Low-VOC coating systems for oak furniture: cost has been the main reason for widespread use of relatively inexpensive nitrocellulose lacquers in the Dutch furniture industry. That's going to change in the near future as furniture manufacturers will be forced to use lacquers with lower amounts of solvents.
This situation is going to change considerably though in the near future. Furniture manufacturers will have to use other lacquers with a lower amount of solvent. This is due to the incorporation of the European solvent directive in Dutch law as well as additional agreements on increased environmental restrictions and enforced health regulations. The European solvent directive is only operative for companies using more than 15 tons of solvent a year, which involves only 20 out of 2,500 Dutch furniture producers. The latter two aspects will have an effect on nearly every Dutch furniture producer handling lacquers.
Most likely an agreement will be signed between the Dutch Ministry for Social Affairs and Employment and the Association of Dutch Furniture Producers (CBM) to reduce health and safety risks for employees in the furniture industry. This agreement involves different topics such as lifting, exposure to saw dust and Chronic Toxic Encephalopathy (CTE). CTE is a disturbance of the nerve system caused by long-term exposure to solvents. The Dutch ministry would like to minimize the possibility of CTE by limiting the amount of solvents in coatings rather then using closed systems or demanding personal protection. These regulations are set for the whole wood processing industry (joinery, parquetry, indoor painting and furniture). Prior to setting the limit on the amount of solvents in lacquers used in the furniture industry, several research projects were done in order to determine the possible use of low-VOC coatings by the furniture industry. The first was to set up criteria for the use of low-solvent coating systems and determine the properties of these alternative systems. The second project was an assessment of the real exposure to solvents, which will be repeated every three years to monitor the results of planned actions for reduction.
Results from the first research project showed that low-VOC coatings are suitable for certain applications. However, for applications which demand a high chemical resistance and higher layer thickness, as well as for the wood species of oak, the use of low-VOC coatings is still an obstacle.
Approximately 10-15% of Dutch furniture producers use oak for their products. The traditional coating system on this type of furniture consists of a coloring nitrocellulose or naphtha stain with two additional layers of nitrocellulose lacquer. After spraying, the coloring stain is rubbed-out with a cloth, causing dark colored vessels and a lighter colored wood surface. During this activity, the exposure to solvents is high.
The aim of a third research project, presented in this paper was to examine which low-VOC coating systems are suitable for 3D oak furniture. The research was divided into the following three parts:
A: Examination of the aesthetical and qualitative performance of coatings systems;
B: Small-scale practical experiments and VOC exposure measurements; and
C: Large-scale experiments.
During the entire research project, the aesthetical acceptance of the coating systems by the furniture producers played an important role.
Materials and Methods
Part A: Examination of the aesthetical and qualitative performance of coatings systems.
Twenty different coating systems were applied to European and American oak chair seats. The coatings were selected by the coating manufacturers and consisted of only commercially available systems. The application of the coatings was carried out at one location and the applied coating weight was measured. The following systems were used:
* Five systems with both a waterborne (WB) stain and lacquer;
* Five systems with a WB stain and a polyurethane or acid-curing lacquer;
* Three systems with a solvent-based stain and a WB lacquer;
* Four reference systems with both a nitrocellulose stain and lacquer;
* One system with a WB stain and nitrocellulose lacquer;
* One system with a high solid stain and acid curing lacquer; and
* One system with a nitrocellulose stain and polyurethane lacquer.
The stains were applied in an E-50 standard color (dark brown). Application was carried out with an air-mix spray gun.
After drying the lacquers, the seats were visually examined by 12 furniture producers, eight paint producers and three other people involved in the project. Scores from one to five were given on color, gloss, brightness, filling and roughness. For the furniture producers, the main criterion was if the product would be acceptable for sale.
The technical quality of the lacquers was assessed on the following criteria and standards:
* Resistance to chemicals according to DIN 68861 class 1B;
* Pencil hardness (Wolff Wilborn) according to ISO 3270;
* Scratch resistance according to EN 438-2 using a Taber Shear Scratch Tester;
* Adhesion according to ISO 2409;
* Cold-check resistance according to ASTM D 1211 using a Votch test chamber; and
* Resistance to discoloration using a Q-panel Q-Sun Xenon Tester (one week exposure at 0.35 W/m2/320nm and 50[degrees] C and 55% RH). The color-change (dE value) was measured with a Minolta CM 2600d spectrophotometer.
Part B: Small-scale practice test and exposure measurements.
