QQ: 859818709

Thorlabs PF10-03-P01P Ø25.4 mm Back Side Polished Protected Silver Mirror

Thorlabs PF10-03-P01P Ø25.4 mm Back Side Polished Protected Silver Mirror

Price on request
Category:

Description

Product Overview

The Thorlabs PF10-03-P01P is a Ø25.4 mm (Ø1 in) plano mirror with a protected silver film on the front and an optical polish on the back. The front face carries the working beam and is figured to λ/10 at 632.8 nm with 10-5 scratch-dig. The rear face is polished to λ/4 at 632.8 nm with 40-20 scratch-dig, and it is that second surface which defines the part. Average reflectance is 96% from 400 to 700 nm, 97.5% from 700 nm to 2 µm, and stays above 95% out to 20 µm. The substrate is fused silica, 6.0 mm thick with a tolerance of plus or minus 0.2 mm, the body is Ø25.4 mm with a +0.0 / -0.1 mm tolerance, the clear aperture is Ø22.9 mm, and the coating is specified from 0 to 45 degrees incidence.

A back-side-polished mirror exists for one reason: it lets you measure the beam while it turns the beam. No metal film is perfectly opaque. Inside the coating band a small fraction of the incident light passes through the silver, crosses the substrate and arrives at the rear face. On an ordinary mirror that face is fine ground, so the light leaves as a wide scatter and is lost. Here the rear face is an optical surface held to λ/4, so the transmitted light emerges as a low-divergence beam that can be caught on a photodiode a few centimetres behind the mount. The manufacturer specifies this leakage as well below one percent of the incident power inside the band, which is small enough not to disturb the reflected beam and large enough to drive a biased detector through a transimpedance amplifier.

The rear polish is not a cosmetic upgrade, and it is worth being precise about why. A fine-ground surface scatters the transmitted light into a cone of tens of degrees. A detector placed behind it then collects a fraction that depends on its distance, its active area and the exact roughness of the particular disc, so the reading is a function of the mechanics rather than of the laser. A λ/4 surface instead delivers a beam whose divergence is set by the incident beam, so a fixed aperture in front of the detector passes a stable fraction and the reading tracks the laser. The 40-20 scratch grade matters for the same reason: scratches and pits scatter, and scatter that misses the detector looks exactly like a drop in laser power. A sampling surface that loses light unpredictably is worse than no sampling surface at all.

The two faces are specified to different standards because they do different jobs. The front carries the beam quality of the reflected path, so it gets the tight figure and the tight scratch grade: λ/10 and 10-5, one grade better than the 40-20 of the ordinary frosted-back PF10-03-P01. The rear never touches the working beam; it only has to deliver a collectable sample and to keep its own reflection a clean, predictable ghost, so λ/4 and 40-20 is the right amount of money to spend there. Reading the table the other way round, as though the rear were the inferior surface, misses the point of the part entirely.

The limitation is the band, and it is a real one. Below one percent transmission only holds inside the specified 450 nm to 20 µm band, where the silver film is thick enough optically to be nearly opaque. Outside that window the film transmits a much larger and strongly wavelength-dependent fraction, so the ratio between the rear sample and the reflected power is no longer a stable calibration: a coefficient measured at 1064 nm does not transfer to 532 nm, and neither transfers to a wavelength where the film is partially transparent. There is a second effect at the edges. The rear surface has its own Fresnel reflection of a few percent per interface, and where the film stops blocking, that reflection returns into the substrate and can walk back toward the source. A sampling mirror is a monitor inside its band, not a universal power probe.

The Ø1 in format is the reason this version exists alongside the Ø19 mm PF07-03-P01P. The two share the same coating specification and the same back-side polish, but the 25.4 mm body is what the majority of mirror mounts, kinematic seats, periscope blocks and scan-head turning positions were designed around, and it provides 22.9 mm of clear aperture instead of 17.1 mm. On a beam of 8 to 12 mm that difference decides whether the sampling area is fully inside the coated zone or partly in the edge roll-off, and a sample taken from the edge of a coating is not the same sample as one taken from the centre.

