Improved processes and dry film resists are meeting the ever-increasing demands of high-tech PCB manufacturing.

Laser direct imaging (LDI) technology has developed to the point where there is little doubt that it can be successfully implemented in large-scale production. Enabled by high speed and specially optimized dry film photoresist compositions, more than 120 systems are now in operation globally, giving some significant benefits to its users. Solid state laser technology has also made operation less costly than for the early gas laser systems. Reliability and performance has improved at the same time. A key benefit of LDI is production flexibility because there is no penalty for running small series vs. longer series. There is essentially no set-up time. The image to be printed is wired straight into the computer "brain" of the laser imager, and is available for printing on the photoresist virtually instantly.

LDI can also enable certain technologies based on its ability to very accurately place images on the board. The image placement tolerance for the Orbotech Paragon is +/- 0.0005", which makes it ideal for sequential or selective imaging of features such as buried resistors or outerlayer selective finishing. Specially designed LDI dry film photoresists are now available in thicknesses up to 100 µm, making it possible to laminate resist over circuitry, or conform to surfaces with very high profile for selective surface metallization.

Figure 1
FIGURE 1. The new LDI dry film resist Riston LDI530 extends the use of LDI down to 30 µm lines/spaces with standard equipment.

LDI dry film resists have been improved significantly over three generations in terms of photospeed, resolution and process latitude. The benefit vs. conventional dry film resists are mainly in very high photospeed, now requiring an energy of less than 10 mJ/cm2 for LDI films vs. typically more than 30 mJ/cm2 for conventional films. The key secondary benefit of consistent sidewall quality and proper polymerization through the full thickness of the resist is also significant. This is especially true for plating applications where an improperly polymerized resist can cause excessive leaching of organic material into the plating bath. Line width consistency can also be affected negatively by improper polymerization. Resistance to plating and etching chemicals is negatively impacted by low grade of polymerization at the base of the resist sidewall. The resolution of an LDI system (image and resist) depends largely on the addressability and spot size of the laser hitting the resist surface and the resolution properties of the photoresist. The newest LDI resist formulation is capable of extending the use of LDI technology towards 30 µm features using standard equipment (Figure 1).

Semi-Additive Processing

Semi-additive processing has been in use in Japan for some time to enable production of advanced IC substrates. It can be seen as a solution to etch factor issues typical when conventional subtractive technology is used to make fine lines below 40 µm features. However, a significant benefit also lies in the uniformity of conductors created by semi-additive processes. A conductor created with the conventional subtractive process (Figure 2) is much less square than the conductor cross-section created by the semi-additive process (Figure 3). A conventional cross-section has more of a conductor "foot" created by the isotropic nature of the typical subtractive etching process. The size of this foot can be hard to precisely control and cause problems for boards that need tight tolerances such as for exact impedance control and boards that need exceptional signal integrity.

Figure 2
FIGURE 2. Cross-section of conductors created using conventional subtractive etching process.

Figure 3
FIGURE 3. Cross-section of conductors created using semi-additive process.

Specially designed dry film photoresists have been developed for semi-additive processing with high resolution down to below 15 µm lines/spaces using 25 µm thick resist. These resists have wide process latitude to enable very precise line width control over the working range (Figure 4).

The desired image from the phototool is reproduced with stunning accuracy on the photoresist over a wide range of exposure (Figure 5). Less than +/- 1 µm variation on a 20 µm line is achieved over 10 Stouffer Steps (Stouffer 41 StepTablet).

Figure 4
FIGURE 4. New dry film resist Riston JSF for semi-additive process provides resolution of less than 15 mm lines/spaces over the exposure working range (100-150mJ/cm2).

Figure 5
FIGURE 5. Phototool reproducibility for Riston JSF showing line width vs. grade of polymerization using Stouffer 41 StepTablet.

The semi-additive process itself starts with a high-density inner layer core - a dielectric layer is applied to the core and microvias are formed (Figure 6). A thin (2-3 µm) electroless copper layer is then plated; to this, photoresist is applied and imaged. Copper is electroplated (20 µm) after which the resist is stripped. A differential etching process is then used to remove the electroless plated copper, leaving approximately 19 µm of copper on the surface.

Figure 6
FIGURE 6. Schematic of a typical semi-additive process.

Figure 7
FIGURE 7. Final 15 µm lines/spaces.

The final characteristic that is important in a semi-additive dry film resist is clean stripping. Removing the resist cleanly is key to achieving the highly uniform plated 15 µm lines and spaces shown in Figure 7.

