Scratch resistance of automobile clearcoats: chemistry and characterization on the micro- and nanoscale.

The scratch resistance of polymeric clearcoats intended for automobile exterior surfaces was tested with scanning 3D nanoindentation and laboratory car wash simulation (AMTEC). It could be shown by using "real world" samples with comparable scratch resistance that the applied methods were suitable for discriminating even small differences in scratch resistance with high accuracy. In the area of physical characterization of scratch resistance, the results of this investigation demonstrate the ability of microscopic, single-contact testing methods to reproduce the macroscopically experienced results.

By using a combination of different scratch methods, insight into the fundamentals, e.g., chemistry, that are responsible for good or bad scratch resistance, is achieved. One key for this goal is the imaging and analytical evaluation of the damage pattern after the indent, which leads to additional results and models presented in this article.

It was found that an essential key to obtaining scratch resistant surfaces is a strengthening of rubber-like elasticity as well as minimizing interactions between individual polymeric chains.

Keywords: Clearcoat, hardness, scratch resistance, mar resistance, scratch testing, nanoindentation, AFM, laboratory car wash (AMTEC), elastic plastic recovery, pile-up, sink-in, surface analysis, mechanical properties, automotive--OEM

**********

The term "scratch or mar resistance"* is used in a wide range of applications like optical elements (eyeglasses, mirrors, or lenses), magnetic data storage (hard-disc or tape), and a variety of high-gloss surface finishes. In all cases, a thin surface layer should ensure protection against mechanical damage. While mainly inorganic hard-coatings are used to achieve the goal in the first two applications mentioned, this article will focus on the new possibilities when using flexible polymeric or hybrid organic-inorganic-composite coatings. To ensure a tight link to one major class of material systems we will concentrate on polymeric clearcoat systems which are currently used by major automobile manufacturers worldwide.

Scratch resistance of automotive clearcoats is of great interest to car manufacturers since improvements can lead to a better value perception of the car. For the purpose of optimizing clearcoat formulations, scratch resistance has to be specified because various test methods exist. Depending on the forces and velocities applied as well as depending on the fact that some tests are performed under water and others at dry conditions, various results are obtained. In this article the focus will only be on the resistance of coating systems against scratches caused by car washing plants. Testing this property accurately in field tests means tremendous temporal and financial effort. Therefore, the well known laboratory car wash test equipment AMTEC (1) was employed on a lab scale. No further tests, such as Crockmeter, (5) Rota-Hub, or other test equipment used to evaluate scratch resistance were considered by the authors in this article.

The aim of this work is to show correlations between physical material properties of various OEM clearcoats with car wash scratch performance and to discuss the results in view of current models described in the literature.

LITERATURE

From a practical point of view, scratch resistance is easily defined as the ability of the outermost surface to withstand mechanical damage during daily use and retain a flawless optical appearance. Unfortunately the time period for the term "daily use" is counted in months or years. This is the main driving force for an urgent quest of scientific and industrial groups to develop a reliable and fast laboratory test to simulate scratching and measure scratch resistance of polymeric coatings.

To overcome this problem, numerous empiric tests have been proposed which can be subsumed as "multiple-asperity contact multiple-scratch methods." In short, some sort of abrasive body (sandpaper, brushes, different sorts of abrasives--even laser copy paper) is moved with specified normal force several times across the sample surface. Classification of scratch resistance is then done by visual inspection, specular gloss (reflection), change in color (grayscale for assessing change in color (2)), or abrasive material loss detection. Examples of this category are test methods with the Taber Abraser, (3,4) the Crockmeter, (5,6) the laboratory car wash system AMTEC-Kistler, (7,8) the test equipment Rota-Hub, and many more standardized or not standardized test methods. All these tests have at least one drawback in common--they typically cause more severe damage to the surface than typical real life cycles when generating optical visible and distinguishable damage in short time periods. This can lead to nonpraxis-relevant results. Additionally, the scratch parameters (e.g., force, velocity) are averaged over ill-defined scratch centers. The same holds for the scratch detection. As nearly

every automotive company uses different tests or test parameters, which sense different but unknown aspects of scratch resistance, no common benchmark or scale for scratch resistance is generally accepted. For chemical companies this means investing in a lot of different test methods. Finally, none of these tests lead to an understanding of the basics of scratch resistance or establish any link to the material properties of the used polymer.

