Nanoseeding via Dual Surface Modification of Alkyl Monolayer for Site

Giin-Shan Chen , Sung-Te Chen , Yenying W. Chen , and Yen-Che Hsu ... Tran , Quang The Phan , Marek Balazinski , Le Thanh Son , Gul Zeb , Xuan Tuan Le...
0 downloads 0 Views 3MB Size
Download PDF
ARTICLE pubs.acs.org/Langmuir

Nanoseeding via Dual Surface Modification of Alkyl Monolayer for Site-Controlled Electroless Metallization Sung-Te Chen† and Giin-Shan Chen*,‡ † ‡

Department of Electronic Engineering, Hsiuping University of Science and Technology, Dali 412, Taichung, Taiwan Department of Materials Science and Engineering, Feng Chia University, Taichung 407, Taiwan ABSTRACT:

In this work, an attempt to fabricate nanostructured metallization patterns on SiO2 dielectric layers is made by using plasmapatterned self-assembled monolayers (SAMs), in conjunction with a novel aqueous seeding and electroless process. Taking octadecyltrichlorosilane (OTS) as a test material, the authors demonstrate that optimizing the N2 H2 plasma conditions leads to the successive conversion of the topmost aliphatic chains of alkyl SAMs to carboxyl (COOH) and hydroxyl (C OH) functional groups, which was previously found in alkyl SAMs only by exposure to “oxygen-based” plasma. Further modifying the plasmaexposed (either COOH or C OH terminated) regions with an aqueous solution (SC-1) creates surface functionalities that are viable for site-controlled metallic seeding (e.g., Co or Ni) with an adsorption selectivity of greater than 1000:1. Neither the combination of costly PdCl2 and complex additives nor the demerits of the associated aqueous chemistry (e.g., seed agglomeration and seed sparseness) are involved. Therefore, the seed particles are only 3 nm in size. Simultaneously, there are sufficient particle densities previously unattainable for electroless deposition to trigger highly resolved Cu metallization patterns with a film thickness of less than 10 nm. The formation of the seed-adsorbing sites is discussed, based on a plasma-dissociated, water-mediated chemical oxidation route.

1. INTRODUCTION In modern microelectronics industry, there is a high demand for creating on dielectric layers a pattern of metallic films with well-defined resolution and featured thickness in the nanometerscale regime. For instance, Cu-based metallization films with an overall thickness of less than 60 nm have to be built as interconnection lines within SiO2 or low-k dielectric layers for the sub-45-nm microelectronics technology nodes.1 Typically, the state-of-the-art technique for patterning a Cu-based metallization film relies on the Damascence process, including resist coating, photolithography, metallization (metallic barrier and Cu), and chemical-mechanical polishing.2 Electroless deposition has become one of the leading technologies for microelectronics metallization, following the finding of using an in-line electrochemical (electroless) process to sequentially deposit a Co- (or Ni-) based barrier layer, Cu thin-film interconnect, and Co-based capping layer, to form a metal-encapsulated, highly endurable Cu-conducting line.3,4 Electroless deposition is based on the reduction of metallic ions from a plating solution onto a dielectric substrate to be metalized without the application of an external electric current. This reaction normally requires the use of a catalyst that serves as r 2011 American Chemical Society

a temporary electron route between the metallic ions and reducing agent, ultimately initiating the autocatalytic metallization process. Since metallization is inhibited on areas not covered by the catalyst, selective metallization can be realized in the desired regions by electroless deposition with the aid of localized catalysts. Indeed, a key approach to selective metallization by electroless deposition is based on patterning self-assembled monolayers (SAMs) to manipulate the distribution of catalysts. As is wellknown, ultraviolet (UV) light is the most commonly used irradiation source to modify the surface structures of various SAMs made of aminopropyltriethoxysilane, alkanethiolate, block copolymer micelle, and so forth, tailoring their terminal headgroups (e.g., thiol and amine) to different chemical functionalities,5 7 resulting in a preferential binding of catalytic particles through covalent coupling and/or electrostatic interaction.6,8 The silane-based moieties in alkylsilanes are readily tethered to hydrolyzed oxide surfaces, enabling the formation of wellordered, uniform SAMs for micro- and nanofabrication applications. Received: June 5, 2011 Revised: July 26, 2011 Published: August 15, 2011 12143

