Carbon nanotube purification method

Abstract

[Problem] To provide a carbon nanotube purification method capable of suppressing shortening of carbon nanotubes. [Solution] A carbon nanotube purification method comprising a first step for obtaining a first dispersion liquid from a first mixture containing a CNT mixture I that contains two or more types of carbon nanotubes having different chiralities, an aqueous medium, and a first surfactant having a sterane backbone, and a second step for separating a CNT mixture II from the first dispersion liquid, wherein: the first step includes performing a high shear treatment on the first mixture; and the CNT mixture II contains, at a proportion higher than that in the CNT mixture I, a carbon nanotube having a chirality (n1,m1) of 1 (n1 represents an integer and m1 represents an integer of n1 or less).

Description

Method for purifying carbon nanotubes

[0001] This disclosure relates to a method for purifying carbon nanotubes.

[0002] Carbon nanotubes (hereinafter also referred to as "CNTs") are expected to be a key material for next-generation electronics due to their excellent optical properties, conductivity, and mechanical strength, and are being actively researched and developed. CNTs have a structure in which graphene sheets, in which carbon atoms are bonded in a hexagonal lattice, are wound into a tube. The electrical properties of CNTs differ depending on their winding direction (chiral angle) and tube thickness (diameter), and they can be metallic or semiconducting, and even among semiconducting CNTs, the band gap can differ.

[0003] Various methods are known for synthesizing carbon nanotubes (CNTs), including laser evaporation, arc discharge, and chemical vapor deposition (CVD). However, currently, it is not possible to synthesize CNTs with a single structure using any of these methods. Therefore, CNTs usually exist as a mixture containing various types of CNTs with different structures. For electronic applications of CNTs, homogeneous electrical properties are a crucial key, and methods for separating and purifying CNTs are strongly desired.

[0004] Various methods have been reported for separating carbon nanotubes (CNTs), but the effectiveness and performance of these methods vary greatly depending on the solvent used and the dispersant used to disperse the CNTs in the solvent. In particular, methods using water as the solvent and surfactants as the dispersant, such as gel separation, density gradient ultracentrifugation, and liquid-liquid two-phase separation, are very important because they are more economical and can separate a wider variety of chiralities compared to methods using polymers or DNA as dispersants. Therefore, the development of technologies for separation methods using water and surfactants is a very important topic.

[0005] Patent Document 1 and Non-Patent Document 3 describe an aqueous solution for structural separation of carbon nanotubes containing at least one selected from the group consisting of solubilized lithocholic acid and solubilized lithocholic acid isomers, as one of the gel separation methods using water and a surfactant, and describe a separation and recovery method using the aqueous solution for structural separation.

[0006] Non-patent document 1 describes density gradient ultracentrifugation, and non-patent document 2 describes liquid-liquid two-phase separation.

[0007] International Publication No. 2020 / 022414

[0008] Nature Nanotechnology 1, (2006)60-65Nature Communications 7, (2016)12056ACS Applied Nano Materials 3, (2020)11289

[0009] CNT purification is a method of separating CNTs dispersed in a solvent, and it is necessary to disperse the CNTs before separation. However, ultrasonic dispersion, which has been used in most conventional methods, is known to cause CNT shortening. When CNTs are shortened, the number of contact points between CNTs increases, and resistance increases, especially when using thin films of CNTs that are optimal for the application. This is undesirable because it can lead to increased energy loss when applied to electronic materials. A purification method is needed that combines dispersion that can suppress CNT shortening with separation that can maintain the length of the CNTs.

[0010] This disclosure is made in view of the above circumstances and aims to provide a carbon nanotube purification method that can suppress the shortening of carbon nanotubes.

[0011] This disclosure provides the following aspects: [1] A first step of obtaining a first dispersion from a first mixture comprising a CNT mixture I containing two or more carbon nanotubes with different chiralities, an aqueous medium, and a first surfactant having a sterane skeleton; and a second step of separating a CNT mixture II from the first dispersion, wherein the first step includes subjecting the first mixture to high shear treatment, and the CNT mixture II has a chirality of 1 (n １ ,m １ ) (however, n １ m is an integer, １ is, n １the proportion of carbon nanotubes having (which is an integer below) is higher than that in CNT mixture I, a method for purifying carbon nanotubes. [2] The method for purification according to [1], wherein the surfactant having a sterane skeleton includes one or more selected from the group consisting of cholic acid, deoxycholic acid, glycocholic acid, glycodeoxycholic acid, taurocholic acid, taurodeoxycholic acid, hyodeoxycholic acid, chenodeoxycholic acid, and alkali metal salts of the foregoing. [3] The method for purification according to [1] or [2], wherein the high shear treatment is performed at a temperature not higher than the critical micelle temperature of the first surfactant. [4] The separation method according to any one of [1] to [3], wherein the first step comprises subjecting the first mixture to high shear treatment to obtain a second mixture, and further centrifuging the second mixture and recovering a supernatant to obtain the first dispersion. [5] The method for purification according to [4], wherein the CNT mixture I includes a precipitate recovered by centrifuging the second mixture. [6] The method for purification according to any one of [1] to [5], wherein the separation of the CNT mixture II is performed by one separation method selected from the group consisting of gel separation, density gradient ultracentrifugation, and liquid-liquid two-phase separation. [7] The method for purification according to any one of [1] to [6], wherein the separation of the CNT mixture II is performed by a method comprising: obtaining, from the first dispersion, a second dispersion having a different surfactant composition from that of the first dispersion; and separating the second dispersion by a gel separation method. [8] The method for purification according to any one of [1] to [7], wherein the second dispersion comprises a second surfactant, and the second surfactant includes one or more selected from the group consisting of dodecyl sulfuric acid, cholic acid, lithocholic acid, deoxycholic acid, and alkali metal salts of the foregoing. [9] The carbon nanotube having the chirality (n １ , m １ ) is a single-walled carbon nanotube, the method for purification according to any one of [1] to [8].

[10] The carbon nanotube having the chirality (n １ , m １ ) comprises semiconducting carbon nanotubes, the method for purification according to any one of [1] to [9].

[11] The chirality (n １,m １ The purification method according to

[10] , wherein (6,4), (6,5), (7,3), (7,5), (8,3), (8,4), (11,-5), (10,3), (9,2), (9,4), (9,1), (10,0), (7,5), (8,4), (11,0), (10,2), (11,1), (12,1), (11,3), (7,6), (10,5), (9,5), (8,6), (8,7) and (9,7).

[12] The chirality (n １ ,m １ A purification method according to any one of [1] to [9], wherein the carbon nanotube having ) is a metallic carbon nanotube.

[13] The chirality (n １ ,m １ The purification method according to

[12] , wherein (7,4), (6,6), (10,1), (9,3), (8,5), (7,7), (11,2), (10,4), (9,6), and (8,8) is one or more selected from (7,4), (6,6), and (8,8).

[14] The average length L of the carbon nanotubes contained in the CNT mixture II ２ The average length L of the carbon nanotubes contained in the CNT mixture I is... １ Ratio to (L) ２ / L １ The purification method according to any one of [1] to

[13] , wherein the chirality (n １ ,m １ ) (however, n １ m is an integer, １ is, n １ A carbon nanotube composition having a content of 50% or more of carbon nanotubes having the following integers:

[16] The chirality (n １ ,m １The carbon nanotube composition according to

[15] , wherein (6,4), (6,5), (7,3), (7,5), (8,3), (8,4), (11,-5), (9,1), (10,0), (9,2), (9,4), and (10,3).

[17] The carbon nanotube composition according to

[15] or

[16] , wherein the average length of the carbon nanotubes contained in the carbon nanotube composition is 0.5 μm or more. The purification method according to any one of [1] to [8], wherein the carbon nanotube is a single-walled carbon nanotube.

[0012] According to this disclosure, it is possible to provide a carbon nanotube purification method that can suppress the shortening of carbon nanotubes.

