A BEGINNER’S LOOK AT ENAMELLING ON SILVER

ALLOY

Common Silver Alloys for Enamelling

Sterling silver (7.5% copper in a minimum of 92.5% silver) is widely used for silver
articles including jewellery and for enamelling the same. The copper content increases
the strength and hardness.

Britannia silver (a minimum silver content of 95.8% and the remainder largely
copper) is less widely used but is especially suitable for taking stamps and dies. Being
softer than Sterling silver, it is more susceptible to bowing unless counter enamelled.
However, some experts state that enamel colours are truer when fired on this alloy. As
to cost, there is virtually no difference.

It is reported that the presence of impurities even at low levels can affect the adhesion
of enamels. In particular, Selenium and Teluriam can cause problems. To avoid this
potential problem, reputable silver suppliers use only fine silver and pure copper in
the initial alloy casting process.

Temperature Characteristics of Silver and Common Alloys

Type Melting Point Solidus Point

0 0
C C
Fine Silver 962 962
Britannia Silver 930 870
Sterling Silver 910 800

Table 1. Temperature Characteristic of Silver Alloys for Enamelling

Although the melting point of fine silver is precise at 961.930C, the other values in the
above table are probably only accurate to  50C depending on the exact alloy
composition and. manufacturing process.

If a molten mass on the kiln floor is to be avoided, the combination of kiln

temperature and firing time must be such that the silver nowhere approaches the
melting point.

Firing Temperatures for Sterling and Britannia Silver

That one should not fire pure silver or its alloys to melting temperature is obvious, but
there is a further consideration that is particular to the alloys. This concerns the
solidus temperature (see third column in Table 1), an aspect clearly and carefully
explained in the Thompson Enamel Workbook.

In the annealed condition, both Sterling and Britannia silver contain a mix of two
discrete constituents: one is composed of pure silver crystals, and the other silver-
copper crystals. At room temperature, Britannia silver has around 85% of its mass as
pure silver crystals and the remaining 15% are silver-copper crystals. For Sterling
silver, 73% is pure silver, and 27% silver-copper.

1
When these silver alloys are heated and reach the solidus temperature, the silver-
copper crystals become liquid. If the temperature then exceeds the solidus, this
eutectic is able to absorb more of the pure silver crystals. The higher the temperature,
the more silver becomes integrated into the silver-copper eutectic. Finally at the
melting point of the alloy, all the silver is in solution with the copper.

Thus, unlike pure silver, alloys of silver-copper change from solid to liquid over a
temperature band. They are essentially solid below the solidus point shown in Table
1, are fully molten at the melting point, and in between are in a transition or “mushy”
state.

The region above the solidus is clearly one of metallurgical change, and no doubt this
explains why Thompson Enamel state that silver alloys should not be allowed to
exceed the solidus temperature. Skilled enamellers recognise that the change of
crystal structure is gradual, and they have the experience to know just how high they
can fire a particular silver alloy. Beginners, however, should limit the alloy
temperatures to no more than those shown in the third column of Table 1.

Thompson Enamel, in fact, recommend Sterling silver be taken no higher than 7900C
to allow for inaccuracies of pyrometers and other variables in firing. As a
consequence, attention has to be paid to some enamels that are hard firing, as they
may not reach maturity easily. Presumably, applying the same criterion to Britannia
silver means one should limit the temperature of the metal to 8600C.

Although the melting temperatures of Britannia silver and Sterling silver are quite
close, the difference in solidus temperature is considerable. Perhaps it is
understandable why Britannia silver holds an attraction for the relatively unskilled,
because the safe working temperature is considerably higher.

Annealing
When enamelling on pure copper, it is common practice to anneal the metal first to
relieve stresses and any work hardening and to burn off surface products such as oils
and greases. This is done by heating it to about two thirds of its melting temperature
and then, after slight air- cooling, it is quenched in water. Apparently, allowing any
metal to air-cool will not entirely soften it. An undesirable consequence of the
annealing is that the surface is converted to black cupric oxide, CuO, and below it, red
cuprous oxide, Cu2O. The quenching can cause some of the oxide to come away and
pickling in an appropriate acid will finally removes these oxides.