Single Contact Testing

To get a more quantitative view of the processes involved during scratch testing, methods using one single contact, i.e., a sharp tip, have been used. Typically these tests use position (depth) and force sensing detectors to produce a defined damage to the sample surface. Two types of single contact measurement systems can be distinguished: depth sensing normal to the sample surface (indentation) (9-11) and lateral scratch set-ups. (12-19) More recently, with the advent of scanning force microscopy, new possibilities to control force and movement on a small (sub-[micro]m) scale became available. As a result, the new generation of scratch test equipment combines both lateral and normal control with the ability to image the resulting damage with resolutions far better than optical microscopy could provide. (20,21)

Not only the geometry of the scratch set-up but also the geometry of the single scratch tip, the normal loads applied, and the velocities used vary over some orders of magnitude. Tip sizes start in the mm-range where typically bent or grinded steel tips are applied and go down to a few nanometers. For fine tips, usually diamond or diamond-coated compound material is used with spherical or 3- or 4-sided pyramid geometries. Consequently, the normal loads cover forces starting from Newton down to [micro]N. Since all these tests are affected by the rheological behavior of the specimen surface, the speed for indentation or scratch testing will significantly influence the obtained results. Typical velocities start from very slow penetrations at about 0.1 [micro]m/sec (mainly limited by thermal drift) and end at high speed scratching at 100 mm/sec.

It is obvious that a standardized test method has not yet been established. In addition, the results from different measurements or groups cannot currently be compared on one single scale. A good overview on different methods for generating scratches and evaluating scratch resistance can be found in references 22 and 23.

Mechanisms

To achieve a rather complete understanding of the mechanisms leading to scratch resistance, decisive physical parameters have to be identified. A comprehensive review on recent literature is given by Shen et al. in several publications. (24-26) Besides others, a so-called two-mechanism three-response model is discussed. Briefly, this means that on mechanical (scratch) impact, a surface can respond in three different ways: First, a complete elastic recovery of the surface might occur. In this case, no visible damage will remain. The other possibilities are that the surface responds by plastic deformation or fracture-type deformation. The latter responses can be translated into terms of the two scratch mechanisms leading to abrasive type scratches or plastic deformation type scratches. In this case, scratch resistant surfaces need not be too hard but rather elastic, similar to the behavior of rubber. It has been shown in literature (27) and will be shown here that both types of scratches, plastic deformation type scratches and abrasive scratches, occur during car wash. The balance, i.e., which scratch type is predominant, strongly depends on the coating material.

Further systematic investigations focused on the impact of both types of scratches on subjective visual evaluation. (28) Results show that plastic deformation type scratches lead to a worse visual impression than abrasive type scratches.

HYPOTHESIS/CHEMISTRY

To translate the results described above into terms of a rough model, the theory of rubber elasticity might be helpful. It is beyond the scope of this article to optimize clearcoat networks by modeling and simulation techniques since the most important essentials of rubber elasticity can be extracted from theory and translated into terms of chemistry. For coating applications it is important that rubber elasticity is a purely entropic-driven effect. Thus, elasticity is enhanced by enlarging the probability of various conformations. Consequently, chains between crosslinks have to be as flexible as possible and the rotational barriers between the segments have to be as small as possible. Additionally, all interactions between different chains have to be reduced to maximize the number of accessible conformations. This is also true for all sorts of entanglements between polymer chains. These principles must be adapted without reducing the network crosslinking density in order to optimize the clearcoat material with regard to scratch resistance.

