Synthesis of biocompatible surfaces by nanotechnology methods

SYNTHESIS OF BIOCOMPATIBLE SURFACES BY NANOTECHNOLOGY METHODS 697Albumin(1)Fibrinogen(2)Carbon cluster(3)K ads K desK ads K desK ads K desK ads Kdes1of albumin molecules in competition with fibrinogenmolecules.As can be seen from Fig. 1a, both albumin andfibrinogen molecules adsorb on the initial polymersurface. Fibrinogen molecules, which are 4–5 timesgreater than albumin molecules (~50 nm against~9 nm, respectively), are, for thermodynamic reasons,almost irreversibly adsorbed so that they stick to thesubstrate with their maximum surface area. In contrast,part of the modified surface (Fig. 1b) is occupiedby carbon clusters so that fibrinogen molecules have(a)(b)Fig. 1. Schematic diagram illustrating the adsorption–desorptionof (1) albumin and (2) fibrinogen molecules on the(a) initial and (b) modified polymer surface containingcarbon clusters (3). Arrows indicate the relative rate constantsof adsorption (K ads ) and desorption (K des ) for albuminand fibrinogen.32no area to settle and stick, whereas significantlysmaller albumin molecules can readily adsorb on carbon-freesurface sites. Practical implementation ofthis adsorption scheme would improve the hemocompatibilityof polymer surfaces and eliminate the formationof thrombuses.In order to provide conditions for the selective(competitive) adsorption of albumin and fibrinogenmolecules as depicted in Fig. 1, it is necessary to producethe nanostructurization of the substrate surfaceby carbon clusters so that the dimensions of open(adsorption-accessible) surface regions between theclusters are be on the order of several units to severaldozen nanometers.As is known, the ion–plasma deposition of nanostructuralcarbon coatings in a pulsed rather than continuousregime makes it possible to quite simply controlthe degree of carbon supersaturation (Δg) in thegas phase and, hence, the mechanism of initial growthstages [5]. Indeed, as low Δg values, the carbon filmgrowth may proceed via the layered or two-dimensional(2D) mechanism of nucleation and growth. Theprobability of this process is characterized by the isobaric–isothermalpotential ΔG 2 ~ 1/Δg. As the Δgvalue increases, the mechanism can change from 2Dto 3D, in which case ΔG 3 ~ 1/Δg 2 (power dependence)and the conditions are more nonequilibrium. This circumstanceimplies a greater possibility that the carbon-coatedsurface morphology will be influenced bythe varying conditions of the technological process. Inaddition, the pulsed regime facilitates the removal ofheat from the substrate surface, since the period oftime between pulses is an order of magnitude greater(as characterized by the off/on ratio) than the pulseduration. The parameters of this technological processcan be optimized by studying the surface morphologyand introducing changes in the process so as to obtainthe desired surface relief.Figure 2 schematically illustrates the dynamics offilm nucleation and growth according to the Volmer–Weber (island growth) mechanism. At the stage of car-(a) (b) (c) (d)Fig. 2. Schematic diagram of the island nucleation and film growth mechanism: (a) nucleation and growth of carbon clusterswithout mutual interactions; (b) lateral growth and island formation; (c) coalesce of islands at a sufficiently large surface coverageand the formation of a network with electrically continuous structure capable of conducting electric current (percolation threshold);(d) formation of a continuous film.NANOTECHNOLOGIES IN RUSSIA Vol. 5 Nos. 9–10 2010

698ALEKHIN et al.bon coating formation, it is necessary to determine theso-called percolation threshold (see Fig. 2c), wherebya carbon network that forms an electrically continuousstructure capable of conducting electric current isformed, typically when the substrate is covered up to70–80%. Why is determining the percolation thresholdso important? Because this point at the stage ofcarbon deposition corresponds to the formation of amosaic carbon nanostructure with cluster dimensionsin a nanometer range. This situation leads to a changein the physicochemical and biomedical properties,thus making it possible to control the hemocompatibilityof the carbon-modified surface [7, 8]. It is highlyprobable that the percolation threshold determines thefinal stage of formation of a hemocompatible