With the ever-increasing drive for miniaturisation, microjoining is critical to the successful performance of assemblies in a wide variety of industries. These include medical products, electronics, telecommunications, defence, automotive, aerospace, and power generation, and cover a wide range of microjoining applications. The increasing complexity and miniaturisation of devices used by these industries demands sophisticated joining processes with low defect rates as they strive to decrease the size of their assemblies whilst increasing performance. This article defines microjoining, the processes commonly used and some of the issues concerned with joining very small components. In recognition that worldwide there is strong growth in the requirement for miniaturised assemblies by these many diverse industries, CSIRO in Adelaide, South Australia commenced a new Microjoining initiative. This article then describes some of the work that has so far been conducted.

MICROJOINING RESEARCH AND DEVELOPMENT AT CSIRO

S.N. DOE

1. INTRODUCTION As microtechnology advances the need for miniaturised components is rapidly increasing. The field of microjoining has come into existence because of this driving force. Although there is no dictionary definition of microjoining, there is consensus that joining sheet materials of less than 0.5 mm in thickness, or tubular materials of less than 1 mm in diameter, constitutes microjoining. Microjoining is a growth area and critical to the successful performance of many components in a wide variety of industries. These include medical products, telecommunications, electronics, defence, automotive, aerospace and power generation, and the components cover a wide range of microjoining applications by welding, brazing, soldering, or adhesive bonding. Environmental concerns such as solvent and lead use are also driving research in this area. The increasing complexity and miniaturisation of devices used by these industries demands sophisticated joining processes and low defect rates as manufacturers strive to decrease the size of their assemblies whilst increasing performance and decreasing cost. Microjoining has emerged as a major part of micro-assembly and micro-mechanics technologies and is receiving increasing interest1. As well as problems in joining together components, there are additional concerns with the sheer scale of some of these components. With pieces smaller than the width of a human hair being common the use of microscopes as a joining aid is frequent. 2. CSIRO Microjoining Facility Recognising the value of a designated Microjoining facility, CSIRO Microjoining was initiated in the later half of 2000. A laboratory area was prepared as a nominally clean work place, fully air conditioned with a HEPA filter extraction system. Investment in equipment and resources was planned to allow the group to develop, compete with, and complement other microjoining facilities or research organizations. A multi-disciplinary team, including scientists, engineers and technicians covering the fields of metallurgy, welding, brazing and soldering, power beams and electrical design was assembled. This was necessary to support the diverse fields in which projects would be developed. Recruitment of staff was completed in January 2001 and since then the team has continued to develop its capabilities and conduct market research. The Microjoining team is presently establishing itself as a resource in Australia, and is currently the only one in Australia. It has affiliations with WTIA (The Welding Technology Institute of Australia), and also with other microjoining establishments around the world. Some of the recent projects are described in more detail in Section 6. 3. MICROJOINING TECHNOLOGY The techniques available for microjoining are much the same as in other joining industries with similarly controlled variables such as voltage, current and travel speed. Joining processes can be divided into four main groups dependent upon how the heat is applied and the effect of the heat.

3.1 Fusion Welding

In this method a heat source is used to melt the joint and form a bridge between the components. In microjoining, close control of the energy is required if damage to the work piece is to be avoided. The processes that lend themselves most readily to microjoining are: Gas Tungsten Arc (GTA), plasma, and power beams (laser and electron beam). These processes offer high precision and fine control.2 3

3.1.1 Arc Welding With the advent of transistorised power sources that can give regulated and controllable outputs of less than 1 amp, arc welding has become an important joining process utilising a versatile and inexpensive power source. Of the available methods of arc welding two processes stand out: GTA and Plasma welding. By pulsing the current in either of these processes the heat input can be further lowered, but this is at the expense of an increased number of welding variables. The peak pulse controls the penetration whilst the background pulse allows solidification without extinguishing the arc. Tooling is extremely important with these processes. Failure to utilise this correctly will result in joint opening and burn-through. When joints are correctly designed and appropriate tooling is used excellent joints are produced. 3.1.2 Power Beam Welding Laser (LBW) and electron beam welding (EBW) processes have long been identified as being excellent microjoining tools, capable of very fine control over power and positioning. They can focus a high energy beam onto a very small spot size allowing deep penetration with little distortion. EBW is almost always conducted under a vacuum unlike LBW which utilises an inert gas atmosphere to protect the joint area. For this reason EBW provides probably the highest quality joints but at the expense of throughput due to pumping time. The size of the vacuum chamber also limits the size of the component. In addition in EBW, as the beam focus diameter tends to be of similar dimension to the depth of penetration, and hence not very deeply penetrating. For this reason the process is more likely to be conduction limited welding rather than deeply penetrating, which limits some of the attraction of the process4. Lasers however can focus their beam energy down to very small sizes, typically of the order of tens of microns. LBW therefore offers greater advantages for microjoining where low penetration is required5. The laser energy can also be pulsed to reduce thermal input and hence distortion. Lasers can be used for similar applications to arc and electron beam welding, however due to the decreasing cost of lasers they are continually finding new application areas. Lasers are also finding use in electronics manufacture where they can be used to reduce the soldering time and hence heat input6 7

