pubs.acs.org/Langmuir © 2009 American Chemical Society
Impact of Nanoscale Roughness of Titanium Thin Film Surfaces on Bacterial Retention )
Elena P. Ivanova,*,† Vi Khanh Truong,† James Y. Wang,‡ Christopher C. Berndt,‡,§ Robert T. Jones, Iman I. Yusuf,^ Ian Peake,^ Heinrich W. Schmidt,^ Christopher Fluke,# David Barnes,# and Russell J. Crawford†
)
† Faculty Life and Social Sciences and ‡Faculty of Engineering and Industrial Sciences, IRIS, Swinburne University of Technology, P.O. Box 218, Hawthorn, Victoria, 3122, Australia, §State University of New York at Stony Brook, Stony Brook, New York 11794, Centre for Materials and Surface Science, Department of Physics, La Trobe University, Victoria 3086, Australia, ^School of Computer Science and Information Technology, RMIT University, GPO Box 2476 V, Melbourne, Victoria 3001, Australia, and #Centre for Astrophysics and Supercomputing, Swinburne University of Technology, P.O. Box 218, Hawthorn, Victoria, 3122 Australia
Received July 17, 2009. Revised Manuscript Received August 31, 2009 Two human pathogenic bacteria, Staphylococcus aureus CIP 68.5 and Pseudomonas aeruginosa ATCC 9025, were adsorbed onto surfaces containing Ti thin films of varying thickness to determine the extent to which nanoscale surface roughness influences the extent of bacterial attachment. A magnetron sputter thin film system was used to deposit titanium films with thicknesses of 3, 12, and 150 nm on glass substrata with corresponding surface roughness parameters of Rq 1.6, 1.2, and 0.7 nm (on a 4 μm 4 μm scanning area). The chemical composition, wettability, and surface architecture of titanium thin films were characterized using X-ray photoelectron spectroscopy, contact angle measurements, atomic force microscopy, three-dimensional interactive visualization, and statistical approximation of the topographic profiles. Investigation of the dynamic evolution of the Ti thin film topographic parameters indicated that three commonly used parameters, Ra, Rq, and Rmax, were insufficient to effectively characterize the nanoscale rough/smooth surfaces. Two additional parameters, Rskw and Rkur, which describe the statistical distributions of roughness character, were found to be useful for evaluating the surface architecture. Analysis of bacterial retention profiles indicated that bacteria responded differently to the surfaces on a scale of less than 1 nm change in the Ra and Rq Ti thin film surface roughness parameters by (i) an increased number of retained cells by a factor of 2-3, and (ii) an elevated level of secretion of extracellular polymeric substances.
Introduction Titanium, in different forms, is commonly used in indwelling devices including orthopedic and dental prostheses and cardiac valves, maxillofacial surgery and vascular stents, because of its high biocompatibility, low toxicity and high corrosion resistance.1,2 The use of these implants is diversifying and increasing, as is research into the use of titanium that has been subjected to surface modification to increase biocompatibility.3 It is well-documented that biofilm formation by human pathogenic bacteria on medical implants can be dramatic, leading to failure of the device, often resulting in the necessity to surgically remove the implant. This can be associated with systemic infection, loss of organ or limb function, amputation, or death.3-5 It has been determined that some bacteria residing in biofilms have a reduced susceptibility to antimicrobial agents compared to planktonic bacteria of the same species, measured by comparing the minimum biofilm eradication concentration (MBEC) with the *Corresponding author. Address: Swinburne University of Technology, P.O. Box 218, Hawthorn, Victoria 3122, Australia. Tel. þ613-92145137. Fax: þ613-92145921. E-mail:
[email protected]. (1) Whitehead, K. A.; Verran, J. Int. Biodeterior. Biodegrad. 2007, 60, 74–80. (2) Jeyachandran, Y. L.; Karunagaran, B.; Narayandass, S. K.; Mangalaraj, D.; Jenkins, T. E.; Martin, P. J. Mater. Sci. Eng., A 2006, 431, 277–284. (3) Hudson, M. C.; Ramp, W. K.; Frankenburg, K. P. FEMS Microbiol. Lett. 1999, 173, 279–284. (4) Ong, Y. L.; Razatos, A.; Georgiou, G.; Sharma, M. M. Langmuir 1999, 15, 2719–2725. (5) Diaz, C.; Cortizo, M. C.; Schilardi, P. L.; de Saravia, S. G. G.; de Mele, M. A. F. L. Mater. Res. 2007, 10, 11–14.
