Patent Application
20250054755
Semiconducting carbon nanotubes (CNTs) are one-dimensional materials with exceptional electronic properties. Field effect transistors fabricated with CNT-based channels have been shown to exhibit current density greater than silicon and gallium arsenide transistors. (Brady, Gerald J., et al., “Quasi-ballistic carbon nanotube array transistors with current density exceeding Si and GaAs”, Science Advances, 2017, Volume 2, Issue 9) CNTs' high current carrying capacity and the ability to easily modulate their conductivity due to their small diameter make them promising materials for next generation communication and logic devices. Maximizing the performance of CNT-based transistors requires that the CNTs in the transistors are aligned and arranged into densely packed monolayer arrays.
High purity semiconducting CNTs are available as suspensions in aqueous and organic solvents, referred to as inks. The fabrication and deposition of aligned arrays of CNTs onto substrates from inks has been investigated using shear, vacuum filtration, directed evaporation, dielectrophoresis, evaporative self-assembly, blown-bubble assembly, gas flow self-assembly, spin-coating, contact-printing assembly, elastomeric release, dimension-limited self-alignment, DNA directed assembly, and Langmuir-Blodgett and -Schaefer methods, among others.
Some alignment methods accumulate CNTs at the interface between two immiscible liquids or between a liquid and a gas. CNTs can self-align at such interfaces, and the aligned CNTs can be picked up or deposited onto substrates. For example, it has been observed that spreading thin layers of polymer wrapped CNTs dispersed in organic solvents, like chloroform, on top of a water bath results in the self-alignment of the CNTs at the interface between the chloroform and water (Joo, Yongho, et al., “Dose-Controlled, Floating Evaporative Self-assembly and Alignment of Semiconducting Carbon Nanotubes from Organic Solvents,” Langmuir, 2014, Volume 30, Issue 12, pages 3460-3466), (Jinkins, Katherine R., et al., “Nanotube Alignment Mechanism in Floating Evaporative Self-Assembly, Langmuir, 2017, Volume 33, Issue 46, pages 13407-13414.). Using this effect, CNT assemblies can be transferred to large-area substrates by continuously flowing thin layers of the CNT ink over the surface of an aqueous subphase and simultaneously lifting a target substrate through the flowing ink layer at constant velocity—in a process termed tangential flow interfacial self-assembly (TaFISA).
Methods of forming films of aligned elongated nanoparticles, films made using the methods, and electronic devices, such as transistors, that incorporate the films are provided. Also provided are methods of forming a patterned substrate for use in the film-forming methods.
One embodiment of a method of forming a film of aligned elongated nanoparticles on a substrate includes the steps of: providing a substrate having at least one liquid film-adsorbing surface region bounded by at least one liquid film-repelling surface region; forming a liquid film on the at least one liquid film-adsorbing surface region, wherein the liquid film is bounded by the at least one liquid film-repelling surface region: contacting a suspension that comprises dispersed elongated nanoparticles with the at least one liquid film, wherein the suspension is immiscible with the at least one liquid film, such that the at least one liquid film and the suspension form an interface and elongated nanoparticles are transferred from the suspension to a surface of the at least one liquid film at the interface; and allowing the at least one liquid film to dissipate, whereby the elongated nanoparticles are deposited on the at least one liquid film-adsorbing surface region along a contact line defined by: the liquid film: the substrate; and the suspension or air, as the liquid film dissipates.
One embodiment of a method of patterning a silicon dioxide surface of a substrate with one or more hydrophilic surface regions bounded by one or more less hydrophilic surface regions includes the steps of: forming a layer of a resist on the silicon dioxide surface; patterning the layer of resist to expose one or more regions of the silicon dioxide surface through the layer of resist: depositing a layer of yttrium metal or a layer of scandium metal over the layer of resist and the one or more exposed regions of the silicon dioxide surface to form yttrium metal-coated or scandium metal-coated surface regions on the silicon dioxide surface: removing the patterned layer of resist from the silicon dioxide surface: oxidizing the yttrium metal or scandium metal on the silicon dioxide surface: forming a layer of material having a lower hydrophilicity than the silicon dioxide on the silicon dioxide surface and over the oxidized yttrium or oxidized scandium; and removing the oxidized yttrium or oxidized scandium from the silicon dioxide surface to form one or more hydrophilic silicon dioxide surface region bounded by the material having the lower hydrophilicity than the silicon dioxide.
One embodiment of a field effect transistor includes: a source electrode: a drain electrode: a gate electrode; and a conducting channel in electrical contact with the source electrode and the drain electrode, the conducting channel comprising a film comprising aligned semiconducting single-walled carbon nanotubes. Some embodiments of the field effect transistor are characterized by a current density of at least 1.8 mA μm−1 and a transconductance of at least 1.14 mS μm−1, as measured at 0.6 V drain voltage.
Other principal features and advantages of the invention will become apparent to those skilled in the art upon review of the following drawings, the detailed description, and the appended claims.
Illustrative embodiments of the invention will hereafter be described with reference to the accompanying drawings, wherein like numerals denote like elements.
The images in
Methods of forming films of aligned elongated nanoparticles, films made using the methods, and electronic devices, such as transistors, that incorporate the films are provided. Also provided are methods of forming a patterned substrate for use in the film-forming methods.
In the methods, elongated nanoparticles, such as carbon nanotubes, floating at the surface of a liquid film are deposited onto a surface of a substrate as the liquid film dissipates on the surface. The alignment and deposition of the elongated nanoparticles occurs along a contact line that is defined either by the liquid film, the substrate, and a second liquid that is immiscible with the liquid film or by the liquid film, the substrate, and air. As the liquid film dissipates, the contact line recedes across the surface of the substrate and the elongated nanoparticles are deposited from the contact line onto the substrate surface in the form of a nanoparticle film.
