Article Industrial & Engineering Highly Concentrated Chemistry Research is published by the Aqueous Dispersions American Chemical
Subscriber access provided by Stockholm Society. 1155 Sixteenth Street N.W., University Library Washington, DC 20036 Published by American Chemical Society. Copyright © American
of Carbon Nanotubes for Flexible Industrialand & Engineering Chemistry Research is published by the Conductive Fibers American Chemical
Subscriber access provided by Stockholm Society. 1155 Sixteenth Street N.W., University Library Washington, DC 20036 Published by American Chemical Society. Copyright © American
Laurent Maillaud, Robert James Headrick, Vida Jamali, Industrial & Engineering Julien Maillaud, Dmitri Chemistry Research published by the Tsentalovich,is Wilfrid Neri, American Chemical E.Amram Bengio, Francesca Subscriber access provided by Stockholm Society. 1155 Sixteenth Street N.W., University Library Washington, DC 20036 Published by American Chemical Society. Copyright © American
Mirri, Olga Kleinerman, Yeshayahu Talmon, Philippe Industrial & Engineering Poulin, and Matteo Pasquali Chemistry Research is published by the
Ind. Eng. Chem. Res., Just American Chemical Accepted Manuscript DOI: Subscriber access provided by• Stockholm Society. 1155 Sixteenth Street N.W., University Library Washington, DC 20036 Published by American Chemical Society. Copyright © American
10.1021/acs.iecr.7b03973 • Publication Date (Web): 22 Feb 2018 & Engineering DownloadedIndustrial from http:// Chemistry Research pubs.acs.org on February 22,the2018 is published by American Chemical
Subscriber access provided by Stockholm Society. 1155 Sixteenth Street N.W., University Library Washington, DC 20036 Published by American Chemical Society. Copyright © American
Just Accepted
& Engineering “Just Accepted” Industrial manuscripts have been p Chemistry Research online prior to technical editing, formatting is published by the Chemical Society providesAmerican “Just Accepted” as a serv
Subscriber access provided by Stockholm Society. 1155 Sixteenth Street N.W., University Library Washington, DC 20036 Published by American Chemical Society. Copyright © American
of scientific material as soon as possible full in PDF format accompanied by an H Industrial & Engineering peer reviewed, but should not be consid Chemistry Research Digital Object Identifier (DOI®). “Just Acc is published by the American Chemical the “Just Accepted” Web site may not in Subscriber access provided by Stockholm Society. 1155 Sixteenth Street N.W., University Library Washington, DC 20036 Published by American Chemical Society. Copyright © American
a manuscript is technically edited and fo site and published as an ASAP article. Industrial to the manuscript text& Engineering and/or graphics Chemistry Research ethical guidelines that apply is published by the to the journ American from Chemical consequences arising the use of in Subscriber access provided by Stockholm Society. 1155 Sixteenth Street N.W., University Library Washington, DC 20036 Published by American Chemical Society. Copyright © American
Dialysis Wet spinning Page Industrial 1 of 21 & Engineering Chemistry Research
1 2 3 0.28 wt. % 1.8 wt. % CNT fiber 4 5 ACS Paragon Plus Environment 6 7 8 µm 200 µm 200 10 µm
Industrial & Engineering Chemistry Research 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Highly Concentrated Aqueous Dispersions of Carbon Nanotubes for Flexible and Conductive Fibers Laurent Maillaud1, Robert J. Headrick1, Vida Jamali1, Julien Maillaud1, Dmitri E. Tsentalovich1, Wilfrid Neri2, E. Amram Bengio1, Francesca Mirri1, Olga Kleinerman3, Yeshayahu Talmon3, Philippe Poulin2, Matteo Pasquali1* 1
Department of Chemical and Biomolecular Engineering, Department of Chemistry, Department
of Materials Science & NanoEngineering, The Smalley-Curl Institute, Rice University, Houston, Texas, 77005, United States. 2
Centre de Recherche Paul Pascal - CNRS, 115 Avenue du Dr. Schweitzer, 33600 Pessac,
France. 3
Department of Chemical Engineering and the Russell Berrie Nanotechnology Institute (RBNI),
Technion-Israel Institute of Technology, Haifa 3200003, Israel
Keywords: carbon nanotubes, dispersion, sonication, dialysis, fibers, wet spinning
Abstract Dispersing carbon nanotubes (CNTs) using surfactants into water requires ultrasonication that supplies mechanical energy to de-bundle and exfoliate CNTs. However, sonication is known to damage CNTs and to cut them into short fragments. Also, the CNT concentration in water dispersion is typically limited to up to 1.0 wt.%. Here, we show that by using a sulfuric acid pretreatment, we can enhance the de-bundling of CNTs and reduce subsequent sonication to achieve homogeneous dispersions without damaging CNTs. Additionally, using a progressive and 1 ACS Paragon Plus Environment
Page 2 of 21
Page 3 of 21 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Industrial & Engineering Chemistry Research
controlled dialysis, we are able to increase the CNT concentration up to 1.8 wt.%.
We
demonstrate that such highly concentrated dispersions can be used as spin dopes to fabricate continuous fibers. Our fibers have an electrical conductivity up to 580 kS/m, a tensile strength of ~1 GPa, and a Young’s modulus of 123 GPa, exceeding the mechanical properties of related fibers made from conventional surfactant-stabilized dispersions of sonicated CNTs.
