Improved Cure Speed and Efficiency of UV Curing Systems Using Silicone Acrylates. Verbesserung der Härtungsgeschwindigkeit und der Aushärtung bei der UV Härtung in Siliconacrylatsystemen. Hardi Döhler Industrial Specialties, Goldschmidt AG 45127 Essen, Goldschmidtstraße 100, Germany Tel.: +49 201 173 2668, Fax: +49 201 173 1933 E-mail: [email protected]

Abstract: Oxygen inhibition in the free radical UV curing of acrylate systems is a well known obstacle to cure. This paper will summarise the efforts made to overcome this problem and introduce an improved initiator system that allows for a better and faster cure of silicone acrylates. Zusammenfassung: Sauerstoffinhibierung ist ein bekanntes Problem bei der radikalischen UV Härtung von Acrylatsystemen. Dieser Vortrag faßt die bisher gemachten Anstrengungen zur Kompensierung dieses Nachteiles zusammen und stellt ein verbessertes Initiatorsystem vor. Damit wurde bei der Härtung von Siliconacrylaten eine schnellere und vollständigere Aushärtung erreicht.

Introduction This paper is going to deal with the curing performance of the radical curing silicone acrylate system. Firstly we will look how silicone acrylate systems compare to both organic acrylate systems and epoxy type UV curing silicones. Then we will discuss the mechanism of UV induced radical polymerisation and the role of oxygen inhibition. Finally, recently investigated improvements for radical UV curing silicone systems will be introduced.

Comparison of silicone and organic UV curing acrylate systems We O

CH3 C

O

CH3

O

CH3 CH3

CH3

H3C SiO Si O Si CH3 CH3

R m

O O

CH3

O

distinguish

between

organic acrylates and silicone

O

acrylates simply by looking if the O

Organic acrylate O Si CH3 (Bisphenol-A epoxyacrylate) CH3 and silicone acrylate n (i.e. Goldschmidt RC Silicone)

products

contain

any

silicone

groups. In fact, most of the commonly used UV systems are silicone free products. All the major

suppliers,

BASF,

UCB, 1

Sartomer and others, focus on 100 % organic products. There are only a few fields of applications for silicone acrylates such as silicone release coatings and reactive performance additives. The latter would be added to organic systems in small amounts to result in improved wetting, slip or defoaming characteristics of the organic acrylate system. Silicone acrylates are used as 100 % products for release coatings. These products are supplied almost exclusively by Goldschmidt. One major difference between organic and silicone acrylates is the content of acrylic groups and therefore their reactivity in the UV curing process. Organic coatings are widely used as varnishes or inks to provide a good cured coating with sufficient hardness, flexibility and surface scratch resistance. Silicone release coatings, however, in addition to hardness and flexibility must provide good release stability against adhesive materials. The acrylic content in organic acrylates is usually significant higher compared to that in silicone acrylates. It is possible to manufacture

silicone

acrylates with very high acrylic content, but this will

lead

to

release

coatings

that

will

perform

well

in

not easy

release applications. The challenge is to adjust the acrylic content to find the

Acrylate Easy Release Silicone Controlled Release Silicone

Trade Name Acrylic Content RC 902 4% RC 715 5%

Hexanedioldiacrylate Tripropyleneglycoldiacrylate Aliphatic Urethane Acrylate, hexafunct. Glycerinetriacrylate, ethoxylated (5,5 EO) Bisphenol A Epoxydiacrylate Aromatic Urethane Acrylate, difunctional

HDDA TPGDA Ebecryl 1290 SR 9021 Ebecryl 600 Ebecryl 210

64% 48% 43% 38% 29% 10%

Acrylic content of some comercial acrylates

right balance between release properties and the reactivity in the UV curing process. Organic acrylates, on the other hand, can be highly acrylated, achieving high cure speeds without harming their performance in the cured coating. It is well known, that all acrylate systems suffer oxygen inhibition. The mechanism will be discussed later, but the result is a loss of photoinitiator and acrylic groups that cannot take part in the polymerisation but form hydroperoxides. Organic acrylate systems compensate this loss by a so-called “chemical inerting”. This is merely to utilise an excess of acrylic groups, photoinitiator and UV power. As long as the coating is curing with sufficient hardness, it does not matter if some of the reactive groups are lost in the formation of hydroperoxides. However, this is quite different in a silicone system. Here, the acrylic group content and the photoinitiator concentration cannot be increased, as this would impact the release properties. In addition, hydroperoxides at the surface of the silicone coating are definitely unwanted, as they may chemically react with adhesive components and harm the 2

release stability. Thus, oxygen inhibition has a major impact to the quality of the silicone release coating. The best way to avoid hydroperoxide formation and to guarantee full cure is to reduce the oxygen content. Therefore, nitrogen inerting is absolutely necessary.

