Bio-Medical Materials and Engineering 24 (2014) 1595–1607 DOI 10.3233/BME-140964 IOS Press
1595
Effect of heat treatment of wood on the morphology, surface roughness and penetration of simulated and human blood J. Rekola a,b,c,∗ , L.V.J. Lassila a,b , S. Nganga a,b , A. Ylä-Soininmäki a,b , G.J.P. Fleming d , R. Grenman c , A.J. Aho a,b,e and P.K. Vallittu a,b a
Department of Biomaterials Science, University of Turku, Turku, Finland Biocity Turku Biomaterials Research Program, Turku Clinical Biomaterial Centre, Turku, Finland c Department of Otorhinolaryngology – Head and Neck Surgery, Turku University Hospital and University of Turku, Turku, Finland d Materials Science Unit, Division of Oral Biosciences, Dublin Dental University Hospital, Trinity College Dublin, Ireland e Department of Orthopaedics and Traumatology, Turku University Hospital, Turku, Finland b
Received 1 December 2011 Accepted 1 July 2013 Abstract. BACKGROUND: Wood has been used as a model material for the development of novel fiber-reinforced composite bone substitute biomaterials. In previous studies heat treatment of wood was perceived to significantly increase the osteoconductivity of implanted wood material. AIM: The objective of this study was to examine some of the changing attributes of wood materials that may contribute to improved biological responses gained with heat treatment. METHODS: Untreated and 140◦ C and 200◦ C heat-treated downy birch (Betula pubescens Ehrh.) were used as the wood materials. Surface roughness and the effect of pre-measurement grinding were measured with contact and non-contact profilometry. Liquid interaction was assessed with a dipping test using two manufactured liquids (simulated blood) as well as human blood. SEM was used to visualize possible heat treatment-induced changes in the hierarchical structure of wood. RESULTS: The surface roughness was observed to significantly decrease with heat treatment. Grinding methods had more influence on the surface contour and roughness than heat treatment. The penetration of the human blood in the 200◦ C heattreated exceeded that in the untreated and 140◦ C heat-treated materials. SEM showed no significant change due to heat treatment in the dry-state morphology of the wood. DISCUSSION: The results of the liquid penetration test support previous findings in literature concerning the effects of heat treatment on the biological response to implanted wood. Heat-treatment has only a marginal effect on the surface contour of wood. The highly specialized liquid conveyance system of wood may serve as a biomimetic model for the further development of tailored fiber-composite materials. Keywords: Natural fiber-composite, heat treatment, roughness, liquid interaction, biomimetic
*
Address for correspondence: J. Rekola, Department of Otorhinolaryngology and Head and Neck Surgery, Turku University Hospital, Kiinamyllynkatu 4-8, 20520 Turku, Finland. Tel.: +358 2 3130000; E-mail: [email protected]. 0959-2989/14/$27.50 © 2014 – IOS Press and the authors. All rights reserved
1596
J. Rekola et al. / Heat treatment effects on wood morphology, surface roughness and liquid penetration
1. Introduction The surface roughness and contour of a biomaterial are some of the attributes that can affect the interfacial integration of an implanted material and host bone. Implant surface modification has been reported to have a significant influence on the bone response [1–10]. The implant surface roughness increases mechanical interlocking, the active surface area and changes the wettability of the surface [1,3,7,11]. It also provides a more suitable basis for the positive biological bone response leading to increased osteoconductivity [4,5]. A biomaterial’s interaction with body fluids determines many of the properties that influence the osteoconductivity of the implant material, for example cell adhesion and spreading. In this aspect, the hydrophilicity and wettability of a biomaterial have been considered positively [12,13]. Conversely, hydrophilic surfaces also are more prone to promote bacterial colonization [14]. Liquid conveyance features are beneficial for bulk bone substitutes and penetration of the liquids enables biological activity, for example bone forming and immunological regulation inside the implanted material. On the surface level the absorption and drainage of liquid into the material can, to some extent, be simplified to a capillary action, where the possible empty spaces between fibers act as capillary tubes. The liquid influx becomes more complex at greater depths into the porous material where interconnectivity between the pores form complicated liquid conveyance systems. Fiber-reinforced composite (FRC) materials have been under development for non-metallic loadbearing implants [3–5,15–22]. The future fields of application could also include orthopedic and head and neck surgery as well as oral and maxillofacial surgery and dentistry, in which some FRC technologies have been embraced in clinical usage [23–25]. Recently wood, a natural fiber composite, has been introduced as a model material for developing synthetic FRCs [26,27]. The study design for using wood as a model material has included the use of heat treatment for managing the properties of the wood. Heat treatment is a method introduced by the wood industry to improve the biological and physical endurance of wood against the elements. In vivo studies have shown that heat treatment increases significantly the osteoconductivity of the implanted wood material [27,28]. Wood has a hierarchical structure, with highly evolved liquid conveyance features, which yields a greater biomechanical strength than the solid from which it is composed [29]. In vitro studies have shown, that the biomechanics of wood can be modified considerably using heat treatment [30]. Due to the aforementioned observations it has been concluded that heat-treated wood may be used as model material for biomaterial research in several ways; as a biomimetic model, as a bone substitute candidate, as a model material for mechanical testing and as a possible scaffold for tissue engineering. In fact, the unique anatomy of wood (see Fig. 1) has already been used in biomaterial scaffold applications and a novel approach for bulk bone substitutes [31–34]. The aim of this study was to evaluate changes that take place in a wood material during heat treatment, to better understand the biological response reported previously in the literature [27,28]. To facilitate a sensible comparison, the exact same materials (untreated, 140◦ C and 200◦ C heat-treated birch wood) as in previous studies were used. The effect of heat treatment on the surface topography of wood was measured and reported. The effect of heat treatment on liquid interaction, wetting and wicking (or liquid penetration), was measured using two manufactured liquids (simulated blood) as well as human blood. Contact angle measurements were not performed due to adequate prior knowledge available in the literature about the effects of heat treatment concerning hydrophilicity. To further understand the surface topography as well as the structures facilitating the liquid conveyance system, the wood morphology
J. Rekola et al. / Heat treatment effects on wood morphology, surface roughness and liquid penetration
(a)
1597
(b)
(c) Fig. 1. Typical structure of birch wood. The vessel (A) is the main water conducting unit in birch. The length of a single vessel cell is 0.2–0.6 mm and diameter ca. 100 µm. Vessels are separated from each other with a bar-like (scalariform) structure called perforation plate. Thus, they form long tubes. Vessels are perpendicularly connected with ray parenchyma cells (B), through small holes called pits (+). Rays function as storage cells and enable tangential water transport. They are less than 0.1 mm in length and less than 30 µm in diameter. The most common cell type in birch xylem is a fiber (C), which can be separated into tracheids and libriform fibers. They provide the main support to the structure and are 0.4–1.6 mm in length and 10–40 µm in diameter. The location of the larger magnification picture is indicated with a white box.
was imaged with Scanning Electron Microscopy (SEM) and the influence of heat treatment on liquid transport into inner structures of wood was theorized.
2. Materials and methods The European downy birch (Betula pubescens Ehrh.) wood used in the study was cut from the surface wood of larger blocks (approx 30 × 10 × 5 cm) of untreated, 140◦ C heat-treated and 200◦ C heat-treated (for a time period of two hours).
