31.1  How does remote sensing of chemical pollutants work?. 

The are several techniques, but the one of most interest to the public is 
the system being used to identify grossly polluting vehicles. The system 
consists of an infra-red source on one side of the road, and a detector 
system on the other. The collimated beam of IR is directed at a gas filter 
radiometer equipped with two liquid-nitrogen-cooled indium antimide 
photovoltaic detectors. The beam is split, and passes through a 4.3um 
bandpass filter to isolate the CO2 spectral region, a 4.6um filter to 
isolates the CO region, and a third filter to isolate the HC region. 
A non-absorbing region is also used to compensate for signal strength. 
There are various specific enhancements, such as the spinning gas-filter 
correlation cell in the University of Denver FEAT ( Fuel Efficiency 
Automobile Test ) system used to cost-effectively identify grossly-polluting 
vehicles [1]. "Optical remote sensing for air pollutants - review " by 
M.Simonds et al [2], provides a good introduction to the diverse range of 
instruments used for remote sensing of pollutants. 
 
31.2  How does a Lava Lamp work?.

Contributed by: Jim Webb <jnw4347@email.unc.edu>

A container filled with clear or dyed liquid contains a non-water-soluble 
substance (the "lava") that's just a little bit denser (heavier), and has 
a greater thermal coefficient of expansion, than the liquid around it. 
Thus, it settles to the bottom of the container. A heat source at the 
bottom of the container warms the substance, making it expand and become 
less dense than the liquid around it. Thus, it rises. As it moves away 
from the heat source, it cools, contracts a bit, and becomes (once again) 
heavier than the medium. Thus, it falls. Heavy, light, heavy, light. 
Sounds like a Milan Kundera novel. 
(Actually, to be more precise: dense, less dense, dense, less dense.)

31.3 How do I make a Lava Lamp?.

Contributed by: Jim Webb <jnw4347@email.unc.edu>

Method 1. A new, easy, simple, cheap lava lamp recipe

Use mineral oil as the lava. Use 90% isopropyl alcohol (which most 
drugstores can easily order) and 70% isopropyl alcohol (grocery-store 
rubbing alcohol) for the other ingredient. In 90% alcohol the mineral oil 
will sink to the bottom; slowly add the 70% alcohol (gently mixing all 
the while; take your time) until the oil seems lighter and is about to 
"jump" off the bottom. Use the two alcohols to adjust the responsiveness 
of the "lava." 

This mixture is placed in a closed container (the "lava lamp shape" is 
not required, although something fairly tall is good) and situated over a 
40-watt bulb. If the "lava" tends to collect at the top, try putting a 
dimmer on the bulb, or a fan at the top of the container.

To dye the lava, use an oil-based dye like artists' oil paints or a 
chopped-up sharpie marker. To dye the liquid around it, use food 
coloring. 

Two suggestions for better performance: 1) Agitation will tend to make 
the mineral oil form small bubbles unlike the large blobs we're all used 
to. The addition of a hydrophobic solvent to the mixture will help the 
lava coalesce. Turpentine and other paint solvents work well. To make 
sure what you use is hydrophobic, put some on your hand (if it's so toxic 
you can't put it on your hand, do you want to put it in a container that 
could break all over your room/desk/office?) and run a little water on 
it. If the water beads, it should work fine. 2) For faster warm-up time, 
add some antifreeze or (I've not tried it) liquid soap. Too much will 
cloud the alcohol. Keep in mind that the addition of these chemicals may 
necessitate your readjusting the 90% to 70% alcohol mixture.

Method 2. The "official" way - from a patent [3].

The patent itself is not very specific as to proportions of ingredients. 
The solid component (i.e., the waxy-looking stuff that bubbles) is said 
to consist of "a mineral oil such as Ondina 17 (R.T.M.) with a light 
paraffin, carbon tetrachloride, a dye and paraffin wax." 

The medium this waxy stuff moves in is roughly 70/30% (by volume) water 
and a liquid which will raise the coefficient of cubic thermal expansion, 
and generally make the whole thing work better. The patent recommends 
propylene glycol for this; however, glycerol, ethylene glycol, and 
polyethylene glycol (aka PEG) are also mentioned as being sufficient.

