- Email: [email protected]
Time and cost effective methods for testing chemical resistance of aluminium metallic pigmented powder coatings
Progress in Organic Coatings 63 (2008) 49–54
Contents lists available at ScienceDirect
Progress in Organic Coatings journal homepage: www.elsevier.com/locate/porgcoat
Time and cost effective methods for testing chemical resistance of aluminium metallic pigmented powder coatings a,∗ ˚ Jostein Mardalen , John Erik Lein b , Helene Bolm c , d Merete Hallenstvet , Volker Rekowski e a
SINTEF Petroleum Research, N-7465 Trondheim, Norway SINTEF Materials and Chemistry, N-7465 Trondheim, Norway DuPont Powder Coatings Scandinavia AB, S-593 25 V¨ astervik, Sweden d Hydro Aluminium Rolled Products AS, P.O. Box A, N-3081 Holmestrand, Norway e DuPont Performance Coatings GmbH, Horbeller Street 15, D-50858 K¨ oln, Germany b c
a r t i c l e
i n f o
Article history: Received 3 January 2008 Accepted 17 April 2008 Keywords: Powder coatings Chemical stability Metallic pigments Pigment coating
a b s t r a c t Polyester based powder coatings containing different types of aluminium metallic flake pigments have been investigated with respect to their chemical stability in acid environments. The metallic flakes are made chemically stable by covering them in silica. The degree of silica coverage and the silica morphology are far more important for the chemical stability of the pigments than the silica thickness. The acid resistance of the final powder coating is found to depend on the pigment embedment depth, on the chemical composition and morphology of the powder coating, and on the silica coating of the pigment. The latter being the single most important factor in our study. New chemical resistance tests have been developed. One for the testing of the metallic pigments alone before embedding the pigment in a coating, and another for the cured coating containing these pigments. Additionally TEM cross-sections have been used for investigating the quality of the silica coating on pigments, and optical absorption for comparison of pigment embedment depth. This approach to powder coating development emphasizing the measurement of the quality of each component separately has proven to be robust, reliable, and time and cost effective. By testing of separate components the need for time-consuming quality cheques of cured coatings is reduced while maintaining the ability to predict final paint properties. Further, the uncertainty related to testing many process parameters simultaneously is considerably reduced. © 2008 Published by Elsevier B.V.
1. Introduction Powder coating has been established as one of the most efficient coating techniques available. It offers advantages as no or very little volatile organic compounds (VOC), easy application procedures and very high efficiency due to recycling of excess powder. Powder coating has a low impact on health and environment. The level of toxicity has been reduced, and the problems associated with explosive hazard of powder in air have been solved [1–3]. These facts combined with very good economics have made powder coating an attractive coating technology. It has been claimed that the development in the powder coating industry mainly has focused on cost and health, environment and
∗ Corresponding author. Tel.: +47 98 28 39 10; fax: +47 73 59 11 02. ˚ E-mail address: [email protected] (J. Mardalen). 0300-9440/$ – see front matter © 2008 Published by Elsevier B.V. doi:10.1016/j.porgcoat.2008.04.007
safety issues rather than new product development [2]. The development of new products and new manufacturing techniques has been slower than predicted some years ago. One important reason for this is that the test procedures needed for development work are slow (time-consuming exposures, such as the Florida test), expensive and not sufficiently robust. The comprehensive test programs as specified by GSB [4], Qualicoat [5] and others are designed for the purpose of quality assurance of coating products entering the market, but are far to time and cost ineffective to be suitable for R&D work. These test are (mainly) on finished coatings, and do not address optimization of the quality of single components or constituents in the coating. The tests are not suitable to monitor the effect of changing one single parameter because there are too many uncertain factors. This can to some extent be encountered for by statistics, but this will further increase development cost and time consumption. There is, therefore, a general demand for robust, reliable, quick and cheap experiments isolating single coating components.
50
J. M˚ ardalen et al. / Progress in Organic Coatings 63 (2008) 49–54
Table 1 Overview of the different commercially available aluminium flake pigments. The grain size “Fine” refers to D50 values <20 m and “Coarse” refers to D50 > 30 m Pigment no.
Metal
Type
Shape
Surface coating
Size
Predicted chemical durability
I II III IV V VI
Aluminium Aluminium Aluminium Aluminium Aluminium Aluminium
Non-leafing Non-leafing Non-leafing Non-leafing Non-leafing Non-leafing
Flake Flake Flake Flake Flake Flake
Silica Silica Silica Silica Silica Silica
Fine Fine Fine Coarse Fine Fine
Standard Standard Improved Improved Extra reinforced Extra reinforced
The “predicted chemical durability” is based on information from the suppliers.
Metallic effect coating is an established technology [6] and aluminium pigments for powder coatings applications have been focused by the pigment suppliers lately. The leafing pigments, having a tendency to float to the surface of the coating and to orient parallel with the coating surface, give a bright silver-like finish. The disadvantage with leafing aluminium pigments is the poor rubbing, weathering and chemical resistance unless a clear topcoat is applied. Semi-leafing aluminium flake pigments have a potential of giving powder coatings with good in-service properties without having to apply a clear topcoat. This put some requirements on the pigment properties and has been the reason why pigment suppliers have worked extensively on pigment coating optimization. Many suppliers have focused on rubbing, weathering and chemical stability together with the powder application properties. New variants of organic and inorganic pigment protective layers have been developed recently, including sol–gel encapsulations techniques [7–9]. Some of these have also been introduced to the market. In the present work we report on a powder coating development work where we have used non-leafing aluminium flake pigments protected by an inorganic silica coating, and where the requirements for rubbing, weathering and chemical stability has been normative. The chemical stability of the coating has been important to monitor. The Acidified Salt Spray (ASS) test, originally designed for corrosion measurements, has been used for this in combination with colour measurements. This test is, however, very time and cost consuming. In the present work we have (i) developed a new procedure for testing the chemical stability of the pigment, (ii) successfully used transmission electron microscopy (TEM) to visualize and evaluate the quality of the pigment coating, (iii) developed a rapid and cost effective test for the chemical stability of the cured coating and (iv) suggested the use of optical absorption to qualitatively measure the pigment embedment depth. The development and utilization of these techniques did not only speed-up the development process, they also gave us a far more detailed knowledge on how different parameters and properties influenced the final coated product. The measurements and tests were far more robust and reliable for R&D work than standardized quality tests.