[0013] Figure 1 shows the optical absorption spectra of a 1% DOC CNT dispersion before and after ultracentrifugation when using CoMoCAT-CNT. Figure 2 is a graph showing the dependence of the yield of ultracentrifugation on the type and concentration of surfactant and time when using CoMoCAT-CNT. Figure 3 shows the optical absorption spectra of CNTs eluted stepwise at various LC concentrations when using CoMoCAT-CNT. Figure 4 shows the optical absorption spectra of separated (6,5)CNTs and (11,-5)CNTs when using CoMoCAT-CNT. Figure 5 shows the circular dichroism spectra of separated (6,5)CNTs and (11,-5)CNTs when using CoMoCAT-CNT. Figure 6 shows the optical absorption spectra of various (n,m)CNTs when using CoMoCAT-CNT. Figure 7 shows, from left to right, the AFM images of the mixed CNT sample when using CoMoCAT-CNT, the AFM image of the (6,5) CNT sample separated after SFM dispersion, and the AFM image of the CNT sample dispersed by sonication when using CoMoCAT-CNT. Figure 8 is a histogram of the diameter calculated from the AFM images of the separated (6,5) CNTs when using CoMoCAT-CNT. Figure 9 is a graph showing the dependence of ultracentrifugation yield on the type of surfactant and time. Figure 10 is the optical absorption spectrum of CNTs eluted using 0.5% SC / 0.5% SDS when using HiPco-CNT. Figure 11 is a diagram showing an example of the experimental procedure. Figure 12 is the optical absorption spectrum of CNTs eluted using 0.5% SC / 0.5% SDS / 0.06% LC when using HiPco-CNT. Figure 13 is a diagram showing an example of the experimental procedure. Figure 14 shows the optical absorption spectrum of CNTs eluted using 0.4% SC / 0.6% SDS when HiPco-CNTs were used. Figure 15 shows the optical absorption spectra of CNTs eluted stepwise at various LC and DOC concentrations when HiPco-CNTs were used. Figure 16 shows the AFM image of the CNT sample when HiPco-CNTs were used.

[0014] The method for purifying carbon nanotubes of the present disclosure comprises: a first step of obtaining a first dispersion from a first mixture containing a CNT mixture I including two or more types of carbon nanotubes having different chiralities, an aqueous medium, and a first surfactant having a sterane skeleton; and a second step of separating a CNT mixture II from the first dispersion, wherein the first step includes subjecting the first mixture to a high-shear treatment, and the proportion of carbon nanotubes having one chirality (n １ , m １ ) (wherein n １ is an integer, m １ is an integer no greater than n １ ) in the CNT mixture II is higher than that in the CNT mixture I,

[0015] According to the purification method of the present disclosure, shortening of CNTs can be suppressed. Although the present disclosure should not be construed as being limited, the reason why the purification method of the present disclosure can achieve such an effect is considered as follows. That is, CNTs usually exist in a bundled state, where two or more CNTs aggregate to form a bundle. In order to separate and purify CNTs, it is considered necessary to separate CNTs from the bundle and disperse them. In the purification method of the present disclosure, when preparing a dispersion of CNTs, a predetermined surfactant is used and a high-shear treatment is performed in an aqueous medium. The high-shear treatment is a milder dispersion method than sonication, and is expected to be able to suppress shortening more than ultrasonic dispersion. When the separation method does not depend on the length of CNTs, it is expected that the length of CNTs after purification is also maintained.

[0016] In the present disclosure, a carbon nanotube having a chirality of (n, m) may be referred to as (n, m)CNT or the like in some cases.

[0017] (First Step) In the first step, a first dispersion is obtained from a first mixture containing a mixture I including two or more types of carbon nanotubes having different chiralities, an aqueous medium, and a first surfactant having a sterane skeleton. The first step includes subjecting the first mixture to a high-shear treatment.

[0018] The CNTs in Mixture I are not particularly limited as long as they contain two or more types having different chiralities (n, m). CNTs usually exist as bundles, and a bundle may contain tens to hundreds of CNTs. The CNTs may include metallic CNTs and semiconducting CNTs.

[0019] Examples of metallic CNTs include those in which, for the chirality (n, m), the difference between n and m (n-m) is a multiple of 3.

[0020] Examples of semiconducting CNTs include CNTs that do not fall under the category of metallic CNTs. Preferred examples of semiconducting CNT chiralities include (6,4), (6,5), (7,3), (7,5), (8,3), (8,4), (11,-5), (10,3), (9,2), (9,4), (9,1), (10,0), (7,5), (8,4), (11,0), (10,2), (11,1), (12,1), (11,3), (7,6), (10,5), (9,5), (8,6), (8,7) and (9,7), and more preferred examples include (6,4), (6,5), (7,3), (7,5), (8,3), (8,4) and (11,-5).

[0021] In one aspect, Mixture I preferably contains at least semiconducting CNTs. The content of semiconducting CNTs in CNT Mixture I is preferably 1 to 70% by mass, 2 to 60% by mass, or 3 to 50% by mass.

[0022] In this embodiment, mixture I preferably contains CNTs having chirality selected from (6,4), (6,5), (7,3), (7,5), (8,3), (8,4), (11,-5), (10,3), (9,2), (9,4), (9,1), (10,0), (7,5), (8,4), (11,0), (10,2), (11,1), (12,1), (11,3), (7,6), (10,5), (9,5), (8,6), (8,7) and (9,7), and more preferably contains CNTs having chirality selected from (6,4), (6,5), (7,3), (7,5), (8,3), (8,4), (11,-5), (10,3), (9,2) and (9,4). The content of these chiral CNTs in CNT mixture I is preferably 1 to 70% by mass, 2 to 60% by mass, or 3 to 50% by mass.

[0023] In another embodiment, mixture I may also preferably contain at least metallic CNTs. The content of metallic CNTs in CNT mixture I is preferably 1 to 70% by mass, 2 to 60% by mass, or 3 to 50% by mass.

[0024] In this embodiment, mixture I preferably contains, but is not limited to, CNTs having chiralities selected from (7,4), (6,6), (10,1), (9,3), (8,5), (7,7), (11,2), (10,4), (9,6), and (8,8). The content of these chiral CNTs in CNT mixture I is preferably 1 to 70% by mass, 2 to 60% by mass, or 3 to 50% by mass.

[0025] The average diameter of the CNTs in mixture I is preferably 0.5 to 5 nm, 0.6 to 3 nm, or 0.7 to 2.5 nm. The average diameter of the CNTs is determined based on the chirality (n, m) and the lattice length a (0.249 nm) of the hexagonal lattice of graphene, where a{(n ２ +m ２ +mn) １／２ You can calculate it as} / π.

[0026] The average length of the CNTs in mixture I is preferably 500 nm to 2 μm, 700 nm to 1.5 μm, or 800 nm to 1.3 μm.

[0027] The CNTs in mixture I preferably include single-walled CNTs. The proportion of single-walled CNTs in mixture I is preferably 80-100% by mass, 90-100% by mass, or 95-100% by mass.

[0028] The CNT content in CNT mixture I is preferably 90-100% by mass, 95-100% by mass, or 99-100% by mass. Mixture I may typically be a mixture of CNTs, but it may also contain impurities.

[0029] The content of CNT mixture I in the first mixture is preferably 0.001 to 3% by mass, 0.005 to 1% by mass, or 0.01 to 0.5% by mass.

[0030] The aqueous medium contains at least water and may optionally contain a hydrophilic solvent such as alcohol (e.g., methanol, ethanol) and a surfactant (e.g., a nonionic dispersant such as Triton X-100). The water content in the aqueous medium is preferably 90-100% by mass, 95-100% by mass, or 99-100% by mass.

[0031] The first surfactant has a sterane skeleton (cyclopentanoperhydrophenanthrene skeleton). Preferably, the first surfactant includes a compound in which a hydrophilic group such as a hydroxyl group, a carboxyl group, or a metal salt thereof is bonded to the sterane skeleton.