This mixture is placed in a closed container (the "lava lamp shape" is 
not required, although something fairly tall is good) and situated over a 
40-watt bulb. If the "lava" tends to collect at the top, try putting a 
dimmer on the bulb, or a fan at the top of the container.

Method 3. The "less official" way - from Popular Electronics [4]. 

Several non-water-soluble chemicals fall under the category of being 
"just a little bit heavier" than water, and are still viscous enough to 
form bubbles, not be terribly poisonous, and have a great enough 
coefficient of expansion. Among them: Benzyl alcohol (Specific Gravity 
1.043 g/cm3), Cinnamyl Alcohol (SG 1.04), Diethyl phthalate (SG 1.121) 
and Ethyl Salicylate (SG 1.13). [The specific gravity of distilled water 
is 1.000.]

Hubscher recommends using Benzyl Alcohol, which is used in the 
manufacture of perfume and (in one of its forms) as a food additive. It can 
be obtained from chemical or laboratory supply houses (check your yellow 
pages); the cheapest I could find it for was $25 for 500 ml (probably 2, 
maybe 3 regular-sized lava lamps' worth). An oil-soluble dye is nice to 
color the "lava"; Hubscher soaked the benzyl in a chopped up red felt-tip 
pen and said it worked great. [Benzyl alcohol is "relatively harmless", 
but don't drink it, and avoid touching & breathing it.] 

Hubscher found that the benzyl and the water alone didn't do much, so he 
raised the specific gravity of the water a little bit by adding table 
salt. A 4.8% salt solution (put 48 grams of salt in a container and fill 
it up to one liter with water) has a specific gravity of about 1.032, 
closer to benzyl's 1.043. I find that the salt tends to cloud the water a 
bit.. you might want to experiment with other additives. (Antifreeze? 
Vinegar?)

This is put into a closed container and placed above a 40-watt bulb, as 
above. Either way, I would suggest using distilled water and consider 
sterilising the container by immersing it in boiling water for a few 
minutes.. algae growing in lava lamps is not very hip.

Caveat: Some of these chemicals are not good for you. Caveat 2: Some of 
these companies are not good for you if they find you've been infringing 
on their patent rights and trying to sell your new line of "magma 
lights." Be careful.

31.4  What is Goretex?.

Goretex is a dispersion-polymerised PTFE that is patented by W.L.Gore and
Associates [5]. It is classed as a stretched semi-crystalline film, and is
produced by extrusion under stress ( faster take-up rate than extrusion 
rate ). The extrudate is stretched below the melting temperature, often
in the presence of an aromatic hydrocarbon that swells the amorphous region,
creating porosity. The hydrophobic nature of the PTFE means that liquid
water is repelled from the pores, whereas water vapour can pass through.
It is important to realise that once the PTFE pores are filled with liquid
water, the fabric can allow liquid water to pass though until it is dry
again. Thus Goretex-containing fabrics ( such as Nomex/Goretex - which 
consists of an outer aramid fabric, a central Goretex layer, and a cotton 
backing ) should never be used as protection from chemicals as many will
pass straight through. Any water-miscible solvent ( eg alcohol ) can fill 
the pores, and then liquid water can displace it and continue to rapidly 
pass through until the fabric is fully dried out.

31.5  What causes an automobile airbag to inflate?.

The final cause is the production of nitrogen from 10s of grams of sodium
azide, but there are some extra chemicals involved along the way.
Sodium azide is toxic, The airbag inflators are aluminium-encased units
that contain an igniter (squib), gas generating pellets ( or wafers of 
sodium azide propellant ), and filters to screen out combustion products.
The electrical signal ignites a few milligrams of initiator pyrotechnic 
material. The pyrotechnic material then ignites several grams of booster 
material, which ignites the tens of grams of sodium azide, and the azide
burns very rapidly to produce nitrogen gas and sodium.  

The sodium azide is pelletised to control the rate of gas generation by
controlling its surface area. The free sodium would form sodium
hydroxide when it contacts  the water in people's noses, mouths, and
eyes so, to prevent this, the manufacturers mix in chemicals that will
produce sodium salts ( silicates, aluminates, borates ) on combustion.

Inflator units also often have a layer of matted material of alumina and
silica called Fiberfrax in the particulate filter. The Fiberfrax mat reacts
with most of the remaining free sodium in the generated gas. A typical
reaction pathway is as follows [6];-

            300C
   2 NaN3  ------>  2 Na  +  3 N2
  10 Na  +  2 KNO3  ------>  K2O  +  5 Na20  +  N2
  K2O  +  Na2O  +  SiO2  ------>  alkaline silicate glass.     

