Vacuum Web Coating

Web Cleaning - Tacky rolls update

I was recently at the SVC annual technical conference where I listened to a paper given by Sheila Hamilton about the Teknek tacky roll technology.  I have long been enthusiastic about this technique for removing debris from webs or foils and think that it is underused within the vacuum coating industry. 

The process uses an elastomer roll that is in contact with the web surface where debris is removed from the web and sticks to the tacky roll. This process can remove debris of size down to 0.3 microns.  This is a simple process but effective.  As you can imaging webs have a high debris level and so it would be easy for the tacky roll to quickly become clogged with debris and thus lose the effectiveness as the tacky surface is covered in debris. To recover the tacky surface this roll is usually in contact with a roll with a higher tack that accumulates the debris. This roll will also become clogged but it is like an onion, built up of many layers that can be peeled off revealing successive layers of high tack surface.

As most pinholes are caused by debris being coated and moved after metallization the reduction in debris levels is an easy way of reducing the number of pinholes in the coatings.  As with any cleaning technique it is important to do the cleaning at the correct point in the process as well as to consider cleaning both sides and not just the side that is to be metallized.

If the roll of material is to be cleaned on a winder before it reaches the metallizer but if this is done it is important that the web is kept clean following the use of the tacky roll.  As polymers winding over rolls produces an electrostatic charge the web can attract airborne debris and so it is critical that between the tacky roll and rewinding the web is kept in a clean environment such as under a positive pressure clean air hood or, if the application demands it, in a cleanroom.

One approach I have been advocating for a number of years now is to incorporate the tacky rolls into the vacuum system. I have tried to encourage vacuum system suppliers to develop this process but the progress has been slow. I know that some companies (2 in the Far East & 1 in Europe) have incorporate the process and I now hear that Ulvac have started to offer systems that include the tacky roll. There are limitations though.  The systems they have added it too are all ones where the roll length is short and so they do not have to peel off layers from the high tack roll and this considerably simplifies the process.  Until this automated peeling process is developed to be robust for production so that the roll length can be increased to the long lengths used in modern metallizers I expect that these roll will only be used on the higher technology processes where roll lengths are short enough and the webs already clean enough that neither tacky roll becomes clogged by a single roll.

However I regard this as encouraging. Many people did not believe that the process would work well in vacuum. It is interesting that the elastomers were developed with the space programme in mind and so the problem of outgassing and loss of tack had already been addressed and was not the limiting factor that some expected.  Reports of some tack rolls ‘drying out’ are probably as a result of having them exposed to the plasma cleaning process, which because of the continual bombardment of the surface will break bonds on the surface and degrade the performance.  This is simple to prevent by suitably shielding the plasma from the tacky rolls.  I say simply but I suspect that if you want to add the tacky rolls as a retrofit item then space will be limited and shielding may be more difficult. New machines will be able to be designed with this in mind and suitable shielding should not be a problem.

Another piece of information that I was not aware of before is the high temperature use of the tacky rolls. The elastomers are stable o high temperatures and have also been used as a cleaning roll for hot embossing shims.  These embossing shims can pick up debris, oligomers, additives or fillers as the come in contact with the polymer surface and these can accumulate in the grooves of the metal shim reducing the quality of future embossing. Cleaning the shim is always an issue. Some use release agents to lower the shim surface energy and prevent this pickup but this can transfer onto the polymer and make the subsequent coatings hard to adhere to the surface. Thus the tacky roll method of cleaning has an advantage of cleaning the shim but not adversely affecting any subsequent adhesion.

So as you can see my enthusiasm for this tacky roll technology has not been diminished but has rather been endorsed with the news of the first machines being sold with this technology included.  I am still convinced that it is only a matter of time before this process is available for metallizers and for those wishing to produce pinhole free coating will be an essential part of any new metallizer specification.  I suspect that this process will be speeded up if any of you who are considering buying a new metallizer start to ask is the tacky roll cleaning available in the metallizer.  If the machine builders get enough people asking about this technology they will speed up their development to make sure it becomes available to meet their customer needs.

Until then watch this space & I will periodically report on progress.

A problem of un-metallized streaks

I would like to inform you that we are receiving metallized BOPP film from outside and facing the problem of unmetallized streaks on the surface which is more visual after the lamination to other printed substrate. This sort of defects we are getting normally and have made complaint and also claim for the rejection due to this defect. I am requesting you to pass the information why the unmetallized streaks appear on the surface to some part of the metallized BOPP film.

