so a mirror that has been in a mount for two years can still be identified when it is time to decide whether to keep or retire it.
Technical Specifications
| Attribute | Value |
|---|---|
| Part Number | PF20-03-M03-10 |
| Pack size | 10 mirrors |
| Optic in Pack | PF20-03-M03, identical to the single-unit version |
| Coating | Unprotected gold, no dielectric overcoat |
| Spectral Range | 800 nm to 20 µm |
| Average Reflectance | 97% from 800 nm to 20 µm |
| Diameter | 50.8 mm (2 in), +0.0 / -0.1 mm |
| Thickness | 12.0 mm (0.47 in), ±0.2 mm |
| Substrate | Fused silica, CTE 0.55 × 10⁻⁶/°C |
| Surface Flatness | λ/10 at 633 nm, peak to valley |
| Surface Quality | 40-20 scratch-dig |
| Clear Aperture | 90% of diameter (45.7 mm) |
| Angle of Incidence | 0° to 45° |
| Rear Surface | Fine ground, part number laser-engraved |
| Pulsed Damage Threshold | 4 J/cm² at 10.6 µm, 100 ns, 1 Hz, Ø0.435 mm |
| CW Damage Threshold | 1000 W/cm at 10.6 µm, Ø1.18 mm |
| Cleaning | Not cleanable; blow-off only |
| Storage of Unused Units | Original packaging, desiccated cabinet, never stacked, never paper-wrapped |
The damage figures are measured on clean optics and, as Thorlabs points out, the threshold for a given coating does not vary with the diameter of the part. What varies is the cleanliness of the surface you actually put in the beam. A pack of ten does not raise the threshold of any individual mirror; it simply makes it practical to replace one that has fallen below it.
Applications
This section describes where a ten-piece box earns its place, rather than where a gold mirror is used in general. The distinction matters: a single unit answers the question “what do I put at this fold”, and a box answers the question “how do I keep a beamline that changes every few months running”.
- Rebuilding a chirped-pulse amplifier after a compressor change. Moving from 1200 line/mm to 1500 line/mm gratings changes the angle of incidence and therefore the fold geometry on both sides of the compressor. A box of ten covers the four to six folds that have to move, with margin for the one that gets touched during alignment.
- Inserting an additional amplification stage. A new stage adds a stretcher-to-amplifier fold, an amplifier-to-compressor fold and usually a beam-height transition through a periscope, which is two more reflections. Five or six mirrors disappear from the box in one afternoon.
- Multi-pass amplifiers and stretchers. A four-pass or eight-pass arrangement concentrates many reflections into a small volume, where mirrors sit close together and are handled repeatedly. This is the configuration in which bare gold is most likely to be damaged and least likely to be replaceable on short notice.
- Long delay lines and beam transport. A delay line that has to reach several metres needs a dozen or more folds. Each one is a small loss, and the chain is only as good as its worst mirror, so having identical spares on the shelf keeps the chain uniform.
- Teaching and shared-user laboratories. Where a beamline is rebuilt by different people every semester, the attrition rate of unprotected optics is high and predictable. A box converts an unpredictable procurement event into a planned one.
- Prototyping a new layout before committing to protected optics. Build the geometry with bare gold, confirm the routing and the beam heights, then decide which folds deserve a protected coating. The bare gold units that were not damaged remain in service.
Selection Guide
Plan the box against the layout rather than against the budget. The sequence that works is to count folds first, then add the ones that will move, then add a handling allowance.
- Count the fixed folds. A typical femtosecond amplifier with a stretcher, one regenerative or multi-pass stage, a compressor and a delivery periscope lands between ten and sixteen reflections. Write the number down; it is the floor for the box size.
- Add the folds that will move. Any fold downstream of a grating angle adjustment will be repositioned whenever the dispersion is retuned. In a system that is tuned often, treat those as consumable positions rather than as installed hardware.
- Add a handling allowance of about twenty percent. The mirrors lost during a rebuild are not lost to laser damage; they are lost to a glove touching a face, a ring slipping, or an optic being set down on a bench. One spare in five is a realistic figure.
- Confirm the diameter. The three round sizes in this family are Ø1/2 in (12.7 mm), Ø1 in (25.4 mm) and Ø2 in (50.8 mm), and there is no Ø7 mm option. The two-inch size on this page carries a 45.7 mm clear aperture and is the right choice for an expanded beam or a large-collimated beam from a multi-stage amplifier; the one-inch size is usually adequate inside a compressor where the beam is small.
- Decide whether the pack is the right unit at all. If the layout is stable and the mirrors are protected, a single protected unit is the economical answer and the bare gold box is the wrong purchase. The box is for layouts that change.
Series Comparison
Within the gold family the choice is narrow, and it is decided by bandwidth versus peak reflectance rather than by price. The table below covers only the gold coatings, because those are the ones that compete for the folds in an ultrafast layout.
| Code | Coating | Range | Ravg | Role in an ultrafast layout |
|---|---|---|---|---|
| M01 | Protected gold | 800 nm-20 µm | 96% | Fixed folds in a stable layout, where the mirror will be cleaned and kept |
| M02 | MIR enhanced gold | 2-20 µm | 98% | Mid-infrared OPA idler or DFG output, where the band sits entirely above 2 µm |
| M03 | Unprotected gold | 800 nm-20 µm | 97% | Folds that must not add dispersion or rotate polarization; consumed rather than maintained |
| L01 | Unprotected gold, line optimised | 10.6 µm line | 99% | A CO₂ alignment or pump beam sharing the bench, where only one line matters |
If the beamline is broadband, the M03 is the only gold option that combines the full 800 nm to 20 µm range with the absence of an overcoat. The M01 matches the range but adds the overcoat; the M02 gives up everything below 2 µm to gain one point of reflectance; the L01 gives up the entire band to gain two points at a single line. For a system that has to fold a 700-900 nm pulse and a 3 µm idler through the same routing, the M03 is the only one of the four that does not force a compromise somewhere.
