Please use this identifier to cite or link to this item:
|Title:||Laser structuring of ultra-fine circuit lines in printed circuit boards|
|Subjects:||Hong Kong Polytechnic University -- Dissertations|
Printed circuits -- Mathematical models
|Publisher:||The Hong Kong Polytechnic University|
|Abstract:||Laser structuring technique emerged in recent years for the need of fabricating fine circuit lines and spaces in printed circuit board. Most of the previous work only introduced laser structuring as a new method in the fabrication of fine circuit lines and mentioned that the width of circuit line can be reduced under 50 μm or below with this technique. Laser structuring technique will have a prosperous future only when the relationship between process parameters and fabrication results are deeply understood. In this work, the study focused on the control, optimization of process parameters and the prediction of circuit geometry in laser structuring technique. Also, the minimum width of circuit lines and circuit spaces fabricated with laser structuring technique that can pass quality and reliability tests were investigated. In laser structuring of circuit lines (and circuit spaces), there are a number of process parameters. This study firstly focused on investigating the influences of the process parameters in the three main steps of this technique - the tin electroplating step, the laser writing step and the copper etching step. The effect of electroplating time and current density on the thickness of the tin layer in the tin electroplating step, the effect of laser power, repetition rate, number of repetitions and laser bite size of a frequency-tripled Nd:YAG laser (355-nm wavelength) to the geometry of the circuits in the laser writing step, the effect of time and circuit geometry (after the laser writing step) to final geometry of the circuits in the copper etching step were investigated. From these investigations, relations between the complex process parameters to the fabrication results in laser structuring have been identified. Excimer laser (KrF, 248 nm wavelength) was used as another laser source in laser structuring. Excimer laser shows its advantages when the total width of circuit line and space was as fine as below 60 μm, which cannot be achieved by Nd:YAG laser. When the total width of circuit line and circuit space was above 60 μm, it was more efficient to use Nd:YAG laser as laser source whose scanning velocity was about 120 times faster than Excimer laser.|
Laser writing is the key step in the fabrication of fine circuit lines and circuit spaces in laser structuring. The laser parameters in laser writing step will directly influence the geometry of the circuit space. A set of experiments was designed using the Taguchi methodology, in which all parameters of interest are varied over a specified range, to obtain the quantitative estimations of the various parameters and the main parameters affecting the geometry of the circuit space. According to the analysis of variance (ANOVA), the main characteristic parameters that influence the geometry of the circuit space - depth of space, width of space and resolidification height - were investigated. Laser power and number of repetition are the most important parameters in the laser structuring process that have influences on all the experimental results-the space width, the space depth and the resolidification height. Some other parameters should also be considered in the experiment process according to the ANOVA analysis such as repetition rate and bite size. A model of this non-linear system (laser writing step) might help the industry on the applicability of the laser structuring technique. Artificial neural network (ANN) is a useful tool for predicting outcome results and optimizing the operation parameters of the aforesaid non-linear process. ANNs are the fist time to be used in the study of laser writing step in laser structuring technique. Two artificial neural network (ANN) models were designed - ANN prediction model and ANN optimization model. The ANN prediction model is used to predict the geometry of the circuit space and the ANN optimization model is used to select optimum process parameters that can achieve an acceptable geometry for the circuit space in the laser writing step. In the ANN prediction model, the errors between the prediction results and experimental results were within ten percent. The optimum results from the ANN optimization model also matched the experimental results very well. With a new approach, by combining the characteristic parameters of circuit space predicted by the ANN prediction model and the mathematical description of the Gaussian profile, the geometry of the cross-sectional profile of the circuit space with the consideration of resolidification height in the laser writing step can be calculated. After the laser writing and copper etching steps (tin layer was also stripped away), the final circuit lines and circuit spaces fabricated were tested using the quality and reliability tests - electrical open/short test, peel test and surface insulation resistance test (SIR test). The minimum width of circuit lines and spaces that passed the open/short test was as narrow as 25±5/40±5 μm. From the investigation, it was found that the variation of the width of the circuit lines and spaces was caused by the bite size of the laser, a splashing problem of melting materials in the laser writing step and a tin-layer cracking problem caused by etching in the copper etching step. When the variance of the widths of the circuit lines is ±5 μm, the variance of the width of the circuit line is just matching the IPC standard (IPC 6011) of ±20%. The copper circuit lines are found to be separate from the dielectric layer when the widths of circuit lines are under 10 μm. So the adhesion quality of circuit lines fabricated by laser structuring is acceptable only when the width of circuit lines is above 10 μm. When the widths of circuit spaces were above 45 μm, the resistance between isolated copper lines was above 10⁶ ohm from the beginning to the end of the SIR test. The minimum width of circuit spaces that can pass the SIR test was 45 μm. In general, the minimum widths of circuit lines and circuit spaces with good quality and reliability fabricated by laser structuring were 25 μm and 45 μm respectively. The project is significant for both applied and academic fields. This study contributes to the understanding of the laser structuring technology and is of benefit in the fabrication of very fine line circuits in advanced printed circuit board industry.
|Description:||xxvi, 241 leaves : ill. ; 31 cm.|
PolyU Library Call No.: [THS] LG51 .H577P ISE 2007 Zhang
|Rights:||All rights reserved.|
|Appears in Collections:||ISE Theses|
PolyU Electronic Theses
Files in This Item:
|b20940518_link.htm||For PolyU Users||162 B||HTML||View/Open|
|b20940518_ir.pdf||For All Users (Non-printable)||8.38 MB||Adobe PDF||View/Open|
All items in the PolyU Institutional Repository are protected by copyright, with all rights reserved, unless otherwise indicated. No item in the PolyU IR may be reproduced for commercial or resale purposes.