METHOD OF PRODUCING A HIGH DENSITY PATTERN OF ISOLATED CLUSTERS
FIELD OF THE INVENTION
The present invention relates to production and manipulation of isolated clusters formed on the surface of a substrate, in particular but not exclusively nanoclusters.
The present specification will make reference to several published articles, of which the contents are herein incorporated by reference.
BACKGROUND OF THE INVENTION
In the booming field of nanotechnologies, efforts are being made to produce nanostructures usable, for example: - in nanoelectronics, for instance under the form of Quantum Dots (QD's), single electron transistors, nanophotonic devices, etc.;
- in the manufacturing of chemical products such as highly specialized catalysts; and
- in the field of superconductivity as well. Such applications require mastering the size, distribution and adhesion of nanoaggregates.
Precise control of nanocluster properties, such as size, lateral position and positional stability, is required in applications such as supported heterogeneous catalysts, magnetic and optical data storage media, and others where size, lateral position and shape influence the optical, magnetic, electronic and/or chemical properties. Precise control of nanocluster alignment and stability
is also required in the fabrication of nanostructures and in the fabrication of material requiring accurate alignment to achieve their optimum conductivity potential.
A scanning probe microscope, in tunnelling and atomic force manifestations thereof, is known to be extremely useful for the fabrication, characterization and manipulation of nanostructures. Investigations in this area include Scanning Tunnelling Microscope (STM) tip movement of weakly absorbed xenon atoms on a nickel surface [1] and the sled-type movement of C6o islands by an Atomic Force Microscopy (AFM) tip during image scanning [2]. Schaefer et al. [3] first demonstrated that the AFM tip can deliberately displace Au clusters on a surface. The controlled manipulation of GaAs nanoparticles on a GaAs surface with AFM was further used in particle-by-particle nanofabrication [4]. Some recent interesting results include local surface modification by an AFM tip [5], formation of a high area density oxide array on Ti [6], an extension thereof to Si3N [7], and manipulation of individual single-walled carbon nanotubes on a SiO2 surface [8] and multi-walled carbon nanotubes on a Highly Oriented Pyrolytic Graphite (HOPG) surface [9]. Quate and co-workers [6] were able to achieve an area density of oxide of 1.6 terabits-in"2 on Ti using single- walled C nanotubes grown on an AFM cantilever, thereby paving the way to a new approach for ultrahigh density data recording using AFM oxidation.
More recently, it was shown to be possible to manipulate double stranded DeoxyriboNucleic Acid (DNA) by an AFM tip on a molecular level [10], and even to manipulate single proteins and DNA molecules by pulling and twisting [11]. It has also been demonstrated that AFM may be used to control individual nano- whisker nucleation at specific sites [12], as well as the use of the tip as a scribe or punch [13] in a soft material. A recent study [14] contemplates a probe control software and experimental mechanism for accomplishing many of these manipulations.
Furthermore, the preparation, electronic structure and physical properties
of nanoparticles have been topics of active research effort over the past decade because of their size-dependent optical and electronic properties [39, 40]. Supported metal clusters on the surface of relatively non-reactive substrates, such as Al203 and amorphous carbon, have been extensively investigated by X- ray Photoelectron Spectroscopy (XPS) [41-45], STM/AFM [46-48] and other surface sensitive techniques [49-54]. The coalescence of the clusters on supported surfaces is an important aspect in the construction and assembly of nanostructures.
Although electron beam lithography may be used in the fabrication of structures with dimensions in the range of nanometers, the low throughput of serial lithography methods represents a severe bottleneck. Alternative methods, such as those based on self-organization to form large arrays of well-defined QD's, have opened new pathways for nanostructure fabrication. The very recently discovered low energy ion sputtering at normal incidence has been demonstrated to be an attractive alternative to the Stranski-Krastanov (SK) growth mode [55], opening a promising route for die parallel fabrication of uniform semiconductor QD's ordered in a hexagonal array [56-57]. However, the two-dimensional (2D) dots are still not isolated or supported by the surface of the substrate, and the processing issues are not yet solved. In addition, the controlled size and surface distribution of metal clusters were found to be very important in determining the electrical properties of Multiple tunnel Junction Devices (MJD) based on clusters such as QD's [58].
Also, there is a longstanding interest in surface modification of polymeric materials through the use of low and middle kinetic energy N2 + ion beam and N2 plasma treatments [26-33]. An important goal of the surface modification is the adhesion of deposited thin metal films to a substrate. It was recently shown that N (Nitrogen)-containing groups, grafted by N2 plasma surface treatment [18, 34] onto Dow Cyclotene 3022 polymer, which is a low permittivity polymer, are characterised by an excellent adhesion of evaporated copper (Cu) thin films, as measured by MicroScratch® and - peel tests. The importance of the surface
grafting of N-containing species through the use of low energy (3 keV) N2 + implantation into Dow Cyclotene 3022 polymer at an angle of 55° with respect to the surface was subsequently determined. The use of such ion beams leads to only a weak improvement of Cu adhesion due to an XPS-determined lack of N- containing groups at the outer surface of the substrate [35]. This is explained by the fact that ion implantation at 3 keV places the N deeper into the polymer material than a plasma treatment does, as determined by a Stopping and Range of Ions in Matter (SRIM) calculation [36]. Thus, while the substrate contains more N following ion beam treatment, less N is available at the surface where it can react with the metal [34].
While N2 plasma chambers are now used in microelectronics fabrication, the use of ion beams for the introduction of N-containing groups may be required due to specific conditions. Therefore, efforts are presently made to control, i.e. limit the ion beam implantation depth.
SUMMARY OF THE INVENTION
In accordance with the present invention, there is provided a method of producing a high density pattern of isolated clusters on a surface of a substrate, comprising: treating the surface of the substrate through reactive plasma to produce element-adhesion sites distributed on the surface of the substrate in accordance with the pattern; depositing a first transition series element on the surface of the substrate; and forming, by diffusion and/or coalescence, clusters of the deposited element on the element-adhesion sites.
The present invention is also concerned with a method of manipulating the size of clusters of a first transition series element on a surface of a substrate, comprising: irradiating the clusters with a low energy ion beam to cause coalescence of the clusters; and in response to the cluster irradiation with a low energy ion beam, diffusing the clusters on the surface of the substrate to thereby
change the size of the clusters.
Further in accordance with the present invention, there is provided a method of producing clusters on a surface of a substrate, comprising: depositing a first transition series element on the surface of the substrate; forming clusters of the deposited element; and manipulating a size of the clusters on the surface of the substrate, comprising irradiating the clusters with a low energy ion beam to cause coalescence of the clusters through diffusion on the surface of the substrate.
The present invention still further relates to a method of enhancing adhesion of clusters to a surface of a substrate, comprising: depositing a first transition series element on the surface of the substrate; forming, by diffusion and/or coalescence, clusters of the deposited element on the surface of the substrate; and, before deposition of the first transition series element and formation of the clusters of the deposited element, irradiating the surface of the substrate with a low energy ion beam to modify the surface of the substrate in view of enhancing adhesion of the clusters of the deposited element to the modified surface of the substrate, wherein irradiating the surface of the substrate comprises applying the low energy ion beam to the surface of the substrate at a grazing angle with respect to the surface of the substrate.
