[Rockhounds] Fw: SOFTWARE FOR INDENTIFYING MINERALS

[DonH]

Hi Horst,


Having contributed to two recently published books on mineralogy, I can 
tell you that software for determinative mineralogy won't always get you 
far, depending on the type of minerals you are trying to identify, and 
how exact you want the ID to be; e.g., will you be satisfied with the 
group name (e.g., apatite) or do you need to know the exact species 
within the group.  Also, will you be satisfied with 5 or 6 minerals it 
could be or do you absolutely need to know the exact mineral.  I think 
you are someone who collects micromounts and rare species.  As you 
probably know, there are well over 4,000 valid species and thousands of 
varieties; many of these will be very difficult to identify.

In Geology 101, we give the students a tray of minerals to ID by color, 
streak, hardness, cleavage, luster, etc.  These are all unique and are 
all common rock-forming minerals.  In mineralogy lab, the first day, 
they do the same exercise, except a few of the minerals are very similar 
and cannot be identified by determinative properties.  Then we introduce 
them to refractive index, relief, pleochroism, optic class, etc., by 
which they can identify many more minerals.  After that we give them an 
exercise in which they can identify 6 out of 8 mineral grains--assuming 
they have been paying attention--but the last two have such similar 
optical properties that they can't be identified with certainty.  After 
that they learn the principles of EDS, XRD, etc.

This book recently published by the Mineralogical Record contains a lot 
of information on how to ID minerals at home.

   http://www.minrec.org/bookdetail.asp?id=48

(I contributed to the book but receive no compensation for sales).  The 
book does have a disk with useful software on it, but the reader still 
needs to do some analytical work in the kitchen first.  Color is a very 
variable property, and many silicate minerals have a colorless streak. 
However, specific gravity and optical properties add far more value to 
the ID process. The book gives instructions on how to obtain those 
values, at least to 1 or 2 decimal places.

There are plenty of minerals that require multiple advanced analytical 
techniques to ID.  For example, telling apart all the amphiboles 
requires both chemistry and structure, and then there is still some 
interpretation and assumption involved (one reason being that electron 
beam instruments can't measure (OH-), so it must be assumed by 
difference).  Some of them are distinctive, but many are not, and it 
would be fruitless to try to positively identify all the amphiboles by 
simple determinative techniques.  On a similar note, I recently analyzed 
eight samples of fluorescent apatites from Franklin & Sterling Hill, NJ. 
  All but one were mixed phosphate/arsenate apatites, and I expected at 
least 2 or 3 of them to be the same, but they were all different.  The 
microprobe operator had a fit when I asked him to configure a scan for 
zinc and lead in addition to the usual suspects, and asked me three 
times if I was sure there was significant arsenic (I finally had to show 
him the Franklin & Sterling Hill apatites in the Handbook of Mineralogy 
  to prove they really exist).  When all was done, I was getting about 
95-98% weight totals, which I usually don't accept as valid, but in this 
case it means that 2-5 weight% of the specimens were elements for which 
we didn't even analyze.  After he corrected the data for drift and ran 
the results through the microprobe software, the printout gave an 
identity for a mineral I've never heard about.  The calculated 
stoichiometry was so bad that it simply cannot exist in this 
manifestation of the universe.  I have to resort to doing the 
calculations by hand.  In the end, I can provide a first-order 
approximation of which 2 or 3 specific minerals each one might be, but 
until I get some help from someone who is familiar with analyzing these 
species (after all, I'm still a student, and these are somewhat 
difficult), I refuse to try to assign a positive ID.  Yet you will find 
any number of people who would look at these, squint, rub their chin, 
hold a finger in the air, and declare a definite identity.  The most 
difficult thing for a collector or geologist to say is "I have no idea," 
but ironically it is the phrase that should be uttered more than any 
other.  Sometimes you just can't tell without instrument analysis, and 
sometimes, not even then.

Okay, back to the question at hand: how much dirty work are you willing 
to do on the specimen, and how willing are you to accept uncertain results?


Fondly yours,
Don