Converting Quarterly title
The Official
Technical Magazine
small AIMCAL logo

Vacuum Web Coating

Blogmaster: Dr. Charles A. Bishop


Surface energy is an important factor when trying to manage web surfaces and the general aim is to increase the surface energy of webs with the expectation that with it will come increased adhesion. Often this is true but it does not have to be the case. It is true that if the surface energy is low then the coating may not wet or cover the whole surface and so the adhesion will be reduced. However there are different types of adhesion. There is the low adhesion that is due to proximity of the two materials such as Van der Waals force and the most adhesion that this can produce will be when the two surfaces are in full contact. Reduced contact area will lower this type of adhesion proportionally.

Usually what everybody wants is the very high adhesion that is only obtained when there are direct chemical bonds between the coating and the bulk polymer. To obtain this type of adhesion it will require the breaking of some of the polymer chains and chemically modifying the chain ends to encourage the depositing coating to attach directly to these polymer chains. As the polymer chains are entangled in the rest of the bulk polymer it is difficult to pull out these chains and hence the only way to pull the coating off is to break some of these bonds. This type of failure takes a lot more energy than it does to separate out the coatings only bonded by proximity.

Where the confusion comes is that in order to chemically modify the polymer surface to raise the surface energy or to chemically change the surface to enable direct chemical bonds to the polymer the process is similar and sometimes identical. When we plasma treat the polymer surface to increase the surface energy we will be changing the surface chemistry usually by substituting some of the carbon atoms with oxygen atoms  When we want to increase the adhesion we frequently use oxygen to act as the chemical bridge between the polymer and aluminium metallization. This is done using vacuum plasma where typically there is an argon/oxygen gas mixture with the argon bombarding the polymer and giving some chain scission and having broken the chain the oxygen bonds to the chain ends. Aluminium has a great affinity for oxygen and it will bond well to oxygen and so this will provide strong anchor points for the coating.  Between these anchor points there will still be the proximity bonding but by comparison this is a very minor contribution to adhesion.

So if we have modified the surface to optimize the chemical bonding and so maximise the adhesion we will also have significantly increase the proportion of oxygen on the surface and in so doing we will have increase the surface energy.   

It is possible to have a high surface energy and low adhesion. The plasma treatment, when started, will break bonds and the oxygen will begin to become attached to the surface. This process is not self limiting and the plasma will continue to break bonds long after the adhesion has been optimized. This continued treatment will keep breaking more and more bonds at the surface into ever small fragments. These small fragments instead of being entangled into the bulk of the polymer are now poorly attached and so although the aluminium can be well adhered to the oxygen and to the polymer the short length of the polymer chain fragment means that it has little cohesion to the bulk polymer. The surface energy will still be high as there is still a high proportion of oxygen present. Thus surface energy does not directly equate to adhesion. 

If you want a more detailed view of this topic please look up the paper given by Prof. Steven Abbott at the AIMCAL European Conference in Prague earlier this year or try to catch one of his Webinars that he gives periodically via AIMCAL.

Posted in: Adhesion


#638 mustafa homran
Friday, September 28, 2012 5:30 PM
its true if we keep metalized film in cold room for tow days
it will keep wetting tension high ? or the opist is true?
#642 Charles Bishop
Tuesday, October 02, 2012 4:48 AM
Keeping the polymer web cool will slow down the migration of any low molecular weight to the surface. However keeping the web cool only slows this process down it does not stop it altogether.

With metallized film there is the additional factor that the metal surface has an extremely high surface energy and when it comes into contact with the back surface of the film it will be energetically favourable for the low molecular weight material present on the back surface to transfer onto the front surface. Storing at a cool temperature will slow this down but not stop it altogether.

The problem will vary from film grade to film grade or from supplier to supplier as each supplier will have a different polymer recipe and even with the same supplier may have a different recipe for different grades. Hence it is impossible for me to know what to expect of your particular film.