Introduction Carbon nanotubes (CNTs) are of great interest because of their unique combination of high aspect-ratio, ballistic electron transport, high current-carrying capacity, and high mechanical strength and stiffness.1 Considering these characteristics, macroscopic CNT fibers are promising to replace metallic electrical wires2, and to yield the first truly synthetic electrical conductors that can be deployed on a large scale. However, the widespread use of CNTs is still challenging due to the processing difficulties, which have hindered the ability to take full advantage of their properties at the macroscopic scale. Wet spinning has been shown to be an effective way for processing meter-long fibers.3 This method consists of dispersing CNTs in solution, extruding the CNT dispersion out of a spinneret, and coagulating in a bath to form a solid fiber. This approach was the first method for producing CNT fibers,4 and is reminiscent of the scalable polymer fiber synthesis processes used in industry for decades. A large amount of research has been done to improve fiber properties and to scale-up the wet spinning process of CNT fibers.4-8 One of the most successful approaches relies on dissolving CNTs in superacids to obtain concentrated CNT liquid crystals for fiber spinning, which was under development by Teijin and is currently being commercialized by DexMat.3, 8 The high CNT mechanical and electrical fiber performance results from the fact that the CNTs are not damaged during dissolution and that the CNT concentration is high enough to produce 2 ACS Paragon Plus Environment
Industrial & Engineering Chemistry Research 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 4 of 21
dense and ordered fibers. Undamaged and long CNTs allow electrical conduction with fewer junctions and associated contact resistance, leading to more conductive fibers.9 In addition, longer CNTs also provide better mechanical stress transfer. As a result, the obtained fibers are expected to display improved mechanical properties.10,
11
However, processing in superacids
requires ventilated facilities as well as appropriate handling of used solvents. Moreover, superacids require adaptation of all the equipment in direct contact with the acid so as to prevent corrosion of components such as mixers, injection chamber and piston, and the spinneret, which is very important for controlling the fiber diameter and morphology. An efficient and easy way to disperse CNTs consists of mixing them with surfactants in water and supplying mechanical energy, generally via sonication. However, sonication induces the scission of CNTs and is not efficient for producing homogeneous dispersions at relatively high CNT concentration (> 1 wt.%) unless a long sonication time (≥ 1h) is used.12,
13
. For these
reasons, CNTs are shortened and both mechanical and electrical properties of the original CNT material are compromised.14-16 Therefore, obtaining a high CNT concentration in surfactant stabilized dispersions to produce high performance CNT fibers has until now remained a challenge.6, 15-18 Here, we report a new method to produce highly concentrated aqueous dispersions of CNTs stabilized by surfactant which allows the production of high performance CNT fibers by wet spinning. We show that by using a sulfuric acid pre-treatment, only a short sonication time at low power is required to achieve a homogenous CNT aqueous dispersion at ~ 0.3 wt.%. Moreover, using a dialysis process, we are able to progressively extract water from the CNT dispersion and achieve high concentrations of CNTs up to 1.8 wt.%. The highly concentrated dispersion shows a stable nematic liquid crystalline phase. Fibers spun under optimized conditions exhibited electrical conductivity up to 580 kS/m, modulus of 123 GPa, and mechanical tensile strength of ~ 3 ACS Paragon Plus Environment
Page 5 of 21 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Industrial & Engineering Chemistry Research
1 GPa. These performances largely exceed the mechanical properties of related fibers made from conventional surfactant stabilized dispersions of sonicated CNTs.16, 19
Experimental CNT dispersions preparation and setup Single-walled nanotubes (SWCNTs) obtained from SouthWest NanoTechnologies (batch CG301X, diameter, d ~ 0.8 nm) were used. To facilitate the individualization of CNTs by sonication, the SWCNTs were first stirred for 48 h in oleum (20 % free SO3, purchased from Sigma-Aldrich) (see Figure S1 for the Raman data) at a concentration of 10 mg/mL, neutralized with water, and filtered to form a wet cake. The neutral wet cake of SWCNTs was diluted with water containing sodium deoxycholate surfactant (SDOC, grade 97 %, purchased from SigmaAldrich) so that the final concentrations were 0.3 wt.% of SWCNTs and 1.0 wt.% SDOC in 10 mL of water. Multiple 10 mL dispersions were sonicated for 20 min at 15 W using a tip sonicator (Misonix S-4000). Then, the 10 mL mixtures were combined in the same vial and centrifuged for 30 min at 3000 g to remove the largest residual CNT aggregates. Finally, a dialysis tubing membrane (diameter = 16 mm, molecular weight cut-off (MWCO) = 10,000 Da, purchased from Thermo Scientific) was filled with the final centrifuged SWCNT dispersion. A 1.5 L aqueous dialysis solution was prepared containing 5-10 wt.% of dextran (average mol wt. = 150,000 g/mol, purchased from Sigma-Aldrich) and 1.0 wt.% SDOC. The dialysis tubing membrane (containing the sonicated and centrifuged SWCNTs in 1.0 wt.% SDOC) was placed into the dialysis reservoir (containing 5-10 wt.% dextran and 1.0 wt.% SDOC) and gently stirred for several days. This process gradually removes the water from the SWCNT dispersion (increasing the SWCNT concentration) while keeping the SDOC concentration constant. The SWCNT concentration change was monitored with absorption spectroscopy using Beer’s Law. When the 4 ACS Paragon Plus Environment
Industrial & Engineering Chemistry Research 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
