Characterization of new yellow cadmium-free gold brazing alloys

M. DABAL, M. MAGRINI (D.I.M.E.G., University of Padova, Italy) M. POLIERO (Leg. Or., Bressanvido, Vicenza, Italy)

Introduction Production of a full range of jewellery is usually obtained using gold brazing alloys (solders) of the same gold content (caratage) as the materials that must be joined. Generally, a brazing alloy (often called solder in the jewellery industry) facilitates a permanent bonding of two metallic pieces by heating at a temperature higher than the Liquidus temperature of the brazing alloys, but lower than the Solidus temperature of the parent materials to be joined. A high liquidus temperature of the brazing material can cause some problems because any high temperature treatment decreases the mechanical properties of the parent alloy by annealing and can fade the surface of the jewellery. Historically, jewellery brazing alloys (solders) have been obtained by alloying gold-silver-copper alloys with additions of metals such as zinc (Zn) and cadmium (Cd). Low-melting metals such as Cd are employed to decrease the melting range and to impart desirable wetting and flow characteristics. However, cadmium is very toxic to human organs. The fumes generated during the melting process contain CdO that is very irritant to the respiratory system and can be

absorbed through inhalation. Several studies have implicated exposure to CdO fumes as an industrial health hazard that must be controlled, and the International Agency for Research on Cancer has classified Cd as a carcinogen. The temperatures involved in brazing often exceed the Cd boiling point (797C) and the vapor pressure of Cd is very high under these conditions. Thus, there is a requirement by the industry for cadmium-free brazing alloys. Substitution of Cd in jewellery brazing alloys can be done by use of other low melting point metals; the most applicable are tin (Sn), indium (In) and gallium (Ga) as well as zinc (1,2). In and Ga have very low melting points (156C and 30C respectively) and high boiling points (2080C and 2403C). Additions of a few percent of these elements depress the melting range of Au-AgCu alloys, but the solubility of In and Ga is limited and in multicomponent systems, especially in silver-rich alloys, these elements do not significantly affect the liquidus temperatures but depress the solidus temperatures, resulting in a widening of the melting range. Whilst a number of cadmium-free brazing alloys are commercially

available at all caratages, there is little published on the performance of such solders. This paper reports on work carried out to characterise the properties of yellow gold jewellery cadmium-free brazing alloys containing indium and gallium in order to determine whether cadmium substitution by these elements adversely affects performance. Experimental procedures Yellow gold brazing alloys of different composition and gold content (18, 14 and 9 ct) were prepared in the form of 1 mm and 0.5 mm diameters wires and as foils. The chemical composition of the alloys are given in Table 1. Three different compositions of 18 and 14 ct gold alloys and two compositions of 9 ct gold alloys were prepared. Samples of wire were annealed at 550C for 20 mins in air. The alloys were characterised by metallography of the wire, by microhardness measurements on a Leitz-Wertzlar microhardness tester, tensile tests on an Instrom 1121 machine and differential thermal analysis using a Netsche Differential Scanning Calorimeter with a heating rate of 2.5 K/min. Differential thermal analysis is a powerful technique for evaluating changes in physical-chemical properties as a function of temperature because it records exothermic or endothermic behaviour. Fluidity and wetting tests were performed according to Normandeau (1), using brazing alloy foils of dimension 14 x 3 x 0.25 mm, prepolished and deoxidized. These foils were positioned between two foils of conventional jewellery alloys of the same caratage with a spacing of 0.25 mm. The samples were heated in a furnace in a nitrogen atmosphere at a temperature 35C higher than the solidus temperature of the jewellery

Table 1. Composition of brazing alloys

Ag% 18 Kt /1 18 Kt /2 18 Kt /3 14 Kt /1 14 Kt /2 14 Kt /3 9 Kt /1 9 Kt /2 6.25 5.75 55.25 13.33 14.16 14.16 31.88 29.38 Cu% 8.50 9.50 12.25 15.00 14.25 14.58 18.13 19.38 Zn% 5.50 6.00 6.50 8.75 9.17 10.00 8.12 10.62 In% 4.75 3.75 1.00 4.58 3.75 2.92 3.12 2.50 Ga% 1.25 0.62

