CR-Scientific Minerals & Earth Science newsletter
Terms of UseJune / July 2003
HTML version
In this issue:
I. "What's That Rock?"
II. Benzene Contamination of Water - with experiment
III. Lab project: Cobalt Nitrate & Zinc
IV. News, and the future
I. "What's That Rock, and Why is it Sitting on Your Workbench?"
Many
readers have surely
faced this question or something like it. Often it's followed by "What
are
you going to do with all these rocks?" A mineral specimen (or a
few
dozen) might sit for months on the workbench amid papers and other
clutter.
"Actually,
I like
to call them minerals. The rocks are out in the driveway...""Uh,
whatever. So
what are you going to do with all these rocks?""I'm going
to find
out what they're made of, what else?" The chemist always wants
to
take things apart and scrutinze them, fiddle with them. The average
person
seems to find this odd, even unsettling."Why would
you want
to do that? Uhhh, nevermind, I think I'll go watch TV...""Oh, okay,
talk to
you later..." The last few words trail off: replaced in the collector's
mind, perhaps, with thoughts of rare mineral crystals, 3-color
fluorescent
skarn assemblages, bubbling flasks of HCl, colorful ammine complexes,
etc...
Yes, eventually he or she finds time to "do something with all these
rocks". Displaying them on a shelf and explaining their subtleties to
anyone who cares to listen is another favorite.Where does
one begin,
though, in the search for the chemical composition of a mineral or
other
object? The variety of tests to which the chemist could subject an
unidentified
mineral fragment is very large. Discounting the amount of time it would
take to
conduct chemical tests for each and every metal that occurs naturally
in
significant amounts (even in moderately common minerals), there's
another
problem: chemical tests tend to consume the sample, so the greater
number of
tests one wishes to conduct, the greater will be the amount of sample
required.
Quantitative analysis in the hands of any but the most experienced can
easily
require more than the available sample size- suppose, for example, the
only
fragment of a given mineral is no larger than a fourth of a rice grain.
Qualitative analysis is not quite so demanding, but it
pays to use micro
or semi-micro techniques anyway. The
analyst who runs
(let alone has access to) EDS, atomic absorption, or mass spectrometry
equipment might cheerfully point out that instrumental analysis
requires only
the tiniest speck of mineral, but we're talking about experiments that
are
within the means of the avocational mineralogist or collector. You'd
have to
save your pennies for a long time to buy an electron microprobe
setup...
you could certainly send your unidentified sample out to
Excalibur
Minerals where they do EDS analysis for a certain price, but we're
presuming here that (like the author) you're just itching to have a go
with
test tubes, chemicals, and spot plates.In order
to determine
what something contains, it helps to know what's in the thing before
you set
out to test it. This sounds contradictory, yes-- but knowing in
advance, at
least roughly, what a mineral probably contains or doesn't
contain can
do away with a great deal of unnecessary work. Then again, we'll see
later why
it isn't a good idea to presume too much.Suppose
the specimen is
a mica, readily classed as such by its appearance and "flaky" habit
(perfect basal cleavage). This tells us several things from the start.
First
off, it's a silicate; second, it probably doesn't contain much Pb, Cu,
Sn, Ag,
V, Co, or the like (unless it's one of a few very rare micas); third,
it's
certain the mineral contains at least one of the following: Mg,
Na, Ca,
K, Al. It may also contain Li, Ba, B, and / or F.Since it's
a mica (and
therefore a silicate), it won't dissolve readily or at all in acid1. It's still worth a try, but don't
count on
favorable results. The insolubility is going to be helpful in this
case,
though. Calcite from the original rock sample, which may still be mixed
in
between the mica grains, can be removed by soaking the mica in
hydrochloric
acid, washing with distilled water three or four of times, and
evaporating to
dryness.2 To
solubilize the mica,
it will probably be necessary to fuse it on charcoal with alkali
carbonate or
bisulfate. The resulting fusion, when cool, can be dissolved in
distilled water
or HCl and the analysis started from there. It's likely you can skip
the
sulfide precipitation step which would normally be done using H2S
or
ammonium sulfide; that is, if you're fairly sure the "hydrogen sulfide
group" metals are absent. You may want to start with NaOH or KOH
solution,
looking for hydroxide precipitates. Any precipitate that doesn't
dissolve in an
excess of the alkali hydroxide can be separated from ones that do, and
so
forth. Halfway
through the
analytical scheme you've devised, some information may surface that
suggests a
change in course. The author had this happen; a pink precipitate from
treatment
with 5% aqueous ammonia suggested something that hadn't been
anticipated in the
sample. Based on this result, a drop of the suspension was put in a
small test
tube, dissolved in a couple drops of concentrated H2SO4,
and treated according to the persulfate oxidation method of Feigl
(1958). The
test was negative for Mn, despite a sensitivity of 0.1 microgram Mn and
a
dilution limit of 1:500000 (1958). The procedure for cobalt was not
done in its
entirety, but the other characteristics of a cobalt hydroxide
precipitate were
not present (i.e., the pink giving way to blue upon addition of
concentrated
acid; the solution giving a dark precipitate with H2O2;
etc.). The only remaining suspect was therefore chromium.Chromium
can form a
grey-white precipitate which does in fact become pale pink with excess
ammonia.
