Granulation Method

Note that while there are many 'secret' recipes and granulation was 'lost' this was as far as I can tell a result of gentleman scholars and acadamics in the late 19th and early 20th century who were too snobbish and ethnocentric to realize that much of the world still does granulation, that peasants across Europe.

13 Minute Read

By Charles Lewton-BrainMore from this author

Instead of granulation I personally strongly suggest using poppy seeds or small polyethylene balls (available in Germany in graduated sizes), supergluing them onto a wax and then casting your object…however:

Note that while there are many 'secret' recipes and granulation was 'lost' this was as far as I can tell a result of gentleman scholars and acadamics in the late 19th and early 20th century who were too snobbish and ethnocentric to realize that much of the world still does granulation, that peasants across Europe (where these very academics were writing and publishing) continued to do it (Norway, Central Europe) but they didn't look at them, instead choosing to believe in its loss. This resulted in the 1920's with a number of people experiementing with recipes, some being totally secret (a well known metalsmith whose work I admire a lot has still to my knowledge not been willing to share his 'secret' since the 1960's).

Another point, granulation in written theory often is deemed to work because a copper containing salt is used as an addition agent to the sticky material that is used to position the balls. The copper is then supposed to alloy with the contact point of the ball tangent to the metal thus momentarily lowering its melting temperature (sometimes called 'eutectic soldering in books) which is what bonds them. Scanning Electron Microscope analyses of ancient granulations show however no difference in alloy between ball, contact point and surface metal.

While it is conceivable that this is due to atomic diffusion over time in ancient artifacts I think it backs up a theory which follows-that the copper additions have nothing to do with forming a eutectic alloy but are instead ways of 'dirtying' the metal surface and changing it's surface tension so that when you have the metal up near its fusing point (because fine silver has no melt interval and just goes wet and melts this may make fine silver work a little tricker than using sterling) the balls are held together in spheres by surface tension until their contact points fuse onto the surface below. While the following extract from 'Theorie und Praxis des Goldschmiedes' downplays gold carburization as a granulation method (heat gold granules in charcoal a long time, they go black, granulate with them-I've seen this done) I think that approach too backs up this theory.

Cutting metal: roll the sheet out thin-this give control of volume, cut crosswise with snips to make 'pallions' like are used for solder. Another way is to make jump rings of thin wire. It should be noted that making jump rings is the only way to accurately guarentee uniform volumes to the balls.

Making balls: I have used the traditional approach of layering charcoal powder and bits of metal in and iron box or a crucible and heating in a kiln. If you get too hot the granules burn their way down and form a pool. Oops. If you do it right then you gently run water through it when done which washes out the charcoal dust leaving the balls behind and then one dries and sorts the balls by size (using stacked mesh sifters or stacked containers with graduated holes getting smaller in each lower container).

John Cogswell came up with a great way of making granules or little balls using a charcoal block. He was using them for granulation but there are lots of uses for little spheres. One makes a metal frame for the block as shown in the diagram. The block is tilted on a prop of some kind to about 30 degrees and a small pan of water is placed next to it. When one heats up small chips of metal as soon as they become round they roll down the surface of the block, maintaining their shape and then landing in the water. The frame guides them off the block into the pan. I have drawn jump rings rather than chips of metal because jump rings are the only way to easily and accurately guarantee identical metal volumes when making small balls. This is a quick method for small quantities of spheres. Larger quantities are usually made by layering the metal pieces in charcoal powder and heating in a kiln for a time, later washing away the ash. Cogswell wrote his Masters thesis on sterling silver granulation and a small version of it can be found in McCreight's edited book 'Metals Technic".


Tim McCreight
Brynmorgen Press
33 Woodland Road
Cape Elizabeth, Maine,
04101, USA

Harold O'Connor did a wonderful series of pins using reticulated silver with 18k Gold granules bonded to them in relationship to the wrinkles and then the whole darkened, very nice.

What follows is an extract from one of the best books on goldsmithing I know that deals with granulation. It is roughly translated

Extract from a Section on Granulation from the book:
Theorie und Praxis des Goldschmieds, by Erhard Brepohl,
VEB FachBuchVerlag Leipzig, DDR, 1978 Pages 277-280,
(rough translation: Lewton-Brain 1989).

There is a new edition out of this book which skips the obligatory capitalist bashing required of East German books before the wall came down.

2.3.2. Granulation The nature of granulation

One still imagines behind the word something strange, mystical or at least full of secrets even though the principle of this technique is quite simple. "Spheres 'as small as dust' are to be fixed to a base plate. The technical difficulty comes from the fact that due to the smallness of the spheres the use of all standard soldering techniques is excluded.