3.2 Solid State Bonding

Here there is no melting of material. Joints are made by plastic flow occurring at the interface to bring the components into intimate contact and thereby form a bond. Many different processes are available, but those used in microjoining tend to be based upon ultrasonic vibration or friction welding to join dissimilar materials. Diffusion bonding either in the liquid or solid phase is another process that has also been successfully used on a microjoining scale. 3.2.1 Ultrasonic Bonding Ultrasonic bonding is a solid phase joining process that relies upon displacing the interfacial oxides and contaminants whilst at the same time using pressure to form a bond. There is a slight temperature rise during bonding. Ultrasonic bonding is probably the most important process in the electronics industry where it is used as a means of first level interconnect between semiconductor chips and package pin-

outs. In this process there is a high frequency vibration with a low pressure applied to cause the plastic flow required. There are two variants of ultrasonic wire bonding – ball bonding and wedge bonding8. Wedge bonding is primarily performed using aluminium wire. The wire is wedge bonded at one point using ultrasonic energy, then drawn out in a loop then similarly wedge bonded at the other end. It is usually performed at ambient temperature, unlike ball bonding which is characterised as a thermosonic process, meaning heat (typically 150°C) is applied during the bonding process. Today the most used method is ball bonding with gold wire. This process works by forming a small ball on the end of the wire. This ball is bonded as the first joint, then the wire is drawn out in an arc before attaching this as a wedge bond. Unlike wedge bonding, ball bonding has the advantage of being able to be drawn out in any direction. 3.2.2 Diffusion Bonding Diffusion bonds are made by using intermediate heat (typically 0.4 of melting temperature) for a period of time, and cause minimal distortion to components. Dissimilar metal joints can also be made this way without intermetallic compounds being formed. During the time at temperature diffusion occurs between the surfaces until a joint is formed. There are two types of diffusion bonding: solid phase and liquid phase. During liquid phase diffusion bonding a low melting point eutectic is formed by diffusion between the component parts. 3.2.3 Friction Welding Friction Welding may be used to join together a wide range of materials including non–metals, however the geometry of components is fairly limiting. In the microjoining field it has been used to join components to heat sinks, specifically in the electronics industry where it has been used to directly attach aluminium heat sinks to alumina substrates5. It is also commonly used for attaching tubes or rods to bulk or sheet components.

3.3 Soldering / Brazing

During soldering or brazing, a third lower melting point material is drawn into the gap between the components by capillary action. There is no melting of the components themselves. The difference between soldering and brazing is defined by the melting point of the filler metal, with brazing considered to be occurring at temperatures greater than 450°C.9 The gap through which the solder or braze alloy is drawn should be small so that maximum strength is imparted to the joint. Soldered and brazed joints gain their strength from the wetting of the parent metals by the molten solder or braze metal. Any contamination would cause a reduction in wetting and consequent lowering of joint strength. Both processes may be used to join similar or dissimilar materials by selection of the correct filler alloy, or by using coatings. Brazing is also finding use as a means of joining metals to non-metals. Soldering is one of the primary joining processes used by the electronics and electrical industries, and may be used in the manufacture of such products as printed circuit boards, control systems or audio systems. Solder joints have to provide electrical, thermal, and mechanical functions. Traditionally solders containing lead have been used in various applications. However, concerns about toxicity and health hazards means that there is a drive to develop and use lead free solders. This may cause problems with the traditional soldering methods and new techniques may be required.

3.4 Adhesive Bonding

Adhesive bonding relies upon the attractive forces between the molecules at the surface of the adhesive and those of the surfaces to be joined. The larger the molecules the better the adhesion, which is why organic adhesives are commonly used. The liquid adhesive is used to wet the surfaces to be bonded. This is then cured, or hardened, to form a solid bond. This curing can be aided by applying temperatures of approximately 150°C. Like brazing or soldering, the thickness of the adhesive should be minimal to provide a strong bond. Recent developments with adhesives has increased the scope for their use. By doping the adhesive with metal (typically silver) it is possible to make them electrically or thermally conductive10. The development of thermally conductive adhesives has aided heat dissipation in electronic devices, thereby increasing component lifetime, and electrically conductive adhesives are being considered as replacements for solders as they have the potential to remove lead from the assembly. However it is not just a case of simply changing to electrically conductive adhesives as various other factors, such as electrical conductivity, lifetime, component finishes and design issues would need further assessment. 4. DESIGN FOR MICROJOINING Like any joining process there are many important considerations to take into account during selection, but process control becomes critical when joining small components. Not surprisingly different processes may often be selected for the same application. Consideration would then need to be given to such factors as capital expenditure, staff skills, and ease of mechanisation or automation. The major considerations are:

• heat input • fit-up • consistency / repeatability • other factors

4.1 Heat Input

With such thin and delicate components, precise control over the welding parameters must be exercised. Too much power will result in over-penetration and potential damage to delicate components. Often the welding is conducted near to components that may be damaged by the heat from welding. Very tight control of the heat input is required in this case. When welding metals of high thermal conductivity to metals of low thermal conductivity, or thick to thin metals, correct joint design and heat sinks are needed to achieve a thermal balance.