desired SWCNT concentration was reached, the concentrated SWCNT dispersion was removed from the dialysis tubing membrane and placed into a glass vial. Using a wet spinning process, the concentrated SWCNT dispersion was extruded through a spinneret (130 µm diameter) into acetone (coagulant bath) to form the fiber. The fiber was collected onto a winding drum. Note that the linear velocity of the drum was equal or higher than the average extrusion speed at the spinneret in order to align CNTs by stretching the filament. Fibers were dried at room temperature for 24 hrs. Overnight water washes were further applied to remove residual surfactant. After each washing, fibers were dried in an oven at 80 °C for 12 hrs. In order to reach higher conductivity, washed 1.8 wt.% fibers were doped by sublimating iodine at 200 °C for 24 hours in a sealed oven. The fibers were then cooled down to room temperature and washed with ethanol to remove any excess iodine. Characterization CNT dispersion concentration was measured by absorption spectroscopy using Beer’s law, after having determined the extinction coefficient at 900 nm (Shimadzu UV-1800).20 Calibration curves were used with known CNT concentrations in the dilute regime (0.001 to 0.01 wt.% of CNTs) for this purpose. Concentrated dispersions after dialysis were diluted by a factor of 100, and then sonicated for 2 hours before characterization. Fiber tensile strength and elongation-atbreak were determined from tensile tests (at least 7 tests) performed at 10 mm/s on 20 mm long single filaments using a mechanical testing instrument (Zwick Z2.5/TN1S). Tensile stress was calculated by normalization of the tensile force with the fiber average cross-sectional area as measured by light microscopy on five different points along the fiber (see Figure S2 in Supporting Information) using a LEICA DM 2500P microscope and 10x/0.25 HI plan or 40X/0.65 HI plan objectives. Fiber resistance 𝑅 was measured by four-point probe measurements using a Keithley 2400 source meter. The four gold-coated probes were pressed onto the fibers to 5 ACS Paragon Plus Environment
Page 6 of 21
Page 7 of 21 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Industrial & Engineering Chemistry Research
ensure good contacts without damaging the fibers. The two central probes were spaced by 6.9 cm and external current delivering probes were 1 cm from each central probe. Then, the fiber resistivity ρ and conductivity σ were calculated with the following formula: σ = ρ-1 =
L RS
where 𝑆
and 𝐿 are the cross sectional area and the length of the fiber, respectively. Fiber morphology was observed using SEM (FEI Quanta 400 ESEM FEG) at an accelerating voltage between 1.5 and 5 kV. Bulk density of the fibers was calculated by measuring the linear density through weighing ten-meter-long filaments and using the cross-sectional area of fibers determined by light microscopy. The density was 1 g.cm-3 for as-spun fibers, 1.2 g.cm-3 for washed fibers and 1.3 g.cm-3 for iodine doped fibers.
Results and Discussion Highly concentrated aqueous dispersions of CNTs prepared by dialysis In order to study the effect of fuming sulfuric acid pre-treatment on CNT dispersions, we prepared two 10 mL samples of 0.3 wt.% SWCNTs (diameter d ~ 0.8 nm and aspect ratio L/d ~ 3000) and 1 wt.% SDOC as stabilizing surfactant in water. One sample was prepared from SWCNTs that were not treated and the other one was prepared from SWCNTs that were stirred with a magnetic bar in fuming sulfuric acid (oleum) for 48 hours, then filtered on membrane and washed with water. We found that at this concentration, the untreated SWCNTs were well dispersed (no aggregates over 1 µm were visible by light microscopy) after 1 hour of harsh tip sonication at 30 W. However, oleum treated SWCNTs were well dispersed after being sonicated for only 20 min at 15 W. This milder and shorter sonication was sufficient to obtain an aggregatefree and homogeneous dispersion. In fact, oleum protonates the SWCNTs surface and promotes the SWCNT de-bundling due to repulsive interactions. Based on earlier work on acid-spun 6 ACS Paragon Plus Environment
Industrial & Engineering Chemistry Research 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
SWCNT fibers coagulated in water,3 we believe that a small amount of acid remains intercalated within the SWCNTs and keeps the SWCNTs partially exfoliated. Therefore, as experimentally observed, much weaker/shorter sonication is necessary for further de-bundling. This weaker/shorter sonication prevents SWCNTs from excessive shortening. The average length of the SWCNTs that were pretreated with oleum was estimated to be 1020 ± 130 nm by statistical analysis of cryogenic transmission electron microscopy (cryo-TEM) images as described in previous work21 (cryo-TEM images in Figure S3 and bootstrap analysis results in Figure S4 in Supporting Information). Moreover, cryo-TEM images show straight SWCNTs which suggests that the pristine structure of the SWCNTs was not affected by sonication.22 Next, we developed a novel route to reach high SWCNT concentrations in surfactant-stabilized dispersions up to 1.8 wt.%. After sonication, ultracentrifugation and water evaporation can be used to concentrate SWCNT dispersions. However, ultracentrifugation leads to SWCNT aggregates that pack at the bottom of the vial. To remove these large aggregates, the dispersions generally need to be re-dispersed via sonication which further damages the CNTs. Alternatively, concentrating dispersions via water evaporation increases both SWCNT and SDOC concentrations, resulting in attractive depletion forces between the CNTs.23 These interactions can destabilize the dispersions at high CNT concentrations by forming large CNT aggregates. Lastly, a high surfactant concentration can compromise fiber properties and affect fiber formation during the coagulation step. Figure 1a shows the schematic of a membrane dialysis process known as the osmotic stress set up24, that we used to achieve high-concentration SWCNT dispersions. This method is based on the exchange of water molecules and solutes between the sample and a reservoir. The 0.28 wt.% SWCNT dispersion (about 10 wt.% of SWCNTs are lost after centrifugation of the initial dispersion) and 1 wt.% of SDOC, is placed in a dialysis bag (Figure 1b). The bag has pores with 7 ACS Paragon Plus Environment
Page 8 of 21
Page 9 of 21 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Industrial & Engineering Chemistry Research
a cut-off length that only allows the exchange of water molecules, ions, and surfactants until osmotic equilibrium is reached. The reservoir, of ~1 L of solution, contains the same concentration of SDOC as the original SWCNT dispersion. Thus, the SDOC concentration remains the same throughout the dialysis. To extract water from the dispersion, a high molecular weight stressing polymer in water is added to the reservoir. The migration of water molecules through the membrane is forced to equilibrate the water chemical potentials on either side of the membrane bag. Between 5 and 10 wt.% of dextran was used as stressing polymer. After a few days of dialysis (depending on the desired final SWCNT concentration), a significant decrease of the bag volume was observed. At the end of the process, a viscous SWCNT paste was removed from the bag (Figure 1c).