dx.doi.org/10.1021/la202104z | Langmuir 2011, 27, 12143–12148

Langmuir For instance, SAMs of octadecyltrichlorosilane [OTS; CH3(CH2)17SiCl3] and of its derivatives [e.g., octadecyltrimethoxysilane; CH3(CH2)17Si(OCH3)3] can be patterned by UV-light, a scanning probe, anodization, plasma, microcontact printing, and so forth. These patterns subsequently act as a resist mask for pattern transfer9,10 or templates for micro- and nanostructured fabrication.11 13 To the best of our knowledge, the majority of studies on plasma treatments of alkyl (including alkylsilane) SAMs focus on plasma-induced pathways leading to surface functionalization;14 19 none of these works attempt to fabricate patterns of seeds for site-selective electroless metallization.14 20 As an example, the OTS-SAM is often used as a model material for assessing the surface modification of polyolefins by O214 or NH315 plasma because of the close resemblance of its constituent unit (methylene) to the olefin monomers. This is somewhat surprising, given that plasma treatment is a facile process to quickly activate chemically inert organic monolayers (e.g., methylterminated SAMs) that result in potential applications to nanostructured fabrication and biomolecular recognition.21 23 We have used N2 H2 mixed gases to replace toxic, irritating ammonia (NH3) as a vacuum plasma source to create a passivation layer on Si-based dielectric layers, which acts as a diffusion barrier against Cu diffusion.24 Herein, attempts are made to tether an alkylsilane-SAM to dielectric layers and to use such a vacuum plasma process to manipulate the surface chemical moieties of the SAM for site-selective metallization using a novel aqueous seeding and electroless process.25 OTS-SAMs are selected as a test case because the behavior of metallization for alkyl SAMs through plasma exposure is rarely documented in terms of siteselective seeding for micro- or nanostructured metallization. It will be shown that the synergy of the dual surface modification (i.e., N2 H2 plasma followed by aqueous solution treatment) is to tailor the topmost surface regions of the hydrophobic OTSSAMs into hydroxyl functional groups with negative charging (or hydrogen bonding) and superhydrophilicity. Doing so can provide points for anchoring ultrafine catalytic particles (3 nm in size) with extremely high adsorption selectivity and distribution density and thus realize the electroless metallization of ultrathin films with a thickness of less than 10 nm.

2. EXPERIMENTAL SECTION 2.1. SAM Preparation and Characterization. 500-nm-thick thermally oxidized (SiO2) dielectric layers on Si (100) wafers were used as substrates. The substrates were cleaned by an SC-1 (NH4OH/H2O2/ H2O = 1:1:5) aqueous solution at 85 °C for 10 min, rinsed thoroughly with ethanol and deionized water successively in an ultrasonic cleaner, and finally dried by a stream of nitrogen gas. Octadecyltrichlorosilane (OTS) with a 95% volume concentration was purchased from SigmaAldrich and used without further purification. The carefully cleaned substrates were immersed in a freshly prepared 25 °C toluene solution containing 1 mM OTS for a duration ranging from 10 s to 4 min, removed from this solution, rinsed twice with toluene and deionized water, and blown dry by a stream of nitrogen. All reagents used were of analytical grade and used as received. Fractional coverage of the OTS monolayers was monitored using an automatic contact-angle analyzer (FTA125, First Ten Angstroms, Inc., USA) to measure the static contact angles of water droplets. A carefully cleaned oxidized Si wafer typically has a contact angle value (θ) of 10°, while one covered by an OTS-SAM is hydrophobic (θ g 110°).26,27 Adsorption of the OTS molecules in the monolayer was further assessed by near-edge X-ray absorption fine structure (NEXAFS) spectroscopy of

ARTICLE

Figure 1. Process steps for site-selective catalytic seeding and electroless deposition starting from dual surface modification by localized vacuum plasma (a,b) and SC-1 aqueous solution (c), followed by cationic seed adsorption (d), cationic seed reduction (e), and selective electroless metallization (f). the C K-edge using the synchrotron radiation source from the 20 A1 beamline (70 1200 eV) at the National Synchrotron Radiation Research Center, Taiwan. The spectra were obtained at the normal incidence configuration. The binding energy scales were referenced to 285.0 eV, which is associated with C 1s spectra of hydrocarbon (CHx) from adventitious contamination.