Langmuir 2010, 26(3), 1973–1982
minimum inhibitory concentration (MIC).6,7 Biofilm removal from indwelling devices is an issue of concern. An estimate by the Centre for Disease Control and Prevention (CDCP) indicated that 65% of human bacterial infections involve biofilm formation.8 For example, Troodle et al., showed biofilm formation by Staphylococcus aureus, Escherichia coli, and Pseudomonas aeruginosa on all samples collected from catheters removed from patients.9 S. aureus strains in particular are reported to be significant contributors to infections associated with orthopedic implants.8 Therefore, understanding the retention of human pathogens on host tissues and inanimate surfaces is an important step in efforts to reduce infection. There has been a number of conflicting research results published concerning the influence of substratum surface roughness on bacterial attachment, and whether the roughness has a positive, negative, or neutral impact on bacterial adhesion. In addition, there has been little work reported that directly links the surface architecture of the substrate to the extent of bacterial attachment.7,10,11 The aim of this study was to investigate whether (6) Donlan, R. M. Clin. Infect. Dis. 2001, 33, 1387–1392. (7) An, Y. H.; Friedman, R. J.; Draughn, R. A.; Smith, E. A.; Nicholson, J. H.; John, J. F. J. Microbiol. Methods 1995, 24, 29–40. (8) Tortora, G. J.; Funke, B. R.; Case, C. L. Microbiology: An Introduction, 8th ed.; Pearson Education, Inc., Benjamin Cummings: San Francisco, CA, 2004. (9) Troidle, L.; Finkelstein, F. Ann. Clin. Microbiol. Antimicrob. 2006, 5, 1–7. (10) Whitehead, K. A.; Colligon, J.; Verran, J. Colloids Surf., B: Biointerfaces 2005, 41, 129–138. (11) Mitik-Dineva, N.; Wang, J.; Truong, V. K.; Stoddart, P.; Malherbe, F.; Crawford, R. J.; Ivanova, E. P. Curr. Microbiol. 2008, 58, 268–273.
Published on Web 10/20/2009
DOI: 10.1021/la902623c
1973
Article
Ivanova et al.
nanoscale roughness of titanium thin film surfaces affects the nonspecific attachment and subsequent retention of two bacterial species, P. aeruginosa and S. aureus.
Experimental Section Titanium Thin Film Preparation. The titanium thin films of 3, 12, or 150 nm thickness (henceforth referred to as 3, 12, or 150 nm films) were prepared onto precleaned plain microscope slide substrates (Biolab Scientific, Ltd.) using a Kurt J Lesker CMS-18 magnetron sputtering thin film deposition system in direct current mode at an argon gas pressure of 4 mTorr and a power of 150 kW. The base pressure of the system is below 5 10-8 Torr. Glass slides were cleaned using 0.25% potassium hydroxide (KOH)), flushed with distilled water, soaked in 5 M HNO3 for 10 min, flushed with distilled water, soaked in 5 M KOH for 10 min, flushed with water, and finally blow-dried using 99.99% purity nitrogen gas. Commercial-grade 99.99% pure titanium (ASTM grade-2 titanium) was used to deposit titanium thin films of 3, 12, and 150 nm thickness. The deposition conditions used were as described elsewhere,12 and the thickness of the film was controlled by the period of sputtering, calculated from the calibrated deposition rate.12 Contact Angle Measurements. The contact angles of different solvents on titanium disks were measured using the sessile drop method.13-15 Three solvents, Milli-Q water, formamide (Sigma), and diidomethane (Sigma) were used. An FTA1000 (First Ten A˚ngstroms, Inc.) instrument was used to measure the contact angles at room temperature (ca. 23 C) in air. An average of at least five measurements was taken for each solvent and titanium surface. Each measurement of a particular contact angle was recorded in 50 images in 2 s with a Pelco model PCHM 575-4 camera, and the contact angle was determined as a result of images analyzed using the FTA Windows Mode 4 software. The average contact angle for each of the three solvents on each surface was used to calculate the surface free energy and its components, based on the Lewis acid/base method.13-15 Statistical data processing was performed using the SPSS 16.0 program (SPSS, Inc., Chicago, IL). Single independent group t-tests were performed to evaluate the consistency of surface roughness parameters. Titanium Thin Film Surface Characterization. A scanning probe microscope (SPM) (Solver P7LS, NT-MDT) was used to obtain images of the surface morphology and to quantitatively measure and analyze the surface roughness of metallic surfaces on the nanometer scale. The analysis was performed in the semicontact mode, which reduces the interaction between the tip and sample, thus avoiding the destructive action of lateral forces present in the contact mode. Carbon “whisker” type silicon cantilevers (NSC05, NT-MDT) with a spring constant of 11 N/m, tip radius of curvature of 10 nm, aspect ratio of 10:1, and resonance frequency of 150 kHz were used to obtain topographic resolution. Scanning was performed perpendicular to the axis of the cantilever at a rate of typically 1 Hz. Image processing of the raw topographical data was performed with first-order horizontal and vertical