mosaiccluster structure on the substrate surface. In this study,the percolation threshold was determined as a functionof the number N and the repetition frequency f ofpulses of the carbon plasma generator. The thicknessof the deposited carbon coating varied within 0.3–15 nm. The proposed technology was experimentallyimplemented and patented [6].DETERMINING THE TECHNOLOGICALPARAMETERS AND PHYSICALAND BIOMEDICAL PROPERTIESOF PLASMA-DEPOSITED CARBONNANOSTRUCTURESWe have experimentally determined the optimumtechnological parameters for the process of LDPE andPU modification by carbon clusters to reach a percolationthreshold as the pulse repetition frequency f =1 Hz and their number N = 50, which corresponded toan effective carbon coating thickness of ~7.5 nm. Thesurface morphology of samples was studied by scanningelectron microscopy (SEM) and scanning probemicroscopy (SPM).The SEM measurements were performed on aCamscan 4 (Cambridge Instruments) microscopewith a typical magnification of ×1000–5000, whichcorresponds to the size of visible fields on a level ofseveral tens of microns. By studying the characteristicfeatures of morphology, it is possible to evaluate theoptimum technological parameters, introduce thenecessary corrections, and obtain the surface reliefnecessary for further research. It was established thatthe relief of modified polymer surfaces was significantlydependent on the number N of carbon plasmapulses and their repetition frequency f. The dynamicsof variation of the surface relief exhibited the followingtrends. At a small amount of the deposited material(N = 2–10), the necessary surface morphology (correspondingto Fig. 2c) is not formed. In this case, thecarbon-modified polymer surface has a globular structure,i.e., exhibits separate clusters with dimensionswithin 100–200 nm. Large values of the pulse repetitionfrequency and number ( f ≥ 10 Hz, N ≥ 100) correspondedto the formation of structures representingeither fibrils (3000–4000 nm long) or continuouschains. The most favorable regimes were those with f =0.3–1.0 Hz and N = 30–50.Direct evidence for the formation of a carbon clusterstructure on the LDPE surface was provided by theresults of SPM investigations in air at room temperature.These measurements were performed on aSolver TM Pro M (NT-MDT Company) instrumentoperating in a tapping mode with 2D and 3D imaging.The cantilever had an elasticity coefficient of about20 N/m and a resonance frequency of 131.851 kHz;the point probe had a tip radius of several nanometers.The minimum spatial resolution of the microscope, asevaluated by the size of the minimum visible elementsof the sample surface, was about 10 nm.Figures 3 and 4 present typical 2D images obtainedby an SPM examination of LDPE and PU modified atvarious repetition frequencies ( f = 0.1–1.0 Hz) andnumbers (N = 30–50) of carbon plasma pulses. As thepulse repetition frequency was increased from 0.1 to1.0 Hz and/or their number was increased from 2 to50, the size of the deposited carbon clusters or theiraggregates, as well as the degree of surface coverage bycarbon, increased to 70–80%. This led to the formationof a mosaic carbon network with an electricallyconducting structure.Thus, based on our results, it was concluded thatthe optimum pulse repetition frequency and carbondeposition velocity for the formation of the desiredcluster structure on the polymer surface are 0.3–1 Hzand 0.1–0.2 nm/s, respectively.Additional information on the structure of modifiedcarbon coatings on polymer substrates wasobtained by mathematically processing their Ramanspectra [9]. An analysis of parameters such as theRaman shift and the heights and widths of model spectrumcomponents (Fig. 5), which were determinedusing the results of Tamba et al. [10], allowed us tomake the following conclusions concerning the structureof layers formed on polymer substrates. It was evidentthat both sp 2 and sp 3 hybridization is present incarbon deposits. According to the general principles ofthermodynamics, structural elements with the sp 2 typeof hybridization form small-size clusters. Of all thepossible clusters of this kind, the most probable aresix-member rings, which can be either separate orconnected into sets of several rings. These rings maypossess angular misorientation and can be shaped likeirregular hexagons. These clusters are bound by elementsof the sp 3 hybridization and can be chaoticallyoriented relative to each other.Medico-technical investigations were performed atthe Research Institute of Transplantology and ArtificialOrgans (Ministry of Public Health of the RussianFederation, Moscow), which involved (i) an analysisof the thrombocyte adhesion with allowance for theirmorphological characteristics and (ii) a determinationNANOTECHNOLOGIES IN RUSSIA Vol. 5 Nos. 9–10 2010