After granulation was forgotten for a long time it became possible first in this century to revive it again . Some adept fellows took it upon themselves to solve this problem with weeks of long, tiring work. Even though every goldsmith has tried at some time to fuse some balls onto a sheet only a few have truly mastered the technique. Only is rare cases have the experts revealed anything of the secret. One can explain the silence in that they wanted to utilize the results of the lengthy experiments themselves. This is a view that springs from Capitalist competition but has nothing to do with socialist principles of work!

On the other side an important granulator says thus; 'If the granulation were such a simple thing that any dolt could get granules onto the sheet without any trouble the technique would be a senseless game!' This touches on the main problem with granulation, as even this technique is only a medium which the goldsmith can call on in order to realize his artistic ideas. Only he who feels artistically strong enough in order to fill the technique with their own lives should attempt to draw and paint with golden granules on a golden ground. The various methods of granulation

To get one's bearings the main methods that are given in the literature are assembled here. This critical compilation should help the learner to search for his methods which despite all the published details on this areas this does not save him from many experiments which he must then translate into workshop practice; and finally it all depends upon the skill of the goldsmith! The most complete publication on granulation comes from Marc Rosenberg1, who described a procedure that was used by F. Stanger:

Small pieces similar to solder pallions are cut from an alloy of high gold content. One mixes them in a small crucible with charcoal dust. The crucible is heated long enough that all the pallions draw up to form balls. Then one heats them at red heat for some time in the charcoal dust where the temperature should lie some degrees under the Solidus point. Allegedly the surface of the balls is enriched with carbon, gold carbide is formed which should reduce the melting point of the surface of the balls to some 900oC. One does not need any flux but instead glues the granules on to the plate with some spit.

According to Littledale2 the granules are made in a similar way. One needs a copper hydroxide solution which serves as a 'chemical solder'. When any kind of copper salt solution reacts with sodium hydroxide a bright blue precipitate of copper hydroxide (Cu(OH)2 is formed which one filters out and while it is still damp is mixed with gum Arabic, gum tragacanth or clay in the proportion of 1:1 and then diluted with a little water (Note=an alternative is Klyrfire® thgat is used for enamelling). One sets up the granules using this solution whereby it is important to ensure that the solution does not run next to and around the granules but instead sits directly under them. Therefore one paints the balls onto the ground with it like ink. At red heat the glue breaks down to form carbon, the copper hydroxide into copper oxide. The oxide is reduced to metallic copper at 850oC, it alloys with the gold and exploiting the minimum in the gold-copper system an easy flowing alloy, the 'chemical solder' makes the join possible between granule and base plate. It is again otherwise according to E. Frey3: According to this granules and base metal are alloyed with some excess of copper. the granules are produced in charcoal dust but one then anneals the base plate and the granules thoroughly to obtain a thick oxide layer on them. The granules are set out using some Fluoron. At red heat the flux reduces the oxide portion. So one obtains as with the procedure from Littledale the easy flowing gold copper alloy at the boundary surfaces of the parts being joined which makes possible the union. In all these procedures the effect may be traced back to a change of the melting point. One article4 goes even so far as to claim that the melt interval in granulation alloys should be at least 30oC, if possible 60-80oC. In contradiction to this is the fact that the first successful granulation work was carried out with fine gold which can have absolutely no melt interval while yet Au 333 with it's great melt interval is unsuitable! Littledale and Frey want only to lower the melting point of the surface of the granule in that a low melting gold-copper alloy forms on it while the core should stay solid.

  1. Rosenberg, M: Geschichte der Goldschmiedekunst, Bd. 2: Granulation. Frankfurt (Main): Verlag Keller 1918
  2. British Patent Nr. 415181/1934
  3. Treskow, E.: Geschicte, Kunst and Technik der Granulation, Diebeners Goldschmiede-Jahrbuch 1959, Page 30, Stuttgart: Verlag Diebener 1959
    4 Deutsche Goldschmiedezeitung (1951), Number. 6, Page 139)

If however one truly only needs a copper rich surface one could cover the granules and metal sheet with copper much more easily.

Finally the view of Stanger may be contradicted as the gold has almost no ability to dissolve carbon and that small amount is given up again upon cooling. The gold carbide compound can only be made with difficulty; it is unlikely that it can be formed by simply heating in charcoal dust!