Finally, the present invention is concerned with a method using contact mode atomic force microscopy for (a) locally cleaning a surface of a substrate on which clusters of a first transition series element have been formed and (b) assembling the clusters, comprising applying an atomic force microscopy tip to the surface of the substrate, and scanning with the atomic force microscopy tip a region of the surface of the substrate, wherein scanning of the region of the surface of the substrate with the atomic force microscopy tip is performed in a same direction to form a line of clusters at an edge of the scanned region.
The foregoing and other objects, advantages and features of the present
invention will become more apparent upon reading of the following non- restrictive description of illustrative embodiments thereof, given by way of example only with reference to the accompanying drawings.
BRIEF DESCRIPTION OF THE DRAWINGS
In the appended drawings:
Figure 1 is a flowchart of a first illustrative embodiment of method according to the present invention, for producing a high density pattern of isolated clusters on a surface of a substrate;
Figure 2a is a contact mode AFM image of clusters formed from a nominal 32 A of evaporated Cu, for a number of one (1 ) scan;
Figure 2b is a contact mode AFM image of clusters formed from a nominal 32 A of evaporated Cu, for a number of two (2) scans;
Figure 2c is a contact mode AFM image of clusters formed from a nominal 32 A of evaporated Cu, for a number of three (3) scans;
Figure 2d is a contact mode AFM image of clusters formed from a nominal 32 A of evaporated Cu, for a number of four (4) scans;
Figure 2e is a contact mode AFM image of clusters formed from a nominal 32 A of evaporated Cu, for a number of five (5) scans;
Figure 3 is a close-up image of the area of Figure 2e;
Figure 4 is a plot of the Root-Mean-Square (RMS) surface roughness versus the number of scans for the images of Figures 2a-2e;
Figure 5a shows a contact mode AFM 3 μm by 3 μm image of Cu clusters formed from a nominal 32A thick layer of evaporated Cu on untreated Dow Cyclotene 3022 polymer material, after 10 scans for a set point value of -2.0V;
Figure 5b shows a contact mode AFM 3 μm by 3 μm image of Cu clusters formed from a nominal 32A thick layer of evaporated Cu on untreated Dow Cyclotene 3022 polymer material, after 10 scans for a set point value of -1.0V;
Figure 5c shows a contact mode AFM 3 μm by 3 μm image of Cu clusters formed from a nominal 32A thick layer of evaporated Cu on untreated Dow Cyclotene 3022 polymer material, after 10 scans for a set point value of 0V;
Figure 5d shows a contact mode AFM 3 μm by 3 μm image of Cu clusters formed from a nominal 32A thick layer of evaporated Cu on untreated Dow Cyclotene 3022 polymer material, after 10 scans for a set point value of +2.5V;
Figure 5e shows a contact mode AFM 3 μm by 3 μm image of Cu clusters formed from a nominal 32A thick layer of evaporated Cu on untreated Dow Cyclotene 3022 polymer material, after 10 scans for a set point value of +5.0V;
Figure 6 is a 3-D isometric representation of Figure 5e;
Figure 7 is a plot of the particle density as a function of the number of scans, for the different set point values of Figures 5a-5e;
Figure 8 is a plot of the average height of a line formed by an assembly of
clusters after 10 scans, as a function of the set point value;
Figure 9a is a plot of a Cu cluster spectral property (Cu 2p3/2 binding energy) as a function of ion irradiation time and normal Cu thickness;
Figure 9b is a plot of a Cu cluster spectral property (Cu 2p3 2 FWHM (Full Width at Half Maximum)) as a function of ion irradiation time and normal Cu thickness;
Figure 9c is a plot of a Cu cluster spectral property (Cu 3d binding energy) as a function of ion irradiation time and normal Cu thickness;
Figure 9d is a plot of a Cu cluster spectral property (Cu 3d FWHM) as a function of ion irradiation time and normal Cu thickness;
Figure 10a is a plot of the evolution of Cu cluster size, as determined by XPS peak intensity during Ar+ (Argon+) ion bombardment;
I
Figure 10b is a plot of the evolution of Cu cluster size, as determined by XPS peak intensity during Ne+ (Neon+) ion bombardment;
Figure 11 is a plot of the Cu mass thickness versus Ar+ ion bombardment time;
Figure 12 is a plot of the Cu cluster density versus Ar+ ion bombardment time;
Figure 13 is a plot of the Cu coverage versus Ar+ ion bombardment time;
Figure 14a is a plot of the Cu cluster average size for Ar+-treated HOPG surfaces before and after a 10-second pulse of irradiation;
Figure 14b is a plot of the Cu cluster average size for Ne+-treated HOPG surfaces before and after a 10-second pulse of irradiation;
Figure 15 are simulated curves of the dependence of Cu cluster average size versus ion irradiation time and primary ion current density;
Figure 16a is a plot of a N profile in polymethyl methacrylate (PMMA) after ion beam implantation of 3 keV N2 + at an incident angle of 5° with respect to the surface of the substrate, as computed by SRIM simulation;
Figure 16b is a plot of N . profiles in PMMA after ion beam implantation of 3 keV N2 + at an incident angle of 15° with respect to the surface of the substrate, as computed by SRIM simulation;
Figure 16c is a plot of N profiles in PMMA after ion beam implantation of
3 keV N2 + at an incident angle of 45° with respect to the surface of the substrate, as computed by SRIM simulations;
Figure 16d is a plot of N profiles in PMMA after ion beam implantation of 3 keV N2 + at 65° from the sample surface, as computed by SRIM simulations;
Figure 17 is a plot showing the initial average Cu cluster size (coalescence behavior) for evaporated Cu on Dow Cyclotene 3022 polymer substrates after different surface treatments;
Figure 18 is a flowchart of a second illustrative embodiment of method according to the present invention, for manipulating the size of clusters of a first transition series element on a surface of a substrate;
Figure 19 is a flowchart of a third illustrative embodiment of method according to the present invention, for enhancing adhesion of clusters to a surface of a substrate; and
Figure 20 is a flowchart of a fourth illustrative embodiment of method according to the present invention, using contact mode atomic force microscopy for (a) locally cleaning a surface of a substrate on which clusters of a first transistion series element have been formed and (b) assembling the clusters.
DETAILED DESCRIPTION OF THE ILLUSTRATIVE EMBODIMENTS
Although the illustrative embodiments of the present invention will be described mainly in relation to Cu, it should be kept in mind that the present invention equally applies to other first transition series elements including metals such as zinc (Zn), nickel (Ni), iron (Fe), cobalt (Co), etc., and their alloys. Regarding the substrate, it may be made of a material such as a polymeric insulator material, low permittivity polymers, oxides, HOPG, PMMA, Dow Cyclotene 3022 polymer, silk, other low permittivity materials, etc.
