High density flexible multilayer

State of the art

Flexible materials are already being used today as base substrates for electronic assemblies. Typical materials range form polyimides for high-end applications down to cheap paper-like or plastic materials like PET for very cost-sensitive applications. In this project we will concentrate on these “high-end” type flex substrate materials like polyimide, LCP (liquid crystal polymer) etc. Manufacturers of these types of flex circuits detect an increasing interest in a wide variety of fields for flex technology.

The development of high density and high performance flexes is mainly driven by applications as LCD, HDD, DVD and consumer products, such as camcorders or foldable mobile phones. The needs for the interconnection of LCDs, especially color LCDs and the increasing demands on the resolution are pushing the limits of the technology for single and double sided flexes. High volume production reaches a pitch of 50 µm for single sided and 70 µm for double sided flexes. An example of a LCD interconnection flex with a pitch of 50 µm is given in the picture below. Multilayers are mainly required in high end consumer products like camcorders and the leading edge of technology was used for the motherboard of the Playstation 2 which was a 10 layer flex build up with a 6 layer core and two sequential build ups. All features known from rigid boards, such as surface via, via on pad or staggered vias are possible. Only a very restricted number of manufacturers in the Far East can provide these technologies and the European manufacturers are far behind the capabilities of the Far East located companies especially if high volume production is required.

Although enjoying an increasing popularity current flex substrate and assembly technologies still have serious restrictions:

Normally a flex substrate is a 3-layer laminate where a Cu sheet (layer 1, thickness typically minimum 12µm) is laminated onto a polyimide carrier (layer 2) using an adhesive (layer 3). Thinner Cu is not possible because of mechanical constraints (danger of breaking of the thin Cu layer during the lamination process). Although widely used these 3-layer laminates have following limitations:

• The conductor pitch is limited to minimum 40-50 µm, because the underetching of the Cu during wet etching process is proportional (and about equal) to Cu thickness
• The Cu thickness puts a lower limit to the bending radius of the substrate
• The presence of an adhesive has negative influence on feasibility and reliability of certain advanced assembly technologies, like e.g. adhesive flip-chip (melting of laminate adhesive during flip-chip thermocompression step)
• Only a limited number of sources (2 Asian) for adhesives are available; that fact increases the price for 3-layer laminates

Also 2-layer laminates are offered on the market, where the Cu is deposited onto the polyimide without the use of adhesive. These laminates are expensive because of the complex production process. The problem is the adhesion of the Cu (or any other metal) onto the polyimide surface. The adhesion strength of a metal onto a polyimide is normally very low. What is currently done in production environment to produce 2-layer laminates with sufficient adhesion strength, is first modifying the polyimide surface by e.g. plasma treatment, followed by a thin-film deposition (sputtering) of Cu onto the modified (more reactive) polyimide, followed by further increasing the Cu thickness by electrolytic plating. In this case thin Cu layers with very fine pitch capability are possible, but it is clear that the use of vacuum processes (plasma treatment, sputtering) increases the price of the product considerably. Another production method for 2-layer laminates, used in Japan, is the deposition of liquid polyimide layers onto Cu films followed by an imidization process at high temperatures in a nitrogen atmosphere. This is a cheaper production process, but the restriction here is also the minimum required Cu thickness for mechanical reasons, thus limiting the minimum achievable pitch. The lamination of two single sided materials using a thermoplastic polyimide adhesive is used to get double sided materials. Today such 2-layer laminates have a minimum Cu thickness of 12 µm.

So a decrease of the price of the base substrate is a vital issue for wider use of high-end flex, and especially cheap high-end flex base substrates with very fine pitch capabilities (thin Cu conductor layers) do not exist so far.

On this type of flex only pure electronic assembly is performed up to flip-chip components. Embedding of passive or active components in flexible printed circuits is at least not state of the art and is, according to the knowledge of the consortium, not performed anywhere in the world so far, certainly not on a commercial basis. The limitation to two component layers (front and back side of the flex laminate) limits the compactness of the circuit.