SYNTHESIS OF BIOCOMPATIBLE SURFACES BY NANOTECHNOLOGY METHODS 699(a)40(a)55nm4003002001000 100 200 300(b)400 nm4003002001000 100 200 300(c)400 nm4000nm400nm40nmnm0.80.60.40.20 0.2 0.4 0.6(b)0.80.80.60.40.20 0.2 0.4 0.6(c)0.80.8nm045nm0453002001000 100 200 300 400nm0nmFig. 3. 2D images obtained by SPM examination of LDPEsurface modified by N = 30–50 pulses of carbon plasma ata repetition frequency of f = (a) 0.1, (b) 0.3, and (c) 1.0 Hz.nm0.60.40.2nm00 0.2 0.4 0.6 0.8nmFig. 4. 2D images obtained by a SPM examination of thePU surface modified by N = 30–50 pulses of carbonplasma at a repetition frequency of f = (a) 0.1, (b) 0.3, and(c) 1.0 Hz.of the degree of activation of the complement systembased on the general hemolytic activity of the complementin human blood serum. The results showed (seetable) that the proposed modification of the polymersurface by carbon leads to a decrease in the thrombocyte(platelet) adhesion and a drop in the rate constantK ind of induced activation of the complementsystem. These characteristic were calculated using thetimes of the half-lysis of sensitized goat erythrocytesby human blood serum before and after incubationwith a sample. The surface is classified as hemocompatible,provided that its K ind does not exceed 1.5. Itshould be noted that the initial polymer surfaces hadK ind = 5.1 and 5.5 for LDPE and PU, respectively. Ananalysis of data in the table shows that the proposedmodification significantly improves the hemocompatibilityof the polymer surface at an affective carbonfilm thickness of 5–12 nm and cluster dimensionswithin 50–120 nm.NANOTECHNOLOGIES IN RUSSIA Vol. 5 Nos. 9–10 2010

700ALEKHIN et al.Intensite, rel. units30201008001000 1200 1400 1600 1800Wavenumber, cm −1Fig. 5. Analysis of the Raman spectrum of a carbon coatingshowing general line envelope and its components correspondingto sp 2 hybridization, sp 3 hybridization, and misorientedstates, respectively.NANOSTRUCTURAL MODIFICATIONOF POLY(METHYL METHACRYLATE)SURFACE BY A COMBINATIONOF PLASMACHEMICAL TREATMENTAND VACUUM ULTRAVIOLET IRRADIATIONPoly(methyl methacrylate) (PMMA) is now widelyused in practical transplantology as a base material forvarious implants [11].Plasmachemical Processing of PMMA SurfacePreviously, Vasilets et al. [12, 13] showed that plasmachemicalprocessing and vacuum ultraviolet (VUV)irradiation modify the properties of near-surface layersof medical polymers. The destruction of polymers andthe formation of active chemical radicals and speciesunder these conditions make it possible to control thebiocompatibility of the polymers. In particular, themodification of the near-surface layers of a polymerleads to a change in its nanoroughness. In this context,we have studied the effect of plasmachemical treatmentin an oxygen-containing medium at a small partialpressure (~2 Pa) followed by VUV irradiation onthe PMMA surface roughness in a nanometer range.The initial PMMA film with a molecular mass of8 × 10 4 and a thickness of about 0.8 μm was depositedby centrifuging onto a polished Si(100) single crystalsurface (Fig. 6a) and then nanostructured by beingtreated for 20 s in an RF (13.56 MHz) plasma. Theparameters of the surface roughness and geometry ofthe surface features (nanohills and nanograins) onPMMA films were studied by SPM in the atomic forcemicroscopy (AFM) mode on a sample surface area of1.4 × 1.4 μm 2 . The dimensions of the features weredetermined by the computer-aided recognition of thesurface relief, which was performed using the methodof feature-oriented scanning in the virtual regime[14].It was established that the plasmachemical treatmentof a PMMA film led to the formation of flat nanograins(Fig. 6b) with average lateral dimensions ofabout 66 nm, an average height of about 1.8 nm, andan average distance of about 104 nm between adjacentgrains. The number of nanograins per unit surface areawas