Comparison of acrylic and epoxy type silicones There are two types of UV curable silicone release coatings available: UV catalysed cationic curing epoxy silicones and UV initiated free radical curing silicone acrylates. Here, the term “silicone” is somewhat misleading as both systems are actually organo-modified silicones (OMS). The position of the organic modification as well as the ratio between silicone CH3 CH 3

CH 3

H3C SiO Si O Si CH3 CH 3

CH 3

CH3 CH3

O Si CH3

H3C SiO Si O Si

CH 3

CH3 CH3

m

n

the quality of the product.

O Si CH3

R m

n

O

organic

moieties are important for

CH3

CH3

and

While the silicone content of

CH3

acrylic and epoxy silicones can be very similar, the

O

synthesis of acrylic OMS

O

allows the manufacture of

Epoxy silicone and silicone acrylate, both on chain modified

end

chain

functionalities

with high silicone content and sufficient high acrylic content. The synthesis of epoxy silicones, however, needs on chain modification to achieve a sufficient functionality. As already discussed, the major drawback for acrylic type silicones is the need for nitrogen inerting. One of the drawbacks of epoxy type silicones is the need for on chain modification, which eventually results in higher release levels. In addition, the crosslinking mechanism of cationic type silicones leads to some problems, especially post-curing and poisoning. The free radical curing mechanism is extremely fast. rel. Intensity UV Exposure (flash)

1,0

0,8

0,6

Epoxy Silicone

0,4

Silicone Acrylate 0,2

0

1

2

3

4

5

6

7

8

9

10

Time / sec

Cure speed of epoxy and acrylate system measured via real time IR

3

Within a few milliseconds - the time the web is in the inerted UV chamber - the curing is completed and there will be almost no post-curing. That makes the free radical technology preferable for high-speed curing and in-line conversion processes. The curing of epoxy type silicones is initiated by UV light and will continue to cure well beyond the UV unit and may only be fully completed hours after initiation. If adhesive coating or rewind comes too soon (i.e. line speed too high), the silicones may not have reached a sufficient stage of polymerisation and thus continue to harden within the rewind roll or in adhesive contact. This can result into variable and unstable release levels. Extremely strong UV induced acids start the cationic curing of epoxy silicones. These acids are subject to several poisoning reactions. Printed surfaces, LDPE, PVC and most papers often cannot be siliconised due to poisoning ingredients which reduce the efficiency of the acidic catalyst. Air humidity can have a similar effect. Nevertheless, cationic curing silicones have their fields of applications. In case of advanced applications, however, radical curing silicones are called for. It is then important to handle the oxygen inhibition in an appropriate fashion. UV units with proper nitrogen inerting are now state-of-the-art.

Mechanism of radical UV curing For UV curing, we need UV light of the appropriate energy to cleave a photoinitiator into radicals.

For

silicone

acrylates,

this

photoinitiator

is

often

2-Hydroxy-2-methyl-1-

â

phenylpropane-1-one (HMPP), e.g. Darocur 1173 from Ciba Additives. This photoinitiator is an α-cleaver, breaking up next to the carbonyl group, which is absorbing the UV light, into two radicals. Another photoinitiator with O CH3 C C OH

UV

O C

+

CH3 Photo Decomposition of Darocur 1173

CH3

similar photo decomposition mechanism

C OH

is

CH3

TEGOâ PC

750,

a

premixed

photoinitiator/ silicone blend. The most important UV range for the photo decomposition is about 230 –

270 nm. When a photon of the right energy is absorbed by the carbonyl group, the exited state has a lifetime of about 30 ns. Either the photoinitiator breaks into radicals or falls back into ground state. This means, that 30 ns beyond the UV light area, there will be no new supply of radicals. At 200 m/min web speed, this is a 10 µm distance. This, and the oxygen termination in air, explains why the reaction is completed within the reaction chamber. After α-cleavage there is a number of subsequent reactions. When first generated, the radicals are very close and trapped in the surrounding silicone matrix. In this so-called solvent cage there will be a high rate of recombination and hydrogen abstraction. The latter leads to the formation of Benzaldehyde and Acetone, the recombination to the original 4

molecule. The yield of radicals being released from the solvent cage does not depend on the initial concentration of photoinitiator or the UV light dose. However, there may be a higher yield if the radicals are better soluble in the matrix and therefore can more easily escape the solvent cage. Off cage reactions include recombination of monomeric radicals now forming the original molecule, Pinakole and Benzile. This will again result in a loss of start radicals. Off cage recombination will be depending on the initial concentration of photoinitiator and on photon concentration (UV light per m²). If there are more radicals off cage available, they will more often recombine and not start radical polymerisation. This loss is the reason that •

double photoinitiator content cannot guarantee double line speed

•

two UV lamps at half power can reach higher speeds than one at full power

•

focussed UV light is not as good as diffused UV light.

