Nanto Two critical events observed on Cu films on glass substrate in the microscratch test S. Baba, Y. Yamaguchi, M. Ogawa and T. Iwamori, Y. Nagayama, Y. Yamagata and Y. Such stresses can be quite large, leading to a variety of effects, including de-adhesion. The structure and thermodynamics of surfaces in thin films are briefly reviewed, and then it is shown how surface thermodynamic parameters can be used to describe a variety of intrinsic stress behaviors. Keywords: Intrinsic stress; surface thermodynamics: thin films.
This stress can be quite large, often exceeding the yield stress of the material in bulk form, and can lead to deleterious effects such as cracking, spalling and de-adhesion.
However, it is sometimes necessary, or even desirable, for a thin film to be under stress. For example, it is generally required for electronic material applications that a semiconductor film be grown epitaxially, i. If the in-plane equilibrium lattice spacings of the film and substrate are different, the film will be under stress in order to achieve this lattice matching. Another example where an intrinsic stress is desirable concerns a material that has a thin film coating in a state of compressive stress that can result in enhanced fracture and fatigue resistance compared to the uncoated material.
In addition to the film-substrate interface and the free surface solid-vapor interface of the film, there can be grain bounda"Tel. Cammarata ries in polycrystalline films and interlayer interfaces in multilayered thin films. These surfaces can have a significant effect on the mechanical behavior of thin films in general and the intrinsic stress in particular.
As mentioned above, there can be an epitaxial relationship between a thin film and the substrate, leading to lattice matching at the film-substrate interface. If the lattice matching is perfect, resulting in a defect-free interface, the interface is referred to as coherent. If the film and substrate have different equilibrium lattice spacings, and the substrate is much thicker than the film, the film will have to be coherently strained in order for it to be in perfect atomic registry with the substrate.
Let af and a, denote the bulk equilibrium in-plane lattice spacings of the film and substrate, respectively. For a film perfectly lattice matched to the substrate, the in-plane coherency strain is equal to the misfit. As long as the misfit is not too large, it is generally possible to grow a coherently strained epitaxial film, at least at smaller thicknesses [l].
There is a critical thickness above which it is thermodynamically favorable for the film to elastically relax, resulting in a loss of the perfect lattice matching at the film-substrate interface. One way in which this can occur is by the formation of an array of dislocations at the film-substrate interface that can accommodate some or all of the misfit. As long as the spacing of these misfit dislocations is not too small, so that there is still a significant amount of residual lattice matching at the film-substrate interface, the interface is said to be semicoherent.
If the misfit dislocation spacing is less than a few lattice spacings, or if there is no epitaxial relationship between the film and substrate, the interface is called incoherent. There are two thermodynamic quantities associated with the reversible work to change the area of the surface . One of these is the surface free energy which can be defined by setting the reversible work to create new surface of area A equal to. For simplicity, it will be assumed that the surface stress is isotropic and can be taken as a scalarf this is valid for a surface which displays a three-fold or higher rotational symmetry.
Unlike the surface free energy, which must be positive otherwise solids would spontaneously cleave the surface stress can be positive or negative. For finite size solids in mechanical equilibrium, the surface stress will induce a volume elastic strain relative to a bulk solid .
Similarly, for a thin disk of thickness t, a surface stressfacting on the top and bottom surfaces will result in a radial Laplace pressure of 2flt. Because of this Laplace pressure, the lattice spacing in the interior of the solid at equilibrium will be different from the equilibrium bulk spacing. As with a free solid surface, a solid-solid interface, such as that between a thin film and the substrate, has an associated surface free energy that will be referred to as the interface free energy.
Since the phases on either side of a solid-solid interface can be independently strained, resulting in different strain states at the interface, there are two surface stresses that can be associated with this interface . These surface stresses for solid-solid boundaries will be referred to as interface stresses. For a thin film-substrate interface, it is convenient to define these interface stresses in the following manner. Such a strain would lead to a change in the misfit dislocation density at a semicoherent interface. Suppose a film-substrate system with a semicoherent film-substrate interface has a misfit m.