In contrast, a silver alloy is normally supplied in an annealed condition using a

protective inert atmosphere to exclude any oxidation products. Only subsequent work
hardening by the user, which alters the size and boundary composition of the silver
and silver/copper crystals, would dictate the need to restore the alloy to its annealed
state.

Silver alloys anneal rapidly and one authority states that 6500C is a preferred upper
limit. Although silver oxide forms on the surface when heated, it decomposes above
about 2000C. However, the silver-copper crystals in the alloy mix will oxidise to
produce copper oxide at the surface. Because of the low percentage of copper in the
alloy, the effect is not as extreme or as visible as when annealing pure copper.

2
Nevertheless, a dark grey staining will usually appear and this is commonly called
firestain. It can be avoided by prior coating of the work piece with a proprietary
fluxing compound that restricts the access of oxygen to the metal surface.
Alternatively, the firestain can be removed by pickling after the annealing process.

Further Treatment of Silver Alloys Prior to Enamelling

Again, when enamelling on copper, most transparent enamels benefit from being fired
over a layer of flux, since cuprous oxide will form under an applied coloured
transparent enamel, and will invariably muddy it. The flux layer, if fired correctly,
will absorb the oxide, seal the surface against further oxidation, and leave a clear base
for laying the coloured enamels.

In the case of silver alloys, the amount of copper oxide that will be produced on the
surface, upon heating, is small. Indeed, many transparent enamels can be applied
directly to Britannia and Sterling silver without being degraded. Red transparent
enamels and some rose and pinks apparently are in a special category that can require
the use of a flux layer– more about that later.

Even so, it appears to be good practice to remove, or significantly reduce, the copper
content that resides in the silver-copper crystals at the surface and thereby the copper
oxide. The aim is to produce a surface that is pure silver. This, presumably, is felt to
be a better solution than the use of a pre coat of flux since the final result is never
quite so clear or brilliant as a direct application of transparent enamel on silver. In any
event, why apply a flux coat if it can be avoided in some way.

The literature mentions two main methods of creating a pure silver layer and there
seem to be advocates of both methods. Thus the beginner is faced with a choice. A
reading of the literature has produced the following summary of both methods.

(a) Repeated Heating and Pickling

As mentioned earlier, if a silver alloy piece has had to be annealed by the user
because of work hardening, the surface layer of silver-copper crystals will already
have become oxidised, unless a fluxing compound has been used. This also applies
when findings have been soldered to the work piece. The surface oxide can be
removed by pickling. An effective pickle is dilute sulphuric acid or a derivative of it.
Hot, dilute sulphuric acid is quicker and safer than the concentrated acid in dissolving
copper oxides and it does not attack the base metal underneath. If this heating and
pickling process is then repeated a few times, it will cause a reduction each time in
any remaining surface copper. What remains is a surface that is pure silver and no
copper, a condition that would appear to allow maximum clarity to transparent
enamels applied without a pre-layer of flux. Some authorities recommend repeating
the process up to ten times to be certain the surface copper has been removed by this
process of oxidation and pickling. This, then, is one treatment favoured by many in
the preparation of silver alloys for enamelling. After washing, the silver alloy has a
dull, matt surface that needs brightening in some way. Alternatives mentioned in the
literature include the sparing use of a soft brass brush or a glass fibre brush, both
being used with plenty of water and detergent.

3
 (b) Nitric Acid Treatment
Another technique is by immersion in concentrated nitric acid. There are three hurdles
for the novice to clear: - acquiring the acid, storing it, and using it. One supplier is
Walsh’s who list 70% concentrated nitric acid in three litre containers. European
Directives prohibit sending this through the normal channels and so it must be
collected. It should be kept in a locked cabinet in accordance with the COSH
regulations. When using it, the recommended practice is to wear protective gloves and
apron and to work in a well-ventilated area, as the fumes are noxious.