2.2. Plasma Modification and Surface-Bonding Analysis. The OTS-coated substrates were exposed to the N2 H2 vacuum plasma under the same gaseous mixture, excitation, and discharge conditions described previously,25 but herein, the power density was significantly reduced from 4 to less than 0.2 W cm 2, and the exposure time ranged from a few minutes to less than one second because the OTS-SAMs were only ∼2.7 nm thick.15 Background pressure of the vacuum chamber was 0.5 Pa and optical emission spectroscopy (Emicon system, Plasus, Germany) was employed to monitor the plasma species during the plasma treatments. Impact of the plasma pretreatments on the surface bonding structure was characterized by NEXAFS spectroscopy of the C and N K-edges, together with wetting analysis by water contact-angle measurements. These characterizations allowed the plasma process window to be tuned for the topmost regions of the OTS-SAMs in order to yield appropriate polar groups with superhydrophilicity (θ e 5°) and, when treated by the SC-1 solution, to become preferentially negatively charged. Site-selective seeding and electroless deposition were then performed on the dualsurface-modified regions using the process steps in section 2.3.

2.3. Site-Selective Seeding and Electroless Metallization. Utilizing the dual surface modification (vacuum plasma along with aqueous chemistry), the process to create patterned OTS-SAMs with viable functionality, for site-selective seeding and electroless deposition, is depicted in the six process steps of Figure 1. First, the surfaces of the substrate to be metallized were modified by chemisorption of an OTSSAM using the optimum conditions found in section 2.1. The OTSSAM coated dielectric samples were covered by a copper TEM grid (400 mesh, square type), allowing the unmasked regions to be functionalized by exposure to the vacuum plasma using the optimum process window derived from section 2.2. Doing so resulted in an introduction of hydrophilic functional groups on the topmost region of the SAMs (steps (a) and (b)). Then, the SC-1 aqueous solution was applied in step (c) in an attempt to tailor the hydrophilic functional groups of the plasmaexposed regions to negatively charged sites. Upon immersion in a metal salt (e.g., Co(NO3)2 or Ni(NO3)2) aqueous solution (pH = 5 or 6) in step (d), the negatively charged sites provided points to attract metallic cations in the confined regions. Subsequently, after a complete reduction of the attracted cationic particles by a reducing agent in step (e), the catalytic particles acted as a template for site-selective electroless deposition of Cu or Co-based metallization films, as shown in step (f). 12144

dx.doi.org/10.1021/la202104z |Langmuir 2011, 27, 12143–12148

Langmuir

Figure 2. Evolution of water contact angle for the dielectric samples after immersion in OTS-contained toluene solution for various durations. Electroless deposition of Cu and Co-based alloys was performed in cupric sulfate formaldehyde and cobaltous sulfate sodium hypophosphite solutions, respectively. The process conditions of the SC-1 surface modification, seeding, and electroless deposition, as well as constituents of the baths, are detailed elsewhere.28 The bonding state of the adsorbed metallic cations was analyzed by X-ray photoelectron spectroscopy (PHI 1600, Ulvac, Japan) using Al (Kα) X-ray line at 1486.6 eV as a probe under a background pressure of less than 7  10 8 Pa. Microstructures of the seeds and the plated films were analyzed by a high-resolution scanning electron microscope (SEM; S4800, Hitachi) equipped with an energy-dispersive X-ray spectrometer (EMAX400, Horiba), together with a transmission electron microscope (TEM; 2000, JEOL) operated at an accelerating voltage of 200 kV.

3. RESULTS AND DISCUSSION Deposition of OTS-SAMs on the SiO2 dielectric layers was first monitored by measuring the static contact angles of water droplets on the dielectric layers that were treated at the OTScontained toluene solution for various times. Figure 2 shows that the SiO2 dielectric layers, after thorough clearing by SC-1 and deionized water, were terminated by hydroxyl groups29 and thus became hydrophilic with a water contact angle of 10°. Upon adsorption onto the OH-terminated surfaces, the OTS molecules were tethered to the dielectric layers by a hydrolysis condensation mechanism.30 Therefore, the water contact angle gradually increased due to the accumulation of the nonpolar methyl (CH3) terminal units. Notably, the contact angle was measured five times on different parts of the samples and had a maximum standard deviation of 2° with respect to the mean value. A monolayer was finally formed after immersion for 180 s and thus gave a saturated value of 110°, concurring closely with literature data.26,27 Figure 3a displays a set of carbon K-edge spectra recorded from OTS-SAM coated substrates before a1 and after N2 H2 plasma treatments for 0.5 and 2.0 s (see a2 and a3, respectively). Spectrum a1 reveals that the substrate sample coated with the hydrophobic OTS-SAM indeed exhibited a sharp peak at 288.1 eV and a broad peak at 292.8 eV, in respective agreement with early assignments of the resonance transitions from its constituent C H and C C bonds.31 While it is difficult to discern the detailed changes of the edge structures from the as-recorded spectra in Figure 3a, evolution of the associated surface bonding structures by the plasma exposure was revealed by subtracting the normalized spectrum of a1 from spectra a2 or a3, as displayed in Figure 3b,c, respectively. Figure 3b shows that brief treatment for

ARTICLE

Figure 3. Set of carbon K-edge spectra recorded from OTS-SAM coated substrates before (a1) and after N2 H2 plasma treatments for 0.5 s (a2) and 2.0 s (a3). The spectra for (a2) and (a3) after subtracting the normalized (a1) were shown in (b) and (c), respectively. The dashed line represented the zero level of difference.

