Btech Thermoplastic Film Bonding

Revised: Nov. 1, 2007

Thermoplastic Anisotropic Conductive Film Bonding

Section A.  Low Volume Handling and Processing of NTP & TP Series Adhesives

General

This guideline covers the handling and processing procedures that are suitable for small batch usage of these products.  Large-scale production usage of the products requires techniques that are suitable for automatic handling and high speed processing, but require too much setup to be practical for a small scale or experimental application (please see Section B for Flip Chip Bonder Process Development).

Product Description
TP-1/-2/-3 are advanced ACF adhesives that have electrically/thermally conductive nickel fibers running through the Z-axis thickness of a thin thermoplastic film adhesive.  The nickel fibers have a thin electrical insulation coating that limits electrical conductivity to the Z-axis direction and allows 11micron pitch.  The adhesive is produced as a 50-200 micron (0.002-0.008 inch) thick film.  It is produced to a specific fiber volume.  The second dash number indicates the fiber volume (e.g. TP-2-40 has a 40% fiber volume).  The NTP-1/-2/-3 series use the same thermoplastic resins as the TP series, however the nickel fibers do not have a special extra electrical insulation coating and, thus, they can be used down to a 200 micron pitch limit.

The fibers have a small tilt that allows proper consolidation of the adhesive during bonding.  This is typically an angle of 5 degrees off the Z-axis that runs vertically through the thickness of the adhesive film.  To facilitate alignment of circuits to the adhesive, the direction of the tilt is shown on the release paper.  There is also a small 45 degree angle cut into the top right corner with the fibers tilting away from the corner along the right edge.  The observed tilt angle for a specific lot of film adhesive is shown after the lot number (e.g. Lot 14-175, 5° has an observed angle of 5 degrees).  If the tilt direction indicators are lost the best way to determine the tilt direction is to examine it under a microscope.  At 400X with through transmission light, a distinct shadow can be seen attached to the fiber ends in the opposite direction from the tilt.  For a reflected light microscope examination, slightly trim two edges that are 90 degrees to each other by placing a single sided razor blade over the adhesive film near the edge and taping the back of the razor to cut a clean edge.  When both of these edges are examined at 100X one edge will show the tilt direction with fibers having a very distinct tilt.  The other will show fibers straight up and down.

For small quantities the adhesive film typically comes between release papers clamped together in a package.  The sheets may be stacked on top of each other.

Storage
TP/NTP adhesive films are storable in a controlled environment at room temperature.

Like all thin adhesives, the adhesive may warp and pucker if stored free standing.  It should be stored between two sheets of release paper with a light dead weight on top or returned to the same clamping box it was shipped in with the original bubble wrap plies top and bottom.  Carefully putting the adhesive between two sheets of release paper and putting a dead weight on top normally eliminates any puckering or warping.

Handling
Read the MSDS before handling.  Powder free gloves should be worn to avoid contamination of the bond line.

The adhesive is subject to distortion and splitting if handled improperly.  The adhesive can be easily cut by tapping on a one sided razor placed over it.  This should be done on a clean surface such as glass, plastic or release paper.  Like all adhesive films it will pick up dirt from any surface it touches.

Experimental Bonding
The surfaces to be bonded should be clean enough to meet normal electronic assembly standards.  TP/NTP will accommodate the surface flatness of most common electronic assemblies.  Thicker versions of the film may be required for large bonding areas or surfaces with poor flatness.

TP/NTP has been designed to accommodate local intrusions (from flex circuit traces, die pads, etc.) up to 15% of the Z-axis thickness without compromising Z-axis conductivity.  For example, 100 micron (0.004 inch) thick film can be successfully bonded to a 15 micron thick trace (with a flat substrate on the other side) or to a 10 micron thick trace on one side and a 5 micron pad on the other side.  However these intrusions cause the nickel fibers to tilt over further so alignment compensation is needed for small pad connections. 100 micron (0.004 inch) thick TP/NTP with 5 degree fiber tilt has a 9 micron offset from top to bottom; that increases to ~40 microns when connecting flat pads and consolidating the bond line by 10% during bonding.  This offset increases to 65 microns when connecting a 15 micron thick trace (height above the substrate) to a flat (same level as substrate) pad.