Before turning this hypothesis into chemistry we have to look at the different types of chemical systems employed for clearcoats. Most common resin types for 1K clearcoats are melamines in combination with acrylics and/or blocked isocyanates. In some cases carbamates are also combined with melamines. Depending on the melamine type, it is possible to build networks having high crosslinking densities, thus showing good scratch resistance. In addition, rather low rotational barriers are obtained because of the formation of either linkages. This contributes to rubber-like properties. However, pure melamines behave poorly both in long-term weathering tests and in resistance tests against liquids. (29) Therefore, reactive co-resins with blocked isocyanates and/or acrylics are often used to chemically improve melamines. Both lead to improved weathering and resistance against liquids of the clearcoats. However, when blocked isocyanates are employed, the resulting urethane or urea networks exhibit very strong intermolecular interactions and are, therefore, not the best choice for highly elastic networks. Moreover, since typical isocyanate-based crosslinkers have rather high molecular weights, only medium crosslinking densities can be obtained. Another aspect is the fact that isocyanates undergo a lot of side reactions, e.g., reactions with humidity from the air or with residual water in the basecoats. These uncontrolled reactions often limit the amount of melamines which can be accepted without negatively affecting other properties, in particular, resistance against liquids. Therefore, optimization of 1K clearcoats with regard to scratch resistance is rather difficult.

A common chemistry for 2K clearcoats is a combination of hydroxyl functional acrylics with polyisocyanates. As in 1K systems, the formation of urethane or urea networks built by isocyanate water reactions is not desirable. Optimizing the scratch performance by choosing highly flexible isocyanate derivates, such as hexamethylenediisocyanate-based polyisocyanates, is also limited by the rather high molecular weight of the available crosslinkers. Additionally, high amounts of highly flexible polyisocyanates decrease resistance against liquids.

There are several alternative approaches to avoid this dilemma. A really challenging but also very promising technology is to cure the clearcoats by using UV-irradiation. This technique uses low molecular weight acrylic functional resins as raw materials to create networks having high crosslinking densities. Strong intermolecular interactions can be avoided by choosing the appropriate chemistry of the raw materials. Extreme crosslinking densities can be achieved by employing highly branched high functional acrylic resins.

One drawback of this technique is the fact that automobile bodies are complex three-dimensional objects where shadow zones cannot be avoided. Thus, only combinations of UV- and thermal-cured coatings seem to be the technically feasible technique until now. (30) Nevertheless, the performance of 100% UV-cured clearcoats is also included in this article to ensure a complete overview.

A different technology, as well as a different crosslinking chemistry, is revealed in powder clearcoats. The most common approach is based on epoxy-carboxy chemistry and exhibits worse scratch resistance. Because resins for powders have to fulfill various additional requirements, in particular very defined melting behavior, low viscosity in the melt, and many more, there is limited room left for the optimization of scratch resistance. Due to the required rather high molecular weight of the crosslinker raw materials, high crosslinking densities cannot be achieved.

Further approaches to optimize the scratch resistance of thermal-curable clearcoats are focused on the incorporation of nanoparticles into conventional binders, (31) such as acrylic resins. Since this is a rather new research field and only a few products based on this technique are available on the market, we only include one research product in this article. This clearcoat is particularly optimized with respect to pronounced elastic properties by means of nanoparticles functionalized by grafts on their surface.

Other recent approaches, e.g., the addition of low molecular weight polysiloxanes to 2K clearcoat formulations, (32) are beyond the scope of this investigation. Instead, an accurate, systematic investigation of the correlation between AMTEC scratch resistance and corresponding physical parameters is presented. To the best of our knowledge, until now, only results of small sample series were published covering particular chemistries or very few samples. A series of eight different clearcoats was investigated here, in order to cover as many different clearcoat chemistries as possible. To be as close to practice as possible, most of the clearcoats are derivatives from actual products which are presently employed in car manufacturing.