1598
J. Rekola et al. / Heat treatment effects on wood morphology, surface roughness and liquid penetration
2.1. SEM analysis The wood was cut in half longitudinally and carbon sputter coated (SCD 050; Bal-Tec, Balzers, Liechtenstein). High vacuum SEM (JSM-5500; Jeol Ltd., Tokyo, Japan) images were taken to evaluate the anatomy and microstructure of the surface and the sub-surface of the wood samples and the changes induced by heat treatment. One observer made the examinations visually. 2.2. Surface profilometry Twenty pieces of wood (of approximately 5 × 5 × 7 mm) in size were cut in half for each treatment group, untreated wood and wood heat-treated at 140◦ C and 200◦ C. The cut surfaces were then ground using a disc grinder LaboPol-21 (Struers, Westlake, OH, USA, 300 rpm) along the orientation of the fibers with silicon carbide grinding paper (grit 180) with equal pressure to remove any protruding fibers that could hinder the profilometric analysis. For each condition investigated (untreated or heat-treated at 140◦ C and 200◦ C), 10 specimens were measured (longitudinally) along the length of the fibers and a further 10 specimens were measured perpendicular to the fiber orientation. The ends of the specimens were measured from five different pieces of wood from each wood group (untreated and heat-treated at 140◦ C and 200◦ C). All of the pieces of wood were measured five times, and ground with grinding paper between every measurement. The surface roughness was measured employing contact and non-contact methodological approaches. The contact surface roughness testing was accomplished by Surftest 301 device (Mitutoyo Corporation, Japan) with an 0.8 mm sample length longitudinally to the fibers as well as over the ends of the fibers and 0.4 mm sample length perpendicular to the fibers. Every measurement run consisted of five sequential runs, thus the total number of measuring events was 50 (longitudinal and tangential) and 125 for the end surface of the specimen. To enable a methodological comparison, a non-contact surface roughness test was also used: one piece of wood from each of the heat treatment groups, after being ground manually with rotating motion with the silicon carbide grinding paper, was tested with a non-contact optical profilometer (Talysurf CLI 2000, Taylor–Hobson Precision, Leicester, England). A chromatic length aberration gauge of 3 mm was employed. The resolution was 40 nm in the vertical (z-direction) and measurements were noted every 10 µm (x-directions) and every 2 µm (y-directions). The specimens were scanned at a speed of 1 mm/s over an area of 3 × 5 mm2 . Mean surface roughness (Ra value) and standard deviation were calculated by the TalyMap analysis software package (Taylor–Hobson Precision, Leicester, England). 2.3. Liquid penetration test To demonstrate the penetration of liquids into each wood group (untreated and heat-treated at 140◦ C and 200◦ C) mainly by capillary forces, a dipping test was performed. For each liquid solution, six cylindrical specimens (4.1 mm diameter and 25 mm length) from each wood group were cut with a dental burr. At the end of the specimens, the cut surface resulting from sawing was ground first with 180 grit (FEPA) and finally with 1200 grit grinding paper using a disc grinder. The samples were fixed and dipped in a vertical alignment position for 1 min at ambient temperature in the test liquids, with the exception of human blood which was performed at 37◦ C. Pictures were taken from the dipped samples and liquid penetration depth was measured with Image J analyzer (scale 1568 pixel/10 mm in 25% zoom view). Three different liquids were used. A starch solution was prepared by mixing 150 ml deionised water with 3 g of starch, then rapidly heated to 100◦ C and boiled for three minutes. After slowly cooling the liquid, methylene blue was added for colouring. A glycerol–ethanol–water solution (GEW) was prepared by mixing 103.5 g of glycerol, 33.9 g of water and 21.9 g of ethanol with methylene blue added
J. Rekola et al. / Heat treatment effects on wood morphology, surface roughness and liquid penetration
1599
for colouring. The human blood was bought from the Finnish Red Cross (allowance 53/2010, valid until 44/2011), and consisted of the blood of 17 A+ donors. The blood was supplied in 10 ml BD vacutainers with K2 EDTA added as an anti-coagulant. The average hemoglobin was 154 g/l. The densities of the solutions were measured with a digital density-meter APDMA45 (Anton Paar High-precision instruments, Ashland, USA). The viscosities were measured with automated micro viscometer Anton Paar AMVn (Anton Paar High-precision instruments, Ashland, USA) and analysed using VisioLab for AMVn. 2.4. Statistical methods Shapiro–Wilk testing was conducted to explore the normality of the data distribution. Where the data was normally distributed, one-way ANOVA was used for between group testing, followed by Bonferroni corrected post hoc multiple comparison. In non-normally distributed results, the non-parametric Kruskal–Wallis test was performed to analyze the statistical significance of the differences between groups. The Mann–Whitney U-test was then applied to evaluate the differences between the individual groups. Statistical analysis was not performed for the non-contact profilometry, because the method was chosen for illustrative purposes and therefore there was only one sample per treatment group. 3. Results 3.1. SEM In the SEM images, the heat treatment-induced changes were quite inconspicuous. At smaller magnifications, the surface morphology and the inner anatomy of the heat-treated wood did not show differences compared with the untreated wood. However, at larger magnifications, it was observed that the cell walls in the wood heat-treated at 200◦ C were more discontinuous with microscopic cracks, and in places appeared to form lamellae (Fig. 2). This was the only observation of differences by SEM between the untreated and heat-treated (at 140◦ C and 200◦ C) groups.
Fig. 2. In places the surface of the cell walls of the 200◦ C heat-treated wood was broken (showed microcracks) and formed thin lamellae (arrow). Although sporadic cracking of cell walls was observed in every treatment group due to the shrinking induced by drying, this kind of lamellar breaking was detected only it the 200◦ C heat-treated wood. The lamellar breaking was seen scarcely and only in places where the sample cut surface was in the right angle (perpendicular) to the fibers.
1600
J. Rekola et al. / Heat treatment effects on wood morphology, surface roughness and liquid penetration
3.2. Surface profilometry In contact profilometry the arithmetical mean deviation (Ra value) for the longitudinal and perpendicular measurements was highest in the untreated group, Ra value of 3.53 (1.36) µm. The Ra values decreased with heat treatment. The mean Ra values and associated standard deviations reduced to 2.57 (0.89) µm and 2.02 (0.58) µm following heat treatment at 140 and 200◦ C, respectively. A Kruskal–Wallis test revealed statistically significant difference in Ra values across the three groups (χ2 = 22.89, p < 0.001). Additional Mann–Whitney U-tests with correction for multiple comparisons disclosed all individual between-groups differences to be statistically significant. For the end surfaces of the specimens the Ra values were 2.20 (0.81) µm for the untreated wood groups, 1.57 (0.62) µm for the 140◦ C heat-treated and 1.25 (0.38) µm for the 200◦ C heat-treated group. A Kruskal–Wallis test showed a statistically significant difference in Ra values across the three groups (χ2 = 23.50, p < 0.001). The Mann–Whitney U-test indicated all between-groups differences to be statistically significant. For the non-contact profilometry, the mean roughness was highest in the specimen which was heattreated at 200◦ C, with a Ra value of 24.33 (3.28) µm. An intermediary Ra value (13.93 (1.58) µm) was obtained on heat treatment at 140◦ C and the untreated wood material had a Ra value of 4.24 (0.59) µm. The graphical illustration of the surface topography is depicted in Fig. 3. 3.3. Liquid penetration test The results of the liquid penetration (dipping) tests are depicted in Fig. 4. The densities and viscosities of the test liquids are reported in Table 1. Being normally distributed, the results for the penetration of human blood were analyzed with a one-way ANOVA, which revealed statistically significant differences between groups (F = 27.850, p < 0.001). A Bonferroni corrected post hoc comparison revealed the 200◦ C heat-treated group differed from the untreated and 140◦ C heat-treated groups with statistical significance, while there was no statistically significant difference between the untreated and 140◦ C heat-treated groups. For GEW solution dipping results, non-parametric testing was used. Kruskal–Wallis test revealed statistically significant differences between groups (χ2 = 7.94, p = 0.019). Individual group comparisons revealed the 200◦ C heat-treated group differed from the untreated and 140◦ C heat-treated groups with statistical significance (p = 0.010 and p = 0.025 for untreated and 140◦ C heat-treated groups, respectively), while there was no statistically significant difference between the untreated and 140◦ C heat-treated groups. Starch solution dipping results were analyzed with a one-way ANOVA, which revealed statistically significant differences between all groups (F = 11.511, p = 0.001). A Bonferroni corrected post hoc analysis revealed statistically relevant differences between the untreated and 140◦ C group (p = 0.001). The differences between 200◦ C group and both untreated and 140◦ C group did not reach statistical significance. In a gross picture analysis, the penetrations of all of the liquids were maximal near the surface of the samples. In the simulated blood liquids, methylene blue coloration was separated from the solutions, not penetrating as fast as the rest of the liquid into the wood structures. The level of methylene blue was nevertheless measured as the definite penetration depth, because the clear solution penetration depth was considered too difficult to determine. In fact, in the case of the GEW, the clear portion of the liquid reached the end of the samples in all of the birch wood (untreated and heat-treated at 140◦ C and 200◦ C) groups.
J. Rekola et al. / Heat treatment effects on wood morphology, surface roughness and liquid penetration
1601
(a)
(b)
(c)
(d) Fig. 3. A graph showing the results of the contact profilometry as well as illustrations of the surface analysis of non-contact profilometry. The sand paper grinding shows a significant effect on the surface with the non-contact method. The surface of the 200◦ C heat-treated wood is distinctly more uneven, with numerous spikes representing protruding fibers. The normal fiber orientation seen on the untreated and 140◦ C wood has become indistinguishable on the surface of the 200◦ C heat-treated material. (Colors are visible in the online version of the article; http://dx.doi.org/10.3233/BME-140964.)
1602
J. Rekola et al. / Heat treatment effects on wood morphology, surface roughness and liquid penetration
(a)
(b)
(c) Fig. 4. (a) The results of the liquid penetration test. The actual dipping depth is not included in the results. (b) Samples after dipping. 200◦ C heat-treated (A), 140◦ C heat-treated (B) and untreated (C) dipped in (from left to right) starch solution, GEW solution and human blood. (c) Samples cross sections after dipping: 200◦ C heat-treated, 140◦ C heat-treated and untreated dipped in (from left to right) starch solution, GEW solution and human blood. In the gross analysis of cut surface only the hydrophobic GEW solution had significant visible penetration to the inner structures. This is largely due to the short dipping time of 1 min. (Colors are visible in the online version of the article; http://dx.doi.org/10.3233/BME-140964.)