2. Experimental 2.1. Samples Six different commercially available aluminium based metallic pigments have been investigated in this study (Table 1). The pigments are all of a non-leafing type and are protected by a silica coating. Silica is throughout this work conveniently used as a description of a Si-containing protective layer on the pigment particles. The silica might be in different microstructural and chemical forms including various oxides and hydroxides. The chemical stability of the pigments is defined into three different categories based on information given by the suppliers. These pigments are all added to the same polyester based powder coating.
The powder coating used is based on polyester binder, curing agent, filler and additives according to a commercial formula used by DuPont. The components are mixed using a single screw Buss extruder, and grinded to medium PSD of 40 m. The grinded powder was mixed with the commercial available aluminium flake pigments in Table 1 with a predetermined concentration. By using different process conditions, like various bonding techniques or dry blending methods when mixing base powder and the aluminium pigments together, it is possible to achieve different embedment depths of the aluminium pigments. Two different mixing procedures have been used. Procedure 1 gave a deep pigment embedment and Procedure 2 gave a shallow and more surface oriented pigment embedment. Application of the coatings has been done using a handheld electrostatic powder gun and cured in a convection oven for 10 min at 200 ◦ C giving a film thickness of 80–100 m. 2.2. Measurements The chemical stability of the pigments was tested by exposing 0.3 g pigment to 100 ml 10% w/w HCl at 25 ◦ C for 3 h. The amount of gas evolved was measured during the exposure. A visual evaluation of the pigments before and after exposure in HCl was also done. 10% w/w HCl is prepared from concentrated (37%) HCl by adding 285.8 g acid to 714.2 g distilled water giving 1000 g solution. Chemical resistance of the aluminium-pigmented coatings was investigated using an Acidified Salt Spray (ASS) test following the ASTM G85–94 standard [10]. This test is originally designed for measuring corrosion resistance of coated aluminium. In the present work, however, it is used for monitoring the degradation of the aluminium flake pigments during exposure to an acidified salt spray atmosphere. A new cost and time effective chemical resistance test for the cured powder coating is proposed. The test is denoted the HCl dipping test and consists of dipping the samples in 10% w/w HCl for some hours. As will be discussed in the result section there is a very good correlation between the ASS test and the HCl dipping test. Chemical degradation of the samples is monitored during exposure both to ASS and HCl by measuring the L-value in standard colour measurements [11]. Samples are taken out of the cabinet/bath, and they are rinsed thoroughly in tap water followed by distilled water before drying. This rinsing procedure is arresting the degradation process. Exposure time is restarted on re-entrance of the samples in the ASS chamber or HCl bath. The L-value is used as a quantitative measure of how much visible light a surface will reflect. Degradation of a surface will normally result in a decrease in the measured L-value, i.e., the surface becomes darker. A reduction in L-value by 1 percentage point is normally not visible for our eyes. The L-values given are averages for 3 repeated readings on three predefined locations on three parallel samples. Optical absorption (OA) measurements are in this paper used as a qualitative measure for the average pigment embedment depth. OA is performed in reflection geometry at = 8.4 m. This wave-
J. M˚ ardalen et al. / Progress in Organic Coatings 63 (2008) 49–54
51
length is away from typical absorption lines of the polymer coating, but in a regime where the general absorption is found to scale well with the polymer thickness. Cross-section micrographs have also been used to visualize and measure pigment embedment depth, but we have found that these local spot tests are not sufficiently representative for the average embedment depth. A Philips CM-30 Transmission Electron Microscope (TEM) was used for cross-sectional examination of the pigments and their protective coatings. The pigments are best analyzed when embedded in a polymer matrix. The simplest procedure is, therefore, to investigate pigments in the final cured powder coating, and to strengthen the samples by further embedment in an epoxy resin. Cross-section samples for TEM analyses were made by cutting thin foils of the embedded samples using a Reichert-Jung ultramicrotome. Final sectioning was performed with a MicroStar diamond knife. 3. Results and discussion
Fig. 1. Chemical stability of pigments in hydrochloric acid. Gas evolution vs. exposure time.
3.1. Chemical stability of pigments The chemical stability of the pigments was quantified by monitoring the gas volume developed during exposure to hydrochloric acid as shown in Fig. 1. There is a large difference in the susceptibility for degradation (etching). Pigments I and II with a predicted “standard” chemical durability show considerable gas evolution. After some time the gassing curves level off indicating that all material has reacted. This was confirmed by visual inspection. The two pigments with “improved” chemical durability (pigments III and IV) perform considerably better than pigments I and II. The two “extra reinforced” (the two best) pigments show only very little gas evolution, indicating an excellent chemical stability in hydrochloric acid.