One believes that the granules let themselves be welded down undamaged because one has produced a lower melting surface layer on the granules which melts before its core does so that the solid core prevents the melting of the granule. With granulation it is not so much that there is a change in the thermal characteristics but instead one must be much more concerned to increase the natural surface tension (see section and to reduce the wetting ability of the material. One requires the opposite from a solder that must spread out well on the soldered surface; a low surface tension and good wettability. Every goldsmith knows from their own experience that even a solder that is distinguished by great wettability remains lying as a molten ball upon the soldering place when the flux does not prevent surface oxidation! Here lies also the explanation of the methods of Littledale and Frey, whose copper oxide layer actually only increased the natural surface tension. Stanger's carbon covering is then effective only as a crust which holds the molten granule together and prevents it's flowing.

Even when one were to assume that the outer region of the granule melts for example 50oC below the core to it pleases the practitioner to ask himself how long it takes before the core too of this spotless granule has reached the melting temperature! According to this he would have to have the entire workpiece just so hot that everywhere only the surface of the granules melts but the cores do not reach melting temperature! In fact it can only be that the entire granule melts; one must however be concerned to find ways through special treatment of the surface of the granule, by alloy additions and other procedures to hold the molten granule together without the possibility of affecting the weld at the contact point between granule and recipient surface.

The procedure published by Prof. Ungerer1 seems to be particularly useful and will be enlarged upon by further advice to obtain a basic procedure:

the granules and recipient metal have the same fine content, Au 750 or higher. For making the granules one rolls out a sheet as thin as possible. One can cut small pallions from is as when making preparing solder. So as to obtain equally sized parts the little pieces can also be stamped out using a blanking punch. One moistens the pallions and mixes them with a miniature whisk into the charcoal powder this fills up a small crucible. Every pallion is covered with charcoal dust through this and can no longer melt itself together with other pallions. So, as was described in the procedure from Stanger they are melted and heated for a time at red heat in the charcoal dust. One empties the contents of the crucible out into a porcelain dish so as to wash out the charcoal dust. The granules are dried but they remain blackened. One sorts the granules according to size in a granule sieve. (The sieve one makes oneself in which one prepares several sieve inserts which can be stacked on on top of the other. They are so drilled through that the hole size is reduced from one sieve to the next. The granules can then be separated into the inserts according to their size.)

One sets the granules onto the base plate whose thickness is approximately the same as the diameter of the granules (figure 238).using diluted Fluoron.

  • 1 Ungerer: Die Technik des Granulierens.- Zeitung für Goldschmiede (1943), Number 1, page 2.)

One places the prepared plate onto the soldering support already discussed several times (see figure 110). One brings it up finally to the required heat using uniform heating whereby the heat should particularly reach the underside of the plate. The granules and surface of the plate have to be equally shiny, one holds this state for a moment so that the melting metal parts at the contact point between granule and recipient metal can weld.

Figure 238 Attaching granules to the recipient metal

a) correct: granules of the same thickness are fused to each other as much as to the recipient base plate.
b) wrong: the granules are melted too much and the spaces between them are too small.
c) wrong: the granules are of different sizes and have varying distances between them

Therefore watching for the correct moment in time is the true secret of granulation, only through long experience will one obtain the right feel for it. If one heats too lone the granules run or sink too deep into the base sheet; if one does not heat long enough the don not join well enough and fall off. Prof. Ungerer says legitimately at the end of his paper: 'The control of this beautiful technique must be won with difficulty, it does not fall into anyone's lap without effort.'. Figure 239 shows a neck piece whose granulated gold decoration contrasts with the blackened silver background.

By Charles Lewton-Brain – Copyright © Charles Lewton-Brain 1997
All rights reserved internationally. Copyright © Charles Lewton-Brain. Users have permission to download the information and share it as long as no money is made. No commercial use of this information is allowed without permission in writing from Charles Lewton-Brain.

You assume all responsibility and risk for the use of the safety resources available on or through this web page. The International Gem Society LLC does not assume any liability for the materials, information and opinions provided on, or available through, this web page. No advice or information provided by this website shall create any warranty. Reliance on such advice, information or the content of this web page is solely at your own risk, including without limitation any safety guidelines, resources or precautions, or any other information related to safety that may be available on or through this web page. The International Gem Society LLC disclaims any liability for injury, death or damages resulting from the use thereof.

Charles Lewton-Brain

Master goldsmith Charles Lewton-Brain trained, studied and worked in Germany, Canada and the United States to learn the skills he uses. Charles Lewton-Brain is one of the original creators of Ganoksin.

The All-In-One Jewelry Making Solution At Your Fingertips

When you join the Ganoksin community, you get the tools you need to take your work to the next level.

Become a Member

Trusted Jewelry Making Information & Techniques

Sign up to receive the latest articles, techniques, and inspirations with our free newsletter.