It is stated that when a silver-copper alloy is briefly dipped in the acid, the metal turns
grey or black and slowly fades to light grey/white. The piece is removed from the acid
and dipped into hot water to remove the water-soluble by-products of the reaction.
The process is repeated until dipping in the acid results in no further blackening.
Concentrated nitric acid has a great affinity for attacking copper and so one presumes
the action is one of removing any surface copper leaving pure silver. Dilute nitric acid
is not so effective, as it will attack the silver in the alloy much faster than it attacks the
copper. Nevertheless, it is stressed that the piece should not be left too long in the
concentrated acid since the whole silver alloy will be progressively etched away.
Needless to say, the piece must then be washed thoroughly to remove all traces of the
acid. One method is to first neutralise the acid with a solution of soda crystals (sodium
carbonate), followed by rinsing in water. As with the previously described Heating
and Pickling technique, the silver alloy now has a matt appearance that can be
brightened before enamelling.

So, in respect of creating a surface that is pure silver, both the Heating and Pickle
method, and the Nitric Acid treatment appear, at first sight, to achieve the same result.
Some authorities even suggest doing both i.e. heat and pickle several times and then
dip in the nitric acid.

The respective actions of the two acids can be seen in the following table. The
definitive work is by Mr Woodrow Carpenter in the very first issue of “Glass on
Metal”.

Metal/Oxide Hot dilute Cold concentrated

Sulphuric acid Nitric Acid
Cupric oxide CuO 32 25
Cuprous oxide Cu2O 4 3
Copper 0 160
Silver 0 1

Table 2. Relative Solubilities of copper oxides, copper, and silver in acids

The solubility of silver in cold concentrated nitric acid is taken as the reference point.
Thus, this same acid will attack copper some 160 times faster. It is clear that the hot
dilute sulphuric acid treatment will remove the surface oxides only and hence it is a
safe, time independent treatment. Equally this table shows the nitric acid treatment
will eventually attack all elements making up the silver alloy.

4
Some Initial Tests Using the Nitric Acid Treatment
Having purchased some annealed Britannia silver sheet (0.9mm thick), I made some
test samples, 10mm wide and 35mm long. I immersed the samples in the concentrated
nitric acid to observe the result. I found it convenient to place the test sample in a
plastic sieve with a handle so that it could be immersed easily in the acid. I also found
it effective to brush the silver alloy surface with a feather when it was in the acid,
rather than swilling the acid over it. This technique results in deposits collecting in the
acid. They seem to do no harm, but ultimately the acid will become exhausted and
need replacing. I then lifted the sieve out of the acid and transferred the sample to a
hot water rinse.

To my surprise, when placed in the acid, my samples only turned a light straw colour
and after subsequent washing in the water, the surface was reasonably white. The
absence of any blackening was puzzling. After delving into the literature again, I
discovered a fact about silver and its alloys that seemed to provide an explanation.

When silver or its alloys are heated in air, silver oxide forms on the surface. Above
about 2000 C, this oxide decomposes. There is, however, the potential for oxidation
below the surface. This is because silver is porous to oxygen when it is heated, and so
the underlying material will also oxidise. In the case of pure silver, silver oxide forms,
although it is not clear how much, if any of it, decomposes when it is below the
surface. In the case of the silver alloys, both silver oxide and cuprous oxide will be
produced. Certainly the latter does not decompose at temperature. It is reported that
the oxygen can penetrate to a depth in excess of 0.02mm if a sample is held at
annealing temperature for an excessive time (say one hour). If such a sample were
then to be pickled in sulphuric acid, any surface oxidation that might be present is
removed, but not oxides below the surface. However, subsequent dipping in
concentrated nitric acid, will eat away the surface, exposing any sub-surface oxides
which themselves will be dissolved away. The blackening indicates the presence of
copper oxide (Cu2O).
Despite much searching of the literature, I have not been able to determine the
precise nature and extent of these oxides that form below the surface when silver or
its alloys are heated, or the chemical reaction of the nitric acid upon them.

Nevertheless, from the above observations, it would seem that the Britannia silver I
purchased was supplied in an annealed condition, using a protective inert atmosphere
to avoid any possible sub-surface oxidation products. The nitric acid is attacking the
metal alloy but there are no oxides to react with and cause blackening.