alloys for 5 minutes. After cooling in air, the brazed joints were examined to determine the fluidity of the brazing alloys. Serious errors can occur if the surfaces of the assembled foils are not the same in all tests. For this reason, we prepared a mirror-like surface for all samples by the usual metallographic polishing methods. Metallographic analysis was carried

out on sections of brazed joints to determine erosion behaviour, the presence of any microporosity or fractures. Microhardness profiles were made to evaluate coupling resistance. Results Metallographic analysis of the alloys shows a substantial work hardening resulting from the drawing of the wires, Figure 1. Annealed alloys exhibit an increase in grain size with a partial rearrangement of the structure, Figure 2 (3). The results of microhardness analyses, summarized in Table 2, confirm the substantial work hardening of the drawn wires. These measurements show that the alloys with a critical degree of deformation ( cr) show hardness values higher than 200 HV (theoretical limit for malleable alloys); moreover, annealing at 550C for 20 minutes results in a severe decrease in hardness, due to lowering of the internal strain by recrystallisation. The critical amount of deformation is the minimum amount of

work hardening necessary for recrystallization. Among the more important mechanical properties that are of interest to goldsmiths are the yield strength, the strain and the ductility characteristics. In Table 3, the yield strength and Youngs modulus results are summarised. The 9 ct alloys do

Figure 1 - 18 Kt alloy work hardened

Table 2. Microhardness values of brazing alloy wires

Alloys 18 Kt-1 hardened 18 Kt-1 annealed 18 Kt-2 hardened 18 Kt-2 annealed 18 Kt-3 hardened 18 Kt-3 annealed 14 Kt-1 hardened Microhardness (HV) 244 150.3 254 144.8 234 160 274.3 182.2 Alloys 14 Kt-2 hardened 14 Kt-2 annealed 14 Kt-3 hardened 14 Kt-3 annealed 9 Kt-1 annealed 9 Kt-2 hardened 9 Kt-2 annealed Microhardness (HV) 295 177.3 227.5 189.5 167 170.5 168.7

Figure 2 - 18 Kt alloy annealed

14 Kt-1 annealed

Table 3. Yield strength and Youngs modulus values of brazing alloys

Alloys 18 Kt-1 hardened 18 Kt-1 annealed 18 Kt-2 hardened 18 Kt-2 annealed 18 Kt-3 hardened 18 Kt-3 annealed 14 Kt-1 hardened 14 Kt-1 annealed Yield strength (Kg/mm2) 16 15.4 20 14.1 13.8 13.6 16.2 13.6 Youngs modulus (Kg/mm2) 636 50.1 415 42.3 431 50.2 450 150 Alloys 14 Kt-2 hardened 14 Kt-2 annealed 14 Kt-3 hardened 14 Kt-3 annealed 9 Kt-1 annealed 9 Kt-2 hardened 9 Kt-2 annealed Yield strength (Kg/mm2) 21 17.4 22.7 12.5 21.8 17.3 16.4 Youngs modulus (Kg/mm2) 221 55.7 1274 385 99.5 85.5 79

Table 4. Thermal characteristics of brazing alloys

Alloys 18 Kt /1 18 Kt /2 18 Kt /3 14 Kt /1 14 Kt /2 14 Kt /3 9 Kt /1 9 Kt /2 273.9 T recovery (C) 291.4 298.9 322.4 348.3 335 323.6 T solidus (C) 730 682 792 669 660 668 637 658 T liquidus (C) 765 767 829 741 745 748 702 721 Tl Ts (C) 35 85 37 72 85 80 65 63

not show much reduction in properties after annealing as the annealing temperature is insufficient or cooling not sufficiently rapid. The preferential grain orientation obtained by drawing increases the elastic modulus when it is measured in stress direction. The yield strength of the work hardened alloys is lower than in the annealed condition, suggesting some hardening due to phase changes on cooling after

annealing. Cadmium-containing alloys (1) show yield strengths slightly greater than the cadmiumfree ones, but the latter show a greater ductility that makes them more suitable for jewellery fabrication. Moreover, an increase in In and Ga contents decreases the yield strength, as shown in Figure 3. The results of the differential thermal analyses are given in Table 4. All the samples show a broad