J. Rekola et al. / Heat treatment effects on wood morphology, surface roughness and liquid penetration
1603
Table 1 The density and viscosity of the tested liquids Liquid Starch solution GEW Human blood
Density (g/cm3 ) 1.0005 1.1247 1.0598
Dynamic viscosity (mPa · s) 5.892 ± 0.4638 21.301 ± 1.152 4.9151 ± 0.0696
4. Discussion 4.1. Wood morphology The lamellar breaking of the wood cell walls observed in this study confirm the findings in the literature describing the effects of heat treatment [35]. On drying the wood, microcracks appear in the cell walls due to excessive shrinkage. The cell wall consists of lamellae formed by microfibrils. During heat treatment these layers can become loosened and break, forming the observed protruded lamellae [36]. This phenomenon can yield more relative surface as well as form more protruding fibers at the free surface. The nature of the high vacuum SEM that was used in this study meant that the wood material had to be analyzed as dry as possible. It appeared that the gross anatomy of the dry wood material does not change significantly during the heat treatment. In a wet state, however, the behaviour of the wood is considerably different due to heat treatment protocols [30]. The ability of birch wood to absorb water into the cell walls decreases with heat treatment, leading to reduced swelling [30,37]. This leads to an altered wet state morphology between heat-treated groups; at the higher heat treatment (200◦ C) group liquid remains proportionally more in the hollow channels i.e. cell lumens and not within the cell walls. Although this situation can be said to resemble more the liquid distribution found within a living bone, the biological relevance of this phenomenon cannot be concluded without additional studies and further evidence. The diameter of the open canals (or pores) is a critical factor in determining the quality of the tissue growth into the free space. The literature suggests that the optimal pore size of a biomaterial for mineralized bone in-growth is 100–400 µm, although pore sizes of 50–125 µm have been found to be sufficient for bone tissue in-growth [38]. The inter-connected canalicular structure of wood (Fig. 1) allows the conduction of liquids especially along the cell axis. However, there are number of structures in the anatomy of wood that also allow tangential movement of liquid molecules, i.e. tangential rays and pores. This liquid conducting system (called superapoplasm) has been sophisticated by evolution in nature and there probably is considerable information to be gleaned for biomaterial research. 4.2. Surface profilometry The general rule of thumb that the rougher the surface, the better the implant material, is not true per se. Although a larger surface area means an increase surface contact surface, the specific structural architecture, i.e. the contour of the surface, has a significant effect on the distribution of the shear forces [39]. The results of the contact profilometry of this study are in line with the previously published literature, where similar tests have been conducted with other wood species treated with different heat treatment methodologies [40–43]. The mechanism of the heat treatment-induced decrease in roughness is not clear.
1604
J. Rekola et al. / Heat treatment effects on wood morphology, surface roughness and liquid penetration
One suggestion is that high temperatures alter lignin to a thermoplastic condition which in turn densifies the cell surfaces [41]. The partial disintegration of wood cell walls may also have an influence on surface roughness. Grinding has a significant influence on the surface, when measuring on a micrometer scale, which is why the number of repeated measurements has to be high. The protruding fibers most likely give way to the contact stylus, which can partially explain the large difference in results between contact and non-contact methodological approaches. The effect of grinding is clearly illustrated with the non-contact profilometry. Different grinding pre-treatment, namely rotational manual grinding with silicon carbide abrasive paper not only increased the roughness, but also changed the gross profile and texture of the surface. The non-contact test illustrates how easy it is to modify the surface of the wood, which can be desirable in bone substitute biomaterials. As far as the literature is concerned, the contact stylus method seems to be the gold standard in wood surface profilometry. For biological applications, however, the three dimensional non-contact method can possibly yield significantly more information, especially in fiberous materials as it also illustrates the contour of the measured surface, whereas the two dimensional straight-line measurement of a contact stylus, for instance used in this study, fails to do so. Overall, albeit being statistically significant, the change in surface roughness between the untreated and heat-treated (140 and 200◦ C) wood is however so inconclusive that it probably does not bear much if any relevance biologically. 4.3. Liquid penetration test As the gross canalicular structure of the wood does not change with heat treatment, the wicking effect depends on the viscosity of the solutions and on the hydrophobic/hydrophilic nature of the wood and liquids tested. The literature suggests that the degradation and cross-linking of cellulose and hemicellulose reduce the hydrophilicity. Wood is mostly hydrophobic when heat-treated at 190◦ C. At elevated temperatures, the formation of hydrophilic degradation products seems to again transform the wood to be more hydrophilic in nature [44]. The most hydrophilic liquid employed in this study (i.e. the starch solution) interacted best with the untreated wood, whereas the GEW solution and human blood, being relatively more hydrophobic in nature, had better interaction with the 200◦ C heat-treated wood. The separation of methylene blue coloration from the solution indicates that larger molecules and particles travel slightly slower into the wood structure than small water molecules. This finding may be related to the pore size of the material. In previous in vivo studies, some bone formation has been found inside implanted 200◦ C heat-treated wood implants, indicating that the pore size is sufficient for the penetration of undifferentiated mesenchymal cells [27]. However, the presence of bigger (>100 µm) canals would probably be beneficial in this aspect. As mentioned previously, the ability for dimensional changes diminishes with heat treatment and when submerged into simulated body fluid for a sufficient amount of time with the untreated wood absorbing the greatest volume of water [30]. The 1-min dipping time used in this work excludes the effect of cell wall swelling, leaving the wood’s canalicular conveyance ability and the liquid–material interaction as the affecting attributes. For the testing of optimal biological response, the result of the human blood dipping is the most interesting. The penetration of the blood in the 200◦ C heat-treated wood exceeded that in the untreated and 140◦ C heat-treated wood. This result is in conjunction with previous in vivo and in vitro tests, where the 200◦ C heat-treated wood was the most osteoconductive as well as the most prone to hydroxyl apatite formation, when compared with untreated and 140◦ C heat-treated wood [28, 30]. The heat treatment seems to have a positive effect on the interaction between a bulk of wood and blood – most likely due to the heat treatment – induced chemical changes in the wood polymer. The
J. Rekola et al. / Heat treatment effects on wood morphology, surface roughness and liquid penetration
1605
microstructural alterations as well as the changes in surface roughness are likely to have only a limited influence on this phenomenon.
5. Conclusions The following conclusions can be drawn, bearing in mind the limitations of the study: (1) Birch wood samples show typical hardwood structure with vessels and fibers. The dry-statemorphology of these structures does not change significantly during heat treatment. The specified liquid conducting system of wood could yield useful information for the development of synthetic fiber biomaterials. (2) Heat treatment reduces slightly the surface roughness of wood when measured with contact profilometry. The surface of wood is very prone to the effects of grinding which underlines the importance of cautious preprocessing when characterizing the surface topography of a fiberous biomaterial. The increase in the osteoconductivity of wood associated with heat treatment is more likely due to the chemical changes in the wood material than to the change in surface roughness. (3) Heat treatment most likely changes the liquid distribution in the deeper structures of wood samples by reducing the absorption of liquids into the cell walls, while leaving the channel structure unchanged. (4) Heat treatment increases the penetration of liquids with increased hydrophobicity, including human blood, into the structure of wood.
Acknowledgements We would like to thank Mikko Salomäki from Analytical Chemistry Institute/Arcanum, Turku, Finland for analysing the viscosities and densities of the liquids used in this study and Adam Dowling from the Materials Science Unit, Division of Oral Biosciences, Dublin Dental University Hospital, Trinity College Dublin, Ireland, for performing the non-contact profilometry analyses. Docent Pekka Saranpää from the Finnish Forest Research Institute, METLA, Vantaa, is acknowledged for his contribution in the writing of the text.
References [1] S. Hansson and M. Norton, The relation between surface roughness and interfacial shear strength for bone-anchored implants. A mathematical model, J. Biomech. 32(8) (1999), 829–836. [2] L. Rasmusson, J. Roos and H. Bystedt, A 10-year follow-up study of titanium dioxide-blasted implants, Clin. Implant Dent. Relat. Res. 7(1) (2005), 36–42. [3] R.H. Mattila, M.A. Puska, L.V.J. Lassila and P.K. Vallittu, Fibre-reinforced composite implant: in vitro mechanical interlocking with bone model material and residual monomer analysis, Journal of Materials Science 41(13) (2006), 4321– 4326. [4] R.H. Mattila, P. Laurila, J. Rekola, J. Gunn, L.V.J. Lassila, T. Mäntylä et al., Bone attachment to glass-fibre-reinforced composite implant with porous surface, Acta Biomater. 5(5) (2009), 1639–1646. [5] A. Aho, M. Hautamäki, R.H. Mattila, P. Alander, N. Strandberg, J. Rekola et al., Surface porous fibre-reinforced composite bulk bone substitute, Cell Tissue Bank 5(4) (2004), 213–221. [6] L. Cooper, A role for surface topography in creating and maintaining bone at titanium endosseous implants, J. Prosthet. Dent. 84(5) (2000), 522–534.