We therefore prefer to form radicals uniformly over a wider area, i.e. have two UV lamps at moderate UV power with diffused reflector systems. Focussed reflector systems with a maximum UV power output will lead to a peak formation of radicals with a relatively higher

UV Power [W/m²]

UV Radiation with a 120 W/cm Lamp

16000

14000

12000

Focussed Reflector in Microwave Lamps

10000

8000

6000

Diffused Reflector in Arc Lamps

4000

2000

0 0

20

40

60

80

100

120

Web Pass under Lamp [mm]

rate of radical recombination. It is the absolute amount of radicals formed that will be important, not having them all at once. Surviving radicals will eventually seek an acrylic group to start the free radical polymerisation. At this point, the solubility of the photoinitiator radicals is important for their ability to move around in the silicone matrix in order to find the acrylic groups and convert them to a very high degree.

5

Novel Photoinitiator for RC Silicones Darocurâ 1173 is a widely available and well investigated UV photoinitiator. It is a liquid and thus easy to blend with silicone systems. In search of suitable photoinitiators for silicone acrylates, it is most important to have them in a liquid and silicone mixable form. Although Darocurâ 1173 and its resulting radicals blend with the silicone system, they are not really hydrophobic molecules. Thus, they may not easily escape the solvent cage and may not be able to freely move around within the silicone matrix to react with the acrylic groups. We therefore concentrated our investigations to find a more hydrophobic photoinitiator. This new photoinitiator will be available from Goldschmidt under the trade name TEGOâ A16. Besides the hydrophobic characteristic, the new photoinitiator should have a high UV absorption in the range where medium pressure mercury lamps generate UV light. The

UV Spectra of Photoinitiators 1,5 1,4

Darocur 1173

1,3

TEGO A16

1,2

Hg Spectrum Arc Lamp

1,1 Absorbtion

1,0 0,9 0,8 0,7 0,6 0,5 0,4 0,3 0,2 0,1 0,0 220

230

240

250

260

270

280

290

300

310

Wavelength in nm

absorption is not important in the range of 200 to 230 nm as the silicones themselves absorb in this range. Above 230 nm there are many UV peaks in the UV spectra of the lamps. The more of these peaks are absorbed by the photoinitiator, the better. The UV spectra shows that TEGOâ A16 is utilising more of the UV peaks from standard UV lamps, especially the peak at 255 nm. Even though the absorption of UV light is important, the yield of radical conversion from the exited state is even more so. Eventually, all these factors determine the cure speed potential of a photoinitiator. To measure the cure speed it is best to run a coating trial. To set up such a trial, one can increase the machine speed or reduce the UV output to the critical 6

stage, where the degree of cure is not fully developed. The degree of cure can be measured very well with the subsequent adhesion test. It will give us an indication of how many

Different Photoinitiators in RC 902/RC 711 70:30 with one 120 W/cm UV Lamp

Subsequent Adhesion [%]

100,0 ~ 300 m/min line speed 95,0

~ 80 m/min line speed

90,0 85,0 80,0 75,0 70,0 Darocur 1173 2%

PC 750 15%

TEGO A 16 2%

Photoinitiator and Content

uncrosslinked (and therefore transferable) silicone molecules are left. By running different photoinitiators under comparable conditions we can see the advantage of using a more hydrophobic photoinitiator system. The release value, and more important the stability of the release level with time, will tell us about the conversion rate of the acrylic groups (and the presence of hydroperoxides). With TEGOâ A16, the stability of release against aggressive adhesives such as solvent borne acrylics, was further improved. Another advantage of TEGOâ A16 is that this photoinitiator does not produce Benzaldehyde as a photoreaction by-product. This almond flavoured odor is often present when using Darocurâ 1173 (in combination either with silicone or organic acrylates). Furthermore, Benzaldehyde deposites inside the reaction chamber, especially causing the quartz plates to cloud. With TEGOâ A16, this problem is much reduced and the odor is eliminated.