[0032] The molecular weight of the first surfactant is preferably 260 to 800, 300 to 700, or 350 to 650.

[0033] The solubility of the first surfactant in water is preferably 5 to 300 g / L, 10 to 250 g / L, or 20 to 200 g / L at 1,013 hPa and 25°C.

[0034] The critical micelle temperature of the first surfactant is preferably 0°C to 90°C, 5°C to 70°C, or 10°C to 60°C at 1,013 hPa and 25°C.

[0035] Examples of the first surfactant include any surfactant having a sterane skeleton, such as cholic acid, deoxycholic acid, glycocholic acid, glycodeoxycholic acid, taurocholic acid, taurodeoxycholic acid, hyodeoxycholic acid, chenodeoxycholic acid, and alkali metal salts thereof. In a preferred embodiment, examples of the first surfactant include cholic acid, deoxycholic acid, glycocholic acid, glycodeoxycholic acid, taurocholic acid, taurodeoxycholic acid, and alkali metal salts thereof.

[0036] The content of the first surfactant is preferably 0.1 to 5% by mass, 0.5 to 4% by mass, or 0.8 to 3% by mass, relative to the first mixture.

[0037] The total content of CNT mixture I, aqueous medium, and first surfactant in the first mixture is preferably 80 to 100% by mass, 90 to 100% by mass, or 95 to 100% by mass.

[0038] The first mixture may contain additives and the like in addition to CNT mixture I, an aqueous medium, and the first surfactant.

[0039] High shear treatment refers to a process that applies and disperses shear force. The rotational speed, which is a parameter related to shear force, is preferably 100 to 50,000 rpm, 300 to 30,000 rpm, or 500 to 20,000 rpm.

[0040] The high-shear treatment is preferably carried out at a temperature below the critical micelle temperature of the first surfactant, and more preferably at 5 to 40°C, 10 to 30°C, or 15 to 25°C. The temperature during the high-shear treatment may be, for example, the temperature of the first mixture.

[0041] The high-shear treatment may preferably be carried out for 1 to 150 hours, 3 to 120 hours, or 5 to 90 hours.

[0042] The second mixture obtained by subjecting the first mixture to high shear treatment may be used as the first dispersion as is, or the second mixture may be further centrifuged to collect the supernatant, which may be used as the first dispersion. By centrifuging, the bundles of CNTs and the dispersed CNTs can be separated more effectively.

[0043] The relative centrifugal force during centrifugation may be 10,000 to 500,000 × g, 50,000 to 400,000 × g, or 100,000 to 300,000 × g, and centrifugation may be performed for 10 to 120 minutes, 20 to 100 minutes, or 30 to 60 minutes.

[0044] When the second mixture is further centrifuged, the resulting precipitate may be used as CNT mixture I. In this case, only the precipitate may be used as CNT mixture I, or the precipitate and other CNTs may be used as CNT mixture I.

[0045] (Second step) In the second step, a chirality of 1 (n) is obtained from the first dispersion. １ ,m １ ) (however, n １ m is an integer, １ is, n １ Separate CNT mixture II, in which the proportion of carbon nanotubes having the following integers is higher than that of CNT mixture I.

[0046] As a method for separating CNT mixture II, one separation method selected from gel separation, density gradient ultracentrifugation, and liquid-liquid two-layer separation is preferred, with gel separation being more preferred.

[0047] The gel separation method utilizes the fact that the adsorption behavior between the gel and CNTs differs depending on the structure of the CNTs (diameter, chirality) to separate the CNT mixture II.

[0048] Examples of gels used in the gel separation method include carbohydrate-based gels such as dextran-based gels (Cefakryl: a homopolymer of allyldextran and N,N'-methylenebisacrylamide, GE Healthcare); agarose gels; starch gels; acrylamide gels; and mixtures thereof. Dextran-based gels are preferred, and allyldextran-based gels are more preferred.

[0049] The gel is typically particulate (bead-like), and the average particle size is preferably 25-75 μm, 35-65 μm, or 40-60 μm. The average particle size may be the cumulative 50% volume diameter measured by dynamic light scattering.

[0050] The gel concentration (volume ratio of gel beads to aqueous solution of the analyte, gel / aqueous solution) is preferably 0.01 to 25% as the final concentration.

[0051] The gel separation method may be carried out by a column method, in which the gel is packed into a column and a CNT dispersion (first or second dispersion) is passed through the column, or by a batch method, in which the CNT dispersion (first or second dispersion) is placed in a container packed with gel, and the CNTs adsorbed to the gel or not adsorbed to the gel are collected. In the column method, the solvent can be delivered to the column by gravity using an open column, or by pumping the solvent into a sealed column. With separation using a pump, it is possible to increase the flow rate and perform large-scale processing. Automated separation using a chromatography apparatus is also possible.

[0052] The CNT dispersion used for gel separation may be a first dispersion, or a second dispersion with a different surfactant composition from the first dispersion. Among the first surfactants, sodium deoxycholate has a high affinity between the hydrophobic portion and CNTs, and during gel separation, CNTs micelletized by the first surfactant may not be easily adsorbed onto the gel. By using a second or third dispersion with an adjusted surfactant composition, it is thought that CNTs having a predetermined structure (chirality, diameter) can be preferentially adsorbed onto the gel and separated more efficiently.

[0053] In one embodiment, the second dispersion can be produced by removing at least a portion of the first surfactant from the first dispersion and adding the second surfactant. In another embodiment, the second dispersion can be produced by removing at least a portion of the first surfactant from the first dispersion, adding a pre-surfactant to obtain a pre-dispersion, and then removing the pre-surfactant from the pre-dispersion and adding the second surfactant.

[0054] The removal of the first surfactant from the first dispersion and the removal of the preliminary surfactant from the preliminary dispersion can be carried out by ultrafiltration. That is, at least a portion of the aqueous medium and the first surfactant contained in the first dispersion is passed through an ultrafiltration filter, and the CNTs and the remainder of the aqueous medium and the first surfactant are recovered. Centrifugation may be performed simultaneously with ultrafiltration. The molecular weight cutoff of the ultrafiltration filter can be selected according to the type of first surfactant, and is preferably 1,000 Da to 500,000 Da, 30,000 Da to 200,000 Da, or 50,000 Da to 150,000 Da.

[0055] The second surfactant is C １０－１４ Examples include alkyl sulfates (preferably dodecyl sulfate), dodecanoyl sarcosine, dodecanoic acid, surfactants having a sterane skeleton (however, cholic acid is not included), n-dodecylphosphocholine, and alkali metal salts thereof. Among these, one or more selected from dodecyl sulfate, cholic acid, lithocholic acid, deoxycholic acid, and alkali metal salts thereof are preferred, one or more selected from sodium dodecyl sulfate (SDS), sodium cholate (SC), sodium lithocholic acid (LC), and sodium deoxycholic acid (DOC) are preferred, and one or more of sodium dodecyl sulfate (SDS), sodium cholate (SC), and sodium lithocholic acid (LC) are more preferred.

[0056] As the preliminary surfactant, any of the compounds exemplified as the secondary surfactant can be used, preferably one or more selected from dodecyl sulfate, cholic acid, lithocholic acid, and their alkali metal salts, preferably one or more selected from sodium dodecyl sulfate (SDS), sodium cholate (SC), sodium lithocholate (LC), and sodium deoxycholate (DOC), and more preferably one or more of sodium dodecyl sulfate (SDS), sodium cholate (SC), and sodium lithocholate (LC).

[0057] The second surfactant or a portion of the pre-surfactant may be the same as the first surfactant, and it is sufficient that the composition of the surfactant contained in the first dispersion is different from the composition of the surfactant contained in the second dispersion or the pre-surfactant.

[0058] The second surfactant or preliminary surfactant may be added as an aqueous solution. When the second surfactant or preliminary surfactant is added as an aqueous solution, its concentration is preferably 0.01 to 5% by mass, 0.1 to 3% by mass, or 0.5 to 2% by mass.