1606
J. Rekola et al. / Heat treatment effects on wood morphology, surface roughness and liquid penetration
[7] S. Hansson, Surface roughness parameters as predictors of anchorage strength in bone: a critical analysis, J. Biomech. 33(10) (2000), 1297–1303. [8] T. Albrektsson and A. Wennerberg, Oral implant surfaces: Part 1 – review focusing on topographic and chemical properties of different surfaces and in vivo responses to them, Int. J. Prosthodont. 17(5) (2004), 536–543. [9] T. Albrektsson and A. Wennerberg, Oral implant surfaces: Part 2 – review focusing on clinical knowledge of different surfaces, Int. J. Prosthodont. 17(5) (2004), 544–564. [10] C. Ivanoff, G. Widmark, C. Johansson and A. Wennerberg, Histologic evaluation of bone response to oxidized and turned titanium micro-implants in human jawbone, Int. J. Oral Maxillofac. Implants 18(3) (2003), 341–348. [11] P. de Gennes, F. Brochard-Wyart and D. Quéré, Capillarity and Wetting Phenomena: Drops, Bubbles, Pearls and Waves, Springer-Verlag, 2004. [12] R.J. LaPorte, Hydrophilic Polymer Coatings for Medical Devices – Structure/Properties, Development, Manufacture and Applications, Technomic Publishing Company, Lancaster, USA, 1997. [13] T.G. Ruardy, H.E. Moorlag, J.M. Schakenraad, H.C. van der Mei and H.J. Busscher, Growth of fibroblasts and endothelial cells on wettability gradient surfaces, Journal of Colloid and Interface Science 188 (1997), 209–217. [14] W.-Y. Su, S.-C. Wu, L. Wu, X.-X. Wang, P. Yue and X.-Z. Fu, Surface ant-bacterial adhesion characteristics of TiO2 coated PMMA, Chinese Journal of Structural Chemistry 28 (2009), 1497–1502. [15] S. Tuusa, M. Peltola, T. Tirri, L.V.J. Lassila and P.K. Vallittu, Comparison of two fiber-reinforced composite structures as implant material in calvarial bone defect, Bioceramics, Key Engineering Materials 361–363 (2007), 471–474. [16] M. Hautamäki, A.J. Aho, P. Alander, J. Rekola, J. Gunn, N. Strandberg et al., Repair of bone segment defects with surface porous fiber-reinforced polymethyl methacrylate (PMMA) composite prosthesis: histomorphometric incorporation model and characterization by SEM, Acta Orthop. 79(4) (2008), 555–564. [17] S. Tuusa, M. Peltola, T. Tirri, L.V.J. Lassila and P.K. Vallittu, Frontal bone defect repair with experimental glass-fiberreinforced composite with bioactive glass granule coating, Journal of Biomedical Materials Research Part B – Applied Biomaterials 82B(1) (2007), 149–155. [18] A. Ballo, L.V.J. Lassila, T. Närhi and P.K. Vallittu, In vitro mechanical testing of glass fiber-reinforced composite used as dental implants, J. Contemp. Dent. Pract. 9(2) (2008), 41–48. [19] D. Zhao, N. Moritz, P. Laurila, R.H. Mattila, L.V.J. Lassila, N. Strandberg et al., Development of a multi-component fiber-reinforced composite implant for load-sharing conditions, Med. Eng. Phys. 31(4) (2009), 461–469. [20] S. Tuusa, M. Peltola, T. Tirri, M.A. Puska, M. Röytta, H. Aho et al., Reconstruction of critical size calvarial bone defects in rabbits with glass-fiber-reinforced composite with bioactive glass granule coating, Journal of Biomedical Materials Research Part B – Applied Biomaterials 84B(2) (2008), 510–519. [21] A. Ballo, A. Kokkari, V. Meretoja, L.V.J. Lassila, P.K. Vallittu and T. Närhi, Osteoblast proliferation and maturation on bioactive fiber-reinforced composite surface, J. Mater. Sci. Mater. Med. 19(10) (2008), 3169–3177. [22] A. Ballo, E. Akca, T. Ozen, L.V.J. Lassila, P.K. Vallittu and T. Närhi, Bone tissue responses to glass fiber-reinforced composite implants – a histomorphometric study, Clin. Oral Implants Res. 20(6) (2009), 608–615. [23] P.K. Vallittu, Survival rates of resin-bonded, glass fiber-reinforced composite fixed partial dentures with a mean follow-up of 42 months: a pilot study, J. Prosthet. Dent. 91(3) (2004), 241–246. [24] A. Jokstad, M. Gökçe and C. Hjortsjö, A systematic review of the scientific documentation of fixed partial dentures made from fiber-reinforced polymer to replace missing teeth, Int. J. Prosthodont. 18(6) (2005), 489–496. [25] M.A. Freilich, J.C. Meiers, J.P. Duncan, K.A. Eckrote and A.J. Goldberg, Clinical evaluation of fiber-reinforced fixed bridges, J. Am. Dent. Assoc. 133(11) (2002), 1524–1534; quiz 40–41. [26] J. Rekola, A.J. Aho, P. Viitaniemi, A. Yli-Urpo, M. Hautamäki and J. Kukkonen, Wood–bone – Modified wood as a bone substitute, Suomen Ortopedia ja Traumatologia (SOT) 24 (2001), 542–544 (in Finish). [27] A.J. Aho, J. Rekola, J. Matinlinna, J. Gunn, T. Tirri, P. Viitaniemi et al., Natural composite of wood as replacement material for ostechondral bone defects, J. Biomed. Mater. Res. B Appl. Biomater. 83(1) (2007), 64–71. [28] J. Rekola, A.J. Aho, J. Gunn, J. Matinlinna, J. Hirvonen, P. Viitaniemi et al., The effect of heat treatment of wood on osteoconductivity, Acta Biomater. 5(5) (2009), 1596–1604. [29] L. Gibson, Biomechanics of cellular solids, J. Biomech. 38(3) (2005), 377–399. [30] J. Rekola, L.V.J. Lassila, J. Hirvonen, M. Lahdenperä, R. Grenman, A.J. Aho and P.K. Vallittu, Effects of heat treatment of wood on hydroxylapatite type mineral precipitation and biomechanical properties in vitro, J. Mater. Sci. Mater. Med. 21(8) (2010), 2345–2354. [31] A. Tampieri, S. Sprio, A. Ruffini, G. Celotti, I.G. Lesci and N. Roveri, From wood to bone: multi-step process to convert wood hierarchial structures into biomimetic hydroxyapatite scaffolds for bone tissue engineering, J. Mater. Chem. 19 (2009), 4973–4980. [32] P. González, J. Serra, S. Liste, S. Chiussi, B. León, M. Pérez-Amor et al., New biomorphic SiC ceramics coated with bioactive glass for biomedical applications, Biomaterials 24(26) (2003), 4827–4832. [33] M. Singh, Environment conscious ceramics (Ecoceramics), Ceram. Sci. Eng. Proc. 21 (2000), 39–44.
J. Rekola et al. / Heat treatment effects on wood morphology, surface roughness and liquid penetration
1607
[34] A. de Carlos, J.P. Borrajo, J. Serra, P. Gonzalez and B. Leon, Behaviour of MG-63 osteoblast-like cells on wood-based biomorphic SiC ceramics coated with bioactive glass, Journal of Materials Science – Materials in Medicine 17(6) (2006), 523–529. [35] P. Viitaniemi and S. Jämsä, The modification of wood by heat treatment, VTT Publication 814 (1996), 16–24 (in Finish). [36] Z. Fillo and T. Peres, Submicroscopic investigation on fibre material from beech wood heated at 120◦ C, Holztechnol. 11 (1970), 270–273. [37] F. Kollman and A. Schneider, Über das Sorptionsverhalten wärmebehandelter Hölzer, Holz. Roh. Werkst. 21(3) (1963), 77–85. [38] A. Itälä, H. Ylänen, C. Ekholm, K. Karlsson and H. Aro, Pore diameter of more than 100 µm is not requisite for bone ingrowth in rabbits, J. Biomed. Mater. Res. 58(6) (2001), 679–683. [39] S. Hansson, The dental implant meets bone – a clash of two paradigms, Applied Osseointegration Research 5 (2006), 5–17. [40] G. Gunduz, S. Korkut and D.S. Korkut, The effects of heat treatment on physical and technological properties and surface roughness of Camiyani Black Pine (Pinus nigra Arn. subsp pallasiana var. pallasiana) wood, Bioresource Technology 99(7) (2008), 2275–2280. [41] D.S. Korkut, S. Korkut, I. Bekar, M. Budakci, T. Dilik and N. Cakicier, The effects of heat treatment on the physical properties and surface roughness of Turkish Hazel (Corylus colurna L.) wood, International Journal of Molecular Sciences 9(9) (2008), 1772–1783. [42] S. Korkut and M. Budakci, The effects of high-temperature heat-treatment on physical properties and surface roughness of rowan (Sorbus aucuparia l.) wood, Wood Research 55(1) (2010), 67–78. [43] N. Ayrilmis and J.E. Winandy, Effects of post heat-treatment on surface characteristics and adhesive bonding performance of medium density fiberboard, Materials and Manufacturing Processes 24(5) (2009), 594–599. [44] H. Pecina and O. Paprzycki, Wechselbeziehungen zwischen der Temperaturbehandlung des Holzes und seiner Benetzbarkeit, Holzforsch. Holzverwert. 40 (1988), 5–8.