Figure 1. (a) Schematic of the dialysis process showing how the CNT dispersion is concentrated by progressively transferring water from the dispersion to the reservoir. (b) Photograph of the SWCNT dispersion in the dialysis bag at the beginning of the dialysis. (c) Photograph of a concentrated SWCNT dispersion (1.8 wt.%) after removal from the dialysis bag. The SWCNT concentration was measured by absorption spectroscopy throughout the dialysis process. As shown in Figure 2a, the concentration increases as the dialysis proceeds, from 0.28 8 ACS Paragon Plus Environment
Industrial & Engineering Chemistry Research 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 10 of 21
wt.% to 1.8 wt.%. To the best of our knowledge, this is the highest SWCNT (~ 1 µm long) concentration ever reported for homogeneous surfactant stabilized aqueous dispersions.14,
25, 26
Polarized optical micrographs of the 0.28 wt.% and 1.8 wt.% samples are presented in Figure 2c and 2b, respectively. At 0.28 wt.% of SWCNT, the dispersion is a biphasic system with an isotropic phase in equilibrium with a nematic liquid-crystalline phase. The formation of liquid crystalline phase shows that the SWCNTs are homogenously suspended in water.27 The liquid crystal phase emerges in the form of spindle-shaped liquid crystal droplets (tactoids) in coexistence with the isotropic phase. Previous work has shown the formation of tactoids in aqueous dispersion of bile salt stabilized CNTs.28 These tactoids possessed a homogenous optical texture indicating that CNTs are uniformly aligned inside the tactoids. Most recently, it was shown that tactoids formed in solutions of CNTs in superacid are characterized by two different type of optical texture depending on the size of the tactoids. Small tactoids showed a uniform change in brightness upon the rotation of the crossed polarizers of the microscope (homogenous tactoids), while larger tactoids showed a dark cross at the center when the tactoid is aligned with one of the polarizers (bipolar tactoids).29 The observation of both types of optical texture and a continuous transition from one to the other as a function of droplet size was attributed to the exceptional length of the CNTs that shifts the transition to a length scale observable in an optical microscope.29 Tactoids formed in our aqueous SWCNT dispersions are long (up to 80 µm) with both uniform and bipolar optical textures (Figure 2c-g). The immediate formation of tactoids with both uniform and bipolar configurations within this size range is consistent with what has been reported on other CNT systems,29 suggesting that the CNT length is on the order of one micrometer, in agreement with our cryo-TEM measurements (Figure S3). Moreover, the appearance of nematic tactoids coexisting with the isotropic phase at this low CNT concentration confirms that the SWCNTs are relatively long as the isotropic to nematic phase transition 9 ACS Paragon Plus Environment
Page 11 of 21 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Industrial & Engineering Chemistry Research
concentration in rod-like systems is inversely proportional to their aspect ratio.30,
31
After 120
hours of dialysis, the dispersion was concentrated to 1.8 wt.% SWCNTs. A fully liquid crystalline phase was observed (Figure 2b), and the dispersion was still homogeneous and free of SWCNT aggregates.
Figure 2. (a) Evolution of the absorbance of the SWCNT dispersion at different times of dialysis measured by absorption spectroscopy. The inset shows the SWCNT concentration starting at 0.28 wt.% before dialysis and increasing to 0.37, 0.48, 0.66, 0.99, and 1.8 wt.% after 24h, 48h, 72h, 96h and 120h of dialysis, respectively. (b) and (c) are cross-polarized light micrographs of the 1.8 and 0.28 wt.% SWCNT dispersions, respectively. Red crossed arrows show the directions of the crossed polarizers. (d-e) Polarized optical micrograph of a homogenous tactoid with a uniform
10 ACS Paragon Plus Environment
Industrial & Engineering Chemistry Research 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
optical texture. (f-g) Polarized optical micrograph of a bipolar tactoid with a dark cross at the center, when the tactoid is aligned with one of the polarizers. Scale bars in (d) to (g) are 20 µm.
