Vacuum Web Coating

Question concerning the barrier measurement vs sampling position in a slit roll.

We metallized OPP in Aluminium deposit chamber and I observed a variation in WVTR and OTR data with sampling position in the slit roll :

1. Why barrier seems to be better more deeper in the roll than near the surface ?

2. Where is the best location to take a representative sample : on the roll "out-of-chamber" or on the slit roll ?

Answer from CAB.   

Polymer film as it is wound over rolls will build up a static charge that can attract airborne dust to the surface.  This surface dust will get wound into the roll of film & will be carried through the vacuum coating process. Any dust that is coated may get moved after the coating process & leave behind an uncoated area called a pinhole or pin window.

Now every time the roll is unwound & rewound it can attract more dust each time. Thus from the film line the mill roll may be slit on the production machine into two to make two half mill rolls, each of these will then get unwound & rewound as they are converted from ½ mill rolls into whatever size rolls you are using.   The areas of greatest contamination are likely to be the beginning & end of the rolls as the speed is accelerated & decelerated between stationary & full winding speed. 

This extra contamination at the beginning & end of each roll will be cumulative.  Each machine may well accelerate/decelerate the rolls at different rates & so there is likely to be a different length of higher contamination from each process. 

Thus when it comes to deciding where to take measurements from it is worth going back to each process & estimating how much film is used before the speed is constant and how much film is used on the slow down to stop.  This will give you some idea of the minimum distance you need to be into the roll before you might get consistent, representative results.   It is also worth considering that surface contamination can be transferred from one surface to another & so these acceleration/deceleration distances are a minimum because it is possible that debris from these regions can be spread further into the roll during winding.

The greater the number of time the roll is wound the more surface contamination will be generated & the further into the roll it can have moved.

Thus, it is always better to minimise the amount a roll is handled (rewound) wherever possible.

Question re. plasma treater

I am a student attending the University of Waterloo. I am now on my work term working for a company producing BOPP.

This company has a plasma treatment unit (in a vacuum)(within a metallizer [for 250 cm rolls]) that does not function well. Due to some software problems, we can only restrict the Plasma Power up to around 2kW (maximum when fixed: around 11kW). Plasma Power exceeding around 2.5kW will cause the arc value to increase significantly (which is undesirable).

My job is to optimize the effectiveness of plasma treatment (in a vacuum) process with a constant gas composition being of 80% Argon and 20% oxygen (in one cylinder). Are the only other parameters from which I can alter, record and graph against optical density, OTR, and WVTR values:

gas flow rate
plasma power (within the limit)
vacuum level (If so, how will it effect the process?)

Moreover, to test for the effect of the plasma treatment process, what properties, other than optical density, OTR, and WVTR can we measure to get QUANTATIVE results?

Any other tips?   Please advise.

My response is in the continuation, please feel free to add your comments too.

Follow-on questions.

Follow-on questions.

Few more queries are :

1.       Could you please explain the about tacky roll which you are recommending to use prior to metallization.

2.       Is there is any online test method which can help to tell % contaminated film or the amount of contamination so that we get assurance to use a pre-treatment prior to metallization.

3.       What are the economical process that we can use prior to metallization for cleaning film surface ,is plasma treatment is sufficient for it or how we can judge which roll needs pre treatment.

4.       If our metallized film passing 610 pressure sensitive tape test then can we assure our customers that bond adhesion between coated layer & film is sufficient for any further treatment.

5.       What type of equipment we can use to eliminate aluminium dust which forms cloud after opening metallizer chamber.

The answers are in the continuing section.

On-line monitoring of PVD coatings

The question was as follows.

How would you measure in-line the thickness of a PVD coating (knowing its optical characteristics)? Is it possible for a roll-to-roll process? Do you need to have a stable distance between detector and coil? Do you know people who do that?

            If the optical characteristics are known then it is usually a matter of choosing the appropriate method or region to give the necessary sensitivity to control the process.

            The coil & detector can only measure the conductivity of the coating and the optical properties are inferred from this.  This usually requires a calibration graph generated from coatings produced from the same process.  If the coil & detector are on different sides of the web then the measurement is less sensitive to the position of the web between the coil & detector.  However if the coil & detector are on the same side of the web & the measurement is made in reflection then the distance becomes critical & variations can become a large source of error.  Hence it is common for the measurements, if made in reflection, to be made with the web passing over a non-conducting roll.  Passing the web over the non-conducting roll precisely defines the position of the web & so the coil & detector (usually made concentric to make the head compact) can be placed close to the web without the risk of contacting the surface.  The precise positioning including being able to position the head close to the web surface allows the signal to noise to be improved and hence the accuracy & reproducibility of the measurement to be good.

            If you are depositing a non-conducting coating then it is necessary to look for an optical measurement.  Often a simple transmittance measurement is relatively insensitive and so it is common to look for a wavelength range where there is a change in performance that is more sensitive to process changes.

            An easy one to look at would be the deposition of silica.  Silica is used for transparent barrier coatings, low refractive index layers in optical stacks, anti-reflection layer, etc.  Typically there are some trade-offs that are used in depositing silica.  The most common one relates to the cost.  The deposition rate & source material cost can be affected by the source material stoichiometry.  Silica coatings can be produce in a variety of stoichiometries from SiOx where x can vary from 1 to 2.

The performance of these coatings will vary both in barrier performance as well as colour.  The colour derives from the absorption at the blue end of the visible spectrum causing the coatings to appear yellow.  Thus rather than scan the whole visible spectrum it is possible to have a single wavelength monitored that will be highly sensitive to the stoichiometry.  This single wavelength measurement may not give a thickness measurement if only a single layer is being deposited. This may need a full wavelength scan to view the interference fringes to give a measurement of thickness. Also if the coating is very thin if will not produce any fringes & so thickness will not be easily measured.

            For thin metals there will be a plasma edge that can be measured in the same way. The sharpness of the edge is related to the deposition rate, coating quality & thickness and so using measurements at carefully chosen wavelengths will give information on the coating quality as well as optical performance & the optical performance can be related to the thickness of the coating using a calibration graph.

            Similarly for multilayers choosing specific wavelengths that relate to the desired coating performance can be used to control the coating performance & thickness.

           With optical measurements the sensitivity of the measurement is less affected by the distance of the source & detector but can be affected by changing light levels from extraneous sources.  This can include the light from any process plasma as well as light arriving from outside the system via the view ports.  The light from outside can vary (differences in sunlight, view ports being obscured or not, etc) and so it is more important that this source/detector system uses techniques to eliminate this source of potential error.

            If you want a contact that could help then try the following.

Mr. Tim EMMERICH
NAGY Messsysteme GmbH
Tel: +49-7032-76670 * Fax: +49-7032-72189
emmerich@nagy-instruments.de
http://www.NAGY-Instruments.de 

They make systems for measuring optical density as well as resistivity using the eddy current method.

Dust levels in metallizer & factory.

The following comments/questions were made about the dust levels and I suspect this will be common to many metallizers.

1.Many efforts are made to keep the factory dust free but still lot of dust is seen in the factory.

2.In a metallizer itself lot of dust in powder form is seen.

3.What should be the acceptable dust / particle size in a factory where metallizing is done which will not affect the barrier properties of metallized film?

4.What is the economical way of cleaning the film before metallizing?