Skip to main content
NCBI home page
ACS AuthorChoice logoLink to ACS AuthorChoice
. 2020 Feb 24;20(3):1846–1854. doi: 10.1021/acs.nanolett.9b05066

Dendrimer-Au Nanoparticle Network Covered Alumina Membrane for Ion Rectification and Enhanced Bioanalysis

Chen Wang †,‡,*, Xiao-Ping Zhao †, Fei-Fei Liu †, Yuming Chen §, Xing-Hua Xia ‡,*, Ju Li §,*
PMCID: PMC7307958  PMID: 32091912

Abstract

graphic file with name nl9b05066_0006.jpg

Ion transport in an artificial asymmetric nanoporous membrane, which is similar to biological ion channels, can be used for biosensing. Here, a dendrimer-Au nanoparticle network (DAN) is in situ assembled on a nanoporous anodic aluminum oxide (AAO) surface, forming a DAN/AAO hybrid membrane. Benefiting from the high surface area and anion selectivity of DAN, the prepared DAN/AAO hybrid presents selective ion transport. Under a bias potential, a diode-like current–potential (I–V) response is observed. The obtained ionic current rectification (ICR) property can be tuned by the ion valence and pH value of the electrolyte. The rectified ionic current endows the as-prepared DAN/AAO hybrid with the ability of enhanced bioanalysis. Sensitive capture and detection of circulating tumor cells (CTCs) with a detection limit of 80 cells mL–1 as well as excellent reusability can be achieved.

Keywords: dendrimer-Au nanoparticle network, hybrid membrane, ion selectivity, ionic current rectification, enhanced bioanalysis

Ionic current rectification (ICR) is seen in biological or artificial nanochannels,1−4 where a nonlinear and asymmetric current–potential (I–V) curve appears due to asymmetric ionic transport. Generally, ICR can occur in the cases of symmetry breaking in the structure, surface charge distribution, bulk electrolyte concentration, or combinations of these factors.5−7 The correlation between ICR properties and the degree of symmetry breaking was precisely demonstrated, taking the DNA-stuffed nanochannel as the model system.4 Up to now, a variety of techniques have been developed for the fabrication of ICR devices such as direct modification, plasma grafting, and ion sputtering,8−16 in nanofluidic devices including nanopores with an asymmetric geometry,8,9 homogeneous nanochannels with asymmetric surface modifications,10,11 hybrids composed of varied nanomaterials,12−14 and dynamic curvature nanochannels.15 For example, by laminating the polystyrene-b-poly(4-vinylpyridine) membrane and polyethylene terephthalate membrane, a heterogeneous polymer membrane displaying a highly ICR characteristic was constructed.16 Recently, some inert or conductive nanochannels/nanopores have been directly fabricated on thin metal films, on which detection of nanoparticles,17,18 biomolecules,19 or asymmetric ionic transport20 have been successfully performed.

The network formed by functional nanomaterials and ligands provides a simple, robust, and highly efficient fabrication technique for the construction of novel nanoporous structures.21−24 Compared to traditional solid nanoporous materials, the network has a more exposed surface area. For example, by grafting gold spheres on a glass nanopipette with the DNA linker, a pH-responsive structure was successfully fabricated.24 Owing to the tunable distance within Au nanoparticles (AuNPs) by the DNA linker, the fabricated nanodevice presents varied ICR properties under different pH values. The nonuniform geometry and surface charge density endow the formed device with tunable ICR properties. These interesting works provide new strategies to design and develop multifunctional heterogeneous hybrid devices with a high ICR response for seawater desalination,13 energy conversion,16 and bioanalysis/sensors.8,9

Sensitive and selective cell capture and detection play important roles in the prognosis and treatment of cancer.25 Recently, we developed a nanochannel–ion channel hybrid with an ICR property for the efficient capture and detection of circulating tumor cells (CTCs).26 Due to the unique ICR characteristic, the CTC concentrations ranging from 1 × 102 to 2 × 106 cells mL–1 can be successfully detected with a detection limit of 100 cells mL–1. However, in that work, only the outer surface of the anodic aluminum oxide (AAO) membrane can be used. The fewer surface areas, as well as a relatively smooth surface, is a big challenge for efficient cell capture. In addition, the ion channel dimensions are fixed, and cannot be tuned flexibly. The super small size of the ion channels may prevent the relatively bigger drug molecules or negatively charged molecules from passing through, which will limit its applications in cell downstream analysis, such as drug transport, screening, and cancer therapy.

