Ancient gold granulation process

There was a mention of this a little while ago concerning old work
found with no eutectic material by SEM. I felt that joining by
diffusion shouldn’t be that difficult and therefore a pre technical
era product would not be really unusual. A good empiricist (ala T.
Edison) could develop this . And in the pre technical era all learning
was empirical. The critical Temperature to begin diffusion is 1/2 the
temperature between absolute zero and the melting point There is a
little flexibility in this exact point but it will be reasonably
correct. This temperature for gold would be only 668 K or or 395 C
with the melting point at 1063 C. This is a time at temperature process
. The lower the temperature the slower the process. the higher the
temperature the faster the reaction until the melting mess occurs.
These processes would .best be done in a reducing atmosphere ( at
least in a more primitive era). This would not be difficult to achieve
either by a smoky fire or by enclosing the part in a clay maybe mixed
with dung… The dung clay mix is typical of primitive lost wax work so
this assumption seems reasonable. Oppi Untracht puts the diffusion
temperature at 890 C with some variation with the work done over a
charcoal bed (CO reducing atmosphere). Never done this probably
never will. Seem resonable Peter (Rowe)? Jesse

Jesse, All most all ancient gold contained other metals like copper
and silver to some degree. The addition of copper to the area of the
joint was the most common way of creating the bond and if the amount
of copper is small enough and the temperature of diffusion is high
enough the copper will disperse and the result will appear to be a
uniform piece of metal. The copper added will be indistinguishable
from the copper in the alloy.


James Binnion Metal Arts
4701 San Leandro St #18
Oakland, CA 94601

I had an old goldsmith teach this to me years ago. I don’t know how
would test out, I suspect it would show an alloying condition at the
joint rather than a pure diffusion joint. At any rate, what the
lesson was: Use a fish based glue to secure all the granules. Then
wether your heat source was a torch, kiln, charcoal or what ever the
granules adhered. He said he also used powdered malachite, (because
of the copper it would definately form an alloying joint). I suspect
there is a high metal content in those old glues but I never got it
tested and don’t have any around anymore. For what ever that bit of
is worth to all you. Norman

The granulation secret has always fascinated me and if microscopic
examination reveals no copper alloy in the joins then obviously the
secret hasn’t really been rediscovered. Does anyone really know if
there was any difference in their ordinary soldering and granulation?
Was it the same process? In some close up photos in my jewelry history
books it looks in some areas that the granulation and soldering have
melted together in one area. There is also the ancient name for
malachite which means something like flux or something, don’t have
that reference right here, and malachite is a copper mineral, perfect
for soldering and/or granulation? Brennan mentioned he thought
eutectic bonding (spelling?) could be accomplished by holding the
piece at a certain temerature over time…so has anyone tried this?
Someone else thought maybe pressure was involved, but I seriously
doubt that as not all granulation was done on flat pieces. Oppi’s
book has a picture of a piece of ancient Indian jewelry with extremely
tiny, medium and large gold granulation all on one piece that is just
out of this world. Granulation seemed fairly widespread as far as I
can tell during one period of history so maybe the secret was much
easier than what we use today…it sure looks like a daunting
haphazard procedure today that require infinite patience…too much
for me…Dave

No dissagreement on the actual metal content ,but the use of a copper
salt in the binder is to form a local copper enrichment to form the
eutectic. If the copper was in the parent metals you wouldn’t really
form the same thing. Depeneding on the EDX on the SEM you can see
extreamly low levels . The newer and more costly the toy the lower
they go with better baselines. Diffusion works without the eutectic
formation. Look at a powdered metal text or talk to someone from the
valley accross the Bay.They live and die on diffusion effects. The
eutectic makes it easier and faster is not the enabling factor for the
process. Jesse

I have been waiting for this discussion for quite a while now. I have
been using cupric carbonate mixed with flux for my last granulation
projects with satisfactory results. To those on Orchid that gave me
the necessary and the impetus to do it, thank you. It does
seem to me to be just another form of soldering, however. Years ago, I
read an account and explanation that I would like clarified if anyone
around here is familiar. Here is the short version. The writer claimed
that by cooking gold granules in charcoal, it was possible to form
gold carbide on the surfaces of the granules. Gold carbide in this
explanation has a lower melting point than gold. When properly fluxed,
glued into place, and heated agqin, the gold carbide would melt and
the carbon would burn off forming a perfect weld. Any takers?

I do not belive that gold forms a carbide. And most carbides I know
of have high melting points they are ceramics.