0.5 s induced CdC (287.0 eV) and CdN (285.5 eV) peaks, interpreted as an entanglement/breakdown of the well-ordered alkyl chains14 and nitrogen functionalization of the OTS. An additional peak centered at 288.7 eV appeared, attributed to the formation of OdC—OH (carboxyl) polar groups conceivably by the chemical interaction between the cracked hydrocarbon and the plasma activated oxygen-based species residing in the vacuum chamber. Figure 3c reveals that treating for 2 s resulted in an expected further nitrogen functionalization of the CdN bonds, corresponding to the emergence of the N K-edge at 401.0 eV (spectrum not shown). An unexpected introduction of the prominent C—OH groups (289.1 eV), accompanied by the vanishing of the OdC—OH groups, onto the topmost surface region was also observed. Prolonged exposure for 10 s led to a complete ablation of the OTS-SAM. The associated C K-edge spectrum (not shown here) thus solely contained signals from hydrocarbon contamination, identical to that of the as-prepared SiO2 dielectric samples without coating of an OTS-SAM. Thus, NEXAFS analysis as a function of the plasma treatment time confirms that the N2 H2 plasma treatments of the OTS films lead to a progressive nitridation (i.e., CdN formation) and the successive conversion of the dissociated hydrocarbon within the aliphatic chains (C—C, C—H, and CdC) to double-bond OdC—OH groups and finally single-bonded (C—OH) polar functional groups. As is well-known, the choice of plasma gases determines the nature of the functional groups to be incorporated onto the material surface. Generally, exposure of alkyl (e.g., alkylsilanol, alkanethiol, and alkene-derived) SAMs and polymers (e.g., polyethylene) to O2, air,14,17 19 or CO216 plasma leads to the generation of surface oxygenated functionalities (e.g., OH, CdO, or O—CdO polar groups), whereas exposure of those SAMs to NH3 or N2 plasma introduces nitrogen functionalization or amino ( NH2) groups.15 This study found that, upon the exposure of the N2 H2 plasma, the topmost molecular chains of the alkylsilane (OTS) monolayers were successively broken down into OdC—OH and C—OH hydrophilic groups at exposure times of 0.5 and 2.0 s. Indeed, the appearance of the new oxygenated species, as found herein, is typical of alkyl SAMs when exposed to “oxygen-based” plasma. Notably, the plasma treatments were performed in a vacuum system with a background pressure of ∼0.5 Pa under a total absolute pressure of N2 H2 mixed gases at 10 Pa. It is generally known from residual gas analysis that a vacuum environment under such a background pressure is predominantly ∼90% (or higher) from the adsorbed H2O vapor. In situ monitoring of the molecular/atomic species 12145

dx.doi.org/10.1021/la202104z |Langmuir 2011, 27, 12143–12148

Langmuir

Figure 4. SEM micrographs showing that, after seeding and electroless deposition of Cu for 8 s, numerous tiny particles were present on the plasma exposed region (a), whereas scant amounts of particles were found on the grid-covered (pristine OTS-SAM) region (b).