Herein, we reported a new ionic rectification device fabricated by assembling the AuNPs network on the surface of the AAO membrane for CTCs analysis. Dendrimer (poly(amidoamine) (PAMAM)) and carbon disulfide (CS2) were used as the linkers to connect AuNPs, forming a dendrimer–nanoparticle network (denoted as DAN in the following work) film. This cross-linked DAN film can be in situ assembled onto the surface of AAO by chemical coupling of the −N–C bond in DAN with amine-functionalized AAO, forming the DAN/AAO hybrid. The high surface-to-volume ratio, as well as exposed functional groups on the DAN film, will provide abundant active sites for molecular anchoring, which is of critical importance to efficient and selective cell capture. The symmetry breaking by covering DAN on one side of AAO endows the hybrid with an excellent ICR property, ensuring highly sensitive cell detection. Finally, the captured cells can be released with nearly no damage under the action of benzonase, endowing the hybrid with the ability to be reused with excellent durability. Therefore, the present DAN/AAO hybrid is expected to behave as a robust and sensitive biosensor toward cell analysis and detection. Besides, the AuNPs in DAN film possesses an unique localized surface plasmon resonance (LSPR) characteristic that would enable highly accurate characterizations such as surface-enhanced Raman scattering (SERS)27 and electrochemistry28 and open an avenue for combining fluidics with plasmonics.

Results and Discussion

Fabrication and Characterization of the DAN/AAO Hybrid

The fabrication process is illustrated in Figure 1A. AAO was prepared using the method described previously.29 To obtain the amine-functionalized AAO, AAO was immersed into a 1% (3-aminopropyl) triethoxysilane (APTES) solution for 12 h, followed by rinsing with ethanol and drying in nitrogen gas. The prepared amine-functionalized AAO was then placed at the bottom of a small vial containing a mixture consisting of 4-dimethylaminopyridine (DMAP) stabilized AuNPs (∼7 nm),30 CS2 methanol solution, and PAMAM dendrimer methanol solution. The gold nanoparticles were assembled on the top surface of AAO linked by CS2, forming the DAN film. Following the completion of the reaction, the remaining solution was removed, and the DAN/AAO hybrid was obtained.

Figure 1.

Figure 1

Open in a new tab

(A) Schematic illustration of the preparation of the DAN/AAO hybrid. (B–F) SEM images of the DAN/AAO hybrid. (B) Top of the DAN/AAO hybrid. (C) Cross section of AAO. (D) DAN film. (E) Cross section of the DAN/AAO hybrid. The inset is the enlarged image of the cross section. (F) The thickness of the whole DAN/AAO hybrid. (G) The contact angle of the DAN film. (H–I) XPS spectra of the DAN/AAO hybrid ((H) XPS curve of APTES; (I) XPS curve of AuNPs). (J) UV–vis spectra of the prepared AuNPs and DAN film. Black curve: pure AuNPs solution. Red curve: AuNPs containing the PAMAM dendrimer and CS2 solution. The inset shows the TEM image of the prepared AuNPs; the scale bar is 20 nm. (K) SAXRD patterns for the DAN composites.