Continuous SWCNT fibers by wet spinning To show the potential of high concentration CNT aqueous dispersions for material manufacturing, we produced fibers from dispersions containing 0.3, 1.0, and 1.8 wt.% SWCNT (called 0.3 wt.% fibers, 1.0 wt.% fibers, and 1.8 wt.% fibers, respectively) by a scalable wet spinning set up, previously described by Behabtu et al. 3. The liquid crystalline dispersions were first filtered to remove any solid particles, then loaded into a stainless steel chamber, and extruded by a piston under high pressure (~ 6 atm) through a spinneret (130 µm hole diameter) immersed in an acetone coagulation bath. The fiber filaments were kept in tension during coagulation by collecting them on a drum rotating at a speed equal or higher than the extrusion speed, which would further align the CNTs within the filaments resulting in densely packed fibers. Depending on the SWCNT concentration, continuous fibers could be spun (video in Figure S5). Finally, fibers were dried at room temperature and washed overnight in water to remove the surfactant. Note that various coagulants were tested, including organic solvents (ether, isopropyl alcohol, and ethanol), acids (nitric, sulfuric, and hydrochloric) and bases (sodium hydroxide and potassium hydroxide). However, these coagulants induced coagulation rates either too fast (brittle fibers and clogging of spinnerets) or too slow (deformation of the fibers upon contact with the drum), detrimental for obtaining well-ordered and continuous fibers. The best fiber morphology was found when using acetone as coagulant. Optimization of the spinning process was made by varying the draw ratio defined as the ratio between average translational velocity of the spinning dope at the spinneret and the linear speed of the drum
11 ACS Paragon Plus Environment
Page 12 of 21
Page 13 of 21 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Industrial & Engineering Chemistry Research
collecting the fiber. The draw ratio was considered optimal when the maximal linear speed of the drum was reached for a given extrusion velocity at the spinneret without fiber breakage. Spinning fibers from 0.3 wt.% SWCNT dispersions was possible. However due to the low SWCNT concentration, it was impossible to get long (> 1 meter) and homogeneous fibers (scanning electron microscopy (SEM) images in Figure S6), regardless of the draw ratio. Therefore, the obtained fragments of fibers were not characterized in details. Figure 3 shows the fiber morphology observed by SEM for 1.0 wt.% (top) and 1.8 wt.% (bottom) fibers, before and after washing. We observed that 1.0 wt.% fibers are thicker than 1.8 wt.% fibers. Coagulation of the dispersions concentrated at 1.0 wt.% SWCNT was more difficult than at 1.8 wt.% SWCNT, resulting in a lower maximum draw ratio at 1.0 wt.%. Thus, the ability to stretch the 1.0 wt.% fiber and induce CNT alignment and packing was lower. The maximum draw ratio was 1.1 for 1.0 wt.% and 1.65 for 1.8 wt.% SWCNT. The influence of the draw ratio on SWCNT alignment is apparent when comparing the images in Figure 3b and 3d. Washed fibers made at draw ratio of 1.1 exhibit SWCNT bundles perpendicular to the fiber axis (Figure 3b), whereas these defects do not appear in fibers made with a draw ratio of 1.65 (Figure 3d). These images also allow us to observe the effect of the water washing on the fiber morphology. SWCNT bundles are more clearly seen after surfactant removal. Moreover, the decrease of about 10 % of the fiber diameter after washing is due to the surfactant removal, which leads to fiber densification. The average fiber density is 1.0 g.cm-3 before washing and 1.2 g.cm-3 after washing.
The CNT fiber mechanical properties were tested via tensile strength measurements. Representative stress-strain curves of fibers made with 1.0 wt.% (solid lines) and 1.8 wt.% (dotted lines) are presented in Figure 4. Stress-strain curves for four samples of each type of fiber 12 ACS Paragon Plus Environment
Industrial & Engineering Chemistry Research 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
are reported in Figure S7 in Supporting Information. Consistent results from multiple break tests show that both the increase of SWCNT concentration and water washing improved the mechanical properties of the fibers. Fibers made with 1.8 wt.% SWCNTs have better mechanical properties than those made with 1.0 wt.% SWCNTs, showing thereby the critical importance of spinning fibers from highly concentrated dispersions. 1.0 wt.% washed fibers have a breaking strength of 249 ± 31 MPa and a Young’s modulus of 87 ± 19 GPa, whereas 1.8 wt.% washed fibers have a breaking strength of 959 ± 11 MPa and a Young’s modulus of 123 ± 15 GPa.
as-spun
washed
a
b 5 μm
1.0 wt.%
10 μm
d
c 1.8 wt.%
Figure 3. SEM micrographs of SWCNT fibers made from dispersions containing 1.0 wt.% (a and b) and 1.8 wt.% (c and d) SWCNTs. Left images show as-spun fibers and right images show fibers after water washing.