James Binnion Metal Arts
4701 San Leandro St #18
Oakland, CA 94601

I am glad you wrote this as I neglected to mention in my note that
the goldsmith made his granules in charcoal then ran them thru a stone
sieve. Thanks for keeping all the facts out there. Norman

You know some of these so called “lost” processes are still being
used by traditional peoples, I wonder if anyone has seen how its done
overseas by guys working on little charcoal fires with blowpipes? Do
they use powdered malachite?

Do any of you know what the diameter or percentage of the sphere
forms the actual contact between the bead and the substrate, and also
to the adjacent bead? Do the spheres inter penetrate, touch on a
tangent, or is there a meniscus between the spheres? Thanks, Will

Try this article at Ganoksin:


Charles Lewton-Brain/Brain Press
Box 1624, Ste M, Calgary, Alberta, T2P 2L7, Canada
Tel: 403-263-3955 Fax: 403-283-9053 Email: @Charles_Lewton-Brain

No taker .The mechanism is very well defined. All powdered metalurgy
is based on this mechanism called sintering. There is no MAGIC the
science is established ,you just have to look in the right places. You
are right in that what you are doing is similar to soldering in that
soldering is a diffusion based phemominum where the filler metals are
diffused into the base material at the joint… dfluxes generaly are
glass formers that dissole oxides and clean the surface as well as
keeping the surface shielded from oxugen. Ancient people were just
as smart as present ones ,but they didn’t have the tremendous legacy
of widely available and education that we enjoy. They
observed results of process and passed down their knoweldge slowly one
to one. They didn’t have the tools to dig deep. Sometimes the
explainations of why are wrong, but the process still works. Jesse

Jesse. I have spent the past 16 years learning about diffusion
lamination of precious metals (I make mokume jewelry) I studied it
as much as I can without the aid of a getting materials degree. My
lamination process is done without any component reaching liquidus. I
also understand the analytical abilities of a SEM and its decedents I
used to work in a semiconductor fab lab on the other side of the bay.
I would find it extremely unlikely that granulation could be done by
way of diffusion and not some form of local fusion. There are ways
to fuse high purity gold without high heat (the kumboo thread that is
going on right now is a good example)) however they all rely on some
form of pressure or extremely clean inert or reducing atmosphere for
extended periods of time (hours). I do not belive that the ancients
could create that kind of atmosphere and the pressure would have left
very distinct marks on the surfaces. The only way I can think of
that might be possible is to heat the gold to the point where there
is both a solid phase and liquid phase present and then cool it
back down. It would be very hard to do this without melting the piece
as the difference between solidus and liquidus is only 10-20F .


James Binnion Metal Arts
4701 San Leandro St #18
Oakland, CA 94601

“Aside from the technical difficulties of making the physical bond
between the grains and the metal sheet, the Etruscans used incredibly
tiny granules (0.14mm as opposed to 2mm grains used by the Sumerians,
1.1-0.4mm grains used by the Trojans and 0.25mm grains used by the
Greeks). Some Etruscan jewels appear to be dusted with minute gold
beads covering a substantial surface area or arranged in tidy
patterns. Each granule maintains its own integrity; metal isn’t
welled up around it, yet it’s fused securely to the metal underneath.”(
Granulation: reviving an ancient technique, Elise B. Misiorowski)

We are talking about granules no bigger than half a grain of sand and
smaller. Has anyone succeeded in using the eutectic process to bond
such minute granules? If so, kindly tell. It was this very fact that
suggested to me(I’m not yet aware of others on the same line of
thought, please be so kind as to give me some references) that the
process may not have involved melting(i.e. reaching a eutectic melting
point) Would one even be able to discern any noticeable “flash” of
flowing eutectic between these miniscule granules. To control a torch
flame in such a way as to remain below puddling point of these
granules I imagine would be difficult. Whether it be copper alloy or
gold carbide any near melting temperatures would seem to pose a
daunting challenge for the success of granulating such small
components. As far as making the small granules it was suggested that
this might be difficult in an age that didn’t have draw plates. A
plausible technique was the use of gold shavings(as opposed to our
jump rings or pallions) layerd in the charcoal. But let me run
another possibility by you all. Brennan got me looking up powder
metallurgy and in this modern process there are quite a few techniques
for creating metal powder, size of grain completely controllable.
Only one stuck out as being an available possibility for the Etruscan
masters. In PM they pour the molten metal through an orifice at a
controlled rate and in turn this stream is subjected to a controlled
jet of gas or air in effect “spraying” the metal. A cruder form of
this could have one etruscan pouring a “steady” stream of molten gold
from a crucible and another with a blowpipe providing the "steady"
gust to separate the stream into small particles(over a basin of water
maybe). Likely it would involve some sorting. Dave’s “Indian jewelry
with extremely tiny, medium and large gold granulation” suggests a
very decorative sorting. :slight_smile: As far as the idea of diffusion and
pressure mentioned in the Tribute to the Etruscan post. I hadn’t
figured that the encasing in clay ( brennan suggested addition of
dung) would be limited to flat pieces. In powder metallurgy the
powder is bonded under some 50-100lbs/ at diffusion
temperatures, though I think that this may be overkill for the
aspirations of a goldsmith. How much pressure would the clay exert on
the expanding metal within before the clay cracks? we won’t know
until one of you tries it or until I finish making my furnace. One
thing we do know is that with diffusion a lower temperature only slows
down the process so there is no need to fret over having to keep the
clay from cracking at the "speedy temperatures. The very articles
made by the Etruscans suggest that these craftsman were already very
familiar with the virtue of patience:)