The methods can produce large-area (e.g., wafer-scale) patterned films of elongated nanoparticles with exceptionally high nanoparticle alignment and linear packing densities in a continuous and scalable process. The degree of alignment of the elongated nanoparticles in the deposited films refers to their degree of alignment along their long axes within the film, which can be quantified using the methods described in the Example. In some embodiments, the deposited films have an elongated nanoparticle mean degree of alignment of ±10° or better. (That is, the mean of the magnitude of misalignment for the elongated nanoparticles in the deposited film is no greater than 10°.) This includes elongated nanoparticle films in which the elongated nanoparticles have a mean degree of alignment of +6° or better.
The density of the elongated nanoparticles in the films can be measured by their linear packing density, which can be quantified in terms of the number of elongated nanoparticles per μm and measured as described in the Example. In some embodiments, the method described herein deposit films having an elongated nanoparticle linear packing density of at least 30 elongated nanoparticles/μm. This includes embodiments in which the films have an elongated nanoparticle linear packing density of at least 50 elongated nanoparticles/μm, at least 100 elongated nanoparticles/μm, and at least about 200 elongated nanoparticles/μm.
The high nanoparticle packing densities and degree of alignment that can be achieved using the methods render films made by the methods suitable for use in a variety of high-performance electronic applications, including as channel layers in FETs. The high density and degree of alignment that can be achieved in semiconducting single-walled carbon nanotube (s-SWCNT) films made using the present methods, enables the fabrication of FETs having a current density of at least 1.8 mA μm−1 and a transconductance of at least 1.14 mS μm−1 (measured at 0.6 V drain voltage). This includes FETs having a current density of up to 2.2 mA μm−1 and a transconductance of up to 1.4 mS μm−1, or even higher (measured at 0.6 V drain voltage). More details regarding the fabrication and characterization of such high-performance FET are provided in the Example.
The methods are carried out on a substrate having a surface with at least one liquid film-adsorbing surface region, the boundaries of which are defined by one or more liquid film-repelling surface regions. The liquid film-adsorbing and liquid-film-repelling surface regions are characterized by a difference in their hydrophilicity, such that a liquid applied to the surface adsorbs to the liquid film-adsorbing regions, but not to the liquid film repelling regions. Depending upon the liquid being used, the liquid film-forming regions may by more hydrophilic than the liquid film-repelling regions or vice versa. For example, if the liquid films comprise an aqueous medium or comprise a hydrophilic organic solvent, the liquid film-adsorbing regions will be more hydrophilic than the liquid film-repelling regions. (For the purposes of this disclosure, the term “aqueous medium” refers to pure water, as well as aqueous solutions in which one or more solutes are dissolved in water and aqueous solutions that include one or more organic solvents as minority cosolvents. For example, solutes that depress the melting point of water or antibacterial agents may be dissolved in water to provide an aqueous medium.) However, if the liquid films comprise a hydrophobic organic solvent, the liquid film-adsorbing regions will be less hydrophilic than the liquid film-repelling regions. It should be noted that the description of the liquid film adsorbing and liquid film repelling regions as having “a difference in their hydrophilicity” does not imply that both regions are hydrophilic (i.e., having a water contact angle of 0° or greater). Rather, it is possible for both the liquid film-adsorbing and liquid film-repelling regions to be hydrophilic, for both the liquid film-adsorbing and liquid film-repelling regions to be hydrophobic (i.e., having a water contact angle of 0° or greater), for the liquid film-adsorbing regions to be hydrophilic and the liquid film-repelling regions to be hydrophobic, or for the liquid film-adsorbing regions to be hydrophobic and the liquid film-repelling regions to be hydrophilic. The key is that the difference in the hydrophilicities of the two regions be sufficiently different to allow for selective liquid film retention on the liquid film-adsorbing regions.
As used herein, a hydrophilic surface region is a surface region having a water contact angle of equal to or less than 90° and a hydrophobic surface region is a surface region having a water contact angle (WCA) of greater than 90°, as measured using the sessile drop method at the temperature at which film formation is being conducted. For film deposition at room temperature, this temperature would be ˜23° C. For purposes of illustration only, in some embodiments of the methods both the liquid film-adsorbing and liquid film-repelling surface regions are hydrophilic by the above definition of hydrophilicity, but the liquid film-adsorbing regions are more hydrophilic. For example, a liquid film-adsorbing region can have a WCA of less than 10° and a liquid film-repelling region can have a WCA in the range from 25° to 60° or vice versa.