[0059] When obtaining a second dispersion directly from a first dispersion, the second dispersion can be obtained by performing the following steps once or more times: (i) obtaining a dispersion from the first dispersion (or a new dispersion to which the second surfactant has been added) by ultrafiltration to remove at least a portion of the aqueous medium and the first surfactant; and (ii) obtaining a new dispersion by adding an aqueous solution of the second surfactant of a desired composition to the dispersion obtained in step (i). Steps (i) and (ii) may be performed once, or repeated two to twenty times, or five to fifteen times.

[0060] When obtaining a second dispersion from a first dispersion via a preliminary dispersion, a preliminary dispersion can be obtained by performing step (iii): a step of obtaining a dispersion from the first dispersion (or a new dispersion to which a surfactant has been added) by ultrafiltration to remove at least a portion of the aqueous medium and the first surfactant, and step (iv): a step of obtaining a new dispersion by adding an aqueous solution of a preliminary surfactant of a target composition to the dispersion obtained in step (iii), once or more times. The second dispersion can be obtained by performing step (v): a step of obtaining a dispersion from the preliminary dispersion (or a new dispersion to which a surfactant has been added) by ultrafiltration to remove at least a portion of the aqueous medium and the preliminary surfactant, and step (vi): a step of obtaining a new dispersion by adding an aqueous solution of a second surfactant of a target composition to the dispersion obtained in step (v), once or more times. Steps (iii) and (iv) may be performed once, or repeated two to twenty times, or five to fifteen times. Steps (v) and (vi) may be performed once, or repeated two to twenty times, or five to fifteen times.

[0061] The concentration of the surfactant (second surfactant) in the second dispersion is preferably 0.01 to 5% by mass, 0.1 to 3% by mass, or 0.5 to 2% by mass.

[0062] In one embodiment, the second dispersion preferably contains sodium dodecyl sulfate (SDS), sodium cholate (SC), and sodium lithocholate (LC). In this embodiment, the mass ratio (SDS:SC) of sodium dodecyl sulfate (SDS) to sodium cholate (SC) in the second dispersion is preferably 20:80 to 80:20, 40:60 to 60:40, or 45:55 to 55:45. In the same embodiment, the amount of sodium lithocholate (LC) in the second dispersion is preferably 0.01 to 10 parts by mass, 0.05 to 5 parts by mass, or 0.1 to 1 part by mass per 10 parts by mass of the total of sodium dodecyl sulfate (SDS) and sodium cholate (SC).

[0063] In another embodiment, the second dispersion preferably contains sodium dodecyl sulfate (SDS), sodium cholate (SC), and sodium deoxycholate (DOC). In this embodiment, the mass ratio (SDS:SC) of sodium dodecyl sulfate (SDS) to sodium cholate (SC) in the second dispersion is preferably 20:80 to 80:20, 40:60 to 60:40, or 45:55 to 55:45. In the same embodiment, the amount of sodium deoxycholate (DOC) in the second dispersion is preferably 0.01 to 10 parts by mass, 0.05 to 5 parts by mass, or 0.1 to 1 part by mass per 10 parts by mass of the total of sodium dodecyl sulfate (SDS) and sodium cholate (SC).

[0064] In density gradient ultracentrifugation, the way in which surfactants wrap around (adsorb) to CNTs differs depending on the structure of the CNTs (diameter, chirality), and as a result, the specific gravity of the micellized CNTs differs. This is utilized to separate CNTs (Non-Patent Literature 1). In other words, CNT mixture II can be separated by ultracentrifugation of the first or second dispersion in a density gradient medium. The preparation of the density gradient medium and the ultracentrifugation can be carried out by conventionally known methods.

[0065] In the liquid-liquid two-phase separation method, CNTs are separated by utilizing the phenomenon in which an aqueous solution containing two types of polymers separates into two phases while maintaining their aqueous solution state under certain conditions (Non-Patent Literature 2). That is, by using a polymer that preferentially adsorbs one type of CNT contained in the first or second dispersion, and another polymer that can separate from this polymer in a liquid-liquid two-phase separation, CNT mixture II can be separated.

[0066] The resulting CNT mixture II has a chirality of 1 (n １ ,m １ ) (however, n １ m is an integer, １ is, n １The proportion of carbon nanotubes having the following integers (hereinafter also referred to as "specific CNTs") is higher than that of CNT mixture I. In other words, by carrying out steps 1 and 2, a CNT mixture can be obtained in which the purity of CNTs having specific chirality is increased.

[0067] The above chirality (n １ ,m １ The chirality of the CNT is not particularly limited, but is preferably of the semiconductor type, more preferably one selected from (6,4), (6,5), (7,3), (7,5), (8,3), (8,4), (11,-5), (10,3), (9,2), (9,4), (9,1), (10,0), (7,5), (8,4), (11,0), (10,2), (11,1), (12,1), (11,3), (7,6), (10,5), (9,5), (8,6), (8,7), and (9,7), and even more preferably one selected from (6,5), (11,-5), (7,3), (6,4), (8,3), (7,5), and (8,4). The chirality of the CNT can be confirmed by ultraviolet-visible near-infrared spectroscopy.

[0068] In another embodiment, the chirality (n １ ,m １ The surfactant is not particularly limited, but is preferably of the metallic type, and more preferably one selected from (7,4), (6,6), (10,1), (9,3), (8,5), (7,7), (11,2), (10,4), (9,6), and (8,8). The chirality of the CNTs can be confirmed by ultraviolet-visible near-infrared spectroscopy. The purification method of this disclosure makes it possible to separate not only semiconductor-type CNTs but also metallic-type CNTs depending on the composition of the surfactant, and is expected to have various applications.

[0069] The content of specific CNTs (chirality purity) in CNT mixture II is preferably 50% or more, 70% or more, or 80% or more. The upper limit is 100% or less, and may be 99% or less or 98% or less. The content of specific CNTs (chirality purity) can be confirmed by ultraviolet-visible near-infrared spectroscopy.

[0070] The ratio of the content of specific CNTs in CNT mixture II to the content of specific CNTs in CNT mixture I (chirality purity in CNT mixture II / chirality purity in CNT mixture I) is greater than 1, may be 1.2 or higher, 1.5 or higher, or 1.8 or higher, and may be 50 or lower, 10 or lower, or 5 or lower.

[0071] Average length L of CNTs in CNT mixture II ２ Preferably, the particle size is 0.5 μm or more, 0.8 μm or more, or 0.9 μm or more, and the upper limit is not particularly limited, but may be 10 μm or less, 5 μm or less, or 3 μm or less.

[0072] The average length L of the CNTs contained in CNT mixture II ２ And the average length L of the CNTs contained in CNT mixture I. １ Ratio to (L) ２ / L １ The ratio (L) is preferably 0.5 or more, 0.8 or more, or 0.9 or more, and the upper limit is not particularly limited, but is 3 or less, 2 or less, or 1.5 or less. For reasons that are not clear, it is thought that long CNTs may be selectively included in CNT mixture II by steps 1 and 2, and the above ratio (L) ２ / L １ It is possible that this value may exceed 1.

[0073] The carbon nanotube purification method disclosed herein can suppress the shortening of carbon nanotubes and can be suitably applied to various applications in various fields, such as thin-film transistors, light-emitting devices, sensors, etc.

[0074] The present invention will be further described in detail by the following examples, but the present invention is not limited thereto.

[0075] [Preparation of CNT dispersion: CoMoCAT-CNT] 100 mg of CoMoCAT-CNT (SG65i, Aldrich, diameter 0.75 ± 0.15 nm) was mixed with 1 g / 100 mL or 2 g / 100 mL of SC (sodium cholate, TCI) or 1 g / 100 mL or 2 g / 100 mL of DOC (sodium deoxycholate, Wako) aqueous solution (100 mL). The solution was placed in a bottle and mixed for 12, 24, and 48 hours using a high-shear mixer (SFM, Silverson, L5M-A) while cooling in cold water. The dispersion obtained by SFM dispersion was subjected to ultracentrifugation (200,000 × g, 45 min), and 80% of the supernatant was recovered.