It is worth to observe that the coarser of the two “improved” pigments (pigment IV) performs well and also better than expected when judged from the information given by the supplier in Table 1. This apparent discrepancy is certainly influenced by the pigment size. Coarser pigments will have a smaller effective surface area for a given weight, and will consequently react slower. Thus, the suggested pigment test works best when comparing pigments of equal size. Note also that there is a certain delay before the gas evolution starts. This is particularly evident for pigments I, II and III. This is most probably linked to the fact that it takes some time for the acid to penetrate the protective silica layer.
Fig. 2. TEM cross-sections of flat, flake-shaped aluminium pigment particles. The protective silica coatings are visible on the aluminium surfaces. There are some variation of the thickness and the quality of the silica coating. Note that the coarse pigment (IV) is shown with another magnification than the rest.
52
J. M˚ ardalen et al. / Progress in Organic Coatings 63 (2008) 49–54
Table 2 Observations of the silica coating of the different aluminium pigments from the TEM investigations (i.e., Fig. 2) Pigment
Silica thickness (nm)
Silica morphology
Silica coverage
I II III IV V VI
10–30 10 20 50 5 10
“Fluffy” Slightly “fluffy” Probably some cracks Thick with some cracks (?) Thin and compact Thin and compact
Partially covering Partially covering Fully covering Fully covering Fully covering Fully covering
In order to find the reason for the large difference in the chemical stability TEM cross-section images were made (Fig. 2). The TEM investigations revealed information on the thickness and morphology of the silica coatings. The main results are summarized in Table 2. The TEM images show that the large difference in the chemical stability of the aluminium pigments is mainly determined by the quality of the silica protective coating. Comparing the chemical stability results from Fig. 1 with the silica coating quality (Fig. 2 and Table 2) there are some evident correlations. The silica coating on pigments I and II seems to adhere poorly to the aluminium pigment resulting in only partial coverage and a sort of “fluffy” or porous appearance in the TEM cross-section images. In some places the silica even seems to be “dissolved” or lifted out from the pigment surface and into the polymer-coating matrix. Other pigments (not shown here) are even fluffier and have even poorer chemical stability than I and II. Improved chemical stability is achieved for more compact and better covering silica coatings (III–VI). Even though pigment IV performs better than III we should not conclude that thicker silica coatings generally perform better. The pigment size is also coming into account as described earlier. In fact it seems that the two thinnest coatings (V and VI) are the best ones, showing that the quality of the silica and the degree of silica coverage is more important than the thickness of the silica coating layer. The permeability for hydrochloric acid in silica is obviously very low given a good morphology. Note also that the silica coating is very thin, 5–10 nm, for the most stable pigments (V and VI). We believe that this thin coating is the reason for pigments V and VI show some reaction with HCl already in the beginning of exposure, and do not show a distinct start-up-time as pigments I–III do (see Fig. 1). 3.2. Chemical stability of the powder coating In Fig. 3 the ASS and the HCl dipping tests are compared. The L-values are monitored for the different samples as a function of exposure times. The different coatings perform similarly in both tests and the qualitative ranking between the coatings is equal. The two coatings containing the “standard” pigments (I and II) degrade during exposure, typically after 200 h in ASS and after 8 h during HCl dipping (Fig. 3). The coatings containing pigments III and IV show considerably less degradation, and the coatings with the best pigments (V and VI) are somewhat better than the intermediate with the same pigment size (III). The powder coating containing the coarse pigment with a thicker silica coating (IV) also shows excellent chemical stability, comparable to pigments V and VI. This agrees with the results from the pigment test in Fig. 1. Coarser pigments will naturally be less reactive, and it also seems that a thicker silica layer of the “improved” quality is beneficial. However, comparing with the “extra reinforced” thin silica coatings (V and VI) the conclusion has to be that the quality of this protective coating is far more important than its thickness.
Fig. 3. (a) ASS test and (b) HCl dip test of coated samples made by Procedure 2, shallow pigment location.
Another effect of the larger pigment size is a lower initial L-value, 69 for the pigment IV coating vs. 76–82 for the other pigments. The excellent correlation between the ASS and the HCl tests seems to be independent of the pigment size. Fig. 3 clearly demonstrates the time and cost benefit of using HCl dipping as a technique for accelerated testing. The testing times can be reduced to about 1/30 compared to ASS. This enables a considerable speed-up of development of new coatings since the chemical stability testing often is the time limiting factor. In addition, the HCl dipping can be done without expensive and rather immobile laboratory facilities. It can even be simplified further either by just using visual inspection and comparison with references, and by applying droplets of chemicals on a coated panels rather than dipping (c.f. [8]). The authors have found a similar correlation when comparing standard Mortar test (24 h) [Ref. [4], Ch. 9.18, pp. 35] with dipping in NaOH for about 3 h. The Mortar test can thus also be speeded up considerably. The results are not shown here. When comparing the chemical stability of the pigment (Fig. 1) with the chemical stability of the cured powder coatings (Fig. 3), one find that the coating of the pigment is a decisive property for making chemically stable coatings. For the coatings reported in Fig. 3 it is probably also the most decisive parameter. However, a more chemically stable coating can be made if the pigments are embedded deeper in the polymer. A thicker polymer layer will then protect the aluminium flake pigments. This effect is visualized in Fig. 4 where the two pigment mixing procedures are compared. Procedure 1 gives a pigment embedment deeper into the polymer matrix (c.f. Fig. 5) and consequently a better chemical stability than Procedure 2.