Answer

Unmetallized streaks.

It would be interesting to know more about the unmetallized 'streaks'.

Are they irregular in shape or parallel stripes?

What is the size of the streaks?

To me the word streaks suggest the streaks are anything from a few centimetres in length to a meter or more and irregular in shape.  Is this correct?

How wide are the rolls? 

What position are the streaks found, are the randomly anywhere across the whole width, and do they start immediately on the roll, are they found throughout the whole roll or are they only found part-way through a roll?

One possibility is that during metallization the heat load is close to causing the web to balloon off the cooled deposition drum. As the web sees the heat load from the deposition source the polymer wants to expand and if this expansion is too much the web buckles off the drum and if not controlled this will result in a 'tramline or railroad line'. This is seen as a parallel stripe of material that has a thinner coating on it. The thinner coating is because as the web lifts off the deposition drum the polymer expands and thus the surface area is greater than it was when on the drum and so the same depositing metal has to cover a larger surface and so the coating is thinner. Also the temperature is likely to be higher and so the sticking coefficient of the aluminium will be lower and so not all the metal will stick and this further reduces the coated thickness.

If these thinner coated sections appear and disappear this may be because the process is just on the knife edge of too much heat. A little more heat and the web would permanently lift off the drum and a parallel line would be established and possibly lead to a wrinkle in the roll.

All aluminium coatings oxidise. The thickness of the oxide layer can depend on many things including deposition rate and coating structure.  The thinner deposition area will have a lower deposition rate and so will have slightly higher oxygen content and also because of being thinner the surface oxide will potentially have a greater effect on transmittance than on the thicker areas.

Often the areas where wrinkles or tramlines appear are after the process has been running for some time and the heat load has gradually increased over time. Also they tend to be towards the centre of the roll. Near the edges of the polymer web it is possible that any expansion of the polymer web can overcome the polymer to drum friction and the web slips laterally on the drum and so relieves the lateral tension. However in the centre of the web the friction is too much and such expansion is less likely. Hence of ten the tramlines are only seen in the centre section.

Thus my questions about where the streaks are found within the roll help me to diagnose the probable cause of the defects. 

I hope this answer helps explain what might be occurring.

Moisture & metallizing

Does moisture affect the metallization of Aluminium? 

We laminate Metalized polyester with PEG coated paper and metal peel off.  Once the oxidation starts in metallization will it destroy the whole metal with passage of time or not?

Will metalized MPET will be oxidised if we laminate it with PEG coated Paper.   The ratio of PEG coated paper is 80% PEG and 20% water.  We observe white spots in metalized paper.  Please try to find root cause and suggest us what measures can safe us.

ANSWER
Yes moisture can affect the performance of aluminium metallized films.  This starts with the film before metallization. Most polymers contain moisture as well as the water in the air that is trapped between layers as it is wound into a roll.  This is all carried into the vacuum system. Even when the system is pumped out to a low base pressure there will still be enough oxygen and water in the system that a monolayer of oxygen can form on a surface in less than 1 second.  Thus all aluminium metallized films have a proportion of oxide in them, usually of the order 1% - 2%.  If the vacuum
system has a leak this can be somewhat worse.

The aluminium is a metal that forms an oxide on the surface that acts as a good barrier layer and prevents further rapid oxidation.

The adhesion of the aluminium to the polymer web is dependent on a number of things. This can include the polymer quality, contamination, the storage conditions (humidity and temperature), any surface treatment, the type of treatment, the age of any treatment as well as the process conditions such as deposition pressure.  When laminating another layer to the surface it will depend on the relative adhesion strengths and the residual stress following the lamination. If the residual stress is large the adhesion of the aluminium needs to be higher than if the residual stress is low.

The speed of oxidation and amount of oxidation depends on the thickness of the aluminium layer. If the aluminium is very thin and the adhesion is poor the there will be very little aluminium left once the surface has been oxidised. If the aluminium is thicker then even after the surface oxidation there will be sufficient metal left to give a long lifetime for the rest of the aluminium.

Boats and spitting

What causes spitting from evaporation boats?

ANSWER

Generally spitting is associated with changes to the pool shape and size.  The boat temperature needs to be stable and this not only means the current and voltage need to be stable but also the wire feed rate needs to be stable.