during the plasma pretreatment experiments by optical emission spectroscopy clearly showed the existence of OH (e.g., 306.7 308.9 nm)32 34 and O34 species which had intensities comparable to those of the NH or N plasma species. This finding lends support to the view that, upon desorption, the residual H2O molecules indeed are cracked by the plasma into fragments, e.g., O and OH radicals. These radicals then attract carbon and hydrogen radicals liberated from the OTS surface, subsequently reacting with each other to successively yield the COOH or C OH polar groups, as observed in Figure 3b,c. The accumulation of these polar groups was accompanied by the formation of a waterwetted surface, and thus the associated contact angle significantly reduced from 110° (the pristine OTS-SAMs) to 5° (the hydroxyl terminated SAMs). The studies presented so far allow the optimum trade-off between the monolayers’ degradation and functionalization using the N2 H2 plasma treatment. Under the plasma process window, a duration of no longer than 2 s is required to crack the topmost regions of unmasked monolayers into hydrocarbon fragments, which are successively converted into COOH (0.5 s) and C OH (2.0 s) polar groups through a plasma-dissociated, water-mediated chemical oxidation path. By using the plasma to expose the OTS-coated substrates through a TEM metal grid, the topmost regions of the CH3-terminated OTS monolayers exposed to the plasma were selectively dissociated into waterwetted COOH or C OH groups (Figure 1, steps (b) and (c)). Then, the patterned OTS-SAMs were sequentially treated by the process steps from (c) to (f), starting from surface modification using the SC-1 solution to form negatively charged surfaces with superhydrophilicity (step (c)). After processed by steps (d) and (e) for seeding, SEM imaging analysis, together with energydispersive X-ray microanalysis, showed that X-ray signals of the metallic (e.g., Co or Ni) seeds were obviously detected from the plasma-exposed regions, whereas the signals were obscure in the grid-covered regions (spectra not shown here). This difference indicates site-selective adsorption of seeds. Concurrently, SEM micrographs in Figure 4 confirm that, after electroless deposition of Cu for 8 s by step (f) of Figure 1, many tiny particles existed on the plasma-exposed regions (Figure 4a), but the number of particles was scant on the grid-masked regions (Figure 4b). Notably, as estimated from Figure 4, the seed density for the plasma exposed region is on an order of 1015 m 2, whereas that for the mask-covered region is only 6  1011 m 2. Therefore, the site-controlled seeding process yields a seed-adsorption selectivity of greater than 1000: 1. Figure 4 demonstrates that, after the sequential treatments in aqueous solutions of SC-1 (deprotonation) and nickel (or cobalt) nitrate (catalyst adsorption) by the process steps (c) to

ARTICLE

(e) in Figure 1, both the plasma-induced COOH (0.5 s) and C OH (2.0 s) terminated alkyl chains can effectively bind the catalytic species. First, we discuss the COOH terminated case. As is generally known, protons in hydroxyl groups are detached by an aqueous solution with pH values greater than the pH value of its isoelectric point (pHiep) or its pKa (acid dissociation constant).35,36 The values of pHiep and pKa for carboxylic groups are typically less than 4.37 39 Therefore, during the treatment of the highly basic SC-1 aqueous solution, protons in the COOH functional groups of the plasma-exposed regions indeed are liberated into the solution, and thus the associated surfaces become negatively charged and water wetted (θ = 5°) by OdC O groups. Conversely, the regions of the monolayer, protected by the grids from the plasma exposure, lack wetting (θ = 110°) and remain neutral in charge due to the presence of the nonpolar CH3 headgroups. Upon immersion in an Ni(NO3)2 aqueous solution (pH g 5) by step (d) of Figure 1, the hydrated metallic cations were attracted onto the negatively charged OdC O regions by electrostatic force. In contrast, the adsorption of the metallic cations was inhibited in the OTS-covered regions due to the lack of wetting and mutual electrostatic interaction. Subsequently, after the reduction treatment by step (e) of Figure 1, the catalytic particles were reduced into neutral form and preferentially adsorbed onto the plasma modified OTS-SAM regions (compare Figure 4a and b). By soaking the seed-bearing dielectric samples in an ultrasonic cleaner, the seeds weakly residing on the OTS-covered regions were completely removed, while those on the plasma-exposed regions remained intact. The mechanism invoked for the adsorption of the catalytic species on the C OH terminated alkyl chain is different from that of the COOH terminated one, and deserves further elucidation. Nitrogen contained alcoholic groups have high pKa values of typically ∼10.39 Exposure of the deprotonated alkyl hydroxyl species (C O ) to the aqueous seeding solution (Figure 1, step (d)) thus leads to rapid reprotonation. Therefore, it seems unlikely that many Co(II) or Ni(II) species electrostatically bind to these alkoxides. Indeed, X-ray photoelectron spectroscopy revealed that the sample, after adsorbing Ni(II) species, exhibited a 2p3/2 main peak at 855.6 eV, along with a broad satellite peak at ∼862.0 eV, characteristics of Ni(OH)2.40 This finding indicates that, for the case of C OH terminated alkyl chains, Co(II) and Ni(II) species form complexes with the terminal alcohol species of the monolayer, conceivably by hydrogen bonding between C OH and Ni OH, instead of via electrostatic force. As demonstrated in Figure 5a, the seeds on the micropatterned substrates thus act as a template for the site-selective fabrication of highly resolved Cu metallization patterns using process step (f) of Figure 1. It is evident from the inset in Figure 5a that, for the OTS monolayers patterned by the optimum plasma condition, the metallization pattern was a replica of the grid mask, with the square width (30 μm) and line width (10 μm) essentially identical to the featured sizes of the grids. This sameness implies that the penetration of the plasma-activated radicals and photon radiation into the areas shielded by the grids occurs to a negligible extent. There are many reports on site-selective electroless metallization for a variety of nonconductive substrates with the aid of catalytic particles such as Pd, Au, and Cu.6,13,41,42 Since metallization is confined to the specific surface regions having the plating catalysts, the accuracy of site-selective adsorption and the precise controls over the size and distribution density of the colloidal catalysts are crucial factors to achieving highly resolved 12146