[0076] Furthermore, in the case where DOC was used as the surfactant, a recycling experiment was conducted to redisperse the CNTs precipitated by ultracentrifugation. 100 mL of a 1% or 2% DOC solution was mixed with the precipitate, the solution was placed in a bottle, and the same SFM dispersion and ultracentrifugation were performed, with 80% of the supernatant being recovered.

[0077] [Optical Absorption Spectrum Measurement and Yield Calculation by Ultracentrifugation] Figure 1 shows the optical absorption spectra of a CNT dispersion of 1 g / 100 mL DOC before and after ultracentrifugation when using CoMoCAT-CNT. CNTs have different peak wavelengths of optical absorption depending on their (n,m) diameter and chiral angle. The synthesized CNTs are a mixture of CNTs with various diameters and chiral angles, and their optical absorption spectrum is observed as a superposition of the peaks of this mixture. In the case of CoMoCAT SG65i, the main component among the synthesized (n,m) is (6,5), and the yield was calculated at the peak wavelength (approximately 984 nm) derived from (6,5). The yield after ultracentrifugation can be calculated by taking the ratio of the product of the absorbance (concentration) and volume of the CNTs (vertical axis in Figure 1) before and after ultracentrifugation. For example, in Figure 1, it changes from 86 to 28. When expressed as a percentage, this means that of the dispersed CNTs, approximately 33% (28 / 86) with good dispersibility remained in the supernatant, while the remaining 67% (58 / 86) precipitated. In this case, the ultracentrifugation yield can be calculated to be approximately 33% by mass. A higher ultracentrifugation yield is more economical because it increases the amount of CNTs that can be used for column separation, as described later.

[0078] Figure 2 shows the dependence of the yield of ultracentrifugation using CoMoCAT-CNTs on the type, concentration, and time of the surfactant. For all surfactants, the yield increased with longer time, indicating that dispersibility improved with time. On the other hand, the yield depended heavily on the type of surfactant, with highly hydrophobic DOC yielding a higher yield than less hydrophobic SC. This suggests that the hydrophobic DOC, with its high affinity for the hydrophobic surface of the CNTs, covers them through hydrophobic interactions, making them easier to disperse in the solvent. Higher surfactant concentrations resulted in a higher yield of CNTs obtained in the supernatant. This can also be attributed to the fact that the CNT surface is more covered, making it easier to disperse in the solvent. The maximum yield obtained was 34.9% when treated with 2 g / 100 mL DOC for 48 hours. In the recycling experiment, it was found that 20% of the CNTs that precipitated in the first treatment remained in the supernatant after redispersion. It is expected that the amount of CNTs obtained as supernatant by ultracentrifugation can be increased by repeating the recycling process.

[0079] [Pretreatment for Column Separation] When using CoMoCAT-CNTs, the CNT dispersion dispersed in SC solution using SFM was pretreated to adjust the surfactant composition by mixing in SDS (sodium dodecyl sulfate), a surfactant necessary for CNT column separation (J. Am. Chem. Soc., 2011, 133, 44, 17610-17613). Dilution was performed by adding SDS solution to adjust the concentration to 0.5 g / 100 mL SDS + 0.5 g / 100 mL SC solution.

[0080] In the case of CoMoCAT-CNTs, when the CNT dispersion was dispersed in DOC solution using SFM, a pretreatment was performed to reduce the concentration of DOC before the dilution operation by mixing with the same SDS as described above. Highly hydrophobic DOC coats the surface of the CNTs and weakens the interaction between the column and the CNTs, making it unsuitable for CNT column separation. Therefore, the surfactant was replaced from DOC to SC to reduce the concentration of DOC. The surfactant replacement was performed by ultrafiltration using an ultrafiltration filter unit (Millipore, Amicon 100kDa, 15 mL). 15 mL of the CNT dispersion was placed in the upper part of the filter and centrifuged at 4000 rpm. Due to centrifugation, the surfactant and water pass through the filter to the lower part, but the CNTs do not, so only the CNTs are concentrated in the upper part. Approximately half of the surfactant solution was transferred to the bottom, and approximately half of that amount of solution containing 1 g / 100 mL of SC was added to the top by pipetting. This process was repeated 10 times to gradually reduce the DOC concentration in the surfactant. Assuming that the exchange of half the solution was repeated 10 times, the estimated DOC concentration when replacing 2 g / 100 mL of DOC is approximately 0.007 g / 100 mL. However, this is an ideal estimate assuming that the half volume is accurate and the DOC concentration of the solution transferred to the bottom and the solution remaining at the top are equal. In reality, the volume is not exactly half, and DOC is less likely to pass through the filter than water and becomes concentrated, so the actual concentration will be higher than 0.007 g / 100 mL. Subsequently, as described above, pretreatment was performed to adjust the composition of the surfactant by mixing in SDS. Dilution was performed by adding SDS solution and water to adjust to a 0.5 g / 100 mL SDS + 0.5% mass SC solution (containing a small amount of DOC). In this process, to minimize the impact on the separation of DOC, the volume of the original solution was adjusted to approximately five times its original volume through dilution.

[0081] [Column Preparation and Separation] Separation was performed using a high-performance liquid chromatography (HPLC) system (Cytiva). Separation was carried out at 20°C. Gel beads (Cefacryl S-100, Cytiva) were used as the column support. For the columns used, either a small-scale column for exploring separation conditions (length 10 cm, inner diameter 1.0 cm, Tricorn, Cytiva) or a large-scale column for large-scale separation (length 20 cm, inner diameter 2.6 cm, Hiscale, Cytiva) was used.

[0082] First, experiments were conducted using a small-scale column to explore the optimal conditions. Gel beads were packed into the column, deionized water was passed through it, and then equilibrated with a 0.5 g / 100 mL SDS / 0.5 g / 100 mL SC aqueous solution of the same concentration as the dispersant. A CNT dispersion with a surfactant concentration of 0.5 g / 100 mL SDS / 0.5 g / 100 mL SC was then added. The CNT dispersion used was one in which SFM dispersion was performed with SC 1 g / 100 mL, DOC 1 g / 100 mL, or DOC 2 g / 100 mL. After that, a 0.5 g / 100 mL SDS / 0.5 g / 100 mL SC aqueous solution of the same concentration as the dispersant was added to recover the eluted metal CNTs, and the column was washed until the solution became colorless and transparent. The above procedure is shown in Figure 11.

[0083] Figure 10 shows the optical absorption spectrum of the CNTs eluted as described above. The presence of a peak in the range of 300-500 nm indicates that metallic CNTs were separated.

[0084] Next, an aqueous solution containing LC (sodium lithocholate) in the same concentration as the dispersant was added to the column to elute the semiconductor CNTs adsorbed on the column. The above procedure is shown in Figure 11. The type of semiconductor CNT eluted changes depending on how the LC concentration is varied. Semiconductor CNT elution was performed in the following two ways.

[0085] As the initial elution method, a 0.5 g / 100 mL SDS / 0.5 g / 100 mL SC / 0.006 g / 100 mL LC aqueous solution was added, and the CNTs eluted from the column were recovered.

[0086] Figure 12 shows the optical absorption spectra of separated CNTs when CoMoCAT-CNTs were used. The results are shown when a CNT dispersion dispersed by SFM in 1 g / 100 mL DOC solution was used. Peaks were observed in the ranges of 500-700 nm and 900-1,300 nm, indicating that semiconductor CNTs were separated. Various peaks were observed here, indicating that CNTs with peaks at different wavelengths were eluted. As mentioned above, in CNTs, the peak wavelength of optical absorption differs depending on the (n, m), diameter, and chiral angle of the CNT, so this indicates that CNTs of different (n, m) sizes were eluted together. This is thought to be because the LC concentration was relatively high from the beginning, rather than gradually increasing it step by step, causing semiconductor CNTs with various chiralities to be eluted simultaneously. Although the chiralities are mixed, they are semiconductor CNTs and can therefore be used for semiconductor applications.