It can be easy to find that the thickness monitors respond to thickness changes and change the wire feed and this correction may be too much and the thickness monitor then re-corrects for this change with a further change. This can appear as an almost continuously changing wire feed rate. This will mean that the molten pool will be changing size and shape. The oxide from the surface of the aluminium wire, as well as any impurities, will collect as a skin or crud on the surface of the molten pool and often will collect around the edges of the pool. This collection of material is what usually is ejected and seen as spits. As the pool size changes this material either covers a new area of the evaporation boat or is left stranded on a drying part of the boat and can be thrown off as a spit.  Thus maintaining stability is regarded as the key factor in lowering the number of spits, assuming all other things are the same.

There can be other factors that also contribute such as the age of the wire, as older wire may have more oxide on the surface, and the purity of the wire.

One of you questions refers to MOC of the boat.  I take this to mean method of control?  If so you need to check the capabilities of the machines and possibly their history. If the machines are of different ages they may not have equivalent capabilities and so the same control process may not be available. If the machines are nominally identical it may be that they were used by different teams for completely different products that had different requirements and so different control methods were adopted for the different products on the different machines.  This could account for the original differences and if they now produce the same product these methods of control may simply have been continued on because that is what the operators are most familiar with.

The performance and feed rates can be calculated using the equations that are given in the AIMCAL Metallizing Technical Reference 4th Edn. book.  This is available from AIMCAL and most companies that are members of AIMCAL have copies. However if you need on if you contact AIMCAL at www.aimcal.org they sell then at ~$25 each + postage and packing.

Uniformity of pumping, system symmetry.

For those of you depositing metals this is going to be of less interest but for those of you either reactively depositing coatings or thinking about developing a reactive process this should be of great interest.

I recently went to work on a vacuum system that had been designed as a research and development machine capable of depositing coatings by reactive deposition. Typical coatings that the machine was designed for included titania, alumina and the transparent conducting coatings such as indium tin oxide (ITO).

The system had been designed with coating uniformity in mind and the pumping was arranged around the chamber along the centreline of the web path. The gas input was through manifolds that were similarly positioned to give a uniform flow across the width of the web. Thus the general arrangement of the sputtering and other deposition sources, pumping and gas input could be expected to give uniform films.

Needless to say the reason I am writing this note is that the coatings were anything but uniform.  Originally some ITO coatings were deposited from ceramic sputtering targets and these were reasonably uniformly conducting across the width and fairly easily optimised. However when it was tried to deposit ITO coatings from a metal target it was impossible to optimise the conductivity and when the resistivity was measured in more detail it became apparent that there was a different resistivity on each edge. When a minimum was reached on one edge the other was more than 20x higher and vice versa.

On a detailed review of the system, checking the symmetry of all the equipment it became obvious what was the cause of the problem.  The original design included a cryocoil in the winding zone to getter the water vapour from the roll of polymer and this too was uniformly distributed across the web to give uniform pumping. There was some work done on the best positioning of cryocoils/cryopanels in the vacuum system and having cryocoils in both the winding and deposition zones was shown to be better than having a single cryocoil in either the winding or deposition zone.  To take advantage of this information additional cryopumping was added to the deposition zones. Unfortunately at this point the symmetry was lost as the cryocoil was positioned on only on one side between the original pumps and the chamber wall. To make this worse as the coils were behind the sputtering sources and so could be expected to see some heat they were protected by being protected by a metal box.  The net result of this was that the conductance to the original pump was restricted on the side that had the cryocoil.

We know the end result was skewed pumping to the deposition zone such that the pressure at each end of the sputtering source would be different and so the requirement for reactive gas would be different. Hence, as the reactive gas was being uniformly introduced across the whole width, the optimum point to give the correct stoichiometry for a coating with minimum resistivity would be different at each end of the sputtering target. 

The solution to the problem is simply to change the shape of the cryocoil in the deposition zone to have an equal surface area on either side of the existing pumping orifice. This would reduce the volume the existing coil takes and so the protective radiation shield could be moved further back and so most of the restriction in conductance would be removed. Thus the uniformity would be returned and the pumping speed increased at the same time.

This has yet to be done to prove the above solution is correct as the work has to be planned and fitted into programmed downtime. However we are in no doubt that we have found the source of the non-uniformity and how to correct it.