J. M˚ ardalen et al. / Progress in Organic Coatings 63 (2008) 49–54
53
ing testing procedures beyond the described HCl-test. However, the small discrepancy between ASS and HCl is tolerable for our purpose and is not jeopardizing the usefulness of the HCl dipping test. The HCl curves in Figs. 3(b) and 4(b) show more “oscillations” in the L-values than the ASS counterparts. This looks a bit strange especially since the error bars indicate quite accurate readings. This is probably linked to these samples showing some L-variation across the sample (unevenness) which again is linked to the fact that the HCl dipping is a much faster process. There is obviously a trade-off between the required testing time and the testing quality, but used for semi-quantitative testing of the acidic resistance of a coating the HCl dipping test is by far good enough. The influence of the silica coating and the pigment embedment depth on the chemical stability of the powder-coatings has been demonstrated above. Another factor decisive for the chemical stability is of course the chemical composition and the morphology of the polymer coating itself. In the present work a state of the art polyester-based powder coating has been employed, and further optimization of the polymer is beyond the scope of this publication. 4. Conclusions
Fig. 4. (a) ASS test and (b) HCl dip test. Comparison of the two procedures for pigment mixing. The difference in chemical stability is linked to the pigment embedment depth and is controlled by the pigment mixing process, Procedure 1 (deep embedment) and Procedure 2 (shallow embedment).
Comparing the ASS and the HCl dipping tests for pigments with different embedment depths, Fig. 4(a) and (b), one observes that the ASS test distinguishes more clearly between the embedment depths than HCl-dipping does. This shows that the ASS test is slightly more dependent on the chemical composition and morphology of the polymer coating matrix than the HCl-dipping test is. This is understandable also from the general penetration time in the two cases, and means that one should be careful by accelerat-
Dipping of aluminium metallic pigmented coatings in 10% (w/w) hydrochloric acid is a cost and time effective alternative to testing chemical stability by acidified salt spray (ASS) test. A standardized 1000 h ASS test can be replaced by a much simpler HCl dipping test lasting about 32 h, allowing development work to be considerably speeded up. The chemical stability of the coating is determined by a combination of the chemical protection of the pigments by their coating (in the present case silica), the embedment depth of pigment in the polymer coating matrix, and the chemical composition and morphology of the polymer coating matrix. The chemical stability of the pigment themselves can be tested by monitoring the gas evolution and the colour change when exposing pigment particles to hydrochloric acid. Combined with TEM cross-section images, revealing information on the thickness and quality of the pigment coating, provides valuable information on the pigment quality and is decisive for the choice of pigment. The quality of the aluminium pigment coating is the single most important factor determining chemical resistance of the coating systems we have been investigating. Best protection is achieved with dense (not porous or “fluffy”) and fully covering silica coating on the pigments. The extra reinforced pigments are performing considerably better than other pigments at relatively shallow pigment embedment. Even a silica coating of 5–10 nm is sufficient for protection in acid environments. The influence of the pigment coating is smaller when the pigments are located deeper in the powder coating. Optical absorption measurements have been used to qualitatively measure the pigment embedment depth. A general R&D strategy of testing each constituent of the rather complex coating system separately has as far as possible been employed in the present work. This strategy gives more unambiguous and reliable results, it is more robust, and it reduces the need for many repeated tests for improved statistics. Acknowledgements
Fig. 5. Optical absorption measurements of the standard (I and II) and the extra reinforced (V and VI) pigments. Procedure 1 mixing of the powder and the pigment shows systematically higher absorbance meaning that the pigments are embedded deeper down in the coating. This correlates with the better chemical resistance of Procedure 1 mixed coatings.
This work has been partially financially by the Norwegian Research Council. The authors are indebted to Bjørn Steinar Tanem, John Walmsley and Ton Helvoort (SINTEF/NTNU) for their excellent
54
J. M˚ ardalen et al. / Progress in Organic Coatings 63 (2008) 49–54
sample preparation and their comprehensive TEM cross-section ¨ (DuPont) for skilful powder appliinvestigations, to Maria Nystrom cation and chemical resistance testing and to Lars Johnsen (SINTEF) for L-value measurements. References [1] Z.W. Wicks Jr., F.N. Jones, S.P. Pappas, Organic Coatings, Science and Technology, 2nd ed., Wiley Interscience, New York, 1999. [2] M. Cowley, Polymers Paint Colour Journal (March 1 2005). [3] T.A. Mishev, R. Van der Linde, Progress in Organic Coatings 34 (1998). [4] GSB International: International Quality Regulations For the coating of aluminium building components, Edition May 2007, www.gsb-international.de.
[5] Qualicoat: Specifications for a Quality Label for Paint, Laquer and Powder Coatings on Aluminium for Architectural Applications, 11th ed. Approved 17.11.2005, www.qualicoat.net. [6] CEPE European Council of Paint, Printing Inks and artists’ Colours Industry: Guidance note on the use of metallic effect powder coatings, http://www.cepe.org. [7] Hui Liu, Hongqi Ye, Yingchao Zhang, Colloids and Surfaces A: Physicochemical and Engineering Aspects 315 (1–3) (2008) 1–6. [8] Ulrich-Andreas Hirth: Aluminium Pigments for Powder coating Applications ¨ Paper V.3 from the 8th Nurnberg Congress 2005, and U. Hirth, Focus Powder Coatings 6 (2005) 2–3. [9] R. Supplit, U. Schubert, Corrosion Science 49 (2007) 3325–3332. [10] ASTM G 85 – 94. Standard Practice for Modified Salt Spray (Fog) Testing. [11] Reference [4] Ch. 9.19 pp. 40 and ISO 7724 and DIN 5033.