There are apparently also corn starch and talcum powder used as lubricants 
in the bag, and if the bag explodes these are the powders that contaminate 
people - not the toxic chemicals in the inflator.  

One article quotes 160 grams  of propellant for a drivers-side bag 
( 60 litres of gas) and 450 grams for a passengers-side bags 
( which are 3-5 times larger) . I suspect that may include all of the
above ingredients in the igniter, but not the bag lubricants. 

The bag fills until it reaches slightly above atmospheric pressure, and
the manufacturers now control the bag inflation speed to 90-200mph, which 
is less than the early models - because they were too violent and could
harm occupants. The actual sequence goes something like:-

0  - Impact
15 - 20 milliseconds - sensors signal severe frontal collision.
18 - 23 milliseconds - pyrotechnic squib fired
21 - 27 milliseconds - nylon bag inflates
45 - 50 milliseconds - the driver ( who has moved forward 5 inches)
                       slams into the fully inflated bag
85 -100 milliseconds - the driver "rides the bag down" as the air
                       cushion deflates.

Recently, there have been calls to change the crash testing procedures to
allow the test dummy to be belted in, as seat belt usage is now about 67%.
Having a belted dummy would permit the use of slower inflating airbags, as
the deaths of 30 children ( up to Dec. 1996 ) have been attributed to the 
speed of inflation of the larger passenger-side bag. Early in 1997, the 
US NHTSA finally permitted depowering and/or disabling of passenger-side
airbags. A major airbag supplier is Breed Automotive, Boonton Township, N.J.

More details can be found in specialist articles [7-9], and research is 
continuing into alternative inflation mechanisms, such as compressed gases.
There has been extensive work over the last decade on "hybrid" airbag 
systems. These two-stage systems often use cylinders of compressed gas, 
which can be released at ambient temperatures for situations where low-speed 
deployment is appropriate, or the gas can be rapidly heated for high-speed
deployment. 

31.6  How hazardous is spilt mercury?.

First step - ensure any broken thermometer actually contained mercury, as 
many only contain alcohol. Mercury has an appreciable vapour pressure at 
ambient temperatures, thus if the mercury has split somewhere warm and with 
limited air circulation, then vapour concentrations can accumulate. When 
mercury drops any distance onto a surface, it splatters into hundreds of 
minute globules, resulting in a large surface area. The major hazard is 
the mercury vapour produced from the spill. Mercury usually ends up in carpet 
or cracks in the surface, and so really is only a significant hazard to 
children crawling around the floor. Do not over-react. If the location is 
relatively cool and well-ventilated, there is little danger to adults. Remove
as much mercury as conveniently possible, and just remember when toddlers 
come visiting that there is a slight potential hazard if the area is not 
well-ventilated and is warm. Obviously, if you increase the ventilation, the 
concentrations will decrease faster. The USA ACGIH TLV for mercury vapour is 
0.05mg/m3, whilst the DFG ( Germany ) limit is 0.01mg/m3, and the vapour 
pressure of mercury at 25C is 0.0018mm. At 25C, the equilibrium concentration
would be about 20mg/m3, which is 400 times the permitted TLV. It is unlikely 
that this equilibrium would be reached in areas where there are significant 
airflows, unless the mercury had been finely dispersed ( as in a blown 
manometer, or dropped onto a very rough surface ).

Mercury vapour is rapidly oxidised to divalent ionic mercury by the tissues 
of the body. Human volunteers exposed to tracer doses of elemental Hg 
demonstrated first order kinetics for excretion with a half life of 60 days. 
The lethal concentration for humans is apparently not known, but acute 
mercurialism has resulted from exposures to concentrations within the range 
1.2 - 8.5mg/m3. The human organism is able to absorb and excrete substantial 
amounts of mercury, in some cases as high as 2 mg/day without exhibiting
any abnormal symptoms or physical signs [10].