I think this offers an excellent lesson on why it is essential to have symmetry about the web centreline for all aspects of the system and process.  Adding equipment that restricts the conductance only on one side of the web or adding additional pumping such as a cryocoil and placing in a convenient space in the chamber rather than symmetrically across the web can result in non-uniform coatings.  I emphasise that this will only be a significant problem for reactive processes. Although minor variations may be measured across the web in other coatings for most applications these are negligible.  When depositing ITO from a ceramic target, where some reactive gas is included to compensate for any dissociation during the sputtering process, the dynamics of the process are such that working with an excess of reactive gas is not as critical as it is with the full reactive process converting a metal target to the desired compound.  Hence the non-uniformity of the coating will be similar to plain metal coatings rather than full reactive deposition.

Although it is less important to have symmetry of pumping in metal deposition it will become increasingly important as the width of deposition systems continues to increase. Already we have machines of 4.45m wide. If these are pumped from one end only imaging what the pressure difference will be from one end to the other. This pressure difference will become greater and greater as the width continues to increase and this will be seen as producing different deposition rates from resistance heated boats at each end of the system.  A simple solution is to pump from each end although there will still be a pressure gradient from the web centreline to each edge.  A better solution would be to distribute the pumping across the whole width. It is worth noting that as customer specifications continue to increase, particularly for high technology applications, it can be expected that there will be an increasingly tight specification for resistivity which may require an improvement in system symmetry.

As you might have guessed this is a hobbyhorse of mine. I have worked on too many systems that were never designed with symmetry in mind but were expected to produce coatings by a reactive process that have been a major challenge. Mostly the difficulties could have been eliminated by better design making the processing much easier and the productivity of the system much higher. Thus I find it disappointing that even on new machines I still see some of the same design errors.

Lower than expected barrier performance

I represent a converter who is trying to achieve good barrier properties (wvtr<0.2,otr<0.5) using metallized PET as part of the laminate. We have our own metallizer and the highest OD we can achieve on a single pass is 3.0.However, at this OD, we are unable to achieve the above noted barrier properties. Are there other methodologies that are available to enable to hit these properties. We have heard of double side metallized films utilizing plasma and chemical treating? Are these possible solutions and have you had experience with these types of films? Secondly, we are also concerned with loss of properties with downstream processing like rewinding and slitting. Are there special processing requirements for finished metallized film in order not to degrade the barriers achieved during metallizing?

Answer

The biggest limitation to producing good barrier films is generally pinholes.

Pinholes are produced primarily by debris that remains on the film surface that is larger than the very thin metal coating you deposit so that after metallization if the debris is moved it leaves behind a pinhole.

A second source of pinholes is spitting from the boat. This can be caused by a combination of reasons such as a low purity aluminium wire or the wire having a thick oxide coating on as well as an unstable pool of molten metal that as the pool size changes encourages spitting from the ends of the collection of crud built up from the wire oxide and impurities.

Another source of pinholes is any pickoff. This is where any high spots on the reverse surface presses hard against the freshly deposited metal and in some cases it overcomes the adhesion and picks off the metal from one surface and it transfers to the second surface. Again leaving a pinhole.  Typically this is associated with large fillers and a hard wound roll.

If the incoming roll is cleaned to a high standard it is possible to reduce the number of pinholes and so improve the barrier performance.

This can be done by using techniques such as the tacky roll type method of removing debris.  All polymer film will be covered with debris. This is partly because the polymer film as it is wound generates an electrostatic charge, which attracts airborne debris to the surface, this can include slitter dust. Thus if the roll is cleaned just before the final rewind before the roll then goes into the vacuum system this can help. If it is done earlier it can become recontaminated very easily. Ideally after cleaning the film should be in under a clean filtered air hood to limit recontamination. Also the film should be cleaned on both sides otherwise the debris from the back surface can become transferred onto the front surface as the film is rewound.

Similarly it may be possible to improve the wire purity and make sure the wire surface has a minimum of oxide present and may also be possible to improve the wire feed control to reduce some of the pool variations.  This will help reduce any spitting problems

Other possible things that can be looked at is to see if it is possible to wind the material between metallization and lamination with fewer or no front surface rolls in order to minimise the change of moving the debris and so limiting the number of pinholes that appear. Although it is better to clean the material and not have the debris present but this may not be easy to accomplish.