Mechanism of oxygen inhibition Photoinitiator radicals or acrylic macroradicals react with oxygen to form peroxy radicals. The reaction rate for the oxygen termination of carbon-centred radicals is very high. In fact, this reaction is so fast that all dissolved oxygen will be consumed even before the radical polymerisation of acrylic double bonds begins[1]. The addition of oxygen to radicals results in the transformation of highly reactive carbon-centred radicals into less reactive peroxy

7

radicals, which do not participate in the polymerisation process. Photoinitiator radicals and acrylic groups are therefore removed from the crosslinking process. Furthermore, peroxy radicals will abstract hydrogen from the environment, tertiary

or

preferably secondary

from

carbons

resulting in hydroperoxides and carbon

centred

Radical Polymerisation Cycle P- I (Photoinitiator)

UV

Radical Polymerisation Oxygen

radicals.

Eventually, these radicals will add

R· + ROOH

oxygen, form new peroxy radicals

ROO·

and abstract hydrogen in a fast chain

process,

known

as R- H

autooxidation. For example, in polypropylene

degradation,

an

excess of oxygen can result in cascade formation of hydroperoxides. It is assumed that hydroperoxides (and residual acrylic groups) are responsible for unwanted release properties such as high release values and unstable release values. The best way to avoid hydroperoxide formation and to guarantee full cure is to reduce the oxygen content. As mentioned before, nitrogen inerting is absolutely necessary for silicone acrylate systems. But nitrogen blanketing has not been the only effort to treat oxygen inhibition.

Oxygen scavenging methods Many attempts have been made to overcome oxygen inhibition. As mentioned before, organic acrylic products provide a sufficient high amount of acrylic groups, high photoinitiator concentration and high UV dose for no other reason than to chemically absorb all dissolved and boundary layer oxygen. It is also very common to include tertiary amines in the formulation. They expose their αhydrogens to the peroxy radical even more than tertiary carbons. The resulting new radicals will attract oxygen, thus consume oxygen by the mass formation of aminofunctional peroxides[2]. As mentioned before, hydroperoxides do not harm the properties of organic acrylates but they are suspect to cause release instability in silicone acrylate release coatings. Tertiary amines are therefore not useful to enhance the degree of cure in silicone acrylates. A number of attempts have been made which have not been approved for UV curable formulations. One such method is the triplet forming of oxygen with suitable radiation (red light)

and

methylene

blue

as

[3] sensitizer .

The

subsequent

reaction

with

1,38

Diphenylisobenzofurane is said to be suitable for consumption of dissolved oxygen in UV curing processes. Most promising yet is work done by Miller et al[4]. Unfortunately, N-Vinylpyrrolidone (NVP) is involved which is considered to be carcinogenic. Although the reaction mechanism is not fully understood, NVP may also act as hydrogen donor and thus consume oxygen again by peroxide formation. However, in case of NVP the hydroperoxides would be crosslinked into the polymer network as NVP undergoes a radical polymerisation. The oxygen scavenging effect of phosphites is known and they are used as flameretardants (in form of phosphonates). Phosphites are also commonly used as antioxidants in plastics. The oxygen scavenging works via the oxidation of phosphorus from its 3 valent to 5 valent stage to form phosphates: (RO-)3P + ½ O2 Þ (RO-)3P=O Oxygen scavenging for the stabilisation of plastics is a long term effect, a slow reaction in the weathering of outdoor plastics. In UV curing, this reaction has to be extremely fast to be efficient. Besides their use as plastic stabilisers and flame-retardants, phosphites are rarely described

as

an

oxygen

[5]

scavenger . They have not yet found their way into acrylic

Radical Polymerisation Cycle with Phosphite P- I (Photoinitiator)

UV curable formulations.

UV

Radical Polymerisation

In our attempts to develop a better

initiation

system

Oxygen

for

silicone acrylates, we have now

R· + ·OH

R· + ROOH

ROO·

discovered that the reaction rate of phosphites with some oxygen donors compete

is

high with

enough the

to

Phosphate

Phosphite

R- H

oxygen

addition to carbon centred radicals. In the presence of phosphites or other oxygen scavenging antioxidants, peroxy radicals are transformed into hydroxyl and carbon radicals which participate again in the radical curing of acrylics and thus enhance the degree of cure. The degree of crosslinking in an enriched oxygen environment is considerably improved and the formation of hydroperoxides reduced. This results in a much better curing performance of silicone acrylate systems at higher levels of residual oxygen. Phosphites are also capable of enhancing the degree of cure in organic acrylic systems.