What the sample is good for is relative measurement. The rear photodiode sees a fixed fraction of whatever passes that point in the layout, so it tracks laser drift, window contamination upstream, fibre coupling loss and shutter timing, all of which change the power arriving at the mirror. It is not an absolute power meter: the fraction varies slightly between coating runs and with angle, so the number has meaning once it has been calibrated against a real measurement at the workpiece. Treated that way it is the cheapest closed-loop sensor in a laser system, because it costs one photodiode and no additional optical element in the working path.

Key Features

  • Rear face polished to λ/4 at 632.8 nm, 40-20 scratch-dig: this is the sampling surface, and it is the only reason the part costs more than the frosted-back version
  • Rear transmission well below 1% inside the coating band: enough to drive a photodiode with a transimpedance amplifier, far too little to measurably dim the reflected beam
  • Front face at λ/10 with 10-5 scratch-dig: one scratch grade better than the frosted-back PF10-03-P01 carrying the same silver film
  • Protected silver, 96% from 400 to 700 nm and 97.5% from 700 nm to 2 µm, above 95% out to 20 µm: broadband enough to sample a visible pointer and an infrared working beam with one optic
  • Ø25.4 mm body with Ø22.9 mm clear aperture: the standard one-inch mount format, with enough coated area that a 10 mm beam samples from the centre rather than the edge
  • 6.0 mm fused silica, plus or minus 0.2 mm: the rear face sits a defined distance behind the coating, which keeps the sample geometry reproducible from unit to unit
  • Specified from 0 to 45 degrees: covers a normal-incidence pickoff and the usual 45 degree fold with the same part number
  • No added element in the working path: a pickoff plate brings two surfaces, a wedge and its own wavefront; the polished rear face brings none of that to the reflected beam

Technical Specifications

Model PF10-03-P01P
Type Plano mirror, back side polished
Coating Protected silver (silver film with protective overcoat)
Wavelength range 450 nm – 20 µm
Average reflectance, 400 – 700 nm 96%
Average reflectance, 700 nm – 2 µm 97.5%
Average reflectance, 2 – 20 µm Above 95%
Angle of incidence 0 to 45 degrees
Substrate Fused silica
Diameter Ø25.4 mm (Ø1 in)
Diameter tolerance +0.0 mm / -0.1 mm
Thickness 6.0 mm (0.24 in)
Thickness tolerance plus or minus 0.2 mm
Front surface flatness λ/10 at 632.8 nm over the clear aperture
Front surface quality 10-5 scratch-dig
Rear surface flatness λ/4 at 632.8 nm over the clear aperture
Rear surface quality 40-20 scratch-dig
Clear aperture Ø22.9 mm (90% of diameter)
Rear transmission Well below 1% inside the coating band; sampling only, not a power specification
Damage threshold Coating level for protected silver: 0.225 J/cm² at 800 nm (99 fs, 1 kHz); 1 J/cm² at 1064 nm (10 ns, 10 Hz); CW 500 W/cm at 1.07 µm and 1500 W/cm at 10.6 µm
Mounting Standard Ø1 in optic mounts, kinematic seats and retaining rings
Source Manufacturer published series specification table

Applications

The part earns its cost wherever a fold is already in the layout and a power reading is wanted from that same location. The list below is ordered from the most common case to the more specialised ones.

  • Live power monitoring at a fold you already have: the sampling happens on the mirror that was going to be in the layout anyway, so the monitor costs one photodiode and no new optical surface in the working beam
  • Closed-loop stabilisation: the rear photodiode feeds an acousto-optic modulator driver or a pump current controller, holding delivered power constant through warm-up, through a long shift and through window contamination upstream
  • Measurement at the end of the delivery path: a pickoff after the last turning mirror reports what actually arrives, which is the number that matters when a process drifts and nobody can tell whether the laser or the delivery optics changed
  • Diagnostics where a beamsplitter would spoil the result: a plate adds two surfaces, a wedge and its own transmitted wavefront; the rear polish adds nothing to the reflected beam, so an interferometric or tightly focused path is left alone
  • OEM laser heads and marking systems: one part number replaces a fold mirror plus a beamsplitter plus the alignment labour of installing the beamsplitter, which is a bill-of-materials reduction rather than an optical one
  • Process validation and service work: a temporary mirror with a detector behind it turns a subjective judgement about beam quality into a number that can be written on a service report