[0087] As the next elution method, a 5 g / 100 mL SDS / 0.5 g / 100 mL SC / 0.0005 g / 100 mL LC aqueous solution was added, and the CNTs eluted from the column were recovered. Then, while maintaining the 0.5 g / 100 mL SDS / 0.5 g / 100 mL SC concentration, the LC concentration was gradually increased to 0.008 g / 100 mL, and the same procedure was performed.

[0088] Figure 3 shows the optical absorption spectra of separated CNTs when CoMoCAT-CNTs were used. The results shown are for a CNT dispersion dispersed by SFM in 1 g / 100 mL DOC solution. The LC concentration is indicated for each spectrum, showing the LC concentration at which the CNTs were eluted. It can be seen that CNTs with peaks at different wavelengths were eluted at different LC concentrations. As mentioned above, the peak wavelength of optical absorption in CNTs differs depending on the (n, m), diameter, and chiral angle of the CNT, so this indicates that different (n, m) CNTs were eluted at different LC concentrations.

[0089] [Measurement of optical absorption spectra of separated (6,5)CNTs and (11,-5)CNTs] Next, having determined the separation conditions using a small-scale column, we moved to a large-scale column to perform large-scale separation of (6,5) and (11,5), which are the main components of CoMoCAT SG65i and are enantiomers. Figure 4 shows the optical absorption spectra of the separated (6,5)CNTs and (11,-5)CNTs. (11,-5) was eluted at LC 0.00295% and (6,5) at LC 0.00335%, which are slightly different values ​​from the small-scale measurements. The peaks shown are those originating from the band structures of (6,5) and (11,-5) (E11, E22, E33, etc.) and those originating from the subbands of (6,5) and (11,-5), respectively. Other peaks originating from (n,m) were only faintly observed, such as E11 at (6,4) and E11 at (9,1). The chirality purity estimated from the peak area was 93.3% for (6,5) and 87.4% for (11,-5), indicating that the obtained CNTs were of high purity and single chirality.

[0090] [Measurement of circular dichroism spectra of separated (6,5)CNT and (11,-5)CNT] Circular dichroism (CD) depends on the difference in absorptivity between left-circularly polarized and right-circularly polarized light. It is 0 for non-enantiomers, but a finite value for enantiomers. Also, values ​​with opposite signs are obtained for right-handed and left-handed polarized light. Figure 5 shows the circular dichroism spectra of separated (6,5)CNT and (11,-5)CNT when using CoMoCAT-CNT (their optical absorption spectra are shown in Figure 5). Clear CD peaks were obtained at optical absorption peaks E22, E33, etc., for (6,5) and (11,-5). E11 is not shown because it is outside the measurement range. Also, opposite signs were obtained for (6,5) and (11,-5). This indicates that (6,5) and (11,-5) are enantiomers. Based on the relationship in a previous study (J. Am. Chem. Soc., 2017, 139, 45, 16068-16071), the enantiomer purity estimated from the peak intensity was 80.5% for (6,5) and 83.8% for (11,-5), indicating that the obtained CNTs are high-purity enantiomers.

[0091] [Measurement of optical absorption spectra of various separated (n,m)CNTs] Figure 6 summarizes the optical absorption spectra of various separated (n,m)CNTs when CoMoCAT-CNTs were used. In this method, different (n,m)CNTs were obtained at different LC concentrations: (7,3) at 0.0020%LC, (6,4) at 0.0022%LC, (11,-5) at 0.0028%LC, (6,5) at 0.0032%LC, (8,3) at 0.0038%LC, (7,5) at 0.0046%LC, and (8,4) at 0.0060%LC. Unlike polymer dispersion in previous studies, this method uses an aqueous surfactant system for separation, and is therefore applicable to various (n,m) CNTs.

[0092] [AFM Measurement and Length Distribution Calculation of Separated Samples] When using CoMoCAT-CNTs, to determine the length of the separated samples, the CNTs were transferred to a substrate and their lengths were evaluated using an atomic force microscope (AFM). The CNT dispersions used were a mixed CNT sample containing various (n,m) types dispersed by SFM in a 2g / 100mL DOC solution, and a (6,5) CNT sample separated after SFM dispersion in a 2g / 100mL DOC solution. For comparison, a CNT sample treated with 1g / 100mL SC sonication was also prepared. This sample was prepared by sonicating a 1g / 100mL SC aqueous solution of CoMoCAT (30mL) at 30% power for 3 hours using a tip-type ultrasonic shredder (Sonicfire 250, Branson, tip diameter: 0.5 inches) while cooling in cold water. Once the CNT dispersions were obtained, substrates with controlled CNT density were prepared for AFM measurement. To ensure homogeneity of the substrate surface on which the CNTs adsorbed, silicon substrates with an oxide film surface-treated with aminopropyltriethoxysilane (APTES) were used. If the CNT density is too low, it becomes difficult to locate the sample during AFM measurement, and if the density is too high, it becomes difficult to calculate the length of a single CNT due to overlapping. To optimize the density and facilitate the measurement of CNT length, parameters such as CNT concentration, surfactant concentration, and pH were adjusted to ensure reproducible CNT transfer. The CNT concentration was adjusted so that the absorbance in the UV region (approximately 280 nm) was 0.3. The surfactant was standardized to DOC, which is easily transferred to the substrate while maintaining dispersibility, and was replaced using limit filtration. The DOC concentration was adjusted to 0.1% or 0.05%. The pH was adjusted to approximately 11. By matching the experimental conditions in this way, we were able to reliably obtain substrates with an appropriate density that facilitates length measurement, and their lengths were calculated by AFM measurements.

[0093] Figure 7 shows AFM images of a mixed CNT sample dispersed by SFM using CoMoCAT-CNT, an AFM image of a (6,5) CNT sample separated after SFM dispersion, and an AFM image of a CNT sample dispersed by ultrasonic waves. It can be seen that with SFM dispersion, CNTs exceeding micrometers in length are obtained and their length is maintained even after separation, while with ultrasonic dispersion, they are shorter than micrometers. Here, when considering the length of the CNTs, it is important to consider whether it is a single CNT or a bundle of CNTs. When considering applications as CNTs, it is desirable for them to be long and single. Therefore, in order to distinguish between single CNTs and bundles, the diameter of the CNTs was calculated from the height information of the AFM.

[0094] Figure 8 shows a histogram of diameters calculated from AFM images of separated (6,5)CNTs using CoMoCAT-CNTs. Compared to the theoretical diameter of 0.75 nm for single chirality (6,5)CNTs, there are clearly larger samples with diameters ranging from 1.36 nm to 2.16 nm. The standard deviation for the same sample is ±0.10 nm, and even considering a measurement error of about the standard deviation, the presence of larger diameter samples cannot be explained. Furthermore, the separated (6,5)CNTs have a chirality purity of 93%, and contain 7% of different (n,m) types. However, the diameter dispersion of the raw material CoMoCAT SG5i is narrow, at about ±0.15 nm, and even considering the presence of different (n,m) impurities, the presence of larger diameter samples cannot be explained. From these results, it can be concluded that these larger diameter samples are bundles of CNTs. Therefore, in this study, CNTs outside the range of a standard deviation of ±0.10 nm in the same sample were assumed to be bundles of CNTs and were not included in the length statistics. Similarly, for mixed CNTs containing different (n,m) diameters, CNTs outside the range of the sum of the standard deviations of ±0.25 nm estimated from the expected diameter distribution and measurement error were assumed to be bundles and were not included in the length statistics. Table 1 shows the average length and standard deviation of the CNTs obtained after removing the influence of such bundles as much as possible. The average length and standard deviation were 1.23 ± 0.75 μm for SFM-dispersed mixed CNTs, 1.03 ± 0.50 μm for SFM-dispersed and separated single chirality (6,5) CNTs, and 0.42 ± 0.08 μm for ultrasonically dispersed CNTs. It was found that SFM-dispersed CNTs were longer than ultrasonically dispersed CNTs and were separated while maintaining their length. This study is the first to demonstrate the micrometer length of enantiomer-separated carbon nanotubes (CNTs).