To characterize the morphology of the fabricated AAO, scanning electron microscopy (SEM) was used. Figure 1B–E indicates the successful formation of the DAN/AAO hybrid. The fabricated AAO has regular pores of ∼40 nm in diameter (Figure 1B,C). The upper layer of AAO was covered by the DAN film (Figure 1D,E). The thickness of the DAN film is around 40 nm (inset of Figure 1E). The entire thickness of the DAN/AAO hybrid is ∼50 μm (Figure 1F). The contact angle on both sides of the hybrid was measured (Figure 1G and Figure S1), indicating the hydrophilicity of the DAN film and the hybrid. High-resolution spectra of X-ray photoelectron spectroscopy (XPS) was used to further prove the successful assembly of DAN. It is notable that the DAN/AAO hybrid exhibits singular signals at 101.4 eV ascribed to −Si–O– from the coupling reagent of APTES (Figure 1H) and two obvious Au 4f peaks at 82.5 and 86.5 eV, respectively, from AuNPs (Figure 1I). In addition, energy dispersive spectroscopy (EDS) mapping of the SEM images (Figure S2) was used to identify the elemental distribution of Au, Si, N, O, and C.31 It can be seen that the Au and Si atoms are dispersed homogeneously in the architecture, indicating the formation of the DAN film on the top of AAO. These results demonstrate the successful formation of the DAN/AAO hybrid.

To verify the formation mechanism of the DAN film, UV–vis spectra were used to examine the interaction between PAMAM and AuNPs. The results are shown in Figure 1K, showing a dominant absorption peak at ∼518 nm (black curve in Figure 1J, an average diameter of 7 ± 1 nm in this work shown in the inset of Figure 1J). This absorption resulted from the characteristic plasmonic absorption of isolated AuNPs. After reaction with CS2 and the PAMAM dendrimer, the absorption peak shifted to ∼647 nm due to the change in the dielectric environment (red curve in Figure 1J). The results are in accordance with the previous work.24 The interparticle spacing for the DAN film was determined by the small-angle X-ray diffraction (SAXRD) patterns (Figure 1K) using the Bragg diffraction eq (2θ = 2.4) is less than 4.5 nm.24 To further confirm the interparticle spacing for the DAN film, AuNPs, PAMAM, and CS2 were assembled in solution to form the same structure of the DAN film. The TEM image (Figure S3) shows that the nanopores exist among neighboring AuNPs in the DNP film with the size of interparticle spacing ∼ 3.6 nm.

Ion Transport Properties of the DAN/AAO Hybrid

The ion selectivity of the DAN/AAO hybrid was examined by cyclic voltammetry (CV) using Ru (NH3)63+ or Fe (CN)63– as the electrochemical probes, respectively, in 1 mM KCl aqueous solution. The electrochemical setup is schematically shown in Figure 2A. Then, 0.5 mM Ru (NH3)63+ or Fe (CN)63– was put into one of the two half cells for CV measurements. Gold, Ag/AgCl, and platinum wire were used as the working electrode, reference electrode, and auxiliary electrode, respectively. Figure 2B,C shows the CVs obtained with bare AAO and the DAN/AAO hybrid. Both Ru(NH3)63+ and Fe (CN)63– could transport through the bare AAO as evidenced by the obvious redox peaks. In comparison with the bare AAO, the DAN/AAO hybrid film shows anion selective transport properties due to the existence of surface charges since only the redox current for the ferricyanide anion can be observed (Figure 2C). There is no redox current observable for the electrochemical probe of Ru(NH3)63+ cation (Figure 2B). This suggests DAN is anion-selective, allowing only anions to pass. Because the pKa of DMAP ligands on the AuNPs surface is 9.6,30 DAN should be positively charged when the pH is less than 9.6. The interstitial spacing in DAN is less than 4.5 nm, so it interacts very strongly with ions that try to pass through the interstices of the network.

Figure 2.

Figure 2

Open in a new tab

(A) Setup for CV measurements. The substrate in the figure is used to fix the DAN/AAO hybrid. CVs of the DAN/AAO hybrid (red curve) and bare AAO (blue curve) electrode in 1 mM KCl containing 0.5 mM Ru(NH3)63+ (B) and Fe(CN)63–(C). The scan rate was 50 mV s–1. (D) Schematic illustration of the setup for the measurement of I–V curves. (E) I–V curve of AAO in 1 mM KCl at pH 7.0. (F) I–V curve of the DAN/AAO hybrid in 1 mM KCl at pH 7.0. (G) Rectification ratio of the DAN/AAO hybrid at different AuNP concentrations (30 mM, 6 mM, 3 mM, 0.6 mM, 0.3 mM). (H, I) Illustration of the ICR mechanism of the DAN/AAO hybrid.

