There are two things one needs to consider:
1. Laser Induced Damage Threshold (LIDT):
The LIDT is the condition at which an optic will destruct due solely to the power of the laser. It turns out LIDT is NOT related to the power density (expressed as watts per unit area), hence all the spurious values quoted with such "authority" in the literature. We cringe when we see competitors quote such values...
The way to categorize a CO2 laser optics LIDT is: Power divided by beam diameter, (the P upon D value) which has units of watts per mm. This LIDT value is the same if you use a small laser focussed to a minuscule spot or use a monster laser with a large beam. The validity of P/D is also predicted by software models and mathematical heat transfer models.
Gold coated copper mirrors as made by Laser Beam Products have a P/D value of 4000W/mm. So if you take a 4KW laser, focuses it to 1mm diameter it will just cause a gold coated copper mirror to destruct. (Or a 2000 W laser in 0.5mm beam diameter, or 40,000W laser in 10mm, or 400W laser in 0.1mm beam diameter they all have a P/D of 4000) . Below this the mirror will not destruct due to the laser power.
For ZnSe AR/AR coated (such as a lens), there is also a known LIDT value, about 5000W/mm.
For silicon mirrors the LIDT is about 300W/mm (pretty poor).
Laser Beam Products has measured hundreds of gold coated copper mirrors in a variety of conditions and the 4000W/mm LIDT is remarkably consistent.
The problem comes with ZnSe used for transmissive optics. The intrinsic LIDT of 5000W/mm is valid for sure, but it is very sensitive to external factors:
- Mounting a lens so there is a small level of mechanical force applied and the LIDT plummets.
- Mounting the lens and overtightening it.>
- Using a high pressure cutting gas can cause the LIDT to fall by 90%.
- Allowing the lens to become dirty.
In practice a ZnSe lens is a borderline choice in an industrial environment with a high powered laser. In a clean room, with skilled design, careful mounting and maintenance it should be OK.
2. Thermal distortion:
A gold coated copper mirror has such high heat conductivity the temperature gradients on the surface are small, and the resulting distortion is insignificant. It also happens in a time scale of milliseconds, so the mirror reacts instantly to the beam being turned on and off.
Like for like a ZnSe has 40 X more optical distortion than a gold coated copper mirror. This can cause the focus to shift and the spot size to alter with changes in power. More annoying is the time scale this happens in, around 20 - 40 seconds. So the laser drifts and shifts quite noticeably during the work.