The simple thing to do is to test the surface energy after metallizing and then test periodically to plot the rate of decline. You can do this at normal temperature and compare to material kept cool and so for your particular film evaluate the difference.


#917 Suresh chandra
Thursday, July 25, 2013 1:05 PM
Dear Sir;
I have two queries:

1. ) Why metal bond strength is high in lower OD say 2.2 OD than high OD say 3.0 OD in same type of metallization say metallization on plain Pet Film.

2.) Why there is need of vaccum in metallization process?

Request you pls. tell us in detail.

Best Regards;
#918 Charles Bishop
Thursday, July 25, 2013 4:44 PM
The first question would suggest that you are seeing differences between the levels of adhesion that you are seeing between your own coatings. If this is so I would throw the question back to you. What is different between the processes you use when you deposit the different thickness coatings?
You will have to make some change if you are to change the thickness of the coating. This could be more power to the boats and so a higher temperature and higher wire feed rate for the same winding speed or the same power and wire feed but a slower winding speed. Whatever you do you will have changed the conditions and so will change the response of several things. If you increase the power you will increase the boat temperature and so you will increase the heating of the whole system and hence increase the outgassing rate. The sudden rise in substrate temperature will be steeper and to a higher temperature which will initially change the outgassing rate and oligomer migration rate and so you change the initial nucleation and growth of the coating. These may only be fine changes but nevertheless they are changes.
Is the PET plasma treated before metallization? If so, if you changed the winding speed to change the coating thickness did you also changed the plasma treatment to compensate for the change in residence time during plasma treatment. If the treatment time is changed the adhesion could also be expected to change. If the process were optimised for the slower process then it could be expected that there would something less than an optimised treatment for the faster winding speed.
The other thing to consider is the adhesion test and the failure mechanism. I would check in finer detail where the plane of failure is. It is always the first assumption that the failure is at the interface but often the metal may be well adhered to the surface it deposits onto but the surface it deposits onto is not well adhered to the bulk polymer and so the true plane of failure is a cohesive failure of a weak boundary layer in the surface of the polymer.
So I would contest that the adhesion does not have to be different. Differences are most usually due to differences, however small, in the process that are ignored, or not monitored or controlled well enough.

There are several reasons why you would not would not deposit aluminium at atmospheric pressure.
The first reason would be safety. Aluminium metal has a great affinity to oxidisers. It is often used in explosives/pyrotechnics. As a powder it can have an autoignition temperature around 760 Deg C. If you have a powder the aluminium as it oxidises will heat the powder because the oxidation reaction is an exothermic reaction. If the powder has a small enough mass compared to the surface area the temperature will rise above the autoignition temperature and the aluminium particle will burn. For solid aluminium this becomes less predictable as the oxide on the surface can restrict the rate of oxidation and if the oxide remains intact the autoignition temperature will relate to the melting of the aluminium oxide which is closer to 2000 deg C. However it the oxide is disrupted and the rapid oxidation of aluminium occurs this will relate to the autoignition of the metal at the lower 760 Deg C. Hence evaporating aluminium from boats at atmospheric pressure could be an exciting process but without managing to put a coating onto the substrate.
The second reason for a vacuum when metalizing aluminium is that aluminium oxidises extremely rapidly. Within the vacuum system it is already oxidising before it has been wound up into the roll. It obtains the oxygen from the partial pressure of water that is present in all vacuum systems as a result of outgassing from all the surfaces.
The third reason would be gas scattering. In a vacuum system the metal is evaporated from the boats and if the vacuum is good the mean free path between collisions will be longer than the distance between the evaporation source and the winding substrate and so the atoms will land on the substrate without collision. If the pressure is increased the atoms will be scattered and more material will be scattered away from the substrate and so the deposition rate would go down. At atmospheric pressure the scattering would be enormous and so the substrate would need to be moved closer to the sources ad so the heat load from radiation ant at atmospheric pressure from convection would increase but the heat of condensation would fall and so you would get a greater heating of the substrate but with less coating being deposited and the coating would be transparent because of oxidation, assuming it did not all catch fire.
The fourth reason again relates to the increase in pressure. The rate of arrival of the depositing coating is reduced and the background gas arrival is massively increased and so the deposited coating would be very porous and the included gas content would be larger, assuming you got any coating at all.
I will stop here. There are other reasons why it would not be and easy or good process.
If you want to deposit materials at atmospheric pressure you can change the process from evaporation to something like atomic layer deposition (ALD) where you build up a coating atom by atom which optimises a chemical approach to deposition. In ALD atmospheric pressure or something very close to atmospheric pressure can be used but the rate can be slow unless larger pumps to rapidly move different gases into and out of the deposition zone are used as many tens of cycles are needed to build up a coating thickness of a few nanometres. To build up a thickness of 30nm would (I think) require something between 50 and 100 cycles (each cycle requires 4 steps, (1st gas, purge, 2nd gas, purge) and either the deposition zone needs to be long or multiple deposition heads are required to work very quickly to build up sufficient thickness to compete with vacuum evaporation. So far vacuum deposition of aluminium is still the cheapest process and most of the process limitations can be managed sufficient to be able to produce acceptable products.
I hope this answers your questions.
#931 Tribhuwan
Saturday, August 10, 2013 4:25 AM
What is the standard counts of pinholes in metallized polyester film of 12 micron?
#934 Charles Bishop
Monday, August 12, 2013 6:56 AM
As far as I am aware there is not a standard for the number and size of pinholes for specific applications.