First illustrative embodiment
According to a first non-restrictive, illustrative embodiment of the present invention, there is provided a method 10 (Figure 1) in which the surface of a substrate is treated by reactive plasma to cause metal clusters to diffuse and adhere to specific sites and to grow to target dimensions. Once this growth is established, at least one other component of an alloy can be deposited so that it diffuses to and alloys with the metal clusters previously deposited. This method 10 comprises, as illustrated in Figure 1 :
Operation 12: a surface of a selected insulator substrate is treated, in a predetermined pattern, with a reactive plasma so as to establish adhesion sites provided with chemical groups to which a subsequently deposited first transition series element adheres to form fixed clusters;
Operation 14: a first transition series element such as a metal is deposited by evaporation on the treated surface of the substrate and clusters of the first transition series element are permitted to form naturally by diffusion and coalescence, at an optimised temperature; and
Operation 16: at least one further element of an alloy is then deposited by evaporation on the surface of the substrate at a given deposition rate and permitted to diffuse to, and alloy with, the previously fixed metal clusters.
Operation 12
In many applications, the insulator material of the substrate will be selected in order to avoid electrical contact between the various clusters. For example, a flexible, polymeric substrate can be used for a number of applications.
The surface of the substrate may be treated according to a predetermined pattern using photolithography. For example, the pattern is first transferred to a photoresist from a pattern-defining mask, and the photoresist is then etched. Finally, the surface of the substrate is exposed to plasma and the photoresist is removed before or after evaporation (Operations 14 and 16).
Alternatively, the surface of the substrate may be treated according to a photoresist procedure to produce an intentional porous array in a silicium (Si) wafer, and using this array as a mask directly applied to the surface of the insulator substrate, through which plasma is used to treat the surface in areas corresponding to the pores.
The use of reactive plasma rather than ion beams confines chemical
modifications at the surface of the substrate. As well known to those of ordinary skill in the art, ion beams, because of their generally higher energies (several keV, compared to approximately 10 eV for plasma treatment), cause a deeper implantation of ions, yielding a chemical modification far less concentrated at the surface than that achieved by reactive plasma.
The plasma surface treatment grafts chemical groups onto the surface of the substrate. These groups are able to form reactive bonds to the first transition series element (for example Cu-amine and Cu-thiol complexes), resulting in the first transition series element to bond in the patterned, treated areas (element- adhesion sites), thereby forming clusters. The first transition series element deposited in untreated areas diffuses and coalesces until it encounters fixed clusters that incorporate it, thus stopping further diffusion.
Operation 14
The control of the temperature during deposition of the first transition series element permits the control of the speed of coalescence.
Among the techniques for deposition of the first transition series element, it was found that evaporation allows deposition of the first transition series element at low energies, without surface degradation and/or solvent contamination, in a cost effective way. The first transition series element is deposited on the surface of the substrate and is then free to diffuse and coalesce and, ultimately, to chemically react with the chemical groups at the treated element-adhesion sites.
However, it should be kept in mind that it is within the scope of the present invention to use deposition techniques other than evaporation, such as sputtering, chemical vapour deposition, atomic layer deposition, electroplating, electrodeless plating, etc., as long as the surface is not destroyed and/or the metal is not deposited below the surface.
Operation 16
Control of the deposition rate of the alloy enables tailoring of the cluster sizes since the faster the deposition rate, the smaller both the initial and the final cluster sizes.
Local surface cleaning and cluster alignment using conventional Atomic Force Microscopy (AFM) will now be described.
For example, by using an AFM tip to sweep and push in 'contact mode, clusters such as Cu clusters prepared by vacuum evaporation onto Dow
Cyclotene 3022 polymer material and a subsequent exposure to atmosphere, can easily be displaced by the AFM tip and assembled at an outer edge of a region scanned by the AFM tip to form a line of Cu clusters.
The force applied by the AFM tip plays can be adjusted to ease motion of the clusters, and that plasma or ion beam treatment of the Dow Cyclotene 3022 polymer surface enhances cluster adhesion thereby hindering this ease of motion ability. Therefore, this can be used as a method of nanofabrication.
Operation 201 (Figure 20)
For example, Cu clusters are deposited by Ultra-High Vacuum (UHV) evaporation onto Dow Cyclotene 3022 polymer (also known as BCB) material, which is a low permittivity polymer [15], according to a known deposition procedure described in references [16-19, 34], Briefly stated, a nominal 32 A layer of Cu is evaporated onto a surface of a BCB substrate, at a deposition rate of 0.3 A per second at room temperature. Cu is deposited as spherical clusters due to a weak interaction with the BCB material [16]. The spherical shape of the clusters is confirmed by cross-sectional Transmission Electron Microscope (TEM) and in-situ XPS analysis. It appears that, at this nominal thickness of Cu,
most of the clusters are isolated.
Operations 202 and 203 (Figure 20)
Cluster imaging by AFM is carried out on a Digital Multimode Scanning
Probe Microscope using standard tips. During these measurements, the integral gain is 2.0, the proportional gain is 3.0 and the deflection set point is varied, as indicated in the accompanying figures. The set point value in a feedback control corresponds to a non-calibrated force between the AFM tip and the BCB surface; the higher this value, the greater the force applied by the tip. The tip scanning rate is 2 Hz and 256 lines are used per image. Similar parameters are used both for imaging and manipulation.
Prior to atmospheric exposure, while still under UHV, a strong cluster diffusion is observed on the BCB surface with a resultant coalescence, as previously demonstrated in the references [20,21], due to a weak Cu-BCB interfacial interaction [16]. However, because the AFM used herein operates in air, the deposited Cu clusters are exposed to atmosphere, where they quickly oxidize. This is confirmed by XPS (not shown), which also indicates that each component of the Cu 2p doublet exhibits a shake-up peak, indicating that Cu is oxidized to Cu2+, as expected.
An initial contact AFM scan (not shown) shows a number of large clusters with heights ranging from about 7 to 26 nm. A 3 μm by 3 μm square at the centre of the image is then selected for further manipulation and is repeatedly scanned for up to ten (10) scans, as illustrated in Figures 2a-2e, where only left-to-right scans are used.
More specifically, Figures 2a-2e are 3 μm by 3 μm contact mode AFM images of clusters formed from a nominal 32 A thick layer of evaporated Cu, as a function of the number of scans (1 scan; 3 scans; 5 scans; 7 scans; 10 scans,
respectively). Figures 2a-2e clearly show that (i) the cluster density decreases with the number of AFM tip scans and that (ii) the clusters are assembled at a right edge of the scanned region to form a line of clusters. Such a line of clusters can form, for example, a "nanowire".
Figure 3 is a 5 μm by 5 μm close-up image of the area shown in Figure 2e. A continuous decrease in the Cu cluster density as the number of scans increases can be observed.
Figure 4 further shows a plot of RMS roughness versus the number of scans, at a set point value of 0, as determined by an AFM software. The roughness values obtained do not seem to reflect the behavior hinted at by Figure 3. A very small variation is observed after 7 scans, despite the evidence found when comparing Figures 2d and 2e. This small variation of the RMS roughness is presumably caused by an equation used to compute it, which emphasizes the fact that the larger clusters make up the major component of the roughness value.