[0095]

[0096] [Calculation of Recovery Rate] When using CoMoCAT-CNT, the final yield of (6,5) and (11,-5)CNT, the main components of SG65i, was calculated from the product of the ultracentrifugation yield and the column separation yield. To clarify the differences in surfactant type and concentration, calculations were performed for 1 g / 100 mL SC and 2 g / 100 mL DOC. As shown in Table 2, the ultracentrifugation yield of the SFM-dispersed CNT dispersion was 7.5% for 1 g / 100 mL SC and 34.9% for 2 g / 100 mL DOC. Next, the yields of (6,5) and (11,-5) CNTs were calculated by taking the product of absorbance (concentration) and volume of the column-separated samples and taking the ratio before and after column separation. The yields were 11.2% with 1 g / 100 mL SC and 20.8% (11.0% + 9.8%) with 2 g / 100 mL DOC. By calculating the product of the ultracentrifugation yield and the column separation yield, the final yields of (6,5) and (11,-5) CNTs were 0.8% with 1 g / 100 mL SC and 7.2% (3.8% + 3.4%) with 2 g / 100 mL DOC. This shows that high-concentration DOC yields approximately nine times higher than low-concentration SC, indicating that it is a more economical method.

[0097]

[0098] [Preparation of CNT dispersion: HiPco-CNT] Purification was performed using HiPco-CNTs (manufactured by Nano Integris, diameter 1.0 ± 0.4 nm). The CoMoCAT method is a method for producing CNTs using a cobalt-molybdenum catalyst and CO gas as the carbon source, allowing for selective growth and enabling the acquisition of specific chiralities (e.g., (6,5)) in high yield. In contrast, HiPco-CNTs obtained by the HiPco method are CNTs produced using iron nanoparticles as a catalyst and CO gas (high pressure) as the carbon source. The HiPco method is suitable for mass production, but tends to have a broad chirality distribution. The reason for conducting experiments using HiPco-CNTs is to demonstrate that the purification method described herein is versatile and can be used not only for CoMoCAT-CNTs but also for CNTs of various diameters. CoMoCAT-CNTs are expected to be applied to optical devices and high-performance transistors where high performance is required. Now that it has been confirmed that separation is possible not only with CoMoCAT-CNTs but also with HiPco-CNTs, their application to conductive films and composite materials, where there is a high demand for mass production and cost reduction, is expected.

[0099] 100 mg of HiPco-CNT (Nano Integris, diameter 1.0 ± 0.4 m) was mixed with 100 mL of either 2 g / 100 mL SC (sodium cholate, TCI) or 2 g / 100 mL DOC (sodium deoxycholate, Wako) aqueous solution. The solution was placed in a bottle and mixed for 12, 24, and 48 hours using a high-shear mixer (SFM, Silverson, L5M-A) while cooling in cold water. The dispersion obtained by SFM dispersion was subjected to ultracentrifugation (200,000 × g, 45 min), and 80% of the supernatant was recovered.

[0100] Figure 9 shows the dependence of the yield of ultracentrifugation using HiPco-CNTs on the type of surfactant and time. The yield was found to be nearly constant for all surfactants. However, the yield did depend on the type of surfactant; with highly hydrophobic DOC, the yield was higher than with less hydrophobic SC. This suggests that the highly hydrophobic DOC, with its high affinity for the hydrophobic surface of the CNTs, coats the CNTs through hydrophobic interactions, making them easier to disperse in the solvent.

[0101] [Pretreatment for Column Separation] When using HiPco-CNTs, a CNT dispersion dispersed in DOC solution using SFM was pretreated by mixing in SDS (sodium dodecyl sulfate), a surfactant necessary for CNT column separation (J. Am. Chem. Soc., 2011, 133, 44, 17610-17613), to adjust the surfactant composition. Dilution was performed by adding SDS solution to prepare a CNT dispersion of 0.4 g / 100 mL SC + 0.6 g / 100 mL SDS.

[0102] In the case of CNT dispersions dispersed by SFM in a DOC solution using HiPco-CNTs, a pretreatment was performed to reduce the concentration of DOC before the dilution operation by mixing with the same SDS as described above. Highly hydrophobic DOC coats the surface of the CNTs and weakens the interaction between the column and the CNTs, making it unsuitable for CNT column separation. Therefore, the surfactant was replaced from DOC to SC to reduce the concentration of DOC. The surfactant replacement was performed by ultrafiltration using an ultrafiltration filter unit (Millipore, Amicon 100kDa, 15 mL). 15 mL of the CNT dispersion was placed in the upper part of the filter and centrifuged at 4000 rpm. Due to centrifugation, the surfactant and water pass through the filter to the bottom, but the CNTs do not, so only the CNTs are concentrated in the upper part. Approximately half of the surfactant solution was transferred to the bottom, and approximately half of that amount of solution containing 1 g / 100 mL of SC was added to the top by pipetting. This process was repeated 10 times to gradually reduce the DOC concentration in the surfactant. Assuming that the exchange of half the solution was repeated 10 times, the estimated DOC concentration when replacing 2 g / 100 mL of DOC is approximately 0.007 g / 100 mL. However, this is an ideal estimate assuming that the half volume is accurate and the DOC concentration of the solution transferred to the bottom and the solution remaining at the top are equal. In reality, the volume is not exactly half, and DOC is less likely to pass through the filter than water and becomes concentrated, so the actual concentration will be higher than 0.007 g / 100 mL. Subsequently, as described above, pretreatment was performed to adjust the surfactant composition by mixing in SDS. Dilution was performed by adding SDS solution and water to prepare a CNT dispersion of 0.4% mass SC solution + 0.6 g / 100 mL SDS (containing a small amount of DOC). In this process, to minimize the impact on the separation of DOC, the volume of the original solution was adjusted to approximately five times its original volume through dilution.

[0103] [Column Preparation and Separation] Similar to CoMoCAT, separation was performed using a high-performance liquid chromatography (HPLC) system (Cytiva). Separation was performed at 17°C. Gel beads (Cefacryl S-100, Cytiva) were used as the column support. For the columns, either a small-scale column for exploring separation conditions (length 10 cm, inner diameter 1.0 cm, Tricorn, Cytiva) or a large-scale column for large-scale separation (length 20 cm, inner diameter 2.6 cm, Hiscale, Cytiva) was used.

[0104] First, experiments were conducted using a small-scale column to explore the optimal conditions. Gel beads were packed into the column, deionized water was passed through it, and then equilibrated with a 0.6 g / 100 mL SDS / 0.4 g / 100 mL SC aqueous solution of the same concentration as the dispersant. A CNT dispersion with a surfactant concentration of 0.6 g / 100 mL SDS / 0.6 g / 100 mL SC was then added. The CNT dispersion used was one in which SFM dispersion was performed with 2 g / 100 mL of DOC. Subsequently, a 0.6 g / 100 mL SDS / 0.4 g / 100 mL SC aqueous solution of the same concentration as the dispersant was added, and after recovering the eluted metal CNTs, the column was washed until the solution became colorless and transparent. The above procedure is shown in Figure 13. Figure 14 shows the light absorption spectrum of the CNTs eluted above. A peak is observed in the range of 400 to 700 nm, indicating that metal CNTs were separated.