4.2 Fit-up

It is not uncommon to have to join together components that are smaller than human hairs or thinner than sheets of paper. This makes them extremely difficult to handle and assemble so that they fit together well enough to allow joining. Accurate fixturing and handling devices are almost mandatory for components of this size. These add to the cost of joint production. Consistency in part machining during mass production will also be essential if the maximum benefit is to be obtained from these fixtures.

4.3 Consistency / repeatability

Repeatability of the joining process is essential for microjoining. With the advent of advanced electronics, motion controllers, and monitoring systems it is possible to accurately

control the joining process. Once a process is developed no changes to parameters should be made without consideration of the effect on the rest of the process.

4.4 Other Factors

As with any application, there are often situations where two or three processes will work for a given material or design11, and a further assessment must be made as to which would be the optimum depending upon various factors that must be considered for any joining process:

• investment in capital equipment • production rate required • degree of automation required • level of training of operator.

5. TYPICAL APPLICATIONS

5.1 Electronics

A very high percentage of manufactured goods rely upon electronic circuit boards and circuitry. Traditionally eutectic or near eutectic lead tin solder is used in the manufacture of these components. There are different ways of applying heat to melt the solder but the methods used almost universally are reflow ovens and wave soldering machines. As components become more complex, more delicate, and packing density increases, there is a higher probability of joining related problems.

5.2 Controls and Sensors

With the current trend in computer control and remote sensing, there is a need to manufacture ever smaller and more intelligent sensors and actuators. Sensors are devices that detect and monitor changes or variations in an object or process relative to a reference or standard. Depending on their intended function, sensors recognize variations in weight, vibration, motion, pressure, colour, heat, light, magnetism, chemistry, etc12. With modern, rapidly developing technology, new sensors and instruments are continuously being developed. Their components are miniaturized, sensitive, and precisely arranged in a very small and compact assembly. Most often they utilize micro-electronics, and to protect their components from the elements of the user environment, sensor devices are usually sealed. This protects devices from contamination, moisture, impact, abrasion, heat, magnetism, etc. By doing this, the sensor is manufactured to give consistent and reliable performance throughout the life of the product. As mechanical assembly (screws, gaskets, clamps, etc.) causes time related failures, welding is often the preferred joining process for both intermediate and final assemblies. When hermetic sealing is required, welding is mandatory13. Materials used to manufacture sensors vary, but include aluminium, stainless steel, titanium or nickel super alloys depending upon the potential environment.

5.3 Bio-medical With the recent rapid advances in medical technology there are a variety of intrusive procedures used in the medical industry today requiring tools, instruments, sensors and components in materials that are inert with respect to reactions with the body. Due to the intrusive nature of surgery these are usually as small as possible.

Product reliability is one of the major concerns with the medical industry, and any manufacturing process including joining must be able to satisfy stringent standards to ensure that in-service failure will not occur14. In addition to components used internally, a large number of microjoining applications can be found in support equipment such as automated drug dispensers.

5.4 Jewellery

There is a small amount of joining required in the jewellery industry, mainly to join precious metals for the fashion industry. Traditionally mechanical fixing, adhesives, brazing and soldering have been used in the jewellery industry15. Joining is also used to provide a colour contrast by application of a different metal slurry in a recess followed by flowing at temperature. 6. CSIRO MICROJOINING PROJECTS A literature survey and thorough market analysis was carried out. From this a number of possible areas for more detailed examination were identified and projects initiated with Australian companies.