about 122 μm –2 . The proposed nanostructuretechnology ensures that these results can be easilyreproduced.The formation of surface nanograins can beexplained by local changes in the PMMA film density.These local changes are related to the presence ofinhomogeneities in most polymers, primarily globules/grainsand lamellae, as well as coils, nodes, twists,entanglements, etc. of molecular chains, which agreeswith the results of other investigations [15, 16]. In thecourse ofPMMA treatment in oxygen-containing plasma,the polymer is subject to significant chemical modification,including the formation of polar functionalgroups [17, 128], carbonate (O 2 C=O), and carbonyl(C=O) groups. In addition, there are significantchanges in the binding energies of carbon atoms invarious molecular groups. The content of oxygen inthe near-surface layer exhibits a significant increaseduring the first seconds of treatment; then the growthslows down and eventually oxygen rather slowly penetratesin depth of the material [17].Treatment of PMMA Film Surfaceby VUV RadiationPlasma-treated PMMA samples were exposed tothe radiation of krypton lamps with a characteristicwavelength of λ = 123.6 nm in vacuum at a residualThe dependence that the cluster size, the rate constant (K ads ) of blood plasma protein adsorption, and the rate constant K indof induced activation of the complement system has on the effective thickness of the carbon-containing layer on LDPE surfaceCarbon filmthickness, nmClustersize, nmK ads , ml/(mg s)K indLDPE Vitur LDPE Vitur12 Surface coverage >80% 0.25 0.1 4.3 ± 0.5 3.5 ± 0.4NANOTECHNOLOGIES IN RUSSIA Vol. 5 Nos. 9–10 2010

SYNTHESIS OF BIOCOMPATIBLE SURFACES BY NANOTECHNOLOGY METHODS 701nm12001000800600400200(a)0 200 400 600 800 1000 1200nmnm2.22.01.81.61.41.21.00.80.60.40.2012001000800600400200(b)0 200 400 600 800 1000 1200nm87654nm3210Fig. 6. AFM images of PMMA films: (a) initial film on silicon substrate; (b) nanostructured surface upon treatment in an oxygencontainingRF plasma.pressure of 2 and 100 Pa for various periods of time(0.5, 1, 2, 5, 10, and 20 min) on a setup described indetail elsewhere [3].The photon energy at λ = 123.6 is about 10 eV,which is large enough to excite polymer molecules andbreak chemical bonds such as C–C, C–H, C–CH 2 ,and C–O [15]. Photolysis leads to the formation oflow-molecular-mass gaseous products such as CO,CO 2 , H 2 , H 2 O, and CH 4 , as well as to high-molecularmassgaseous products such as methyl formate(HCOOCH 3 ), formaldehyde (CH 2 =O), methanol(CH 3 –OH), and methyl methacrylate[CH 2 =C(CH 3 )–COOCH 3 ] [19]. In addition, exposureto VUV leads to the formation of intermolecularcross-links. Thus, the VUV-induced smoothing ofPMMA nanorelief is due to the joint action of severalprocesses, including the photoetching of the polymer,the redeposition of the gaseous products of photolysis,and the formation of intermolecular cross-links. Thedegree of smoothing (at a fixed radiation intensity) isdetermined primarily by the treatment duration.An analysis of our results showed that even 2-minVUV irradiation produced clearly detectable smoothingof the roughnesses on the nanostructured PMMAsurface. After 10-min exposure, the average roughnesssize decreased by a factor of 2.6–3; the average lateralsize of nanograins was reduced by half, and their averageheight decreased by a factor of 15–18. Thus, 10-min VUV treatment rendered the PMMA film surfacepractically smooth.MODIFICATION OF PMMA SURFACEBY SYNCHROTRON RADIATIONIt was also of interest to study the modification ofnear-surface layers of polymers by irradiation in thehard VUV (λ = 50–100 nm) and soft X-ray (λ = 10–50 nm) ranges. The experiments on polymer modificationin the short-wavelength spectral range wereperformed using a small synchrotron ring of the KurchatovSource of Synchrotron Radiation (Moscow).However, since it is still a difficult task to separate thecoherent collimated beams of various wavelengthsfrom the entire synchrotron spectrum, the samples ofPMMA films were exposed to a beam of “white” synchrotronradiation (SR). The white beam representedSR in the entire spectrum with λ = 5–400 nm.The irradiation of PMMA by high-energy photonsof the white SR beam leads to the formation of radicals(via the Norrish type I reaction) by rupturing chemicalbonds (C–C, C–H, C=O) with different