Dr. Rod Beavon, who teaches chemistry at Westminster School in London,
points
this out on his informative website at
http://www.rod.beavon.clara.net/cations.htm.
Though the formation of the pink coloration may not be diagnostic for
Cr, its
presence in the mica analysis suggested further tests with chromium in
mind.
This did not test positive for Mn even after several attempts using a fairly delicate and well-established test. It probably didn't contain Co, either. One clue is that the pink disappeared entirely upon addition of acid, but it reappeared when excess ammonia was again added. At no time did the pink color give way to blue, which Co should have done.
The course
of analysis
becomes even less predictable when one is sure a metal has to
be present
in a sample, yet no supporting evidence surfaces during tests. Consider
a mica
sample that has a pale, dirty green coloration similar to that of
actinolite or
epidote- a sure sign of Fe, one would think. The following evidence,
gathered
during actual analysis of such a sample (the same one tested above), is
not
very supportive of this: 1.) No dirty green or red-brown precipitate
was
observed upon addition of NaOH or ammonium hydroxide, even in amounts
that
raised the pH far into alkalinity; the only colors seen were
greyish-white and,
on standing, pale pink; 2.) The ferrocyanide and the thiocyanate tests
for
Fe+3 were negative in both the solution and the
precipitates; and
3.) the ferricyanide test for Fe+2 was also negative in both
the
solution and the precipitates. Where did the iron go? Was it even there
at all?
Were other ions interfering? Could the mica's dirty-green color have
come not
from iron, but from chromium or some other metal?Perhaps it
wasn't so
useful to make assumptions at the outset. Well, it was useful
in a way-
at least we didn't do tests for Mn or Cr until there were hints they
might be
there. We might still feel confident in skipping the tests for, say,
Ag, Pb,
and Sn altogether, unless we were looking for tiny traces. In the days
when
traditional chemical tests were the only way to know what a mineral
contained,
the analysts might not have floundered about so much in deciding what
analytical schemes to use. For us, the hobbyists, educators, and
amateur
scientists who are rediscovering these half-forgotten techniques,
floundering
about isn't so bad (as long as it's with safety in mind); indeed, the
learning
process is an end in itself.Each
mineral type
presents different possible paths for chemical testing, as well as
different
opportunities for learning a bit of mineralogy and chemistry on the
way. The
example of mica is actually a tricky one when it comes to doing and
interpreting the tests. Since the micas in general don't contain much
in the
way of transition metals (or so we thought at the beginning of
the test
scheme), this leaves out many of the very colorful and highly
diagnostic
reactions3. There are also
quite a few
species with only minor variations in composition. Still, you can
perform your
tests and go down the list of micas, eliminating ones it couldn't
be. No
iron in the sample, for instance, means it cannot be biotite,
zinnwaldite,
siderophyllite, or any other mica that contains iron as an obligate
part of its
structure. No calcium means it can't be margarite or clintonite, and so
on.II. Benzene Contamination - with experiment
Depending
on the
generation in which they were born, readers may recall a story told by
many a
chemistry teacher. "In the old days", the teachers would say,
"benzene was used everywhere. People used to wash their hands in the
stuff… until they found out it caused cancer." Some readers may even
remember when benzene (not "benzine", an entirely different substance
also known as V.M.&P. naptha) was available at hardware stores. Benzene
stands out among
the common hydrocarbons4 as one
of the
most toxic to humans and other higher mammals. Casarett and Doull's
Toxicology
(1991) indicates that the human cytochrome-P450 enzyme system can
readily
transform the benzene molecule into the unstable and reactive benzene
oxide,
which from there can form any of several metabolites thought to be
ultimately
responsible for the compound's toxicity5.