One embodiment of a method of forming a film of aligned elongated nanoparticles is illustrated schematically in
A liquid film 108 is formed on each of the one or more liquid film-adsorbing surface regions 104 by contacting the substrate surface with liquid that adsorbs to the liquid film-adsorbing surface regions to form a liquid film thereon. The substrate surface may be contacted with the liquid by flowing the liquid over the substrate surface, by submerging the substrate in the liquid, by spraying the liquid onto the substrate surface, by printing the liquid onto the substate surface, or by other means. The liquid is selectively retained on the one or more liquid film-adsorbing surface regions, but not on the liquid film-repelling regions, such that the boundaries of liquid film-adsorbing surface regions are defined by the liquid film-repelling surface regions. This is illustrated in
A suspension 109 of elongated nanoparticles 111 in a second liquid that is immiscible with the liquid of liquid film 108, referred to herein as a “nanoparticle ink.” is then contacted with liquid film 108. A suspension can be contacted with a liquid film by flowing the suspension over the liquid film, forming a layer of the suspension over the liquid film, spraying the suspension onto the liquid film, inkjet printing the suspension onto the liquid film, or by other means. When immiscible suspension 109 is in contact with liquid film 108 an interface is formed between the suspension and the liquid film and some elongated nanoparticles 111 from suspension 109 are transferred to liquid film 108 where they “float” on top of liquid film 108 (
The elongated nanoparticles may be deposited onto the substrate in the presence of the immiscible suspension (i.e., while the suspension in contact with the liquid film), in which case the dissipation of the liquid films that form on the liquid film-adsorbing surface regions may be due to the dissolution of liquid into the immiscible liquid of the nanoparticle suspension. In this scenario, the elongated nanoparticles are deposited along a contact line formed by the liquid film, the substrate, and the suspension. Alternatively, the elongated nanoparticles may be deposited onto the substrate after the nanoparticle suspension is no longer in contact with the liquid films that form on the liquid film-adsorbing surface regions (e.g., after a flow of the suspension over the liquid films has stopped or after the liquid films have been extracted from the suspension), in which case the dissipation of the liquid films is due to the evaporation of liquid from the films and the elongated nanoparticles are deposited along a contact line formed by the liquid film, the substrate, and air. In some embodiments of the methods, the deposition of the elongated nanoparticles may occur both in the presence of the nanoparticle suspension (
It should be noted that, the immiscible suspensions used in the methods need not be perfectly immiscible, as some degree of dissolution of the liquid of the liquid film in the suspension is permissible. Thus, an immiscible suspension need only be sufficiently immiscible to allow for nanoparticle film formation based on the principles described herein and on the timescale of the nanoparticle film deposition process.
The embodiment of the method shown in
In other embodiments of the methods, one or more films of aligned elongated nanoparticles are deposited from a two-dimensional (2D) liquid/suspension interface formed between an adsorbed liquid film and an overlying water-immiscible suspension in which the elongated nanoparticles are suspended. In these embodiments, as a substrate patterned with one or more liquid film-adsorbing surface regions is translated across the liquid/suspension interface, some of the liquid is retained on the liquid-film adsorbing surface regions as those regions pass from the layer of the liquid into the layer of the suspension and then into the air. This method is illustrated schematically in
As substrate 202 is extracted from liquid 205 and through suspension 209, a liquid film 208 is retained and immobilized on liquid film-adsorbing stripe 204 (
The elongated nanoparticles in may be deposited onto the substrate while the liquid film is still submerged in the suspension, in which case the dissipation of the liquid film may be due to the dissolution of liquid into the suspension, and the elongated nanoparticles are deposited along a contact line formed by the liquid, the substrate, and the suspension. Alternatively, as shown in
The elongated nanoparticles films can be formed from a variety of elongated nanoparticles, where elongated nanoparticles are characterized by high aspect ratios and nanoscale diameters. For the purposes of this disclosure, a high aspect ratio is defined as a length to diameter ratio of at least 5, including aspect ratios of at least 10, and a nanoscale diameter is defined as a diameter of no greater than 1000 nm, including diameters of no greater than 100 nm, no greater than 10 nm and no greater than 1 nm. Elongated nanoparticles include carbon nanoparticles, metal nanoparticles, and semiconducting nanoparticles, such as single-walled carbon nanotubes (SWCNTs), including both semiconducting SWCNTs (s-SWCNTs) and metallic SWCNTs (m-SWCNTs), multi-walled carbon nanotubes (MWCNTs), other types of single wall and/or multiwall nanotubes such as but not limited to boron nitride nanotubes, graphene nanoribbons, cellulose nanofibers (CNFs) and cellulose nanocrystals, metal nanowires (NWs), semiconducting nanowires such as Si or ZnO nanowires, ceramic nanowires, and nanowires made from organic materials.
The suspensions of the elongated nanoparticles are formed by dispersing the nanoparticles in a liquid. The liquid from which the suspension if formed will depend on the nanoparticles being used and may be an aqueous medium or an organic medium. A wide variety of liquids and mixtures of organic liquids may be used to form the liquid films and the suspension, provided that the liquid used to form the suspension is immiscible with the liquid used to form the liquid films. In some embodiments of the methods, the liquid film is an aqueous medium and the suspension is a non-aqueous organic liquid, or vice versa. Examples of suitable organic liquids include chloroform, toluene, 1,2-dichlorobenzene, carbon tetrachloride, cyclohexane, 1,2-dichloroethane, 1,1,2-trichloroethane, dichloromethane, chlorobenzene, 3-chlorotoluene, o-xylene, m-xylene, p-xylene, and mixtures of two or more thereof.
If the elongated nanoparticles are not sufficiently dispersible in a chosen aqueous medium or non-aqueous organic liquid, the nanoparticles may be surface-treated and/or coated with a material that renders the nanoparticles sufficiently dispersible. For example, s-SWCNTs may be coated in a coating that was used in a pre-sorting step to isolate s-SWCNTs from a mixture of s-SWCNTs and m-SWCNTs. These types of coatings are referred to herein as semiconductor-selective coatings. A number of semiconductor-selective coatings are known, including semiconductor-selective polymer coatings. Descriptions of such polymers can be found, for example, in Nish, A.: Hwang, J. Y.: Doig, J.: Nicholas, R. J. Highly selective dispersion of single-walled carbon nanotubes using aromatic polymers. Nat. Nanotechnol. 2007, 2, 640-6. The semiconductor-selective polymers are typically organic polymers with a high degree of x-conjugation and include polyfluorene derivatives, such as poly(9,9-dialkyl-fluorene) derivatives, and poly(phenyl vinylene) derivatives. Optionally, the coatings can be removed from the elongated nanoparticles after the nanoparticle films have been deposited.