9

Improved cure speed with Goldschmidt´s new photoinitiation system This paper is concentrating on the silicone acrylate system and therefore continues with some practical results found with Goldschmidt´s RC Silicone acrylates. Both subsequent adhesion and release stability in long-term ageing can be enhanced either with TEGOâ A16 (at less than 50 ppm residual oxygen) or the combination of TEGOâ A16 and phosphite (at higher residual oxygen levels). The phosphite stabiliser generates reactive radicals in addition to those coming from the photoinitiator decay. Therefore, the amount of oxygen inhibition by-products on the silicone interface will be reduced. A higher oxygen level will not harm the silicone release properties. The oxygen scavenging is especially effective when the oxygen level is in the range of 50 to 400 ppm. Influence of Residual Oxygen on RC 902/ RC 711 100

Subsequent Adhesion TESA 7475 [%]

90 80

TEGO A16/ Phosphite

70 60

PC 750

50 40 30 20 10 0 0

50

100

150

200

250

300

Residual Oxygen Content [ppm]

The stability of the release level, however, is a more important measure than relying just on the subsequent adhesion. Ageing stability with standard photoinitiators is shown, when the residual oxygen content is safely in the 50 ppm production window. Most inerted UV units run extremely reliable below the recommended 50 ppm level. In fact, a stable inerting process is usually settling between 15 and 30 ppm and will not need any adjustments. However, at higher oxygen content the ageing, especially with hot melt adhesives, is increasingly unstable. With the TEGOâ A16/ phosphite combination, the ageing stability is now considerably improved. The reliability of the inerting process is much increased as it will be possible to run with up to 200 ppm residual oxygen.

10

Ageing of a Standart HotMelt at 40°C with RC 902/RC 711 100

Release [cN/in]

90 80

200 ppm, PC 750

70

200 ppm, TEGO A 16 with Phosphite

60

20 ppm, PC 750

50 40 30 20 10 0

1d

7d

1M

3M

6M

Ageing time

Even rudimentary inerted UV equipment may now be able to cure RC Silicones with acceptable results. One example: a fast running coating line in Europe with an inerted UV unit can reach speeds of approx. 400 m/min to produce coatings with good release properties. At higher speeds, the release properties are influenced negatively by increasing residual oxygen levels. However, with the TEGOâ A16/ phosphite combination, subsequent adhesion values now are fully acceptable on this line up to 700 m/min.

Effect of Phosphite at Coating Lines with Critical Inerting at High Speeds 100

Subsequent Adhesion [%]

90

with Phosphite

80 70

Standard

60 50 40 30 20 10 0

400

700 Line Speed [m/min.]

The new photoinitiator/ phosphite combination is an advantage with well inerted UV units as well. The system will give better stability at the adhesive/silicone interface and thus improve the release performance of Goldschmidt RC Silicones. 11

Outlook/ Summary A new photoinitiator with improved silicone solubility has been made available which is capable to enhance the degree of cure and the release performance of silicone acrylates. The improved reliability of the curing process for silicone acrylates by means of a new oxygen scavenger class has been proven. With the addition of phosphites it is now possible to run UV units at higher oxygen levels, e.g. 200 ppm and still achieve high quality release coatings. This process can possibly reduce nitrogen consumption and thus enable customers to save money. Phosphites can help to further improve the silicone/ adhesive release properties even at perfect inerting conditions. The usefulness of phosphites is particularly important with UV units running at critical inerting levels (low quality inerted systems). It will also be useful in processes where rough substrates such as non-woven materials and machine grade papers (e.g. business forms) are used. Both the possible cost savings and the extra reliability will help to further improve the reputation of RC Silicones.

Literature [1] U. Müller, „New Insights in the Influence of Oxygen on the Photocrosslinking of Silicone Acrylates“, in: Organosilicon Chem. IV, 4th (2000), Meeting Date 1998, 663-666. WileyVCH Verlag GmbH, Weinheim, Germany [2] K.K. Dietliker, Chemistry & Technology of UV & EB Formulation for Coatings, Inks & Paints, Volume 3, Sita Technology Ltd, UK, 1991, pp. 83 [3] C. Decker, „A Novel Method for Consuming Oxygen Instantaneously in Photopolymerizable Films“, in: Macromol. Chem. 180,2027-2030, 1979 [4] C.W.Miller et al, „Analysis of the reduction of oxygen inhibition by N-Vinylimides in free radical photocuring of acrylic formulations“, in: RadTech 2000, Technical proceedings [5] C.R.Mogan et al, „UV generated oxygen scavengers and their effectiveness in photopolymerizable systems“, in: Journal of Radiation Curing, October 1983

12