EXPERIMENTAL SETUP

All samples were applied on test panels and cured within one day. This procedure allows for the analysis of surface and clearcoat aging with time. Laboratory car wash AMTEC testing was performed with new brushes for every panel to be as accurate as possible and avoid any effects due to wear out of the brush endings. To evaluate the influence of aging, scratch tests were performed three days after curing and repeated after approximately one month of aging in a dark environment at room temperature. Table 1 shows the chosen samples with respect to type, curing conditions, and chemistry.

[FIGURE 1 OMITTED]

A nanomechanical test was performed using a Hysitron TriboIndenter[R] system mounted on a Digital Instruments Nanoscope Dimension 3100 AFM. The tip used for indents and scratches was a three-sided pyramidal diamond with a 142.3[degrees] opening angle (Berkovich geometry). A schematic drawing as well as the real dimensions of the tip are shown in Figure 1. Special emphasis was put on choosing a sharp tip together with sharp edges to generate high stresses when damaging the surface during indentation.

With this tip, indents were performed on the sample surface with load, hold, and unload cycles of 10 sec each (see Figure 2). Directly after the indent the surface area was imaged by scanning force microscopy at least five times at intervals of 180 sec. This approach allows for a thorough analysis of the damage pattern and gives insight into the recovery dynamics of the clearcoat.

The classification of the damage pattern was adopted from F.N. Jones et al. (33) In our case we did not use scratches across the surface but instead used normal indents because they can be reproduced better due to defined geometry. Additionally, more information can be gathered and semi-empirical models for the description of this type of test are available. (34,35) Two major improvements compared to the experiments of Jones et al. have been made: Due to the strictly normal set-up of the indentation apparatus, better control of depth and applied forces was achieved compared to a cantilever mounted AFM-tip, which is inclined with respect to the sample surface. Second, a complete automated 3D analysis of the damage pattern was performed. The categories were adopted from Shen et al. (24) and were extended to 3D volumes (see Figure 3):

elastic: ([V.sub.ind] - [V.sub.dit]) / [V.sub.ind] * 100%

plastic: ([V.sub.shs]) / [V.sub.ind] * 100%

fracture: ([V.sub.dit] - [V.sub.shs]) / [V.sub.ind] * 100%

[FIGURE 2 OMITTED]

RESULTS

Damage Mechanisms

In order to test the suitability of the used experimental set-up for the investigation of automotive clearcoats, the response of the polymeric material to surface mechanical damage was tested. Therefore, normal forces and indentation times/speeds were varied within two orders of magnitude. While different speeds do not significantly influence the results, the well known (15,36) transition of material response, elastic[right arrow]plastic[right arrow]fracture, was found (Figure 4). We also want to point out the high amount of elastic response (see Figures 4 and 6) which indicated the high quality and scratch resistance of the chosen samples. The nonlinearity of the elastic response with respect to the applied force was due to the breakdown of perfect rubber-like elasticity. This behavior showed that crosslinks and entanglements of polymer-chains are irreversibly broken during the experiment.

It was important to choose the measurement parameters in such a way that all response types were tested simultaneously. This is a prerequisite for distinguishing between the mechanically very similar clearcoat films used in this article. The diversity of results from different clearcoats is plotted in Figure 5 versus the applied normal force. It is obvious that each clearcoat system responded most differently in the region of plastic and fracture damage. As a consequence, all further experiments were performed at a maximum normal load of 5 mN.

[FIGURE 3 OMITTED]

Correlation with Macroscopic Multiscratch Tests

One of the main challenges addressed here was the attempt to establish a direct link between the microscopic single contact measurements and the optical impression of clearcoat systems after macroscopic multiscratch tests. Presently we could only find a similar approach of correlating surface damage versus optical rating with a suitable amount (more than three samples) of "real" coating systems in the work done by Hara et al. (28)

For the clearcoats referenced in Table 1, the correlation of nanoscale measurements with macroscopic tests is shown in Figure 6. Obviously, clearcoats exhibiting high levels of remaining gloss are characterized by a mainly elastic response pattern. They show nearly no plastic or fracture damage. Linear correlation coefficients of [R.sup.2] > 0.7 indicate a good agreement. In parallel to the findings of reference 28 that plastic-type damage contributes more to optical gloss evaluation than fracture-type damage does, a slightly better correlation was found in our experiments for the plastic portion with the laboratory car wash AMTEC evaluation than for the fracture-type damage patterns: [R.sub.plastic.sup.2]=0.76; [R.sub.fracture.sup.2]=0.71.