Contents lists available at ScienceDirect
Progress in Organic Coatings journal homepage: www.elsevier.com/locate/porgcoat
Time and cost effective methods for testing chemical resistance of aluminium metallic pigmented powder coatings a,∗ ˚ Jostein Mardalen , John Erik Lein b , Helene Bolm c , d Merete Hallenstvet , Volker Rekowski e a
SINTEF Petroleum Research, N-7465 Trondheim, Norway SINTEF Materials and Chemistry, N-7465 Trondheim, Norway DuPont Powder Coatings Scandinavia AB, S-593 25 V¨ astervik, Sweden d Hydro Aluminium Rolled Products AS, P.O. Box A, N-3081 Holmestrand, Norway e DuPont Performance Coatings GmbH, Horbeller Street 15, D-50858 K¨ oln, Germany b c
a r t i c l e
i n f o
Article history: Received 3 January 2008 Accepted 17 April 2008 Keywords: Powder coatings Chemical stability Metallic pigments Pigment coating
a b s t r a c t Polyester based powder coatings containing different types of aluminium metallic flake pigments have been investigated with respect to their chemical stability in acid environments. The metallic flakes are made chemically stable by covering them in silica. The degree of silica coverage and the silica morphology are far more important for the chemical stability of the pigments than the silica thickness. The acid resistance of the final powder coating is found to depend on the pigment embedment depth, on the chemical composition and morphology of the powder coating, and on the silica coating of the pigment. The latter being the single most important factor in our study. New chemical resistance tests have been developed. One for the testing of the metallic pigments alone before embedding the pigment in a coating, and another for the cured coating containing these pigments. Additionally TEM cross-sections have been used for investigating the quality of the silica coating on pigments, and optical absorption for comparison of pigment embedment depth. This approach to powder coating development emphasizing the measurement of the quality of each component separately has proven to be robust, reliable, and time and cost effective. By testing of separate components the need for time-consuming quality cheques of cured coatings is reduced while maintaining the ability to predict final paint properties. Further, the uncertainty related to testing many process parameters simultaneously is considerably reduced. © 2008 Published by Elsevier B.V.
1. Introduction Powder coating has been established as one of the most efficient coating techniques available. It offers advantages as no or very little volatile organic compounds (VOC), easy application procedures and very high efficiency due to recycling of excess powder. Powder coating has a low impact on health and environment. The level of toxicity has been reduced, and the problems associated with explosive hazard of powder in air have been solved [1–3]. These facts combined with very good economics have made powder coating an attractive coating technology. It has been claimed that the development in the powder coating industry mainly has focused on cost and health, environment and
∗ Corresponding author. Tel.: +47 98 28 39 10; fax: +47 73 59 11 02. ˚ E-mail address: [email protected] (J. Mardalen). 0300-9440/$ – see front matter © 2008 Published by Elsevier B.V. doi:10.1016/j.porgcoat.2008.04.007
safety issues rather than new product development [2]. The development of new products and new manufacturing techniques has been slower than predicted some years ago. One important reason for this is that the test procedures needed for development work are slow (time-consuming exposures, such as the Florida test), expensive and not sufficiently robust. The comprehensive test programs as specified by GSB [4], Qualicoat [5] and others are designed for the purpose of quality assurance of coating products entering the market, but are far to time and cost ineffective to be suitable for R&D work. These test are (mainly) on finished coatings, and do not address optimization of the quality of single components or constituents in the coating. The tests are not suitable to monitor the effect of changing one single parameter because there are too many uncertain factors. This can to some extent be encountered for by statistics, but this will further increase development cost and time consumption. There is, therefore, a general demand for robust, reliable, quick and cheap experiments isolating single coating components.
50
J. M˚ ardalen et al. / Progress in Organic Coatings 63 (2008) 49–54
Table 1 Overview of the different commercially available aluminium flake pigments. The grain size “Fine” refers to D50 values <20 m and “Coarse” refers to D50 > 30 m Pigment no.
Metal
Type
Shape
Surface coating
Size
Predicted chemical durability
I II III IV V VI
Aluminium Aluminium Aluminium Aluminium Aluminium Aluminium
Non-leafing Non-leafing Non-leafing Non-leafing Non-leafing Non-leafing
Flake Flake Flake Flake Flake Flake
Silica Silica Silica Silica Silica Silica
Fine Fine Fine Coarse Fine Fine
Standard Standard Improved Improved Extra reinforced Extra reinforced
The “predicted chemical durability” is based on information from the suppliers.
Metallic effect coating is an established technology [6] and aluminium pigments for powder coatings applications have been focused by the pigment suppliers lately. The leafing pigments, having a tendency to float to the surface of the coating and to orient parallel with the coating surface, give a bright silver-like finish. The disadvantage with leafing aluminium pigments is the poor rubbing, weathering and chemical resistance unless a clear topcoat is applied. Semi-leafing aluminium flake pigments have a potential of giving powder coatings with good in-service properties without having to apply a clear topcoat. This put some requirements on the pigment properties and has been the reason why pigment suppliers have worked extensively on pigment coating optimization. Many suppliers have focused on rubbing, weathering and chemical stability together with the powder application properties. New variants of organic and inorganic pigment protective layers have been developed recently, including sol–gel encapsulations techniques [7–9]. Some of these have also been introduced to the market. In the present work we report on a powder coating development work where we have used non-leafing aluminium flake pigments protected by an inorganic silica coating, and where the requirements for rubbing, weathering and chemical stability has been normative. The chemical stability of the coating has been important to monitor. The Acidified Salt Spray (ASS) test, originally designed for corrosion measurements, has been used for this in combination with colour measurements. This test is, however, very time and cost consuming. In the present work we have (i) developed a new procedure for testing the chemical stability of the pigment, (ii) successfully used transmission electron microscopy (TEM) to visualize and evaluate the quality of the pigment coating, (iii) developed a rapid and cost effective test for the chemical stability of the cured coating and (iv) suggested the use of optical absorption to qualitatively measure the pigment embedment depth. The development and utilization of these techniques did not only speed-up the development process, they also gave us a far more detailed knowledge on how different parameters and properties influenced the final coated product. The measurements and tests were far more robust and reliable for R&D work than standardized quality tests.
