Wally Chan is a retired manager from Tandem Computers (now part of Hewlett-Packard) with a strong interest in science and engineering. I asked Wally what he knew about the hazards of mercury as used by audio enthusiasts and he came up this well-researched report. - J. Atwood
Mercury Vapor Rectifiers: Breakage in the Home
by Wally Chan
Mercury quantities and releases from household and industrial uses should be minimized because of the cumulative poison effects [1] [2] [3] [4] [5]. One uncommon specialized use is the mercury vapor rectifier, containing approximately 10ug of mercury vapor at 25 microns pressure (60 degrees C) and a total of .5 g of liquid elemental mercury [6]. If totally absorbed into the body, .5 g of elemental mercury has resulted in death, and approximately 1 to 4 gm is regarded as a lethal dose. But what risk occurs in practical circumstances from broken mercury vapor rectifiers?
In older homes, traces of mercury vapor are emitted from mold growth inhibiting paint additives [7], as well as from past thermometer and other spills. To control chronic exposure risk to continuously exposed householders, especially the young, an ATSDR recommended upper limit for mercury vapor in residences is 1.0 ug/m³; for elemental mercury [8]. This is a tighter limit than the OSHA occupational upper limit .1 mg/m³. The 10 ug of mercury vapor from a working rectifier, if suddenly released into a reasonable sized, multiple m³ volume room, will be well under these safety recommendation limits. However, there is still the question of evaporation of the remaining .5 g of liquid mercury into the room air. It turns out that vapor production is slow enough that chronic exposure over many years is the primary concern.
Consider first the case where the liquid mercury remains in one drop. The evaporation rate of a drop of liquid mercury was studied at room temperature in [9], both from an observed and theoretical standpoint. The undisturbed, long term rate of evaporation of a small drop of .2g appears to be of the order of 1 ug/hr. Measurements suggest that a protective film forms on exposed mercury drops and a loss rate of about .5 ug/hr exists for a 2 gm drop if undisturbed and 1.5 ug/hr for a .2 gm drop if frequently measured, disturbing the film. Theory predicted an order of magnitude faster evaporation and proportionality to surface exposed, so film formation and frequency of film disturbance is more significant than surface area alone. Also, the evaporation of a drop was found to violate the OSHA concentration limit by measurement [10] only within a 1 inch hemisphere surrounding the drop.
Evaporation rate data from [11] indicates that mercury evaporates approximately 10X faster at 60°C compared to room temperature (not including any film or atmospheric pressure effects). If the mercury rectifier envelope cracked but continued to remain warm for a long period of time, evaporation into the home environment could be faster than 1 ug/hr, but is unlikely to exceed 10 ug/hr.
Consider next a hypothetical10×10x8 ft=22.7 m³ room with mixing but no outside air exchange. The 1.0 ug/m³ limit will be reached in 23 hours from evaporation at room temperature, and perhaps an order of magnitude sooner (2 hrs) if the drop remains heated to 60°C. But if the room meets the rather minimal 4.7 L/s=16.9 m³/hr single occupant ventilation needs [12] the hot evaporation contribution cannot exceed (10 ug/hr)/(16.9 m³/hr)=.6 ug/m³. So if the room has enough air exchange and mixing for one or more person’s needs, the initial release plus subsequent hot evaporation cannot exceed the 1.0 ug/m³ vapor concentration chronic exposure limit and is more than two orders of magnitude below the OSHA occupational limit.
Consider now the case where the liquid mercury escapes the envelope, and gets into a form of increased surface area due to multiple droplet formation, adsorption onto carpet fibers or wood, or unfortunate vacuum cleaning. Since the evaporation rate attainable is proportional to the surface area, vapor concentration chronic exposure limits could be exceeded, as described in [13]. If liquid mercury gets soaked into carpet, textiles or wood fiber, or should it form an amalgam with aluminum, the substrate usually must be cut away and discarded as mercury hazardous waste.
Recommendations in the event of breakage:
If a mercury vapor rectifier tube’s envelope should break, turn off the associated equipment and allow it to cool down. Ventilate the room and leave until the equipment cools to room temperature. If there is a spill, follow the recommendations of [14] to clean up liquid mercury. Avoid actions that would break up drops or otherwise increase surface area and evaporation rate. Keep nose more than one inch from the spill at all times. Replace the tube and recycle the broken tube and spill kit as mercury hazardous waste. Enjoy your replacement rectifier’s operation, but also consider any additional measures you might take to avoid future releases.
References and Notes
1. Compact florescent lamps contain 4 mg Hg. Regular florescent lamps contain 40 mg Hg.
2. Breakage and spillage of certain household antiques (pendulum clocks, barometers) present opportunities for release of large amounts of liquid mercury. See for example http://jama.ama-assn.org/cgi/content/full/298/4/397.
3. Attempting to recover metals from heating amalgams in the home environment is a really bad idea. See http://www.cdc.gov/mmwr/preview/mmwrhtml/00014464.htm. Both the people and the house were unrecoverable. For those of you using mercury vapor rectifiers on aluminum chassis, Wikipedia has an interesting entry on amalgams relating to progressive aluminum destruction: http://en.wikipedia.org/wiki/Mercury-aluminum_amalgam.