The concentration elongated nanoparticles in the suspension will depend on the type of nanoparticles being used and the size of the surface area to be covered. Generally, at lower nanoparticle concentrations, less substrate surface area is covered. Substrate surface area coverage increases with increasing nanoparticle concentration in the suspension. At sufficiently high nanoparticle concentrations, the elongated nanoparticles at the liquid/suspension interface can interact and form two-dimensional liquid crystals floating at the surface of the liquid. These floating liquid crystals can then be deposited onto a liquid film-adsorbing surface region as the liquid film dissipates and the contact line recedes across the substrate. By way of illustration, the elongated nanoparticle concentration (e.g., the s-SWCNT concentration) in the suspensions can be in the range from 5 μg/mL to 200 μg/ml. However, lower or higher nanoparticle concentrations can be used.
The substrates upon which the elongated nanoparticle films are deposited may be hydrophilic or hydrophobic substrates. Examples of hydrophilic substrates include silicon and silicon oxide (e.g., SiO2). Other hydrophilic substrate materials that can be used include metal oxides (including, but not limited to, aluminum oxide, hafnium oxide, and lanthanum oxide), silicon nitride, hydrophilic semiconductor materials, such as germanium, and organic materials such as hydrophilic polymers or polymers with grafted or crosslinked materials with hydrophilic functional groups such as poly(ethylene glycol) or zwitterions, on their surfaces. Hydrophobic substrate materials include substrates coated or functionalized with hydrophobic chemical functionalities or coatings, such as HMDS and hydrophobic organic polymers, including those that can be used in flexible electronics applications. Such polymers include, but are not limited to, polydimethylsiloxane (PDMS), polyethersulfone, poly(ethylene terephthalate), and the like. These materials may compose the deposition substrate or may be applied as coatings over a bulk substrate base. The liquid film-adsorbing and/or liquid film-repelling regions can be patterned onto a substrate surface using known methods, such as surface treatment or the deposition of patterned coatings.
In embodiments of the methods that use a barrier wall to define a flow channel (e.g.,
The liquid film-adsorbing surface regions can be various shapes and sizes. Because the elongated nanoparticles are deposited along contact lines having contours defined by the borders of the liquid film-adsorbing surface regions, the deposited elongated nanoparticle films will reflect the shape of the liquid film-adsorbing regions. This enables the deposition of elongated nanoparticle films having complex shapes and varied dimensions. By way of illustration, the liquid film-adsorbing surface regions and the nanoparticle films deposited thereon may have a square or rectangular shape, may form a straight line, may have a curved edge (e.g., circular or ovoid), or may form a line with sharp bends and/or rounded corners.
The nanoparticle films can be deposited as highly uniform, monolayer films over large surface areas. The area over which the uniform nanoparticle films can be formed is not particularly limited and can be sufficiently large to cover the surface of an entire semiconductor wafer. By way of illustration, elongated nanoparticle films can be deposited over surface areas of at least 1 mm2, at least 10 mm2, at least 100 mm2, or at least 1 m2.
Conventional methods of patterning substrate surfaces with liquid film-adsorbing and liquid film-repelling surface regions, such as lithography, can be used. A novel method of forming a deposition substrate 302 having liquid film-adsorbing and liquid film-repelling surface regions starting with a silicon oxide (SiO2) substrate 302 is depicted schematically in
This Example demonstrates methods of using thin films of water on a chemically patterned surface to direct s-SWCNTs into high density films in which the s-SWCNTs are aligned along a solvent-water-substrate or air-water-substrate contact line. To demonstrate this process, narrow hydrophilic stripes on Si/SiO2 wafers were used as water film-adsorbing regions and narrow films of s-SWCNTs were deposited near the edges of patterned hydrophilic stripes (
Films Formed with Low s-SWCNT Concentration:
The sample in
The proposed mechanism by which these s-SWCNT films are formed is described above and shown schematically in
The first part of the investigation confirms the retention of water on the hydrophilic surface regions of the substrate. Surface water in the hydrophilic surface regions was observable by eye during the preparation of the substrate after both the dilute aqueous HCl etching and the DI water rinsing. To confirm that surface water was retained even during the deposition conditions with the chloroform-based s-SWCNT ink flowing across the substrate surface, videos of the surface through transparent quartz substrates were collected, and the retention of water was clearly visible. In the videos, some chloroform solvent was also present on the substrate, though it evaporated quickly, within 1-2 seconds, while the water was retained for longer, up to 3-4 seconds.
The second part of this investigation confirmed that s-SWCNTs were deposited at liquid-liquid-solid contact lines formed by the retained water films on the surface, surrounded by the chloroform ink.