dx.doi.org/10.1021/la202104z |Langmuir 2011, 27, 12143–12148

Langmuir

Figure 5. SEM and TEM analyses demonstrating the feasibility of using the process steps in Figure 1 to fabricate both a well-defined Cu metallization pattern (a) and an ultrathin (8 nm) Co P barrier layer (b), in turn catalyzing the growth of a Cu metallization film.

metallization patterns with a film thickness of, for example, 10 nm. Although widely spread, the activation/sensitization process involves the use of costly PdCl2 and numerous complex additives. When conducted in a conventional manner at a Pd Sn colloidal solution, this process, on one hand, generates catalytic particles with sizes less than 5 nm but with extremely inhomogeneous distribution,43,44 and on the other hand, lacks control over the size and morphology of Pd colloids, thus resulting in (a) an agglomeration of the particle sizes up to 10 nm and (b) a limit of particle densities on the order of 1013 m 2.45 47 An APTS-SAM on SiO2 (or low-k) dielectric layer acts as a trapping/adhesion layer to anchor a uniform distribution of Pd (or Au) particles with a size of 5 nm, acting as a catalyst for electroless deposition of 20-nm-thick Ni B or Ni P diffusion barriers.42,48 In addition, Pd colloidal particles reach a size of 30 nm when anchored by the amino headgroups of a UV-patterned monolayer.36 Notably, the site-selective seeding process presented herein does not incur the demerits of the Pd materials and related aqueous chemistry and can grow tiny catalytic particles other than Pd, such as Cu, Co, or Ni. As evidenced from Figure 4a, the metallic cations were effectively adsorbed by and anchored to the plasma activated sites without agglomeration. Indeed, high-resolution SEM analysis (micrograph not shown here) showed that the catalytic particles had sizes and uniform distribution densities of only 3 nm and 1015 m 2, respectively, which are in sharp contrast to those of the “conventional” seeds. As demonstrated in Figure 5b, the seeds with these refining features have the capacity to serve as a template for electroless deposition of a barrier layer with a thickness of less than 10 nm (using Co P as a test case), a barrier which in turn acts as a catalyst for the electroless metallization of Cu.

4. CONCLUSION The evolution of surface bonding structures for alkyl monolayers under the exposure of N2 H2 vacuum plasma has been examined using OTS-SAMs as a test model. Previous studies have found that (a) nitrogen-based (e.g., NH3 or N2) plasma

ARTICLE

primarily introduces nitrogen functionalization or amino groups and (b) oxygen-based plasma (CO2 or O2) leads to oxygenated surface functionalities (e.g., OH, CdO, or O—CdO polar groups). In contrast, this work provides solid evidence from NEXAFS analysis that the N2 H2 vacuum plasma dissociates the topmost molecular chains of the OTS-SAMs and successively produces COOH and C OH surface functional groups due to the oxygenation of the SAM’s hydrocarbon fragments, by the fragmented species of the residual H2O molecules in the vacuum chamber. Further employing the highly basic (pH g 11.0) aqueous (SC-1) solution detaches the protons from the water-wetted COOH groups of the plasma-patterned regions, forming points of negative electric charging and wettability. It is worth noting that the mechanism invoked for the adsorption of the catalytic species onto the highly basic C OH terminated alkyl layer is hydrogen bonding of the catalyst complexes, e.g., Ni(OH)2. Therefore, the OTS-SAM regions subjected to the dual surface modification act as a template for catalyst growth using the all-wet seeding process. Neither the combination of costly PdCl2 and complex additives nor the demerits of the associated aqueous chemistry (e.g., seed agglomeration, seed sparseness, or seed inhomogeneity) are involved. Therefore, the site-selective seeding process is able to accurately adsorb the catalytic particles (e.g., Co or Ni) and precisely control the size at 3 nm. Simultaneously, there is sufficient particle density for electroless deposition to trigger ultrathin metallization patterns with a thickness of less than 10 nm. The study has presented a nanoseeding process for selectively forming Cu micrometer-patterns with vertical dimensions of less than 10 nm. To narrow down the planar dimensions toward nanometer scales, we have to develop a maskless (or mask) lithographic technique and to overcome the problem of random nucleation on catalyst free areas due to factors such as residual seeds, film contamination, or variation in concentration of plating solution. As shown in this study, a good-quality SAM bearing methyl end groups is lack of wetting, mutual electrostatic force, and hydrogen bonding interaction, and thus could be effective in blocking deposition on the catalyst free regions. These topics are the subjects of another study, and are currently under investigation.