To investigate this a little further, I took another of my Britannia silver test samples,
heated it in air to 6500 C for 90 seconds, pickled it in sulphuric acid to remove any
surface oxides, washed it and then dipped it in the nitric acid. It immediately turned
black but cleared quite quickly. I measured the amount of silver alloy that had been
etched away by the nitric acid – 0.005mm. To confirm that the oxidation is
progressive with time and temperature, I then repeated this experiment with two more
samples. One I heated for 15 minutes at 7000 C, and the other for two minutes at 8000
C. Both turned black in the nitric acid and took many tens of seconds to clear with
constant brushing with the feather. In the former case, 0.02mm had been removed by
the acid and in the latter case 0.015mm.

5
This all seems to indicate that, while both the Heating/Sulphuric acid Pickle method
and the Nitric Acid treatment convert the surface layer of a silver alloy to pure silver,
only the latter treatment will dispose of sub-surface oxidation products. From an
enamelling point of view, one may ask is this important? The only way to find out is
to try both methods and compare the results.

Enamels Tested Using the Alternative Preparation Methods

To gain an insight into any major differences between preparing a silver alloy by
heating and pickling or by the nitric acid treatment, I selected two transparent enamels
with markedly different characteristics. One, a hard firing green, (Soyer 50), that, it is
said, can be laid directly on silver alloy, and the other a medium firing red, (Soyer 8)
that requires a flux undercoat. According to Thomson Enamels, any silver oxide that
forms on heating silver or silver alloys, and which does not decompose, will react
with most transparent reds causing them to turn orange or brown. As a consequence, a
prior flux coat is needed that is not discoloured by silver alloy oxides. It should then
provide a suitable base for the reds. Thus simply removing the copper content from a
silver alloy is not sufficient for the direct laying of red enamels.

There are many such fluxes that are formulated for use on silver. In particular,
Thomson advocate the use of their 2020 as it does not “yellow” on silver if a
sufficient amount has been applied. Another flux, Schauer 2A (ST17), is also
mentioned frequently in the literature as being an effective base for red transparent
enamels. I note, in more recent publications, that Ninomiya N1 flux is an alternative,
and then there is Soyer3 flux for silver and Blythe medium flux C1.

So, I firstly took some of my 35mm long Britannia silver test pieces and scribed them
into three sections. Half the samples were then prepared using the heat and pickle
method (see below), and the remainder were prepared using the nitric acid treatment.
Following this, the right hand end was wet laid with the green enamel, and the
remaining two thirds were laid with Schauer 2A flux. After firing two coats of green
on this right hand third and two coats of flux on the remainder, I then wet laid the left
hand end third with the red enamel and fired the test piece again. A further coat of red
was then laid on this third and fired. Thus, I could observe the performance of the
green enamel on bare silver alloy, the flux, and the red over flux all on one test piece.

As to firing times and temperatures for the enamels, I was conscious of the
information about the solidus temperature and I decided to set the kiln temperature
well below it, (7600 C), and to fire for 1¾ minutes. I also did some tests at a higher
kiln temperature, nearer to the solidus temperature for Britannia silver (860 0 C for 1¾
minutes).

Results Using the Heating and Pickle Preparation

The heating of these samples was done at 6500 C for 90 seconds followed by pickling
in hot sodium bisulphate (pool acid). After each pickle, the surface was lightly
abraded with a soft brass brush using plenty of water and detergent. After the first
anneal, the surface had turned grey (firestain). At least two further anneal/pickle
operations were done and the surface was now matt, with hardly a hint of grey. I
presumed the surface was now sensibly free of copper.

6
 The following table summarises the test conditions and the final appearance of the
enamels.

No. of Kiln Temp. Firing Time Final Appearance of Enamels

0
Sample Anneals C Sec
No.
1 3 760 105 Green is slightly opaque, Flux is slightly yellow, Red
on flux is reddish-brown
2 3 860 105 Green is slightly opaque, Flux is almost clear, Red
on flux is reddish-brown
3 10 760 105 Green is slightly opaque, Flux is yellow-grey, Red
on flux is brown
4 10 860 105 Green is rather opaque, Flux is dark brown, Red on
flux is muddy brown
5 10* 760 105 Green is very slightly opaque, Flux is slightly
yellow, Red on flux is reddish-brown
* This sample was not abraded at any stage of the anneal and pickle process

Table 3. Enamel Tests on Britannia Silver Prepared by Heating and Pickling

Across the range of test samples, the green Soyer 50 enamel was acceptable but not
very transparent. Its opaqueness was apparent from the first firing onwards. It was
slightly clearer (sample 5) when the surface was left untouched ie not abraded after
the pickle. This suggests that even slight abrasion damages the pure silver layer
exposing the copper. It is also possible that the removal of pickle deposits allow easier
access of oxygen.