6.1 Welding thin sheet material

A project to improve upon a currently used manufacturing method was initiated with a telecommunications company. They currently used small stainless steel cans to protect a delicate component, and these are made from machined bar stock material. Analysis of the requirements showed that welding of thin stainless steel sheet could provide a less expensive method to produce the protective container. The requirements of the joint were that it was hermetically sealed, and the weld did not stand proud of the outer surfaces, to allow insertion into another assembly. Low heat input welding processes, laser and micro-plasma, were examined, and trials conducted on 0.2 mm thick type 304 stainless steel that replicated the material and thickness of the components. Initial trials showed that welding material of this thickness would present problems for many of the conventional welding processes due to the lack of rigidity of the components giving unacceptable distortion. Close fitting fixtures were required, both to aid rigidity and to remove heat from the joint area. Failure to have such fixturing resulted in opening of the joint ahead of the weld, and consequent failure to close the joint or weld drop through as shown in Figures 1 and 2. Both the micro-plasma and the laser were able to produce joints satisfying the mechanical and physical requirements, however since the laser is a lower heat input process, it was selected for further work. A joining process was developed that used the CSIRO 500 W pulsed NdYAG laser. The welding of the components was completed successfully as shown in Figure 3. Current work is aimed at determining the internal temperatures that would be experienced by componentry inside the assembly during welding, however control of this should be accomplished by selection of optimum travel speed with respect to pulse frequency.

6.2 Welding fine diameter dissimilar metal wires

A company contacted CSIRO with a requirement to find a way to join together lengths of 0.05 mm diameter Electrolytic Tough Pitch copper wire to 0.35 mm diameter stainless steel wire for a medical application. The copper wire was covered with a 0.006 mm thick enamel coating with a melting point of approximately 160°C. The requirements of the joint were that it would provide an electrical contact and have a joint strength sufficient that it could withstand the rigours of handling prior to encapsulating in a sealant for insertion and withdrawal from the human body. Figure 4 shows the thickness of the copper wire in relation to a human hair. The joining process needed to be able to supply a controlled, small amount of energy, as failure to control the heat input would result in damage or even vaporisation of the components. Trials were conducted at CSIRO and overseas. These early trials soon revealed that laser welding at CSIRO appeared to give the greatest chance of manufacturing repeatable joints, and was a process that could easily be assimilated into a production environment. Further work concentrated upon this process only. A single laser pulse could attach the copper wire to the stainless steel wire as shown in Figure 5, however there was found to be too much expulsion of material, resulting in a reduced cross sectional area of the weld. It was felt that there was too much reduction in area to carry even the low current required. The enamel coating on the copper wire proved troublesome, however a technique was established where the energy from the laser beam was initially used to strip this layer prior to welding. The entire welding process including layer removal to welding is shown in Figure 6. Following the work by CSIRO and with work being conducted elsewhere supporting the findings it appears that the company is on the verge of producing a new medical component that has the potential to revolutionise certain medical procedures. 7. THE FUTURE FOR MICROJOINING Microjoining technology has advanced significantly in recent years, and equipment and processes are available to join very small or thin components. To continue developing in some of the new technology fields, be they electronics, biomedical, instrumentation or sensor, microjoining capabilities also need to advance, and take advantage of other scientific developments like robotics and automated systems to aid repeatability. There will be changes imposed upon industry due to environment concerns that will affect both manufacture and end of life. Some of the materials used are potentially toxic and after use there is concern that when disposed of in landfill sites they could leach out and contaminate the land or water. Continued research is essential to ensure progress of microjoining technology into the future. CSIRO and WTIA are working in collaboration under the Ozweld banner, where the business knowledge and contacts of WTIA is being utilised and combined with the technical and scientific abilities of CSIRO. This is part of the SMART Technet Project initiated to improve the competitiveness of Australian industry through the use of the latest technology.

Figure 1. Weld closure failure due to poorly fixtured components x 6 Figure 2. Weld slump due to excessive heat x 40 Figure 3. Cross section of welded joint made using pulsed laser x 40

Figure 4. Comparison of human hair (left) to 0.05 mm copper wire (right) Figure 5. Laser welded 0.05 mm diameter copper wire to stainless steel wire x 200

Cleaning of the enamel layer using a laser x 200 Further cleaning using the laser x 200 Figure 6. Completed weld of 0.05 mm thick copper wire to stainless steel wire x 200

8. REFERENCES 1 S Dunkerton and N Stockham: TWI Bulletin Jan 1991 2 CC Otter et al: TWI Technology Briefing 710, Aug 2000 3 KJ Ely: Proc Taiwan Int. Welding Conf. 1998, Sep 7-8, Taipei, Taiwan, pp 431-439 4 KI Johnson: TWI Document “Introduction to Microjoining”, 1985 5 SB Dunkerton: Welding Review International, May 1992, pp69 6 E Russell: IPC APEX Proceedings, Paper P-AD/3-1 7 F Legewie et al: G-ICALEO, 1998, pp 43 8 B Hueners: Electronic Packaging and Production, Oct. 1998, pp 55 9 AWS Soldering Handbook, 3rd Edition 10 G Thomas: TWI Bulletin March 2000, pp 25 11 A Cullinson: Welding Journal, May 1996, pp 29 12 J Goward: TWI Bulletin May/June 1999 13 Private Communication with Davidson Pty Ltd 14 N Stockham: Microtechnology for Surgical Tools and Implants 15 N Stockham: Welding and Joining, May 1997, pp14-16