energies andforming simple products such H 2 , CO, CO 2 , and CH 4 .At the same time, long-wavelength photons producethe reverse effect (cross linking). As a result, exposureto white SR is accompanied by the destruction ofPMMA and the subsequent cross linking of the products.Figure 7 shows AFM images of the surface relief ofPMMA films, which show evidence for self-organizationand a decrease in the size of nanostructured particles.An explanation of this phenomenon can bebased on the fact that polymers containing tertiarycarbon atoms and possessing low enthalpy of polymerizationare subject to depolymerization (thermodestruction).In addition, this is accompanied by UVradiation-inducedphotodestruction in the presence ofC=O, which decreases the strength of bonds in thepolymer backbone. Both processes lead to polymerdestruction with a quantitative yield of monomers[15]. Thus the exposure of PMMA to the white SRbeam leads to the appearance of structured monomerson the sample surface.NANOTECHNOLOGIES IN RUSSIA Vol. 5 Nos. 9–10 2010

702ALEKHIN et al.µmµm4.54.03.53.02.52.01.51.00.54.54.03.53.02.52.01.51.00.50 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5(b)0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5(c)nm4.5124.0103.53.082.562.01.541.020.500 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5μmFig. 7. AFM images of PMMA films exposed to white SRbeam for (a) 1 min, (b) 5 min, and (c) 17 min.µm(a)In order to check for the biocompatibility of radiation-modifiedsurfaces, we have studied their hydrophilic–hydrophobicbalance characteristics. Figure 86050403020100nm20151050nmMSR, nm76100905804WettingVUV 703SR6021Roughness 50400 5 10 15 20 25t, minFig. 8. Plots of the mean-square roughness (MSR) size and(wetting) contact angle on PMMA surface versus VUV andSR exposure duration t.Contact angle, degshows plots of the mean-square roughness (MSR) sizeand (wetting) contact angle on the PMMA surfaceversus VUV and SR exposure duration. As can be seen,the MSR ceases to depend on the exposure time afterabout 10-min irradiation and is determined by theparameters (wavelength and intensity) of the radiationsource. In the case of VUV, the contact angle change iscorrelated with the MSR variation on the PMMA surface.Using the exposure to SR, it is possible to simultaneouslymodify both the surface topography (roughness)and its chemical composition by varying only theexposure time. Note that the surface wetting dependsmostly on the chemical composition. These resultsindicate that SR affects the polymer surface roughnessto a much greater extent than VUV radiation does.CREATION OF A BIOACTIVE SURFACEWITH ACCELERATED OSTEOINTEGRATIONABILITY BY THE ATOMIC LAYERDEPOSITION METHODMost dental implants throughout the world aremade of commercially available Grade 4 titanium.The most favorable type of interaction between theimplant and natural bone tissue is commonly acceptedto be osteointegration, which involves a complex ofphysiological reactions directly dependent on the surfacemorphology and chemical composition of theimplant [20].A developed microrelief on the titanium surface withroughness R a ~ 1.4–1.7 is conventionally formed by twomethods: (i) sandblasting by corundum (Al 2 O 3 ) particleswith dimensions ranging from several tens to severalhundred of microns and (ii) etching with variousappropriate acid mixtures [21]. In addition to obeyingmicrorelief requirements, the implant surface must bewettable (hydrophilic), which is ensured by giving it anecessary chemical state [22].It was recently demonstrated [23] that modifyingthe surface of titanium by anatase—one of the mostNANOTECHNOLOGIES IN RUSSIA Vol. 5 Nos. 9–10 2010

SYNTHESIS OF BIOCOMPATIBLE SURFACES BY NANOTECHNOLOGY METHODS 703μm9.58.57.56.55.54.53.52.51.50.5−0.5−1.5−2.5−3.5−4.5−5.5−6.5500600 700 800 900 1000μmFig. 