Patent Application
20140364574
The present invention relates to carbon nanotube (CNT) deposition processes and more particularly, to techniques for CNT solubilization and surface-selective deposition via polymer-mediated assembly.
Carbon nanotubes (CNTs) are promising candidates for many different applications such as sensors, supercapacitors, electrodes, drug-delivery, and digital logic. However, one factor limiting the widespread application of CNTs is that many of these applications would require the selective deposition of CNTs from solution onto specific areas of a substrate without covering the whole substrate with a blanket film of CNTs. The selective deposition of CNTs has been difficult, due to: (1) the low solubility of CNTs in most solvents; (2) the difficulty of adapting traditional solution-based processes for the fabrication of CNT thin films; and (3) the lack of proper techniques capable of yielding high-density deposition of CNTs with high selectivity for certain areas of a substrate.
One possible approach for the controlled deposition of CNTs is to use directed self-assembly (DSA), which is an interesting strategy for the fabrication of artificial supramolecular structures or for the modification of surfaces. DSA has been reported using building blocks as diverse as DNA or nanoparticles.
Even though there are a number of examples for the DSA of CNTs with high surface selectivity, these methods have one or several of the following drawbacks: (1) they require covalent chemical modifications that can affect (and damage) the optical and electronic properties of the CNTs; (2) the CNT deposition yield is low and the films have a very low density of CNTs; (3) current methods rely on the use of surfactants which have weak hydrophobic interactions with the CNTs, and the presence of free surfactant affects the DSA process, lowering the yield of CNT deposition.
For example, H. Park et al., “High-density integration of carbon nanotubes via chemical self-assembly,” Nature Nanotechnology 7, 787-791 (October 2012) (hereinafter “Park”), the contents of which are incorporated by reference herein, describes solubilization of CNTs in water using sodium dodecyl sulfate (SDS) as a surfactant. See also, U.S. Patent Application Publication Number 2013/0082233 A1, filed by Afzali-Ardakani et al., entitled “Selective Placement of Carbon Nanotubes Via Coulombic Attraction of Oppositely Charged Carbon Nanotubes and Self-Assembled Monolayers” (hereinafter “U.S. Patent Application Publication Number 2013/0082233”) the contents of which are incorporated by reference herein.
While the SDS surfactant-based approach is effective in a variety of applications, there are some notable drawbacks to the process. First, the interaction between the SDS and the CNTs is relatively weak, which means that there is a dynamic equilibrium between the surfactant molecules on the CNT surface and the free SDS surfactant in solution. The exact amount of free SDS is dependent on the deposition conditions (concentration, temperature . . . ) and thus can cause irreproducibility in the deposition process. Second, since free SDS will be competing with the SDS-coated CNTs for deposition on the positively charged surface, the presence of free SDS surfactant can lower the yield of CNT deposition. Third, excess surfactant is hard to detect with traditional chemical analysis methods based on electrical and spectroscopic measurements. Fourth, in order to remove the maximum amount of free surfactant possible before CNT deposition, a dialysis is usually required, which is an additional processing step, and which also has the setback of being time/concentration dependent (potentially causing irreproducibility problems from batch to batch). Fifth, removal of excess surfactant usually leads to CNT aggregation, which can cause a decrease in the deposition yield of individual tubes. Sixth, in order to decrease CNT deposition times, and so that reservoir solutions of CNT/surfactant can be used several times for CNT deposition, it is desirable to have CNT solutions with a relatively high concentration. However, high concentrations of CNTs cannot be achieved without using an excess of SDS surfactant, which is detrimental for deposition purposes. Seventh, since the surfactant is not very strongly bound to the surface of the CNTs, the amount of free surfactant can vary over time, which ultimately leads to irreproducibility as well.
Accordingly, improved CNT solubilization and surface-selective deposition techniques that can be used for the reproducible fabrication of homogeneous thin films of CNTs from solution, while coating only specific areas of a substrate would be desirable.
