• The mirrors are not silvered or aluminized (metal coated) but are a type called 'dielectric'. They are made by depositing many alternating layers of hard but transparent materials having different indexes of refraction. The thickness of each is precisely 1/2 the wavelength of the laser light (632.8 nm being the most common for a HeNe laser). This results in reflection by interference with very high (>99.9%) efficiency - much greater than for even the best metal coated mirrors. However, note that for a sufficiently long HeNe tube (one with high enough gain), it would be possible to use a pair of freshly coated or protected aluminum mirrors though performance would be pretty terrible. I've gotten a 10" long HeNe tube with an internal HR and Brewster window at the other end to lase using the aluminized mirror from a barcode scanner - just barely. The first HeNe laser would not have been possible without dielectric mirrors despite its length since the wide bore resulted in very low gain.
• The mirrors may be perfectly flat (planar) or one or both may be spherical (concave - positive - with respect to the cavity) with a typical radius (r = 2 * the focal length) equal to the length of the cavity. The latter is a configuration called 'confocal'. Curved mirrors result in an easier to align more stable configuration but are more expensive than planar mirrors to manufacture. I do not know how significant this might be for the common small narrow bore HeNe tubes - I have found tubes with planar mirrors on both ends as well as those with spherical mirrors on the output end. Some will also have some 'wedge' to minimize production of ghost beams and instability resulting from reflections directly back into the resonator. You may be able to tell which type you have by looking at a reflection off of the inner surfaces of the mirrors at each end (assuming the one at the non-output end is not painted or covered). Assuming the outer surfaces are flat, a concave mirror will reduce the size of the reflection very slightly compared to a planar mirror. If wedge is present, the reflections from the front and back (interior) surface of the mirror will shift apart as you move further away (though this may be tough to see on the Anti Reflection (AR) coated output mirror since the reflection from the AR coated surface will be very weak). To further complicate matters, the front (outer) surface of the mirror at the output-end of the tube may be ground to a (slight) convex shape as well resulting in a positive lens which aids in beam collimation.
• One of the mirrors will be nearly totally reflecting and the other will only be partially reflecting at the laser wavelength. These are called the High Reflector (HR) and Output Coupler (OC) respectively. Note that the HR isn't perfect - there will be a low intensity beam exiting from that end of the tube as well as from the OC end assuming it is not covered with paint or tape. Since the reflection peaks at a single wavelength, this type of mirror actually appears quite transparent to other wavelengths of light. For example, for common HeNe laser tubes, the mirrors transmit blue light quite readily and appear blue when looking down the bore of an UNPOWERED (!!) tube.
• The mirrors usually don't have any 'user' adjustments. However, the cylindrical end pieces are mounted by thinner sections of metal tubing so that some slight changes to alignment may be possible with appropriate fixtures. I do not recommend this because: (1) grabbing the high voltage electrodes is not likely to be pleasant and (2) it is too easy to break the seal if you get carried away. There should be no reason for the alignment to have changed unless you whacked the tube - it was set at the factory. However, if you suspect an alignment problem, it is easy to check. Then, you can decide if attempting an adjustment is worth the risks. However, long high power tubes (i.e., 20 mW and up) may require fixtures to maintain mirror alignment even when the mirrors are internal. Such tubes will not be stable by themselves because thermal expansion will result in enough change in alignment to significantly alter beam power - even to the extent of extinguishing the beam entirely at times! There may even be a 'This Side Up' indication (not related to the orientation for linearly polarized tubes) on the HeNe tube or laser head as gravity affects this as well (the alignment and thus power, not the gas, electrons, ions, or light!) and can significantly affect operation. I do not know if this sort of behavior is common or only likely with tubes that are marginal in some way.
• The main beam will emerge from the partially reflecting mirror but this may be at either end of the tube depending on model. For example, where the tube is enclosed in a metal barrel, the HV connections will be to the anode end and the beam will exit from the cathode end. With this arrangement, the positive output of the power supply and ballast resistor can be very close to the tube anode. The entire barrel (cathode) can be connected to earth ground for safety. There is a slight benefit to having the output coupler mirror at the anode-end of the tube due to the typical long-radius hemispherical cavity configuration. With the bore running almost to the mirror mount, more of the mode volume is inside the bore and thus the gain will be slightly higher. But the difference is only really significant for "other color" HeNe laser tubes which have very low gain and these are more likely to use anode-end output configuration.
• Unlike common metal coated mirrors, these dielectric types are not perfectly reflective. Thus, there will be a weaker beam visible from the non-output end of the tube if that mirror is not covered (blocked or painted over). One use of this is to permit monitoring of laser power for purposes of optical power regulation or other closed loop applications.

• OC from .5 mW, 12.5 cm Melles Griot model 05-LHR-002-246 internal mirror HeNe tube: 99.3 percent.
• OC from 2.25 mW, 26 cm Spectra-Physics model 084-1 internal mirror HeNe tube: 99 percent.
• OC from 20 mW, 75 cm Aerotech model unknown internal mirror HeNe tube: 97.7 percent.
• OC from 50 mW, 177 cm Spectra-Physics model 125 large frame external mirror HeNe laser: 99.4 percent.

Due to the behavior of the photodiode at low light levels, the absolute precision of the readings is somewhat questionable. However, the relative reflectivities of these mirrors is probably reasonably accurate. Note, in particular, the high R of 99.4% for the very long external mirror laser compared to the low R of 97.7% (T of 2.3%) for a shorter internal mirror tube. I expect that in addition to the length of the bore, part of this difference is due to the absence of Brewster window losses in the internal mirror tube resulting in a higher gain so that more energy can be extracted via the OC on each pass.

Mirrors for non-red HeNe lasers must be of even higher quality due to the lower gain on the other spectral lines. The OC will also have higher reflectivity for this reason. For green HeNe tubes (which have the lowest gain of all the visible HeNe wavelengths), the transmission is about 1/10th that of a similar length red tube. For example, the reflectivity of a typical green HeNe tube OC is 99.92 to 99.95 percent (.08 to .05 percent transmission) at 543.5 nm.

Notes on making these measurements:

• Position the sensor far enough from the laser that it doesn't see a significant amount of bore light (incoherent glow from the discharge).
• Block ambient illumination from falling on the sensor.
• Orient the mirror being tested at a very slight angle so light doesn't bounce back to the laser's output mirror.
• Assure that the sensor sees only the main beam and not any of its (possibly multiple) reflections from the mirror surfaces.
• Take a reading with the sensor blocked (the 'dark current') and then subtract it from the actual measurements.
• Average several readings of both the laser and transmitted power to minimize the error introduced due to power variations from mode cycling.

Ion Beam Sputtered (IBS) coatings have a much higher packing density, so they withstand the (i.e., 450 °C) frit sealing temperatures and don't even shift 1 nm. Nowadays, everything is hard sealed, with the exception of the high-end (long precision) Brewster tubes. Hard-sealing a BK-7 window puts a lot of stress on it, and that just isn't acceptable on the high-Q tubes. So, those get fused silica windows optically contacted (lapped and polished surfaces that are vacuum tight.) (In fact, with this type of seal, if there is no adhesive present, the windows can be easily removed from your dead, leaky, or up-to-air tubes by heating the Brewster stem and window with a heat gun. The window can then be popped off with your thumbnail!)