9. Roughness profile of the titanium surface upon sand-blasting and chemical etchingwidely occurring forms of titanium dioxide—leads toa significant improvement of the bioactivity ofimplants. In view of the above data, it was decided tocreate a new combined technology for the formationof biocompatible surfaces of titanium for medicalapplications.Developing Integrated Technologyof Implant Surface ModificationThe samples of titanium were washed in an ultrasonicbath with isopropyl alcohol (special puritygrade) at 50°C; then they were rinsed in ultrahighpuritywater and etched in various inorganic acids. Thequantitative compositions of etchants and the treatmentconditions and duration were experimentallyselected with respect to two criteria: (i) obtaining amicrorelief with MSR ~ 2–3 μm and (ii) ensuring acontact angle for water within 0°–5°.The topology of the etched titanium surface wascharacterized using an Alpha-Step prophilometer(Tencor Instruments), which showed that etched samplespossess a developed relief with a maximum roughnessheight of Z max = 13.4 μm and an average roughnessof R a = 2.61 μm (Fig. 9).Figure 10 shows the influence that the etching timehas on the surface topography and contact angle variation,which was studied using a KSV Instrumentssetup. The insets show SEM images of the sample surfaceupon etching for various periods of times. As canbe seen, a minimum contact angle (~4°) and mostdeveloped surface microrelief are observed upon 30-setching.SEM data confirmed the facts that (i) the unevenedges of surface roughnesses are smoothed by acidetching and (ii) active centers of selective etchingappear on the metal surface, which lead to the formationof a fine porous structure. The surface porositywas studied by SPM on a Solver Pro M (NT-MDTCompany) instrument operating in a tapping mode.The maximum probed area size was 11.5 × 11.5 μm.the lateral pore dimensions were within 2–4 μm(Fig. 11).However, the etching also led to a significant scatterof contact angles, which varied within 5°–90° in someregions. The surface chemical composition of thesesamples was studied by Auger electron spectroscopy(AES) (Perkin-Elmer AES instrument). The results ofthis analysis showed that the surface of samples exhibitinghydrophobic properties upon etching was characterizedby the presence of Ti–C bonds, which wasevidence for the presence of carbon on the surface oftitanium (Fig. 12). It was an increased content ofhydrocarbons on the metal surface which impartedhydrophobic properties to the material surface.Thus, the proposed method allowed a developedmicrorelief to be obtained on the surface of titanium.However, the formation of a carbidelike layer on theetched surface hindered a stable hydrophilic state overthe entire surface of a titanium sample (Fig. 12). Forthis reason it was decided to stabilize the chemicalcomposition of a titanium surface with developedmicrorelief using atomic layer deposition (ALD).NANOTECHNOLOGIES IN RUSSIA Vol. 5 Nos. 9–10 2010

704ALEKHIN et al.3 μm 3 μm 3 μmWetting angle, deg1086420 10 20 30 40 50 60 70Etching time, sFig. 10. Plot of the contact angle versus etching time. Insets show SEM images of the sample surface morphology at the indicatedpoints.(a)μmμm101.61.4861.21.00.8420.60.40.200 2 4 6 8 10 μm(b)μm1.60.80108642106840 2Fig. 11. (a) 2D and (b) 3D images obtained by AFM of atitanium surface etched in an acid mixture.µmµmWhy did we select the ALD method? Because itoffers the following unique possibilities [24–26]:(i) Obtaining regular, highly stoichiometric soliddeposits with the desired crystal structure on a substrateat low temperatures (300–500 K).(ii) Performing phase transformations bypassinghigh energy barriers related to nucleation, thus ensuringa conformal coating within 0.5–1.0 nm.One distinctive feature of the ALD technique is thepulsed supply of reactants to the substrate surface.Between these pulses, the reactor is either purged withan inert gas or evacuated. Correctly choosing the processparameters yields conditions where only a chemisorbedlayer of one reactant is left on the substrate toreact with another reactant that is supplied with thenext pulse. This process type is sometimes referred toas self-saturation or self-control [27]. Multiplyrepeated reaction cycles make the layer-by-layergrowth of a thin film of a preset composition possible,the total thickness of which is controlled by merelychanging the number n of repeated reaction cycles.In this study we used a Sunale-R150 Model verticalACO-type reactor (Picosum Oy) in which titaniumduioxide was deposited using ethoxytitanium [97%Ti(OC 2 H 5 ) 4 ] and water. Since the saturated vaporpressure of Ti(OC 2 H 5 ) 4 is low, it was supplied from asource heated to 150°C. The substrate temperatureduring deposition was 250°C.An ellipsometric analysis of these titanium dioxidefilms showed that their refractive index was 2.48 andthe standard deviation of film thickness relative to thesubstrate surface was within 1.8%. The deposition rateNANOTECHNOLOGIES IN RUSSIA Vol. 5 Nos. 9–10 2010