Yield strength in Kgmm2

20

16.2 2.9

Indium contents %

4.15

Figure 3 - Yield strength of alloys versus Indium contents in 14 Kt alloys

TG/% 0 -5 -1 -1.5 -2 -2.5 20 70 120 170 220 270 Temperature/C 320 370

DTA/uV 10 0 -10

endothermic peak at temperatures between 273 and 349C. This peak denotes some recovery of the material with disappearance of internal stress, as seen in Figure 4. Cadmium-free brazing alloys show liquidus and solidus temperatures values close to those of brazing alloys containing Cd as reported by Normandeau (1). Some samples show increased thermal characteristics since liquidus temperatures are lower than Cd brazing samples. In alloys with In and Ga contents greater than 1%, a decrease of the melting temperatures is evident. The samples 18 Kt/2 and 14 Kt/2, with high percent of Cu and Zn shows a broad melting range. The wetting and fluidity test results are summarized in Table 5 and show that the brazing alloys exhibit a good fluidity when molten, the surface area after flow being 4 times greater than that of the original. Also, all the joints show a good colour match between the base metal and the brazing joint. We evaluated the difference in colour by subjective comparison only as the colour spectrophotometers are not able to measue such small areas. Metallographic examination of the brazed joints showed a good metallurgical bonding between the parent and the brazing metals too. There was no evidence of any erosion of the parent metal by the brazing alloys or of porosity, though in some Table 5. Fluidity test of alloys

-20

Alloys 18 Kt

Si (mm2) Se (mm2) 3.5 3.5 3.5 14.14 17.30 13.32

Si/Se 4.04 4.94 3.80

420

14 Kt 9 Kt

Figure 4 - DTA and TG curve of the 14 Kt/3 samples

samples localized fractures were recorded. We can confirm that cadmium-free brazing alloys show good wetting. The microhardness profiles are given in Figures 5-7. The low microhardness values recorded in the conventional parent alloys around the joint could be caused by heating close to their solidus temperatures during
300 250 HARDNESS HY 200 150 100 50 0 100 200 300 400

furnace brazing operations, expecially for 18 ct alloy assemblies. This heating can allow secondary recrystallization, increasing the grain size, with a decrease in hardness of the materials (4). In the joint areas, the microhardness values are the same as those of the parent and brazing metals before the fusion process, a good characteristic of the joint.

500 600 MICROM

700

800

900

1000

Figure 5 - Microhardness profile of 18 Kt brazing alloy

250 245 240 HARDNESS HY 235 230 225 220 215 210 205 100 200 300 400 500 600 MICROM 700 800 900 1000

Conclusion Cadmium-free gold brazing alloys exhibit better thermal characteristics than the cadmium-containing alloys, although an increase in the melting range of 18 ct alloys is seen. Furthermore, this behaviour does not affect the alloy fluidity which is excellent in many samples. This good characteristic enables a satisfactory filling of the joint cavity. In addition, metallographic analysis shows good bonding between the alloys, an absence of any erosion or porosity, but the presence of localized fractures in some instances. Mechanical behaviour of brazed joints is also satisfactory. Cadmium-free brazing alloys show hardnesses greater than 200 HV after wire drawing, but annealing at 550C for 20 minutes enables recovery of ductility. A lower yield strength and a greater ductility are also recorded, making these materials more suitable for further working operations than the cadmium-containing alloys. These results demonstrate that cadmium in gold jewellery brazing alloys can be completely substituted in brazing alloys by indium and gallium, with little detrimental affect on any of the important materials characteristics. References 1 G. Normandeau, Cadmium free gold brazing alloys, Proceeding of the Santa Fe Symposium, 1989, pub. Met-Chem Research Inc 1990, 179-209 2 G. Normandeau, Cadmium free brazing gold alloys, Gold Technology, No.18, April 1996, 2024 3 A. Langford, Cold work and annealing of karat gold jewelry alloys Proceeding of the Santa Fe Symposium, 1989, pub. Met-Chem Research Inc 1990, 358-360. 4 M.F. Grimwade, Gold Technology, No. 2, 1986, p. 17-19 Acknowledgement This article was originally presented at the Santa Fe Symposium, May 1998.

Figure 6 - Microhardness profile of 14 Kt brazing alloy

200 180 160 140 120 100 80 60 40 20 0 100 200 300 400 500 600 MICROM 700 800 900 1000

Figure 7 - Microhardness profile of 9 Kt brazing alloy

HARDNESS HY