The Dietary uptake for mercury was estimated to be :-
      3 micrograms/day Adults
      1    "        " young children
      1    "        "  infants.
and the adult uptake was estimated to comprise of
      0.3 air via Hg(0), 
      0.1 water via Hg(2+), 
      3 food via Hg(CH3Hg+).
( EPA Mercury Criteria Document 1979 )

The CRC Handbook of Laboratory Safety [11] has a chapter on mercury hazards.
A good discussion of mercury ( and other metals ) is found in "Metals and 
their Compounds in the Environment: Occurrence, Analysis and Biological 
Relevance" [12]. 

The best method of removing spilt mercury is to use a vacuum with a flask 
and pasteur pipette and chase the little globules around the floor while not 
breathing :-).  Seriously, a simple vacuum system, or even a pasteur pipette, 
can remove most of the large globules. There are special commercial vacuum
cleaners, but never use a household one - as the expelled air will contain
mercury vapour, and the fine metal globules will contaminate the cleaner. 
For nooks, crannies, and cracks  - where the mercury is likely to remain 
undisturbed, you can either apply flowers of sulfur ( fine elemental sulfur )
or zinc dust, with vigorous brushing to facilitate contact, and sweep up the 
excess. If the mercury is going to be re-exposed ( by cleaning, foot traffic 
etc., ), then the zinc dust may be preferred because of an apparently faster 
reaction rate. However, if you have a light-coloured carpet, pouring yellow 
or grey powder is not usually an option, and if the location is warm and not 
well-ventilated near ground level, ensure that toddlers do not spend hours 
every day playing there. 

There have been several studies on the best methods to clean up spills, 
including "Vaporisation of Mercury spillage" [13]. The abstract reports " A 
report on an investigation of the problem in laboratories and industries of 
mercury (Hg) vaporisation from small droplets in cracks and floors. The 
efficacy of other fixing agents besides flowers of sulfur was metered. 
The results show that the use of a sulfur, calcium oxide and water mixture 
was the most successful mixture for fixing mercury droplets. A second 
convenient technique is the use of an aerosol hair spray. A chelating soap 
is available in some countries, and this would presumably be the method of 
choice in dealing with spillages."

Another article includes methods based on amalgamating with zinc impregnated 
in a metal sponge or scrubbing pad for picking up mercury [14], and another
investigates substances that can be used to remove spilled mercury - such as
iodised activated carbon, copper or zinc powders, molecular sieves of copper 
or silver ions, and silica gel [15]. 

Dental amalgam is apparently a finely divided powder of a silver, tin, 
and copper alloy that is mixed with the mercury. The setting time probably
is a function of the slow dissolution of the alloy in the mercury due to
the particle size of the powder used. The mass % of each individual metal 
amalgam when mercury is saturated at 20C is Ag = 0.04, Cu = 0.0032, and 
Sn = 0.62, but I've no idea if that is the ratio actually used. I presume 
the ratio may be varied to obtain the desired physical properties, and that
there would be a theoretical excess of the alloy to ensure minimal free
mercury. The actual amount of mercury vapour from dental amalgam is low, but 
directly measurable by sensitive mercury vapour analysers. The significance 
of mercury vapour from dental amalgam to health has been very controversial, 
however there are now practical alternatives in widespread use.  

31.7  Did molasses really kill 21 people in Boston?. 

From: mica@world.std.com (mitchell swartz) Date: Sun, 4 Jul 1993
Subject: Molasses Accident
 [excerpt from the Book of Lists #3 (Wallace et alia)]

  THE GREAT BOSTON MOLASSES FLOOD
  "On Jan. 15, 1919, the workers and residents of Boston's North End, mostly 
  Irish and Italian, were out enjoying the noontime sun of an unseasonably 
  warm day. Suddenly, with only a low rumble of warning, the huge cast-iron 
  tank of the Purity Distilling Company burst open and a great wave of raw 
  black molasses, two stories high, poured down Commercial Street and oozed 
  into the adjacent waterfront area. Neither pedestrians nor horse-drawn 
  wagons could outrun it. Two million gallons of molasses, originally 
  destined for rum, engulfed scores of persons - 21 men, women, and children 
  died of drowning or suffocation, while another 150 were injured. Buildings 
  crumbled, and an elevated train track collapsed. Those horses not
  completely swallowed up were so trapped in the goo they had to be shot by 
  the police. Sightseers who came to see the chaos couldn't help but walk in 
  the molasses. On their way home they spread the sticky substance throughout 
  the city. Boston smelled of molasses for a week, and the harbor ran brown 
  until summer."
   
