10.1.1.1 Fluxing
Where solder paste is not used for flip chip attachment purposes, the use of flux becomes necessary,
as reflected by the "flip chip fluxing and placement" step in the simplified flip chip attachment
process flow chart (see Figure 10.3). This fluxing step can be used for each combination depicted in Figure 10.2. The detailed processes and options are described below.
Dip
A schematic of dip-flux process flow is shown in Figure 10.4. At first, a creamy flux is applied to a rotating disk equipped with a doctor blade. The thickness of flux film, e.g. 50 µ, is controlled by the clearance of the doctor blade (see Figure 10.4(a)). The board now enters the placement tool. Depending on the design, the board may already have SMDs placed on the printed solder paste. The die is picked from a feeder medium, such as waffle pack feeders, tape and reel, surf tape or a direct wafer feeder, and is then imaged and centered by a stationary flip chip camera or an on-board camera in the placement head chassis. Afterwards, the chip is brought to the dip flux module (see Figure 10.4(a)) and dipped into the flux film for a preset amount of time (see
Figure 10.4(b)).
Once dipped, the chip with the bottoms of solder bumps covered by flux is brought to the board
(see Figure 10.4(c) and 10.4(d)) and placed onto the corresponding pads (Figure 10.4(e)). The assembled device is then reflowed in an oven, typically in an inert atmosphere.
The solder wets to the pad to form the joint and self-aligns (see Figure 10.4(f)). Figure 10.5 shows a dip-flux device [4].
Imaging the flip chip prior to dip fluxing has the disadvantage that the mechanical contact of the flip chip bumps with the flux carrier can have a negative impact on placement accuracy. Alternatively, the flip chip imaging step may be carried out after the dip fluxing step. There is a very slight risk that the optical bump image can be adversely influenced by the flux material. Dipping prior
to imaging seems to be the preferable sequence [4].
To avoid the possibility of placing flip chips in the wrong orientation, the programming bump pattern used for recognition should be asymmetric. The minimum flux film thickness depends on the bump height variations within a die. To ensure good soldering of the solder bumps on the die, all the bumps have to be dipped in the flux.
This principle of dip flux is most suitable for high viscosity fluxes. The amount of flux involved in the process is brought to a minimum by fluxing only the bump's underside. This is particularly important for the no-clean process. Dip fluxing is not adequate for fluxes with a high
evaporation rate [4-6].
Spray
The process flow for spray fluxing is shown in Figure 10.6. The PCB is placed in the sprayer sample stage, then sprayed with flux. After spraying, the flip chip is placed on the pads. Depending on the flux solvent system, a period of time is allowed for the volatile solvent to evaporate at ambient or slightly elevated temperature. The assembled board is then sent through the reflow oven
for soldering.
Like dip fluxing, spray fluxing is one of the two most commonly used methods. This sprays a mist of flux over the footprint area. Figure 10.7 shows a flux jetting system
for spray applications [7]. This system can apply a flux quantity of 1-2 µg per mm2 , with 5 percent volume consistency. The throughput is 1500 units per hour.
Figure 10.8 shows a coaxial flux jetting system. This employs a coaxial air column to further force the flux droplets landed on the board to spread out and form a uniform film [7]. It prefers fluxes with a viscosity of 7-30 cps and applies 4 µg per mm2flux on board. Surface tension of the substrate surface is important for control of spray quality.
Most fluxes used for spray fluxing are low solid content fluxes with an alcohol content of 95-98 percent. The fluxes are sprayed on to the substrate before placement and the alcohol evaporates rapidly at room temperature. The remaining flux, when properly formulated, can provide sufficient tackiness to hold the flip chip in place during board handling and reflow.
Brush
The process flow for brush fluxing is almost identical to spray fluxing. The only difference is that the spray action is replaced by brushing. Figure 10.9 shows a brush fluxing setup. A low viscosity flux, such as 80cps, is stored in a reservoir, The brush is wetted by the flux fed through a small tubing with an outlet hidden in the interior of the brush. Upon brushing, the brush sweeps through the flip chip footprint area along a programmed path and deposits a layer of flux. In general, the flux quantity applied is greater than that by spray fluxing.
Dispense
Again, the process flow is identical to spray fluxing. Figure 10.10 shows an example of dispense fluxing setup [7]. Dispense fluxing utilizes the same principle as spray, except that control of volume and flux film formation is poorer.
Stamp
Stamp fluxing is more similar to brush fluxing in principle. The stamp is made of a spongy rigid material and is wetted by the flux fed through the back of the stamp. Figure 10.11 shows a pad/stamp headset.
excerpted by permission from Chapter 10 in
Reflow Soldering Processes and Troubleshooting: SMT, BGA, CSP, and Flip Chip Technologies
by Ning-Cheng Lee Copyright 2002, Newnes, Boston, MA
An Imprint of Elsevier Science
|
