Plasma treatment for surface modification of Wood flour and its use in a composite with PP matrix.
Hamílton Víana1,Renata Antoun Simão2, Derval dos Santos Rosa1, Rafael Cordeiro2
- Rua Santa Adélia, 166. CEP 09.210-170 - Santo André - SP - Brazil
1 – Federal University of ABC (UFABC) – CECS – Laboratory of Biodegradable Materials
2 – Federal University of Rio de Janeiro (UFRJ), Laboratory of Surfaces ans Thin Films, RJ, Brazil
Abstract
The growing importance of composites using natural fibers as reinforcements drives the need for methods to increase the fiber-matrix adhesion, among other challenges, to make these materials high performance and expand its market use. A custom reactor was developed to treat short wood fibers under plasma, while guaranteeing that they are homogeneously treated. Two groups of samples were treated and prepared as composite tensile specimens that were tested and the results were analyzed statistically using factorial design and ANOVA tools. It has been found that SF6 is more effective then CH4 or Air for this purpose, and while exists a minimum time-power condition, increases of elastic modulus of over 25% were found when comparing composites with untreated and treated fibers for 40 minutes at 30 Watts. SEM micrographs of fracture surface show indications of stronger interactions between the two phases, confirming the tensile testing results.
Keywords: Plasma, surface modification, wood, polypropylene.
- Introduction
Composites of thermoplastics reinforced with natural fibers have gained attention of researchers and industries in the past decade1. The reasons for this increase of interest are many among which some can be cited: Increased environmental awareness in both the use of renewable resources with lower carbon footprint and the waste on the end of the material’s life and biodegradability; reducing costs with reinforcement materials that are cheaper than the polymeric matrix; increase in mechanical properties associated with reduction in density of the material, resulting in considerable reduction on the weight of components1, 2, 3, 4, 5, 6. Natural fiber composites, among which Wood Polymer Composites (WPC) excel in all these areas, with the added advantage of not only using traditional fabrication methods for thermoplastics, but also having the potential of reduced equipment wear3, 7, 8.
However, despite all the benefits this material class possesses, many problems must be solved before they can achieve their best properties and reach new markets, truly becoming an industrial reality. The main problems are the poor interaction and bonding between the fibers and matrixes, and the effect of humidity on the properties of the material1, 5, 6, 9. Tough many works attempt to mitigate these problems, much research is still needed to fully understand the mechanisms involved and optimize composites with natural and wood fibers to broad use1.
Many techniques are used in attempt to improve fiber-matrix adhesion, like chemical treatment of the fibers, usually Alkaline or Silane treatments1, 10, 11, and incorporation of functionalizing groups in the matrix, where Maleic Acid is the most widely used4, 5. Another very promising technique is low pressure plasma treatment of the fiber surfaces, that are able to change only the surface layer with little to no process waste, while diminishing the needs for additives on the matrix, that can lower certain properties of the composites, thus overcoming the disadvantages of the two other techniques cited 1, 2, 12, 13.
The conditions of plasma treatments found in literature vary greatly due to different techniques, plasma reactor assembly and geometries, gasses used, fibers treated, matrix with which these will be combined and so forth. The expected result, however, is always an increase in adhesion between the fibers and the matrix, aiming at better mechanical properties of the resulting composite. Many works reported achieving this goal either for tensile strength, elastic modulus or both, as well as for flexural and compression strengths at times1, 2, 12, 13.
- Materials and Methods
For the present work a reactor designed and built in-house was used, allowing for approximately 5 grams of wood fibers or flour to be treated in each batch, thanks to a constant movement to make sure that all fibers are exposed to the plasma for at least part of the treatment time. RF inductive plasma was used, with pressure range of 0,5 – 1 Pa and forward power from 30W to 60W. The aim was to treat wood fibers or flour with lengths between 200µm and 1000 µm and width varying from 50 µm to 300 µm. These would then be introduced in a Polypropylene matrix with no additives, via melt compounding in a mixer model MH-50H from M.H. EQUIPAMENTOS for 15 seconds on low rotation and more 15 seconds on high rotation speed. Right after the melted mass in then compression molded for 15 seconds at 60°C in a press of the same manufacturer, where it solidifies resulting in discs of approximately 2mm thickness, with 17wt% content of wood fibers. The discs were then cut with a cutting tool and hydraulic press into tensile tests specimens with smaller working area, of 15 x 5,3 x 2 mm, that are then tested with an INSTROM tensile testing machine.