Obviously,
benzene no
longer appears on the shelves of hardware stores, at least not in the
USA.
There is enough benzene in gasoline, however (State of Calif., 1988),
that
countless people still unknowingly "wash their hands in the stuff".
The Merck Index states that benzene is absorbed through the skin in
potentially
toxic amounts (Merck & Co., 1983). It is also fully capable of
contaminating groundwater, as well (State of Calif., 1988). Although
this is
against a chemist's intuition, so to speak-- that is, benzene is
"hydrophobic" and shouldn't mix with water at all-- it turns out that
benzene and certain other constituents of gasoline are soluble in water
to a
fair degree. The Merck Index (1983) indicates a solubility of one part
benzene
in 1430 parts water (w/w), which is about 0.7 grams per liter. The
Leaking
Underground Fuel Tank Manual (1988) suggests a solubility of 1.4 grams
per
liter6. Benzene makes up as
much as 5
mass-percent of gasoline, especially in "high-octane" varieties. Even
when the partition coefficient of benzene in a water / gasoline
separation is
taken into account, there should be significant amounts of benzene and
other
aromatics which go into the water layer.One might
devise a
simple experiment such as the following7.
First, some high-octane gasoline is shaken
together with an equal volume of water for about 30 seconds; next, the
water
fraction is drained off and collected using a separatory funnel.
Finally, this
water is tested qualitatively for aromatics (mainly benzene, toluene,
xylenes,
and ethylbenzene) according to the method of Feigl (1956). It may be
surprising
to students that these "hydrophobic" compounds go into the water
fraction as easily as they do… in fact, enormously more
C6H6 will dissolve than the U.S. Environmental
Protection
Agency's upper limit of 0.005 milligrams benzene per liter in drinking
water
(EPA, 2002). A
spectrophotometric
assay would be useful to determine just how much benzene were in a
given water
sample, down to some limit of detection that might not be up to EPA
standards
but would still be quite instructive for a classroom demonstration. The
method
of Snell and Snell (1936) could be used to detect the benzene after
first
converting it to nitrobenzene. A standard curve consisting of known
concentrations of benzene could be plotted and compared to the
unknowns. With
commonly available computer software we can bypass entirely the tedious
algebraic method for determining an equation for this standard curve-
this is
accomplished by using the trendline or linear regression functions
available in
Microsoft Excel or Corel Quattro Pro.Because of
benzene's
water solubility, people whose water comes from wells situated near
underground
gasoline tanks ought to be concerned if the tanks should leak. The
EPA's
website (www.epa.gov) and many, many other sources suggest this is in
fact a
serious concern. Fortunately, benzene and other aromatic compounds can
be
degraded in the environment by various microbes, though the
effectiveness of
this varies depending on several factors (Kazumi, et al., 1997). While
much is
already known in the realms of benzene's chemistry, toxicity, and
biodegradation, these areas are interesting for classroom discussion
and
further study. III. Lab project: Cobalt Nitrate and Zinc Compounds under the Blowpipe
Supplies: safety goggles / face shield; heavy gloves; cobalt nitrate solution (approx. 7%); carbon block; crucible tongs; dropper; propane torch; fragment of a zinc mineral (for comparison, try two zinc minerals, one of which is not a silicate); fragment of an aluminum mineral (such as a feldspar); zinc oxide ointment
General procedure: As with all experiments, you alone are responsible for safety. Protective eyewear and leather / canvas gloves must be worn. Do not perform this experiment near papers or other flammable materials. It should be done outdoors.
1.
Place a
fragment of the mineral sample on the carbon block and heat intensely
with the
oxidizing flame until a coating forms around the heated area. If no
coating
forms after prolonged heating, proceed to step 2.2.
Let the block
cool for at least 1/2 hour. After cooling, add a drop or two of cobalt
nitrate
solution to the coating on the block and then re-heat in the oxidizing
flame.
If no coating is present, let the drop fall directly on the mineral
sample and
re-heat; however, dark-colored minerals won't give very good results in
this
case.3.
Let the block
cool again for 1/2 hour. Note the color, if any, that remains where the
cobalt
nitrate solution was added. 4.