Dry Film, Higher Yields

High yield is critical for achieving lowest possible cost. Good conformation of the dry film resist to the base material during lamination is vital if the fabricator is to achieve high yields. If the dry film resist does not conform to the surface, nicks or open defects increase dramatically while the etching solution seeps in under the resist. Improved conformation of the resist also improves resist adhesion, while the contact area between resist and surface is increased. Conformation and adhesion become more of an issue as the industry moves towards thinner dry film photoresists to achieve higher resolution and to reduce the etch channel aspect ratio.

Various approaches can be taken to improve dry film conformation and adhesion. Wet lamination of dry film photoresist is not new, but is growing significantly in popularity as cost and technology demands increase. Wet lamination primarily helps reduce defects attributed to poor dry film conformation. It also enhances dry film adhesion.

Another key to the renewed success of wet lamination is the reliability of the equipment, the wetting station attached to the dry film laminator. There are currently about 140 installed so-called YieldMaster wet lamination units in operation. The wet lamination process has demonstrated significantly improved yields through a reduction in conductor opens/nicks. The process has also demonstrated improved CpK values for conductor width capability (Figure 8).

Figure 8
FIGURE 8. Conductor width Cpk values for wet lamination vs. conventional dry technology.

Figure 9
FIGURE 9. Viscosity vs. temperature for a newly developed dry film photoresist compared to a standard resist.

Further significant advances in resist conformation have been made in the formulation of the dry film itself. Good balance must be reached between resist flow during lamination and stability of the photoresist under storage conditions. Ideally the product should be formulated to achieve low viscosity at lamination temperature under high shear stress, with high viscosity at room temperature with low/no stress. Figure 9 shows the rheology curve for one of the latest high-tech photoresists. The flow characteristics of this dry film resist improve conformation while maintaining storage stability and quality.

Additional advances in new dry film formulations include improved resist flexibility to meet the need to process less rigid materials. There are patented clean formulations that don't leave sludge behind in developer or stripper equipment. New thinner resists with clearer coversheets enhance resolution and etched line definition while maintaining yields by offering the same good conformation as previous generation thicker resists, and more is in the R&D pipeline.

A number of such offerings in the field of dry film photoresist technology are available in the market, and future offerings are expected to continue this trend. PCD&M

Mats J. Ehlin is sales and marketing manager for DuPont Printed Circuit Materials (Research Triangle Park, NC). He can be reached at This email address is being protected from spambots. You need JavaScript enabled to view it..

REFERENCES

  1. Cox, G.S., "Designing Photoresist to Meet Industry Needs With Quality Functional Deployment," CircuiTree, December 1998.
  2. Cox G. S., "Wide Exposure Latitude Photoresists and Fine Line PWB Manufacturing," IPC Printed Circuits Expo Technical Conference, Long Beach,1999.
  3. Custer, W., Custer Consulting, various magazine publications and ongoing market information.
  4. Dietz, K.H., "Techtalk, Fine Lines in High Yield (Part LXVIII)," CircuiTree, May 2001.
  5. Ehlin M.J., Pressley E.G., "Wet Lamination is Making Waves," PC FAB, April 2001.
  6. Ehlin M.J., "Leading Edge PWB Manufacture - Current Technology Limits and Solutions for Volume Production of Fine-Line PWBs," ABRACI, Sao Paolo, Brazil, September 2004.
  7. Feng T., Lien T. Chiang P., "Outer Layer Fine-Line Pattern Manufacture Using Negative-Type Dry Film Photo Resist," CircuiTree, April 2002.
  8. Maurischat, Klaus, "Riding the New Roller Coaters," PC FAB, November 2000.
  9. Orbotech Ltd., "DP-100SL Direct Imaging System," Technical Data 2005.
  10. Suess T., Wu S.K., Dietz K.H., "Wet Lamination of Dry Film Photoresist," PC FAB, December 1996.
  11. Takahashi, T., "New State of the Art Dry Film technology for Fine Lines in High Yields," IPC Printed Circuits Expo Technical Conference, Long Beach, 2002.
  12. Uno, H., "The Development of Dry Film Photoresist with 15µm Lines/Spaces Resolution for Semi-Additive Processing," Electronic Circuits World Convention, Anaheim, February 2005.
  13. U.S. Patent no. 6,180,319.
  14. U.S. Patent no. 4,069,076.
  15. U.S. Patent no. 4,378,264.
Submit to FacebookSubmit to Google PlusSubmit to TwitterSubmit to LinkedInPrint Article