2. Experimental 2.1. Samples Six different commercially available aluminium based metallic pigments have been investigated in this study (Table 1). The pigments are all of a non-leafing type and are protected by a silica coating. Silica is throughout this work conveniently used as a description of a Si-containing protective layer on the pigment particles. The silica might be in different microstructural and chemical forms including various oxides and hydroxides. The chemical stability of the pigments is defined into three different categories based on information given by the suppliers. These pigments are all added to the same polyester based powder coating.
The powder coating used is based on polyester binder, curing agent, filler and additives according to a commercial formula used by DuPont. The components are mixed using a single screw Buss extruder, and grinded to medium PSD of 40 m. The grinded powder was mixed with the commercial available aluminium flake pigments in Table 1 with a predetermined concentration. By using different process conditions, like various bonding techniques or dry blending methods when mixing base powder and the aluminium pigments together, it is possible to achieve different embedment depths of the aluminium pigments. Two different mixing procedures have been used. Procedure 1 gave a deep pigment embedment and Procedure 2 gave a shallow and more surface oriented pigment embedment. Application of the coatings has been done using a handheld electrostatic powder gun and cured in a convection oven for 10 min at 200 ◦ C giving a film thickness of 80–100 m. 2.2. Measurements The chemical stability of the pigments was tested by exposing 0.3 g pigment to 100 ml 10% w/w HCl at 25 ◦ C for 3 h. The amount of gas evolved was measured during the exposure. A visual evaluation of the pigments before and after exposure in HCl was also done. 10% w/w HCl is prepared from concentrated (37%) HCl by adding 285.8 g acid to 714.2 g distilled water giving 1000 g solution. Chemical resistance of the aluminium-pigmented coatings was investigated using an Acidified Salt Spray (ASS) test following the ASTM G85–94 standard [10]. This test is originally designed for measuring corrosion resistance of coated aluminium. In the present work, however, it is used for monitoring the degradation of the aluminium flake pigments during exposure to an acidified salt spray atmosphere. A new cost and time effective chemical resistance test for the cured powder coating is proposed. The test is denoted the HCl dipping test and consists of dipping the samples in 10% w/w HCl for some hours. As will be discussed in the result section there is a very good correlation between the ASS test and the HCl dipping test. Chemical degradation of the samples is monitored during exposure both to ASS and HCl by measuring the L-value in standard colour measurements [11]. Samples are taken out of the cabinet/bath, and they are rinsed thoroughly in tap water followed by distilled water before drying. This rinsing procedure is arresting the degradation process. Exposure time is restarted on re-entrance of the samples in the ASS chamber or HCl bath. The L-value is used as a quantitative measure of how much visible light a surface will reflect. Degradation of a surface will normally result in a decrease in the measured L-value, i.e., the surface becomes darker. A reduction in L-value by 1 percentage point is normally not visible for our eyes. The L-values given are averages for 3 repeated readings on three predefined locations on three parallel samples. Optical absorption (OA) measurements are in this paper used as a qualitative measure for the average pigment embedment depth. OA is performed in reflection geometry at = 8.4 m. This wave-
J. M˚ ardalen et al. / Progress in Organic Coatings 63 (2008) 49–54
51
length is away from typical absorption lines of the polymer coating, but in a regime where the general absorption is found to scale well with the polymer thickness. Cross-section micrographs have also been used to visualize and measure pigment embedment depth, but we have found that these local spot tests are not sufficiently representative for the average embedment depth. A Philips CM-30 Transmission Electron Microscope (TEM) was used for cross-sectional examination of the pigments and their protective coatings. The pigments are best analyzed when embedded in a polymer matrix. The simplest procedure is, therefore, to investigate pigments in the final cured powder coating, and to strengthen the samples by further embedment in an epoxy resin. Cross-section samples for TEM analyses were made by cutting thin foils of the embedded samples using a Reichert-Jung ultramicrotome. Final sectioning was performed with a MicroStar diamond knife. 3. Results and discussion
Fig. 1. Chemical stability of pigments in hydrochloric acid. Gas evolution vs. exposure time.
3.1. Chemical stability of pigments The chemical stability of the pigments was quantified by monitoring the gas volume developed during exposure to hydrochloric acid as shown in Fig. 1. There is a large difference in the susceptibility for degradation (etching). Pigments I and II with a predicted “standard” chemical durability show considerable gas evolution. After some time the gassing curves level off indicating that all material has reacted. This was confirmed by visual inspection. The two pigments with “improved” chemical durability (pigments III and IV) perform considerably better than pigments I and II. The two “extra reinforced” (the two best) pigments show only very little gas evolution, indicating an excellent chemical stability in hydrochloric acid.