A simple model  for the interface free energy T o f a semicoherent interface of misfit m and coherency strain leads to the expression 6 R. Cammarata Figure 1. Thin film growth modes. A similar analysis for the interface which is stress h indicates that for a semicoherent interface, h is of order ro, consistent with experimental measurements for metal-metal semicoherent interfaces . Growth modes Three basic thin-film growth modes  have been identified see Fig. VolmerWeber growth involves three-dimensional growth of islands that eventually coalesce to form a continuous film.
This growth mode is favored when x Surface effects on intrinsic thin-film stresses 7 where 7. This growth mode is often encountered when there is no epitaxial relationship between the film and substrate for example, crystalline metal films deposited on amorphous substrates. The Stranski-Krastanov mode involves two-dimensional growth for one or two monolayers, followed by three-dimensional island-like growth. The switch-over from two-dimensional to three-dimensional growth appears to be related to effects of stress relaxation. Frank-van der Merwe and Stranski-Krastanov growth modes are favored when 7.
Epitaxial growth For layer-by-layer epitaxial growth, it is expected that the film stress is principally a result of coherency strains when a film with an in-plane equilibrium lattice spacing different from that of the substrate is completely latticed-matched with the substrate at the film-substrate interface. As long as the lattice spacing misfit is not too large, the film-substrate interface will be completely coherent during the initial stage of growth. From a thermodynamic point of view, this is because the work to form misfit dislocations at the interface is greater than the work to elastically strain the film to accommodate the misfit.
Since the volume strain energy of the coherent film is proportional to the film thickness, there will eventually be a critical thickness above which it is thermodynamically favorable to introduce misfit dislocations at the interface to relieve some of the elastic strain energy. An expression for the critical thickness t, can be given in terms of the surface thermodynamic quantities and thin film elastic modulus as [2, wheref, and 7. Usually the contribution of the free film surface has been ignored, which in many but not all cases is an acceptable approximation.
Assuming this to be true, it is possible to calculate the critical thickness by substituting equations 3 and 4 into equation 5 and employing an expression for r, that involves the selfenergies for an array of interface dislocations which completely accommodates the misfit. Cammarata 8 L cI 7.
Real-time wafer curvature measurements during ultrahigh-vacuum deposition onto amorphous SiOz [ 5 ] : a polycrystalline Ag; b amorphous Ge; c polycrystalline Si; d polycrystalline Al. Th is the ratio of the deposition temperature to the melting temperature. For Al, the stress generation as a function of thickness during growth and the stress relaxation as a function of time after deposition was halted at a film thickness of nm are shown. The process of stress relaxation can involve the nucleation of misfit dislocations at the interface or slip of pre-existing dislocations.
It is often found experimentally that it is possible to grow a completely lattice matched film to thicknesses greater than the critical thickness. This is presumably a result of kinetic limitations associated with the stress relaxation process. Nonepitaxial island growth Results [ 5 ] from recent experiments investigating the development of thin film stress during Volmer-Weber island growth by ultrahigh vacuum evaporation onto an amorphous substrate for a variety of film materials are shown in Fig.
It Surface effects on intrinsic thin-film stresses 9 is seen that the stress behavior, which occurs for both crystalline and amorphous films, generally involves an initial compressive stress, followed by a tensile jump. The initial compressive stress regime occurs during the formation and growth of islands before coalescence.
The rapid tensile rise initiates around the onset of coalescence and reaches a maximum when the film becomes continuous. The final compressive stage occurs during further growth of the continuous film. Different mechanisms that have been proposed for each regime that lead to permanent static contributions to the film stress will be reviewed below.
It should be noted that there is often a large dynamic contribution to the stress during growth that relaxes when the deposition is halted see Fig. This effect, which has been attributed to adatom effects , is not discussed here. Consideration is first given to the early stage of island growth [ An isolated island can be modeled as a disk of diameter d and thickness t. As discussed earlier, surface stresses acting on this disk exert a size-dependent Laplace pressure that results in an equilibrium lattice spacing different from the bulk equilibrium lattice spacing.