The flux ranged from almost clear to dark brown and the red over the flux was
generally disappointing. The flux discolouration was also apparent from the first
firing onwards. As regards the red, the first coat was nearly always yellow. After
firing the second coat of red, the colour either became more reddish, (samples1,2,5) or
became brown (samples 3,4).

Increasing the number of heating/pickle operations from 3 to 10 was not beneficial.

The implication here is that there is more extensive oxidation below the surface and
that this does degrade the enamels. In respect of the green enamel, I suspect it is the
copper oxide component in the silver alloy. In the case of the flux and red enamel, it
may be a combination of both copper and silver oxide.

I found, from subsequent tests, that even very modest mechanical cleaning of the
surface does have a deleterious effect. However, as shown by test sample 5, even if
the top surface is undisturbed, the flux and red enamels appear to be affected by the
presence of sub-surface oxides.

Results Using the Nitric Acid Treatment

The samples were first cleaned with a soft brass brush and liquid detergent. They
were then immersed in the acid and brushed with a feather for about thirty seconds.
The samples were transferred to a warm water bath and subsequently washed well
using a detergent/ammonia wash. After this, the surface had a somewhat milky

7
 appearance, lacking the normal reflectivity of silver. Although the previous tests,
using the heating and pickling method, had shown that the enamel quality was
adversely affected by even light cleaning, I nevertheless decided to finally clean up
the surface by light abrasion with a soft brass brush to produce a silvery matt
appearance. I did this because this method of brightening the surface is widely
advocated in the literature. I was also interested in a direct comparison with the
heating and pickle samples. Table 4 summarises the results.

Sample Result of Kiln Temp. Firing Time Appearance of Enamels

0
No. Acid dips C Sec
6 Went 760 105 Green is transparent, Flux is clear, Red on flux is red
slightly except at edges where it is yellow
yellow
7 Went 860 105 Green is less transparent, Flux is clear, Red on flux is
slightly red but more yellowing at edges and enamel is
yellow pulling back from edge
8* Went 760 105 Green is transparent, Flux is clear, Red on flux is red
slightly except at edges where it is yellow
yellow
* This sample was not abraded

Table 4. Enamel Tests on Britannia Silver Prepared by Nitric Acid Treatment

These results show a significant improvement on the heated and pickled samples of
Table 3. The green is transparent, although perhaps lacking in sparkle. The brass
brush imparts a satin finish to the silver alloy and this may be the reason for the
subdued colour. The flux is clear and it was possible to achieve a reasonable red
colour. The yellowing at the edges is an indication of oxygen accessing the silver
surface at this point resulting in excessive silver oxide being taken into solution of the
flux and red enamel. Perhaps enamelling in a cloisson or on silver with a raised edge
would reduce this effect. Although abrading samples No 6 and No 7 may have
damaged the upper silver layer, there is no evidence of degradation of enamel colours
because of this. Perhaps this is because there was not the significant sub-surface
oxidation that results from prolonged prior heating that might otherwise interact with
and degrade the enamels.

Figure 1. Comparison of Sample 4 (top) and Sample 8 (bottom)

8
Figure 1 shows (within the limitations of photographic reproduction) the better result
obtained by nitric acid dip as compared to heating and pickling.

I cannot escape the conclusion from these tests that prolonged heating of a silver-
copper alloy in air prior to laying transparent enamels, and particularly red enamels, is
deleterious. The porosity of silver to oxygen at temperature creates oxide products
that degrade the enamels. The sulphuric acid pickle can only remove surface oxides.
If it is necessary to relieve work hardening by annealing, these oxide products can be
removed by a nitric acid treatment or the surface protected by a fluxing compound. As
a consequence of my test results, I decided to adopt the nitric acid treatment for the
pre-enamelling treatment of the silver/copper alloys.