While washing the fibers in water to remove part of the insulating surfactant was primarily done to improve fiber electrical properties, mechanical properties were also strongly affected. Washed fibers show significantly higher breaking strength and modulus than unwashed fibers, independently of the SWCNT concentration. For example, unwashed 1.8 wt.% fibers show a
13 ACS Paragon Plus Environment
Page 14 of 21
Page 15 of 21 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Industrial & Engineering Chemistry Research
breaking strength of 660 ± 68 MPa and a Young’s modulus of 71 ± 12 GPa, which is about 31 % and 42 % lower than the properties of washed 1.8 wt.% fibers, respectively. The breaking strains for raw and washed 1.0 wt.% fibers are about the same, 0.9 ± 0.2 % and 1.1 ± 0.1 %, respectively. We believe that this SWCNT concentration is not sufficient to ensure strong fiber cohesion (weak draw ratio leads to SWCNT misalignment along the fiber axis) regardless of the presence of surfactant. By contrast, as-spun 1.8 wt.% fibers exhibit a large plastic deformation regime with a strain at break of 5.3 ± 1.2 %. Washed 1.8 wt.% fibers exhibit a strain at break of 1.3 ± 0.5 %. Thus, the surfactant is found to act as a plasticizer, which only shows up at high strain. The lower breaking strain of 1 wt.% fibers in comparison with the 1.8wt.% fibers does not allow us to see the plasticizer effect of the surfactant in case of 1 wt.% fibers. The presence of excess of surfactant between SWCNTs in as-spun 1.8 wt.% fibers leads to reduced van der Waals interactions between SWCNTs. The SWCNTs can therefore more easily slide over each other, resulting in a large plastic deformation regime for the macroscopic fibers. After washing, less surfactant remains between SWCNTs so they are in better van der Waals contact and much harder to break apart. Indeed, the sliding of CNT bundles past each other is the major limiting factor of the mechanical properties of the fibers under tensile loading.32, 33 This phenomenon could explain the fragile behavior and the higher breaking stress of washed fibers compared to as-spun fibers. In some cases, it might be beneficial to have both high strain at break and high electrical conductivity, however, in this system there is a tradeoff between these two properties as surfactant is removed. The specific toughness of as-spun 1.8 wt.% fibers was calculated at 30.9 ± 4.5 J.g-1 and 6.6 ± 1.2 J.g-1 for washed 1.8 wt.% SWCNT fibers. Interestingly, CNT fibers produced in this study are as flexible as textile fibers and could potentially be used as conductive thread for smart clothing.
14 ACS Paragon Plus Environment
Industrial & Engineering Chemistry Research 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 16 of 21
Figure 4. Representative tensile stress vs strain curves of fibers made from dispersions containing 1.8 wt.% (dotted lines) and 1.0 wt.% (solid lines) SWCNT. Black curves (breaking stress ≃ 240 and 710 MPa) show results before water washing and red curves show results (breaking stress ≃ 490 and 990 MPa) after water washing.
Fiber electrical conductivity was measured by a four-point probe on 6.9 cm long filaments at the room temperature. On average, 1.0 wt.% fibers have an electrical conductivity of 13 ± 0.2 kS/m before washing, and 34 ± 4 kS/m after washing. By increasing the concentration of SWCNT to 1.8 wt.%, the conductivity increased to 115 ± 20 kS/m for as-spun fibers and to 295 ± 9 kS/m for washed fibers, confirming that water washing removes insulating surfactant from the fibers. Washed 1.8 wt.% fibers were doped by iodine (a stable CNT dopant3,
34
) and reached a
conductivity of 517 ± 18 kS/m (best value of 580 kS.m-1). Mechanical properties were effectively unchanged by iodine doping (see Figure S8).
15 ACS Paragon Plus Environment
Page 17 of 21 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Industrial & Engineering Chemistry Research
Conclusion In summary, we have demonstrated a new method for dispersing SWCNTs in water using surfactants at high concentrations. Our proposed method requires shorter sonication time and hence is less destructive for SWCNTs. The CNT length is preserved via acid pre-treatment, which facilitates the dispersion of the CNTs in water with minimal supply of mechanical energy. A dialysis process was then used to produce homogeneous and highly concentrated aqueous dispersions of SWCNTs by progressive extraction of water from low concentration dispersions. The concentrated dispersions form homogeneous liquid crystalline phases, well suited for spinning of highly ordered fibers. Our results on the wet-spun fibers from concentrated aqueous dispersions of relatively long SWCNTs shows superior mechanical properties compared to previously reported CNT fibers made from aqueous dispersions of sonicated CNTs (see Figure S9). Although, the obtained values are still lower compared to the properties of the acid spun fibers (average breaking stress of 1GPa and Young’s modulus ≃ 120 GPa3), this work shows that CNT fibers can be produced without the need for corrosive solvents. These CNT fibers could serve as lightweight, chemically stable, and flexible alternatives to heavy, rigid conductive wires currently made out of metals in aerospace applications, wearable electronics, and medical devices.
ASSOCIATED CONTENT Supporting information Raman spectroscopy data for oleum-treated SWCNTs, cross-sectional surface area measurement of SWCNT fibers, cryo-TEM specimen preparation of SWCNT aqueous dispersions, 16 ACS Paragon Plus Environment
Industrial & Engineering Chemistry Research 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
bootstrapped length estimate for SWCNTs, movie of a SWCNT fiber spun continuously, SEM images of a 0.3 wt.% fiber, multiple tensile break tests on SWCNT fibers, tensile strength test on fibers before and after iodine doping, and comparison of the fiber properties reported in this work with the literature. The supporting Information is available free of charge on the ACS Publication website at DOI: XXX. AUTHOR INFORMATION Corresponding author *E-mail
[email protected]. Tel.: +1 713 348 5830. ORCID Matteo Pasquali: 0000-0001-5951-395X Notes The authors declare no competing financial interest.