As demonstrated by the references [18, 34], surface chemical modifications can be used to introduce reactive groups that enhance cluster surface adhesion through the formation of reactive bonds at the interface that can in turn result in immobilizing clusters, for example Cu clusters. In particular, a N2 plasma introduced N-containing reactive groups that bonded to the Cu clusters, yielding clusters of smaller size than on an untreated surface [17-19, 34], and a strong Cu-BCB adhesion [18, 34]. AFM images of a nominal 32 A thick layer of Cu, deposited onto such a N2 plasma-modified surface (not shown), show no change in cluster distribution after 10 scans of the AFM tip under the same conditions as previously employed for obtaining the images of
Figures 2a-2e. A plot (not shown) similar to that of Figure 4 shows that there is no variation in roughness, which indicates that few, if any, Cu clusters have been displaced.
A role that such interfacial interactions play is illustrated in Figures 5a-5e, showing contact mode AFM 3 μm by 3 μm images of Cu clusters formed from a nominal 32A thick layer of evaporated Cu on untreated Dow Cyclotene 3022 polymer material, after 10 scans, for the following set point values: -2.0V (Figure 5a); -1.0V (Figure 5b); 0V (Figure 5c); +2.5V (Figure 5d); +5.0V (Figure 5e). It is obvious that a greater force results in a better cleaning. A 3-D image of the +5.0V set point (Figure 5e) result is shown in Figure 6.
To describe the cleaning process, the "Particle Analysis" software supplied with the AFM system is used. Arbitrarily selecting 15 nm as a minimum particle size, plots of the number of larger particles as a function of the number of scans are drawn, at given values of the set point. As illustrated in Figure 7, these plots indicate a weak cleaning at low values of the set point, with the cleaning ability increasing rapidly as the set point value increases.
As mentioned in the foregoing description, the line of assembled clusters can be used to describe the cleaning and assembling capacity. As shown in Figure 8, the average height of this assembled cluster line is directly proportional to the set point value for a given number of scans.
The reduction of cluster density in the scanned region results from the sweeping effect of the AFM tip, by which clusters are pushed due to a weak interfacial interaction [16, 17], In air, due to the presence of water vapour, the interaction between the Cu clusters and the surface of the substrate comprises contributions from capillary, as well as van der Waals, forces, the latter being by far the larger [23]. However, as previously shown in reference [16], the van der Waals force component itself is small in the present case, and can be easily overcome by the AFM tip so that a cluster is swept in the scanning direction. Moreover, the tip force is not sufficiently strong to damage the substrate.
Therefore, in this first illustrative embodiment of the present invention, there is provided a method to control the manipulation of nanoclusters, using
conventional contact mode AFM, for weak cluster-substrate interactions. Such a method may form the basis for both local cleaning and cluster alignment. For example, this method can be used to clean the untreated areas of the surface of the substrate on which a pattern of element-adhesion sites has been formed. Additionally, with the advent of methods of chemically modifying nanometer- sized surface areas through the use of plasmas [18, 34], lasers [24], and ion beams [25], this method allows effective nanofabrication through contact mode AFM.
Second illustrative embodiment
Turning now to a second non-restrictive, illustrative embodiment of the present invention, a method for the manipulation of the dimension of clusters on a surface of a substrate by low energy ion beam irradiation will now be described.
Generally stated, according to the second illustrative embodiment, low energy ion beams, for example Ar+ and Ne+ ion beams, are used to treat isolated Cu clusters deposited onto HOPG surfaces by evaporation. Following such treatment, cluster size and density changes are evaluated as functions of treatment time, using in-situ XPS. It was found that both Cu mass thickness and surface cluster density decrease with treatment time, as the clusters are sputtered, while weak interaction between the clusters and the treated HOPG surface permits cluster coalescence through diffusion of the clusters on the treated HOPG surface, even at room temperature. The effect of such behavior on the decrease in cluster size, which ultimately stabilizes, depending on the ion beam current density at the kinetic energy used will be discussed. Further, it is shown that short bursts of Ar+ or Ne+ ion beam treatment are not masked by such sputtering effects, and these bursts are found to lead to further enhancement of surface coalescence, as kinetic energy is transferred from the primary ion to the clusters.
Experiments were conducted for Cu evaporated onto fresh, untreated HOPG material, Ar+-treated HOPG material, as well as on Dow Cyclotene 3022 polymer material (low permittivity polymer for which HOPG is a model), and used in-situ XPS peak intensity ratios to indicate the presence of extensive coalescence of the Cu clusters at room temperature in UHV [17, 18, 20, 34]. This coalescence (instability) of the Cu clusters on the surface of the substrate is due to the absence of strong interfacial adhesive interactions [16-19, 21 , 34, 59, 60]. Although the supported cluster size and density can be controlled by conditions of preparation (such as evaporation rate, deposited effective mass thickness, etc.), the controllable parameters are still limited. Here, low energy ion bombardment is used to process nanoscale supported clusters. The experimental results indicate that such treatment may provide yet another means to effectively control and fabricate nanoscale structures.
First, the operations involved in the second illustrative embodiment of the method according to the present invention will now be briefly described with reference to Figure 18.
Operation 181
A sample of type ZYA (ZYA is a high grade HOPG) HOPG material, supplied by SPI, Inc., cleaved with adhesive tape just prior to each experiment is immediately inserted into an X-ray photoelectron spectrometer. Then, as a first treatment, the surface of the sample is treated by ion bombardment for 4 minutes at an energy level of 2 keV because, as was previously shown [21], this produces a saturated level of surface damage that assures valid sample comparison.
Operation 182
Cu deposition on the treated surface of the substrate is then carried out in a preparation chamber, using evaporation, to form Cu clusters on this surface.
The deposition rate is of about 2 A per minute, as measured by a quartz crystal microbalance placed near the sample.
Operation 183
Ar+ or Ne+ beam irradiation, using a VG AG2I cold cathode gun in the instrument preparation chamber, takes place at a pressure of less than 10"9 torr, using an ion beam energy of 2 keV and a pressure of 5 x 106 torr. The angle between the ion beam and the normal to the surface of the substrate is about 33°, which is expected to yield approximately 1 x 1013 ions per cm2 of Ar+ ion current density. Absolute values for Ne+ are not known. All the samples are immediately transferred to an analysis chamber without exposure to the atmosphere.
XPS is carried out in a VG ESCALAB 3 Mark II device, using hon- monochromated Mg Kα X-rays (1253.6 eV). A base pressure in the analysis chamber is less than 10"10 torr. Spectral peaks are separated, following Shirley background removal technique known in the art, by using an in-house non-linear least-mean-squares program, with peak positions and FWHM values as previously described [16-21 , 34, 60].
The results will now be presented in relation to Figures 9 to 15 of the appended drawings.
