The growing interest in and demand for platinum jewelry can be viewed as a renaissance of the 1920's for the metal in the United States. The jewelry industry uses four main classes of platinum alloys.

This paper will focus on the Pt 950 class of alloys - alloys with a minimum of 95% weight platinum. This class is used due to the fact that pure platinum is soft and does not have the mechanical strength for most jewelry applications. The mechanical strength of platinum is increased by alloying a second element with the platinum. Different alloying elements have varying degrees of effect on the mechanical and chemical properties of platinum. Palladium and rhodium have the least effect, while iridium, copper and gold have a stronger effect. Ruthenium and nickel have the greatest effect in increasing the hardness of the resulting platinum alloys. The alloying of the elements with platinum is usually achieved in the liquid phase because the diffusion is 100 times greater in liquids than in solids.  Once a homogeneous mixture of the platinum alloy is achieved in the liquid state, there are three main types of mixtures that can result as the alloy is cooled to the solid state:

Complete solubility - the atoms of metal a and metal b form a random mixture that cannot separate (like water and wine.)

Partial solubility - the metal can absorb only a limited quantity of the atoms of metal b and, likewise, b absorbs only a limited quantity of a. These maximum quantities are called the limit of solubility (like sugar and water.)

Total insolubility - the two metals cannot be mixed. Even if they are miscible in the liquid state at a very high temperature, the atoms of each metal separate when the temperature decreases and they reach the solid ·state. The microstructure is a juxtaposition of pure a particles and pure b particles (like an emulsion of oil and water.)

Therefore, it is important to know the behavior of the mixture of the alloying elements with platinum.

Copper and cobalt both have complete solubility in platinum. A single, homogeneous alloy of the type platinum formed with either of these elements has more predictable and consistent properties than those that are multiphase.  These alloys can be easily rolled or fabricated to form the desired final shape and size. Changes occur in the solid state in these two-alloy systems. The solid state reactions are referred to as age hardening or ordering reactions. An understanding of these reactions and the temperatures at which they occur can aid the jeweler in producing a superior finished product. Annealing of the alloy should be done at a temperature above where the ordered phase is stable. Alloys should also be cooled rapidly through this temperature range. However, the final product can be made harder by heating within the temperature region where the ordering takes place.

There are four ferromagnetic elements that are common: iron, cobalt, nickel and gadolinium. Alloying with these elements can result in a magnetic material at least over part of the alloy range. When choosing an element for alloying with platinum, many properties must be taken into account. The reactivity of the element is very important. If the element is oxidizable, it can be a source of defects in the final product, not only during the melting but also during the casting or brazing of the final piece. Factors that must be kept in mind in choosing an alloying element are the recycling of the alloy, whether the scrap can be reused and the effect of contaminants, the availability of investments and crucibles that will not react with the components, and so on. One very important property for jewelry is its castability, which is the key to good quality for lost-wax pieces. The elements used to improve the castability of karat gold have boiling points far too low, compared with the casting temperature. They will evaporate and will not be effective.

Each system must be tested and its flowability and castability measured. As a rule, it is not usually possible to find an alloy that can meet every requirement: color, hardness, castability, weldability, ductility, small grain size. The jeweler also wants to use the same alloy for cast and fabricated pieces to be sure they have the same color. When the properties of an alloy have been determined, the jeweler has to decide on the best compromise to satisfy the design and the functionality of the final piece. For instance, a ring or bracelet must be more wear-resistant, hence, have a greater hardness than a pendant or a necklace. One must also be aware of other steps in making the jewelry. If the joining of a platinum and gold piece is required, it is important to make sure that the brazing alloy does not contain elements which will embrittle the platinum.  Cadmium, indium, tin and zinc are elements used in brazing alloys for gold which will embrittle platinum.

Gold and platinum pieces that are to be joined should be finished separately. The temperatures used for annealing platinum are too high for gold and polishing platinum is more difficult than polishing gold. Gold must be polished first so it does not lose its color. The main problem when joining these two materials is the different coefficient of expansion. The difference in thermal expansion can result in cracking from the stresses generated. However, the cracking can be prevented by using the proper stress-relieve anneal. Carbon and silicon are two other contaminants to be avoided. However, silicon is often the basic component of the investment materials. Thus, the reuse of metal from the casting tree must be done with the knowledge that this is a potential source of casting defects which can lead to poor quality castings. There are many other sources of these elements in the workshop, for example, the dust in the air, the reducing flame of the torch, the polishing material, and the solder blocks. Each step of the process must be carefully examined and analyzed to avoid contamination.

