Langmuir 2006, 22, 4843-4845
4843
Evaluation of Surface Roughness of Hydrogels by Fractal Texture Analysis during Swelling G. Mabilleau, M. F. Basle´, and D. Chappard* INSERM, EMI 0335 - LHEA, Faculte´ de Me´ decine, 49045 Angers Cedex, France ReceiVed February 7, 2006. In Final Form: March 20, 2006 The surface of a biomaterial reacts in contact with biological fluids. Hydrogels are used to prepare biomaterials. The surface roughness of materials can be explored by several techniques. However, when considering hydrogels, the surface examined in the dry state does not reflect the final conformation. How the surface roughness is affected by swelling has been little explored by quantitative methods. We have evaluated the surface roughness of poly(2hydroxyethyl methacrylate) (i.e., pHEMA) by image analysis. Images of disks, prepared from linear pHEMA, were obtained on a light microscope after various incubation times in saline. Fractal texture analysis was done on images to determine the fractal dimension D. In this study, D exhibited a significant decrease during swelling and was highly correlated with the swelling ratio (r2) 0.994, p < 0.00001). Water uptake by the surface of the polymer affected the surface roughness. Image analysis using fractal algorithms appears to be the most interesting technique for the quantitative exploration of surfaces of hydrated materials that cannot be measured by conventional methods.
1. Introduction Polymers are widely used in medicine to make medical devices, prostheses, cements or coatings, and microbeads and are also used in the encapsulation of cells and the controlled release of active compounds.1 Hydrogels are a class of hydrophilic materials formed by a 3D network of chains, held together by weak cohesive forces in the form of hydrogen or ionic bonds. When the chains are cross linked, these materials exhibit increased mechanical resistance and are able to take up large quantities of water and biological fluids without dissolution. The biocompatibility of hydrogels is due to their ability to mimic tissues in relation to their high hydration and their special surface properties. The swelling of hydrogels changes in response to external conditions such as temperature, pH, ionic strength, and electrical potential.2-4 Hydrogels such as poly(vinyl alcohol) (PVA), poly(N-vinyl-2pyrrolidone) (PVP), and poly(2-hydroxyethyl methacrylate) (pHEMA) that have the capacity to swell in body fluids are employed for the controlled-release delivery of water-soluble drugs.5-9 Hydrogels are divided in two groups: (1) highly swollen hydrogels including cellulose derivatives, PVA, and PVP, among others, and (2) moderately and poorly swollen hydrogels including pHEMA and many of its derivatives. Surfaces and interfaces are areas where material properties can be different from the entire material.10 Scanning electron microscopy (SEM) provides information on the surface of dehydrated materials but does not reflect the biomaterials in * Corresponding author. E-mail:
[email protected]. Tel: (33) 241 73 58 65. (1) Thomas, D. W. AdVanced Biomaterials for Medical Applications, 1st ed.; Springer: Sofia, 2005; Vol. 180. (2) Okano, T.; Bae, Y. H.; Jacobs, H.; Kim, S. W. J. Controlled Release 1990, 11, 255-265. (3) Alhaique, F.; Marchetti, M.; Riccieri, F. M.; Santucci, E. J. Pharm. Pharmacol. 1981, 33, 413-418. (4) Lee, W. F.; Chen, C. F. Polym. Gels Networks 1998, 6, 493-511. (5) Cascone, M. G.; Pot, P. M.; Lazzeri, L.; Zhu, Z. J. Mater. Sci. Mater. Med. 2002, 13, 265-269. (6) Hsiue, G. H.; Chang, R. W.; Wang, C. H.; Lee, S. H. Biomaterials 2003, 24, 2423-2430. (7) Kaneda, Y.; Tsutsumi, Y.; Yoshioka, Y.; Kamada, H.; Yamamoto, Y.; Kodaira, H.; Tsunoda, S.; Okamoto, T.; Mukai, Y.; Shibata, H.; Nakagawa, S.; Mayumi, T. Biomaterials 2004, 25, 3259-3266. (8) Lu, S.; Anseth, K. S. J. Controlled Release 1999, 57, 291-300. (9) Seabra, A. B.; De Oliveira, M. G. Biomaterials 2004, 25, 3773-3782. (10) Morra, M. John Wiley & Sons: Chichester, U.K., 2002.