The bonding process is a function of temperature and pressure and is affected to a lesser degree by heatup rate, surface area, surface flatness and substrate material.  Because temperature is usually the easiest variable to measure and control, we recommend starting your bonding experiments at a nominal point (such as 125-130°C, 50 psi for TP-1/NTP-1; 150-155°C, 50 psi for TP-2/NTP-2; and 185-190°C, 50 psi for TP-3/NTP-3) and varying the temperature at a constant pressure until suitable bonds are produced.  We recommend using two pieces of glass microscope slides to verify that full resin wet-out is produced at the temperature/pressure bonding conditions prior to experimenting with real test packages.  Higher temperature produces higher consolidation of the adhesive and would be the direction to go if incomplete bonding is observed.  The thermoplastic resin will start to degrade during bonding if heated above 150°C for TP-1/NTP-1; 170°C for TP-2/NTP-2 or 210°C for TP-3/NTP-3 (ok for solder reflow exposure).  Lowering the temperature will reduce excessive consolidation as observed by less squeeze-out of the adhesive around the bond edge.  If temperature can’t be varied easily, pressure variation can be used.  Again, higher pressure produces higher consolidation, lower pressure less.  The time it takes to reach temperature and the length of any dwell at temperature will affect how the adhesive consolidates.  Longer time means more consolidation at a given temperature and pressure.  A fixture that takes a long time to heat up will require a lower temperature or pressure.  Since the nickel fibers conduct lots of heat through the Z-axis the actual resin temperature can be somewhat lower than indicated by a thermocouple mounted at the interface.

A suitable bond will have wetted out both surfaces and won’t show squeeze-out of the adhesive around the edge.  This can be determined by: (1) bond line compression that is ~10% of the original adhesive film thickness (for example, 0.004” film will compress down to ~0.0036”); or (2) Z-axis electrical conductivity meets your requirement; or (3) die shear testing shows >500 psi bond strength.  Some applications may require much higher pressure (>300 psi) at ~140°C (for a TP-2/NTP-2 example) or, to avoid higher pressure, a higher bonding temperature (165-170°C for a TP-2/NTP-2 example).

The temperature and pressure should be applied for long enough to ensure that the adhesive film and both substrates reach the proper temperature range.  The bond line should be cooled 30-40°C before releasing pressure.  One-sided bonding (such as application to a wafer before dicing) can be accomplished by using commercial heated roller film laminating equipment or a heated vacuum press.  Sticking the adhesive to one of the substrates first is the best way to assure proper positioning.  This can be accomplished by putting the adhesive on release paper, putting the substrate on top, heating them on a hot plate (within 15-25°C of normal bond temperature) and applying light finger tip pressure to the substrate before removing it from the hot plate.  The adhesive should stick to the substrate enough that it can be trimmed around the edges if desired.

A.  Non-alignment Bonding

Bonding pressure for laboratory experimental applications that do not require precision alignment is most easily accomplished with a simple spring fixture with a block on the end (Figure 1).  A film such as Kapton is sometimes used under the block to protect the substrate.  We use this type of fixture in our own lab…due to thermal absorption it takes about 15 minutes to reach ~150°C bond line temperature (for TP-2/NTP-2 applications) and 3 minutes to cool down below 120°C.

Pressure application with a dead weight is not recommended because the weight must be much larger than the bond area to get enough pressure.  This leads to tipping of the weight, which creates a bond that is over compressed on one edge and debonded on the other.  Very large bond areas of >6 cm2 (1 in2) are best bonded with a hot press or heated roller technique.

            The bond is reworkable by exposing it to 15-25°C above the bond temperature and separating the bond line.  The film can be scraped off the substrates at this temperature or removed at room temperature with isopropyl alcohol.  New film must be used for a new bond.

B.  Precision Alignment Bonding

            Since TP/NTP thermoplastic films are not optically transparent the ability to properly align both sets of pads when using an opaque adhesive film is really the only unique assembly requirement.  Manual or automated heated flip chip bonders with split optical vision for blind alignment offer the best process control.  For example, SUSS MicroTec used their highly accurate FC150 machine for TP-2, they also make a higher thruput model FC250.  Semiconductor Equipment Corp Models 850 and 860 have also been used by TP-2 customers.

