TECHNET Archives

1996

TechNet@IPC.ORG

Options: Use Monospaced Font
Show Text Part by Default
Show All Mail Headers

Message: [<< First] [< Prev] [Next >] [Last >>]
Topic: [<< First] [< Prev] [Next >] [Last >>]
Author: [<< First] [< Prev] [Next >] [Last >>]

Print Reply
Subject:
From:
[log in to unmask] (Roland Jaquet)
Date:
Sat, 27 Jan 1996 08:52:26 +0100
Content-Type:
text/plain
Parts/Attachments:
text/plain (271 lines)
For those of you who have been looking for Infos about E.D. 
I copied part 1/2 of the paper presented in EIPC.

I am interested to know your opinion about ED, Thank you

[log in to unmask]
Roland Jaquet
Tech Manager
Henri Jaquet SA
PCB Manufacturer of prototypes and small to medium batches
Geneva, Switzerland

____________________________________________________

PAPER 1/2 PRESENTED TO EIPC SEMINAR - BASEL, SWITZERLAND, DEC 5-6, 1995

Yield Improvement and Cost Reduction by Implementation of an
Electro-Deposited Positive Photoresist
(P.E.P.R. Positive Electro-deposited PhotoResist)

THE SUPPLIER TECHNICAL ASPECT
Mr. Alan Guttridge, Shipley, UK

 1 INTRODUCTION
Much has already been written about the theoretical benefits of introducing
positive electrodeposited (ED) photoresist processes into the PCB
manufacturing process. However, this paper approaches the subject from a
practical perspective and details the installation and operation of a new
Positive ED photoresist in an automatic line at a medium sized PCB facility.
The paper covers some of the basic theory of the process, but more
importantly, covers the site preparation, logistics, implementation, and
benefits, of installing a Positive ED system. 
Henri Jaquet is a medium sized printed circuit manufacturer based in Geneva
and specialising in fast turnaround low/medium volume medium/high technology
PCBs, both double sided and multilayer. 
Switzerland is a high cost manufacturing area and, although Henri Jaquet SA
has a very loyal customer base, there is continual pressure on margins.
Therefore methods of reducing costs and improving productivity and yields
i.e. remaining competitive, are continually sought.
Henri Jaquet SA employs all the usual methods any good PCB manufacturer
would use to ensure continuous gradual improvement and these methods will
continue. However, a strategic decision was made to make a step change in
capability, in an effort to position the company at the leading edge of PCB
manufacturing technology.
A major part of this step change, although by no means the only part, was
the introduction of a Positive Electrodeposited PhotoResist from Shipley.
It is fully appreciated that there are few circuit designs in circulation at
the moment which require a Positive ED resist, but this is largely because
for the most part they either cannot be manufactured, or the yields are so
low when made by conventional techniques as to make manufacture uneconomic.
Henri Jaquet SA, being a fast turn around prototype PCB supplier, has an
advantage in that it is in regular direct contact with designers who are
looking to the future and can therefore discuss, agree and manufacture
advanced designs very quickly. Once Positive ED technology is a proven
manufacturing tool, new designs, currently not manufacturable with Dry Film
will become common place.

2 FUNDAMENTAL CHEMISTRY OF ED RESIST
Liquid positive working resists based on diazonaphthoquinone chemistry have
long been dominant in the semiconductor industry and are now gaining PCB
acceptance for high volume inner layer manufacture via roller coating methods.
Shipley PEPR 2400 resist uses similar chemistry in a colloidal form that
enables coating by electrodeposition. This coating method allows its use as
a high resolution photoresist for outerlayer circuits as well as for inner
layers.
Electrodeposited resists are supplied as aqueous emulsions, the discrete
micelles (<200nm) each holding all the components of the resist system.
These micelles are suspended in an aqueous media and are stabilised by the
presence of a surface charge. In the presence of an electric field, micelles
migrate toward and coat an electrode.
Electrodeposition can occur at either the anode or the cathode, depending on
the surface charge on the micelle. Anodic formulations, such as the PEPR
2400 resist, have negatively charged micelles and coat onto the anode.
Electrodeposition of an anodic photoresist is an electrochemical process,
comprising three distinct stages: electrophoresis, electrolysis, and
electro-osmosis.
Electrophoresis describes the movement of charged micelles under the
influence of an electric field and is the first stage of the process in
which the resist is transported to the suitably charged substrate to be
coated. Electrolysis of water occurs at the same time as electrophoresis
generating hydrogen and hydroxide ions at the cathode and oxygen and
hydrogen ions at the anode (the substrate). The hydrogen ions generated
neutralise the surface charges on nearby micelles causing them to coalesce
onto the anode surface forming a resist film.
The initial film contains water and occluded ions, which are driven out of
the film by electro-osmosis which also compacts the film and increases its
insulating properties until further electrophoresis and electrolysis stop.