The present invention provides techniques for carbon nanotube (CNT) solubilization and surface-selective deposition via polymer-mediated assembly. In one aspect of the invention, a method for self-assembly of carbon nanotubes on a substrate is provided. The method includes the following steps. A charge is created on one or more surfaces of the substrate. The substrate is contacted with carbon nanotube-polymer assemblies dispersed in a solvent, wherein the carbon nanotube-polymer assemblies include the carbon nanotubes wrapped in a polymer having side chains with charged functional groups, and wherein by way of the contacting step the carbon nanotube-polymer assemblies selectively bind to the charged surfaces of the substrate based on complementary electrostatic interactions between the charged functional groups on the polymer and the charged surfaces of the substrate and thereby self-assemble on the substrate.
In another aspect of the invention, a structure is provided including: carbon nanotubes on a substrate, wherein the substrate includes one or more charged surfaces and wherein the carbon nanotubes are wrapped in a polymer so as to form carbon nanotube-polymer assemblies, the polymer having side chains with charged functional groups, and wherein the carbon nanotube-polymer assemblies are selectively bound to the charged surfaces of the substrate based on complementary electrostatic interactions between the charged functional groups on the polymer and the charged surfaces of the substrate.
A more complete understanding of the present invention, as well as further features and advantages of the present invention, will be obtained by reference to the following detailed description and drawings.
Provided herein are techniques for the selective deposition of thin films of carbon nanotubes (CNTs) on specific surfaces using polymer-mediated directed self-assembly. With the present techniques, a combination of CNT/polymers can be used for CNT deposition and further device fabrication. The polymers have a backbone capable of binding strongly to the surface of the CNTs. These polymers have side-chains with charged functional groups capable of solubilizing the CNTs via interaction with the solvent and these same functional groups can also be used for the selective deposition of CNTs on specific areas of a given substrate. The interactions responsible for this selectivity are complementary electrostatic interactions between the functional groups and those areas of the substrate. See
A substrate 110 is provided having specific areas 112 having a (e.g., positive) surface charge complementary to the charged side chains 104a on the polymer 102. Thus, once the CNT/polymer assemblies 108 are deposited onto the specific areas 112 of the substrate 110, the CNT/polymer assemblies 108 will self-assemble on the substrate (mediated by the polymer)—thus leading to selective deposition of the CNTs on only desired portions of the substrate.
As will be described in detail below, these functional polymers are used to disperse the CNTs, and modular modifications on the chemical structure of the polymer's side-chain have a large impact on the surface properties of the polymer-coated CNTs. These slight chemical modifications highly influence CNT solubility and enable surface-selective CNT deposition from solution using directed self-assembly based on electrostatic interactions involving the functional groups on the side-chain of the polymer, e.g., poly(thiophene)s with functional side-chains bearing negatively charged functional groups.
As shown in
According to an exemplary embodiment, the functional polymers are regioregular polythiophene 202 with charged phosphonate side-chains. See
In one exemplary implementation of the present techniques, poly(thiophene)s 202 with functional side-chains bearing negatively charged functional groups were synthesized from a parent poly(thiophene) polymer 306 with alkylbromide side-chains. See
The desired functional groups were incorporated by post-polymerization modifications using a nucleophilic substitution reaction, which is a modular approach for introducing a wide variety of chemical moieties into the side-chains of conjugated polymers. See L. Zhai et al., “A Simple Method to Generate Side-Chain Derivatives of Regioregular Polythiophene via the GRIM Metathesis and Post-polymerization Functionalization,” Macromolecules, 2003 36, 61-64 (Published December 2002), the entire contents of which are incorporated by reference herein. In order to obtain water-soluble, negatively-charged poly(thiophene) 202 to enable DSA of CNTs using electrostatic interactions, the bromide-substituted poly(thiophene) 306 was subjected to a nucleophilic substitution reaction using triethylphosphite followed by phosphonic ester hydrolysis and deprotonation with a tetraalkylammonium hydroxide salt.
In order to study the effect of the degree of side-chain substitution for polymer 202 on the self-assembly of CNTs, random regioregular polythiophenes with different degrees of side-chain functionalization were synthesized using GRIM polymerization conditions using a mixture of 2,5-dibromo-3-hexylthiophene 304 and 2,5-dibromo-3-(6-bromohexyl)thiophene monomers. Treatment of the resulting copolymer 310 with triethyl phosphite followed by ester hydrolysis yielded the corresponding copolymer 312 with different degrees of phosphonic acid functionalization in the side chain. The resulting degree of side-chain functionalization was determined by the initial feed ratio of the monomers. Copolymers 302 where 5% and 50% of the repeat unit side-chain was substituted were thus obtained.