The ionic transport properties of the DAN/AAO hybrid were investigated using two Ag/AgCl electrodes (Figure 2D). For the pure AAO film, an Ohmic linear I–V curve is observed (Figure 2E). After DAN was modified to the AAO surface, the ionic current displays a diode-like I–V curve (Figure 2F) due to the existence of asymmetry in the structure and surface charge density from the linkers. To achieve the high-performance ICR device, the influencing factors including AuNPs concentrations, channel size, and AAO thickness were investigated. Figure 2G shows the effect of AuNP concentration on ICR (the I–V curves refer to Figure S4). With an increasing AuNP concentration, the ICR ratio (f ≡ I+1 V/I–1 V, I+1 V is denoted as the current value at +1.0 V, I–1 V is denoted as the absolute current value at −1.0 V) significantly increases to a maximum value of 19.3 when the film was prepared with 30 mM AuNPs. It has been reported that decreasing the density of dendrimers between particles will decrease the interparticle spacing.32 To keep the same quantity of dendrimers, an increased AuNPs concentration will decrease the ratio of dendrimers in the DAN layer, resulting in a smaller interparticle spacing in the DAN film. The ion transport is influenced by both ion selectivity and ion conductivity. The ion conductivity of the hybrid will decrease as the interparticle spacing in the DAN film gets smaller, resulting in a decrease of ionic current (Figure S4). However, the ICR property was determined by the distinct transport actions of co-ions and counterions in the nanochannels.10,33 The decreased channel size will contribute to more excellent ion selectivity. Thus, a higher ICR ratio can be expected in the case of a smaller interparticle spacing in the DAN film. The effects of the AAO channel size and thickness (Figure S5 and Figure S6) indicate that the channel size could influence the ICR property efficiently, while AAO thickness does not. The effect of surface charges will be weakened with the increase of channel size, which in turn decreases the ICR efficiency. Interestingly, although there is little influence of the AAO thickness on the dimensionless ICR ratio, it was found that the ionic current decreases slightly as AAO becomes thicker (Figure S6A), caused by the increased electric resistance for thicker AAO. Considering all factors in the fabrication and device performance, the DAN/AAO hybrid prepared by 30 mM AuNPs and AAO with a 40 nm diameter was used in the following measurements.

The ICR mechanism of the present hybrid is schematically shown in Figure 2H,I. While the pKa of pure AAO is around 6, the pKa of the primary amine group of the PAMAM dendrimer and APTES is around 10.0,34 so we expect our amine-functionalized AAO to be positively charged when the pH value is less than 10.0. Under a pH = 7 environment, therefore, both DAN and AAO should be positively charged. When the electric field is applied from the bottom (AAO) side to the top (DAN) side, cations (K+) are driven from bottom to top, and anions (Cl–) from top to bottom. While Cl– can freely pass the anion-selective DAN, K+ would be stopped by the DAN and accumulate inside the AAO nanochannel. To maintain electroneutrality, some Cl– would stay to accompany K+, increasing the ionic strength. The increased ion concentration within the AAO nanochannel results in a reduced electrical double layer (EDL) thickness, and so more Cl– can pass through the hybrid, unencumbered by the charge on amine-functionalized AAO. We called this the “on” state (Figure 2H). In comparison, when the electric field is applied from the top (DAN) side to the bottom (AAO) side, the cations (K+) are driven from top to bottom, which, however, will be excluded by DAN because it is anion-selective. Also, those K+ previously stored inside the nanochannel will be driven out and exit from the bottom. The Cl– are driven from bottom to top, but they flow through DAN, so they are not accumulated inside the AAO tubule either; therefore, the ions within AAO nanochannels are depleted, resulting in a reduced ionic strength, larger EDL, and decreased Cl– current. Under the extreme condition, nearly no Cl– ions flow through the hybrid, and accordingly, the hybrid presents the “off” state (Figure 2I).