Copyright of Bio-Medical Materials & Engineering is the property of IOS Press and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.
1595
Effect of heat treatment of wood on the morphology, surface roughness and penetration of simulated and human blood J. Rekola a,b,c,∗ , L.V.J. Lassila a,b , S. Nganga a,b , A. Ylä-Soininmäki a,b , G.J.P. Fleming d , R. Grenman c , A.J. Aho a,b,e and P.K. Vallittu a,b a
Department of Biomaterials Science, University of Turku, Turku, Finland Biocity Turku Biomaterials Research Program, Turku Clinical Biomaterial Centre, Turku, Finland c Department of Otorhinolaryngology – Head and Neck Surgery, Turku University Hospital and University of Turku, Turku, Finland d Materials Science Unit, Division of Oral Biosciences, Dublin Dental University Hospital, Trinity College Dublin, Ireland e Department of Orthopaedics and Traumatology, Turku University Hospital, Turku, Finland b
Received 1 December 2011 Accepted 1 July 2013 Abstract. BACKGROUND: Wood has been used as a model material for the development of novel fiber-reinforced composite bone substitute biomaterials. In previous studies heat treatment of wood was perceived to significantly increase the osteoconductivity of implanted wood material. AIM: The objective of this study was to examine some of the changing attributes of wood materials that may contribute to improved biological responses gained with heat treatment. METHODS: Untreated and 140◦ C and 200◦ C heat-treated downy birch (Betula pubescens Ehrh.) were used as the wood materials. Surface roughness and the effect of pre-measurement grinding were measured with contact and non-contact profilometry. Liquid interaction was assessed with a dipping test using two manufactured liquids (simulated blood) as well as human blood. SEM was used to visualize possible heat treatment-induced changes in the hierarchical structure of wood. RESULTS: The surface roughness was observed to significantly decrease with heat treatment. Grinding methods had more influence on the surface contour and roughness than heat treatment. The penetration of the human blood in the 200◦ C heattreated exceeded that in the untreated and 140◦ C heat-treated materials. SEM showed no significant change due to heat treatment in the dry-state morphology of the wood. DISCUSSION: The results of the liquid penetration test support previous findings in literature concerning the effects of heat treatment on the biological response to implanted wood. Heat-treatment has only a marginal effect on the surface contour of wood. The highly specialized liquid conveyance system of wood may serve as a biomimetic model for the further development of tailored fiber-composite materials. Keywords: Natural fiber-composite, heat treatment, roughness, liquid interaction, biomimetic
*
Address for correspondence: J. Rekola, Department of Otorhinolaryngology and Head and Neck Surgery, Turku University Hospital, Kiinamyllynkatu 4-8, 20520 Turku, Finland. Tel.: +358 2 3130000; E-mail: [email protected]. 0959-2989/14/$27.50 © 2014 – IOS Press and the authors. All rights reserved
1596
J. Rekola et al. / Heat treatment effects on wood morphology, surface roughness and liquid penetration
1. Introduction The surface roughness and contour of a biomaterial are some of the attributes that can affect the interfacial integration of an implanted material and host bone. Implant surface modification has been reported to have a significant influence on the bone response [1–10]. The implant surface roughness increases mechanical interlocking, the active surface area and changes the wettability of the surface [1,3,7,11]. It also provides a more suitable basis for the positive biological bone response leading to increased osteoconductivity [4,5]. A biomaterial’s interaction with body fluids determines many of the properties that influence the osteoconductivity of the implant material, for example cell adhesion and spreading. In this aspect, the hydrophilicity and wettability of a biomaterial have been considered positively [12,13]. Conversely, hydrophilic surfaces also are more prone to promote bacterial colonization [14]. Liquid conveyance features are beneficial for bulk bone substitutes and penetration of the liquids enables biological activity, for example bone forming and immunological regulation inside the implanted material. On the surface level the absorption and drainage of liquid into the material can, to some extent, be simplified to a capillary action, where the possible empty spaces between fibers act as capillary tubes. The liquid influx becomes more complex at greater depths into the porous material where interconnectivity between the pores form complicated liquid conveyance systems. Fiber-reinforced composite (FRC) materials have been under development for non-metallic loadbearing implants [3–5,15–22]. The future fields of application could also include orthopedic and head and neck surgery as well as oral and maxillofacial surgery and dentistry, in which some FRC technologies have been embraced in clinical usage [23–25]. Recently wood, a natural fiber composite, has been introduced as a model material for developing synthetic FRCs [26,27]. The study design for using wood as a model material has included the use of heat treatment for managing the properties of the wood. Heat treatment is a method introduced by the wood industry to improve the biological and physical endurance of wood against the elements. In vivo studies have shown that heat treatment increases significantly the osteoconductivity of the implanted wood material [27,28]. Wood has a hierarchical structure, with highly evolved liquid conveyance features, which yields a greater biomechanical strength than the solid from which it is composed [29]. In vitro studies have shown, that the biomechanics of wood can be modified considerably using heat treatment [30]. Due to the aforementioned observations it has been concluded that heat-treated wood may be used as model material for biomaterial research in several ways; as a biomimetic model, as a bone substitute candidate, as a model material for mechanical testing and as a possible scaffold for tissue engineering. In fact, the unique anatomy of wood (see Fig. 1) has already been used in biomaterial scaffold applications and a novel approach for bulk bone substitutes [31–34]. The aim of this study was to evaluate changes that take place in a wood material during heat treatment, to better understand the biological response reported previously in the literature [27,28]. To facilitate a sensible comparison, the exact same materials (untreated, 140◦ C and 200◦ C heat-treated birch wood) as in previous studies were used. The effect of heat treatment on the surface topography of wood was measured and reported. The effect of heat treatment on liquid interaction, wetting and wicking (or liquid penetration), was measured using two manufactured liquids (simulated blood) as well as human blood. Contact angle measurements were not performed due to adequate prior knowledge available in the literature about the effects of heat treatment concerning hydrophilicity. To further understand the surface topography as well as the structures facilitating the liquid conveyance system, the wood morphology
J. Rekola et al. / Heat treatment effects on wood morphology, surface roughness and liquid penetration
(a)
1597
(b)
(c) Fig. 1. Typical structure of birch wood. The vessel (A) is the main water conducting unit in birch. The length of a single vessel cell is 0.2–0.6 mm and diameter ca. 100 µm. Vessels are separated from each other with a bar-like (scalariform) structure called perforation plate. Thus, they form long tubes. Vessels are perpendicularly connected with ray parenchyma cells (B), through small holes called pits (+). Rays function as storage cells and enable tangential water transport. They are less than 0.1 mm in length and less than 30 µm in diameter. The most common cell type in birch xylem is a fiber (C), which can be separated into tracheids and libriform fibers. They provide the main support to the structure and are 0.4–1.6 mm in length and 10–40 µm in diameter. The location of the larger magnification picture is indicated with a white box.
was imaged with Scanning Electron Microscopy (SEM) and the influence of heat treatment on liquid transport into inner structures of wood was theorized.
2. Materials and methods The European downy birch (Betula pubescens Ehrh.) wood used in the study was cut from the surface wood of larger blocks (approx 30 × 10 × 5 cm) of untreated, 140◦ C heat-treated and 200◦ C heat-treated (for a time period of two hours).