Where companies work to a standard it is usually defined by their customer to meet a particular need.

In addition to the lack of standards for the number of pinholes there is also a lack of a standard for the equipment to measure the pinholes. The light uniformity and intensity will affect the number of pinholes counted. In some cases people use image analysis to size and count the pinholes whereas others simply count the number of pnholes and categorise the size as big or small.

If the metallized product is to be used as a barrier material there are standards on the measurement of barrier performance and as barrier performance is dependent upon the number as size of defects others will use the barrier performance as a measure of the coating quality.

I think that the Japanese have some publshed numbers for the number and size of pinholes for some if their higher tech applications but these too tend to be used by one company and copied by others so that it is becominng a standard.

If anyone wants to send in some fo the specifications that they have had to work to we can see if there is a common trend or and common levels that everyone is working to or if everyone is working to different specifications.

I will keep looking to see if I can find any standards that are published and if I find any I will post them on the blog.

#952 Marcelo S.
Tuesday, September 03, 2013 9:19 AM
I have observed , that after the lamination process , occurs in some cases a removing the metallization layer that is transfered to another substrate, resulting lamination strength low due to starter of the metal of PET film, this could be related to lamination or metallization process .
#1008 Jijo Raj
Monday, November 04, 2013 10:46 PM
We are facing the problem of metal lifting in bopp film after long aging . Customer complient about the poor bond strength during lamination .
We checked the metallized film sample with scotch tape and didn't find any metal lifting .What is the reason of such kind of metal lifting ?
#1009 Charles Bishop
Tuesday, November 05, 2013 8:56 AM
The Scotch tape test is a very poor test and tells you virtually nothing about the bond strength of the coating to the substrate. If the coating fails using this test all it tells you is that the coating is very very poor. If the coating passes all it tells you is that the adhesion is similar or better than the adhesion of the adhesive on the scotch tape. It is possible to find some higher strength adhesive tape to check if the coating adhesion is higher than for the standard scotch tape but this too gives you no number that you can use as a quality control measure.

Your coating adhesion does not sound as if it has been optimised. If you do any pre-treatment, either before it reaches the metallizer as a corona treatment ro inside the metallizer as a plasma treatment all you may be able to measure is the surface energy and this can be misleading as the surface energy does not tell you what the adhesion is. Initially as the surface energy begins to increase, with treatment time or energy, the adhesion may also increase but once the surface energy has reached a maximum the adhesion may fall and end up lower than it would be if the surface had not been treated at all.