Try the
procedure again with a tiny spot of zinc oxide ointment on the block
(make sure it's cool before you touch it!);
heat
it intensely in the oxidizing flame to burn away all the petrolatum and
other
additives, leaving just the ZnO. Do you notice a color difference when
the ZnO
is hot, as opposed to letting it cool for 1/2 hour or so? Does the
grass-green
color appear with cobalt nitrate?5.
Try the
procedure again with an aluminum mineral instead of a zinc mineral;
this time,
look for a deep blue color instead of a green hue when you add the
cobalt
nitrate solution. Does this procedure work if you use a highly
infusible
aluminum mineral such as corundum? Why, might you guess, or why not?Conclusions & Discussion:
When a
zinc mineral
other than a silicate is heated with the oxidizing flame on a carbon
block, a
fine sublimate of zinc oxide forms around the assay. Cobalt nitrate
solution
reacts with this upon heating to produce a vivid, grass-green color.
Smith
(1953) indicates that tin and antimony can also give greenish
colorations, but
these are indistinct or "dirty". Zinc
silicates will
leave a coating that gives an ultramarine blue coloration rather than
the
expected grass-green. This blue is similar to that given by aluminum
(Smith,
1953).If the
conditions are
right, a coating of zinc oxide can also form when brazing together
metals with
a torch. Some of the zinc in the [brass] brazing rods will volatilize
as the
oxide when the metal is hot enough, re-condensing on the cooler regions
of the
metal as a thin coating. As one might expect, this coating is yellow
when hot
and white when cold. Though the author hasn't yet tested this coating
with the
cobalt nitrate reaction, it should give the grass-green reaction
characteristic
of zinc oxide. It is interesting that brazing rods often have a coating
of
borax on them; this appears to enhance the formation of the zinc oxide
coating
on the cooler metal parts, just as in the case of fusing a mineral with
soda or
borax on charcoal.IV. News and Future Events
At least
two new
articles are up on the site: a brief article on tube testing for
fluorine is on
this page, and preparation of
ammonium sulfide
reagent is on this page. These are
modified
procedures based on those in the classic texts of fifty years ago or
earlier.Don Peck,
a chemist and
avocational mineralogist with many years of experience, has written a
comprehensive book on mineral analysis and identification. Hopefully
when he
publishes the work that we'll be able to offer some copies of it here
on the
site. We've seen the manuscript of Mr. Peck's book and must say it
looks
very informative.Dana
Morong is working
on a compendium of sources, little-known hints, and corrections to
information
found in the classic mineral analysis texts. This, too, sounds quite
promising,
and if possible we'll make it available on this site- however, Mr.
Morong says
he might put his compendium on a disk which could be included in the
back of
Don Peck's book. Either way, we'll try to let you know in the near
future.Once again
we'd like to
mention the Ultra
8V
centrifuge. We've tried it for qualitative mineral analyses and
have found
it very handy. As for making up biological stains... spinning the
preparation
down in the centrifuge seems much less messy than putting a highly
colored
solution through filter paper and getting the dye all over one's hands,
the
bench top, etc. For convenience, the Ultra 8V has variable speed, timed
shutoff, a locking lid, and suction cup feet. It holds up to 8 tubes of
15 mL
capacity. The smaller sleeves, also included with the unit, will hold
12 x 75
mm test tubes nicely.
This unit is great!
Speaking
of test tubes,
additional lab glassware is now
available, with
more sizes and types on the way. A shipment of polypropylene bottles has also arrived-
these
are excellent for storing caustic alkali solutions and other reagents
which
would otherwise attack glass. How many of us have ruined glass
containers with
concentrated NaOH...There's an
ongoing
effort to improve the site's content, organization and layout. Please
also feel
free to submit suggestions for this newsletter. Your feedback is always
appreciated. Feel free to comment on what you'd like to see more of,
what you'd
like to see less of, what you find too detailed, what you find not
detailed
enough, etc.That concludes this issue of the CR-Scientific newsletter.
Until next time, stay safe and have fun.
Email: sales_AT_crscientific_DOT_com
(replace the "_AT_" with an @ symbol and the "_DOT_" with a period. In order to thwart spam-bots, I'm no longer providing a direct link or easily-harvested address.)