Let do and to represent the size of an island when it first becomes firmly attached to the substrate so that a film stress can be generated. As the island grows, the equilibrium spacing will change, but the island is constrained by the substrate not to deform laterally. Thus, the change in equilibrium spacing leads to a latent strain that is manifested as a film stress. As a result, the firmly attached islands will produce a compressive film as they grow, consistent with experiment. It is generally agreed that the rise in tensile stress observed in the different systems shown in Fig. This is consistent with a model that has been popular over the years that involves tensile stress generation resulting from relaxation of grain boundaries formed during coalescence [ and which has been the subject of recent reinterpretations and reformulations [ When two islands impinge upon each other, a grain boundary is formed and two free surfaces disappear.
As the distance between neighboring growing islands is reduced, there will be a critical interaction distance where it will be thermodynamically favorable for the islands to elastically deform and impinge to form a grain boundary. The elastic strain energy created when the islands impinge 10 R. The values that can be obtained from equations 7 and 8 should be viewed as approximate upper and lower limits for the stress contribution from the grain boundary relaxation mechanism. Significant grain growth is sometimes observed during and after island coalescence [16, This process is driven by the reduction in grain boundary area per unit volume of the film because the surface free energy of a grain boundary is positive, and thus reduction in the grain boundary area reduces the free energy of the film.
Since a region near a grain boundary is expected to display a lower atomic density than the interior of the grain, the elimination of grain boundaries leads to a negative latent strain and, therefore, a tensile film stress. It should be noted that an increased grain size often leads to a reduced flow stress, so that increasing the grain size may allow for stress relaxation by plastic flow.
Unlike the initial compressive stress regime during island growth and the tensile stress stage during coalescence, where plausible models have been proposed to explain these behaviors, the origin of the compressive stress generated after coalescence is less clear. One possibility is that it is a continuation of the initial compressive stress that was being generated during island growth that was temporarily masked by the tensile jump during island coalescence.
If this asymptotic value is larger in magnitude than the tensile stress generated by the grain boundary relaxation mechanism, a superposition of the general compressive behavior resulting from surface stress effects with a step-function-like tensile jump at coalescence can qualitatively explain the CTC behavior [ 5 ].
Another proposed mechanism for this late stage compressive stress involves the incorporation of surface adatoms into grain boundaries [ 5 ]. The diffusion of atoms into the grain boundaries is driven by the supersaturation of adatoms that exists during deposition. More studies need to be conducted before a clear understanding of the late stage compressive behavior is understood.
The growth mode is determined by a balance of surface free energies, and will greatly influence the generation of intrinsic stresses. The formation of a coherent lattice-matched interface between a film and substrate with different equilibrium lattice spacings will result in a coherency stress characteristic of epitaxially grown films. While a large amount of the stress results from dynamic processes during deposition, and can be relieved when deposition is halted, there is generally a significant residual stress. The early stage compressive residual stress, generated prior to island coalescence, can be understood as resulting from a Laplace pressure owing to surface stresses.
The tensile jump, which occurs during island coalescence, is generated when islands that are separated by a certain critical distance elastically deform in order to impinge on each other and form a grain boundary. The driving force for this impingement is the lowering in surface energy when a grain boundary is formed and two free surfaces disappear.
Grain growth, driven by the resulting reduction in grain boundary energy, can also contribute to a tensile stress contribution. The origin of the compressive stress generated after the tensile jump is not completely clear, but may be due to a variety of proposed mechanisms, including ones involving surface stresses or incorporation of adatoms at grain boundaries. Matthews Ed. Cammarata, Prog.
- Effective Curriculum for Underserved Gifted Students: A CEC-TAG Educational Resource.
- Top Authors;
- PDF Adhesion Aspects of Thin Films volume 2 Ebook - video dailymotion.
Cammarata, K. Sieradzki and F. Spaepen, J. Academic Press, Boston, MA Floro, E. Chason, R. Cammarata and D. Srolovitz, MRS Bull. Cammarata, T. Trimble and D. Srolovitz, J. VSP, Utrecht Hoffman, Phys. Thin Films 3, 21 1 Dupuy and A. Cachard Eds. Plenum, New York, NY Dojack and R. Hoffman, Thin Solid Films 12,71 Pulker, Thin Solid Films 89, Nix and B.