’ AUTHOR INFORMATION Corresponding Author

*Tel: (886) 4-2452-9008. E-mail: [email protected].

’ ACKNOWLEDGMENT The work was supported in part by the National Science Council, Taiwan, under grant number NSC 99-2221-E164-015 and NSC 99-2221-E-035-011-MY3. ’ REFERENCES (1) International Technology Roadmap for Semiconductors (ITRS) 2009; Available online at http://public.itrs.net/. (2) Edelstein, D.; Heidenreich, J.; Goldblatt, R.; Cote, W.; Uzoh, C.; Lustig, N.; Roper, P.; McDevitt, T.; Motsiff, W.; Simon, A.; Dukovic, J.; Wachnik, R.; Rathore, H.; Schulz, R.; Su, L.; Luce, S.; Slattery, J. Proceedings of the IEEE International Electron Devices Meeting, Washington D.C., USA, 1997. (3) Singer, P. Semicond. Int. 2005, 10 (34), 44–46. (4) Yoshino, M.; Nonaka, Y.; Sasano, J.; Matsuda, I.; ShachamDiamand, Y.; Osaka, T. Electrochim. Acta 2005, 51, 916–920. 12147

dx.doi.org/10.1021/la202104z |Langmuir 2011, 27, 12143–12148

Langmuir (5) Friebel, S.; Aizenberg, J.; Abad, S.; Wiltzius, P. Appl. Phys. Lett. 2000, 77, 2406–2408. (6) Zhu, P.; Masuda, Y.; Koumoto, K. J. Mater. Chem. 2004, 14, 976–981. (7) Gorzolnik, B.; Mela, P.; Moeller, M. Nanotechnology 2006, 17, 5027–5032. (8) Chen, M. S.; Brandow, S. L.; Dulcey, C. S.; Dressick, W. J.; Taylor, G. N.; Bohland, J. F.; Georger, J. H.; Pavelchek, Jr. E. K.; Calvert, J. M. J. Electrochem. Soc. 1999, 146, 1421–1430. (9) Sugimura, H.; Hayashi, K.; Amano, Y.; Takai, O J. Vac. Sci. Technol., A 2001, 19, 1261–1265. (10) Jiang, P.; Li, S. Y.; Sugimura, H.; Takai, O. Appl. Surf. Sci. 2006, 252, 4230–4235. (11) F€arm, E.; Kemell, M.; Ritala, M.; Leskela, M. Thin Solid Films 2008, 517, 972–975. (12) Kim, Y. K.; Park, S. J.; Koo, J. P.; Kim, G. T.; Hong, S.; Ha, J. S. Nanotechnology 2007, 18, 015304. (13) Flavel, B. S.; Yu, J.; Ellis, A. V.; Quinton, J. S.; Shapter, J. G. Nanotechnology 2008, 19, 445301. (14) Unger, W. E. S.; Lippitz, A.; Gross, Th.; Friedrich, J. F.; Woll, Ch.; Nick, L. Langmuir 1999, 15, 1161–1166. (15) Tatoulian, M.; Bouloussa, O.; Moriere, F.; Arefi-Khonsari, F.; Amouroux, J.; Rondelez, F. Langmuir 2004, 20, 10481–10489. (16) Delorme, N.; Bardeau, J. F.; Bulou, A.; Poncin-Epaillard, F. Thin Solid Films 2006, 496, 612–618. (17) Medard, N.; Soutif, J. C.; Poncin-Epaillard, F. Langmuir 2002, 18, 2246–2253. (18) Aureau, D.; Morscheidt, W.; Etcheberry, A.; Vigneron, J.; Ozanam, F.; Allongue, P.; Chazalviel, J. N. J. Phys. Chem. C 2009, 113, 14418–14428. (19) Rosso, M.; Giesbers, M.; Schroen, K.; Zuilhof, H. Langmuir 2010, 26, 866–872. (20) Zabetakis, D.; Dressick, W. J. ACS Appl. Mater. Interfaces 2009, 1, 4–25. (21) Blattler, T. M.; Binkert, A.; Zimmermann, M.; Textor, M.; Voros, J.; Reimhult, E. Nanotechnology 