The third part of this investigation confirmed that s-SWCNTs deposited on the retained surface water films are mobile and do not pass through the surface water film to deposit on the interior of the hydrophilic region while the water is present. It was inferred that this must be the case by looking at the solubility of the s-SWCNTs, the properties of the hydrophilic surface regions in the presence of water, and the morphology of the deposited films. To start, polymer-wrapped s-SWCNTs used are not soluble in water, so when they diffused from the chloroform ink to the chloroform-bulk water interface, they stayed at that interface. The retained surface water films on the substrate could be thin enough to behave differently than bulk water. To check this, a surface with water-adsorbing and water-repelling surface regions was prepared, sparse s-SWCNTs were deposited on them directly from chloroform using no water via spin coating, and then the samples were dipped in water. Before the water exposure, the s-SWCNTs were randomly distributed on the whole surface on both the water-adsorbing hydrophilic and water-repelling, less hydrophilic (relatively hydrophobic) regions. After the water exposure, all the s-SWCNTs on the water-adsorbing hydrophilic surface regions were gone, and all those on the relatively hydrophobic surfaces remained in place. The ability to remove s-SWCNTs with a simple water rinse demonstrates that any s-SWCNT that managed to penetrate the retained water film would not adhere to the hydrophilic surface and would diffuse back to the ink-water film interface or the ink-water film-substrate contact line. When the experiment was repeated with a high packing density of randomly oriented s-SWCNTs that formed an interconnected network, most of the s-SWCNTs on the hydrophilic surface regions were not removed by a simple water rinse, which implies that the adhesion forces between neighboring s-SWCNTs can stabilize the location of deposited s-SWCNTs.
Next, the morphology of the deposited films on the water-repelling region and on the water-adsorbing region was studied (
The final part in this investigation confirmed the self-alignment of the s-SWCNTs in the hydrophilic surface regions into s-SWCNT films that are more closely packed compared to the less closely packed s-SWCNT films deposited onto the relatively hydrophobic surface regions. In this Example, high-resolution AFM imaging with very sharp probes having a tip radius close to 1 nm was used to examine the s-SWCNT films. With the high-resolution probes and very careful imaging conditions, individual s-SWCNTs that have a pitch of 3.5 to 4 nm can be resolved. It was observed that the high packing density films near the borders of the hydrophilic surface regions (
The more fully packed morphology on the hydrophilic surface regions can be explained as follows. Because the hydrophilic surface regions retain water films, the contact line on the surface near the border of the hydrophilic and relatively hydrophobic regions is relatively stable, compared to the moving contact line on the uniform relatively hydrophobic regions, and the interaction time between the s-SWCNTs and the substrate is longer than on the relatively hydrophobic surface regions. The extra interaction time afforded by the relatively stable contact line can explain the very close packing of the films. This extra time would allow the s-SWCNTs to migrate right up to the previously deposited s-SWCNT, self-align to it, and form the closely packed films with very few gaps.
The high-resolution imaging allowed for the measurement of the degree of alignment as well as the packing density. Each s-SWCNT in a 325 nm square portion of the film was manually traced to determine the length and angle of each visible s-SWCNT segment. Histograms were then made and these lengths and angles, using a Gaussian fit and the standard deviation of the high packing density films on hydrophilic regions, were ˜3.5°. The packing density results are summarized in Table 1.
More concentrated inks (>20 μg mL−1) can deposit more s-SWCNTs at the ink-water interface where the s-SWCNTs can interact and form two-dimensional liquid crystals. These interfacial liquid crystals also can be aligned by the shear force of the flowing ink. When this interfacial liquid crystal is deposited on a hydrophilic surface region, it is transferred to the surface of the retained water film instead of being directly deposited on the substrate surface. When the hydrophilic surface region is patterned into stripes oriented to the liquid crystal's alignment direction, the retained water film-substrate contact lines constrain the liquid crystal and preserve the global alignment of the s-SWCNT film. The self-alignment process condenses the constrained liquid crystals into high packing density films. In narrow hydrophilic stripes (<=˜100 μm) the condensed s-SWCNT film typically deposits near the first edge of the hydrophilic surface region to pass through the ink-water interface and leaves a gap on the bottom edge of the stripe where retained surface water likely prevents the deposition of other s-SWCNTs. In wider stripes (200 and 300 μm) the s-SWCNT film is held in the center of the stripe by connections to the films fixed in place on the relatively hydrophobic surface regions on either end of the stripe.
Large area depositions are desirable for many applications, so films were deposited on a sample with a set of 100 μm by 2 mm stripes spanning a 25×25 mm region, and closely packed and well aligned films were deposited in almost every hydrophilic stripe.
To determine how the improved alignment and increased packing density of the s-SWCNT films affect their electronic performance, top gated FETs were fabricated using a 5 nm HfO2 atomic layer deposition insulator. The films were made using 100 μg mL−1 s-SWCNT ink from the high packing density films (
Hydrophilic Pattern Fabrication: The hydrophilic patterns were made using a lithographic process on thermal oxide coated Si wafers. The desired hydrophilic pattern was formed using an e-beam lithography system with a polymethyl methacry late (PMMA) resist. Very large regions were patterned using a foil shadow mask and UV/ozone chamber to dry etch the resist. After development, the patterned wafers were cleaned in a UV/ozone chamber and then 2 or 3 nm of yttrium metal was deposited in a thermal evaporator that was built into a nitrogen glovebox. After removal from the glovebox and completing a standard liftoff process using hot acetone, the substrates were again placed in a UV/ozone chamber at 90° C. for 8 minutes to clean and oxidize all the surfaces, as well as to oxidize the deposited yttrium metal. The samples were then treated with HMDS using either a vapor or liquid process. The exposure time and concentration were calibrated to partially functionalize the non-patterned regions to a water contact angle of 35-45 degrees. Alternatively, the samples were prepared, by overexposure to HMDS to produce WCA values greater than 45 degrees, then using short exposures in a UV/ozone chamber to lower the WCA to the target of 35-45 degrees. Just before the s-SWCNT deposition process, the oxidized yttrium thin films, now ˜3-5 nm thick due to oxidation, were removed using a 5 second dip in dilute aqueous hydrochloric acid (1:25 v/v, ˜0.1 molarity), followed by a quick dip/rinse in clean DI water then dried with compressed nitrogen. At this point, the prepared samples had a 35-45° WCA in the non-patterned regions and a very hydrophilic surface (<10° WCA) in the regions that previously had the oxidized yttrium thin film.