1598
J. Rekola et al. / Heat treatment effects on wood morphology, surface roughness and liquid penetration
2.1. SEM analysis The wood was cut in half longitudinally and carbon sputter coated (SCD 050; Bal-Tec, Balzers, Liechtenstein). High vacuum SEM (JSM-5500; Jeol Ltd., Tokyo, Japan) images were taken to evaluate the anatomy and microstructure of the surface and the sub-surface of the wood samples and the changes induced by heat treatment. One observer made the examinations visually. 2.2. Surface profilometry Twenty pieces of wood (of approximately 5 × 5 × 7 mm) in size were cut in half for each treatment group, untreated wood and wood heat-treated at 140◦ C and 200◦ C. The cut surfaces were then ground using a disc grinder LaboPol-21 (Struers, Westlake, OH, USA, 300 rpm) along the orientation of the fibers with silicon carbide grinding paper (grit 180) with equal pressure to remove any protruding fibers that could hinder the profilometric analysis. For each condition investigated (untreated or heat-treated at 140◦ C and 200◦ C), 10 specimens were measured (longitudinally) along the length of the fibers and a further 10 specimens were measured perpendicular to the fiber orientation. The ends of the specimens were measured from five different pieces of wood from each wood group (untreated and heat-treated at 140◦ C and 200◦ C). All of the pieces of wood were measured five times, and ground with grinding paper between every measurement. The surface roughness was measured employing contact and non-contact methodological approaches. The contact surface roughness testing was accomplished by Surftest 301 device (Mitutoyo Corporation, Japan) with an 0.8 mm sample length longitudinally to the fibers as well as over the ends of the fibers and 0.4 mm sample length perpendicular to the fibers. Every measurement run consisted of five sequential runs, thus the total number of measuring events was 50 (longitudinal and tangential) and 125 for the end surface of the specimen. To enable a methodological comparison, a non-contact surface roughness test was also used: one piece of wood from each of the heat treatment groups, after being ground manually with rotating motion with the silicon carbide grinding paper, was tested with a non-contact optical profilometer (Talysurf CLI 2000, Taylor–Hobson Precision, Leicester, England). A chromatic length aberration gauge of 3 mm was employed. The resolution was 40 nm in the vertical (z-direction) and measurements were noted every 10 µm (x-directions) and every 2 µm (y-directions). The specimens were scanned at a speed of 1 mm/s over an area of 3 × 5 mm2 . Mean surface roughness (Ra value) and standard deviation were calculated by the TalyMap analysis software package (Taylor–Hobson Precision, Leicester, England). 2.3. Liquid penetration test To demonstrate the penetration of liquids into each wood group (untreated and heat-treated at 140◦ C and 200◦ C) mainly by capillary forces, a dipping test was performed. For each liquid solution, six cylindrical specimens (4.1 mm diameter and 25 mm length) from each wood group were cut with a dental burr. At the end of the specimens, the cut surface resulting from sawing was ground first with 180 grit (FEPA) and finally with 1200 grit grinding paper using a disc grinder. The samples were fixed and dipped in a vertical alignment position for 1 min at ambient temperature in the test liquids, with the exception of human blood which was performed at 37◦ C. Pictures were taken from the dipped samples and liquid penetration depth was measured with Image J analyzer (scale 1568 pixel/10 mm in 25% zoom view). Three different liquids were used. A starch solution was prepared by mixing 150 ml deionised water with 3 g of starch, then rapidly heated to 100◦ C and boiled for three minutes. After slowly cooling the liquid, methylene blue was added for colouring. A glycerol–ethanol–water solution (GEW) was prepared by mixing 103.5 g of glycerol, 33.9 g of water and 21.9 g of ethanol with methylene blue added
J. Rekola et al. / Heat treatment effects on wood morphology, surface roughness and liquid penetration
1599
for colouring. The human blood was bought from the Finnish Red Cross (allowance 53/2010, valid until 44/2011), and consisted of the blood of 17 A+ donors. The blood was supplied in 10 ml BD vacutainers with K2 EDTA added as an anti-coagulant. The average hemoglobin was 154 g/l. The densities of the solutions were measured with a digital density-meter APDMA45 (Anton Paar High-precision instruments, Ashland, USA). The viscosities were measured with automated micro viscometer Anton Paar AMVn (Anton Paar High-precision instruments, Ashland, USA) and analysed using VisioLab for AMVn. 2.4. Statistical methods Shapiro–Wilk testing was conducted to explore the normality of the data distribution. Where the data was normally distributed, one-way ANOVA was used for between group testing, followed by Bonferroni corrected post hoc multiple comparison. In non-normally distributed results, the non-parametric Kruskal–Wallis test was performed to analyze the statistical significance of the differences between groups. The Mann–Whitney U-test was then applied to evaluate the differences between the individual groups. Statistical analysis was not performed for the non-contact profilometry, because the method was chosen for illustrative purposes and therefore there was only one sample per treatment group. 3. Results 3.1. SEM In the SEM images, the heat treatment-induced changes were quite inconspicuous. At smaller magnifications, the surface morphology and the inner anatomy of the heat-treated wood did not show differences compared with the untreated wood. However, at larger magnifications, it was observed that the cell walls in the wood heat-treated at 200◦ C were more discontinuous with microscopic cracks, and in places appeared to form lamellae (Fig. 2). This was the only observation of differences by SEM between the untreated and heat-treated (at 140◦ C and 200◦ C) groups.
Fig. 2. In places the surface of the cell walls of the 200◦ C heat-treated wood was broken (showed microcracks) and formed thin lamellae (arrow). Although sporadic cracking of cell walls was observed in every treatment group due to the shrinking induced by drying, this kind of lamellar breaking was detected only it the 200◦ C heat-treated wood. The lamellar breaking was seen scarcely and only in places where the sample cut surface was in the right angle (perpendicular) to the fibers.
1600
J. Rekola et al. / Heat treatment effects on wood morphology, surface roughness and liquid penetration
3.2. Surface profilometry In contact profilometry the arithmetical mean deviation (Ra value) for the longitudinal and perpendicular measurements was highest in the untreated group, Ra value of 3.53 (1.36) µm. The Ra values decreased with heat treatment. The mean Ra values and associated standard deviations reduced to 2.57 (0.89) µm and 2.02 (0.58) µm following heat treatment at 140 and 200◦ C, respectively. A Kruskal–Wallis test revealed statistically significant difference in Ra values across the three groups (χ2 = 22.89, p < 0.001). Additional Mann–Whitney U-tests with correction for multiple comparisons disclosed all individual between-groups differences to be statistically significant. For the end surfaces of the specimens the Ra values were 2.20 (0.81) µm for the untreated wood groups, 1.57 (0.62) µm for the 140◦ C heat-treated and 1.25 (0.38) µm for the 200◦ C heat-treated group. A Kruskal–Wallis test showed a statistically significant difference in Ra values across the three groups (χ2 = 23.50, p < 0.001). The Mann–Whitney U-test indicated all between-groups differences to be statistically significant. For the non-contact profilometry, the mean roughness was highest in the specimen which was heattreated at 200◦ C, with a Ra value of 24.33 (3.28) µm. An intermediary Ra value (13.93 (1.58) µm) was obtained on heat treatment at 140◦ C and the untreated wood material had a Ra value of 4.24 (0.59) µm. The graphical illustration of the surface topography is depicted in Fig. 3. 3.3. Liquid penetration test The results of the liquid penetration (dipping) tests are depicted in Fig. 4. The densities and viscosities of the test liquids are reported in Table 1. Being normally distributed, the results for the penetration of human blood were analyzed with a one-way ANOVA, which revealed statistically significant differences between groups (F = 27.850, p < 0.001). A Bonferroni corrected post hoc comparison revealed the 200◦ C heat-treated group differed from the untreated and 140◦ C heat-treated groups with statistical significance, while there was no statistically significant difference between the untreated and 140◦ C heat-treated groups. For GEW solution dipping results, non-parametric testing was used. Kruskal–Wallis test revealed statistically significant differences between groups (χ2 = 7.94, p = 0.019). Individual group comparisons revealed the 200◦ C heat-treated group differed from the untreated and 140◦ C heat-treated groups with statistical significance (p = 0.010 and p = 0.025 for untreated and 140◦ C heat-treated groups, respectively), while there was no statistically significant difference between the untreated and 140◦ C heat-treated groups. Starch solution dipping results were analyzed with a one-way ANOVA, which revealed statistically significant differences between all groups (F = 11.511, p = 0.001). A Bonferroni corrected post hoc analysis revealed statistically relevant differences between the untreated and 140◦ C group (p = 0.001). The differences between 200◦ C group and both untreated and 140◦ C group did not reach statistical significance. In a gross picture analysis, the penetrations of all of the liquids were maximal near the surface of the samples. In the simulated blood liquids, methylene blue coloration was separated from the solutions, not penetrating as fast as the rest of the liquid into the wood structures. The level of methylene blue was nevertheless measured as the definite penetration depth, because the clear solution penetration depth was considered too difficult to determine. In fact, in the case of the GEW, the clear portion of the liquid reached the end of the samples in all of the birch wood (untreated and heat-treated at 140◦ C and 200◦ C) groups.
J. Rekola et al. / Heat treatment effects on wood morphology, surface roughness and liquid penetration
1601
(a)
(b)
(c)
(d) Fig. 3. A graph showing the results of the contact profilometry as well as illustrations of the surface analysis of non-contact profilometry. The sand paper grinding shows a significant effect on the surface with the non-contact method. The surface of the 200◦ C heat-treated wood is distinctly more uneven, with numerous spikes representing protruding fibers. The normal fiber orientation seen on the untreated and 140◦ C wood has become indistinguishable on the surface of the 200◦ C heat-treated material. (Colors are visible in the online version of the article; http://dx.doi.org/10.3233/BME-140964.)
1602
J. Rekola et al. / Heat treatment effects on wood morphology, surface roughness and liquid penetration
(a)
(b)
(c) Fig. 4. (a) The results of the liquid penetration test. The actual dipping depth is not included in the results. (b) Samples after dipping. 200◦ C heat-treated (A), 140◦ C heat-treated (B) and untreated (C) dipped in (from left to right) starch solution, GEW solution and human blood. (c) Samples cross sections after dipping: 200◦ C heat-treated, 140◦ C heat-treated and untreated dipped in (from left to right) starch solution, GEW solution and human blood. In the gross analysis of cut surface only the hydrophobic GEW solution had significant visible penetration to the inner structures. This is largely due to the short dipping time of 1 min. (Colors are visible in the online version of the article; http://dx.doi.org/10.3233/BME-140964.)