Two groups of samples were prepared, the first group consisted of fibers treated of the same time and power condition (30 minutes, 50 Watts) with different gases, to verify which is the most effective to increase matrix-fiber adhesion. Atmospheric air, CH4 and SF6 were used. On the second group, samples were treated only with SF6 for a specific set of conditions based on a 2-level factorial design, using 2 factors or variables. The variables tested were power and treatment time, with low levels being 10 minutes and 30W, and the high levels were 40 minutes and 60W. There was also a treatment with the same conditions as the previous group, to confirm the result found. All data of treated and untreated wood+PP composites were statistically analyzed for the null hypothesis using the one-way ANOVA test with 95% confidence interval (α=0,05) on the software Minitab 16.1.
- Results and discussion
Tensile testing conducted on the first group of samples revealed that the addition of 17wt% of untreated wood fiber in PP matrix increased the elastic modulus from 408MPa (Standard Deviation: 18,6MPa) to 537MPa (S.D.: 35,5MPa), an increase of 32%. One-way ANOVA analysis of the elastic modulus of the composites shows that CH4 and atmospheric air plasma have no measurable effect on this property, as it can be seen by the overlapping confidence intervals on image 1. SF6 plasma, on the other hand, displays a considerable increase in elastic modulus compared to untreated wood fiber, way beyond any statistical doubt, as it can also be seen on image 1. The new value (663MPa) is 23% higher than untreated wood fiber and 63% higher than pure PP.
Figure1 - ANOVA analysis of composite tensile specimens tested from group 1.
Given the increase in elastic modulus achieved with SF6 plasma, as well as to better understand the influence of the time and power on the reactor used, further analysis was made using a 2-level factorial design with 2 factors (power and time). Image 2 shows the graphical representation of this design.
Figure 2 – 2-level factorial design for the second group of samples, made to analyze the effect of time and power on the resulting properties.
The analysis of variance table for elastic modulus is shown on Table 1. It can be seen that both variables are relevant to the surface treatment process, as well as the interaction between both. It can also be concluded that there is a strong curvature associated with the results, which is a consequence of the experimental parameters chosen for this test.
The results presented on image 3 for the second groups of tensile test specimens shows clearly that the low levels chosen have statistically no effect on the results, compared to untreated wood fibers composite (the case of 10 minutes-30W will be addressed shortly). The samples treated with both variables at high level, on the other hand, show a great increase in elastic modulus, similar to the samples treated on the same condition as the first group (30 minutes-50W). What can be concluded from this is that the minimum time-power condition for effective surface plasma treatment of wood fibers with low pressure SF6 gas is close to 30min-50W. So, the curvature indicated on the ANOVA table derives from the initial slope, bellow which no effect can be observed.
Table 1 –ANOVA table shows the effects of the variables and probability of their influence on the results. P is the probability of that parameter NOT affecting the results.
Source / DF / Seq SS / Adj SS / Adj MS / F / PMain Effects / 2 / 75073 / 75072,8 / 37536 / 21,46 / 0
Time / 1 / 43292 / 43292,4 / 43292 / 24,75 / 0
Power / 1 / 31780 / 31780,4 / 31780 / 18,17 / 0
2-Way Interactions / 1 / 31225 / 31224,8 / 31225 / 17,85 / 0
Time*Power / 1 / 31225 / 31224,8 / 31225 / 17,85 / 0
Curvature / 1 / 7299 / 7299,5 / 7299 / 4,17 / 0,054
Residual Error / 20 / 34989 / 34988,8 / 1749
Pure Error / 20 / 34989 / 34988,8 / 1749
Total / 24 / 148586
Figure 3 –ANOVA analysis of composite tensile specimens tested from group 2.
