Single-step electrochemical functionalization of double-walled carbon nanotube (DWCNT) membranes and the demonstration of ionic rectification
- First Online:
- Cite this article as:
- Zhan, X., Wu, J., Chen, Z. et al. Nanoscale Res Lett (2013) 8: 279. doi:10.1186/1556-276X-8-279
- 2.2k Downloads
Carbon nanotube (CNT) membranes allow the mimicking of natural ion channels for applications in drug delivery and chemical separation. Double-walled carbon nanotube membranes were simply functionalized with dye in a single step instead of the previous two-step functionalization. Non-faradic electrochemical impedance spectra indicated that the functionalized gatekeeper by single-step modification can be actuated to mimic the protein channel under bias. This functional chemistry was proven by a highly efficient ion rectification, wherein the highest experimental rectification factor of ferricyanide was up to 14.4. One-step functionalization by electrooxidation of amine provides a simple and promising functionalization chemistry for the application of CNT membranes.
KeywordsCarbon nanotube membranesRectificationVoltage gatekeeperAmine electrooxidation
Electrochemical impedance spectra
A protein channel embedded in a cell membrane functions as a natural regulator in the biological system. Conformational change of proteins actuated by voltage can open or close the gate of the channel, which regulates ion permeation with high selectivity [1–4]. It inspires researchers to develop artificial nanopores and nanochannels in response to external signals (voltage, pH, temperature, light, etc.) by mimicking natural ion channels . Transmembrane voltage is an excellent stimulus to open or close the gate of a nanodevice since it is not aggressive, is tunable, and can act over a short time scale . Therefore, it can modulate ionic flux and rectify ionic transport current through the nanochannel/nanopore. These nanodevices acting as rectifier enable the possible applications in single-molecule sensing and separation [7–10]. Carbon nanotube (CNT) membranes offer a fast fluid platform. The fluid velocity of a carbon nanotube membrane is 10,000 times faster than the conventional membrane of similar pore size due to atomically smooth graphite core [11, 12]. Moreover, the CNT membranes have far more mechanical strength than lipid bilayer films, thus providing an exciting opportunity for chemical separation, drug delivery, and other applications [13, 14]. Carbon nanotube membranes can imitate ion channels with functionalized molecules acting as mimetic gatekeepers. Chemical functionalization of molecules (biotin , phosphorylation , and charged dye ) at the entrance of the CNT core enables the modest modulation of ionic transportation. Further study had shown that the steric hindrance of gatekeepers at the pore entrance can be controlled with voltage . Negative bias repels the anionic tethered molecules away from the CNT entrance, opening the channel, while positive bias pulls the anionic tethered molecules into the pore, thus closing the channel. The voltage-gated carbon nanotube membranes have been successfully applied in drug delivery. CNT membranes enable the programmable delivery of the addictive drug nicotine into the human skin in vitro for abuse treatment . Neutral caffeine can also be pumped through CNT membranes via a highly efficient electroosmotic flow that is 100-fold more power efficient compared to conventional materials such as anodized aluminum oxide membranes .
To achieve gatekeeper activity on CNT membranes, there needs to be a high functional density only at the CNT tips or pore entrances [12, 21]. This has been largely achieved with a two-step process, wherein diazonium grafting first creates carboxyl groups at the CNT tips followed by carbodiimide coupling chemistry [17, 22]. Diazonium grafting generates highly reactive radicals that covalently react with the electrode or subsequent organic layer on the surface under mild solvent and temperature conditions [23, 24]. However, it is difficult to control the amount of carboxylate groups on the CNT tip due to polymerization during diazonium grafting [24, 25]. In principle, grafting reaction is self-limiting when an insulating polymer layer stops the electrochemical reduction of diazonium salt. However, with ionic functional groups (such as carboxylates), the reaction can proliferate and block carbon nanotubes. Another complication of the diazonium approach is that it generally requires two-step functionalization since the diazonium formation reaction is not compatible with many functional groups that would be required on the gatekeeper. This adds complication and reduces the overall yield. Electrochemical oxidation of amine to coat carbon fiber surface predates diazonium grafting with its first report in 1990 . It enables immobilization of various primary amine-containing molecules on different electrode surfaces [27–31]. The electrografted layer is characterized by atomic force microscopy, X-ray photoelectron spectroscopy, ellipsometry, time-of-flight secondary ion mass spectrometry, and electrochemistry methods [32–34]. Amine electrochemical oxidation greatly simplifies the surface modification process since it does not need complicated synthesis and surface chemistry. Even large molecules including dendrimers and metal-ligand complex can be directly functionalized on a conductive surface in a single step [35–38]. Electrografting of amine offers a simple and efficient functional chemistry for CNT applications. Electrografting of amine provides binding sites on CNTs for the coating of Pt-Ru and Ag nanoparticles that exhibit excellent electrocatalytic activity [39, 40]. The more controllable electrochemical grafting of the fluorinated aminobenzoic acid layer enables the Pt monolayer deposition on CNT buckypaper. The highest record of mass activity has been achieved at 2,711 A g−1 in methanol oxidation .