Effect of pH on Ion Transport

The pH value has a significant influence on the ion transport behavior of the present DAN/AAO hybrid. Figure 3A shows the measured I–V curves of the DAN/AAO hybrid under different pH values. Figure 3B is the rectification ratio calculated from Figure 3A, which decreases as the pH value changes from 3.0 to 10.0, then rapidly declines to f < 1 (reversed ICR direction) at pH 11.0. The pH-tunable ICR property is schematically illustrated in Figure 3C. As has been described above, the present hybrid is positively charged when the pH value is less than 10.0. At a lower pH, the DAN/AAO hybrid surface will carry more positive charges that attract more anions passing through the DAN/AAO hybrid. Thus, a higher ICR property can be expected. With increasing the pH value to 10.0, the ionic current and ICR ratio decrease due to the declined positive surface charge. It is furthermore observed that the I–V property of the DAN/AAO hybrid sustains reversion at pH 11.0 (f = 0.3). The amino groups are neutral at a high pH,15 but the acylamion groups from PAMAM may partially hydrolyze at a high pH value (pH 11.0) because the pKa of acylamion groups is around 10.0.35 Therefore, the surface of AAO nanochannels is neutral, but the DAN film might be negatively charged in this condition. The ionic current will be dominated by cations (K+) rather than Cl– ions. Under this circumstance, the I–V properties of the DAN/AAO hybrid is reversed: a high conducting (“on”) state for V < 0 and a low conducting (“off”) state for V > 0. When the solution pH is equal to pKa, the hybrid is electroneutral, and thus nearly no ICR will be observed. To keep a constant pH value, the effect of temperature on the mass transport property of the hybrid was also investigated (Figure S7). The currents of both “on” and “off” states increase as the temperature increases. However, the current of the “on” state increases more significantly than that in the “off” state, resulting in an increased ICR ratio at a higher temperature. This result is in accordance with previous work.36

Figure 3.

Figure 3

Open in a new tab

I–V curves measured on the DAN/AAO hybrid in solutions with different pH values. (B) Rectification ratio of the DAN/AAO hybrid in solutions with different pH values. (C) Mechanism of ion transport through the DAN/AAO hybrid in solutions with varied pH values (brown spheres represent anions; blue spheres represent cations). (D) The ζ potential of DAN fragments at different pH values.

To further quantify the charge change at different pH, the ζ potential of the self-assembled DAN fragments was determined by mixing AuNPs, CS2, and PAMAM dendrimer. As shown in Figure 3D, the measured ζ potentials changed from 38.0 mV to 2.71 mV, indicating a decreased positive charge density from pH 3.0 to pH 10.0. As expected, the ζ potential changes to −12.7 mV at pH 11.0. The results demonstrate that the solution pH can regulate the surface charges of the DAN/AAO hybrid via protonization/deprotonization, which regulates the ICR properties.

Effect of Ions Concentration and Valence

To further illustrate the ionic transport properties of the DAN/AAO hybrid, different electrolyte concentrations at pH 7.0 were added into two half cells. As shown in Figure 4A, the DAN/AAO hybrid shows different rectified currents for different ion concentrations. The transmembrane current increases with increasing the ion concentration from 0.001 mM to 1 M due to the increased amount of ions through the DAN/AAO hybrid. The ICR ratio (f) increases to 65.4 at 100 mM solution; then, it decreases with further increasing of the ion concentrations (Figure 4B). It is clear the ICR ratio reaches a maximum value when the surface charges are balanced by the counterions in the EDL.37 Note that the thickness of electrical double layers is affiliated with the ionic Debye length, λD, with λD ∝1/(cb)1/2, where cb is the bulk ionic concentration.38 Therefore, upon increasing the ionic strength to 1 M, the thickness of EDL decreased, suggesting the decrease of ionic selectivity due to the decline of the EDL screening effect, and thus, the ICR ratio declines. For a system with lower electrolyte concentrations, although the EDL overlapping occurs, the ICR decreases due to fewer ions transported across the channel,37,39 thus leading to a decline of f.