Clemens, J. Freund and E. Chason, J. Sheldon, A. Lau and A. Rajamani, J. Seel, C. Thompson, S. Hearne and J. Floro, J. Chaudhari, J. Doerner and W. Nix, CRC Crit. Solid State Muter. Adhesion Aspects of Thin Films, Vol. DLC exhibits hardness, wear resistance and chemical inertness properties close to those of diamond. IJnfortunately, DLC films delaminate due to internal compressive stress. This paper describes processing and characterization of functionally gradient diamond-like carbonmetal nanocomposite films on Ti-6AV alloy, which is commonly used in biomedical and aerospace applications.
Internal stresses in diamond-like carbon thin films were reduced via incorporation of elements that form carbides e. These materials were produced using a novel pulsed laser deposition process that incorporates a multicomponent rotating target. Transmission electron microscopy of the DLC-metal nanocomposite films revealed that these films self-assembled into particulate or layered nanocomposite structures that possessed a high fraction of sp'-hybridized carbon atoms. Nanoindentation testing of the DLC-metal nanocomposite films demonstrated that these films possessed high hardness and Young's modulus values of approximately 35 GPa and GPa, respectively.
These DLC-metal nanocomposite films can be optimized for specific medical applications: for example, DLC-silver nanocomposites have been shown to possess antimicrobial properties. Keywords: Diamond-like carbon: pulsed laser deposition; functionally gradient materials; antimicrobial coatings. One biomaterial coating with tremendous potential is diamond-like carbon DLC. Narayan The term diamond-like carbon DLC describes hydrogen-free hard carbon solids that possess a cross-linked, non-crystalline network of sp2- and sp3-hybridized carbon atoms [l].
DLC also offers transparency to light ranging from deep ultraviolet to far infrared. In addition, DLC films are amorphous, atomically smooth and do not contain open corrosion paths to the underlying substrate. DLC has been accepted as an ideal coating material for use in implantable medical devices over the last 10 years.
Allergy, carcinogenicity, wear, corrosion, oxidation and metal ion release can all be eliminated by coating a metal or polymer implant with DLC. DLC thin films have current and potential applications in cardiovascular cardiac stent and heart valve , orthopaedic knee joint and hip joint and ophthalmic intraocular lens and artificial retina areas. DLC has been shown by a number of investigators to be fully biocompatible with all cell types.
DLC does not cause blood coagulation. For example, Jones et al. DLC coatings did not cause hemolysis, platelet activation, or thrombus formation. In addition, fibrinogen adsorption on DLC is much lower than fibrinogen adsorption on metals or polymers commonly used in blood-contacting applications. Several vascular and cellular processes, including fibroblast proliferation, collagen synthesis and blood vessel proliferation, lead to the formation of an avascular connective tissue capsule.
The connective tissue capsule consists of several different cellular layers, including an inner layer of macrophages, a concentric layer of fibrous tissue and fibroblasts pm and an outer vascularized tissue layer. Many metals including steel, Co-Cr-Mo alloy and Ti-6AV alloy and polymers including poly methy1 methacrylate , polyurethane and polyethylene trigger the formation of relatively thick interfacial layers [7, DLC films elicit a minimal or nonexistent fibrous capsule; such minimal encapsulation ensures that medical device function will not be diminished.
In addition, a DLC film hermetically seals an implant, preventing the release of metal ions or monomer to surrounding tissues. Poor adhesion is the sole practical limitation preventing widespread application of DLC thin films. DLC films commonly possess large internal compressive stresses that can exceed 10 GPa. The magnitude of these internal stresses can be correlated with the fraction of sp3-hybridized carbon atoms [ Friedmann et al.
These films were then cooled to room temperature to allow for further deposition. Adherent films of 1. Both Raman spectroscopy and electron energy loss spectroscopy data from single-layer annealed specimens revealed only subtle microstructural and chemical changes compared with unannealed films. The main advantage of this thermal annealing technique for compressive stress reduction is that pure DLC films can be prepared.
However, there are several drawbacks to high-temperature annealing of DLC films. First, no polymer substrate can undergo this annealing process. Also, Kustas et al. Alternatives to high-temperature annealing must be found in order to deposit DLC thin films on metals and polymers used in medical devices. We propose the use of functionally gradient films, in which the concentration of metal systematically varies from one film interface to the next Fig. On the other hand, a high concentration of metal is desired near the film-substrate interface to improve adhesion with the substrate and to reduce internal compressive stress.