Seven coating systems were selected for Part B, based on the results of the aesthetic assessment of Part A. Each coating system was applied to 30 U.S. oak chairs. The chairs were taken from the normal production line of a furniture producer involved in the project.
The stains were applied with an air-mix spray gun and the lacquers were applied by electrostatic spraying, both by hand. Before applying the second lacquer layer, the chairs were sanded automatically (using a glass medium) and by hand. The applied amounts of coating per chair and total use of coating were registered. In addition, parameters that could effect the final result such as temperature, relative humidity, wood moisture content and airspeed were noted. The tests were carried out in a well-equipped paint room of a furniture manufacturer. During the tests no normal production took place.
The VOC-exposure of the sprayer and the person that rubs out the stain were measured using active carbon and tenax absorption media. With a pump, a defined amount air was sucked through the absorption media. Desorption found place with 0.5% of methanol in C[O.sub.2]. Solvent concentrations were analyzed with GS-MS. From the solvents concentration the exposure index (EI) was calculated. This value is the sum of the concentration of each volatile compound divided by its MAC-level. Legal limits are exceeded when the exposure index is higher than one.
An aesthetic evaluation was done comparable to Part A.
Part C: Large-scale practice test.
Four coating systems were selected for Part C, based on the results of the aesthetic assessment and exposure of Part B. Part C was performed similar to Part B with the following alterations:
* For each coating system, 40 U.S. oak chairs were used. One coating system was used on the chairs of and produced by one specific furniture producer.
* The application of the lacquers was carried out with an electrostatic spray gun attached to a spray robot.
* Apart from VOC-exposure, formaldehyde and isocyanate exposure was measured on a fixed placed in front of the spraying booth. Personal exposure was measured at the person that shifted the chairs.
The tested low-VOC lacquers were compared with the common nitrocellulose finishes of the furniture producers. The test chairs were transported to the normal production location of the furniture producers.
Results of Part A
Part A: Examination of the aesthetic and qualitative performance of coatings systems.
The VOC emission per square meter was calculated from the applied amount of coating on each seat. Results are shown in Figure 1. The full waterborne systems had 12-19 grams VOC emission per [m.sup.2], while complete nitrocellulose systems had an emission of 105-120 grams/[m.sup.2]. Combining a waterborne stain with a solventborne lacquer results in a higher VOC emission than the combination of a solventborne stain with a waterborne lacquer, due to the fact that two layers of lacquer are applied.
[FIGURE 1 OMITTED]
At the aesthetic assessment, the complete waterborne systems scored very low. Figure 2 shows that the best score of a waterborne system is 13% of acceptance of the involved critics. Four complete waterborne systems scored zero percent of acceptance. The most acceptable systems were the complete solventborne systems (best score is 88%). Of the combined waterborne-solventborne systems, the systems with a waterborne stain had a higher score compared to systems with a waterborne lacquer.
[FIGURE 2 OMITTED]
The assessment of the general impression had a considerable influence on the evaluation of the separate parameters. When the general impression was considered to be "unacceptable," each individual parameter gained a low score.
Figure 3 shows that for all aesthetical parameters, the furniture producers were more critical in their opinion than the coating producers. Furthermore, the furniture producers were the most univocal in their judgement.
[FIGURE 3 OMITTED]
Most lacquers showed a good chemical resistance (no damage in class 1B is often requirement for normal furniture use). The waterborne lacquers were slightly discolored by ethanol (48%). Two acid curing and one waterborne lacquer showed discoloration due to the impact of tea.
The nitrocellulose lacquers were the most sensible for temperature changes in the cold-check test. After 40 cycles, the first nitrocellulose systems showed cracks. Four of the seven waterborne systems showed cracks after 50-80 cycles. In general, 60 cycles without damage is a sufficient result.
The nitrocellulose lacquers showed less resistance to scratching compared to other lacquers. Remarkably, some of the waterborne lacquers showed the same scratching resistance as the polyurethane and acid curing lacquers.
The hardness of the polyurethane lacquers was comparable to the tested waterborne and nitrocellulose lacquers. One acid curing lacquer gave a better result, the other acid curing lacquers scored less.
Systems for Part B
Based on these results, the following systems were regarded as being good enough for further research:
* One complete waterborne system;
* Three systems with a waterborne stain and a acid reactive lacquer;
* One system with a waterborne stain and a polyurethane lacquer;
* One system with a high solid stain and a acid reactive lacquer; and
* One complete nitrocellulose system (reference).
Of the above mentioned systems, the waterborne system had the lowest aesthetical acceptance (13%). The other systems combined high aesthetical acceptance with good technical performance and reduced VOC emissions.