Notes:
1 Some readers would say "what do you mean it won't dissolve in acid? Just use HF". No, thanks. Anything that can dissolve glass is worth staying away from if possible. This writer is not feeling bold enough to work with HF, save that miniscule amount that might be liberated during mineral tests.
Attack by
HCl, H2SO4, or
HNO3 may not occur noticeably within a few minutes or even hours, but
leaving
the sample immersed in the concentrated acid for a week or more (with a
cover
to prevent evaporation) is sometimes helpful. Phlogopite will dissolve
slowly
in HCl when left this way. A 10 x 75 test tube of HCl is sitting on the
writer's workbench, and in it is a bit of mica that is nearly
dissolved...
after about two months.2 A sample was treated in this manner and found to have significant calcite clinging to it; the calcite impurities were too small to see with the unaided eye, but the HCl made them stand out plainly with a tiny stream of CO2 bubbles. Within about 30 seconds the bubbling stopped, leaving only the mica flakes.
3 As mentioned, the one, common exception in micas is of course iron. The ferrocyanide and thiocyanate spot tests (both for ferric iron) and the ferricyanide test (for ferrous iron) are available.
4 We're not including benzo[a]pyrene or the like in a list of "common" hydrocarbons. Though benzo[a]pyrene is widely encountered in trace amounts, it's [hopefully] not in large quantity in any one place.
5 The literature points to phenol as one of the products; this arises from a simple non-enzymatic rearrangement of the benzene oxide (Amdur, et al., 1991). Phenol is one molecule that ought not to be knocking about inside living cells.
6 The reason for the discrepancy is unclear but is assumed to be due to differing laboratory conditions; however, one can probably assume that the true value is somewhere between the given ones at room temperature. In other words, about a gram per liter should dissolve- a very significant amount from a toxicological standpoint.
7 Keep in mind that gasoline is toxic and must be handled cautiously. It will penetrate polyethylene or latex gloves in short order. Aside from gasoline's toxicity, there's the extreme fire hazard. While you're at the gas station filling your tank, it's easy to forget just how dangerous (!) this material is. When you're in the lab or anywhere gasoline is outside your car's fuel tank, pay constant and careful attention to the hazard.
Works Cited / Suggested Reading:
Amdur, Mary, J. Doull, and C. Klaassen, eds. Casarett and Doull's Toxicology: The Basic Science of Poisons. New York: Pergamon Press, 1991.
Anger, V., and Feigl, F. Spot Tests in Inorganic Analysis. Amsterdam: Elsevier, 1972.
Beavon, Rod. "Inorganic Cation Analysis", web article at http://www.rod.beavon.clara.net/cations.htm. 2000.
Chamot, Emile, and Mason, Clyde. Handbook of Chemical Microscopy. London: Chapman & Hall, 1940.
Environmental Protection Agency. National Primary Drinking Water Standards. EPA publication 816-F-02-013, 2002.
Feigl, F. Spot Tests in Inorganic Analysis. Amsterdam: Elsevier, 1958.
Feigl, Fritz. Spot Tests in Organic Analysis. Amsterdam: Elsevier, 1956.
Hillebrand, W., and Lundell, G. Applied Inorganic Analysis. New York: John Wiley and Sons, 1929.
Kazumi, J., et al. "Anaerobic Degradation of Benzene in Diverse Anoxic Environments". Environmental Science & Technology 1997, 31, 813-818.
Merck & Co., Inc. The Merck Index, Tenth Edition. Rahway, New Jersey, 1983.
Smith, Orsino C. Identification and Qualitative Analysis of Minerals. Princeton: Van Nostrand, 1953.
Snell, F., and Snell, C. Colorimetric Methods of Analysis. New York: D. Van Nostrand, 1936.
State of California Leaking Fuel Task Force. LUFT, Leaking Underground Fuel Tank Manual: Guidelines for Site Assessment, Cleanup, and Underground Storage Tank Closure. California, 1988.
While the information in this newsletter is thought to be accurate to the best of the writer's current knowledge, it is not guaranteed to be free of errors or to be suitable for any particular use. The procedures and experiments outlined within can be dangerous or even fatal if carried out improperly. If you choose to attempt any of them, you proceed entirely at your own risk.
You may distribute this newsletter freely, provided the contents of the file are not truncated or altered in any way. Please email the address given above if you find any errors or omissions or would simply like to make a suggestion.
Terms of Use
Lab Glassware
Main Page