We have developed functionally gradient DLC-metal nanocomposite films using a novel single-target pulsed laser deposition technique. Functionally gradient DLC-metal nanocomposite films were developed with biofunctional metals. For example, silver exhibits anti-microbial and antiinflammatory properties. Nanocrystalline silver has demonstrated an unsurpassed anti-microbial spectrum, with anti-microbial function against different pathogens.
In addition, nanocrystalline silver provides broad-spectrum fungicidal action [ l l , Functionally gradient DLC film design. The high metal atom Concentration at the filmsubstrate interface provides improved adhesion to the underlying substrate. Narayan Transmission electron microscopy was used to determine the film microstructure. Rutherford backscattering spectroscopy, X-ray photoelectron spectroscopy, visible Raman spectroscopy and electron energy loss spectroscopy were used to assess carbon-bonding characteristics.
Scratch adhesion testing, nanoindentation testing and wear testing were performed in order to determine the mechanical and tribological properties of these films. Anti-microbial testing was performed in order to assess the anti-microbial properties of the DLC-silver nanocomposite film. The substrates were then cleaned with acetone and methanol for 10 min each in an ultrasonic cleaner.
The silicon substrates were placed alongside the Ti-6AV substrates, allowing for simultaneous deposition on both materials. The sample holder was loaded into the PLD chamber, shown schematically in Fig. A high-purity graphite pellet was used as the target, and its surface was partially covered by small pieces of various metals, including silver, copper, titanium and silicon Fig.
The depositions were conducted for 40 min at room temperature at a chamber pressure of Torr. The target-to-substrate distance was maintained at 4. The graphitemetal target was rotated at 5 rpm. The energy density of the laser pulse was approx. The different deposition conditions and time intervals are displayed in Table 1.
The sample names are denoted by the metal component of the DLC-metal nanocomposite. The number of metal pieces on the graphite target and the relative amount of metal ablation are provided. Much larger amounts of metal were introduced in these functionally gradient DLC-metal films than in the initial films. Structural characterization was performed on the thin films deposited on silicon substrates. Several cross-sectional transmission electron microscopy samples were prepared. These samples were examined in a Topcon B unit with a point resolution of 0.
Pulsed laser deposition system. Metal Piece Graphite Target Figure 3. Schematic of target configuration used in this study. Narayan the electron diffraction pattern was used to obtain short-range structural information. Film morphology and crystallinity were determined by high-resolution transmission electron microscopy. Heavy metal atoms in the DLC matrix can be studied in detail, since scattering power or contrast depends upon atomic number squared Z2. Parallel electron energy loss spectroscopy was used to obtain information about the carbon bonding in the DLC-metal nanocomposite films; loss spectra were collected from zero up to eV energy loss.
Raman spectroscopy was performed to assess bonding configuration and internal compressive stress within the DLC and DLC-metal nanocomposite films.
- To Rise Above the Fences: Poetry.
- Account Options.
- Lesson Plans The Feminine Mystique.
- Adhesion Aspects of Thin Films, volume 2: Adhesion Aspects of Thin Films - Google книги.
Mechanical and tribological testing was performed on functionally gradient DLC-silver and functionally gradient DLC-titanium nanocomposite films. A DCM dynamic contact module head was used in these tests. These scratch tests were performed under a linearly increasing load; the maximum load used was 1.
The scratch length was set to 3 , and the scratch speed was set to 3 d m i n. A diamond tip 20 pm tip diameter, Rockwell C geometry was used for these tests. Alumina and Cr6 bearings of 6 mrn diameter were used as static wear partners. The amplitude of the wear track was set to 6 mm, and the scratch speed was set to 3 c d s. Normal loads of 3, 7 and 10 N were applied. Finally, anti-microbial testing was performed on a DLC-silver nanocomposite film and a silicon piece using a modified disk diffusion test.
The bright contrast corresponds to the buckling edges. A regular, sinusoidal pattern was noted throughout the film. The buckling appears to originate at the specimen edges and other defects, and spreads quickly over the entire film. Buckling appeared at film thicknesses exceeding 50 nm and buckling size increased with film thickness. The generation of buckling in DLC films is also related to post-deposition environmental factors.