The primary hypothesis of this paper is that the efficiency of voltage gatekeeping can be enhanced to obtain high on/off ratio using electrooxidation of amine in one step. The conformational changes of tethered dye molecules under bias will be identified by non-faradic electrochemical impedance spectroscopy (EIS) measurements. The EIS spectra can prove the effectiveness of this single-step functionalization on double-walled carbon nanotube (DWCNT) membranes. Transmembrane ionic rectification will be measured to compare the efficiency of gatekeeping. Stronger rectification indicates more efficient gatekeeping. The gatekeeper density is still unknown in our previous work. This can be quantified by dye assay on glassy carbon due to its similar structure with CNTs. A single-step modification may give higher overall functional density over a complicated two-step modification.
Fabrication of double-walled carbon nanotube membranes
Modification of DWCNT membranes
To avoid grafting in the inner core of CNTs, CNT membranes were placed in U-tube fittings under a 2-cm inner solution column pressure. In two-step functionalization, as-prepared DWCNT membranes were first modified by flow electrochemical grafting with 5-mM 4-carboxy phenyl diazonium tetrafluoroborate/0.1-M KCl solution at −0.6 V for 2 min. In the next step, Direct Blue 71 dye (Sigma-Aldrich) was coupled with the carboxyl group on the tip of CNTs with carbodiimide chemistry: 10 mg of ethyl-(N′,N′-dimethylamino) propylcarbodiimide hydrochloride and 5 mg of N-hydroxysulfosuccinimide were dissolved into 4 ml of 50-mM Direct Blue 71 dye in 0.1 M 2-(N-morpholino) ethane sulfonic acid buffer for 12 h at ambient temperature.
In one-step functionalization, Direct Blue 71 dye, which has a primary amine, was directly grafted to CNT by electrooxidation of amine. Electrografting was carried out under a constant potential of 1.0 V using a potentiostat (E-corder 410, eDAQ, Denistone East, Australia) in the three-electrode cell. The CNT membrane, with sputtered Pd/Au film (approximately 30-nm thick) on the membrane's back side, was used as the working electrode; Pt wire was the counter electrode, and the reference electrode was Ag/AgCl. Before electrografting, the ethanol solution of 0.1 M LiClO4/1 mM direct blue was purged by argon gas for 15 min to remove adsorbed oxygen in the solution.
Rectification experimental setup
The schematic of the ionic rectification setup is shown in Additional file 1: Figure S1. Both U-tube sides were filled with potassium ferricyanide solution. The working electrode (W.E) was DWCNT membrane coated with 30-nm-thick Pd/Au film; the reference electrode (R.E) was Ag/AgCl electrode. Voltage was controlled using an E-Corder 410 potentiostat. The counter electrode was a sintered Ag/AgCl electrode purchased from IVM Company (Healdsburg, CA, USA). The membrane area was approximately 0.07 cm2. Linear scan was from −0.60 to +0.60 V with the scan rate at 50 mV/s.
Dye assay quantification of carboxyl and sulfonate density on glassy carbon
Toluidine blue O was reported to quantify the carboxyl group density on the polymer film. Our dye assay method was similar to that of previous reports [43, 44]. Glassy carbon was incubated in 0.2-mM toluidine blue O (TBO, Sigma-Aldrich) solution at pH 10 and at room temperature for 1 h to adsorb positively charged dye onto the anionic carboxylate or sulfonate group. The glassy carbon was then rinsed with NaOH (pH 10) solution and further incubated in 0.1-mM NaOH (pH 10) solution for 5 min to remove physisorbed TBO dye. The adsorbed TBO on anionic glassy carbon was removed from the HCl solution (pH 1). The concentration of desorbed TBO in the HCl solution was determined by the absorbance at 632 nm using Ocean Optics (Dunedin, FL, USA) USB 4000 UV–vis spectrometer. The calculation of carboxyl or sulfonate density was based on the assumption that positively charged TBO binds with carboxylate or sulfonate groups at 1:1 ratio on glassy carbon.