ACKNOWLEDGMENTS Research was supported by Air Force Office of Scientific Research (AFOSR) grants FA9550-151-0370 and FA9550-12-1-0035, the Robert A. Welch Foundation (C-1668), and the Unites States-Israel Binational Science Foundation grant 2012223; RJH was supported by a NASA Space Technology Research Fellowship (NSTRF14), grant number NNX14AL71H. The cryoTEM was performed at the Laboratory for Electron Microscopy of Soft Materials, supported by the Technion Russell Berrie Nanotechnology Institute (RBNI).
REFERENCES (1) Saito, R.; Dresselhaus, G.; Dresselhaus, M. S. Physical properties of carbon nanotubes, World scientific: 1998. 17 ACS Paragon Plus Environment
Page 18 of 21
Page 19 of 21 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Industrial & Engineering Chemistry Research
(2) Lekawa-‐Raus, A.; Patmore, J.; Kurzepa, L.; Bulmer, J.; Koziol, K. Electrical properties of carbon nanotube based fibers and their future use in electrical wiring. Adv. Funct. Mater. 2014, 24 (24), 3661-3682. (3) Behabtu, N.; Young, C. C.; Tsentalovich, D. E.; Kleinerman, O.; Wang, X.; Ma, A. W. K.; Bengio, E. A.; ter Waarbeek, R. F.; de Jong, J. J.; Hoogerwerf, R. E.; Fairchild, S. B.; Ferguson, J. B.; Maruyama, B.; Kono, J.; Talmon, Y.; Cohen, Y.; Otto, M. J.; Pasquali, M. Strong, Light, Multifunctional Fibers of Carbon Nanotubes with Ultrahigh Conductivity. Science 2013, 339 (6116), 182-186. (4) Vigolo, B.; Penicaud, A.; Coulon, C.; Sauder, C.; Pailler, R.; Journet, C.; Bernier, P.; Poulin, P. Macroscopic fibers and ribbons of oriented carbon nanotubes. Science 2000, 290 (5495), 13311334. (5) Davis, V. A.; Parra-Vasquez, A. N. G.; Green, M. J.; Rai, P. K.; Behabtu, N.; Prieto, V.; Booker, R. D.; Schmidt, J.; Kesselman, E.; Zhou, W.; Fan, H.; Adams, W. W.; Hauge, R. H.; Fischer, J. E.; Cohen, Y.; Talmon, Y.; Smalley, R. E.; Pasquali, M. True solutions of singlewalled carbon nanotubes for assembly into macroscopic materials. Nat. Nanotechnol. 2009, 4 (12), 830-834, DOI: 10.1038/nnano.2009.302. (6) Steinmetz, J.; Glerup, M.; Paillet, M.; Bernier, P.; Holzinger, M. Production of pure nanotube fibers using a modified wet-spinning method. Carbon 2005, 43 (11), 2397-2400. (7) Dalton, A. B.; Collins, S.; Muñoz, E.; Razal, J. M.; Ebron, V. H.; Ferraris, J. P.; Coleman, J. N.; Kim, B. G.; Baughman, R. H. Super-tough carbon-nanotube fibres. Nature 2003, 423 (6941), 703-703. (8) Ericson, L. M.; Fan, H.; Peng, H. Q.; Davis, V. A.; Zhou, W.; Sulpizio, J.; Wang, Y. H.; Booker, R.; Vavro, J.; Guthy, C.; Parra-Vasquez, A. N. G.; Kim, M. J.; Ramesh, S.; Saini, R. K.; Kittrell, C.; Lavin, G.; Schmidt, H.; Adams, W. W.; Billups, W. E.; Pasquali, M.; Hwang, W. F.; Hauge, R. H.; Fischer, J. E.; Smalley, R. E. Macroscopic, neat, single-walled carbon nanotube fibers. Science 2004, 305 (5689), 1447-1450. (9) Hecht, D.; Hu, L.; Grüner, G. Conductivity scaling with bundle length and diameter in single walled carbon nanotube networks. Appl. Phys. Lett. 2006, 89 (13), 133112. (10) Yakobson, B.; Samsonidze, G.; Samsonidze, G. Atomistic theory of mechanical relaxation in fullerene nanotubes. Carbon 2000, 38 (11), 1675-1680. (11) Gspann, T. S.; Montinaro, N.; Pantano, A.; Elliott, J. A.; Windle, A. H. Mechanical properties of carbon nanotube fibres: St Venant’s principle at the limit and the role of imperfections. Carbon 2015, 93, 1021-1033. (12) Lucas, A.; Zakri, C.; Maugey, M.; Pasquali, M.; Van Der Schoot, P.; Poulin, P. Kinetics of nanotube and microfiber scission under sonication. J. Phys. Chem. C 2009, 113 (48), 2059920605. (13) Lu, K.; Lago, R.; Chen, Y.; Green, M.; Harris, P.; Tsang, S. Mechanical damage of carbon nanotubes by ultrasound. Carbon 1996, 34 (6), 814-816. (14) Yu, J.; Grossiord, N.; Koning, C. E.; Loos, J. Controlling the dispersion of multi-wall carbon nanotubes in aqueous surfactant solution. Carbon 2007, 45 (3), 618-623. (15) Barisci, J. N.; Tahhan, M.; Wallace, G. G.; Badaire, S.; Vaugien, T.; Maugey, M.; Poulin, P. Properties of Carbon Nanotube Fibers Spun from DNA-‐Stabilized Dispersions. Adv. Funct. Mater. 2004, 14 (2), 133-138. (16) Wu, X.; Morimoto, T.; Mukai, K.; Asaka, K.; Okazaki, T. Relationship between Mechanical and Electrical Properties of Continuous Polymer-Free Carbon Nanotube Fibers by Wet-Spinning Method and Nanotube-Length Estimated by Far-Infrared Spectroscopy. J. Phys. Chem. C 2016, 120 (36), 20419-20427. 18 ACS Paragon Plus Environment