“real life” conditions when they are placed into biologic conditions of hydration.11 The environmental scanning electron microscope (ESEM) in the wet mode and a gaseous secondary electron detector have been proposed in the analysis of poly(ethylene glycol) hydrogels and hydrated gelatins.12,13 The technique is useful for following changes in surface topography but does not provide measurements of the surface roughness. Atomic force microscopy (AFM) allows the study of material surfaces in aqueous environments simulating in vivo conditions.14-16 Surface roughness can be measured by AFM on the nanoscopic level and depends of the scan size. However, the method has been seldom used for hydrogels because it is applicable only to thin films; it is not applicable on highly hydrated hydrogels because the probe is “limed” in the highly swollen layers of the material.17 In the same way, profilometry (contact or optical), which has often been used to study the surface roughness of biomaterials in the dried or hydrated state, cannot be used in highly hydrated gels. We have investigated the evolution of surface roughness on large areas of pHEMA during its swelling phase; pHEMA is a polymer with high hydrophilicity.18 Fractal texture analysis on microphotographs was chosen with the “skyscrapers” algorithm. This method was suggested for digitized mammography in search of malignant microcalcifications.19 It was found to be suitable for measuring the roughness of titanium test pieces from SEM images of their surface, and the fractal dimension (D) was found to be highly correlated with contact profilometry data.20 The surface is the most reactive area of a biomaterial when it is in (11) Mohan, Y. M.; Murthy, P. S. K.; Sreeramulu, J.; Raju, K. M. J. Appl. Polym. Sci. 2005, 98, 302-314. (12) Gattas-Asfura, K. M.; Weisman, E.; Andreopoulos, F. M.; Micic, M.; Muller, B.; Sirpal, S.; Pham, S. M.; Leblanc, R. M. Biomacromolecules 2005, 6, 1503-1509. (13) Micic, M.; Zheng, Y.; Moy, V.; Zhang, X.-H.; Andreopoulos, F. M.; Leblanc, R. M. Colloids Surf., B 2003, 27, 147-158. (14) Baguet, J.; Sommer, F.; Duc, T. M. Biomaterials 1993, 14, 279-284. (15) Siedlecki, C. A.; Marchant, R. E. Biomaterials 1998, 19, 441-454. (16) Mabilleau, G.; Moreau, M. F.; Filmon, R.; Basle´, M. F.; Chappard, D. Biomaterials 2004, 25, 5155-5162. (17) Guryca, V.; Pacakova, V.; Tlust’akova, M.; Stulik, K.; Michalek, J. J. Sep. Sci. 2004, 27, 1121-1129. (18) Monthe´ard, J. P.; Kahovec, J.; Chappard, D. In Desk Reference of Functional Polymers: Syntheses and Applications; Arshady, R., Ed.; American Chemical Society: Washington, DC, 1997; pp 699-717. (19) Caldwell, C. B.; Stapleton, S. J.; Holdsworth, D. W.; Jong, R. A.; Weiser, W. J.; Cooke, G.; Yaffe, M. J. Phys. Med. Biol. 1990, 35, 235-247.
10.1021/la060368v CCC: $33.50 © 2006 American Chemical Society Published on Web 04/15/2006
4844 Langmuir, Vol. 22, No. 10, 2006
Mabilleau et al.