Figures 9a to 9d show Cu cluster spectral properties, respectively Cu
2p3/2 binding energy; Cu 2p3/2 FWHM; Cu 3d binding energy; and Cu 3d FWHM, as a function of ion irradiation time in seconds. The initial Cu equivalent thickness is 8 A. For comparison, Cu cluster growth during initial deposition is also given as obtained by XPS on Cu deposition [21 , 59]. The differences correspond to cluster growth [41-45, 16-19, 21 , 34, 59-61]. Clearly, the trends on deposition are the inverse of those on the Ar+ ion beam irradiation treatment subsequent to Cu deposition, indicating that the subsequent ion beam treatment
reduces the size of the Cu clusters.
Using a technique previously developed and published [19] for determining cluster size from XPS peak intensity ratios, Cu cluster size evolution on ion bombardment are obtained; as presented in Figures 10a and 10b. More specifically, Figures 10a and 10b illustrate the evolution of Cu cluster size, as determined by XPS peak intensity, during Ar+ ion bombardment and during Ne+ ion bombardment respectively. The low and high Ne÷ current densities of Figure 10b are obtained by pressure variation; the pressures used being respectively 5x10"6 and 3x10"5 torr. The results indicate that the cluster size decreases rapidly before stabilizing. In the example of Figure 10a, the Cu cluster average size stabilizes in a range comprised between 1.4 and 1.5 nm. In fact, it can be said that stabilization occurs at a slightly smaller size, taking into account the fact that coalescence occurs in the period of time between the ion beam treatment and the XPS measurement. The results for the two types of ion beams indicate that stabilization depends on the ion beam current density (see for example Figure 10b). at a chosen kinetic energy, and that, at least for these two types of ion beams, this does not appear to depend strongly on the type of ion used.
The change of effective Cu overlayer mass thickness may be determined from the change of the average Cu cluster size, using the following relation:
where la is an XPS peak intensity from Cu clusters, l°a is an XPS peak intensity from the Cu clusters prior to ion beam treatment, /s is an XPS peak intensity from the substrate, and /°s is an XPS peak intensity from the substrate prior to ion beam treatment, d is an average effective Cu cluster size, λa is an inelastic mean free path or attenuation length for the Cu clusters, and λs an inelastic
mean free path or attenuation length for the substrate, θ is the Cu coverage, which is defined as follows:
@X d (2)
where ω is a mass thickness of Cu.
Figure 11 shows a plot of Cu mass thickness as a function of Ar+ irradiation time. The change of Cu mass thickness ω as a function of ion beam treatment (ion bombardment) time, as illustrated in Figure 11 , is given by the following equation:
It can be seen that the Cu mass thickness decreases logarithmically with ion beam treatment (ion bombardment) time t as follows:
ω =ωo-B\θgt (4)
where ωo is the Cu mass thickness prior to ion beam treatment, β is the slope of the plot of Figure 11 and is dependent on experimental conditions.
Figure 12 illustrates the evolution of the Cu cluster density during Ar+ ion beam treatment (ion bombardment) as determined using the following equation:
ω n d3 = ω or n th
7 (5) wi n =n0e ' (6)
where n0 is the initial Cu cluster size, and β is a constant dependent on the particular experimental conditions.
Figure 13 further illustrates the Cu coverage variation as a function of ion beam irradiation obtained from Equation (2).
Short bursts of ion beam treatment are also found to manifest an effect on cluster coalescence behavior. Figures 14a-14b show the dependence of cluster size after deposition and subsequent short time (10 seconds) Ar+ and Ne+ ion irradiation bursts, respectively. It can be seen that there are obviously two stages for Cu cluster growth (coalescence processing): in a first stage labelled l, the as-deposited Cu cluster coalescence rate surges, then in a second stage labelled II, the cluster coalescence rate further increases steadily with time, on short-time ion irradiation bursts. That is, short bursts of ion irradiation enhance Cu cluster coalescence. Both the coalescence processes in Figures 14a and 14b are described by the following power law:
d - t 00
where k is a constant, which is related to the initial cluster size, and α depends on the ion treatment condition and average cluster size [17, 20]. Clearly, short- time ion bombardment bursts result in an increase in the value of α.
No new peaks appeared in the Cu 2P3/2 XPS spectra for as-deposited Cu or during Ar+ ion bombardment, indicating that no Cu-C bond formation occurred at the Cu-HOPG interface. The binding energy shifts of the Cu 2P32 spectra and the FWHM changes, from Figures 9a and 9b, are then attributed to Cu cluster size changes [41-45, 16, 60, 21 , 19, 61]. More specifically, the average size of Cu clusters decreases during ion beam treatment (ion bombardment), which is consistent with the Cu cluster size estimations obtained from XPS peak intensity ratios of Figures 10a and 10b. The FWHM increase of the Cu 2P3 2 spectrum of
Figure 9b, during ion bombardment, is also consistent with the results observed on Cu clusters on graphite [43-44], TiO2 [49] and AI2O3 [61]; it is generally observed for various first transition series element clusters and is attributed to lifetime broadening and screening of the photoelectrons [62].
It is interesting to note that, during ion bombardment, the Cu clusters rapidly reduce in size, stabilizing at approximately 1.4 nm, as shown by Figure 2a. A consideration of Figures 10a, 10b and 13 indicates that there is clearly a competition between atomic etching and Cu cluster coalescence during Ar+ ion sputtering. As illustrated by Figures 11 and 12, there are continuous decreases in the Cu cluster mass thickness and density during ion beam treatment. This permits one to analyze the change in Cu cluster size as the sum of two competing processes as follows:
d(d) d dm + ■ 3 d
(8) dt dt dt
where the first term on the right hand side is the cluster mass reduction due to ion sputtering, and the second term is the increase of coalescence.
For the sputtering process, the following relation applies:
3 dm SJM
0) dt
where S is the sputtering yield of Cu clusters, J is the ion beam current density, M is the molar mass of Cu, p is the Cu volume density and A is the Avogadro's constant.
From Equation (6), the following equation is obtained:
Because the effective thickness change may be non-linear, from Equation (3), the effective cluster sputtering yield may not be a constant, and:
nA dm P A ft d δd _ pA dω
S = (11)
JM dt JM dt 3JM dt
where dm is the mass loss of the cluster per unit surface area. Therefore,
From Equation (4), the following equation is obtained:
Equation (13) indicates that the Cu cluster size depends on the competition between the sputtering and coalescence processes of the cluster during Ar+ ion bombardment. The coalescence of the cluster results in a continuous increase of Cu cluster size, while sputtering results in a continuous decrease, with the proviso that the sputtering rate of the ion beam decreases with treatment time.
Figure 15 shows simulated curves of the dependence of Cu cluster size on irradiation time and primary ion current density, using Equation (13). A higher ion beam current will give a higher sputtering rate, and this gives a larger B value. The value of B used to fit the data of Figure 15 is 1.5 for low beam currents and 3.0 for high beam currents, which is close to the value of 3.5 obtained from Figure 11. The simulation indicates that (1) there is clearly a
saturation limit of size on ion sputtering the clusters, (2) the greater the ion current density, the shorter the time for reaching this stabilized Cu cluster size, and (3) the stabilized size of the clusters is dependent on the primary ion current density. That is, the higher the ion current density, the smaller the saturation cluster size. These results are in agreement with the experimental result of Figure .10b. This is confirmed in Figure 9, where the initial few moments of sputtering appear to be more important than coalescence on Cu cluster growth. While a similar result could have been directly arrived. at from a combination of Equations (4)-(6), it is herein chosen to derive results based on sound physical principles.