  From this we see 21 people were killed, the half life was fairly short for 
  the contaminants. Long term effects were probably negligible.

31.8  What is the active ingredient in mothballs?.

Mothballs were originally made from camphor ( C10H16O, [76-22-2], MP 176C,
BP 204C ), or naphthalene ( C10H8, [91-20-3],  MP 82C, BP 218C ),
but para-dichlorobenzene ( C6H4Cl2, [106-46-7], MP 55C, BP 173C ), became 
cheaply available as an unwanted by-product of ortho-dichlorobenzene
production, and thus became the most common active ingredient. However 
para-dichlorobenzene is also a suspected carcinogen, and naphthalene 
has again become a common active ingredient. Consequently, the best
method of finding the active ingredient is to read the label on the packet,
Note that adding mothballs to modern gasolines will not increase the octane 
rating of the fuel - refer to the Gasoline FAQ posted in rec.autos.tech for 
more details.  

31.9  Is vinegar just acetic acid?.

Most countries have food regulations that permit the use of acetic acid as 
clearly-labelled "synthetic white vinegar". Most vinegars are actually malt 
vinegars ( fermented ), and synthetic acetic acid is not allowed to be sold 
as Malt Vinegar. Most natural, unfortified, malt vinegars are appropriately 
labelled. The classification can get rather messy when bulk suppliers dilute 
malt vinegar concentrates with acetic acid, which itself could either be 
synthetic, or from another fermentation process. Regulations usually require 
any addition of acetic acid to be clearly marked on the label, and the 
product is not normally legally sold as pure "malt vinegar". The amount of 
acetic acid in "natural" malt, cider, or wine vinegars usually ranges from 
4% - 6%, but some examples can have up to approximately 20%. Vinegar is 
produced by the exothermic aerobic bacterial oxidation of ethanol to acetic 
acid via acetaldehyde.

31.10 What are the different grades of laboratory water?. 

There are several techniques used in chemical laboratories to obtain the
required purity of water. There are several grading systems for water, but
the most well-known is the ASTM system, although certain applications (HPLC)
often require purer water than ASTM Type I, consequently additional
treatments such as ultrafiltration and UV oxidation may also be used to 
reduce concentrations of uncontrolled impurities, such as organics.

ASTM Type                                    I         II        III
Specific Conductance   (max. uMhos/cm.)    <0.06      <1.0       <1.0
Specific Resistance    (min. Mohms/cm.)   >16.67      >1.0       >1.0
Total Matter           ( max. mg/l )       <0.1       <0.1       <1.0
Silicate               ( max. mg/l )        N/D        N/D        0.01
KMnO4 Reduction        ( min. mins )      >60.0      >60.0      >10.0

Type                                         A          B          C
Colony Count (Colony forming units/ml)    0 Bacteria   <10      <100 
pH                                          NA         NA       6.2-7.5 
 
The techniques to purify natural waters - which may be almost saturated 
with some contaminants - are frequently used in combination to obtain high 
purity laboratory water. Some purification techniques use less energy than 
distilling the water, and may be used in combination where large volumes of 
"pure" water are required. The design of purified water systems, and the
materials used for construction, are selected according to the important
contaminants of the water. For some applications, 316L stainless steel may 
be required, whereas other applications may require polyvinylidene difluoride
and polytetrafluoroethylene materials. Systems are carefully designed to 
minimise the volume of water remaining static and in "dead ends" - where 
microbes could grow.    

The first treatment is usually a coarse physical filtration using a depth
filter that can remove undissolved large particles and other insoluble
material in the feed water. 

For smaller volumes, distillation is the pretreatment method of choice.
Distilled water is water that has been boiled in a still and the vapour 
condensed to obtained distilled water. While many impurities are removed 
( especially dissolved and undissolved inorganics that make water "hard", 
most organisms, etc. ), some impurities do remain ( volatile and some 
non-volatile organics, dissolved gases, and trace quantities of fine 
particulates ). Distilled water has lost many of the ionic species that 
provided a pH buffer effect so, as it dissolves some CO2 from the air 
during condensation and storage, the pH moves to around 5.5 ( usually from 
close to the neutral pH of 7.0 ). Distilled water has the vast majority of 
impurities removed, but often those residual compounds still make it 
unsuitable for demanding applications, so there are alternative methods of 
purifying water to remove specific undesirable species. 