Ad Meiorum Dei Gloriam

They will touch on a tangent which is the only way a sphere can touch another or
a flat surface. In other words a point contact.

Remember this topic started with a statement about maybe only one
sample and whether it was possible . It doesn’t mean anything against
other practices that use a copper salt to form a eutectic at the bond.
This make the job much easier. Jesse

The writer claimed that by cooking gold granules in charcoal, it
was possible to form gold carbide on the surfaces of the granules.
Gold carbide in this explanation has a lower melting point than

This is an old myth born by authors like Marc Rosenberg with his
influentual book from 1910-25. It was perpetuated without any
scientific proof. Goldcarbide appears very rarely and is not constant.
It therefore could not be the explanation for the granulation process.
And it was not logical, either, for, as anyone knows, carbides have
instead very high melting points.

The European point of view: the granulation process HAS been
discovered in all details with the means of natural sciences. There
have been analyses of granules, sheet and the bonding area, and copper
concentrations have been detected in the bonding areas up to 15 times
higher than in the granules and the sheet. So they did use copper
salts, which has always been described in the literature of following
centuries (Plinius, Theophilus and others). Soldering was different,
because the granulation bond is not strong enough for parts exposed to
direct force. If you are interested in details you may read:
Granulation by Jochem Wolters (in German) or very good and compact in
Italian language: Granulazione etrusca by E. Formigli and G. Nestler.
Best wishes.

India is the place where the process is probably still practiced on
the “curbside” in a pattern as close to ancient methods as anyplace.
Untract seems to be the authority in English at least. He has his two
metal oriented books and a more recent one on jewelry from India. I
think he lives in Finland now. Maybe you could try to find him ? I
have two older and have to really look at them again. The new one is
at the local library and maybe I will be lucky enough to check it out
and see what’s there. gold should be an ideal material to make small
particles since you could pound out very thin leaf and cut this into
very small pieces to granulate in a hot carbon bed. Separation by wet
flotation should work even with really fine stuff. even air flotation
ala 3M Super abrasive grading. As I feel never say never even if you

Since you are in Arizona and I am in Texas maybe I can visit if you
set up to try this I will dig more, although I thought I was going to
go away. Jesse

This thread got started with a report of granulation that was tested
by a scanning electron microscope an with x-ray dispersion analysis
without finding copper. The technique should easily see copper if it
was there and the machine,operator and technique were any good. You
have to assume this. The queston was could this have been done. I
don’t see why not ,but the copper definately makes the job easier, and
as a result would be the thing to do. Its purely an acaademic thing.

Hi, An ancient dental tech :o) once taught me how to fuse a small
higher melting point piece of gold to a lower melting point piece.
We took a piece of thick masonry with a smooth glaze on one side that
fit flat against a concrete block and put them in the freezer
overnight. The piece to be soldered to was put in a thin (1/4" thick
piece of soldering investment) and also put into the freezer. When
you are ready, you place the piece to be soldered in place and hold
it with a little flux. Using a very hot flame, glance it on an angle
until the surface of the invested metal begins to melt. Immediately
get the torch away from it and voila it is stuck. You get only 1 try
without starting over. Hope this is of some use.



Skip Meister
Orchid Jewelry Listserve Member
A day without sunshine is like night!
N.R.A. Endowment
"No man’s life, liberty or fortune is safe…while our legislature is in session."
Benjamin Franklin