As shown in Figures 14a and 14b, coalescence is significantly enhanced by ion beam irradiation, irrespective of the ion. For example, without irradiation, α ~ 0.11 while after 10 seconds of Ar+ ion irradiation (2 keV), α ~ 0.31. This enhancement of mobility may be explained by the fact that the primary ions transfer kinetic energy during ion bombardment, and this added energy is used to overcome the surface interaction potential barrier (surface friction), leading to a higher coalescence rate. While at this moment, a relationship between the enhanced Cu cluster coalescence with recently reported metal surface - topography changes induced by ion bombardment cannot be proposed [64-67], it is noted that both processes involve mass transport on sputtering.
The previously described competition between sputtering and coalescence during ion beam treatment leads to the non-linear sputtering rate found in Figure 11. Thus, both Cu coverage and cluster density undergo nonlinear decreases during ion bombardment, as previously speculated following Equation (13).
Using recent results [17, 18, 34] that N2 plasma surface modification, with a kinetic energy of bombardment less 10 eV [18, 34], can extensively reduce Cu cluster surface diffusion, even at 350°C, and combining them with lithographic technology, the Cu cluster growth and surface diffusion can be controlled to
fabricate nanostructures, such as QD's and lines.
It is noted that Birtcher and coworkers [68, 69] recently reported that in- situ TEM showed no coalescence of Au clusters, deposited onto amorphous carbon, after MeV heavy ion sputtering of an Au target. While this may be due to experimental artefacts or interpretation, the result appears to disagree with earlier results [70, 71], which found strong coalescence of Au clusters produced by evaporation, as well as those subsequently irradiated by MeV heavy ions; such heavy ion irradiation was found to enhance cluster coalescence, much as in the present case.
In summary, according to the second illustrative embodiment of the present invention, the sputtering behavior of Ar+ and Ne+ ions on nanoscale Cu clusters, evaporated onto the HOPG surface, is investigated by in-situ XPS. It is shown that there is an extensive competition between mass loss by sputtering and cluster coalescence, on cluster size. It is further found that Cu cluster coalescence may be enhanced by short-time ion bombardment bursts, due to the transfer of kinetic energy and the absence of strong cluster-substrate
' interactions. The stabilized size of Cu clusters on the amorphous carbon surface, after extensive sputtering, is controlled by the ion beam density under the given sputtering conditions. The coalescence of Cu clusters, increased by ion treatment, results in changes of Cu mass thickness, coverage and cluster density. Those of ordinary skill in the art will appreciate that the present results are expected to be useful for nanoscale cluster and particle processing and control in future electronic applications based on nanostructure formation.
Third illustrative embodiment
Turning now to a third non-restrictive illustrative embodiment of the present invention, a method is described to enhance adhesion of clusters to a
• modified surface of a substrate using low energy N2 + beam irradiation at grazing angles. Since the ion beam diameter can be made very small, local area
treatment can be performed to form, for example, a nanometer pattern for the adhesion of Cu clusters and, therefore, a distribution of the Cu clusters on the surface of the substrate in accordance with this pattern.
As illustrated in Figure 19, the third illustrative embodiment of the present invention comprises:
Operation 191
A first transition series element is deposited on the surface of the substrate.
Operation 192
Clusters of the deposited element are formed on the surface of the substrate by diffusion and/or coalescence.
Operation 193
Before deposition of the first transition series element and formation of the clusters of the deposited element, the surface of the substrate is irradiated with a low energy ion beam applied at a grazing angle with respect to said surface of the substrate.
Through experiments intending to enrich the N content at the outer surface of a Dow Cyclotene 3022 polymer material by using a 3 keV N + ion beam to implant N+ ions at various angles, it was demonstrated that, both with in-situ XPS analysis of the modified surface of the substrate and the initial coalescence behavior of Cu clusters deposited by evaporation, that a lower angle (grazing angle) between the N2 + ion beam and the surface of the substrate leads to improved N enrichment at the surface.
Atom implantation using ion beams at a grazing angle with respect to the surface of the substrate was tested with an N2 + ion beam initial incident angle, varied down by 5° with respect to the surface using a specially designed sample holder.
Figures 16a-16b show N2 + implantation profiles into polymethyl methacrylate (PMMA) obtained at several incident angles by SRIM simulations [36]. PMMA has a density very similar to that of Dow Cyclotene 3022 polymer and is available in the TRIM program. A Coulomb explosion is assumed to occur [37], which, in this energy range, leads to the implantation of two N+ ions, each at half the original energy. As can be seen in Figures 16a, the nitrogen distribution peak is significantly shifted toward the outer surface (0 A) of the substrate when the incident angle θ tends to 5° in comparison with the results obtained with incident angles θ of 15° (Figure 16b), 45° (Figure 16c) and 65° (Figure 16d).
XPS spectra (not shown) were obtained for the C 1s, O 1s, Si 2p and N 1s orbitals. The peak shapes show no differences compared with the XPS spectra obtained at an implantation angle θ of 55°. Indeed, two N 1s components [35] at 399.5 and 400.5 eV are in the same 2:1 peak area ratio. However, the intensities are substantially changed, as shown in Table 1 below:
TABLE 1
Considering the relative concentrations obtained using appropriate sensitivity factors, it appears that the relative concentration of N increases when
the implantation angle θ decreases. Such a result means that, even under circumstances under which a certain amount of substrate is lost on sputtering, the N surface concentration can approach that achieved by a plasma treatment.
The initial coalescence kinetics of Cu clusters on these modified surfaces is determined in-situ by using XPS, as' previously described [17, 19, 20], using above Equation (7):
d=kta (7)
where d refers to a mean cluster size, t is the time, and k and α are constants.
The values of the constant α correspond to slopes of the lines in Figure 17, and are given in Table 2 further below.
The different cluster coalescence behaviors observed result from different extents of interaction between the cluster and the surface of the substrate [20].
The inventors named in the present patent application recently found a direct correlation between macroscopic adhesion, as determined by MicroScratch® testing, and nanoscopic adhesion, as determined by cluster coalescence behavior [38]. This correlation is used herein to determine MicroScratch® critical loads [18] for the coalescence behaviors observed, which prove to match those found directly [18], as listed in Table 2 below:
TABLE 2
The results clearly demonstrate that N2 + implantation at a grazing angle θ
of 5° yields an enhanced adhesion compared with N2 implantation at an angle θ of 55°. As can be easily shown from Figures 16a-16b and Table 1 , such a result is due to a higher concentration of N-containing groups at the outer surface of the substrate, although the adhesion achieved is not as high as that obtained with a plasma treatment [18, 32-34].
At least three reasons for this difference in adhesion between ion beam and plasma treatments are considered.