No buckling occurred as long as the films were kept in a vacuum. Exposure of the film to humidity or other gaseous species initiated the buckling process. This delamination process was also accelerated by an increase in ambient humidity. It has been proposed that gas atoms diffuse into the interface between the film and the substrate, and initiate the delamination process. These factors suggest internal compressive stress is the source of DLC film delamination. It is interesting to note that the DLC film on silicon does not immediately delaminate, while DLC film on Ti-6AV delaminates immediately after exposure to ambient humidity.
DLC on silicon forms a silicon carbide interfacial Figure 4. The silicon carbide interfacial layer provides better DLC film adhesion than the titanium carbide interfacial layer. This result suggests the importance of interfacial bonding in promoting adhesion of a DLC film to a given substrate. It appears that internal compressive stresses are minimized in these DLC-metal films. Transmission electron microscopy was also performed on the DLC-metal nanocomposite films. The DLC-copper composites were notable in that they appeared speckled Fig. This speckling indicates segregation of copper into a separate phase.
Figure 5. Figure 6. Adhesion properties of DLC-metal nanocomposite films 21 3. Z-contrast scanning transmission electron microscopy Z-contrast scanning transmission electron microscopy provides unique information on nanostructured composite materials.
The Z-contrast signal is collected from a high angle annular detector, and the electron signals scattered through large angles typically 75 to m a d are analyzed. Contrast is proportional to the atomic number Z squared. For example, the silverkarbon contrast is over 1. Non-carbide-forming elements, such as silver, platinum and copper, were dispersed as nearly spherical metal clusters in the DLC matrix Figs 7 and 8. Nanodiffraction and STEM imaging reveal that silver, platinum and copper form nanocrystalline particles, with an average crystal size that varies between 3 and 5 nm.
Figure 8 demonstrates atomically sharp boundaries between the silver nanoparticle and the hard carbon matrix. The large random particles that are observed in these micrographs are artifacts of the ion milling process used in transmission electron microscopy sample preparation. Dark field cross-sectional images of functionally gradient DLC-titanium and functionally gradient DLC-silver nanocomposites are shown in Fig. The bright regions correspond to the higher atomic number titanium and silver regions, while the dark regions correspond to the DLC matrix.
From kinetic considerations, the formation of coherently-strained nanometersized metal particles having a narrow size distribution is favored, because Ostwald ripening is not favored under these conditions. From thermodynamic considerations, the total energy within the DLC-metal nanocomposite system includes elastic energy, surface energy, interface energy and the edge energy of nanometer-sized metal particles.
This DLC-metal nanocomposite system minimizes its total energy when a periodically ordered array of three-dimensional, coherently-strained nanometer-sized dots is formed. Figure 7. Bright field Z-contrast image of DLC-silver nanocomposite film. Narajan Figure 8. Dark-field Z-contrast image of DLC-silver nanocomposite film. Figure 9. The presence of metal carbides in DLC-metal composites containing carbide-forming metals was corroborated with election energy loss spectroscopy.
For example, the low loss titanium carbide peak was observed at 22 eV. Electron energy loss spectroscopy Electron energy loss spectra between to 3 10 eV were acquired. The sp3 fraction was determined from the K edge loss spectra using an empirical technique . Radial distribution function Electron diffraction provides high intensity beams and large scattering cross sections, which assist in the characterization of amorphous and nanocrystalline materials. Radial distribution function RDF analysis of the electron diffraction pattern provides short-range structural information on amorphous materials.
The function G r gives the most probable distances between atoms in a sample. Specifically, radial distribution function analysis provides the first and second coordination numbers, and the first and second nearest atomic neighbor distances. The values for amorphous carbon can be compared with those of diamond and graphite. Since DLC possesses both fourfold and threefold atomic coordinations, the first and second coordination spheres in the radial distribution function have values between those of graphite 3, 6 and those of diamond 4, Figure 10 shows the radial distribution function for the DLC The shaded areas in the radial distribution function represent the data used in calculating the second and first nearest neighbor distances.