J. Rekola et al. / Heat treatment effects on wood morphology, surface roughness and liquid penetration
1603
Table 1 The density and viscosity of the tested liquids Liquid Starch solution GEW Human blood
Density (g/cm3 ) 1.0005 1.1247 1.0598
Dynamic viscosity (mPa · s) 5.892 ± 0.4638 21.301 ± 1.152 4.9151 ± 0.0696
4. Discussion 4.1. Wood morphology The lamellar breaking of the wood cell walls observed in this study confirm the findings in the literature describing the effects of heat treatment [35]. On drying the wood, microcracks appear in the cell walls due to excessive shrinkage. The cell wall consists of lamellae formed by microfibrils. During heat treatment these layers can become loosened and break, forming the observed protruded lamellae [36]. This phenomenon can yield more relative surface as well as form more protruding fibers at the free surface. The nature of the high vacuum SEM that was used in this study meant that the wood material had to be analyzed as dry as possible. It appeared that the gross anatomy of the dry wood material does not change significantly during the heat treatment. In a wet state, however, the behaviour of the wood is considerably different due to heat treatment protocols [30]. The ability of birch wood to absorb water into the cell walls decreases with heat treatment, leading to reduced swelling [30,37]. This leads to an altered wet state morphology between heat-treated groups; at the higher heat treatment (200◦ C) group liquid remains proportionally more in the hollow channels i.e. cell lumens and not within the cell walls. Although this situation can be said to resemble more the liquid distribution found within a living bone, the biological relevance of this phenomenon cannot be concluded without additional studies and further evidence. The diameter of the open canals (or pores) is a critical factor in determining the quality of the tissue growth into the free space. The literature suggests that the optimal pore size of a biomaterial for mineralized bone in-growth is 100–400 µm, although pore sizes of 50–125 µm have been found to be sufficient for bone tissue in-growth [38]. The inter-connected canalicular structure of wood (Fig. 1) allows the conduction of liquids especially along the cell axis. However, there are number of structures in the anatomy of wood that also allow tangential movement of liquid molecules, i.e. tangential rays and pores. This liquid conducting system (called superapoplasm) has been sophisticated by evolution in nature and there probably is considerable information to be gleaned for biomaterial research. 4.2. Surface profilometry The general rule of thumb that the rougher the surface, the better the implant material, is not true per se. Although a larger surface area means an increase surface contact surface, the specific structural architecture, i.e. the contour of the surface, has a significant effect on the distribution of the shear forces [39]. The results of the contact profilometry of this study are in line with the previously published literature, where similar tests have been conducted with other wood species treated with different heat treatment methodologies [40–43]. The mechanism of the heat treatment-induced decrease in roughness is not clear.
1604
J. Rekola et al. / Heat treatment effects on wood morphology, surface roughness and liquid penetration
One suggestion is that high temperatures alter lignin to a thermoplastic condition which in turn densifies the cell surfaces [41]. The partial disintegration of wood cell walls may also have an influence on surface roughness. Grinding has a significant influence on the surface, when measuring on a micrometer scale, which is why the number of repeated measurements has to be high. The protruding fibers most likely give way to the contact stylus, which can partially explain the large difference in results between contact and non-contact methodological approaches. The effect of grinding is clearly illustrated with the non-contact profilometry. Different grinding pre-treatment, namely rotational manual grinding with silicon carbide abrasive paper not only increased the roughness, but also changed the gross profile and texture of the surface. The non-contact test illustrates how easy it is to modify the surface of the wood, which can be desirable in bone substitute biomaterials. As far as the literature is concerned, the contact stylus method seems to be the gold standard in wood surface profilometry. For biological applications, however, the three dimensional non-contact method can possibly yield significantly more information, especially in fiberous materials as it also illustrates the contour of the measured surface, whereas the two dimensional straight-line measurement of a contact stylus, for instance used in this study, fails to do so. Overall, albeit being statistically significant, the change in surface roughness between the untreated and heat-treated (140 and 200◦ C) wood is however so inconclusive that it probably does not bear much if any relevance biologically. 4.3. Liquid penetration test As the gross canalicular structure of the wood does not change with heat treatment, the wicking effect depends on the viscosity of the solutions and on the hydrophobic/hydrophilic nature of the wood and liquids tested. The literature suggests that the degradation and cross-linking of cellulose and hemicellulose reduce the hydrophilicity. Wood is mostly hydrophobic when heat-treated at 190◦ C. At elevated temperatures, the formation of hydrophilic degradation products seems to again transform the wood to be more hydrophilic in nature [44]. The most hydrophilic liquid employed in this study (i.e. the starch solution) interacted best with the untreated wood, whereas the GEW solution and human blood, being relatively more hydrophobic in nature, had better interaction with the 200◦ C heat-treated wood. The separation of methylene blue coloration from the solution indicates that larger molecules and particles travel slightly slower into the wood structure than small water molecules. This finding may be related to the pore size of the material. In previous in vivo studies, some bone formation has been found inside implanted 200◦ C heat-treated wood implants, indicating that the pore size is sufficient for the penetration of undifferentiated mesenchymal cells [27]. However, the presence of bigger (>100 µm) canals would probably be beneficial in this aspect. As mentioned previously, the ability for dimensional changes diminishes with heat treatment and when submerged into simulated body fluid for a sufficient amount of time with the untreated wood absorbing the greatest volume of water [30]. The 1-min dipping time used in this work excludes the effect of cell wall swelling, leaving the wood’s canalicular conveyance ability and the liquid–material interaction as the affecting attributes. For the testing of optimal biological response, the result of the human blood dipping is the most interesting. The penetration of the blood in the 200◦ C heat-treated wood exceeded that in the untreated and 140◦ C heat-treated wood. This result is in conjunction with previous in vivo and in vitro tests, where the 200◦ C heat-treated wood was the most osteoconductive as well as the most prone to hydroxyl apatite formation, when compared with untreated and 140◦ C heat-treated wood [28, 30]. The heat treatment seems to have a positive effect on the interaction between a bulk of wood and blood – most likely due to the heat treatment – induced chemical changes in the wood polymer. The
J. Rekola et al. / Heat treatment effects on wood morphology, surface roughness and liquid penetration
1605
microstructural alterations as well as the changes in surface roughness are likely to have only a limited influence on this phenomenon.
5. Conclusions The following conclusions can be drawn, bearing in mind the limitations of the study: (1) Birch wood samples show typical hardwood structure with vessels and fibers. The dry-statemorphology of these structures does not change significantly during heat treatment. The specified liquid conducting system of wood could yield useful information for the development of synthetic fiber biomaterials. (2) Heat treatment reduces slightly the surface roughness of wood when measured with contact profilometry. The surface of wood is very prone to the effects of grinding which underlines the importance of cautious preprocessing when characterizing the surface topography of a fiberous biomaterial. The increase in the osteoconductivity of wood associated with heat treatment is more likely due to the chemical changes in the wood material than to the change in surface roughness. (3) Heat treatment most likely changes the liquid distribution in the deeper structures of wood samples by reducing the absorption of liquids into the cell walls, while leaving the channel structure unchanged. (4) Heat treatment increases the penetration of liquids with increased hydrophobicity, including human blood, into the structure of wood.
Acknowledgements We would like to thank Mikko Salomäki from Analytical Chemistry Institute/Arcanum, Turku, Finland for analysing the viscosities and densities of the liquids used in this study and Adam Dowling from the Materials Science Unit, Division of Oral Biosciences, Dublin Dental University Hospital, Trinity College Dublin, Ireland, for performing the non-contact profilometry analyses. Docent Pekka Saranpää from the Finnish Forest Research Institute, METLA, Vantaa, is acknowledged for his contribution in the writing of the text.
References [1] S. Hansson and M. Norton, The relation between surface roughness and interfacial shear strength for bone-anchored implants. A mathematical model, J. Biomech. 32(8) (1999), 829–836. [2] L. Rasmusson, J. Roos and H. Bystedt, A 10-year follow-up study of titanium dioxide-blasted implants, Clin. Implant Dent. Relat. Res. 7(1) (2005), 36–42. [3] R.H. Mattila, M.A. Puska, L.V.J. Lassila and P.K. Vallittu, Fibre-reinforced composite implant: in vitro mechanical interlocking with bone model material and residual monomer analysis, Journal of Materials Science 41(13) (2006), 4321– 4326. [4] R.H. Mattila, P. Laurila, J. Rekola, J. Gunn, L.V.J. Lassila, T. Mäntylä et al., Bone attachment to glass-fibre-reinforced composite implant with porous surface, Acta Biomater. 5(5) (2009), 1639–1646. [5] A. Aho, M. Hautamäki, R.H. Mattila, P. Alander, N. Strandberg, J. Rekola et al., Surface porous fibre-reinforced composite bulk bone substitute, Cell Tissue Bank 5(4) (2004), 213–221. [6] L. Cooper, A role for surface topography in creating and maintaining bone at titanium endosseous implants, J. Prosthet. Dent. 84(5) (2000), 522–534.