Next, in step 404, CNTs (e.g., single-walled carbon nanotubes (SWCNTs) are contacted with the polymer dispersion (from step 402). For example, the CNTs can be added to the polymer dispersion and sonication used to aid in mixing. As described above, the present functional polymers have charged side chains and a backbone capable of wrapping around CNTs to form CNT/polymer assemblies. Thus these CNT/polymer assemblies are formed in step 404, dispersed in the solvent.
According to an exemplary implementation of the present techniques, once mixed, centrifugation was used to remove insoluble CNT bundles, and excess polymer in the supernatant was removed by filtration of the CNT dispersion through a 0.22 micrometer cellulose acetate filter and washing with excess water. Excess polymer in the filtrate solution can be easily detected by UV-Vis spectroscopy. The CNT/polymer assemblies in the filter were resonicated in water and the resulting aqueous dispersion was stable over time. Concentrations of up to several mg/ml of SWCNTs in water could be obtained this way.
As provided above, a goal of the present process is the surface-selective, polymer-mediated assembly of the CNTs on the surface of a given substrate. This is based on the presence of complementary charges on the functionalized polymer and on the substrate. Thus, in step 406, a surface charge is created on those surfaces of the substrate on which assembly of the CNTs is desired. It is notable that while in the instant example surface-selective deposition is employed to deposit CNTs on certain areas of the substrate, if so desired the same processes described herein can be employed to deposit CNTs on the entire surface of a substrate. According to an exemplary embodiment, when the CNT/polymer assemblies contain negatively charged side chains, the substrate can be selectively coated with a monolayer of positively charged molecules. It is notable that the steps of methodology 400 do not need to be performed in the order depicted in
In one exemplary embodiment, the surface charge is selectively created on one or more surfaces of the substrate by forming a monolayer of a compound bearing a (positive or negative) charge on those surfaces. Further, the present techniques can leverage the fact that certain compounds, such as hydroxamic acid, selectively interact with certain metal oxides or other materials (silicon nitride, aluminum oxide) over silicon dioxide. See, for example, J. P. Folkers et al., “Self-assembled monolayers of long-chain hydroxamic acids on the native oxides of metals,” Langmuir, 11, 813-824 (March 1995) (hereinafter “Folkers”), and H. Park et al., “High-density integration of carbon nanotubes via chemical self-assembly,” Nature Nanotech., 7, 787-791 (December 2012). The entire contents of each of the foregoing references are incorporated by reference herein. See also, U.S. Patent Application Publication Number 2013/0082233 which describes an exemplary process for decorating a surface of a substrate with a charge. In one exemplary embodiment, the monolayer includes a positively charged pyridinium salt bearing a hydroxamic acid moiety, NMPI (4-(Nhydroxycarboxamido)-1-methylpyridinium iodide).
Hydroxamic acids and phosphonic acids selectively bind to surfaces which are relatively basic, but they do not bind to surfaces which are more acidic. This can be thought of as an acid/base reaction. For instance, hafnium oxide (HfO2) is relatively basic, so hydroxamic/phosphonic acid binds to it, and silicon dioxide is relatively acidic, so hydroxamic/phosphonic acid does not bind to it. Other examples of surfaces to which hydroxamic acids and phosphonic acids bind (selectively over binding to silicon dioxide) are silicon nitride and aluminum oxide.
It is believed that surfaces with an isoelectric point greater than the pKa of the acid (hydroxamic, phosphonic) used for the self-assembly will give better directed self-assembly in general, due to deprotonation of the acid. See, for example, Folkers. This is true for silicon nitride, aluminum oxide, hafnium oxide. Conversely, surfaces with an isoelectric point less than the pKa of the acid (hydroxamic, phosphonic) used for the self-assembly will give worse/no directed self-assembly. This is true for silicon dioxide.
Thus, in general, according to the present techniques—those surfaces for which self-assembly of the monolayer is desired will be formed from a material having an isoelectric point greater than the pKa of the acid (hydroxamic, phosphonic) used for the self-assembly, and those surfaces for which self-assembly of the monolayer is not desired will be formed from a material having an isoelectric point that is less than the pKa of the acid. Non-limiting examples include hafnium oxide and silicon dioxide, respectively.