FAQ
Q1. If I open the box and use four mirrors now, will the remaining six still be good in six months?
Yes, provided they stay in the original packaging inside a desiccated cabinet and are not stacked or wrapped. Gold does not oxidise, so shelf life is not limited by chemistry; it is limited by contamination and by contact. A mirror that has never left its case and has never been touched is as good in six months as it is today. The failure mode to avoid is the well-intentioned habit of taking a spare out to inspect it, setting it face-down on a bench, and putting it back.
Q2. Should I pre-clean the spares before putting them into service?
No. Cleaning is the damage mechanism, not the remedy. A mirror that comes out of the case with a clean surface should go straight into the mount. Blow-off with clean dry air or dry nitrogen is the only step that is ever appropriate, and even that should be reserved for visible dust. If a spare looks contaminated, it is a candidate for retirement, not for a cleaning procedure.
Q3. What are the signs that a bare gold mirror in service has reached the end of its life?
Three signals are reliable. First, a fingerprint or smudge that is visible under oblique illumination: the organic film absorbs, and the loss is permanent. Second, a diffuse haze or a settled aerosol pattern, which scatters and cannot be blown off. Third, an edge chip or a lifted coating at the rim, usually caused by a retaining ring, which grows with thermal cycling. Any one of these is sufficient; there is no partial recovery path for a bare gold surface.
Q4. How do I store the unused mirrors so that they are still usable?
In the original case, in a desiccated cabinet, standing on edge or held by the case’s own support, never face-down on any surface, never stacked one on another, and never wrapped in paper or tissue. Paper and tissue shed fibres and carry abrasive particles, and both scratch gold on contact. Keep the box away from hardware drawers, where loose screws and washers are the usual source of the scratch that ends a mirror’s life.
Q5. Can I mix protected and unprotected gold mirrors in the same beamline?
You can, and many laboratories do, but the mixing has to be deliberate. If the reason for using bare gold is the absence of overcoat dispersion or overcoat-induced polarization change, then a single protected mirror in the chain reintroduces exactly the term you were avoiding, and the benefit of the other eleven is lost. Use bare gold for the whole chain, or use protected gold throughout and account for the overcoat in the dispersion budget.
Q6. Does the ten-piece pack carry a different specification from the single unit?
No. The optic in the pack is the PF20-03-M03, with the same 50.8 mm diameter, the same 12.0 mm thickness, the same 97 percent average reflectance from 800 nm to 20 µm, the same λ/10 at 633 nm figure, the same 40-20 surface quality and the same 4 J/cm² and 1000 W/cm damage thresholds. The pack changes the quantity and the packaging, nothing else.
Usage Precautions
Treat the box as a consumables store, not as a shelf of instruments. The discipline that keeps a ten-pack useful is almost entirely about what happens before a mirror is mounted.
- Never touch the coated face, even with gloves. Gloves protect against skin oils, not against the lint and abrasive dust that gloves pick up. Hold the optic by the ground rim between thumb and forefinger, and put it into the mount before doing anything else with that hand.
- Open one case at a time. Opening all ten cases at once to check the shipment turns a controlled store into ten exposed surfaces on a bench. Verify the count by weight or by the sealed cases, not by unwrapping every optic.
- Do not stack, do not paper-wrap, do not lay face-down. These three habits account for most of the preventable losses in an ultrafast laboratory. A bare gold mirror resting face-down on a bench is being scratched by its own weight.
- Store desiccated but not sealed wet. A dry cabinet at room temperature is right. Condensation from a cold-to-warm transfer concentrates airborne contamination on the surface, so let a case come to room temperature before opening it.
- Respect the damage thresholds as cleanliness-dependent. The published 4 J/cm² pulsed and 1000 W/cm CW figures are measured on clean optics. Thorlabs notes that contamination lowers the threshold substantially, and on bare gold the contaminant cannot be removed, so the effective threshold after an incident is unknown.
- Retire early rather than late. The whole point of holding ten is that the marginal mirror can be discarded without a procurement delay. A mirror that is suspected of contamination should leave the beamline immediately; keeping it in service to save a unit is how a beamline acquires an unexplained loss and an unexplained scatter background.
- Record which serial positions are in service. The engraved part number identifies the model, not the individual unit, so keep a simple log of installation dates per fold. It is the only way to know how long a given position has been running and to correlate a change in throughput with a change in hardware.
Disclaimer
All specifications on this page are compiled from Thorlabs’ publicly available product documentation and are presented for selection guidance only. Where this page and the manufacturer’s current official datasheet differ, the official datasheet governs. Coating specifications, damage thresholds and packaging quantities may be revised by the manufacturer without notice. Damage thresholds are measured values under the stated test conditions and do not constitute a performance guarantee for any particular application; the user is responsible for verifying suitability at the actual wavelength, pulse duration, repetition rate and beam diameter. Pack contents are as supplied by the manufacturer; please confirm the current pack configuration with our sales team when ordering. Prices and delivery times are quoted separately and are not implied by this page.
Contact Us
If you are planning a beamline rebuild and want help sizing a pack against your fold count, or need the current lead time for the ten-piece configuration, our team can assist.
Hotline: +86 159-2722-0467
E-mail: 859818709@qq.com | Website: www.soar-laser.cn
We can also supply the single-unit PF20-03-M03 for evaluation, the protected gold PF20-03-M01 where cleanability matters more than dispersion, and two-inch mounts to suit the 12.0 mm body.









Reviews
There are no reviews yet.