2008, 19, 075301. (22) Weng, C. C.; Liao, J. D.; Wu, Y. T.; Tsai, S. C.; Chen, C. H.; Zharnikov, M. J. Vac. Sci. Technol., B 2009, 27, 1949–1957. (23) Chao, C. C.; Wang, T. C.; Ho, R. M.; Georgopanos, P.; Avgeropoulos, A.; Thomas, E. L. ACS Nano 2010, 4, 2088–2094. (24) Chen, S. T.; Chen, G. S.; Yang, T. J J. Electrochem. Soc. 2003, 150, F194–F199. (25) Chen, S. T.; Hsieh, Y. H.; Shih, Y. C.; Hsu, P. W.; Chen, G. S. Electrochem. Commun. 2007, 9, 2764–2767. (26) Dong, J.; Wang., A.; Simon, Ng. K. Y.; Mao, G. Thin Solid Films 2006, 515, 2116–2122. (27) Wang, Y.; Lieberman, M. Langmuir 2003, 19, 1159–1167. (28) Chen, S. T.; Chen, G. S. J. Electrochem. Soc. 2004, 151, D99–D105. (29) Zhuravlev, L. T. Langmuir 1987, 3, 316–318. (30) Sagiv, J. J. Am. Chem. Soc. 1980, 102, 92–98. (31) Peters, R. D.; Nealey, P. F.; Crain, J. N.; Himpsel, F. J. Langmuir 2002, 18, 1250–1256. (32) Jang, I. Y.; Lee, J. K.; Lee, S. H.; Huh, S. M.; Kwon, H. J.; Kim, S. S.; Cho, H. K.; Moon, S. H. J. Electrochem. Soc. 2008, 155, D614–D617. (33) Ionita, M. D.; Teodorescu, M.; Stancu, C.; Stancu, E. C.; Ionita, E. R.; Moldovana, A.; Acsente, T.; Bazavan, M.; Dinescu, G. J. Optoelectron. Adv. Mater. 2010, 12, 777–782. (34) SpecLine 2.0, Database Software, PLASUS, Germany, 2008. (35) Parks, G. A. Chem. Rev. 1965, 65, 177–198. (36) Masuda, Y.; Kondo, M.; Koumoto, K. Cryst. Growth Des. 2009, 9, 555–561. (37) Shyue, J. J.; De Guire, M. R. Langmuir 2004, 20, 8693–8698. (38) Schweiss, R.; Pleul, D.; Simon, F.; Janke, A.; Welzel, P. B.; Voit, B.; Knoll, W.; Werner, C. J. Phys. Chem. B 2004, 108, 2910–2917. (39) Lide, D. R. In Handbook of Chemistry and Physics, 90th ed.; CRC Press: New York, 2009; Section 8. (40) Mu, S; Wu, D.; Wang, Y.; Wu, Z.; Yang, X.; Yang, W. Appl. Mater. Interface 2010, 2, 111–118.

ARTICLE

(41) Yoshino, M.; Masuda, T.; Yokoshima, T.; Sasano, J.; ShachamDiamand, Y.; Matsuda, I.; Osaka, T.; Hagiwara, Y.; Sato, I. J. Electrochem. Soc. 2007, 154, D122–D125. (42) Asher, T.; Inberg, A.; Glickman, E.; Fishelson, N.; ShachamDiamand, Y. Electrochim. Acta 2009, 54, 6053–6057. (43) Burrell, M. C.; Smith, G. A.; Chera, J. J. Surf. Interface Anal. 1988, 11, 160–164. (44) Froment, M.; Queau, E.; Martin, J. R.; Stremsdoerfer, G. J. Electrochem. Soc. 1995, 142, 3373–3377. (45) Chang, S. Y.; Lin, C. W.; Hsu, H.; Fang, J. H.; Lin, S. J. J. Electrochem. Soc. 2004, 151, C81–C88. (46) Lee, C. H.; Kim, J. J. J. Vac. Sci. Technol., B 2005, 23, 475–479. (47) Lau, P. P.; Wong, C. C. Thin Solid Films 2009, 517, 1939–1943. (48) Osaka, T.; Yoshino, M. Electrochim. Acta 2007, 53, 271–277.

12148

dx.doi.org/10.1021/la202104z |Langmuir 2011, 27, 12143–12148