[FIGURE 4 OMITTED]

Effect of Curing Chemistry

Figure 7 shows the results for the individual clearcoat systems under investigation sorted from best (left) to worst (right). It shows that, apart from the UV-monocure system #8, the polymer films responded very similar to the generated damage. Variations of elastic, plastic, and fracture response were within 10%. Special attention should be paid to systems #5 and #6 where the effect of high scratch resistance can either be achieved by using partly UV-curing technology or by conventional thermal curing and adequate melamine resins. Another special case is clearcoat #4, which had good values for elastic (high) and plastic (low) response but seemed to be vulnerable to fracture-type damage. Here, high elasticity had been achieved by high crosslinking density. However, as the used polymer chains were too stiff, the resulting coating had a very high E-modulus and became brittle. System #7 was clearly a state-of-the-art high scratch resistant system with conventional thermal curing which uses inorganic nanocomposites to reinforce a nearly perfect flexible rubber-like polymer network. On the other hand, system #8 showed what is possible with controlled UV-activated crosslinking.

[FIGURE 5 OMITTED]

The differences in curing/baking mechanisms and chemistry can be seen in Figure 8. In this figure the change within the clearcoat during an aging period of 35 days at room temperature (40% relative humidity, in darkness) is shown. As expected, slow crosslinking reactions of not completely converted active sites led to an increase in network density and thus to higher elastic values. Surprisingly, it was not the plastic part, as we had expected from our hypothetical model, but the fracture part of the response which was reduced during this aging process. On the other hand, the curing type directly influenced the aging behavior. For 1K systems (#2, #3), the crosslinking was not perfectly finished after curing, which led to high changes during the aging process. This is a quite new result as typically all 1K systems should react completely upon thermal curing. We correlate this behavior with the fact that both clearcoats used blocked isocyanates as their hardener. For that reason, the thermal initiation seems not fully sufficient to activate all blocked cyano-functionalities, which in turn slowly react during aging. UV-activated systems do not show such pronounced aging phenomena as can be seen with sample #8 (100% UV) or #6 (dual-cure), which showed reduced aging compared to #5 and #7.

[FIGURE 6 OMITTED]

[FIGURE 7 OMITTED]

[FIGURE 8 OMITTED]

The 2K clearcoats under investigation were considered to postcure dramatically, since it is well known that during the thermal curing processes a complete conversion of the isocyanates does not normally occur. Due to high reactivity of the isocyanates, the major part of the postcuring process might have been completed before we performed the first experiments (three days after panel preparation). We conclude that the remaining postcuring process had only limited impact on the mechanical properties of the clearcoats.

DISCUSSION AND SUMMARY

Scratch resistance, especially of polymers, is both a materials property and field of science which is very complex and not easily accessible from either the chemical side or from the physical point of view.

In terms of chemistry, we followed the hypothesis that fracture and abrasion of the polymer cannot be completely avoided due to their low elastic modulus, which is one or two orders of magnitude lower than for ceramic surfaces. Therefore, we successfully concentrated on the avoidance of plastic (pile-up) deformation. Highly elastic coatings can be achieved by increasing the polymeric network density, e.g., by using melamine crosslinks in 2K recipes (#5) or by UV-curing (#8).

However, there are additional challenges one has to face by using those approaches: the network can get brittle and/or the chemical resistance (acids, alkalines, etc.) can decrease due to increased vulnerability to swelling. It is therefore important to find a compromise in clearcoat optimization between excellent scratch performance and other properties. Indeed, sample #8 showed cracks even immediately after application and more pronounced damage during weathering tests. Sample #5 failed the chemical resistance tests.