Selection Guide

Two questions decide whether this is the right part. Does the layout need a power reading at this point, and does the beam sit inside the coating band. If the answer to the first is no, the frosted-back mirror is cheaper and produces no stray rear beam to deal with.

Requirement Part Why
Fold the beam and sample it, one-inch format PF10-03-P01P (this page) Rear polished to λ/4, under one percent of the beam reaches the rear face inside the band
Fold the beam, no sampling, lowest cost PF10-03-P01 Same silver film and substrate, frosted rear face, front face at 40-20
The same sampling design in a 19 mm body PF07-03-P01P Same coating and back polish, Ø17.1 mm clear aperture, Polaris fixed-mount format
A sample with no fold at all Beamsplitter plate or wedge Use when the beam must continue straight and the sample must come off at an angle
Reflectance above 99% over a narrow band Dielectric laser-line mirror Silver cannot reach 99%; the dielectric stack can, but only inside a restricted band and angle range
Work below 450 nm PF10-03-F01 UV-enhanced aluminum covers 250 to 450 nm; silver is out of specification there and the sampling ratio is not defined
Work above 2 µm with maximum infrared reflectance PF10-03-M02 Mid-infrared enhanced gold reaches 98%, but no back-side-polished version exists in that coating

Detector choice follows from the leakage figure. Under one percent of a few watts is tens of milliwatts, which suits a photodiode with a transimpedance amplifier and a neutral density filter on the front; under one percent of a few hundred milliwatts is a few milliwatts, which suits the same detector without the filter. The one arrangement to avoid is a detector that saturates, because a saturated reading looks like a stable reading and the loop will stop correcting without announcing it. Set the gain so that the expected power sits at roughly half scale at the start of a shift.

Series Comparison

The back-side-polished option sits inside the ordinary one-inch metal mirror family, so the useful comparison is between the two rear-face finishes and then between the coatings.

Item Front face Rear face Ravg Role
PF10-03-P01P λ/10, 10-5 λ/4, 40-20 96% / 97.5% Fold plus live sample
PF10-03-P01 λ/10, 40-20 Fine ground 97% / 95% Fold only
PF07-03-P01P λ/10, 10-5 λ/4, 40-20 96% / 97.5% Fold plus sample, 19 mm
PF10-03-G01 λ/10, 40-20 Fine ground 90% / 95% Fold, moisture tolerant
PF10-03-M01 λ/10, 40-20 Fine ground 96% Fold above 800 nm
Broadband dielectric λ/10, 20-10 Polished or ground 99% Narrow band, high reflectance

The single most common mistake in this comparison is ordering the P01P for the front-face grade and then blocking the rear face with a plate or a blob of paint to stop a stray reflection. That defeats the purpose and pays for a polish nobody uses. If the rear beam is inconvenient, the correct part is the frosted-back P01, which scatters it away by design and costs less.

FAQ

Q: How much light actually comes out of the back?

A: Well below one percent of the incident power inside the coating band, and the manufacturer does not publish a tighter number because it varies with wavelength, with angle and between coating runs. Treat it as a proportional sample rather than a known fraction: calibrate the detector once against a real power measurement at the same point, then use the rear reading to track change.

Q: Why is the rear face only λ/4 when the front is λ/10?

A: Because the rear face never carries the working beam. It only has to deliver a sample that can be collected and to keep its own reflection predictable, and λ/4 with 40-20 does both. Spending λ/10 money on a surface that only feeds a monitor would raise the price without changing a single measurement.

Q: Can I use it outside 450 nm to 20 µm?