It is worth to observe that the coarser of the two “improved” pigments (pigment IV) performs well and also better than expected when judged from the information given by the supplier in Table 1. This apparent discrepancy is certainly influenced by the pigment size. Coarser pigments will have a smaller effective surface area for a given weight, and will consequently react slower. Thus, the suggested pigment test works best when comparing pigments of equal size. Note also that there is a certain delay before the gas evolution starts. This is particularly evident for pigments I, II and III. This is most probably linked to the fact that it takes some time for the acid to penetrate the protective silica layer.
Fig. 2. TEM cross-sections of flat, flake-shaped aluminium pigment particles. The protective silica coatings are visible on the aluminium surfaces. There are some variation of the thickness and the quality of the silica coating. Note that the coarse pigment (IV) is shown with another magnification than the rest.
52
J. M˚ ardalen et al. / Progress in Organic Coatings 63 (2008) 49–54
Table 2 Observations of the silica coating of the different aluminium pigments from the TEM investigations (i.e., Fig. 2) Pigment
Silica thickness (nm)
Silica morphology
Silica coverage
I II III IV V VI
10–30 10 20 50 5 10
“Fluffy” Slightly “fluffy” Probably some cracks Thick with some cracks (?) Thin and compact Thin and compact
Partially covering Partially covering Fully covering Fully covering Fully covering Fully covering
In order to find the reason for the large difference in the chemical stability TEM cross-section images were made (Fig. 2). The TEM investigations revealed information on the thickness and morphology of the silica coatings. The main results are summarized in Table 2. The TEM images show that the large difference in the chemical stability of the aluminium pigments is mainly determined by the quality of the silica protective coating. Comparing the chemical stability results from Fig. 1 with the silica coating quality (Fig. 2 and Table 2) there are some evident correlations. The silica coating on pigments I and II seems to adhere poorly to the aluminium pigment resulting in only partial coverage and a sort of “fluffy” or porous appearance in the TEM cross-section images. In some places the silica even seems to be “dissolved” or lifted out from the pigment surface and into the polymer-coating matrix. Other pigments (not shown here) are even fluffier and have even poorer chemical stability than I and II. Improved chemical stability is achieved for more compact and better covering silica coatings (III–VI). Even though pigment IV performs better than III we should not conclude that thicker silica coatings generally perform better. The pigment size is also coming into account as described earlier. In fact it seems that the two thinnest coatings (V and VI) are the best ones, showing that the quality of the silica and the degree of silica coverage is more important than the thickness of the silica coating layer. The permeability for hydrochloric acid in silica is obviously very low given a good morphology. Note also that the silica coating is very thin, 5–10 nm, for the most stable pigments (V and VI). We believe that this thin coating is the reason for pigments V and VI show some reaction with HCl already in the beginning of exposure, and do not show a distinct start-up-time as pigments I–III do (see Fig. 1). 3.2. Chemical stability of the powder coating In Fig. 3 the ASS and the HCl dipping tests are compared. The L-values are monitored for the different samples as a function of exposure times. The different coatings perform similarly in both tests and the qualitative ranking between the coatings is equal. The two coatings containing the “standard” pigments (I and II) degrade during exposure, typically after 200 h in ASS and after 8 h during HCl dipping (Fig. 3). The coatings containing pigments III and IV show considerably less degradation, and the coatings with the best pigments (V and VI) are somewhat better than the intermediate with the same pigment size (III). The powder coating containing the coarse pigment with a thicker silica coating (IV) also shows excellent chemical stability, comparable to pigments V and VI. This agrees with the results from the pigment test in Fig. 1. Coarser pigments will naturally be less reactive, and it also seems that a thicker silica layer of the “improved” quality is beneficial. However, comparing with the “extra reinforced” thin silica coatings (V and VI) the conclusion has to be that the quality of this protective coating is far more important than its thickness.
Fig. 3. (a) ASS test and (b) HCl dip test of coated samples made by Procedure 2, shallow pigment location.
Another effect of the larger pigment size is a lower initial L-value, 69 for the pigment IV coating vs. 76–82 for the other pigments. The excellent correlation between the ASS and the HCl tests seems to be independent of the pigment size. Fig. 3 clearly demonstrates the time and cost benefit of using HCl dipping as a technique for accelerated testing. The testing times can be reduced to about 1/30 compared to ASS. This enables a considerable speed-up of development of new coatings since the chemical stability testing often is the time limiting factor. In addition, the HCl dipping can be done without expensive and rather immobile laboratory facilities. It can even be simplified further either by just using visual inspection and comparison with references, and by applying droplets of chemicals on a coated panels rather than dipping (c.f. [8]). The authors have found a similar correlation when comparing standard Mortar test (24 h) [Ref. [4], Ch. 9.18, pp. 35] with dipping in NaOH for about 3 h. The Mortar test can thus also be speeded up considerably. The results are not shown here. When comparing the chemical stability of the pigment (Fig. 1) with the chemical stability of the cured powder coatings (Fig. 3), one find that the coating of the pigment is a decisive property for making chemically stable coatings. For the coatings reported in Fig. 3 it is probably also the most decisive parameter. However, a more chemically stable coating can be made if the pigments are embedded deeper in the polymer. A thicker polymer layer will then protect the aluminium flake pigments. This effect is visualized in Fig. 4 where the two pigment mixing procedures are compared. Procedure 1 gives a pigment embedment deeper into the polymer matrix (c.f. Fig. 5) and consequently a better chemical stability than Procedure 2.