Alloy usage is different all over the world depending on the local customs and the hallmark rules. In Japan, alloys are acceptable with 85% platinum.  In Europe, only 95% platinum (or greater) alloys are allowed. The first platinum alloys were made with the precious elements iridium and palladium, because of their oxidation resistance.

But the iridium alloys have a high melting point and Pd, as has been indicated, has a small effect on the hardness, so non-precious metals were tested The traditional alloy in Europe is 950Pt-50Cu. This platinum-copper alloy (Pt-Cu) has poor casting qualities and an undesirable color. To overcome these deficiencies, a Pt-Co alloy was developed. However, this alloy is weakly magnetic. So, when sorting minute pieces of scrap by magnetic separation, a large amount of the platinum alloy may be lost.

The remainder of this paper will focus on a new alloy that has been developed in order to retain the good properties of both the Pt-Cu and Pt-Co systems and yet overcome their deficiencies. The new alloy is a 950Pt-Cu-Co alloy that is non-magnetic.

The results of detailed experimental procedures conclude that the Pt-Cu-Co alloy has sufficient strength and ductility for most jewelry applications, whether cast or fabricated.

The microstructure of the new alloy appears to be more sound, based on laboratory tests, than the Pt-Ru. The quality is nearly the same as Pt-Co but the new alloy is not magnetic, simplifying scrap recovery.

The color measurements show that the four alloys, Pt-Cu, Pt-Co, Pt-Ru, and Pt-Cu-Co, have nearly the same color. The naked eye would have a difficult time differentiating the color of these four alloys.  The initial data generated in the laboratory indicates the new alloy has promising properties. Costantino Volpe's report on additional casting and jewelry trials that were done at Tiffany to further evaluate the Pt-Cu-Co alloy versus Pt-Ru, is summarized below.  

Comparison of Pt-5Ru and Pt-Co-Cu Platinum Casting Alloys, Part II
In the world of metal and alloy production, platinum is not necessarily a difficult metal to work with. It is fairly soft, yet holds a nice polish. It is one of the few elements that has no oxidation properties, so in fact it is a pleasant metal to cast in that respect, no cover gas or vacuum is needed. Although it has a high melting point, induction machines today have more than enough power to melt it.

The problems that arise when casting platinum jewelry are the result of restrictions that are inherent to the industry. In other industries when alloys are designed for various usage, there is usually more freedom in terms of the alloying elements added and to what degree. This allows an alloy to be designed that has all the necessary properties, including good manufacturing properties. However, in our industry, we are constrained by the fact that our platinum alloy must be at least 95% platinum. This allows us only 5% to work with in terms of adding other elements to give us desired properties. 950 platinum has a very narrow gap between the solidus and liquidus and this causes very quick freezing. This fact forces us to go to high superheats and casting speeds to get the metal into the mold and achieve good filling. Unfortunately these two acts will contribute to shrinkage porosity. The other constraint that we place on ourselves is that of design. Jewelry is designed to look good, not necessarily for ease of manufacture.  Therefore, what property can we change that will make casting these 950 alloys easier?

Pure platinum and platinum-ruthenium alloys tend to have a fairly high surface tension. Metals with high surface tension require more force and time to get into cavities. We can observe this phenomenon very easily in gold casting where the addition of small amounts of zinc produce very good improvements in castability due to the fact that zinc reduces the surface tension of the alloy. A lower surface tension allows the metal to flow against mold cavity walls with less friction, meaning the flow will be quicker and the cavity will fill with less force. The result is we don't have to apply as much superheat or throw it as hard. The tradeoff here is that better wetting of the mold could result in quicker heat loss and without proper mold temperature we can negate the effects of the lower surface tension.  Theoretically, if we can find an element or combination of elements that will reduce the surface tension, then we can increase the castability of the alloy.