The next common treatment is ion-exchange, which involves using a bed of 
resin that exchanges with unwanted dissolved species, such as those that 
cause "hardness" ( calcium, magnesium ) in water. Two resins are used, one 
that exchanges anions ( usually a strong anion exchanger such as Amberlite 
IRA-400 - a quaternary ammonium compound on polystyrene ), and one that 
exchanges cations ( usually a strong cation exchanger such as Amberlite 
IR-120 - a sulfonic acid compound on polystyrene ). These resins can also 
be combined in "mixed bed" resins, such as Amberlite MB-1A, which is a 
mixture of IRA-400 [OH- form] and IR-120 [H+ form]. The porosity of the 
polystyrene-based resin is dependant on the amount of cross-linking, which
is, in turn, dependant on the proportion of divinyl benzene used in the
process. Unfortunately, selectivity of a highly porous resin is inferior
to that of a less porous, more cross-linked, resin, so a balance between
the rate of exchange and the selectivity is sought. Agarose, cellulose, 
or dextran can be used in place of the polystyrene base. Sophisticated
systems can have many beds in sequence, using both stronger and weaker
ion exchange resins. 

The exchange potential for ions depends on a number of factors, including 
molecular size, valency and concentration. In dilute solutions, exchange 
potentials increase with increasing valency, but in concentrated solutions 
the effect of valency is reversed, favouring the absorption of univalent 
ions rather than polyvalent ions. This explains why calcium and magnesium 
can be strongly absorbed from feedwater in softening processes, but then are 
easily removed from the ion exchange resin when concentrated sodium chloride 
is used as regenerant. In dilute solutions, the order of common anion 
exchange potentials on strong anion exchangers is sulfate > chromate > 
citrate > nitrate > phosphate > iodide > chloride. In dilute solutions, the 
order of common cation exchange potentials on strong cation exchangers is 
Fe3+ > Al2+ > Ba2+ > Pb2+ > Ca2+ > Cu2+ > Zn2+ = Mg2+ > NH4+ = K+ > Na+ > 
H+ > Hg2+.          

There are two forms of ion exchange for water purification. To "deionise"
feed water, the resins are in the OH- ( anion exchanger ) and H+ ( cation
exchanger ) forms.  If sodium chloride was present in the feed water, the 
sodium ion would displace the hydrogen ion from the cation resin, while 
the chloride would displace the hydroxyl ion from the anion resin. The 
displaced ions can combine to form water. Separate beds of resins can be 
regenerated using 1 Normal acid ( HCl or H2SO4 ) for strongly-acid cation 
resins, or 1 Normal sodium hydroxide for strongly-basic anion resins.
The amount of regenerant is approximately 150 - 500% of the theoretical
exchange capacity of the bed.

If the intention is to merely "soften" the feed water to reduce deposits, 
the beds can be in the Cl- ( anion exchanger ) and Na+ ( cation exchanger ) 
forms. These are replaced by the dilute polyvalent species in the water that 
rapidly form undesirable insoluble deposits as process water evaporates, 
like calcium, magnesium and sulfate. The beds can be regenerated by passing 
highly concentrated salt ( sodium chloride ) solutions through them until 
all the polyvalent ions on the resins have been replaced. This technique 
produces "soft" process water that used in industry.  

When a dilute feedwater solution containing salt passes through a cation 
exchange resin bed in the hydrogen form, the reaction that occurs is:-
Na+  +  Cl  +  R.SO3H  <=>  H+  +  Cl-  +  R.SO3Na
Obviously, the acidity of the water strongly increases as it moves down the
bed, which inhibits the exchange process. If a mixed bed is used, the 
products soon encounter the anion exchange resin and are also removed:-
H+  +  Cl-  +  R.NH2  <=>  R.NH3  +  Cl-
H+  +  Cl-  +  R.NH3OH  <=>  R.NH3  +  Cl-  +  H2O
Mixed bed resins are usually more efficient than equivalent single beds.

If the water feeding the resin beds has already been distilled ( very common 
in laboratories - the resin beds then last much, much longer, and the 
distillation has also removed other impurities  ), then the water is called 
"distilled and deionised". Laboratory water that has had most of the ionic 
impurities removed will have a high electrical resistance, and is often known 
as "18.3 megohm" water because the electrical resistance is >18,300,000 
ohm/cm, but note that non-ionic impurities may still be present.    