[0105] To the above column, a 0.6 g / 100 mL SDS / 0.4 g / 100 mL SC / 0.010 g / 100 mL LC aqueous solution, which contains LC (sodium lithocholate) in the same concentration as the dispersant, was added. After recovering the CNTs eluted from the column, the LC concentration was gradually increased up to 0.072 g / 100 mL while maintaining the 0.6 g / 100 mL SDS / 0.4 g / 100 mL SC concentration, and the CNTs eluted from the column were recovered in the same manner. Furthermore, for a CNT dispersion of 0.6 g / 100 mL SDS / 0.4 g / 100 mL SC, an aqueous solution of 0.6 g / 100 mL SDS / 0.4 g / 100 mL SC / 0.010 g / 100 mL DOC was added, which consisted of a surfactant of the same concentration as the dispersant plus sodium deoxycholate (DOC). After recovering the CNTs eluted from the column, the same procedure was performed by gradually increasing the DOC concentration up to 0.092 g / 100 mL while maintaining the 0.6 g / 100 mL SDS / 0.4 g / 100 mL SC.

[0106] Figure 15 shows the optical absorption spectra of the separated CNTs. Peaks were observed in the ranges of 500–700 nm and 900–1,300 nm, indicating that semiconductor CNTs were separated. In particular, this method allows for the acquisition of different (n,m) CNTs at different LC and / or DOC concentrations by gradually changing the LC and / or DOC concentrations. Specifically, (9,1) was obtained at 0.6% SDS / 0.4% SC / 0.034% LC, (10,0) at 0.6% SDS / 0.4% SC / 0.0042% LC, and (0.6% SDS / 0. (8,3) was obtained with 4% SC / 0.0057% LC, (9,2) with 0.6% SDS / 0.4% SC / 0.0065% LC, (7,5) with 0.6% SDS / 0.4% SC / 0.0072% LC, (9,4) with 0.6% SDS / 0.4% SC / 0.076% LC, and (10,3) with 0.6% SDS / 0.4% SC / 0.088% DOC. From this point, it can be seen that, unlike polymer dispersion in previous studies, this method is a separation using an aqueous surfactant system, and therefore is applicable to various (n,m) values.

[0107] [AFM Measurement and Length Distribution Calculation of Separated Samples] When using HiPco-CNTs, the length of the separated sample was determined by transferring the CNTs to a substrate and evaluating the CNT length using an atomic force microscope (AFM). The CNT dispersion used was a (6,5) CNT sample separated after SFM dispersion in a DOC 2g / 100mL solution. A substrate with controlled CNT density was prepared for AFM measurement. To homogenize the substrate surface on which the CNTs adsorb, an oxide film-coated silicon substrate surface-treated with aminopropyltriethoxysilane (APTES) was used. To optimize the density and facilitate measurement of CNT length, parameters such as CNT concentration, surfactant concentration, and pH were adjusted to ensure reproducible CNT transfer. The CNT concentration was adjusted so that the absorbance in the UV region (approximately 280 nm) was 0.3. The surfactant was standardized to DOC, which is easily transferred to the substrate while maintaining dispersibility, and was replaced using limit filtration. The concentration of DOC was adjusted to 0.1% or 0.05%. The pH was adjusted to approximately 11. By matching the experimental conditions in this way, substrates with an appropriate density that facilitates length measurement could be reproducibly obtained, and their lengths were calculated by AFM measurement.

[0108] Figure 16 shows the AFM image after SFM dispersion when using HiPco-CNTs. After SFM dispersion, CNTs exceeding micrometers were obtained, similar to the case when using CoMoCAT-CNTs. When using CoMoCAT-CNTs, as shown in Figure 7, the length of the CNTs remained almost unchanged before and after separation; therefore, with HiPco-CNTs, the length before separation (after SFM dispersion) was observed.

[0109] The average length and standard deviation were 1.29 ± 0.82 μm, and the diameter was 0.84 ± 0.11 nm.

[0110] The carbon nanotube purification method disclosed herein can suppress the shortening of carbon nanotubes and can be suitably applied to various applications in various fields, such as thin-film transistors, light-emitting devices, sensors, etc.

Claims

1. The process includes a first step of obtaining a first dispersion from a first mixture comprising a CNT mixture I containing two or more carbon nanotubes with different chiralities, an aqueous medium, and a first surfactant having a sterane skeleton; and a second step of separating a CNT mixture II from the first dispersion, wherein the first step includes subjecting the first mixture to high shear treatment, and the CNT mixture II has a chirality of 1 (n １ ,m １ ) (however, n １ m is an integer, １ is, n １ A method for purifying carbon nanotubes, wherein the proportion of carbon nanotubes having the following integers is higher than that of CNT mixture I.

2. The purification method according to claim 1, wherein the surfactant having a sterane skeleton comprises one or more selected from cholic acid, deoxycholic acid, glycocholic acid, glycodeoxycholic acid, taurocholic acid, taurodeoxycholic acid, hyodeoxycholic acid, chenodeoxycholic acid, and alkali metal salts thereof.

3. The purification method according to claim 1, wherein the high shear treatment is carried out at a temperature below the critical micelle temperature of the first surfactant.

4. The separation method according to claim 1, wherein the first step includes subjecting the first mixture to high shearing to obtain a second mixture, and further centrifuging the second mixture to collect the supernatant liquid to obtain a first dispersion.

5. The purification method according to claim 4, wherein the CNT mixture I includes a precipitate recovered by centrifuging the second mixture.

6. The purification method according to claim 1, wherein the separation of the CNT mixture II is carried out by one separation method selected from gel separation, density gradient ultracentrifugation, and liquid-liquid two-phase separation.

7. The purification method according to claim 1, wherein the separation of the CNT mixture II is carried out by a method comprising the steps of obtaining a second dispersion having a different surfactant composition from the first dispersion from the first dispersion, and separating the second dispersion by a gel separation method.

8. The purification method according to claim 1, wherein the second dispersion contains a second surfactant, and the second surfactant contains one or more selected from dodecyl sulfate, cholic acid, lithocholic acid, deoxycholic acid, and alkali metal salts thereof.

9. The chirality (n １ ,m １ The purification method according to claim 1, wherein the carbon nanotube having ) is a single-walled carbon nanotube.

10. The said chirality (n １ , m １ ) carbon nanotubes having the same include semiconducting carbon nanotubes. The purification method according to claim 1.

11. Chirality (n) of the semiconductor carbon nanotube １ ,m １ The purification method according to claim 10, wherein (6,4), (6,5), (7,3), (7,5), (8,3), (8,4), (11,-5), (10,3), (9,2), (9,4), (9,1), (10,0), (7,5), (8,4), (11,0), (10,2), (11,1), (12,1), (11,3), (7,6), (10,5), (9,5), (8,6), (8,7), and (9,7).

12. The chirality (n １ ,m １ The purification method according to claim 1, wherein the carbon nanotubes having ) include metallic carbon nanotubes.

13. Chirality (n) of the metallic carbon nanotube １ ,m １ The purification method according to claim 12, wherein (7,4), (6,6), (10,1), (9,3), (8,5), (7,7), (11,2), (10,4), (9,6), and (8,8) is one or more selected from (7,4), (6,6), and (8,8).

14. Average length L of carbon nanotubes contained in the CNT mixture II. ２ And the average length L of the carbon nanotubes contained in the CNT mixture I. １ Ratio to (L) ２ / L １ The purification method according to claim 1, wherein the value of is 0.5 or more.

15. Chirality of 1 (n １ ,m １ ) (however, n １ m is an integer, １ is, n １ A carbon nanotube composition having a carbon nanotube content of 50% or more that has the following integers.

16. The chirality (n １ ,m １ The carbon nanotube composition according to claim 15, wherein (6,4), (6,5), (7,3), (7,5), (8,3), (8,4), (11,-5), (9,1), (10,0), (9,2), (9,4), and (10,3).

17. The carbon nanotube composition according to claim 15 or 16, wherein the average length of the carbon nanotubes contained in the carbon nanotube composition is 0.5 μm or more.

Images