On the second group, the three repeated composite tests (pure PP, Untreated fibers and SF6 treated fibers for 30min at 50W) presented statistically the same results as in the first group, confirming its results. The mean PP elastic modulus is however lower on the second group (373,7MPa, S.D.:35,5), and would be considered different if a 90% confidence interval was chosen, resulting in greater percentage increase on the properties for this group.
Although it cannot be statistically affirmed that the 40min-60W samples (E=656,2MPa) performed better than the 30min-50W (E=627,2MPa), especially due to the high standard deviation on this last case, the mean value of the first is higher with low deviation. It is believed that a higher power associated with longer treatment time will render a higher elastic modulus for the wood-PP composite. Two effects are believed to be responsible for this increase observed: Higher surface roughness on a nano-scale, increasing surface area and mechanical anchoring points, and; addition of fluorine groups on the surface of the wood fiber, as reported before, rendering it hydrophobic and thus increasing its wettability in the also hydrophobic matrix, resulting in more effective mixing with less defects and higher interaction forces between the two phases14, 15.
For the fibers treated for 40minutes at 30W, the power on the plasma was not sufficient to change the surface enough so that the fiber-matrix interaction would be increased. The energy of the ions colliding with the surface was probably not enough to introduce the fluorine groups, either by not being able to break the cellulose and lignin chains, or by not being able to produce the adequate ion based on the SF6 gas.
When the power was at the high level of 60W but the time was only 10 minutes, what is believed that made the resulting elastic modulus be statistically the same as the untreated fibers was the fact that the exposure time of each fiber to the plasma is lower than the total treatment time. Considering that the fibers are in constant motion to allow for all of them to come in contact with the plasma, and that this movement is fast enough when compared to the total treatment time, it can be assumed that all fibers are exposed for approximately the same dose of plasma along the whole process, although possibly with a statistical distribution of exposures. Since the amount loaded in the reactor is at least 1 order of magnitude more than a monolayer of fibers, so to speak, the mean exposure time is, in a crude estimate, 10 times less than the total time. This would give approximately 1 minute of plasma exposure for each fiber or less, not being enough to affect the surface in an effective way for the great majority of fibers, resulting in an elastic modulus very close to the untreated case, as it is seen on the data of image 3.
It is currently unknown the reason why fibers treated for 10 minutes at 30W presented composite elastic modulus lower than that of untreated fibers composite, but two possibilities will be investigated in further work: A problem during the melting compound or the compression molding, generating a defective disc from which the samples were extracted, or; the humidity content change on the fiber due to the vacuum of the plasma chamber influenced negatively the fiber-matrix interaction.
Scanning electron microscope images were taken from the tensile test fracture surface of composites with untreated fibers (image 4) and fibers treated for 30 minutes at 50W (image 5). On the first it can be seen that the wood fibers suffered pull-out from the matrix and there are both, fibers with no evidence of adhesion to the PP and holes with the texture of the fibers that were pulled out, also showing very little interaction between them.
Figure 4 – SEM micrographs of tensile test fracture surface of composite made with untreated fibers.
Figure 5 – SEM micrographs of tensile test fracture surface of composite made with fibers treated for 30min at 50W under SF6 plasma.
On the second case, image 5, where the fibers were treated with SF6, it can be seen that there are less apparent fibers and smaller pull-out lengths, and in many cases it can be seen that some fibers are almost fully inside the polymer, even on highly deformed areas like it is shown. All these are indications of a stronger interaction between fibers and matrix and higher interface shear strength, confirming the tensile testing results that after the plasma treatment with SF6 under adequate conditions, the elastic modulus increases2.