1606
J. Rekola et al. / Heat treatment effects on wood morphology, surface roughness and liquid penetration
[7] S. Hansson, Surface roughness parameters as predictors of anchorage strength in bone: a critical analysis, J. Biomech. 33(10) (2000), 1297–1303. [8] T. Albrektsson and A. Wennerberg, Oral implant surfaces: Part 1 – review focusing on topographic and chemical properties of different surfaces and in vivo responses to them, Int. J. Prosthodont. 17(5) (2004), 536–543. [9] T. Albrektsson and A. Wennerberg, Oral implant surfaces: Part 2 – review focusing on clinical knowledge of different surfaces, Int. J. Prosthodont. 17(5) (2004), 544–564. [10] C. Ivanoff, G. Widmark, C. Johansson and A. Wennerberg, Histologic evaluation of bone response to oxidized and turned titanium micro-implants in human jawbone, Int. J. Oral Maxillofac. Implants 18(3) (2003), 341–348. [11] P. de Gennes, F. Brochard-Wyart and D. Quéré, Capillarity and Wetting Phenomena: Drops, Bubbles, Pearls and Waves, Springer-Verlag, 2004. [12] R.J. LaPorte, Hydrophilic Polymer Coatings for Medical Devices – Structure/Properties, Development, Manufacture and Applications, Technomic Publishing Company, Lancaster, USA, 1997. [13] T.G. Ruardy, H.E. Moorlag, J.M. Schakenraad, H.C. van der Mei and H.J. Busscher, Growth of fibroblasts and endothelial cells on wettability gradient surfaces, Journal of Colloid and Interface Science 188 (1997), 209–217. [14] W.-Y. Su, S.-C. Wu, L. Wu, X.-X. Wang, P. Yue and X.-Z. Fu, Surface ant-bacterial adhesion characteristics of TiO2 coated PMMA, Chinese Journal of Structural Chemistry 28 (2009), 1497–1502. [15] S. Tuusa, M. Peltola, T. Tirri, L.V.J. Lassila and P.K. Vallittu, Comparison of two fiber-reinforced composite structures as implant material in calvarial bone defect, Bioceramics, Key Engineering Materials 361–363 (2007), 471–474. [16] M. Hautamäki, A.J. Aho, P. Alander, J. Rekola, J. Gunn, N. Strandberg et al., Repair of bone segment defects with surface porous fiber-reinforced polymethyl methacrylate (PMMA) composite prosthesis: histomorphometric incorporation model and characterization by SEM, Acta Orthop. 79(4) (2008), 555–564. [17] S. Tuusa, M. Peltola, T. Tirri, L.V.J. Lassila and P.K. Vallittu, Frontal bone defect repair with experimental glass-fiberreinforced composite with bioactive glass granule coating, Journal of Biomedical Materials Research Part B – Applied Biomaterials 82B(1) (2007), 149–155. [18] A. Ballo, L.V.J. Lassila, T. Närhi and P.K. Vallittu, In vitro mechanical testing of glass fiber-reinforced composite used as dental implants, J. Contemp. Dent. Pract. 9(2) (2008), 41–48. [19] D. Zhao, N. Moritz, P. Laurila, R.H. Mattila, L.V.J. Lassila, N. Strandberg et al., Development of a multi-component fiber-reinforced composite implant for load-sharing conditions, Med. Eng. Phys. 31(4) (2009), 461–469. [20] S. Tuusa, M. Peltola, T. Tirri, M.A. Puska, M. Röytta, H. Aho et al., Reconstruction of critical size calvarial bone defects in rabbits with glass-fiber-reinforced composite with bioactive glass granule coating, Journal of Biomedical Materials Research Part B – Applied Biomaterials 84B(2) (2008), 510–519. [21] A. Ballo, A. Kokkari, V. Meretoja, L.V.J. Lassila, P.K. Vallittu and T. Närhi, Osteoblast proliferation and maturation on bioactive fiber-reinforced composite surface, J. Mater. Sci. Mater. Med. 19(10) (2008), 3169–3177. [22] A. Ballo, E. Akca, T. Ozen, L.V.J. Lassila, P.K. Vallittu and T. Närhi, Bone tissue responses to glass fiber-reinforced composite implants – a histomorphometric study, Clin. Oral Implants Res. 20(6) (2009), 608–615. [23] P.K. Vallittu, Survival rates of resin-bonded, glass fiber-reinforced composite fixed partial dentures with a mean follow-up of 42 months: a pilot study, J. Prosthet. Dent. 91(3) (2004), 241–246. [24] A. Jokstad, M. Gökçe and C. Hjortsjö, A systematic review of the scientific documentation of fixed partial dentures made from fiber-reinforced polymer to replace missing teeth, Int. J. Prosthodont. 18(6) (2005), 489–496. [25] M.A. Freilich, J.C. Meiers, J.P. Duncan, K.A. Eckrote and A.J. Goldberg, Clinical evaluation of fiber-reinforced fixed bridges, J. Am. Dent. Assoc. 133(11) (2002), 1524–1534; quiz 40–41. [26] J. Rekola, A.J. Aho, P. Viitaniemi, A. Yli-Urpo, M. Hautamäki and J. Kukkonen, Wood–bone – Modified wood as a bone substitute, Suomen Ortopedia ja Traumatologia (SOT) 24 (2001), 542–544 (in Finish). [27] A.J. Aho, J. Rekola, J. Matinlinna, J. Gunn, T. Tirri, P. Viitaniemi et al., Natural composite of wood as replacement material for ostechondral bone defects, J. Biomed. Mater. Res. B Appl. Biomater. 83(1) (2007), 64–71. [28] J. Rekola, A.J. Aho, J. Gunn, J. Matinlinna, J. Hirvonen, P. Viitaniemi et al., The effect of heat treatment of wood on osteoconductivity, Acta Biomater. 5(5) (2009), 1596–1604. [29] L. Gibson, Biomechanics of cellular solids, J. Biomech. 38(3) (2005), 377–399. [30] J. Rekola, L.V.J. Lassila, J. Hirvonen, M. Lahdenperä, R. Grenman, A.J. Aho and P.K. Vallittu, Effects of heat treatment of wood on hydroxylapatite type mineral precipitation and biomechanical properties in vitro, J. Mater. Sci. Mater. Med. 21(8) (2010), 2345–2354. [31] A. Tampieri, S. Sprio, A. Ruffini, G. Celotti, I.G. Lesci and N. Roveri, From wood to bone: multi-step process to convert wood hierarchial structures into biomimetic hydroxyapatite scaffolds for bone tissue engineering, J. Mater. Chem. 19 (2009), 4973–4980. [32] P. González, J. Serra, S. Liste, S. Chiussi, B. León, M. Pérez-Amor et al., New biomorphic SiC ceramics coated with bioactive glass for biomedical applications, Biomaterials 24(26) (2003), 4827–4832. [33] M. Singh, Environment conscious ceramics (Ecoceramics), Ceram. Sci. Eng. Proc. 21 (2000), 39–44.
J. Rekola et al. / Heat treatment effects on wood morphology, surface roughness and liquid penetration
1607
[34] A. de Carlos, J.P. Borrajo, J. Serra, P. Gonzalez and B. Leon, Behaviour of MG-63 osteoblast-like cells on wood-based biomorphic SiC ceramics coated with bioactive glass, Journal of Materials Science – Materials in Medicine 17(6) (2006), 523–529. [35] P. Viitaniemi and S. Jämsä, The modification of wood by heat treatment, VTT Publication 814 (1996), 16–24 (in Finish). [36] Z. Fillo and T. Peres, Submicroscopic investigation on fibre material from beech wood heated at 120◦ C, Holztechnol. 11 (1970), 270–273. [37] F. Kollman and A. Schneider, Über das Sorptionsverhalten wärmebehandelter Hölzer, Holz. Roh. Werkst. 21(3) (1963), 77–85. [38] A. Itälä, H. Ylänen, C. Ekholm, K. Karlsson and H. Aro, Pore diameter of more than 100 µm is not requisite for bone ingrowth in rabbits, J. Biomed. Mater. Res. 58(6) (2001), 679–683. [39] S. Hansson, The dental implant meets bone – a clash of two paradigms, Applied Osseointegration Research 5 (2006), 5–17. [40] G. Gunduz, S. Korkut and D.S. Korkut, The effects of heat treatment on physical and technological properties and surface roughness of Camiyani Black Pine (Pinus nigra Arn. subsp pallasiana var. pallasiana) wood, Bioresource Technology 99(7) (2008), 2275–2280. [41] D.S. Korkut, S. Korkut, I. Bekar, M. Budakci, T. Dilik and N. Cakicier, The effects of heat treatment on the physical properties and surface roughness of Turkish Hazel (Corylus colurna L.) wood, International Journal of Molecular Sciences 9(9) (2008), 1772–1783. [42] S. Korkut and M. Budakci, The effects of high-temperature heat-treatment on physical properties and surface roughness of rowan (Sorbus aucuparia l.) wood, Wood Research 55(1) (2010), 67–78. [43] N. Ayrilmis and J.E. Winandy, Effects of post heat-treatment on surface characteristics and adhesive bonding performance of medium density fiberboard, Materials and Manufacturing Processes 24(5) (2009), 594–599. [44] H. Pecina and O. Paprzycki, Wechselbeziehungen zwischen der Temperaturbehandlung des Holzes und seiner Benetzbarkeit, Holzforsch. Holzverwert. 40 (1988), 5–8.
Copyright of Bio-Medical Materials & Engineering is the property of IOS Press and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.