contact with biological fluids. In this study, we have prepared disks of linear pHEMA by radical polymerization. Disks were allowed to swell in saline; the determination of water uptake (swelling ratio) and texture analysis were performed at regular time intervals. 2. Experimental Section Preparation of Polymer Disks. Commercial HEMA contains residual methacrylic acid and cross linkers due to the fabrication process. The polymerization inhibitor 4-methoxyphenol (added by the manufacturer before shipping at a concentration of 350 ppm) also needs to be removed. HEMA was purified and distilled under reduced pressure (5 × 10-2 mbar, 70 °C). All reagents were obtained from Sigma-Aldrich Chemical (Illkirsh, France). The linear polymer was prepared by bulk polymerization. Briefly, the polymerization mixture was composed of HEMA (10 mL) and 0.2 g of benzoyl peroxide (BPO) used as an initiator. The mixture was accelerated by N,N-dimethyl-p-toluidine (NNDMPT) in 100:1 mol/mol BPO/NNDMPT. Monomers were polymerized at 4 °C for 2 h in polypropylene wells (Delta Microscopies, Labege, France). Evaluation of Swelling Behavior. The swelling behavior was evaluated by the blot-and-weight technique. All disks were weighed to determine the dry weight (Wd), and then they were allowed to swell in sterile saline at a constant temperature (22 °C). Experiments were done with six disks of polymer, each sample being immersed in 50 mL of saline. After immersion at different incubation intervals, disks were blotted to remove excess saline and were weighed to evaluate the wet weight (Ww). After weighing, disks were immersed in saline to carry on the swelling process. The swelling degree was evaluated by quantifying the water uptake using the equation Ww - Wd × 100 water uptake ) Wd Measurement of Surface Roughness. Surface roughness was evaluated by texture analysis of microscopic images using a fractal algorithm (the “skyscraper” algorithm).19-21 The surfaces of six disks of pHEMA were photographed in the dried state and after 5, 15, 30, 60, 240, and 720 min of immersion in sterile saline. Disks were placed on a glass slide and imaged with episcopic illumination on a Leica DMR D microscope (Leica, Rueil-Malmaison, France) at a magnification of 100×. The disks’ surfaces were illuminated using an external fiber optic illuminator providing cold light to limit the evaporation of water. The light was positioned at a 30° angle with respect to the surface of the disks to increase the details of the surface. All samples were examined under the same experimental conditions. Numeric microphotographs were obtained with an Olympus Camedia C3040 (Olympus, Rueil-Malmaison, France) with a fixed aperture. A region of interest composed of 512 pixels × 512 pixels was trimmed by overimposing a mask on the micrographs using Adobe Photoshop software, release 8 (Adobe, Edinburgh, Scotland). Images were analyzed with laboratory-made software written in Visual Basic (Microsoft Corp., Redmond, WA). The accuracy of the measurements was previously assessed from computer-simulated images of known fractal dimension.20,22 Images of the hydrogel surfaces were in the bmp format and coded on 8 bits (i.e., in 256 gray levels; black ) 0, white ) 255). Pixels, which constitute an image A, can be considered to be skyscrapers whose heights are represented by the gray level. The roof of a skyscraper is a square of side . The surface area of the image A() is obtained by measuring the sum of the top surface (2) and the sum of the exposed lateral sides of the skyscrapers. The gray levels of adjacent pixels are then averaged in squares of ) 2, 4, 8, 16, and 32 pixels to produce new images, and A() is calculated for each according to A() )
∑ + ∑[abs[Z(x, y) - Z(x + 1, y)] + 2
abs[Z(x, y) - Z(x, y + 1)]
Figure 1. Effects of hydration on pHEMA disks. (A) Percentage of water uptake and evolution of the fractal dimension with the time of hydration. (B) Correlation between the fractal dimension and water uptake. where Z(x, y) is the height of a skyscraper in the x, y plane. abs is the absolute value. The fractal dimension of the surface (D) was determined by plotting a graph of log A() versus log . The linear regression line was computed only on the aligned points by the least-squares method. The fractal dimension was obtained as D ) 2 - slope. Statistical Analysis. Statistical analysis was performed using Systat statistical software, release 11 (SPSS Inc., Chicago, IL). All results were expressed as mean ( standard error of the mean. The analysis of variance (ANOVA) was used to compare the differences between the groups. Differences were considered to be significant at p < 0.05.
3. Results and Discussion Figure 1A shows the time course evolution of the swelling ratio, determined by the percentage of water uptake, and the evolution of D measured at the surface of the same disks. pHEMA swelled in saline to reach 5.04 ( 0.87% after 5 min and 19.76 ( 1.62% after 240 min. Our results are similar to those reported by others.23 D measured on the dry polymer was 2.403 ( 0.012. After 5 min of immersion in saline, D exhibited a significant decrease of -0.9%; it continued to decrease significantly over the time of study to reach 2.288 ( 0.007 (-4.78%) after 240 min. The correlation between water uptake and D was found to be linear as shown in Figure 1B with r2 ) 0.994. The surface topography of dry polymer appears in Figure 2. For this study, we have chosen the face of the disk that was in direct contact (20) Chappard, D.; Degasne, I.; Hure´, G.; Legrand, E.; Audran, M.; Basle´, M. F. Biomaterials 2003, 24, 1399-1407. (21) Chappard, D.; Chennebault, A.; Moreau, M.; Legrand, E.; Audran, M.; Basle´, M. F. Bone 2001, 28, 72-79. (22) Russ, J. C. Plenum Press: New York, 1994. (23) Brazel, C. S.; Peppas, N. A. Eur. J. Pharm. Biopharm. 2000, 49, 47-58.