Part B: Small-scale practice test and exposition measurements.
In Figures 4 and 5, the VOC emission per chair and exposure indexes is shown. The VOC emission per chair is calculated from the applied coating weight on 30 chairs. The shown exposure is the exposure to one coating layer. Note that the lacquer is normally sprayed in two layers, which will double the stated emission. The exposure index is an average of two measurements during at least 20 minutes spraying time.
[FIGURES 4-5 OMITTED]
The nitrocellulose lacquers contain the highest amount of solvent (750 grams per liter). However, the VOC emission of most acid reactive and the polyurethane lacquers was comparable to the nitrocellulose lacquer due to the difference in applied layer thickness. It was suggested that the sprayer needed more experience spraying lacquers with a higher solid matter content.
As expected, rubbing-out of the nitrocellulose stain resulted in a high exposure index. During spraying of the lacquers, the exposure index was high for the nitrocellulose and acid curing lacquer. It was observed that certain components with low MAC-levels in the lacquer caused high exposure indexes. The relative high exposure index of the acid curing lacquers is caused by (iso)-butanol, which is added for product stabilization and electrical conduction (for electrostatic spraying).
The aesthetic assessment of the coating systems applied to the chairs is given in Figure 6. The acid curing and polyurethane lacquers were judged as "too filling." The nitrocellulose system gained the highest degree of acceptance from the furniture producers, while two systems combining a waterborne and a solventborne layer were fully unacceptable. The best alternative systems were combinations of a waterborne stain with a polyurethane or acid curing lacquer. These systems also scored higher than average on brightness. As was seen in Part A, the coating producers gave higher rankings for visual acceptation compared to the furniture producers.
[FIGURE 6 OMITTED]
Two systems were selected for Part C, the waterborne stain with polyurethane lacquer and the waterborne stain with acid curing lacquer (3). They combine a lower exposure index with a reasonably aesthetic evaluation and a VOC emission reduction of 23-30%.
Part C: Large-scale practice test.
To make a good comparison between the two lacquers, the choice was made for using one single stain for both coating systems. Due to the fact that this stain was not deliverable anymore, a stain with similar properties was selected (based on information of the manufacturer). Furthermore two additional lacquers were added to the test.
The VOC concentration was measured on three different places: a) close to the robot; b) on the person that shifts the chairs directly after coating application; and c) on the cart where the coated chairs were placed on for drying. The formaldehyde emission (acid curing lacquers) and the presence of isocyanate aerosols (polyurethane lacquer) were measured on the same positions.
The values in Figure 7 are average values of two independent measurements. The VOC concentration close to the robot exceeds the allowed maximum value of 1.0. This measurement was probably carried out too close to the robot to be used as a realistic value. Possibly the probe was in direct contact with the coating spray during a short time of the measurement. The measurements on the person that shifts the chairs gave values, which are comparable to normal VOC exposure in a spraying cabin. The AC1 lacquer gave the lowest VOC exposure (0.36) and the AC3 lacquer the highest (0.67). The values are within legal limits.
[FIGURE 7 OMITTED]
The exposure to formaldehyde is given in Figure 8. Comparable to the VOC exposure, the measurement close to the robot gave high and less reliable values (the standard deviation is high for all three lacquers). The formaldehyde concentration at the person that shifts the chairs was less than 1.0 for all lacquers.
[FIGURE 8 OMITTED]
The measured concentration of isocyanate aerosols during the application of the polyurethane lacquer is low (exposure index is less than 0.1 on the person that shifts the chairs).
Concerned furniture producers evaluated all chairs as being too rough. Water within the waterborne stain results in a re-orientation of the grain, which causes the wood fibers to "stand up." They are fixing in that position by the first layer of lacquer. Although the systems were sanded automatically and additionally by hand, the "standing up" of the fibers is still noticeable in the final result. Furthermore, the color of the test chairs was slightly different compared to the chairs from the normal production. The color is not seen as main criteria because it is relatively easy to adjust.
The aesthetical performance was evaluated more positively by coating producers compared to furniture producers. Furniture producers often state this as a problem. The furniture producers on the other hand were more univocal in their judgement. The coating producers know what progress was made since the introduction of waterborne systems and evaluate new products as "better or (more) acceptable," while furniture producers compare new products with their nitrocellulose finished products. A client's desired quality of furniture is difficult to answer.