Background correction was used to aid analysis. Figure 11 contains normalized G r values as a function of distance for the DLC Using the best-fit data interpolation, the first and second neighbor distances are 1. The ratio of the second to the first nearest neighbor for the DLC As seen in Table 2, these values are similar to those for pure DLC and those for pure diamond. This result indicates that metal atoms have a minimal influence on the sp3bonding in DLC-metal nanocomposite films. Narayan Radial distribution function G r as a function of distance of DLC Normalized radial distribution function G r as a function of distance for DLC Table 2.
Adhesion properties of DLC-metal nanocomposite films 25 3. Rutherford backscattering spectroscopy and X-ray photoelectron spectroscopy The composition of DLC-metal nanocomposite films can be estimated using a simple geometric consideration. This estimate can be obtained by taking a ratio of the length of the arc of the ablation path on the metal piece to the perimeter of the circle navigated by the laser beam Table 3.
Unfortunately, this estimate neglects the differences in the ablation rates of graphite and metal. The true metal composition will be less than the geometric estimate, because metals possess higher reflectivity and smaller ablation rates than graphite. The composition of DLC-metal nanocomposite films was obtained experimentally by Rutherford backscattering spectroscopy and X-ray photoelectron spectroscopy.
Rutherford backscattering spectroscopy determined the concentration of copper in the DLC-copper nanocomposite film Cu-1 to be 1.
Adhesion Aspects of Thin Films, Volume 1 - CRC Press Book
X-ray photoelectron spectroscopy data corroborated the Rutherford backscattering spectroscopy findings. X-ray photoelectron spectroscopy gave a concentration of 1. DLC-copper nanocomposite films are quite smooth and nearly entirely free from particulates. On the other hand, a number of particulates of variable size can be observed in DLC-titanium and DLC-silicon nanocomposite films. The morphology and size of these particulates suggest that they were formed from condensed liquid droplets. Splashing takes place in most materials through subsurface boiling or shock-wave ejection of particulates.
The large amount of particulates in DLC-titanium and DLC-silicon nanocomposite films results in the geometric estimations being lower than the experimental values obtained through Rutherford backscattering spectroscopy and X-ray photoelectron spectroscopy. Table 3. Narayan 26 , , I , , , 1. Adhesion properties of DLC-metal nanocomposite films 27 Figure Narayan 28 3.
Raman spectroscopy The Raman spectra of functionally gradient DLC-silver nanocomposite films contain broad peaks, because selection rules for optical transitions are relaxed. All of the spectra show the following: 1 a broad hump centered in the cm-' region, which is known as the G-band, and 2 a small shoulder at cm-', which is known as the D-band. The G-band is the optically allowed EZgzone center mode of crystalline graphite, and is typically observed in DLC films.
The Dband is the A,, mode of graphite. High quality DLC films demonstrate the following: 1 a relatively symmetrical G-band and 2 a lesser D-band, suggesting an absence or a low amount of graphite clusters. The visible Raman spectra for the DLC-titanium nanocomposite reveals increased asymmetry and an increased D-band. These features suggest that the DLC-titanium nanocomposite possesses less tetrahedrally bonded carbon than high-quality DLC. Raman spectroscopy was also used to study the internal stress conditions within the DLC-metal nanocomposite films.
Interatomic separation is correlated with the interatomic force constant, which, in turn, is correlated with the atomic vibrational frequency. The principle by which this data interpretation technique operates is as follows: when a material is stressed, the equilibrium separation between its constituent atoms is altered in a reversible manner.
If the tensile load on the material increases, bond lengths increase, force constants decrease and vibrational frequencies decrease. On the other hand, if the material is subjected to mechanical compression, bond lengths decrease, force constants increase and vibrational frequencies increase. The visible Raman spectrum G-peak positions for DLC-metal nanocomposites were determined by multiple Gaussian fittings. The G-peak for DLC was at Raman spectroscopy of functionally gradient DLCsilver composites also revealed that films with larger silver concentrations exhibited larger shifts in both G- and D-peaks Table 4.
These values suggest that DLC-metal films with higher metal concentrations possess lower amounts of internal compressive stress. Nanoindentation, adhesion and tribological properties Nanoindentation revealed significant substrate effects, due to the presence of a relatively soft substrate and a relatively hard coating. During nanoindentation, the modulus of the coated sample approached that of the uncoated sample at roughly Adhesion properties of DLC-metal nanocomposite films 29 Table 4. Table 5. Substrate effects are observed even at indentation depths of 75 nm.