4. The form of the mercury (elemental liquid metal, elemental vapor, compounds such as methyl mercury) matters regarding absorption and toxicity. For a discussion of health effects see http://cerhr.niehs.nih.gov/common/mercury.html.
5. Because of the extreme vulnerability of the nervous system in both developing fetus and children, pediatricians have recommended elimination of mercury from the home environment including minimizing direct use of mercury in doctor’s medical devices to lessen later environmental releases and the further contamination of the food chain. See http://aappolicy.aappublications.org/cgi/content/full/pediatrics;108/1/197 for discussion of mercury in vaccines, the environment, and effects on the developing nervous system. In addition to the direct exposure concerns, there is a global concern with mercury released in energy use, material goods, and in the waste flow of manufacturing. Even if the trace amounts are well below direct human harm, the concern is that the worldwide aggregate flow to the oceans feeds the methyl mercury chain and it ends up eventually at the top of the food chain, in us. For example, mercury is used in an electrolytic chlorine production process. The chlorine process leaks some mercury into the environment and the derived chlorine containing chemicals contain trace mercury. When these chemicals are used in household and industrial applications, trace mercury is released. Also, durable end products that contain chlorine (with trace mercury) as a component release the mercury contaminant originating from the chlorine production process as the discarded product degrades.
6. Electron Tube Division, Radio Corporation of America. Electron Tube Design, 1962 p805. Mercury content is .5 g total, .01 mg vapor for type 816.
7. Agocs, Etzel, Parrish, Paschal, Campagna, Cohen, Kilbourne, Hesse, “Mercury Exposure from Interior Latex Paint,” New England Journal of Medicine, Oct 18, 1990 p1096-1101.
8. Agency for Toxic Substances and Disease Registry (ATSDR) mercury vapor for residential indoor air not to exceed 1,000 ng/m³. See link cited in [13] below. OSHA, .1 mg/m³ maximum permitted mercury vapor exposure for 8 hr day, 40 hr week.
9. Thomas G. Winter. “The evaporation of a drop of mercury,” Am. J. Phys. 71(8), August 2003 p783-786.
10. Portable Jerome 431-X Mercury Vapor Analyzer instrument used in [9] to directly measure drops evaporating.
11. Dushman and Lafferty. Scientific Foundations of Vacuum Technique, 1962 p 697. Real evaporation will be slower due to film formation and atmospheric pressure. So 10X over room temperature is an upper limit estimate.
12. ASHRAE standard 62-1989 was used at its lowest limit: 10 cfm=4.7 L/s per occupant recommendation. This standard has since been upwardly upgraded. The American Society of Heating, Refrigerating and Air-Conditioning Engineering (ASHRAE) recommends (in its Standard 62-1999, “Ventilation for Acceptable Indoor Air Quality”) that homes receive .35 air changes per hour, but not less than 15 cubic feet per minute (cfm) per person. (From http://www.epa.gov/iaq/homes/hip-ventilation.html). The more conservative 10 cfm number was used for estimating toxicity, to reflect a more tightly shut household. For still more on this topic web search: house air exchange rate.
13. For an example of the household concentrations attainable from a single mercury switch spill from a thermostat see http://www.atsdr.cdc.gov/HAC/pha/watervliet/msa_p1.html. The ATSDR vapor threshold should not be exceeded by the evaporation of a single drop in the 60°C condition if there is normal room ventilation. But increasing the surface area even at room temperature can cause a potentially unsafe condition particularly for the young.
14. Cleanup procedure-see http://www.epa.gov/mercury/spills/index.htm#flourescent.
One of the more eye-catching features in a lot of home-built “extreme” tube audio amplifiers are mercury-vapor rectifiers. Their hazy blue glow that is modulated by the current draw of the amplifier adds to the organic life that attracts people to vacuum tube amps. But concerns about safety have polarized the audio community, with some fearing that their homes may become EPA hazard sites! And, do mercury-vapor rectifiers have a sonic benefit in tube amplification? This article will try to answer this question. A 
The oscillator board is mounted in a separate insulated sub-enclosure and is wired directly to the power supply. This avoids the oscillator bounce at start up but more importantly it keeps the RC components of the oscillator at a fairly steady temperature, reducing frequency drift. This is further improved by using ordinary metal film resistors with a typical tempco of + 100ppm/°C and matching them with polyphenylene sulphide film caps which have a tempco of - 100 ppm°/C. Once the oscillator warms up it holds frequency within 0.01 Hz which is about as much as can be expected with analogue techniques. The output voltage is set at 1.0V.
When the motor arrived I was very surprised - it was one of the famous Papst “flywheel” motors with external rotor (fig 1). As I checked it over it became obvious that it was neither two phase nor a synch motor. The motor had three leads, as do many two phase synch motors, but the resistances of the windings as measured between the three leads were equal. This is typical of three phase motors, a two phase motor will typically show twice the resistance across the outside of the pair of windings as between each winding and the common centre. I couldn’t detect any rotor magnetisation, so the motor either had the best shielding ever or it was not synchronous.