The use of adhesion promoting measures such as plasma pre-treatment if done well, with the process carefully optimised, can improve the adhesion as well as the metal wetting. This has two advantages. The improved wetting means that the metal will spread out on the surface and so will produce a continuous coating at a reduced thickness and this can be seen as either the same OD at a thinner coating thickness or for the same thickness a higher OD. The second advantage is that the coating is less likely to have problems with pick off as a method of producing pinholes.

In addition there may be other changes that can show benefits. If the metallized film is laminated very soon after metallization it may lose more barrier performance than if any further processing is delayed by a day or so.  The aluminium coating is very soft and is more prone to damage if it is handled very soon after metallization. If the roll is stored for a short period of time the native aluminium oxide is allowed to build up on the surface of the aluminium and this is much stranger and any damage is likely to be less.

Where double side metallization wins is that there statistically it is unlikely that any of the pinholes on each of the two aluminium coatings lines up with each other and so the increased tortuous path reduces the gas or moisture transmission.

I hope this gives you some explanation about where the lack of barrier comes from and some possible routes forward.

Deposition of conducting polymers

I need to know

1.         What are the methods used to deposit conducting polymers?

2.         Can we use vacuum coating deposition to deposit conducting polymers on glass surface?

Answer.

To some extent the deposition method depends on the type of conducting polymer and the formulation used.

Many of the conducting polymers have other materials added to provide other properties such as a polymer binder to better adhere to the substrate where there will be extreme flexing of the device or the addition of solvents including water to reduce the final coating thickness.

H.C.Stark gave this description of application methods in the paper they presented at the 2005 AIMCAL Fall Conference.

For small and irregularly shaped parts, spray coating or dip coating is often preferred. For larger substrates, such as plastic sheets or wide-web plastic rolls, other application methods such as gravure printing, roller coating, etc. may be used.  For deposition of patterned PEDT films, ink-jet, screen or nozzle printing can be used.   In addition to this small discrete substrates are often spin coated.

If the polymer has been diluted using a solvent there could be problems in trying to deposit the coating in a vacuum system because of the different vapour pressures.

The OLED polymers are regularly evaporated in vacuum systems and Kurt J. Lesker sell a slot source specifically for the deposition of the OLED polymers in vacuum.  In principle if the conducting polymer does not fractionate it too could be directly evaporated in vacuum using a similar source.  The trick with this type of deposition appears to be using the lowest possible temperature to prevent damage to the polymer and to keep the temperature as even and constant as possible to keep the deposition rate and hence coating thickness constant.

Maximum deposition onto BOPP.

We intend metallising 18-micron plain treated BOPP and 18-micron metallising grade plain treated BOPP. Is there a constraint on the maximum OD levels that one can achieve on a BOPP? I am made to understand that normally BOPP cannot be vacuum metallised beyond 2.2 OD.

Answer.

The limitations on the OD that can be deposited onto BOPP will depend on a number of things.

The critical part of the process is managing the heat load the film sees during the deposition zone.

The film is pulled tight onto the deposition drum. The reason for this is to make a good intimate contact between the polymer web and the cooled surface of the deposition drum. In this way the heat transfer coefficient is increased. The heat transfer coefficient has several components, radiation, convection and conduction.  The radiation is small, the conduction depends on the contact surface area and this will depend on the polymer and deposition drum surface roughness as well as the contact pressure. Thus for any particular film/drum combination it can be seen that contact pressure can have an influence.  This is limited because as the film heats up the yield strength of the film decreases and so the tension that it can withstand before it suffers some permanent dimensional change.  This is the same for all metallizers and so the limitations are similar.  Some machines can cool the deposition drum to a lower temperature than others and so will have an advantage allowing a slightly thicker coating than those with a higher deposition drum temperature.  Also a larger diameter deposition drum can have an advantage as the longer deposition zone allows slightly more time to deposit the coating and remove the heat.

The one factor I have left out was that of conduction. I was found that any absorbed moisture in the film could be an advantage in that it increased the heat transfer coefficient. Under the effect of heat the moisture evaporated from the surface and filled the gaps between the film and the deposition drum and then some additional heat was transferred through convection currents. Some modern machines have deliberately made use of this idea to improve the heat transfer coefficient. Where the film comes in contact with the deposition drum gas is injected into the gap and becomes trapped between the web and drum. This trapped gas provides the means to substantially increase the heat transfer coefficient. This enables either a thicker aluminium layer to be deposited or a faster deposition rate to be used.