- Conclusions
Plasma treatment of short wood fibers with low pressure RF Plasma ofSF6on a in-house built reactor has been confirmed as an effective way to increase fiber-matrix interaction and adhesion, resulting in superior elastic modulus when compared to both, neat Polypropylene (76%) and composite with untreated fibers (25%). The modulus of 656,2MPa was achieved after treating the fibers under constant motion, to guarantee all fibers are treated, for 40 minutes at 60W power. Statistical ANOVA analysis with 95% interval of confidence show that treatment conditions with low power (30W) and/or little time (10 minutes) are ineffective to increase elastic modulus, and a strong curvature on the response data indicate that higher power and longer times should render better mechanical properties, which will be evaluated in future works. It has been suggested that for reactor loads where fibers rest over one another, thus the need for the constant movement, the mean exposure time to plasma is shorter than the total time, explaining the ineffectiveness of treatments of less than 30 minutes. Fracture micrographs of the fractured surface reveal that the fiber-matrix interaction and adhesion indeed increase, as indicated by a smaller amount and shorter length of fiber pull-out.
- Acknowledgements
The authors would like to thank CNPq, ANP, FAPESP and CAPES for the financial support.
- References
[1] - Faruk O, Bledzki AK, Fink H-P, Sain M. Biocomposites Reinforced with Natural Fibers: 2000-2010. Progress in Polymer Science.2012; 2000–2010. In press.doi:10.1016/j.progpolymsci.2012.04.003
[2] - Yuan X, Jayaraman K, Bhattacharyya D. Effects of plasma treatment in enhancing the performance of woodfibre-polypropylene composites. Composites Part A: Applied Science and Manufacturing. 2004;35:1363–1374.
[3] - Herrera-Franco PJ, Valadez-González A. A study of the mechanical properties of short natural-fiber reinforced composite. Composites Part B: Engineering. 2005;36:597–608.
[4] - Dányádi L. et al. Wood flour filled PP composites: Compatibilization and adhesion. Composites Science and Technology.2007; 67:2838–2846.
[5] - Pérez E, Famá L, Pardo SG, Abad MJ,Bernal C. Tensile and fracture behavior of PP/wood flour composites. Composites Part B: Engineering. 2012;43:2795–2800.
[6] - Sobczak L, Lang RW, Haider A. Polypropylene composites with natural fibers and wood – General mechanical property profiles. Composites Science and Technology. 2012; 72:550–557.
[7] - Mohanty AK, Misra M, Hinrichsen G. Biofibres, biodegradable polymers and biocomposites: an overview. Macromolecular Materials and Engineering. 2000;276-277
[8]- Zampaloni M, Pourboghrat F, Yankovich SA, Rodgers BN, Moore J, Drzal LT, et al. Kenaf natural fiber reinforced polypropylene composites: a discussion on manufacturing problems and solutions. Composites Part A. 2007;38:1569–80.
[9] - Stamboulis A, Baillie CA, Garkhail SK, Van Melick GH, Peijs T. Environmental durability of flax fibres and their composites based on polypropylene matrix. Applied Composite Materials. 2000;7:273–94.
[10] - Dányádi L, Móczó J, Pukánszky B. Effect of various surface modifications of wood flour on the properties of PP/wood composites. Composites part A. 2010;41:199–206.
[11] - Sreekumar PA. et al. Effect of fiber surface modification on the mechanical and water absorption characteristics of sisal/polyester composites fabricated by resin transfer molding. Composites Part A: Applied Science and Manufacturing. 2009;40:1777–1784.
[12] - Li R, Ye L, Mai Y-W. Application of plasma technologies in polymer composites : a review of recent developments. Composites part A 1997;28A:73–86.
[13] – Sinha E, Panigrahi S. Effect of Plasma Treatment on Structure, Wettability of Jute Fiber and Flexural Strength of its Composite. Journal of composite Materials. 2009; 43(17):1791-1803
[14] - Santos AEF, Bastos DC, da Fonseca MD, Simão RA. Inducing Surface Hydrophobization on Cornstarch Film by SF6 and HMDSO Plasma Treatment. Carbohydrate Polymers.2012; submitted for publication. doi:10.1016/j.carbpol.2012.08.031
[15] - Suanpoot Pet al. Surface analysis of hydrophobicity of Thai silk treated by SF6 plasma. Surface and Coatings Technology. 2008;202:5543–5549.