Thus, when hydroxamic acid is used for the self-assembly, the starting wafer can be, for example, a semiconductor wafer having a layer of a first material with an isoelectric point that is less than the pKa of hydroxamic acid, such as SiO2. A patterned layer of a second material with an isoelectric point that is greater than the pKa of hydroxamic acid, such as silicon nitride, HfO2 and/or aluminum oxide (Al2O3), can be formed on the first material using conventional lithography and etching techniques. As noted above, it may be instead desired to cover the entire surface of the substrate (as opposed to selective areas) with the CNTs, in which case, the whole surface of the substrate would be decorated with the second material. Basically, the second material should be present anywhere on the substrate where deposition of the CNTs is desired.
Namely, the pattern of the second material will dictate where on the surface of the wafer the CNTs will be present since, 1) the charged monolayer will interact with the second material (and not the first material) on the wafer, and 2) the CNT/polymer assemblies will self-assemble on the monolayer based on coulombic attraction between the complementary surface charges of the CNT/polymer assemblies and the charged monolayer. In this example, according to step 406, the surface charge can be selectively created on the (second material) surfaces of the substrate by contacting the substrate with hydroxamic acid (e.g., immersing the substrate in a hydroxamic acid solution or drop casting the hydroxamic acid on the substrate) which will form the charged monolayer on the (second material) surfaces.
See, for example,
Referring back to
As highlighted above, the present process can be implemented to form a layer of CNTs or an array of individual CNTs on a substrate. When formed as a layer, the CNTs are not regularly aligned, and overlap and/or contact one another. By contrast, in an array, the CNTs are regularly aligned, and do not overlap or contact one another. An array of individual CNTs may be preferable when devices are being fabricated which require individual CNTs to form the device. For instance, when the device features that are capable of interacting with the CNTs specifically are small enough (in the order of magnitude of about 100 nanometers or less), only one CNT fits inside the feature. After deposition of an array of individual CNTs, regular electrode deposition steps used for CMOS fabrication can be used to obtain transistors of each individual CNT in the array. There are instances in which obtaining an array of those individually placed CNTs would be advantageous, for instance if one wants to fabricate arrays of transistors of individual CNTs. This is interesting for digital logic applications based on CNTs.
Whether a layer of CNTs or an array of individual CNTs is formed on the substrate can be regulated by tailoring the above-described process. For instance, to attain an array of individual CNTs (as opposed to a layer), the substrate surface modifications (described above) can be configured such that the features are small enough that only a fraction of the CNTs interacts. Additionally, the carbon nanotube-polymer assembly dispersion can be diluted (e.g., with additional solvent) to reduce the number of CNTs deposited on a given area of the substrate. These techniques can be employed individually, or in combination with one another.
A process for surface-selective polymer-mediated assembly of CNTs via interactions between (e.g., thiol) functional groups on the polymer and surfaces of a substrate coated with a material configured to interact with the functional groups are described, for example, in U.S. patent application Ser. No. 13/912,417, filed on Jun. 7, 2013, entitled “Surface Modification Using Functional Carbon Nanotubes,” the entire contents of which are incorporated by reference herein.
The present techniques have the following advantages: (1) the binding between the present functionalized polymer and the CNTs is via pi-pi stacking, which is relatively strong—this means that there is no dynamic equilibrium between the polymer molecules on the surface of the CNTs and the free polymer in solution; (2) excess polymer can be removed by washing the polymer/CNT assemblies with excess solvent, which is desirable for CNT deposition purposes, because excess free polymer could result in deposition of just the excess polymer on the charged substrate and not deposition of the CNTs, since both are charged species competing for the charged substrate; (3) excess surfactant is easy to detect, since the polymers are colored and can also be fluorescent as is the case with conjugated polymers such as poly(thiophene)s, which allows for their detection using standard spectroscopic techniques; (4) in order to remove the maximum amount of free polymer possible before CNT deposition, dialysis is not required, since the excess polymer can be easily washed away by filtration; (5) removal of excess surfactant usually does not lead to nanotube aggregation, since the fraction of the polymer directly interacting with the CNT walls is hard to remove using common solvents for washing; (6) high concentrations of CNTs in solution can be achieved without having to use excess polymer, since a thin coating of polymer will be in close contact with the CNT walls and preventing the CNTs from aggregating; (7) since the polymer is strongly bound to the surface of the nanotubes, CNT/polymer dispersions are highly stable over time—this means that the amount of free polymer will not vary over time; (8) the side-chain of the polymer can be fine-tuned and modified to promote different kinds of interactions with different substrates.