J. M˚ ardalen et al. / Progress in Organic Coatings 63 (2008) 49–54
53
ing testing procedures beyond the described HCl-test. However, the small discrepancy between ASS and HCl is tolerable for our purpose and is not jeopardizing the usefulness of the HCl dipping test. The HCl curves in Figs. 3(b) and 4(b) show more “oscillations” in the L-values than the ASS counterparts. This looks a bit strange especially since the error bars indicate quite accurate readings. This is probably linked to these samples showing some L-variation across the sample (unevenness) which again is linked to the fact that the HCl dipping is a much faster process. There is obviously a trade-off between the required testing time and the testing quality, but used for semi-quantitative testing of the acidic resistance of a coating the HCl dipping test is by far good enough. The influence of the silica coating and the pigment embedment depth on the chemical stability of the powder-coatings has been demonstrated above. Another factor decisive for the chemical stability is of course the chemical composition and the morphology of the polymer coating itself. In the present work a state of the art polyester-based powder coating has been employed, and further optimization of the polymer is beyond the scope of this publication. 4. Conclusions
Fig. 4. (a) ASS test and (b) HCl dip test. Comparison of the two procedures for pigment mixing. The difference in chemical stability is linked to the pigment embedment depth and is controlled by the pigment mixing process, Procedure 1 (deep embedment) and Procedure 2 (shallow embedment).
Comparing the ASS and the HCl dipping tests for pigments with different embedment depths, Fig. 4(a) and (b), one observes that the ASS test distinguishes more clearly between the embedment depths than HCl-dipping does. This shows that the ASS test is slightly more dependent on the chemical composition and morphology of the polymer coating matrix than the HCl-dipping test is. This is understandable also from the general penetration time in the two cases, and means that one should be careful by accelerat-
Dipping of aluminium metallic pigmented coatings in 10% (w/w) hydrochloric acid is a cost and time effective alternative to testing chemical stability by acidified salt spray (ASS) test. A standardized 1000 h ASS test can be replaced by a much simpler HCl dipping test lasting about 32 h, allowing development work to be considerably speeded up. The chemical stability of the coating is determined by a combination of the chemical protection of the pigments by their coating (in the present case silica), the embedment depth of pigment in the polymer coating matrix, and the chemical composition and morphology of the polymer coating matrix. The chemical stability of the pigment themselves can be tested by monitoring the gas evolution and the colour change when exposing pigment particles to hydrochloric acid. Combined with TEM cross-section images, revealing information on the thickness and quality of the pigment coating, provides valuable information on the pigment quality and is decisive for the choice of pigment. The quality of the aluminium pigment coating is the single most important factor determining chemical resistance of the coating systems we have been investigating. Best protection is achieved with dense (not porous or “fluffy”) and fully covering silica coating on the pigments. The extra reinforced pigments are performing considerably better than other pigments at relatively shallow pigment embedment. Even a silica coating of 5–10 nm is sufficient for protection in acid environments. The influence of the pigment coating is smaller when the pigments are located deeper in the powder coating. Optical absorption measurements have been used to qualitatively measure the pigment embedment depth. A general R&D strategy of testing each constituent of the rather complex coating system separately has as far as possible been employed in the present work. This strategy gives more unambiguous and reliable results, it is more robust, and it reduces the need for many repeated tests for improved statistics. Acknowledgements
Fig. 5. Optical absorption measurements of the standard (I and II) and the extra reinforced (V and VI) pigments. Procedure 1 mixing of the powder and the pigment shows systematically higher absorbance meaning that the pigments are embedded deeper down in the coating. This correlates with the better chemical resistance of Procedure 1 mixed coatings.
This work has been partially financially by the Norwegian Research Council. The authors are indebted to Bjørn Steinar Tanem, John Walmsley and Ton Helvoort (SINTEF/NTNU) for their excellent
54
J. M˚ ardalen et al. / Progress in Organic Coatings 63 (2008) 49–54
sample preparation and their comprehensive TEM cross-section ¨ (DuPont) for skilful powder appliinvestigations, to Maria Nystrom cation and chemical resistance testing and to Lars Johnsen (SINTEF) for L-value measurements. References [1] Z.W. Wicks Jr., F.N. Jones, S.P. Pappas, Organic Coatings, Science and Technology, 2nd ed., Wiley Interscience, New York, 1999. [2] M. Cowley, Polymers Paint Colour Journal (March 1 2005). [3] T.A. Mishev, R. Van der Linde, Progress in Organic Coatings 34 (1998). [4] GSB International: International Quality Regulations For the coating of aluminium building components, Edition May 2007, www.gsb-international.de.
[5] Qualicoat: Specifications for a Quality Label for Paint, Laquer and Powder Coatings on Aluminium for Architectural Applications, 11th ed. Approved 17.11.2005, www.qualicoat.net. [6] CEPE European Council of Paint, Printing Inks and artists’ Colours Industry: Guidance note on the use of metallic effect powder coatings, http://www.cepe.org. [7] Hui Liu, Hongqi Ye, Yingchao Zhang, Colloids and Surfaces A: Physicochemical and Engineering Aspects 315 (1–3) (2008) 1–6. [8] Ulrich-Andreas Hirth: Aluminium Pigments for Powder coating Applications ¨ Paper V.3 from the 8th Nurnberg Congress 2005, and U. Hirth, Focus Powder Coatings 6 (2005) 2–3. [9] R. Supplit, U. Schubert, Corrosion Science 49 (2007) 3325–3332. [10] ASTM G 85 – 94. Standard Practice for Modified Salt Spray (Fog) Testing. [11] Reference [4] Ch. 9.19 pp. 40 and ISO 7724 and DIN 5033.