SYNTHESIS OF BIOCOMPATIBLE SURFACES BY NANOTECHNOLOGY METHODS 705Min:−1664Max:1308dN(E)Ti−C202 214 226 238 250 262Energy, eV274 286Fig. 12. AES spectrum of the acid-etched surface of titanium in the region of poor wetting (where the contact angle varied within5°–90°).was 0.039 nm per cycle. We have prepared sampleswith film thicknesses of 8, 24, and 48 nm, which wereobtained using n = 200, 600, and 1200 reaction cycles,respectively.In addition to high precision in maintaining presetfilm thicknesses, the ALD method also ensures uniqueresults with respect to the uniform (conformal) coatingof large developed surfaces (including those ofcomplicated shapes), which is achieved by virtue of thefact that the reaction proceeds every time in a singlelayer chemisorbed on the substrate surface. We haveexperimentally checked for this on test structuresobtained by the ion-beam processing of a silicon platein which trapezoid-profiled grooves with a base widthof ~2 μm and a depth of ~4 μm were formed and subsequentlycoated by titanium dioxide using the ALDmethod. As can be seen from Fig. 13, the TiO 2 layerthickness is virtually the same over the entire surfacewith a complicated relief.Characterization of TiO 2 Films Obtainedby the ALD Method on TitaniumThe TiO 2 layers on titanium substrates were studiedby Fourier transform IR spectroscopy (FTIR)(Perkin-Elmer Spectrum 100 spectrometer) and X-ray diffraction (XRD) (Rigaku Ultima IY universaldiffractometer), which showed that both the thicknessand structure of TiO 2 layers depend on thenumber of reaction cycles. Figure 14 shows theFTIR spectra measured in the reflection modebefore (curve 1) and after (curves 2–4) TiO 2 depositionby the ALD method. As can be seen, the spectraof both the titanium substrate (curve 1) and TiO 2layer deposited for 200 cycles (curve 2) display abroad absorption band at 700–1000 cm –1 , whichcorresponds to amorphous titanium dioxide [9]. Anincrease in the number of cycles to n > 600 leads tothe appearance of a sharp peak at 870 cm –1 , whichcorresponds to a crystalline TiO 2 with an anatasestructure (curves 3 and 4). According to the FTIRdata, similar changes from the amorphous to polycrystallineanatase structure were observed for thefilms grown by the ALD on silicon substrates.The appearance of a polycrystalline anatase structurein the layer grown by ALD with increasing numberof reaction cycles was also confirmed by the XRDdata (Fig. 15).LayerTiO 2Bottomlayer2 μmFig. 13. SEM image of the surface of silicon with a trapezoidalgroove formed using ion-beam processing followedby TiO 2 deposition by the ALD method.NANOTECHNOLOGIES IN RUSSIA Vol. 5 Nos. 9–10 2010

706ALEKHIN et al.R, %90801703260450401500 1400 1300 1200 1100 1000 900 800 700 600 500Wavenumber, cm −1Fig. 14. FTIR spectra of TiO 2 films on polished titanium substrates: (1) initial Ti substrate; (2–4) TiO 2 layers grown by the ALDmethod using n = 200, 600, and 1200 reaction cycles, respectively.Measurements of the wetting contact angle on thesubstrates with titanium dioxide coatings obtained bythe ALD with n > 600 on acid-etched titanium surfacesshowed easily reproducible results with the contactangle within 0°–5°.Relative intensity, %10090200 cycles80 600 cycles7060501200 cyclesAnatase40302010020 24 28 32 36 40 44 48 52 56TiO 2Ti(101) AnataseFig. 15. XRD patterns obtained for TiO 2 coating formedon titanium by the ALD for n = (1) 200, (2) 600, and(3) 1200 reaction cycles. .3212θ, degThe biocompatibility of modified titanium sampleswas assessed in experiments performed according tostandard method [28] with an evaluation of the abilityof MC3T3-E1 osteoblast cells to proliferate, adhere,and differentiate. The ability to differentiate was eval-Alkaline phosphatase activity,ng/(mg protein)/min (%)1601401201002020202001 2 3 4Fig. 16. Histogram of the alkaline phosphatase activity onmedical titanium samples prepared by different methods:(1) initial Ti surface; (2) ALD of TiO 2 (n = 1500);(3) 1% HF acid etching followed by the ALD of TiO 2 (n =1600); (4) HCl/H 2 SO 4 acid etching followed by the ALDof TiO 2 (n = 1600).NANOTECHNOLOGIES IN RUSSIA Vol. 5 Nos. 9–10 2010