3 PRACTICAL IMPLEMENTATION OF ED RESIST TECHNOLOGY
3.1 Site preparation for introduction of a Positive ED Photoresist
While positive ED resists are primarily designed to replace Dry Film in
panel plate, tent and etch operations, we decided that for logistical and
productivity reasons we would use one resist both for inner and outerlayers
(Positive ED). Therefore, we decided to begin by converting the innerlayer
process to Positive ED. This offered two major benefits:
a) Initial tests could be carried out to define and fine tune the process
using AOI with relatively inexpensive innerlayer material.
b) The logistics of converting from pattern to panel plating and at the same
time introducing a new resist process whilst continuing to meet the tough
delivery schedules of Henri Jaquet SA customers were NOT very appealing. 
The overriding objective was to continue to service the final customer with
on time deliveries while introducing major manufacturing changes and also
training staff. With protecting the customer as our major objective we set
about planning the conversion.
A production size ED coating line is larger and at first sight seems more
complex than a cutsheet laminator however, it is, in essence simply a
plating line which is common to all circuit producers. 
In order to accommodate this line, various pieces of production equipment
had to be moved and a 'clean' room built around the coater. The area in
which the line is situated is rated at class 100,000 and panels are put
directly through a hatch into the imaging area, which is at a class 10,000.
The line illustrated here is capable of producing approximately 70 panels
per hour of 450mm x 610 mm ranging in thickness from 0.15mm to 3.2 mm.
Panels of various thickness and sizes can be processed at the same time; a
significant benefit for a company producing fast turnaround small batches.
Another major change required for the Positive ED resist to give optimum
performance is the conversion from ammoniacal to cupric chloride etching.
Whilst Positive ED resists operate with ammoniacal etchants, the operating
window is narrower than for cupric chloride. In order that the transition
ran as smoothly as possible, this conversion was made first to the
innerlayer Dry Film line. This meant that experience and confidence in acid
etching was gained prior to the introduction of the ED resist.

3.2 The Positive ED Coating Line
The line which Henri Jaquet SA have installed was built by STS and
incorporates a sophisticated software package which monitors, records and
controls all aspects of the process. Control of the process is via a PC
located at the input end of the line. A screen of process step is available,
with the status clearly visible, i.e. temperature, solution level, valves
open or closed etc...

3.3 Manufacture of  PCB’s using Positive ED Photoresist
As stated earlier, a major concern was to protect customers during the
changeover to PEPR.
To this end a test vehicle was designed and run through the Dry Film
process, where AOI data was collected.
The design of this test vehicle had to be thought through very carefully, as
an initial temptation was to fill it with 75 micron lines and spaces to test
the performance of the Positive ED resist. However,  had we used this design
as the benchmark yield on the current process (the starting point for
conversion to Positive ED) it would not be meaningful. A design which gave a
first pass yield through AOI in the mid to high 80% region was employed as
this reflected a realistic production target and previous data suggested
that it was repeatable. Experience also suggested that using more complex
and dense fine line and space designs would have given very variable first
pass yield data making it difficult to set a meaningful baseline.
This first pass yield at AOI was to be our baseline for conversion to ED,
i.e. we would run this same test vehicle through the ED process until we
achieved at least the same first pass yield as for Dry Film, at which point
a full conversion would be made, in the knowledge that the risk to
manufacturing, and thus the final customer was minimal, - we had the data to
support the transition.
In order not to interrupt normal Dry Film production, a second developing
machine was set up adjacent to the existing production machine. This machine
was to be used for characterising the ED process.

3.4 Setting up the process
Surface preparation
For the surface preparation of both inner and outerlayers, and in common
with most other manufacturers both brushing and chemical cleaning techniques
are used with Dry Film. It was a goal of the project to remove this step if
at all possible, thus shortening the production cycle and, perhaps more
importantly, reducing handling to eliminate a major possible cause of yield
loss.
Whilst ED coatings can exhibit better overall adhesion than Dry Film, the
cleanliness of the surface prior to the application of ED is critical,
although the actual topography is not thought to be so important as is the
case with Dry Film.
Tests were carried out using brushed, chemical cleaned and ED line cleaned
panels, where one side was insulated so that no coating took place. After
coating and drying, the panels were processed through the etching machine
and subjected to a light box inspection for pinholes.
No pinholes were present from any of the three pre-clean techniques.
Therefore it as felt that innerlayer panels can be taken directly from the
laminate store and processed through the ED line without any previous treatment.