s-SWCNT Deposition Processes: s-SWCNT inks were made from poly [(9,9-dioctylfluorenyl-2.7-diyl)-alt-(6,6′-[2,2′-bipyridine])](PFO-BPy)-wrapped and sorted arc discharge s-SWCNTs dispersed in chloroform using a previously published procedure. (G. J. Brady, A. J. Way, N. S. Safron, H. T. Evensen, P. Gopalan, M. S. Arnold, Science Advances, 2 (9) (2016).) Two deposition processes were used to make s-SWCNT films. The films made with a TaFISA process were made using a previously published process. (K. R. Jinkins, Aligned 2D carbon nanotube liquid crystals for wafer-scale electronics, Science Advances, 2021, Vol. 7, Issue 37.) Two substrates were inserted in a water bath at 45° with their surfaces facing each other to make a 3 mm wide channel. A needle injected the chloroform-based s-SWCNT ink across the top of the water surface in the channel between the two substrates (4 mL min−1) while the substrates were lifted through the ink-water interface at 40) mm min−1. In this Example, one of the substrates had the hydrophilic chemical patterns and the other, sacrificial substrate was treated with HMDS to have a ˜40° WCA. The films made for calibrating the Raman based measurement of packing density were made using a shear alignment technique using a previously published process. (Jinkins. Katherine R . . . et al. “Substrate-Wide Confined Shear Alignment of Carbon Nanotubes for Thin Film Transistors.” Advanced Electronic Materials 5.2 (2019): 1800593.) The chloroform-based s-SWCNT ink is injected with a high volumetric flow rate into a sub-mm height and ˜1 cm wide and 2.5 cm long channel formed between the substrate and a highly polished metal surface. The high flow rate and confined channel dimensions generate a large amount of shear that aligns the s-SWCNTs in the direction of ink flow as they are deposited on the substrate.
Water Contact Angle Measurement: Static water contact angle measurements were performed using a Dataphysics OCA 15 optical contact angle system. Depending on the surface condition, DI water droplets of 2 to 5 μL were dispensed and the contact angle immediately measured.
Atomic Force Microscopy: AFM images were collected on a Bruker Icon instrument using a TESPA and/or ScanAsist-Air probes for low resolution imaging and SAA-HPI-SS probes for high resolution imaging. High resolution images were collected with square pixel sizes of approximately 0.3 nm.
Transmission Electron Microscopy: s-SWCNT films on Si/SiO2 substrates were sent to Eurofins EAG Laboratories (Santa Clara, CA, USA) for sample preparation and TEM imaging. The samples were prepared using the ex-situ focused ion beam (FIB) lift out technique on a FEI Dual Beam FIB/SEM and were capped with Carbon-based Epoxy/I-C prior to milling. TEM imaging was performed on a FEI Tecnai Osiri FEG/TEM operated at 200 kV in bright-field TEM mode and high-resolution TEM mode.
FET Device Fabrication: A 100 kV E-beam system (Elionix G-100) was used to lithographically pattern devices using a multi-step process. After deposition of alignment marks and s-SWCNT films, unwanted s-SWCNTs were removed using a radiofrequency (RF) oxygen plasma etch. Electrodes were deposited in two steps: first the high precision channel forming electrodes, and second the large contact pads. The conformal 5 nm HfO2 insulator for the top gate was deposited by atomic layer deposition (Veeco Fiji G2 PE-ALD). The top gate metal was deposited on top of that insulating layer.
s-SWCNT Sorting and Ink Fabrication: Single wall arc discharge carbon nanotubes were purchased (Aldrich, 698695) and sorted to isolate semiconducting s-SWCNTs by sonicating with toluene solutions of PFO-BPy (ADS Dyes, ADS153UV, Poly[(9,9-dioctylfluorenyl-2,7-diyl)-alt-co-(6,6-{2,2″-bipyridine})], MWN=˜40 to 100 kDa). A mixture of 120 mg of raw s-SWCNT powder and 60 mL of 2 mg mL−1 PFO-BPy in toluene were sonicated using a Fisher sonic dismembrator model 500 (400 W) for 30 minutes at 40% power in a 150 mL beaker surrounded by a static water bath for cooling. Next, the solution was divided into six tubes that were centrifuged with a swing bucket rotor (Sorval WX, TH-641, Thermo Scientific) at 41,000 rpm (300,000 g, max) for 10 minutes to remove soot and undispersed s-SWCNTs. The top ˜90% of the suspension was collected, and this process was repeated for a total of six batches. The combined suspensions from all six batches were condensed using a rotary evaporator down to ˜50 mL and then centrifuged again (same rotor and speed) for 1 hour to precipitate any poorly dispersed s-SWCNTs. The top suspension was collected, diluted to 60 mL with toluene, and centrifuged again (F65L-6×13 rotor at 25,500 rpm for 24 hours: in 10 mL PTFE tubes with ETFE caps, part 3114-0010, Thermo Scientific Nalgene) to form a pellet of all the suspended semiconducting s-SWCNTs. The pellet was separated from the PFO-BPy rich supernatant which was discarded. The semiconducting s-SWCNT pellets were redispersed in fresh toluene using stirring on a hot plate set at 80° C., and the pelleting process was repeated two or three more times to reduce the mass ratio of PFO-BPy: s-SWCNT to ˜0.5:1 as measured by UV/Vis/NIR absorption spectroscopy. The s-SWCNT ink was prepared by dispersing the final pellet into chloroform (Fisher Scientific, C606-4) and diluting to a concentration of 100 μg s-SWCNT mL−1.