A: You can turn a beam with it, but you should not trust the sample. Outside the band the film transmits a much larger and strongly wavelength-dependent fraction, so the ratio between the rear reading and the reflected power is neither small nor stable, and a calibration taken at one wavelength will not transfer. Below 450 nm the correct optic is the UV-enhanced aluminum mirror, which has no back-side-polished equivalent.

Q: Does the sample depend on the angle of incidence?

A: Yes, mildly. The film is specified from 0 to 45 degrees, and transmission through a thin metal layer rises as the incidence angle increases. If the mirror is installed at 45 degrees and later moved to normal incidence, recalibrate rather than assuming the old coefficient still applies.

Q: Will the rear reflection cause trouble?

A: It can, and it is the one thing to check in the layout. The rear surface reflects a few percent of the transmitted light, which is a tiny absolute amount, but it returns along the axis of the sample beam and can re-enter the substrate or reach the source. Tilt the detector a few degrees off axis, or place it behind an aperture, and the problem disappears.

Q: Is the sample polarisation sensitive?

A: The silver film itself reflects s and p slightly differently away from normal incidence, and the transmitted sample inherits that difference. For power monitoring of an unpolarised or randomly polarised beam it does not matter. For a polarisation-maintaining delivery fibre where the state of polarisation is the measurement, use a pickoff designed for that purpose instead of inferring it from a fold.

Q: Can it replace a beamsplitter in an existing system?

A: Often yes, and that is its main commercial argument. A plate beamsplitter adds two surfaces, a wedge and its own transmitted wavefront to the straight-through path. This part adds nothing to the reflected path, because the sampling surface is behind the coating and outside the beam. Where the original layout used a beamsplitter purely to obtain a power reading, replacing it with a fold plus a rear detector usually improves the working beam as well.

Usage Precautions

  • Handle by the rim only, with gloves or finger cots, and never touch either face. The front is a 10-5 surface and the rear is an optical surface feeding a measurement, so a fingerprint on the back is a calibration error rather than a cosmetic defect.
  • Keep the rear face clean and inspect it when the reading drifts. Dust on the back attenuates the sample, and because the detector cannot tell dust from a weaker laser, a dirty rear face produces a correction in the wrong direction in a closed loop.
  • Set the detector gain so the expected power sits near half scale. A saturated reading looks stable and quietly disables the loop.
  • Tilt the detector slightly off the sample axis, or put an aperture in front of it, to keep the rear-surface reflection out of the measurement and away from the source.
  • Do not paint, tape or block the rear face to suppress the sample beam. If the rear beam is unwanted, the frosted-back PF10-03-P01 is the correct part and costs less.
  • Recalibrate after any change of angle, wavelength or mirror position. The sample fraction is a property of the installed geometry, not of the part alone.
  • Tighten retaining rings only until the disc stops moving. A 6 mm substrate bends under a metal ring, and a bent mirror changes both the reflected wavefront and the sampling geometry.
  • Avoid high-humidity storage. The protective overcoat slows tarnish but does not seal the silver, and edge discolouration shows up first in the region where the sample is taken from.

Disclaimer

Specifications on this page are compiled from the manufacturer published series tables and are offered for selection purposes only; the current official datasheet governs. The rear transmission figure is a qualitative statement about sampling, not a guaranteed power ratio, and it varies with wavelength, angle and coating run. Price and delivery are quoted separately by our sales desk.

Contact Us

Tell us the wavelength, the beam diameter at the mirror, the angle of incidence, the power you expect and what the detector has to drive, and we will confirm whether a rear sample of under one percent is enough signal or whether the layout needs a proper pickoff. If the mount is a Polaris fixed seat rather than a standard one-inch ring, mention it, because the Ø19 mm PF07-03-P01P is the same design in that format. Project and bulk quantities are quoted individually.

Reviews

There are no reviews yet.

Be the first to review “Thorlabs PF10-03-P01P Ø25.4 mm Back Side Polished Protected Silver Mirror”

Your email address will not be published. Required fields are marked *

中文EN