Coating & drying
Given that the pre-clean was as discussed above, the coating process was
quite simple and fast and takes place within about 10 seconds, with
thicknesses of 5-7 microns sufficient to provide a hard pinhole free
deposit. The design of the ED tank allows for solution sparging from both
the sides and the bottom of the bath. Solution movement did not have a major
influence on the thickness variation across the panel although it did affect
uniformity of appearance. Some adjustment of the solution flows was required
to achieve a coating with uniform appearance but this was not too difficult
or time consuming. To a large extent the drying stage was the most critical
part of the process with respect to exposure time, resolution and adhesion.
It must be remembered that the resist is not simply dried, but it is this
stage which give the deposit its final photoresist characteristics, and
careful design of the ovens is essential.

Exposure
Exposure energy requirements for Positive ED resists are higher than for Dry
Films, but the indications were that there is a larger degree of exposure
dose latitude than with Dry Film .
While it is possible to expose a Positive ED resist readily with a standard
5kW exposure unit, and indeed this is the practice at Jaquet, in terms of
productivity a 7 kW unit would be preferred. The major influences on
exposure were drying of the resist after coating and the thickness.
Thickness is monitored by Beta Back Scatter and variations of only ± 0.75
microns can be achieved on a nominal 6 micron coating. This variation is not
sufficient to cause any problems with respect to resolution or development.
There are two distinct aspects to drying; the first is to physically drive
moisture out of the film, and the second is to fully coalesce the micelles
to form a true homogenous hard film of photoresist. The temperature profile
of the oven was important, however in practice the best method of
determining the correct drying profile and cycle was to set the approximate
parameters and process a panel through the system. It was then exposed and
developed using resolution targets and step wedges in various parts of the
panel. By analysis of the results, decisions could be made as to parameter
changes.

Developing
Developing  can be carried out using sodium carbonate, as with Dry Film, or
sodium hydroxide, as is common with other positive resists. Sodium hydroxide
was preferred because it provided a faster development than sodium carbonate.
Assessment of development quality was made by three measures.
a) Resolution.
b) Unexposed resist loss.
c) Step tablet.
Unexposed loss was calculated by measuring the coating thickness after
drying, using Beta back scatter, placing a strip of tape over an area of the
panel and passing through the developing machine, and re-measuring. If the
developing parameters were set correctly the panel should be passable
through the developer 4 times with no more than a 10% loss in unexposed resist.

Etching
Etching is made via cupric chloride and, as in common with most thin liquid
photoresists, will be up to 30% faster than with Dry Film. Whilst this is
not necessarily an advantage with respect to throughput, as in most
facilities etching is not a bottleneck, it is a considerable help in
reducing undercut because of the significantly reduced residence time in the
etchant.

Resist Removal
Proprietary resist removers were not necessary when using PEPR 2400 resist,
since sodium hydroxide, at a higher concentration and temperature than used
for developing, is perfectly satisfactory.

4 BENEFITS OF ED RESIST TECHNOLOGY
4.1 Benefits to manufacturing
1) Less handling, therefore reduced handling damage; the largest cause of
yield loss in most PCB facilities.
2) Increased process capability meaning that existing technology can be made
with higher yields and « state of the art » parts become manufacturable.
3) Reduction in waste from the fact that there is no Mylar removal
necessary, the resist itself is much thinner than Dry Film therefore less
developer and remover are used, and lower resist solids means that there is
less volume of organics to waste treat.
4) As panel plating is used, track heights are more uniform making
soldermask application easier and more cost effective.

4.2 Benefits to customer
The real benefit to the customer is that boards can be designed which are
currently not manufacturable, and no longer will he be able to say that
board manufacturers are holding back the advance of technology,

5 THE FUTURE
As stated in the first part of this paper, landless fine line designs are
already receiving serious consideration and designers are becoming
increasingly confident that they can be manufactured economically and with
good reliability. We feel that the ideal system would be for the Positive ED
coating line to be an extension of the panel plating process, and for
automatic exposure and developing to be incorporated. The ultimate process
would enable the boards to be put on the plating line after drilling and
removed after resist stripping.

 



ATOM RSS1 RSS2