The following non-limiting example illustrates the above-described process for functionalizing the surface of a substrate and polymer-mediated assembly of the CNT/polymer assemblies on the functionalized substrate in accordance with the present techniques. In this example, the substrate employed was a patterned substrate of HfO2 and silicon dioxide SiO2 (see, for example,
The substrate was then exposed to a solution of the above-described CNT/polymer assemblies for two hours, followed by rinsing with water to remove excess solution, and CNT deposition was confirmed by scanning electron micrograph (SEM). See
The CNTs were self-assembled only on the monolayer-coated HfO2 areas of the substrate with very high selectivity, due to the coulombic attraction between the positively charged substrate and the negatively charged CNT/polymer assemblies. The deposition density on HfO2 was very high as shown in the SEM images, which is desired for self-assembly purposes. See, for example,
When substrates that had not been covered with a positively charged monolayer were exposed to the solution of CNT/polymer assemblies, the CNTs were selectively deposited on the HfO2 as observed by SEM, but the density was comparatively low, as shown in
When poly(3-hexylthiophene) (P3HT) was used instead of regioregular polythiophene polymers with charged phosphonate side-chains for the directed self-assembly of CNTs, then no selectivity was observed (see
The effect of the degree of functionalization in the polymer side chain was also studied for the present functionalized polymers. Copolymers with 5% of phosphonic acid in the side-chain (polymer 312, 100*b/(a+b)=5) were only soluble in organic solvents such as CHCl3, and copolymers with 50% phosphonic acid in the side-chain were only slightly soluble in water in their deprotonated form (polymer 302, 100*b/(a+b)=50).
Directed Self-Assembly of CNTs was tested using polymers containing 5% phosphonate side-chains on the HfO2/SiO2 substrates where the HfO2 was coated with a positively charged monolayer, which resulted in non-selective placement of the CNTs. See
On the other hand, when copolymers that contained 50% of phosphonate substitution in the side-chain were used for the directed self-assembly of CNTs from water, selective placement was observed on the HfO2/SiO2 substrates. However, the CNTs were aggregated and the placement density was very low as shown in
The present techniques are further illustrated by way of reference to the following non-limiting example:
Instruments: NMR spectra were obtained on a Bruker Avance (400 MHz). NMR chemical shifts are given in parts per million (ppm) referenced to CHCl3/tetramethylsilane (TMS) (7.24 ppm for 1H). Polymer molecular weights were determined at room temperature on a Waters 2695 GPC system in tetrahydrofuran (THF) at 1.0 mL/min (1 mg/mL sample concentrations), approximate molecular weights were estimated using a polystyrene calibration standard. Scanning Electron Microscopy (SEM) images were obtained using a LEO 1560 at 1 keV, 20 μm aperture. Substrates were cleaned with a plasma cleaner from Harrick Plasma, model PDC-32G. MiliQ water was obtained using a Q-Pod from Milipore. The probe sonicator used for these experiments was a SONICS from Vibra Cell. CNT bundles were removed by centrifugation using an IEC Centra CL2.