Comparison of Some Fluxes for Use with Silver Sensitive Enamels

The red transparent enamel that I tested confirmed the need for a flux undercoat.To
compare the Schauer 2A flux tested with three others that I had to hand, I cut out a
piece of Britannia silver 40mm long by 30mm wide and scribed it into four equal
strips. It was given the nitric acid treatment as already described and finally gently
abraded with the soft brass brush using plenty of detergent and ammonia solution. The
fluxes tested were the Schauer 2A, Ninomiya N1, Soyer 3 and Latham T232. There
was no blackening of the silver during the acid treatment and this is consistent with
the previous findings.

All four fluxes were ground to 150/325, swilled with distilled water until clear, and
wet laid. The test piece was then fired at 7600C for two minutes. Although some straw
colour was evident with the 2A and N1 fluxes, after firing a second flux coat, all four
fluxes were clear. The 2A flux was quite gritty to wet lay and grainy when fired. I
subsequently purchased some fresh lump 2A, ground it and found that, after firing, the
grittiness and grainy appearance was no longer evident. Clearly my original powder
stock had degraded. I then laid a thin coat of the red Soyer 8 enamel over the bottom
half of the test piece and fired it as before.

With one coat of red enamel, all sections had some yellowing particularly at the
edges. I applied a further three thin coats, firing each coat, and the yellowing
progressively reduced. The final result on all fluxes, see Figure 2, was a reasonable
red, albeit with a slightly brownish tinge. From this test, I concluded that there was no
significant difference between the four fluxes when they were fired to a low
temperature.

Figure 2. Fluxes and Red Enamel on Britannia Silver

9
Further Considerations – Firing Rate
It seemed to me that any difficulties in achieving clear transparent enamels,
particularly the reds, is associated with oxides that develop on heating silver and its
alloys. The logical outcome is to minimise the extent of any such oxidation. The ideal
would be to lay down the enamels in such a way that the metal does not oxidise when
heated, or, if this is not possible, to at least restrict the amount of oxidation.

As regards the former possibility, an inert gas atmosphere, enamelling kiln would be
an attractive proposition. Unfortunately, I have not found a kiln of this type that
would be justified, economically. It remains a goal of mine to investigate this further.

Thinking then about restricting the surface and sub-surface oxides, it should be
beneficial to minimise the time that the metal is at temperature. Thus, if the enamel
powder can be brought to maturity in the shortest possible time, it will be least
affected by oxidation of the metal. In fact, I note that some experts consider that a
short, sharp firing results in brighter, clearer enamel colours. The effect may indeed
be associated with minimising the amount of metal or oxide absorbed by enamels. It
might also be an intrinsic property of all enamels that they fire better when brought to
maturity in the shortest possible time.

Whatever the reason, the advocates of high, rapid firing recommend setting the kiln to
around 10000C, and to remove the piece from the kiln when the enamel has reached
maturity or just beyond. If the piece is a small jewellery item, this strategy means it
must be watched closely while it is in the kiln and then quickly removed or else the
silver alloy, and particularly Sterling, will significantly exceed the solidus
temperature. For example, a particular small Sterling silver item with its stilt, placed
in a kiln at 9500C, took 47 seconds to reach 7900C. If this piece had been left in the
kiln for a further 10 seconds, I calculated it would then have reached 8500C.

My present strategy had been to set my kiln to around 7600 C and to allow my
Britannia silver samples to slowly reach maturity. Perhaps 1¾ minutes is rather a long
time, but at least the kiln temperature was such as to not allow the silver to exceed its
solidus temperature. Now, I am about to embark on the rather more adventurous, high
rate of fire approach. Regarding the enamel powder itself, I had already been using
150/325 grains and these lie closer together and reach maturity much quicker than
coarser grains.

With my small kiln, 9500 C was about the maximum temperature that I could achieve.
I calculated that my 10mm by 35 mm by 0.9mm thick Britannia silver alloy test piece
on its small trivet support would have a time constant of around 25 seconds. It is then
possible to construct two heating curves for this piece – see figure 3.