Unfortunately, surface tension measurements are very difficult to perform, especially on high-melting metals. One can try to melt a drop of platinum on a flat surface and then look at its cross-sectional profile, but even this is difficult and not exact because one actually wants to observe the profile in the liquid state. It is therefore easier to measure it qualitatively. This is what we have done by way of these casting studies.  Under identical conditions we can compare how various shapes fill for different alloys. It is important to have as few variables as possible. This means that the pieces have to be placed on the button in exactly the same manner, the flasks should be invested with the same batch of investment, burnout should be in the same oven and cast at the same temperature and speed. Casting at the same temperature and speed can only be accomplished using a modern casting machine with a pyrometer. Trying to judge the temperature by eye can be very difficult, especially when using a new alloy, which may have a different look when molten. The only way to tell with any kind of accuracy is to use an instrument such as an optical pyrometer.

Experimental Procedure
Our casting studies were done in commercially available centrifugal induction casting machines, incorporating typical production processes for platinum alloys. Side-by-side comparisons were done between two alloys, Pt-Ru and a new alloy, Pt-Co-Cu. The melt characteristics of the two alloys were very similar except that the Pt-Co-Cu alloy produced more smoke than the Pt-Ru. The castings were then examined for fill and other general characteristics such as surface roughness, color and porosity. Simple bend tests were also conducted to determine ductility.

Results

Filling
All grids filled, which was unexpected. The Pt-Ru shows flow lines on the surface at lower acceleration, an indication that the metal was at the freezing point as it made contact with the mold wall. The Pt-Co-Cu did not show this characteristic. In coil tests the Pt-Co-Cu alloy showed more filling than the Pt-Ru. The test was repeated and in both cases the coil showed greater filling with the Pt-Co-Cu alloy.

When the gauge size was increased, both alloys filled the coil completely. This could lead us to conclude that the thinner and finer the piece, the more advantageous the Pt-Co-Cu alloy will be.

Surface Quality
Both plates tested filled well. The Pt-Co-Cu alloy showed a very slightly rougher surface than the Pt-Ru. This was the case with all samples cast, regardless of size and shape. The slightly rougher surface was probably due to oxide formation.

Ductility
All rods tested filled completely. Both rod sizes were placed in a vise and bent 90° and back with pliers. The Pt-Ru alloy broke after only two or three bends while the Pt-Co-Cu alloy survived up to 10 bends before breaking. The new alloy showed much more ductility near the fracture area.

The test was repeated several times, since any slight porosity in the rod would greatly influence the results. Repeating the tests produced the same results. Qualitatively, we can say that the Pt-Co-Cu is more ductile than the Pt-Ru.

Porosity
.25-inch-diameter cast rods were ground down halfway along their length in an effort to measure the degree of porosity throughout the piece. The Pt-Co-Cu showed porosity around the sprue junction, while the Pt-Ru alloy showed a more random distribution throughout the piece.

Color
To determine any color difference, rings were cast from the two alloys and polished. Placed side-by-side the two alloys were virtually indistinguishable from each other. Polishers reported that polishing of the Pt-Co-Cu rings took less effort. The tests were done without identifying which set of rings was a different alloy to try to keep the tests objective.

Joining
The rings were cut and then welded together by traditional torch methods. Rings were also laser-welded. The results of the welding are as follows: Pt-Ru rings soldered without any oxidation. The Pt-Co-Cu rings actually started to melt before the weld filler material started to flow completely. The filler material was a thin sheet of platinum placed in the joint. The rings also showed a slight discoloration. After running one hour in a magnetic tumbler, there was still a slight stain evident that had to be hand polished off. Laser welding produced no noticeable surface oxidation with either alloy.

Conclusions
After casting about 50 items of various shapes and sizes from rod to plate to cylinders, the general conclusion seems to be that the Pt-Co-Cu alloy would be easier to cast and process into jewelry. The only serious drawback with this alloy is its tendency to form a surface oxide during welding or soldering operations. If an efficient means of preventing or removing this oxide was made available, then one could conclude that this alloy would be an improvement over the Pt-Ru alloy currently in widespread use.

V3N1

Platinum Alloy Characteristics:
A Comparison of Existing Platinum Casting Alloys with Pt-Cu-Co
Dr. Richard Lanam and Florence Pozarnik Engelhard-CLAL
Costantino Volpe Tiffany & Co

This is an abbreviated version of the original work. For full technical details, please consult the original paper.