Incidentally the gas wedge also reduces the coefficient of friction between the web and the drum allowing the web to move laterally more easily than without the gas present. This reduces the propensity of the web to buckle off the surface causing tramlines (railroad lines), which is also a heat related problem.

Thus there can be quite a large difference in performance between a system that has poor deposition drum cooling and no gas insertion mechanism and a small deposition drum and one that has a lower deposition drum temperature, larger drum diameter as well as a gas wedge insertion unit.

Added to this can be the degree of orientation or more precisely the tensile performance of the BOPP as the higher the tensile performance the greater the tension that can be pulled.

Thus if you are wanting to coat to a high OD it is worth reviewing what machines you have available to see if one is more suitable than the others.

Copper Metallization

Question.

What is critical point if we would copper coating on PET film?

We will modify our aluminium metallizing machine to cooper metallizing, what should we do?

Answer

using existing evaporation boats can be problematic. The copper does not wet the boat well and so can be difficult to control in the same way as aluminium. 

I know that companies have used traditional boats but the boats need to be conditioned and this can take time and everybody seems to work on their own solution to the problem that they keep secret.

Sidrabe, the system building company in Riga, Latvia, sell special boats that are designed specifically for deposition thick copper.  These are designed to be direct replacement boats for the standard boats but may have limitations as the current requirements may be limited by the transformers you have on your system.  These boats use tungsten rods that are wire wrapped down the centre of a tube of refractory metal. The copper is fed into the end of the boat and the motlten copper runs along the rods and is evaporated out through a slot in the tube. Because the vapour comes from the out of the centre of the bundle of the rods there is little or no spitting and the quality of the coating is good.

If you wish to contact Sidrabe a couple of contacts are

Nils Veidemanis             or Victor Kozlov

veidemanis@sidrabe.eu  or vkozlovs@sidrabe.eu

phone +371 7249806      or     +371 67812542

Victor Kozlov presented a paper at the SVC 2006 Annual Technical conference on the use of these sources if you wish to read up about them.

Adding nitrogen as third gas in plasma treater.

Sir, here in our Metallizer we have put a third gas line (N2) for plasma, now we are going to take trails with combination of three gases (Oxy, Argon and Nitrogen). But I don’t have any experience of three gas combination, so can you tell me how I have to start, should I reduce the flow of oxygen and put nitrogen by keeping Ar same OR should I reduce the flow of Argon and keeping oxygen the same and add Nitrogen, please mail me how this Nitrogen work in this plasma system, should I increase current value also?

Answer.

Re. Nitrogen plasma content/treatment, the following papers might have some information that you may find helpful.

1.    Shi M.K. et al

Plasma treatment of PET and acrylic coating surfaces - 1.  In-situ XPS measurements

Journal of Adhesion Science & technology  Vol.14, No 12, 2000  pp 1485-1498

2.    Poncin-Epaillare F. et al

Reactivity of surface groups formed onto a plasma treated polypropylene film

Macromoecular Chemistry & Physics  Vol 200, No.5 May 1999  pp 989-996

3.    Vallon S. et al

Adhesion improvement of silica layers on polypropylene and polycarbonate induced by plasma treatments

Euradh '96. Adhesion '96.  Vol 1 Conference proceedings.

Cambridge, 3-6th Sept 1996 pp 325-330  9(12)4

4.    Nowak S. et al

Surface characterisation & adhesion of plasma treated PP

Polymeric materials science and engineering. Vol 62  Conference proceedings. Boston Mass.  Spring 1990 pp 437-441  012

5.    Andre V. et al

In-situ metallization of PP films pretreated in a nitrogen or ammonia low-pressure plasma

Thin Solid Films  vol 181,  No 1/2 , 10th Dec 1989  pp 451 - 460

There are always a number of options that you can take.

One would to be keep the total pressure constant and also the argon flow constant and then use the reactive gas flow to maintain the total pressure constant. In using two reactive gases these can be added as a fixed ratio.

Alternatively the gasses can be added as fixed flows and the total pressure can be allowed to vary.

I personally would tend to use the first method.  Although the flow of the reactive gas can vary I believe that, as the supply of reactive gas is always expected to be in excess of what is required for the desired effect, small variations of gas flow are not significant.