Raman Spectroscopic Measurement of s-SWCNT Film Packing Density:
Raman spectroscopy was also used to characterize the packing density and alignment of s-SWCNT films, and those results corroborate the high packing density and excellent alignment results. Using Raman spectroscopy to measure the packing density of s-SWCNTs in a film, a calibration curve is necessary. The intensity of s-SWCNT peaks in a Raman spectrum are dependent on many spectrometer and sample prep factors, so care must be taken to ensure the calibration and unknown samples are measured under the same conditions. Two sets of calibration samples were fabricated using a shear force alignment method (ShASAM) described elsewhere. (Jinkins. Katherine R . . . et al. “Substrate-Wide Confined Shear Alignment of Carbon Nanotubes for Thin Film Transistors.” Advanced Electronic Materials 5.2 (2019): 1800593.) The s-SWCNT packing density at marked locations was measured using AFM since that imaging technique does not contaminate the sample surface. This deposition technique was used because it is expected to deposit individual s-SWCNTs in relatively well aligned films. The deposition of individualized s-SWCNTs is important since the calibration relies on manually counting the s-SWCNTs in particular locations to determine the density. If bundles of s-SWCNTs were deposited, then the number of s-SWCNTs present in any location would be undercounted since a feature that looks like a single s-SWCNTs might be 2 or more s-SWCNTs that were not resolved. Raman spectra were collected at those same locations, and the packing density and spectral information was used to generate the calibration curves shown in
Spectra of films high packing density were collected using the same conditions and were found to have a maximum G/Si ratio of ˜17, corresponding to a packing density of ˜210 s-SWCNT μm−1. This indicates a very high packing density that is outside the range of the calibration curve. Extrapolating this calibration curve outside of its measured range provides a rough estimate of the s-SWCNT packing density of greater than 200 s-SWCNTs μm−1.
Polarized Raman spectroscopy can be used to determine the degree of alignment of a s-SWCNT film because of a combination of the one-dimensional nature of s-SWCNTs and the enhancement of the Raman signal due to the s-SWCNT's electronic transition near the energy of the Raman laser. If the excitation laser energy is near to an optically active electronic transition, the intensity of a Raman signal is highly enhanced. In s-SWCNTs laying on a surface, the absorption cross-section of a photon approaching normal to the surface is proportional to the square of the cosine of the angle, a, between the polarization of the photon and the direction of the long axis of the s-SWCNT. When using a polarized filter before the detector, the Raman signal of that s-SWCNT is also proportional to the square of the cosine of the angle between the polarizer and the long axis of the s-SWCNT. Combining a polarized light source and polarized scattered photons, both at the same angle, the Raman signal intensity of a s-SWCNT oriented with angle α to those polarizers is proportional to cos4(α). By knowing the orientation of the polarizers, the Raman signal can be used to determine the orientation of the s-SWCNT by holding the polarizers in a fixed position and using a rotating stage (β) to rotate the s-SWCNT. If a film of s-SWCNTs is measured instead of a single s-SWCNT, then each s-SWCNT in the film will contribute signal intensity based on its angle relative to the polarizers' angles. If the film is perfectly aligned, the Raman signal will be proportional to cos4(β) as the sample is rotated.
Practical considerations of the spectroscopy equipment must be considered to ensure accurate measurements. The polarizers must be aligned at the same angle and that angle must be aligned to the beam splitter. If the polarizers are misaligned, then the signal will be artificially attenuated. If the laser polarization is not aligned to the beam splitter, then the light exiting the beam splitter will be elliptically polarized instead of linearly polarized. The equipment used in this measurement used a half-wave plate to adjust the polarization angle of the laser and an adjustable polarizer in the beam path just before the detector mirrors, gratings, and detector. Both of those optical elements were mounted in holders with stepper motors and were installed in the spectrometer only when needed. Matching the laser and detector polarization angle and matching that angle to the beam splitter requires careful measurements. This data was collected on a confocal Raman microscope using a 50x, long working distance objective, and the spot size was ˜1-3 μm depending on how tightly focused the beam was. This small spot size means the rotating sample stage must be very precisely centered (and flat) or different portions of the film will be measured as the sample is rotated.
After aligning the polarizers, one high packing density and highly aligned film was mounted on a rotating stage and measured at various angles. The intensity of the G peak vs sample rotation is plotted in
The intensity ratio of the s-SWCNT Raman spectra of the film aligned parallel to and perpendicular to the matched laser and detector polarizers has been used to determine the degree of alignment of s-SWCNT films. For the data measured, the ratio is (8465/18)=470 and this corresponds to a wrapped normal distribution with a standard deviation of approximately 9.4°. When comparing the measured ratio of 470 to the expected polarization ratio of the Raman laser (nominally 1000:1) and including all the potential sources of error described above, it becomes apparent that this measurement is likely near the performance limits of this equipment. To see how the measurement limit relates to the expected results for the closely packed films, the film alignment measurements provided by the AFM images were used to calculate the expected parallel to perpendicular ratio (Table 2).
Highly p-doped silicon wafer pieces with ˜15 nm of dry thermal oxide were cleaved to the desired size.
Alignment marks (global and local) were fabricated using e-beam lithography and 0.7/50 nm of Cr/Au using a ˜290 nm thick PMMA resist (475 PMMA A6, 4000 rpm for 60 seconds, softbake at 185° C. for 90 seconds; development using 2:1 (v/v) isopropanol/methyl-isobutyl ketone in an ice bath at 0° C. for 40 seconds; liftoff using hot acetone. These softbake, development, and liftoff processes were the same for all subsequent process steps.).