First, the surface concentration of N-containing groups resulting from a plasma treatment is much higher, because, as noted previously in reference [35], plasma deposition occurs at energies below those needed for bond breaking, which means that all the N is deposited onto the surface of the substrate made, for example of Dow Cyclotene 3022 polymer. Ion implantation, occurring at energies sufficient to break bonds, and ion scattering, which leads to broadened ion ranges and straggling effects, give a deeper N distribution.
Secondly, ion implantation causes sputtering, which leads to a constant erosion of the surface, a new surface constantly replacing an eroded one. Therefore, there is a maximum surface concentration that N can attain, even at grazing angles.
Thirdly, while the N s XPS peaks are at the same energies, respectively 399.5 and 400.5 eV, for both ion implantation and plasma treatments, the ratios differ, from 2:1 on implantation [34, 35] and 1:1 on plasma treatment [35]. Consequently, ion implantation leads to a higher concentration of groups such as -NHCHO, -NCO and -NHNH2, whereas plasma treatment leads to a higher concentration of groups such as — NH2 and -NHOH [34]. The groups contributing to the 400.5 eV peak may form stronger complexes with Cu. On the one hand, Table 1 shows that the O concentration is much greater by plasma treatment. On the other hand, it was previously demonstrated that an increase in interfacial O does not lead to a higher copper adhesion [34].
Therefore, in this third illustrative embodiment of the present invention, a method is provided to enhance adhesion of, for example, Cu to, for example, Dow Cyclotene 3022 polymer by N2 + beam implantation at grazing angles. Although the adhesion achieved does not reach adhesion levels obtained by N2 plasma treatment, it appears that the method allows enough adhesion for specific purposes.
The foregoing description shows that the present invention provides techniques for manipulating the dimensions of supported particles over large ranges of size, from micrometers to nanometers; for manipulating particle surface densities over large ranges of several orders of magnitude; for making multicomponent metal particle arrays through Ar+ ion-beam or plasma enhanced diffusion and coalescence; for using AFM to perform surface cleaning and particle assembly over dimensions ranging from micrometers to nanometers; for using extremely low grazing angle N2 ion beam irradiation to provide benefits of
N2 plasma treatment, such as an increased surface adhesion of Cu clusters for example, with the advantage of a local area treatment, since the ion beam diameter can be made very small.
Possible applications of the illustrative embodiments of the present invention comprise, amongst others:
- Supported metal nanoclusters (1-10 mm in size) can be used in high-density data storage devices, biosensors, single electron transistors, and possibly in the biotech industry for the fabrication of nano-arrays or for microfluidics.
- The clusters, stable to microelectronics processing temperatures, can be probed by any of several scanning probe microscopic techniques such as, for example, magnetic, magnetorestrictive, electric force, capacitance, etc., microscopies. Such probing can be carried out quickly, over a large number of clusters.
- Such arrays or patterns of clusters can function as catalysts for nanotube production, supports for biosensors, etc. Indeed, it is possible to treat the deposited clusters with reactants such as A-R-B, where A reacts with the deposited cluster, R is an organic chain and B is a chemical group that remains available for further reaction.
Although the present invention has been described hereinabove by way of illustrative embodiments thereof, these embodiments can be modified at will, within the scope of the appended claims, without departing from the spirit and nature of the subject invention.
REFERENCES
[1] D. M. Eigler, E. K. Schweizer, Nature (London) 344 (1990) 524.
[2] R. Lϋth, E. Meyer, H. Haefke, L. Howald, W. Gutmannsbauer, H.-J Gϋntherodt. Science 266 (1994) 1979.
[3] D. M. Schaefer, R. Reifenberger, A. Patil, R. P. Andres, Appl. Phys. Lett. 66 (1995) 1012.
[4] T. Junno, S. Anand, K. Deppert, L. Montelius, L. Samuelson, Appl. Phys. Lett. 66 (1995) 3627.
[5] Y. Kim, C. M. Lieber, Science 257 (1992) 275; J. A. Dagata, J. Schneir, H. H. Harary, C. J. Evans, M. T. Postek, J. Bennett, Appl. Phys. Lett. 56 (1990) 2001 ; E. S. Snow, P.M. Campbell, Apple. Phys. Let. 64 (1994) 1932; D. Wang, L. Tsau, K. L. Wang, P. Chow, Appl. Phys. Lett. 74 (1995)1295; S. Guwo, C.-L. Yeh, P.-F. Chen, Y.-C. Chou, T.-T. Chen,T.-S.
Chao, S.-F. Hu, T.-Y. Huang, Appl. Phys. Lett. 74 ( 999) 1090.
[6] E. B. Cooper, S. R. Manalis, H. Fang, H. Dai, K. Matsumoto, S. C. Minne, T. Hunt, C. F. Quate, Appl. Phys. Lett. 75 (1999) 3566.
[7] F. S. Chien, J.-W. Chang, S.-W. Lin, Y.-C. Chou, T. T. Chen, S. Gwo, T.-S. Chao, W.-F. Hsieh, Appl. Phys. Lett. 76 (2000) 360.
[8] H.W.C. Postma, A. Selimeijer, C. Dekker, Advanc. Mater. 12 (2000) 1299.
[9] M. R. Falvo, J. Steele, R. M. Taylor, R. Superfine, Tribol. Lett. 9 (20O0) 73.
[10] D. Anselmetti, J. Fritz, B. Smith, X. Femandez-Busquets, Single Molecules 1 (2000) 53; C. Thelander, L. Samuelson, Nanotechnol. 13 (2002) 108.
[11] T. Strick, J.-F. Allemand, V. Croquette, D. Bensimon, Physics Today 54 (2001) 46.
[12] B. J. Ohisson, M. T. Bjork, M. H. Magnusson, K. Deppert, L. Samuelson, L. R. Wallenberg, Appl. Phys. Lett. 79 (2001) 3335.
[13] U. Kunze, B. Klehn, Advanc. Mater. 11 (1999) 1473; S. Hu, A. Hamidi, S. Altmeyer, K. Koster, B. Spangenberg, H. Kurt, J. Vac. Sci. Technol. B16 (1998) 2822; A. Notargiacomo, V. Foglietti, E. Cia'nci, G. Capellini, M. Adami, P. Faraci, F. Evangelisti, C. Nicolin, Nanotechnol.10 (1999) 458; T.- H. Fang, C.-l. Wang, J.-G. Chang, Nanotechnol. 11(2000)181.
[14] R. Resch, A. Bugacov, C. Baur, B. E. Koel, A. Madhukar, A. A. G. Requicha, P. Will, Appl. Phys. A67 (1998) 265.
[15] S. Poulin, D.-Q. Yang, E. Sacher, C. Hyett, T. H. Ellis, Appl. Surf. Sci. 165 (2000) 15.
[16] D.-Q. Yang, S. Poulin, E. Sacher, C. Hyett, Appl. Surf. Sci. 165 (20O0) 116.
[17] D.-Q. Yang, E. Sacher, in "Metallization of Polymers 2" E. Sacher, ed., Kiuwer, New York, p 99.