10
 1000
900
800
700

Degrees C
600
500
400
300
200
100
0
0 15 30 45 60 75 90 105 120 135 150
Time seconds

Figure 3. Heating Curves for 35x 10x 0.9mm Silver Strip

This shows heating curve B for the silver alloy when the kiln temperature is set at
8000C. This corresponds to the test results reported thus far in which the samples were
withdrawn from the kiln after 105 seconds. Heating curve A is with the kiln set to
9400C, and it can be seen that the high rate of firing means that such samples should
be removed after 45 seconds if they are not to exceed 8000C.

If there is a critical period, it is where the silver alloy starts to oxidise rapidly and
before the enamel fuses to shut off the oxygen supply to the surface. If we assume this
to be between about 2000C and 7500C, then it may be deduced from the figure that a
high rate of fire sample, (curve A), is exposed for around 33 seconds. A low rate of
fire sample, (curve B), is exposed for around 62 seconds – nearly 100% longer. I
needed to see whether this difference in exposure time is significant for enamel
clarity.

High Rate of Fire Tests

First of all, I did some sample enamel tests using Soyer 3 flux for silver to see if the
firing time of around 45 seconds, deduced from my high–rate heating curve (A), was
of the right order. Samples were put in the kiln, (set at 9400C), for 35, 40 and 45
seconds respectively. At 35 seconds, the enamel was just past the crystalline stage, at
40 seconds it was just past orange peel, and at 45 seconds it was smooth. This seemed
reasonable confirmation that a 40 to 45 second firing time was appropriate.

Accordingly, I made up some more samples, similar to those reported on in Table 4.

The surface of the Britannia silver was not abraded but cleaned with an ammonia and
detergent solution using a toothbrush. It was somewhat dull as a consequence. I used
the Soyer 3 flux as an undercoat for the red enamel, the kiln was set at 9400C and the
firing time was 45 seconds. After firing the enamels and upon examination, the flux
was clear and the green was a rich, brilliant colour and superior to the twenty or so
previous results. After firing the first coat of Soyer red over the flux, it was pinky red
with some yellow at the edges. After a second coat was fired, I had a good red and
only slightly orange at the edges. I subsequently repeated these tests using the Schauer
flux with comparable results.

11
 These very limited results do perhaps suggest that the high rate of firing produces a
superior result for the green laid directly on the silver alloy. The flux, when it was
fired quickly, was also very clear. Of course, with large pieces on corresponding
bulky stilts, the time taken to reach maturity will be much longer than that reported
here, but one can understand why it would still seem sensible to fire for the shortest
time that is possible. Despite this, as I am somewhat cautious in my approach, I
remain rather apprehensive about firing Sterling silver at a high rate for fear of
exceeding its rather low solidus temperature.

Surface Treatment after Acid Dip and prior to Laying Enamels

The action of the nitric acid on the Britannia silver results in a dull and sometimes
streaky finish that remained after neutralising the acid and washing the silver. If there
is no subsequent, prolonged heating of the silver alloy in air, then, using the brass
brush to brighten the surface prior to laying the enamels did not seem to affect the
final result. Nevertheless, I wondered if there was another way of brightening the
surface other than by mechanical means. If so, it would remove any concerns I might
have about over vigorous use of the brass brush in exposing the copper component of
the silver alloy.

I eventually settled on a cleaning agent for silver, (Hagerty Silver Foam). I applied
this using a moistened soft toothbrush and the resulting surface finish was very
similar to that achieved using the brass brush. Whether the action is entirely chemical,
or there is a very mild abrasive in the cleaning agent, is not clear. I then washed the
surface thoroughly as before. This is a method of brightening the surface that I am
more attracted to rather than gentle abrasion with a brass or glass fibre brush.

Application to Sterling Silver Alloy

Having established, to my satisfaction, a procedure for preparing Britannia silver and
then a firing regime for the transparent green and red enamel, I wanted to see how
they worked on Sterling silver, the latter having a higher copper content. I purchased
some 1mm thick sheet and used the same size samples as before. I also obtained some
more Britannia silver, (1.1mm thick), to compare with the previous results. I used the
same procedure of nitric acid treatment, subsequent brightening using the silver foam,
and firing at a high rate (9400C for 40 secs). The result is shown in Table 5.