Figure 4.

Figure 4

Open in a new tab

(A) I–V curves of the DAN/AAO hybrid at different concentrations of KCl. (B) Rectification ratio of the DAN/AAO hybrid at different concentrations of KCl. (C) I–V curves of the DAN/AAO hybrid at different electrolytes (KCl, K2SO4, and MgCl2) with the same concentration in 1 mM. (D) Current value at +1.0 V at different electrolytea (KCl, K2SO4, and MgCl2) with different concentrations. The inset was the enlargement of panel D at a low concentration. (E) Rectification ratio of the DAN/AAO hybrid at different electrolytes (KCl, K2SO4, and MgCl2) with the same concentration in 1 mM at a potential from +0.2 V to +1.0 V. (F) Rectification ratio with different electrolytes under different concentrations.

The valence is another important factor affecting the ion transport characteristics in nanofluidics because it is directly related to the charge inversion process.40 In this study, the I–V curves of the DAN/AAO hybrid with different valence counterions were measured under different ion concentrations (Figure 4C,D). As shown in Figure 4C, the ion current displays I(MgCl2) > I(KCl), I(K2SO4) > I(KCl), regardless of the applied potential bias. This phenomenon may correlate with two parameters in principle.40 The first one is ion concentration, which is higher for the divalent salt at a constant concentration. The second possible factor could correspond to the charge of the ions. Figure 4E illustrates that the largest ICR ratio of the DAN/AAO hybrid is obtained in a 1 mM K2SO4 solution regardless of the applied potential bias due to the anion-selective hybrid membrane. Figure 4F demonstrates that f has a local maximum (fmax) with varying salt concentrations. In our study, the fmax of KCl and K2SO4 is achieved at 100 mM and 10 mM, respectively. It is reported that the f could reach its maximum value when the charge on the surface of the DAN/AAO hybrid is balanced by counterions in the EDL.37 For the DAN/AAO hybrid, the inner surface of DAN/AAO exhibits a positive charge from the amino group at pH 7.0. Since the sulfate ion carries charges as high as twice of the chloride ion, the shielding effect will be stronger in the former case. Consequently, the f of the former reaches the maximum value at a lower electrolyte concentration. In a similar way, the electrolyte concentration at which fmax occurs shifts to a lower level when the cation valence increases. The fmax for the case of MgCl2 occurs at 10 mM. These phenomena have been explained previously.41 In addition, if two salts have the same total valences, the salt concentration at which fmax occurs is the same, but the value of fmax depends on ionic diffusivity.40 For K2SO4 (1:2) and MgCl2 (2:1), the salt concentrations at which fmax occur are similar (10 mM). However, the value fmax for the case of K2SO4 (70.3) is larger than the value fmax for MgCl2(55.9). As described above, the as-prepared hybrid membrane is anion-selective, and the anions dominate the ICR property. The charge of a sulfate ion is twice as high as a chloride ion. Keeping a similar liquid flowing rate, more negative charges could be transported through the hybrid in the case of sulfate ions, leading to a higher fmax for K2SO4 than MgCl2. These results demonstrate the important role of ion valence in the charge inversion and high ICR ratio in an asymmetric nanochannel.