Keeping these problems in mind, sample #7 exhibited no serious weaknesses. In this case the use of specifically modified nanoparticles allowed the network strength to be increased without decreasing other relevant properties. Further investigations on long-term stability, such as weathering and resistance against liquids, are not completely finished. Yet until now, no serious drawbacks have been found. Thus, we consider this sample as one promising candidate to be validated on car exteriors in the near future.

In the area of physical scratch characterization the results of this investigation show the ability of microscopic, single contact testing methods to reproduce the macroscopically experienced results. A similar comparison with clearcoats cleaned multiple times in a real car wash facility is being performed and will hopefully confirm those results. However, it is not only the reproducibility and direct correlation to the laboratory test which make the nanoscale experiments valuable. It could be shown that in addition to quantifying the scratch resistance the test set-up used here is more suited to give information about the fundamentals (e.g., chemistry) that are responsible for good or bad performance. One key for this goal is the imaging and analytical evaluation of the damage pattern after the indent, which leads to additional results and models.

References

(1) ISO/FDIS 20566:200, "Paints and Varnishes--Determination of Scratch Resistance of a Coating System Using a Laboratory Car Wash" (also German standard DIN 55665).

(2) ISO 105-A02:1993, "Textiles--Test for Colour Fastness--Part A02: Grey Scale for Assessing Change in Colour."

(3) Gregorovich, B. and Hazan, I., Prog. Org. Coat., 24, 131 (1994).

(4) ISO 7784-2:1997, "Paints and Varnishes--Determination of Resistance to Abrasion--Part 2: Rotating Abrasive Rubber Wheel Method"; ASTM D 4060:2001, "Standard Test Method for Abrasion Resistance of Organic Coatings by the Taber Abraser."

(5) ISO 105-X12:2001, "Textiles--Test for Colour Fastness--Part X12: Colour Fastness to Rubbing," Note: corrected and reprinted in 2002.

(6) Trumbo, D.L., Rudelich, J.C., and Mote., B.E., In Perspectives on New Crops and New Uses, Janick, J. (Ed.), ASHS Press, Alexandria, VA, pp. 267-271, 1999.

(7) http://www.amteckistler.de.

(8) Meier-Westhues, U., Klimmasch, Th., and Tillack, J., Eur. Coat. J., 9, 258 (2002).

(9) VanLandingham, M.R., Villarrubia, J.S., Guthrie, W.F., and Meyers, G.F., Macromol. Symp., 167, 15 (2001).

(10) VanLandingham, M.R., Villarrubia, J.S., and Meyers, G.F., ACS Polymer Preprints, 41, 1412 (2000).

(11) Heischkel, Y., Schwalm, R., Kutschera, M., and Menzel, K., presented at RadTech conference, 2003.

(12) Alers, P. and Consiglio, R., Prog. Org. Coat., 40, 287 (2000).

(13) Barbeau, P., Magny, B., Roche, S., and Loubet, J.-L., Eur. Coat. J., 10, 406 (2002).

(14) Blackman, G.S., Lin, L., and Matheson, R.R., ACS Symposium Series, 1218 (1998).

(15) Briscoe, B.J., Pelillo, E., Ragazzi, F., and Sinha, S.K., Polymer, 39, 2161 (1998).

(16) Gauthier, C. and Schirrer, R., J. Mater. Sci., 35, 2121 (2000).

(17) Jardret, V., Lucas, B.N., Oliver, W., and Ramamurthy, A.C., "Scratch Durability of Automotive Clear Coatings: A Quantitative, Reliable and Robust Methodology," J. COAT. TECHNOL., 72, No. 907, 79 (2000).

(18) Lin, L., Blackman, G.S., Matheson, R.R., Prog. Org. Coat., 40, 85 (2000).

(19) Lin, L., Blackman, G.S., Matheson, R.R., Mater. Sci. Eng., A317, 163 (2001).

(20) Schwalm, R., Beck, E., and Pfau, A., Eur. Coat. J., 1, 107 (2003).