Result of Kiln Temp. Firing Time Appearance of Enamels

0
Sample Acid dips C Sec
Britannia Went 940 40 Green is bright and transparent, Flux is clear, Red on
Silver slightly flux is red with no yellow at edges
0.9mm yellow

Britannia Went 940 40 Green is bright and transparent, Flux is clear, Red on
Silver slightly flux is red with no yellow at edges
1.1mm yellow

Sterling Went 940 40 Green is not quite so bright and transparent, Flux is
Silver slightly slightly less clear, Red on flux is red with no yellow
1.0mm yellow at edges. Sample is significantly bowed.

Table 5. Enamel Tests on Britannia and Sterling Silver Prepared by Nitric Acid
Treatment and Using a High Rate of Fire

12
Although these results indicate that the green and flux are not quite so true when fired
on Sterling silver, the differences are slight and only observable under a bright light
and close scrutiny. Other colours may show up differently and further tests would be
needed to determine this.

I did find, however, that the sterling silver developed a pronounced bowing and I
suspect that I had taken it above the solidus temperature and it had softened.
Nevertheless, I feel happy to use the Sterling silver as an alternative to Britannia
silver, albeit with rather more care about firing rate, and I decided to use this alloy
when I came to do a few more tests on different red transparent enamels.

Back to the Heat and Pickle Method

Having decided to adopt the nitric acid treatment, I had, on the way, found out about
high rate of firing and perhaps not to abrade the surface if possible. Would these latter
procedures have improved the results shown in Table 3? So, I repeated the test shown
as sample 1 in Table 3, but with no abrasion after pickling and using silver foam to
brighten the surface. I also fired the enamels at the high rate, (9400C for 40 secs).

The green was marginally less opaque than sample 1, the flux almost clear and the red
was somewhat less brown. Notwithstanding this, the enamels were still not quite as
good as when the nitric acid treatment was used. It confirmed to me that significant
sub-surface oxidation is detrimental to the appearance of these enamels laid on silver
alloy.

Some Further Tests on Red Enamels

There is little doubt that my investigations have been dominated to some extent by
trying to produce a good red using the Soyer 8 transparent enamel. It made me
wonder if, using the procedures here, I could achieve acceptable results with other
makes of red transparent enamels. So, I obtained a representative selection of seven
other red enamels for testing. They were as follows :-

Soyer red 39, Soyer Light red 41, Schauer Ruby red 8, Schauer Ruby red 1216,
Blythe Red A31, Latham Cherry red 236, Kujaku Ruby red 105A. The latter enamel
is relatively recently available in the UK and is claimed to not require an undercoat of
flux. Sarah Wilson has already reported on Kujaku reds, (GOE Winter Journal 2000)
and found them very satisfactory with and without a flux undercoat.

I purchased some circular Sterling silver stampings that had a retaining edge. I used
the nitric acid treatment without any prior annealing and as expected the stampings
turned a light straw colour and no blackening. I cleaned them with the silver foam and
applied two coats of Soyer 3 flux using the high rate of firing previously described.
The Kujaku enamel was applied directly to the metal without any prior flux coats.

13
The results are shown in Figure 4.

A Soyer 8

B Soyer 39

C Soyer 41

D Schauer 8

E Schauer 1216

F Blyth A31

G Latham 236

H Kujaku 105A

Figure 4. Tests on Some Red Transparent Enamels

Due to the photographic limitations, the Figure does not do justice to the nice clear
colours obtained with all the enamels. There was no yellowing of flux or red and the
result with the Kujaku 105A was particularly impressive since it was laid direct on the
sterling silver without an underlying flux layer. Perhaps because of the retaining edge,
and the geometry of the discs, there was no bowing of these samples.

Conclusion
When I look back at trying to understand the process of enamelling on silver alloy and
my attempts to translate that into a working procedure, I realise it was not quite as
straightforward as I had imagined. Perhaps the most revealing conclusion is the fact
that silver and its alloys are porous to oxygen at enamelling temperatures. If oxides
are allowed to form, they can affect the clarity of transparent enamels. The nitric acid
treatment is an effective method of ensuring the metal is best prepared for the
subsequent application of such enamels. There should be no prolonged heating of the
silver prior to the laying of the enamels.

R.L. Jackson August 2007

14