Table 5 illustrates average hardness and modulus values for several films at nm maximum indentation depth. In general, functionally gradient DLC-silver nanocomposite films demonstrated slightly higher modulus and hardness values than functionally gradient DLC-titanium nanocomposite films. These differences can be primarily attributed to the slightly higher concentration of sp3-hybridized carbon observed in functionally gradient DLCsilver nanocomposites.
The values observed are comparable to values reported by Voevodin et al. Evaluation of the adhesion between coating and substrate was performed using a CSM Microscratch instrument. Scratch adhesion testing is commonly used for determining the integrity of coated substrates [ 15, , The coating-substrate response to scratch testing may be separated into three regimes. In regime one, mild plastic deformation is observed up to tensile cracking. In regime two, higher loads produce both regular and irregular crack patterns. If you are the author of this article you still need to obtain permission to reproduce the whole article in a third party publication with the exception of reproduction of the whole article in a thesis or dissertation.
Information about reproducing material from RSC articles with different licences is available on our Permission Requests page. Fetching data from CrossRef. This may take some time to load. Jump to main content. Jump to site search. Journals Books Databases. Search Advanced. Current Journals. Archive Journals. All Journals. New Titles. Pick and Choose. Literature Updates. For Members. For Librarians. RSS Feeds. Chemistry World. The desorption and solubility properties of porphyrin monomers are similar to C 60 , thus removal of the fullerene layer results in simultaneous desorption or dissolution of the porphyrin layer, for example by immersion in toluene.
We now consider the transfer of porphyrin polymers onto a target dielectric substrate. An extended covalently linked network of TBPP was prepared by sublimation onto a heated substrate as described in the Experimental section. The ordered regions are very similar to those originally reported by Grill et al. If the molecules are deposited with sub-monolayer coverage on a substrate held at room temperature, followed by annealing, small disconnected islands in which monomers are connected in an arrangement with square symmetry are observed.
For the transfer experiments a 15 nm thick layer of C 60 is deposited on a porphyrin polymer derived from Zn-TBPP and the network is transferred to SiO 2 by peeling, gold etching and mechanical transfer as described earlier. These maps, taken over macroscopic areas of 0. Figure 5: Fluorescence emission spectroscopy maps over 0. Due to the thermal stability of the covalent bonds linking neighbouring porphyrins, the residual C 60 can be removed by annealing without removing the polymeric network note that a similar annealing treatment applied to non-polymerised porphyrins results in complete removal of the molecular thin film.
The spectra of transferred porphyrin two-dimensional polymers are similar in shape to those of transferred or sublimed porphyrin monomer monolayers; the peaks are observed at, within experimental error, the same wavelength as the monomer. These observations are consistent with previous studies of arrays of porphyrins coupled by phenyl groups . In conclusion we have shown that C 60 shows an unexpected mechanical adhesion which is sufficiently strong to promote the removal of a metal film from a mica substrate.
Furthermore this route may be used to remove molecular thin films from a metal substrate through a process of mechanical removal followed by etching, and also to transfer them to a dielectric surface. The method is demonstrated for a SiO 2 substrate but is expected to be compatible with other dielectrics. The process is effective for films with thickness as small as a monolayer and has been demonstrated as route to isolate two dimensional polymers formed by on-surface synthesis, allowing an investigation of their functional properties. Commercially supplied terminated gold films on mica Georg Albert, Physical Vapor Deposition are used as substrates and prepared via Ar-sputtering for 30 min at 0.
The sputter-anneal-cycle is repeated until the herringbone reconstruction is clearly observed in STM images. After the mica is removed, the gold is etched using commercial gold etchant supplied by Sigma Aldrich , an aqueous KI solution, for 3 to 5 min. Subsequently the samples are rinsed with de-ionised water to remove excess KI. Anatolie S. Yaron Paz. Mayrhofer, Michael Rohwerder and Andreas Erbe.
Adhesion Aspects of Thin Films, Volume 1
Neil R. Peter H.
Related Adhesion Aspects of Thin Films, volume 2: Adhesion Aspects of Thin Films, volume 2
Copyright 2019 - All Right Reserved