leveling, and the topography and surface profile of the samples were obtained simultaneously (as shown in Figures 1 and 2). In this way, the surface features of the samples were measured with a resolution of a fraction of a nanometer, and the surface roughness of the investigated areas could be statistically analyzed using standard instrument software (LS7-SPM v.8.58). For extended nanosurface characterization beyond algorithms built in to the atomic force microscopy (AFM) package, the image analysis software package Nano-SPAce Topography (NSPAT) (12) Wang, J. Y.; Ghantasala, M. K.; McLean, R. J. Thin Solid Films 2008, 517, 656–660. (13) Van Oss, C. J.; Good, R. J.; Chaudhury, M. K. J. Colloid Interface Sci. 1985, 111, 378–390. (14) Van Oss, C. J.; Good, R. J.; Chaudhury, M. K. Langmuir 1988, 4, 884–891. € (15) Oner, D.; McCarthy, T. J. Langmuir 2000, 16, 7777–7782.
1974 DOI: 10.1021/la902623c
was developed. Around this package the following experiment workflow was carried out: (i) acquisition of nanosurface AFM data together with metadata characterizing sputter density, film thickness, nanoscale grid resolution, and other parameters required either for the analysis itself or for the visualization and interpretation of results; (ii) export of the raw data acquired in the AFM experiments; (iii) reformatting and storage of the raw data with the metadata in a distributed file system; (iv) data analysis using the NSPAT package and native XGrid functionality on a parallel server cluster; and (v) visualization of the results. NSPAT has been implemented to run on a parallel Apple Mac OS 10.5 server cluster dedicated to biomedical image processing and visualization. Six high-end eight-core servers are available directly, and several hundred Macs in student laboratories can be accessed seamlessly in addition through the native Apple Xgrid protocol. The cluster is also connected through Globus16 to the broader worldwide grid computing infrastructure to serve other advanced medical image analysis. While the parallelization was not critical for the few data sets analyzed in this paper, it is significant in our project for processing and visualizing very large amounts of raw data meaningfully and, more importantly, refining our algorithms iteratively and repeatedly running them on an ever growing collection of data sets within a very short period of time. The main functions of NSPAT are these: Mean Plane. For all topographic parameters, the profile is calculated relative to the mean plane. The mean plane is defined as the statistical mean of the data grid. Maximum Height and Depth. Maximum height and depth (relative to the mean plane) are parameters used in several other functions. In addition, some of the functions work relative to a band defined around the mean plane. Skewness Rskw. Skewness is a measure of the symmetry of the height probability density function and can distinguish between wide valleys with narrow sharp peaks versus high plateaus with sharp deep valleys. For example, a Gaussian surface, having a symmetrical shape for the height distribution has a skewness of zero; a plateau honed surface with predominant plateau and sharp deep valleys would tend to have a negative skew, whereas a surface composed of a disproportionate number of sharp steep peaks will have positive skew.17 Kurtosis Rkur. Kurtosis represents a key feature in the probability density function of the height profile, the ‘peakedness’ of the profile. A Gaussian surface has a kurtosis value of 3; a surface that is centrally distributed has a kurtosis value greater than 3; a surface that has a well spread out distribution has a kurtosis value of less than 3.17 Peak Count Rp and Valley Count Rv. The peak and valley count is the number of discernible peaks (or valleys, respectively) relative to a narrow band around the base plane. Data above the band counts toward definite peaks (high), data below counts as definite valleys (low), and data within the band counts as uncertain. The band chosen depends on the properties of the experiment such film thickness. This means valleys and peaks entirely contained within the band are ignored in the peak count as height variations that are “too small to count”. A scanning trajectory of points entirely contained in the band and connecting a low point (definite valley) to a high point (definite peak) indicates an ascent to a peak; one that connects from a high to a low point is a descent from a peak. While the rough peak count may be sufficient for characterizing some aspects of the topography, it ignores, for example, spiral connectivity, where a single peak and a single valley may curl around each other and where unidirectional scans in any two orthogonal directions will recognize this topography as multiple peaks. An exact 3D peak count is (16) Foster, I.; Kesselman, C., The GRID: Blueprint for a New Computing Infrastructure; Morgan Kaufmann: San Francisco, CA, 1999. (17) Gadelmawla, E. S.; Koura, M. M.; Maksoud, T. M. A.; Elewa, I. M.; Soliman, H. H. J. Mater. Process. Technol. 2002, 123, 133–145.