An alternative process that has increasingly replaced ion-exchange is 
reverse-osmosis, which uses osmotic pressure across special membranes to 
remove most of the impurities. It is called reverse-osmosis because the feed 
side is pressurised to drive the purified water through the membrane in the 
opposite direction than would occur if both sides were the same pressure. 
The two common membrane materials are cellulose acetate or polysulfone 
coated with polyamine, and typical rejection characteristics are:- 
                       Monovalent    Divalent    Pyrogens, Bacteria
                         Ions          Ions      Organics > 200 MW
Cellulose Acetate        >88%          >94%            >99%  
Polyamine                >90%          >95%            >99%

The huge advantage of RO is that membranes can easily be maintained 
( occasional chemical sterilisations ), are largely self-cleaning, and can 
produce large amounts of water with no chemical regeneration and minimal 
energy requirements - just the pressure ( 200 psi ) required to push the 
water along the membrane surfaces and improve the osmotic yield. RO is 
commonly used as a pretreatment stage when very pure water is required, and 
for situations where large volumes of reasonably pure water are required. 
 
Organic species and free chlorine are usually removed from water by passing 
the water through a bed of activated carbon where they form a low energy 
chemical link with the carbon. These filters are often installed upstream 
of the ion-exchange and reverse osmosis stages to protect them from chlorine 
and organics in the feed water. Polyamine RO membranes require feedwater 
containing <0.1ppm free chlorine, however cellulose acetate membranes can
tolerate up to 1.5ppm free chlorine.   

The final stage of producing "pure" laboratory water usually involves 
passing the deionised  water through a 0.22um filter, which is sufficiently 
small to remove the vast majority of organisms ( the smallest known 
bacterium is around 0.3um ), thus sterilising the water. 

Recently, ultrafiltration has become popular as a means of reducing pyrogens 
( they are usually lipopolysaccharides from the degradation of gram negative 
bacteria ). They are measured by either injecting a sample into test rabbits
and measuring body temperature increase or by the more sensitive Limulus
Amebocyte Lysate (LAL) test. The internal membrane of an ultrafiltration
system has a pore size of <0.005um. This will remove most particles, 
colloidal silica, and high MW organics such as pyrogens, down to about 
10,000MW. These are usually for cell-culture and DNA research, and are 
located at the point of use, however the ultrafiltration unit has to be 
regularly sanitized to prevent microbial growth.
 
Ultraviolet irradiation can be used as a bactericide (254nm) or to destroy
organics by photo-oxidation (185nm). The water is exposed to UV for periods 
up to 30 minutes, and the UV interacts with dissolved oxygen to produce 
ozone. The ozone promotes hydroxyl radical formation, which result in the 
destruction of organic material. Usually a high intensity, quartz mercury 
vapour lamp is used, and is followed by an ion exchange and organic scavenger
cartridge to collect decomposition products. The product water is very low in
total organic carbon.

Dissolved gases can be removed by passing the water through a vacuum 
degassing module that utilises an inert, gas-permeable membrane surrounded 
by a vacuum to remove dissolved gases from the water.

The purest laboratory water is usually obtained after passing through a 
system that can include reverse osmosis or distillation of the feed water, 
followed by activated carbon to remove chlorine and organics. The water is 
passed through ion exchange resins to remove inorganic ions, through a 
UV oxidation stage, followed by a combined ion exchange and organic scavenger 
cartridge, and finally through a 0.22um filter. An additional stage of vacuum
degassing to remove dissolved gases may be added for some applications - such
as for semiconductor production. 

These pure water systems are regarded as " point-of-use ", because it is 
extremely difficult to prevent the reintroduction of contamination during
storage and distribution. The water is commonly known as " 18.3 Megohm " 
water, because it has a specific resistance greater than 18.3 Megohm-cm 
at 25C. It also contains < 5 ppb of total organic carbon, < 10 ppb of total 
dissolved solids, and < 1 colony forming unit / mL of micro-organisms.
  
Details of laboratory and industrial water-purification processes are 
available in the catalogues of equipment suppliers such as Barnstead [16]
and Millipore [17].