Selective Cell Capture, Detection, and Release

In our DAN/AAO concept, while AAO provides structural support,42 the DAN covering has a large exposed surface area to the external flowing medium and can be biofunctionalized. Au is well-known to be a highly versatile substrate for chemical and biological analyses. The high surface-to-volume ratio and exposed functional groups on the DAN film provide abundant active sites and room for probing molecular anchoring. Therefore, the prepared hybrid is expected to behave as a biosensor. The performance of the fabricated hybrid as a biosensor was then tested including sensitivity, selectivity, and durability. First, the ability of the present hybrid for cell capture and detection was investigated. As shown in Figure 5A, based on previous work,26,43 the aptamers that could recognize the special protein on CTC membranes are first immobilized on the DAN surface. As the sample flows through the hybrid, the target CTCs will be captured due to molecular recognition. The addition of benzonase nuclease leads to the cleavage of CTCs from the hybrid surface. During the course of cell capture and release, the ICR property will vary accordingly. The larger ionic current value with “on” state endows a sensitive capture, and detection of CTCs can be realized. The characterizations of the aptamer immobilization, the optimization of aptamer concentration, and the CTCs capture kinetics are detailed described in Supporting Information and Figures S8–10. The electrochemical assay was then adopted for the CCRF-CEM cell detection of various concentrations. I–V curves with different concentrations of CCRF-CEM cells captured were recorded and are shown in Figure 5B. The ionic current at +1.0 V versus cell concentrations is displayed in Figure 5C. Owing to the increasing blockage effect by the captured cells, the currents decrease with cell concentration. In addition, as shown in the inset of Figure 5C, the current (I) is inversely proportional to the logarithm of cell concentrations from 300 to 1000 cells mL–1. The calibration equation is I [μA] = −0.063log10C + 0.26 (where I is the current response at different cell concentrations. C is the concentration of CCRF-CEM cells in unit of cells/mL) with a correlation coefficient of 0.99. The detection limit is 80 cells mL–1. Compared with other CTCs detection methods (Table S1 in the Supporting Information), the present method could reach a lower detection limit without the need for introducing any signal amplification strategy.

Figure 5.

Figure 5

Open in a new tab

(A) Schematic demonstration of the CTCs capture and release processes and the corresponding I–V curves of the DAN/AAO hybrid. (B) I–V properties of the DAN/AAO hybrid after capturing different concentrations of CCRF-CEM cells (50, 80, 100, 300, 500, 700, 1000, 104, 105, 106, and 107 cells mL–1). (C) Currents at +1.0 V vs CCRF-CEM cell concentration. Inset: the linear calibration plot of the changed current value vs the logarithm of cell concentrations at cell concentrations from 50 to 1000 cells mL–1. (D) Ionic currents at +1.0 V after the capture of different cells. (E) Ionic currents at +1.0 V vs enzyme incubation time. (F) Fluorescence microscopic images of the captured CCRF-CEM cells on the DAN/AAO hybrid, stained by PI/AO. The inset is the fluorescence microscope image of the DAN/AAO hybrid after cells were released. (G) Detection of CTCs in different samples. Results are obtained from PBS buffer and human blood samples, respectively.

To evaluate the selectivity of this present platform, k562 cells and Raji cells were used as the other two types of cell models. With the same capture and detection processes as the CCRF-CEM cells, the electrical current signals were obtained. As shown in Figure 5D, only CCRF-CEM cells show a sharp drop of current, while k562 cells and Raji cells show slight changes. The reason is that only CCRF-CEM cell membranes possess a type of transmembrane receptor, which can recognize with the aptamer sgc8c,44 thereby giving rise to the efficient and selective CCRF-CEM cell capture on the DAN/AAO surface. The captured cells cover the DAN/AAO hybrid surface, which efficiently blocks the ions transport through the hybrid, leading to varied I–V properties. The result confirms the excellent selectivity of the present method for cell capture and detection.