Industrial & Engineering Chemistry Research 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
(17) Kozlov, M. E.; Capps, R. C.; Sampson, W. M.; Ebron, V. H.; Ferraris, J. P.; Baughman, R. H. Spinning Solid and Hollow Polymer-‐Free Carbon Nanotube Fibers. Adv. Mater. 2005, 17 (5), 614-617. (18) Razal, J. M.; Gilmore, K. J.; Wallace, G. G. Carbon Nanotube Biofiber Formation in a Polymer-‐Free Coagulation Bath. Adv. Funct. Mater. 2008, 18 (1), 61-66. (19) Mukai, K.; Asaka, K.; Wu, X.; Morimoto, T.; Okazaki, T.; Saito, T.; Yumura, M. Wet spinning of continuous polymer-free carbon-nanotube fibers with high electrical conductivity and strength. Appl. Phys. Express 2016, 9 (5), 055101. (20) Jeong, S. H.; Kim, K. K.; Jeong, S. J.; An, K. H.; Lee, S. H.; Lee, Y. H. Optical absorption spectroscopy for determining carbon nanotube concentration in solution. Synth. Met. 2007, 157 (13), 570-574. (21) Bengio, E. A.; Tsentalovich, D. E.; Behabtu, N.; Kleinerman, O.; Kesselman, E.; Schmidt, J.; Talmon, Y.; Pasquali, M. Statistical length measurement method by direct imaging of carbon nanotubes. ACS Appl. Mater. Interfaces 2014, 6 (9), 6139-6146. (22) Banerjee, S.; Hemraj-‐Benny, T.; Wong, S. S. Covalent surface chemistry of single-‐walled carbon nanotubes. Adv. Mater. 2005, 17 (1), 17-29. (23) Maillaud, L.; Zakri, C.; Ly, I.; Pénicaud, A.; Poulin, P. Conductivity of transparent electrodes made from interacting nanotubes. Appl. Phys. Lett. 2013, 103 (26), 263106. (24) Bonnet-Gonnet, C.; Belloni, L.; Cabane, B. Osmotic pressure of latex dispersions. Langmuir 1994, 10 (11), 4012-4021. (25) Islam, M.; Rojas, E.; Bergey, D.; Johnson, A.; Yodh, A. High weight fraction surfactant solubilization of single-wall carbon nanotubes in water. Nano lett. 2003, 3 (2), 269-273. (26) Wang, Y.; Iqbal, Z.; Mitra, S. Rapidly functionalized, water-dispersed carbon nanotubes at high concentration. J. Am. Chem. Soc. 2006, 128 (1), 95-99. (27) Puech, N.; Blanc, C.; Grelet, E.; Zamora-Ledezma, C.; Maugey, M.; Zakri, C.; Anglaret, E.; Poulin, P. Highly ordered carbon nanotube nematic liquid crystals. J. Phys. Chem. C 2011, 115 (8), 3272-3278. (28) Puech, N.; Grelet, E.; Poulin, P.; Blanc, C.; Van Der Schoot, P. Nematic droplets in aqueous dispersions of carbon nanotubes. Phys. Rev. E 2010, 82 (2), 020702. (29) Jamali, V.; Behabtu, N.; Senyuk, B.; Lee, J. A.; Smalyukh, II; van der Schoot, P.; Pasquali, M. Experimental realization of crossover in shape and director field of nematic tactoids. Phys. Rev. E 2015, 91 (4), 042507. (30) Onsager, L. The effects of shape on the interaction of colloidal particles. Ann. N. Y. Acad. Sci. 1949, 51 (4), 627-659. (31) Flory, P. J.; Abe, A. Statistical thermodynamics of mixtures of rodlike particles. 1. Theory for polydisperse systems. Macromolecules 1978, 11 (6), 1119-1122. (32) Salvetat, J.-P.; Briggs, G. A. D.; Bonard, J.-M.; Bacsa, R. R.; Kulik, A. J.; Stöckli, T.; Burnham, N. A.; Forró, L. Elastic and shear moduli of single-walled carbon nanotube ropes. Phys. Rev. Lett. 1999, 82 (5), 944. (33) Zhang, X.; Li, Q. Enhancement of friction between carbon nanotubes: an efficient strategy to strengthen fibers. ACS Nano 2009, 4 (1), 312-316. (34) Grigorian, L.; Williams, K.; Fang, S.; Sumanasekera, G.; Loper, A.; Dickey, E.; Pennycook, S.; Eklund, P. Reversible intercalation of charged iodine chains into carbon nanotube ropes. Phys. Rev. Lett. 1998, 80 (25), 5560.
19 ACS Paragon Plus Environment
Page 20 of 21
Page 21 of 21 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Industrial & Engineering Chemistry Research
Table of Contents Graphic
Dialysis
0.28 wt. %
200 µm
Wet spinning
1.8 wt. %
200 µm
CNT fiber
10 µm
20 ACS Paragon Plus Environment