The waterborne stain used in Part C caused relatively strong raising of the grain. Much better results were obtained in Part B, where an other waterborne stain was used with a higher binder and solvent content (up to 100 g/1 VOC compared to 50 g/l in Part B). This stain gave an acceptable low raising of wood fibers. Extra labor time due to more sanding is unacceptable for most furniture producers. The profit and market shares are already under pressure due to the relatively high labor costs compared to competitive countries.
The adjustment of the color seems not to be a technical obstacle. The stain determines the color of the finished product, but the color can only be evaluated as part of the full coating system. It can be time consuming to reach the same color with a WB/PU system as with a nitrocellulose system. The right color, however, is very important for the furniture producers. It seems that the difficulty with waterborne stain used in Part C can be solved by using a stain with a higher binder content. The process of adjusting the color of the stain seems to be just a small step. By doing so the use of low-VOC coatings for classical oak furniture should be possible.
The National Health Council advised to reduce the maximum allowed concentration for n-butylacetate from 710 to 150 mg/[m.sup.3]. With the new value two of the three acid curing lacquers will exceed the legal exposure limits. Butyl acetate is added to the lacquers to tune conductivity, which is necessary for electrostatic spraying. Also for formaldehyde the National Health Council advised a new MAC-level. The advised value is a factor 10 lower that the present value (reduction from 1.5 to 0.15 mg/[m.sup.3]). With this advised MAC-level, the formaldehyde concentration for the second acid curing lacquer would exceed the maximum value.
This research did not directly lead to the selection of a low-VOC coating system that will be applied on oak furniture. The surface roughness of the systems, selected in two assessment rounds and demonstrated in the last part of the project was unacceptable for the furniture producers.
Replacing nitrocellulose (or naphtha) stains with waterborne stains will lead to minimization of VOC exposure during the coating application. Additionally a reduction of the VOC emission to the environment is obtained.
It is possible to lower VOC exposure by replacing nitrocellulose lacquers with polyurethane and acid reactive lacquers. The aesthetic performance of waterborne lacquers was unacceptable. The brightness and filling of the pores differed too much from the reference with a nitrocellulose finish. The chemical resistance (alcohol) of the waterborne lacquers was also less.
Polyurethane or acid reactive lacquers show better aesthetical performance compared to waterborne lacquers. However, the exposure to VOC of polyurethane and acid reactive lacquers differs per product and is not always better at prospect. The products should be selected on VOC concentration as well as (theoretical) exposure risk. An exposure to VOC is highly influenced by some specific solvents with a low MAC-level, which could be difficult to replace by other solvents.
How to Continue?
A project is proposed to arrange the fine-tuning of the process with a coating and furniture producer. Adjusting the color of the stain, balancing the binder and VOC content and surface roughness will be part of this research.
During the time, that the agreement on safety and health improvement will be active, furniture producers will be visited and advised about possible measures to minimize the risk of CTE. Possibly, at the end of the agreement-period, producers will be confronted with new regulations on VOC exposure risk. These regulations will focus on replacement of solvents at the source. One of the proposed regulations is to set a maximum for the allowed amount of solvents in coatings. For a stain, the maximum amount of solvent could be fixed at approximately 100 or 150 grams/liter, forcing the industry to use waterborne stains and preventing high solvent exposure during rubbing-out the stains. For a lacquer, the maximum amount of solvent might be fixed at 400-600 grams/liter. For 3D applications, this will result in a replacement of nitrocellulose lacquers by medium or high solid (polyurethane or acid reactive) or waterborne lacquers. A disadvantage of this proposed regulation is that a correlation between the amount of solvent and the VOC exposure risk is not always clear. Classifying all lacquers on their exposure risk (i.e. theoretical risk calculated from the recipe) would be better, but it would involve a considerable amount of work and would be less practical.
Barend van de Velde of SHR Timber Research can be reached at P.O. Box 497, 6700 Al Wageningen; +31 (0)317 425422 Fax + 31 (0)317 425783; Email: firstname.lastname@example.org. Erwin Beckers can be reached at E-mail: email@example.com.
This paper was initially presented at PRA's Third Wood Coatings Congress, held Oct. 28-30, 2002 in The Hague, The Netherlands. Conference proceedings are available at www.pra.org.uk.
|Printer friendly Cite/link Email Feedback|
|Title Annotation:||coatings industry|
|Author:||van de Velde, Barend; Beckers, Erwin|
|Date:||Jun 1, 2003|
|Previous Article:||Media and milling equipment directory: here is a look at the latest in media and milling technology from select industry suppliers. For more...|
|Next Article:||Show review: ECS 2003: despite the U.S.'s war with Iraq and a sluggish economy, the European coatings show still drew record attendees to Nurnberg.|