There are other tests such as the peel tests or better still the fragmentation test that enables the true adhesion of the coatign to be determined. However this test requires a higher skill level than does the scotch tape test and it also requires a microscope or better still an electron microscope with a micro-tensile stage to carry out the test. A short cut test that you might carry out more quickly and using only a tensile test with an optical microscope is to stretch the sample and look for the onset of cracking, the point at which the cracks first appear across the tensile sample. This will be at a low strain (probably <1%) for poor adhesion but may be higher than 2% for high adhesion samples.

I could never optimise the adhesion using only the scotch tape test and so it is entirely likely that your customer is experiencing the problem because your coating is only just exceeding the scotvch tape test but the true adhesion is still low.
#1189 steven
Monday, April 28, 2014 5:24 AM
We facing the high surface resistivity problem for mpet and pe lamination film after
curing for 7 days. Just PE side
#1455 Amritanshu De
Saturday, August 08, 2015 3:11 AM
I have two questions. How you can achieve High O.D, say 3.2 on polyester ? Is it by twice metalising the same film or by doing metalising in slower speed ?
How the barrier properties & adhesion properties vary with amount of metalisation on polyester & BOPP (keeping the thickness of base film same, say 12 micron)?
#1456 Charles Bishop
Saturday, August 08, 2015 6:38 AM
There are a few options. It is possible to drive the boats harder, increasing the power and so the temperature and so allowing an increase in the wire feed. This keeps the winding speed high but the penalty is the higher cost of boats as the lifetime will be reduced at the higher power. An alternative is to keep the deposition rate constant but reduce the winding speed to give a lower productivity but maintain the boat lifetime.
The key problem to these options is that the heat load to the substrate is increased and it will depend on your system and cooling capacity as to whether the temperature can be controlled and the web pass through the deposition zone undamaged. If there is insufficient cooling of the web then there may be dimensional changes to the web or problems of wrinkling. If you do not have the capacity or option to increasing the web cooling then the other option of depositing the coating in two separate passes may be the only way of achieving the desired thickness. I do know of systems that are able to deposit coatings of up to OD 5.0 in a single pass and so it is possible with the right system design.
There are software packages such as TopWeb from that allows you to start with your existing process and then look at the effect on the substrate of slowing down the winding speed or keeping the winding speed constant and increasing the deposition rate. This will give you an idea of how the substrate temperature changes and who much risk there will be of the web wrinkling.
I hope this helps.

Post Comment

Name (required)

Email (required)

Enter the code shown above:


Charles Bishop photo

Dr. Charles A. Bishop

Charles is a toolmaker by trade after completing a mechanical engineering apprenticeship. He then entered University and obtained a Bachelors degree in materials engineering with a Diploma in Industrial Studies. During his final year he first started work on vacuum based research, helping develop a process for manufacturing titanium based bone implants for tendon location. He went on to obtain a Masters degree and Doctorate following further research into vacuum deposition processes. During this time and as a postgraduate he also worked as a consultant.

Charles next spent time in industry working for various divisions of ICI including polyesters, nylon, Imagedata, Flex Products Inc., and explosives as well as contributing to other projects. In 1998 he took the opportunity to return to consultancy work and set up his own company.

Charles has more than 30 years experience in vacuum deposition mainly onto flexible webs. He has regularly contributed papers to conferences and recently has edited this blog on behalf of AIMCAL as well as being one of their presenters for various webinars and the more formal Converting School courses.

Charles has also published 2 books, Vacuum deposition onto webs, films and foils and Roll-to-roll vacuum deposition of barrier coatings.


The world's web process community.
Join. Participate. Succeed.

USA Office

201 Springs Street
Fort Mill, SC, 29715
Phone: 803-948-9470
Fax: 803-948-9471

Europe Office

PO Box 69533
1060 CA Amsterdam
The Netherlands
Phone: 0031 (0)20 8208635
Fax: 0031 (0)20 8208634

facebook linklinkedin linkyoutube linktwitter linkinstagran