EValuation of Surface Roughness of Hydrogels
Langmuir, Vol. 22, No. 10, 2006 4845
When a dry hydrophilic polymer network is placed in saline, ionizable groups of its macromolecular chains interact with the ionic species in the solvent, thus resulting in an expansion of the network.24 Polymer chains at the surface are the first to swell when they are in contact with saline. After 30 min of hydration, chain relaxation (from the surface to the point of progression) has occurred because of the progression of water. With time, water penetrates deeper and chain relaxation continues, leading to a decrease in surface roughness. Many natural objects are irregular in shape; they possess branchings, holes, and irregularities that considerably increase their complexity.25 Mandelbrot established fractal geometry to describe the complexity of objects found in nature that markedly differ from man-made objects based on Euclidean geometry.26 A fractal object tends to fill up space, and its dimension value is not an integer. For example, a complex curve will cover a large area of the plane with a topological dimension lying between 1 and 2. The “monster curves” described by mathematicians tend to fill the whole plane and have a high fractal dimension. (The Peano curve is 1.67.) In the same way, a very rough surface will occupy a large volume with a dimension between 2 and 3.22,27 When an object has a homogeneous composition, an image of its surface is known to be directly related to the real relief of its surface.28 Material surfaces are often known to be fractal, but surface images are reported as being self-similar rather than selfaffine (a characteristic that is observed only on mathematical objects).22,28 Fractal analysis is a powerful tool in the analysis of subtle changes occurring in objects; we recently found that the removal of medullary lipids influenced the surface roughness of radiographic images of bone, before and after delipidation.29 During the last few decades, several authors have tried to quantify the surface roughness of materials by image analysis.20,30-32 A psychological experiment on the visual appreciation of surface roughness is reported by Pentland.33 Various artificial textures were produced by computer imaging and were submitted to a population of students. They were asked to classify them as more or less rough according to their own visual impression. The visual classification was highly correlated with the fractal dimension measured on the images.
Conclusions In this study, the fractal dimension of a hydrogel surface decreased during swelling and tended to reach 2, the lower topological dimension. The fractal dimension appears to be an interesting parameter related to surface roughness when this parameter cannot be measured by any currently available technique. Acknowledgment. G.M. thanks the French Ministry of Research and Technology for a fellowship. This work was made possible by grants from Contrat de Plan Etat - Re´gion “Pays de la Loire” and INSERM. LA060368V
Figure 2. Evolution of surface roughness of the pHEMA hydrogel with hydration. Image of the disk surface in the dry state and after 30 and 240 min of hydration in saline. At the beginning of the study, the surface roughness of the dry polymer was due to stripes present on the mould, which progressively disappeared during hydration.
with the mould during polymerization. The roughness corresponds to the stripes present on the moulds used to prepare the disks.
(24) Peppas, N. In Biomaterials Science: An Introduction to Materials in Medicine; Ratner, B. D., Hoffman, A. S., Schoen, F. J., Lemons, J. E., Eds.; Elsevier Academic Press: London, 2004; pp 100-107. (25) Cross, S. S. Micron 1994, 25, 101-113. (26) Mandelbrot, B. W. H. Freeman: San Francisco, 1982. (27) Peitgen, H. O.; Saupe, D. Springer-Verlag: New York, 1988. (28) Blacher, S.; Henrioulle, N.; Brouers, F.; Sarychev, A. Acta Stereol. 1997, 16, 19-29. (29) Chappard, D.; Pascaretti-Grizon, F.; Gallois, Y.; Mercier, P.; Basle´, M. F.; Audran, M. Eur. J. Radiol. 2006, doi: 10.1016/j.ejrad.2005.12.033 (30) Carpinteri, A.; Ferrara, G.; Imperato, L. Eng. Fract. Mech. 1994, 48, 673-689. (31) Ng, S. H.; Fairbridge, C.; Kaye, B. H. Langmuir 1987, 3, 340-345. (32) Babadagli, T.; Develi, K. Theor. Appl. Fract. Mech. 2003, 39, 73-88. (33) Pentland, A. P. IEEE Trans. Pattern Anal. Mach. Intell. 1984, 6, 661674.