Langmuir 2010, 26(3), 1973–1982
Ivanova et al.
Article
Figure 1. Typical two-dimensional (2D) AFM images and surface profiles of glass substrata and titanium thin film surfaces from approximately 4 μm 4 μm scanned areas. implemented using a parallel variant of the well-known and efficient maze flooding algorithm, treating inside-band data as maze wall points, distinguishing parallel flooding peaks above and valleys below by different colors (positive and negative numbers), and counting the corresponding number of colors needed. X-ray Photoelectron Spectroscopy (XPS) Analysis. The surface compositions of the titanium-coated glass slides were determined from X-ray photoelectron spectra using a Kratos Axis Ultra DLD spectrometer (Kratos Analytical Ltd., U.K.). The energy scale of the instrument was calibrated by measuring the Au 4f7/2 (Eb = 84.0 eV), Ag 3d5/2 (Eb = 368.3 eV), and Cu 2p3/2 (Eb = 932.7 eV) binding energies for pure metal foils. Langmuir 2010, 26(3), 1973–1982
Spectra were recorded while irradiating the samples with a monochromated Al KR source (hν =1486.6 eV) operating at 150 W. The analysis area was approximately 300 700 μm2. Elements present on the surface of each sample were identified from survey spectra recorded over the energy range 0-1400 eV at intervals of 1 eV and a pass energy of 160 eV. High-resolution spectra were recorded for selected photoelectron peaks (C 1s, O 1s, N 1s, Ti 2p, and Si 2p) at intervals of 0.1 eV and a pass energy of 20 eV. Bacterial Strains. The bacteria used in this study were S. aureus CIP 68.5 and P. aeruginosa ATCC 9025. Bacterial strains were obtained from American Type Culture Collection DOI: 10.1021/la902623c
1975
Article
Ivanova et al.
Figure 2. Three-dimensional (3D) projections of typical AFM images and surface profiles of glass substrata and titanium thin film surfaces from approximately 4 μm 4 μm (bottom) scanned areas. Readers using version 8.0 or higher of Acrobat Reader can enable interactive, 3D views of the data by clicking on the figure panels. Once enabled, 3D mode allows the reader to rotate and zoom the view using the mouse. (ATCC, USA) and Culture Collection of the Institute Pasteur (CIP, France). Bacterial strain stocks were prepared in 20% glycerol nutrient broth (Merck) and stored at -80 C. Both strains were cultured on nutrient agar (Oxoid) and nutrient broth (Oxoid) at room temperature (ca. 22 C). Cellular Surface Charge Measurements. Bacterial cell surface hydrophobicity was evaluated from contact angle measurements on lawns of bacteria using the sessile drop method as described elsewhere.18 In brief, bacterial cells in a buffer (OD(600) = 0.3) were deposited on cellulose acetate membrane filters (Sartorius, 0.2 μm). The wet filters were air-dried at ambient temperature (ca. 22 C) for approximately 30-40 min to attain a “plateau state”. Zeta potential measurements of both strains were obtained by measuring the electrophoretic mobility (EPM) using a standard protocol described elsewhere.19 The EPM was measured as a function of ionic strength by microelectrophoresis using a zeta potential analyzer (ZetaPALS, Brookhaven Instruments Corp, Holtsville, NY). The bacterial cell suspension was freshly (18) Mitik-Dineva, N.; Wang, J.; Mocanasu, R. C.; Stoddart, P. R.; Crawford, R. J.; Ivanova, E. P. Biotech J 2008, 3, 536–544. (19) de Kerchove, A. J.; Elimelech, M. Langmuir 2005, 21, 6462–6472.