Self-Assembly of CNT/polymer 1 from water: For the deposition of CNTs wrapped in the polymer (regioregular polythiophene with charged phosphonate side-chains), 3 milligrams of the phosphonic acid version of the polymer were dispersed in 10 milliliter miliQ water, to which 5 milligrams of tetramethylammonium hydroxide pentahydrate were added. The polymer was sonicated until completely dissolved, and 2 milligrams of CNTs were added. The resulting mixture was sonicated using a probe sonicator for 45 minutes. The CNT dispersion was centrifuged for 30 minutes to remove CNT bundles and the supernatant was filtered through a cellulose acetate filter (0.22 micrometer (μm) pore size). The filter was washed with 100 milliliters of miliQ water to remove excess polymer and tetramethylammonium hydroxide and the residue was resonicated into 10 milliliters of miliQ water with a bath sonicator. The CNT dispersion was sonicated using a probe sonicator for 45 minutes, followed by centrifugation for 30 minutes to remove CNT bundles. The resulting supernatant was used for the directed self-assembly of the CNTs. Patterned substrates of HfO2 on SiO2 were plasma cleaned for 5 minutes and dipped in a 100/1 HF/H2O solution followed by thorough rinsing with water and drying under nitrogen. The HfO2 parts of the substrate were selectively coated with a monolayer of positively charged molecules (NMPI) by immersion for 2 hours in an NMPI solution (24 milligrams of NMPI in 9 milliliters EtOH, 3 milliliters H2O), followed by rinsing with water. Exposure of these substrates to the CNT/polymer solution for 2 hours led to the self assembly of thin films of CNTs with very high density and selectivity in the HfO2 parts of the substrate.
Polymer Synthesis: Unless otherwise noted, all reactions were performed in oven-dried glassware, and under an oxygen-free atmosphere of nitrogen. Polymerizations were carried out using standard Schlenk techniques. Anhydrous solvents and all other chemicals were obtained from Sigma-Aldrich and were used as received. 2,5-dibromo-3-(6-bromohexyl)thiophene2 and NMPI1 were synthesized as previously described in the literature. CNTs (ASP) were obtained from Hanwha Nanotech and used as received.
General procedure for the synthesis of poly(thiophene)s with alkylbromide side-chains: An exemplary process for the synthesis of poly(thiophene)s with alkylbromide side-chains is shown in
For the random copolymer of poly(3-hexylthiophene) and poly(3-(6-bromohexylthiophene)) 310 (a=95%, b=5%), 85% yield. Mn=22K, PDI=1.3, 1H-NMR (CDCl3, ppm): 0.80-1.20 (bt, 3H), 1.20-1.60 (bm. 4.2H), 1.82 (bt, 2.1H), 1.90 (bt, 0.1H), 2.81 (t, 2.1H), 3.41 (bt, 0.1H), 6.96 (bs, 1.1H).
For the random copolymer of poly(3-hexylthiophene) and poly(3-(6-bromohexylthiophene)) 310 (a=50%, b=50%), 85% yield. Mn=23K, PDI=1.3, 1H-NMR (CDCl3, ppm): 0.80-1.20 (bt. 3H), 1.20-1.60 (bm. 10H), 1.82 (bt, 4H), 1.90 (bt, 2H), 2.81 (t, 4H), 3.41 (bt, 2H), 6.96 (bs, 2H).
General procedure for polymer side-chain modification to obtain phosphonated polythiophenes: An exemplary process for polymer side-chain modification to obtain phosphonated polythiophenes is shown in
For the precursor of polymer 202, 1402 (a=0, b=n), 95% yield. Mn=25K, PDI=1.3, 1H-NMR (CDCl3, ppm): 1.3-1.8 (bm, 16H), 2.5 (bt, 2H), 4.1 (bm, 4H), 6.95 (bs, 1H). This polymer with diethylphosphonate side-chains was dissolved in 3 milliliters of dichloromethane and TMS-Br was added drop-wise. The reaction was stirred at room temperature overnight, after which volatiles were removed under vacuum. The resulting polymer was sonicated in 3 milliliters of methanol and stirred at room temperature for 4 hours, after which volatiles were once again removed under vacuum. The resulting polymer with phosphonic acids in all side-chains 312 was insoluble in water and organic solvents but could be taken into aqueous solutions of tetramethylmmonium hydroxide to yield polymer 202.
Although illustrative embodiments of the present invention have been described herein, it is to be understood that the invention is not limited to those precise embodiments, and that various other changes and modifications may be made by one skilled in the art without departing from the scope of the invention.
This application is a continuation of U.S. application Ser. No. 13/912,403 filed on Jun. 7, 2013, the disclosure of which is incorporated by reference herein.
| Number | Date | Country | |
|---|---|---|---|
| Parent | 13912403 | Jun 2013 | US |
| Child | 14031668 | US |