Split optical machine vision systems can be too slow for some production applications, so adding external fiducials to the substrate will enable the use of higher volume bonding processes while still maintaining the necessary precision pad alignment.

thermoplastic adhesive bonding

Section B.  Flip Chip Bonder Process Development for NTP & TP Series Adhesives 

Step 1 – Confirm Aequate Equipment Capability

  • Heated bond head (up to 200°C) with vacuum pickup
  • Bond head pressure capability >5 kg per cm2 of bond area
  • Heated work stage (up to 150°C)
  • Split optical alignment capability

Step 2 – Define Experimental Bonding Parameters

  • Use two thin (<400 micron) microscope slides cut to correct application size.
  • Initial bond trial:  for NTP-2/TP-2 work stage at 130°C; bond head at 175°C; for NTP-1/TP-1 work stage at 105°C; bond head at 150°C; for NTP-3/TP-3 work stage at 180°C; bond head at 225°C.  For all NTP/TP adhesive films apply 3.5 kg per cm2 (0.35MPa or 34N/cm2 or 50 psi) bond area for longer than 6 seconds.
  • Inspect the bond area……if properly bonded (1) there will be no visible voids on the two glass surfaces and (2) the bond line thickness will have decreased ~10% (for example from 100 micron to ~90 micron).
  • (1)  If voids are present then use the same test coupon with the voids and either increase the bond head temperature, or the work stage temperature, or the bond time.  Increasing bond pressure only has a secondary affect on eliminating voids.  For example, NTP-2/TP-2 thermoplastic resin can be heated to 170°C but, since the viscosity decreases rapidly with increasing temperature, bonding pressure may have to be reduced to avoid decreasing the bond line thickness too much.
  • (2)  If the bond line thickness has decreased too much then make a new test coupon and reduce the bond head temperature or pressure or the bond time.  If both the bond head and work stage are set at 135°C for NTP-2/TP-2 the pressure required to produce a void free bond dramatically increases to 7MPa (70 kg/cm2 or 1000 psi).

Step 3 – Bond Experimental Test Package

  • Use very thin (200 micron) microscope cover glass to simulate the actual flip chip die.
  • Use the same bonding parameters from Step 2 that created void-free/proper thickness bonds to bond the thin glass to a real substrate.
  • If voids are present or the bond line thickness has decreased too much then vary the temperatures and bond time until a satisfactory bond is produced.

Step 4 – Bond Real Flip Chip Test Package

  • Important – NTP/TP adhesive films have been designed to replace bumps and underfill.  Z-axis penetrations from bumps and traces will cause the nickel fibers to tilt over more and, for small pad connections, will require X-axis die offset for proper Z-axis alignment.  Higher bonding temperature and pressure will also be required to force the bumps and traces into the NTP/TP products for 100% void-free contact.  Please see Section A of this document for further details. 
  • Align the die and substrate, separate them vertically, ensure that the bond head and work stage are at the proper temperatures, place the NTP/TP preform on the substrate, apply the proper pressure and bond time, remove the vacuum bond head.
  • Use acoustical scanning to confirm lack of bond voids.
  • Use a micrometer to confirm proper bond line thickness.

Ideal Production Configuration and Process

  • Die is non-bumped, pads are a few microns higher than non-conductive surface.
  • Substrate conductive surfaces are a few microns higher than non-conductive surface.
  • NTP/TP film adhesive has been vacuum laminated to either the wafer before dicing or the substrate (for example flex circuits before singulating).  This will reduce the bonding temperatures and pressure required for final package assembly. 

Vacuum lamination equipment options:

  • heated vacuum press.....every PCB shop uses this type of equipment.
  •  vacuum drawer laminator (example Dynachem EN-2400, ~$65k new, used equipment available).....fairly common piece of equipment for wafer producers....more capability and more costly than needed for NTP/TP application.
  • air cooled laminator used for making solar panels.... www.eets.co.uk ~$35k new.   These machines are very simple so many companies build their own.


The basic vacuum lamination production process: 

  • tack a piece of NTP/TP to the wafer with a heated plate (outside diameter is slightly larger than the wafer) 
  • vacuum laminate the NTP/TP to the wafer 
  • run a scribe around the outside of the wafer to trim off excess NTP/TP and expose the dicing streets
  • dice with a 4-6 micron particle diamond saw (or use laser dicer)
btech termobonding powerpoint

 

 

home | data sheets | thermoplastic bonding | order samples | contact us | sales agents
news | links | site map