[18] A. Sadough-Vanini, D.-Q. Yang, L. Martinu, E. Sacher, J. Adhesion 77 (2001) 309.
[19] D.-Q. Yang, M. Meunier, E. Sacher, Appl. Surf Sci. 173 (2001) 134.
[20] D.-Q. Yang, E. Sacher, J. Appl. Phys. 90 (2001) 4768.
[21] D.-Q. Yang, E. Sacher, Surf Sci. 504 (2002) 125.
[22] F. El Feninat, S. Elouatik, T.H. Ellis, E. Sacher, I. Stangel, Appl. Surf Sci. 183 (2001) 205.
[23] X. Wu, M. Meunier, E. Sacher, J. Appl. Phys. 87 (2000) 3618 and references therein.
[24] • D. Popovici, M. Meunier, E. Sacher, J. Adhesion 70 (1999) 155.
[25] D.-Q. Yang, E. Sacher, Appl. Surf. Sci. 195 (2002) 187.
[26] L. Zhang, K. Takata, T. Yasui, H. Tahara, T. Yoshikawa, Mater. Chem. Phys. 54 (1998) 98.
[27] J.-S. Cho, W.-K. Choi, S.-K, Koh, K.-H. Yoon, J. Vac. Sci. Technol. B16 (1998) 1110.
[28] G. Mesyats, Y. Klyachkin, N. Gavrilov, A. Kondyurin, Vacuum 52 (1999) 285.
[29] A. Caballero, D. Einen, J. P. Espinos, A. Fernandez, A. R. Gonzalez-Elipe, Surf. Interf. Anal. 21 (1994) 418.
[30] M. Guemmaz, A. Mosser, J.-J. Grob, R. Stuck, Surf. Coat. Technol. 100- 101 (1997) 353.
[31] D. L. Williamson, J. A. Davis, P. J. Wilbur, Surf. Coat. Tech. 103-104 (1998) 178. i
[32] M. R. Wertheimer, L. Martinu, J. E. Klemberg-Sapieha, G. Czeremuszkin, in Adhesion Promotion Techniques, edited by K. L. Mittal, A. Pizzi, Marcel Dekker, New York, 1999, p. 39.
[33] L. J. Gerenser, J. Vac. Sci. Technol. A6 (1988) 2897.
[34] D.-Q. Yang, L. Martinu, E. Sacher, A. Sadough-Vanini, Appl. Surf. Sci. 177 (2001) 85.
[35] D.-Q. Yang, E. Sacher, Appl. Surf. Sci. 195 (2002) 203.
[36] http://www.srim.org
[37] J. Los, T. R. Groves, in Collison Spectroscopy, edited by R. G. Cooks, Plenum, New York, 1978, Chapter 6.
[38] D.-Q. Yang, E. Sacher, J. Phys. C: Cond. Matt. 14 (2002) 7097.
[39] S. Edelstein, R. C. Cammarata (Eds.), "Nanomaterials: Synthesis, Properties and Applications', top Publishing, Bristol, 1996.
[40] J. H. Fendler (Ed.), "Nanoparticles and Nanostructured Films", Wiley-VCH, Weinheim, Germany, 1998.
[41] W.F. Egelhoff, Jr., G.G. Tibbetts, Phys. Rev. B19 (1979) 5028.
[42] G.K. Wertheirn, Phys. Rev. B36 (1987) 9559.
[43] C.T. Campbell, Surf. Sci. Rept. 27 (1997) 1.
[44] CR. Henry, Surf. Sci. Rept. 31 (1998) 231.
[45] H. Brune, Surf. Sci. Rept. 31 (1998) 125.
[46] E. Ganz, K. Sattler, J. Clarke, Surf. Sci. 219 (1989) 33.
[47] C. W. Whelan, C. J. Barnes, Appl. Surf. Sci. 119 (1997) 288.
[48] W.-G. Lu, H. Wu, Y.-Q. Xiong, Y. Guo, D.-Q. Yang, H.-L. Li, J. Vac. Sci. Technol. B18 (2000) 1156.
[49] U. Diebold, J.-M. Pan, T.E. Madey, Phys. Rev., B47 (1993) 3868.
[50] M. De Crescenzi, M. Diociaiuti, L. Lozzi, P. Picozzi, S. Santucci, Phys. Rev. B35 (1987) 5997.
[51] S. Norrman, T. Andersson, CG. Granqvist, Phys. Rev. B18 (1978) 674. .
[52] F. Parmigiani, G. Samoggia, G.P.-Ferraris, J. Appl. Phys. 57 (1985) 2524.
[53] H. Shirakawa, H. Komiyama, J. Nanoparticles Res. 1 (1999) 17.
[54] L. Zhang, F. Cosandey, R. Persaud, T.E. Madey, Surf. Sci. 439 (1999) 73.
[55] V. A. Schukin, D. Bimberg, Rev. Mod. Phys. 71 (1999) 1125.
[56] S÷ Facsko, T. Dekorsy, C. Koerdt, C Trappe, H. Kurz, A. Vogt, H.L. Hartnagel, Science 285 (1999) 1551.
[57] S. Facsko, H. Kurz, T. Dekorsy, Phys. Rev. B63 (2001) 165329.
[58] E. M. Ford, AH. Ahmed, Appl. Phys. Let. 75 (1999) 421; A. S. Cordan, Y. Leroy, A. Goltzene, A. Pepin, C Vieu, M. Mejias, H. Launois, J. Appl. Phys. 87 (2000) 345; Appl. Phys. Lett. 74 (1999) 3047.
[59] D.-Q. Yang, E. Sacher, Surf. Sci. 516 (2002) 43.
[60] D.-Q. Yang, E. Sacher, Appl. Surf. Sci. 173 (2001) 30.
[61] Y. Wu, E. Garfunkel, T. E. Madey, J. Vac. Sci. Technol. A14 (1996) 1662.
[62] W. F. Egelhoff, Jr., Surf. Sci. Rept. 6 (1987) 253.
[63] P. Deltour, J.-L. Barrat, and P. Jensen, Phys. Rev. Lett. 78 (1997) 4597.
[64] S. Rusponi, G. Costantini, C. Boragno, U. Valbusa, Phys. Rev. Lett. 81 (1998) 2735.
[65] S. Rusponi, G. Costantini, C Boragno, U. Vatbusa, Phys. Rev. Lett. 81 (1998) 4184.
[66] Maekawa, F. Okuyama, Surf. Sci. 481 (2001) L427.
[67] D. Flamm, F. Frost, D. Hirsch, Appl. Surf. Sci. 179 (2001) 95.
[68] R. C Birtcher, S. E. Donnelly, S. Schlutig, Phys. Rev. Lett. 85 (200O) 4968.
[69] L. E. Rehn, R. C Birtcher, S E. Donnelty, P. M. Baldo, L Funk, Phys. Rev. Lett. 87 (2001) 207601.
[70] J. G. Skoferonick, W. B. Phillips, J. Appl. Phys. 38 (1967) 4791 ; W. B. Phillips, E. A. Desloge, J. G. Skoferonick, J. Appl. Phys. 39 (1968) 3210.