(21) Meichsner, G., Burk, T., Zhang, H., Larbig, H., Sander, R., and Kutschera, M., Farbe und Lack, 7 (2005).

(22) Courter, J.L. and Kamenetzky, E.A., Eur. Coat. J., 7, 100 (1999).

(23) VanLandingham, M.R. and Giraud, M., "Scratch and Mar Resistance of Polymeric Materials," unpublished.

(24) Shen, W., Smith, S.M., Jones, F.N., Ji, C., Ryntz, R.A., and Everson, M.P., "Use of a Scanning Probe Microscope to Measure Marring Mechanisms and Microhardness of Crosslinked Coatings," J. COAT. TECHNOL., 69, No. 873, 123 (1997).

(25) Shen, W., Ji, C., Jones, F.N., Everson, M.P., and Ryntz, R.A., Polym. Mater. Sci. Eng., 74, 346-347 (1996).

(26) Shen, W., Ji, C., Jones, F.N., Everson, M.P., and Ryntz, R.A., Surf. Coat. Int., 6, 253 (1996).

(27) Betz, P. and Bartelt, A., Prog. Org. Coat., 22, 27-37 (1992).

(28) Hara, Y., Mori, T., and Fujitani, T., Prog. Org. Coat., 40, 39-47 (2000).

(29) ISO/DIS 2812-5:2004, "Paints and Varnishes--Determination of Resistance to Liquids--Part 5: Method Using a Temperature Gradient in an Oven."

(30) Muhle, J., Fey, Th., and Wulf, M., Farbe und Lack, 109, 18 (2003).

(31) German Patent Application DE 10247359, 29.04.04.

(32) WO 01/09261, 2001.

(33) Jones, F.N., Shen, W., Smith, S.M., Huang, Z., and Ryntz, R.A., Prog. Org. Coat., 34, 119 (1998).

(34) Oliver, W.C. and Pharr, G.M., J. Mater. Res., 7, 1565 (1992).

(35) Pharr, G.M., Mater. Sci. Eng., A, 253, 151 (1998).

(36) Briscoe, B.J., Pelillo, E., Sinha, S.K., and Evans, P.D., Wear, 200, 137 (1996).

M. Kutschera** and R. Sander -- BASF AG*

P. Herrmann, U. Weckenmann, and A. Poppe -- BASF Coatings AG ([dagger])

* Clarification: Throughout this article we use the term "scratch" resistance for the ability of a polymeric surface to withstand multiple cycles in a car washing plant without visible gloss loss. (In some publications this is also referred to as "mar" resistance).

* Polymer Research Laboratory, D-67056 Ludwigshafen, Germany.

([dagger]) Strategic Research, D-48165 Munster, Germany.

** Author to whom correspondence should be addressed. E-mail: [e-mail]. Paper submitted for publication February 2005.

Table 1 -- Clearcoat Systems Used

Sample
Number  Clearcoat             Curing/Baking Condition

#1      powder clearcoat      145[degrees]C/30 min
#2      1K solventborne       135[degrees]C/20 min
#3      waterborne clearcoat  155[degrees]C/23 min
#4      2K solventborne       140[degrees]C/22 min
#5      2K variant            140[degrees] /22 min
#6      2K + UV dual cure     1500 mJ/[cm.sup.2] + 140[degrees]C/14 min
#7      2K evolution          140[degrees]/22 min
#8      UV monocure           1500 mJ/[cm.sup.2] [N.sub.2] inert
                                atmosphere

Sample
Number  Chemistry

#1      epoxy-carboxy
#2      OH-functional acrylic/melamine/blocked isocyanate
#3      OH-functional acrylic--blocked isocyanate
#4      OH-functional acrylic--polyisocyanate
#5      OH-functional acrylic--polyisocyanate melamine
#6      OH-functional resin--polyisocyanate, acrylic functional resin
#7      nanocomposite clearcoat based on acrylic resins
#8      acrylic--functional resins