1976 DOI: 10.1021/la902623c
prepared before the measurement. All measurements were executed in triplicate, and for each sample the final EPM quoted represented the average of five successive ZetaPALS readings, each of which consisted of 14 cycles per run. All data were processed using software that employed the Smoluchowski equation.19,20 Bacterial Growth and Sample Preparation. Prior to each experiment, a fresh bacterial suspension was prepared for each of the strains grown overnight in 100 mL of nutrient broth (Oxoid) (in 0.5 L Erlenmeyer flasks) at 37 C with shaking (120 rpm). Bacterial cells were collected at the logarithmic stage of growth as confirmed by growth curves (data not shown). As cell densities may vary, the cell density of each strain was adjusted to OD(600) = 0.3 to obtain samples that possessed a similar number of cells in each sample. A hemocytometer was used to quantify cell numbers in the adjusted bacterial suspensions before attachment experiments according to the method described by Mather and Roberts.21 An aliquot of 5 mL of bacterial suspension was added (20) Eboigbodin, K. E.; Newton, J. R. A.; Routh, A. F.; Biggs, C. A. Langmuir 2005, 21, 12315–12319. (21) Mather, J.; Roberts, P. Introduction to Cell and Tissue Culture: Theory and Technique; Plenum Press: New York, 1998.
Langmuir 2010, 26(3), 1973–1982
Ivanova et al.
Article
in a sterile Petri dish with samples of glass slides that had been coated with titanium thin films. These were incubated for 18 h at room temperature (ca. 22 C). Sterile nutrient broth (5 mL) was used as a negative control. Samples were handled under sterile conditions until just prior to imaging. After incubation, glass slides with titanium thin films were gently washed with copious amounts of sterilized nanopure H2O (18.2 M Ω cm-1) to remove nonattached cells and left to dry at room temperature for 45 min at 55% humidity. This procedure allowed the bacterial cells to maintain at least a semihydrated state, as confirmed by their morphological appearance, and allowed all imaging experiments to be performed under identical conditions.
Visualization and Quantification of Viable Cells and Extracellular Polymeric Substances (EPSs). Two dyes were
used simultaneously; the first to monitor the level of production of extracellular substances, and the second to enable the visualization of viable bacterial cells retained on the surfaces. An aliquot of Concanavalin Alexa Fluor 488 (Molecular Probes Inc.) dye was added to the suspension, and incubated for 30 min to allow the dye to diffuse thoroughly throughout the sample. This dye selectively binds to R-mannopyranosyl and R-glucopyranosyl residues in EPSs.22 The dye stock solution was prepared by dissolving 5 mg in 5 mL of 0.1 M sodium bicarbonate at pH 8.3 and stored at 20 C. A ratio of 20 μL dye to 100 μL suspension was used to stain any EPS produced. After the initial 30 min incubation period, a Vybrant CFDA SE Cell Tracer Kit (Molecular Probe, Inc.) was used to stain viable cells. Working solutions were prepared by diluting 1 μL of a 10 mM stock solution to 1000 μL in 10 mM phosphate-buffered saline (PBS) solution (pH 7), followed by warming to 37 C. A volume of 20 μL working solution was then added to 100 μL of suspension and incubated at 37 C for 15 min. This dye concentration was found to be sufficient to visualize labeled cells. After incubation, the samples were resuspended in fresh medium at 37 C for 30 min. The samples were then washed with sterilized nanopure H2O (18.2 M Ω cm-1), left to dry for a few hours at room temperature (ca. 22 C, humidity 55%) without additional fixation to prevent the deformation of the cells and analyzed using confocal scanning laser microscopy (CSLM). The CSLM used was an Olympus FluoView FV1000 Spectroscopic Confocal System, which includes an inverted Microscope System OLYMPUS IX81 (20, 40 (oil), 100 (oil) UIS objectives) and operates using multi Ar and HeNe lasers (458, 488, 515, 543, and 633 nm). The system is equipped with a transmitted