Assembly of Fine-Pitch Package Integrated Circuits in a Laminated Transfer Process for E-Textile Applications ^†

Abstract

Laminating screen-printed, flexible transfers is a reliable method of creating electronic connections across a textile. However, their limited resolution means that it is not possible to solder on fine-pitched components. An alternative substrate, etched copper on polyimide is compatible with fine pitch components, but is unsuitable for use over large areas. This work examines methods of combining the two systems to produce e-textiles that combine both their benefits. Tensiometer tests showed that 3M 9087 double-sided tape was the most robust for attaching the polyimide, and electrical interconnects formed by overlapping the transfer on top of the polyimide circuit gave the most stable and lowest resistance connections.

1. Introduction

E-textile garments seek to provide interactive electrical functionality across potentially large textile surfaces in an unobtrusive manner. To achieve this, they require both large-scale conductive structures, compatible with the properties of textiles, e.g., flexible and stretchable, and combine this with small, detailed conductive patterns suitable for attaching small surface mount electronic components and circuits [1]. One of the best methods currently available for producing large-scale, electrical connections across a textile is using flexible, screen-printed laminate transfers [2]. The system involves printing several layers of different functional pastes, including a silver-based conductor, and insulator, protective encapsulation and an adhesive, onto a temporary backing. The structure is then inverted and laminated onto a textile [3,4,5], as illustrated in Figure 1.

This process can be used to create large, garment-scale designs, limited only by the area that can be screen printed. The transfer is capable of being stretched a few percent beyond its original length, and combining this with meandered designs make it suitable for use with stretch fabrics [2].

The disadvantage of this process is the limited resolution of the printing process. Typically, feature sizes below 200 µm are not possible. This precludes the attachment of fine-pitched electronic components onto a transfer. The silver pastes that form the electrical connections are also unable to withstand the >200 °C temperatures needed for soldering, presenting an additional challenge when attaching components. It is possible to use components with wider-pitched connections, attached with low-temperature curing conductive epoxies, but this is not ideal as the larger components are more obtrusive and noticeable to the wearer of an e-textile garment and epoxies are both weaker and not as easy to manufacturer compared with soldering.

An alternative method of producing flexible circuits is using copper on polyimide patterned using a photolithographic etching process. As shown in Figure 2, this process allows feature sizes as low as 100 µm, comparable with those of traditional rigid PCBs, and the materials are compatible with the temperatures used in s standard manufacturing processes such as reflow soldering.

A system incorporating both these techniques, using copper and polyimide modules for mounting the components, interconnected by large, laminated transfer tracks, would allow the benefits of both to be realised. This work investigated various means of connecting the systems together, evaluating both their mechanical and electrical performance.

2. Method

Printed track layouts were designed, and the transfers were produced by Conductive Transfers Ltd. (Barnsley, UK) and tested with custom-made circuits of etched copper on polyamide (GTS Flexible Materials, Ebbw Vale, UK).

First, a series of different adhesives were investigated to find one suitable for bonding polyimide. Polyimide has a very smooth, inert surface and so is difficult to bond adhesively. Five different adhesives were evaluated:

• Loctite 401 cyanoacrylate;
• Loctite 5940 silicone paste adhesive;
• Lohman K2211 double-sided TS tape;
• 3M 9087 white double-sided plastic tape;
• 3M VHB™ series 9473 clear double-sided plastic tape.

These were chosen based on their availability and their advertised ability to adhere to polyimide. Both peeling force and the sheer pulling forces were assessed using a Tinius Olsen H25KS Tensile tester which records the force on a sample as strain is increased.

Following this, three different architectures for combining the two conductors onto the textile were tested. As shown in Figure 3, the polyimide circuit was first placed, copper side down, onto the transfer. It was secured with an adhesive and the electrical connections were made using anisotropic conductive tape, a double-sided adhesive tape that only conducts in one dimension to prevent short circuits between multiple adjacent connections. The second method uses the polyimide circuit placed copper side up and conductive epoxy interconnects extending over the edge of the circuit and down onto the transfer. The final structure again placed the polyimide circuit face up with the conductive transfer extended over the top. The conductive layer of the transfer was exposed on the underside, and it was also connected to the copper with anisotropic tape.

These were again tested in the tensile tester, with the resistance between the transfer and the copper measured as the sample was stretched.

3. Results

The maximum pulling and peeling forces for each adhesive are shown in

Table 1. Maximum pulling and peeling forces withstood by the different adhesives between polyimide.

Adhesive	Maximum Peeling Force/N	Maximum Pulling Force/N
Loctite 401 cyanoacrylate	4.2	95.9
Loctite 5940 silicone paste adhesive	0.9	51.7
Lohman K2211 double-sided TS tape	6.6	54.2
3M 9087 white double-sided plastic tape	9.2	125.9
3M VHB™ series 9473 clear double-sided plastic tape	9.1	118.4

.

In both the pulling and the peeling test, the two 3M tapes showed the highest strength. Of these, the 9087 tape withstood a slightly larger force and was significantly thinner so was chosen as the adhesive for the subsequent tests.

The relationship between resistance and stretch distance for the three connection structures is shown in Figure 4. The third structure, with the polyimide and copper placed under the conductive transfer was able to withstand the greatest stretch without a significant increase in resistance. This design had the advantage of adhering the polyimide directly to the textile which increased its durability because the textile provides a greater bonding surface area. In all tests, the initial resistance was between 1 and 2 Ω meaning that the conductive epoxy and anisotropic tape were initially providing connections of approximately equal conductivity.

4. Conclusions

This work investigated various methods of combining etched copper on polyimide circuits combined with laminated, screen-printed transfers. This methodology enables small, detailed circuit layouts, suitable for mounting fine-pitch components, to be combined with large conductive structures suitable for connecting circuits across a garment.

It was found that double-sided-tape-based adhesives provided the strongest bond to the polyimide, with 3M 9087 giving a combination of high bonding strength (125.9 N) and low thickness (184 µm).

Of the three methods of adhesive structure tested, the arrangement with the polyimide circuit bonded to the fabric and the conductive transfer overlapping above it gave the best conductivity as the samples were stretched. Anisotropic conductive tape was found to be suitable for making the electrical connections.