Results and discussion
Summary of ionic rectification factor on single-step modified DWCNT-dye membrane
7.2 ± 0.3
3.1 ± 0.3
2.4 ± 0.2
6.4 ± 1
2.0 ± 0.1
2.0 ± 0.1
5.6 ± 1
2.3 ± 0.1
1.7 ± 0.1
Comparison of ionic current rectification factor in K3Fe(CN)6solution
Concentration of K3Fe (CN)6
Single-step electrooxidation of amine
Electrochemical grafting of diazonium and coupling of dye
Chemical grafting of diazonium and coupling of dye
3.9 ± 0.8
14.4 ± 0.6
2.9 ± 0.2
4.0 ± 0.4
4.4 ± 0.9
9.8 ± 0.3
2.9 ± 0.2
3.3 ± 0.07
3.4 ± 0.1
8.0 ± 0.4
3.2 ± 0.3
3.6 ± 0.2
Summary of ionic rectification factor on DWCNT membrane after water plasma oxidation to remove gatekeepers
3.2 ± 0.3
1.7 ± 0.2
2.4 ± 0.2
2.8 ± 0.3
1.5 ± 0.07
2.0 ± 0.2
2.4 ± 0.2
1.4 ± 0.0.02
2.0 ± 0.2
Ferricyanide has a well-known redox potential of 0.17 V (vs. Ag/AgCl), and thus, an important control experiment was done to make sure that the observed rectification was not due to faradic current; instead, it was due to transmembrane ionic current. Cyclic voltammetry scans (−0.6 to 0.6 V) showed no redox reaction on both as-made and one-step functionalized DWCNT membranes in 50-mM ferricyanide (Additional file 3: Figure S3). We also did not observe redox reaction on glassy carbon in 2-mM ferricyanide, as seen in the flat curve in Additional file 4: Figure S4A. The much larger conductive area of the glassy carbon electrode compared to 5% DWCNT membrane requires the use of more diluted (2 mM) ferricyanide solution. However, with the supporting 0.5-M electrolyte KCl solution, the oxidation and reduction peaks were observed at 0.29 and 0.06 V, which were similar to those found in reports [30, 50]. The experiment was also repeated with both redox species. In Additional file 4: Figure S4B, no redox peak was found on glassy carbon in 50-mM ferricyanide solution and 25-mM ferricyanide/ferricyanide solution. The control experiments of cyclic voltammetry on DWCNT membrane and glassy carbon ruled out the redox reaction of ferricyanide, which supports the ionic rectification on electrochemically grafted CNT membranes.
Quantification of carboxyl and sulfonate density using dye assay
Modification on glassy carbon
Step 1 in two-step functionalization
Electrochemical grafting of 4-carboxyl phenyl diazonium for 8 min
1.3 × 1015
1.3 × 1015
Step 2 in two-step functionalization
Carbodiimide coupling of dye
2.0 × 1015
1.07 × 1015
0.93 × 1015
Electrochemical grafting of dye by amine oxidation for 8 min
0.9 × 1015
0.9 × 1015
DWCNT membranes were successfully functionalized with dye for ionic rectification by electrooxidation of amine in a single step. Non-faradic (EIS) spectra indicated that the functionalized gatekeeper by one-step modification can be actuated to mimic the protein channel under bias. This functional chemistry was proven to be highly effective on the enhancement of ion rectification, wherein the highest experimental rectification factor of ferricyanide was up to 14.4. The control experiments supported that the observed rectification was a result of transmembrane ionic current instead of electrochemical reaction of ferricyanide. With the decreasing size of ion, we have observed smaller rectification due to partially blocked ion channels. The rectification was decreased with the higher ionic concentration. It suggested that the rectification is attributed to both charge and steric effects at low concentration, while the steric effect is dominant at high concentration. After removing the dye, the DWCNT-dye membrane exhibited no enhancement of rectification. This control experiment supported that the rectification was induced by functionalized dye molecules. The saturated functionalized dye density by a single step was quantified at 2.25 × 1014 molecules/cm2 on glassy carbon by dye assay, the same as that of two-step functionalization. However, no apparent change of rectification was observed for two-step functionalization. The dye molecules on the membrane by single-step functionalization are more responsive to the applied bias due to direct grafting on the conductive surface instead of the grafted organic layer. Another possible reason is that the actual yield of the second step of the two-step modification on CNT membranes may be much less than the calculated 18% yield on glassy carbon. One-step functionalization by electrooxidation of amine provides a simple and promising functionalization chemistry for the application of CNT membranes.
This work was supported by NIDA, #5R01DA018822-05, DOE EPSCoR, DE-FG02-07ER46375, and DARPA, W911NF-09-1-0267. Critical infrastructure provided by the University of KY Center for Nanoscale Science and Engineering.
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.