light differential interference contract attachment and a charge-coupled device (CCD) camera (Cool View FDI). Digital image analysis of the CSLM optical images was performed to quantify the volume of presumed EPS and viable cells. FV10-ASW 1.6 software was used to measure and analyze the intensity integration (counts per second, cps) of a fluorescent dye to determine the amount of each substance present.23,24 The total intensity of a biofilm was considered to be the total intensity of the EPS and viable cells. All samples were scanned at four locations of 126.72 μm 126.72 μm area each. Scanning Electron Microscopy (SEM). In all SEM experiments, titanium discs with adsorbed bacteria were initially sputter-coated with 20 nm gold thin films using a Dynavac CS300 according to a procedure developed previously.25,26 High-resolution images of titanium thin films with the retained bacterial cells were taken using a field-emission SEM (FESEM; ZEISS SUPRA 40 VP) at 3 kV at 1000, 5000 and 20 000 magnification. (22) Goldstein, I. J.; Hollerman, C. E.; Smith, E. E. Biochemistry 1964, 4, 876– 883. (23) Feller, B. E.; Kellis, J. T., Jr; Casc~ao-Pereira, L. G.; Knoll, W.; Robertson, C. R.; Frank, C. W. Langmuir 2008, 24, 12303–12311. (24) Neu, T. R.; Lawrence, J. R. FEMS Microbiol. Ecol. 1997, 24, 11–25. (25) Mitik-Dineva, N.; Wang, J.; Truong, V. K.; Stoddart, P.; Malherbe, F.; Crawford, R. J.; Ivanova, E. P. Curr. Microbiol. 2009, 58, 268–273. (26) Truong, V. K.; Rundell, S.; Lapovok, R.; Estrin, Y.; Wang, J. Y.; Berndt, C. C.; Barnes, D. G.; Fluke, C. J.; Crawford, R. J.; Ivanova, E. P. Appl. Microbiol. Biotechnol. 2009, 83, 925–937.
Langmuir 2010, 26(3), 1973–1982
Figure 3. XPS wide spectra of titanium thin film surfaces of 150, 12, and 3 nm and glass substratum, 0 nm. Images at 1000 and 5000 magnification were used to calculate the number of bacteria adhering to the titanium surfaces. The results were statistically analyzed. Reconstruction of Interactive 3D Images. Interactive 3D visualization of the titanium surfaces was undertaken with a custom C-code and the S2PLOT graphics library.27,28 The input data files were in NT-MDT format and were read into the viewing tool (mdtview) using a modification of the nt-mdt module of Gwyddion by David Necas and Petr Klapetek (http://gwyddion. net/, version 2.12). NT-MDT files were converted into a 3D surface, colored according to height, and displayed with the S2PLOT s2surpa function. Visualizations were exported from mdtview to an intermediate VRML format, with textures for axis labels in TGA format. Textures were converted to PNG format, and the VRML model was imported into Adobe Acrobat 3D version 8 to create an interactive figure, using the approach described by Barnes & Fluke.27 JavasScript commands were used to provide additional functionality. If viewing this paper in electronic format, the interactive Figure 2 can be viewed by mouse clicking on the four panels, provided Adobe Reader Version 8.0 or higher is being used. This opens a window where the surface can be examined interactively using the mouse to control the camera orientation and zoom level.
Results Titanium Surface Characterization. The XPS and AFM analyses and contact angle measurements were performed for Ti thin film surfaces to evaluate elemental composition, surface topography, and the surface wettability of the Ti thin films. The XPS elemental analysis was carried out for both Ti thin films and glass substrata to assess the surface elemental composition, chemical functionality, and Ti coverage. The XPS elemental analysis showed that titanium and oxygen were the most abundant elements, hence indicating that Ti was present as TiO2 (27) Barnes, D.; Fluke, C. New Astron. 2008, 13, 599–605. (28) Barnes, D.; Fluke, C.; Bourke, P.; Parry, O. Publ. Astron. Soc. Aust. 2006, 23, 82–93.
DOI: 10.1021/la902623c
1977
Article
Ivanova et al. Table 1. Atomic Fractionsa of Elementsb Detected on the Surface of Each Sample by XPS
Ti film thickness [nm]
Mg
Zn
Na
O
Ti
Ca
N
K
C
Si
Al
c
5.9 55.5