Yttrium metal stripes (3 nm thick) were fabricated using e-beam lithography with a ˜55 nm thick PMMA resist (950 PMMA C1, 2000 rpm for 60 seconds). UV/ozone cleaning was carried out at 90° C. for 8 minutes (and a 3-minute air purge, also at 90° C.). HMDS surface treatment (vapor or liquid) was used to achieve a water contact angle of ˜40° in regions without oxidized yttrium. Etching of the oxidized yttrium used a 5 second dip in dilute aqueous hydrochloric acid (1:25 v/v, ˜0.1 molar HCl concentration), followed by a dip rinse in DI water and blow dry with nitrogen. Deposition of s-SWCNT films using the TaFISA process was as described above.
The FET channel was defined using an RF oxygen plasma etch. The desired s-SWCNTs were protected with an e-beam patterned PMMA mask (˜100 nm, 950 PMMA C2, 4000 rpm, 60 seconds). Electrodes defining the FET channel were precisely written using e-beam lithography with ˜100 nm thick PMMA resist, (950 PMMA C2, 4000 rpm, 60 seconds) and thermal evaporation of 10/10 nm of Pd/Au. Thick contact pads were deposited on the channel electrodes to ensure low device electrode resistance and were fabricated using e-beam lithography and ˜290 nm thick PMMA resist (475 PMMA A6, 4000 rpm for 60 seconds) and 10/80 nm of Pd/Au.
Atomic Layer Deposition (Veeco Fiji G2 PE-ALD) was used to deposit 5 nm of HfO2 on all surfaces (150° C., 41 cycles alternating Hf-TDMAH [tetrakis(dimethylamido) hafnium (IV)] and water pulses).
Top gate metal was deposited using e-beam lithography with ˜100 nm thick PMMA resist, (950 PMMA C2, 4000 rpm, 60 seconds) and thermal evaporation of 0.7/20 nm of Cr/Au.
Contact to the source and drain electrode pads were made using gold coated tungsten needles in a custom-built probe station. The 10/80 nm Pd/Au on SiO2 electrode pads had an intermediate adhesion strength, and application of moderate pressure with the probe needles allowed scratching/delamination of the metal and penetration of the thin 5 nm HfO2 insulating layer.
Some devices were fabricated with intentional shorts across the channel to measure the electrode resistance which was 20-40 Ohms.
This example demonstrates elongated nanoparticle film deposition processes for delivering water and an s-SWCNTs-containing ink (i.e., an s-SWCNT suspension) to a substrate that differ from those described in Example 1. The principle of the film deposition mechanism is the same for this example as was described in Example 1 and so demonstrates that the methods are not restricted to any particular process for delivering liquids or s-SWCNT inks to the substrate.
The s-SWCNT films were deposited using a variant of the ShASAM process which is described in detail elsewhere. (Jinkins, Katherine R., et al. “Substrate-Wide Confined Shear Alignment of Carbon Nanotubes for Thin Film Transistors.” Advanced Electronic Materials 5.2 (2019): 1800593.) The flow channel apparatus used to in this example is shown in FIG. 1 of Jinkins et al. (2019). The substrate was placed between two metal plates with the target surface facing the smooth polished surface of one plate wall with a gasket and stainless-steel foil spacers between them. The foil spacers are slightly thinner than the gasket so when the screws holding the metal plates together are tightened the gasket is compressed to the thickness of the foil spacers. This forms an open flow channel between the substrate surface and the polished metal surface with a liquid tight seal and gap equal to the thickness of the foil spacers. The polished metal plate has a liquid inlet at one end of the substrate and the gasket is open at the other end, such that a liquid injected into the channel flows across the substrate in one direction from the inlet to the gasket opening. First water was injected into the channel and allowed to drain. Next a set of three liquids were staged in the inlet line: first 0.5 mL of chloroform saturated with water, then 0.25 mL of chloroform containing 125 μm/mL of s-SWCNTs, and last 3.25 mL of dry chloroform. This set of liquids was injected into the channel inlet at a volumetric flow rate of 100 mL/min and then followed with a manual injection of ˜10 mL 2-propanol. The high velocity injection of the s-SWCNT ink into the narrow channel results in a high shear rate of up to 10,000 s−1. On the patterned hydrophilic regions of this sample with retained surface water the s-SWCNTs are deposited on the chloroform-water interface and migrate to the chloroform-water-substrate contact line where they deposit aligned with the direction of the contact line which may or may not be aligned with the direction of liquid flow. On substrates with no patterning and no retained surface water the high shear liquid flow aligns the s-SWCNTs to the direction of the ink flow as they are deposited on the substrate directly from the chloroform ink.
Cross-Polarized Optical Microscopy: Optical images were collected on a Nikon Optiphot-2 microscope equipment with a fixed direction linear polarizing filter in the camera light path and a rotatable linear polarizing filter in the light source path.
The foregoing description of illustrative embodiments of the invention has been presented for purposes of illustration and of description. It is not intended to be exhaustive or to limit the invention to the precise form disclosed, and modifications and variations are possible considering the above teachings or may be acquired from practice of the invention. The embodiments were chosen and described in order to explain the principles of the invention and as practical applications of the invention to enable one skilled in the art to utilize the invention in various embodiments and with various modifications as suited to the particular use contemplated. It is intended that the scope of the invention be defined by the claims appended hereto and their equivalents.
This invention was made with government support under 1727523 awarded by the National Science Foundation. The government has certain rights in the invention.