Finally, the successful release of CTCs from the DAN/AAO hybrid after cell capture is investigated. Benzonase nuclease was used to induce CCRF-CEM cell release.26 To obtain the proper cell release time, the cells captured on the DAN/AAO hybrid were treated by enzyme solution (25 units mL–1 where one unit is the enzyme amount that catalyzes the reaction of 1 nmol of substrate per minute) for 5, 10, 20, 30, 40, and 50 min, respectively. As shown in Figure 5E, with the enzyme incubation time from 0 to 40 min, the current signals first increase and then reach a plateau, which indicates that the captured CCRF-CEM cells have been nearly released completely. Moreover, the cell viability changes during the course of cell capture and release were investigated by the Acridine Orange/Propidium Iodide (AO/PI) assay. As shown in Figure 5F, almost 100% of the cells are alive after being captured on the DAN/AAO hybrid. After release, there are no cells on the DAN/AAO hybrid (inset in Figure 5F). The released CCRF-CEM cells from the DAN/AAO hybrid were collected, and their viability was investigated. The fluorescent images in Figure S11 demonstrate that the benzonase nuclease enzyme shows slight damage to the cells. To investigate the performance of the present platform in a complex real blood sample, CCRF-CEM cells were added in the prepared volunteer blood solution at concentrations of 300, 500, 700, and 1000 cells mL–1. The results are displayed in Figure 5G. A trend similar to that of the PBS solution is observed, verifying the practical application of the CCRF-CEM cell sensor in complex blood samples. The fact that most of the cells captured on DAN/AAO stay alive provide means for multimodal study, for example, by SERS and electrochemistry.

The durability of the prepared hybrid device is a critical factor toward practical bioanalysis. We have repeatedly used one hybrid for CTC capture and detection after cell release. The very similar I–V curves and ICR properties (Figure S12) in a continuous 5 day test demonstrate the excellent stability and reusability of the present hybrid toward CTCs capture, detection, and release.

In conclusion, we have demonstrated an engineered heterogeneous hybrid by chemically assembling DAN onto the top of AAO nanochannels. Due to the high surface-to-volume ratio of the AuNPs and high density of binding sites of PAMAM, the network formed has more surface area and can expose more functional groups, which results in more asymmetries in the structure and surface charge density of the DAN/AAO hybrid, thus leading to obvious rectified ion transport. The presence of amino groups on the DAN/AAO hybrid surface enables the anion-selective hybrid membrane to process ionic rectification and also promotes anion selectivity. Moreover, the ionic rectification properties can be tuned by ion valence and pH of the bulk solution. Under optimal conditions, highly sensitive circulating tumor cell detection based on this heterogeneous structure was successfully achieved with a detection limit of 80 cells mL–1. This device also provides a sensitive multimodal (chemical, electrical, and optical) platform to study cell adhesion and cell behavior in situ.

Acknowledgments

This work was supported by the National Natural Science Foundation of China (21874155, 21635004, and 21575163), the National Key Research and Development Program of China (2017YFA0206500), the Natural Science Foundation of Jiangsu Province (BK20191316), and the Double First-Class University project (CPU2018GY25).

Glossary

Abbreviations

DAN

dendrimer-Au nanoparticle network

AAO

anodic aluminum oxide

ICR

ionic current rectification

CTCs

circulating tumor cells

AuNPs

gold nanoparticles

PAMAM

dendrimer poly(amidoamine)

CS2

carbon disulfide

APTES

(3-aminopropyl) triethoxysilane

DMAP

4-dimethylaminopyridine

f

ICR ratio

fmax

local maximum ICR ratio

Supporting Information Available

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acs.nanolett.9b05066.

  • Materials and reagents, instrumentation, and experimental methods; contact angle of the DAN/AAO hybrid; SEM image of the DAN/AAO hybrid and corresponding EDS elemental mappings; TEM image of the DAN film; I–V curve of the DAN/AAO hybrid with varied AuNPs concentrations; effect of the channel size and thickness of AAO; effect of temperature; XPS spectra of the thiol sgc8c-modified DAN/AAO hybrid; I–V properties of the sgc8c-modified DAN/AAO hybrid; fluorescence images of the captured cells on the hybrid substrate; I–V properties of the DAN/AAO hybrid with different incubation times; the current value at +1.0 V with different incubation times; fluorescence microscope images of cells with different incubation times; fluorescence microscope images of CCRF-CEM cells after releasing; the I–V curves before and after CTCs release; the ionic current values at +1.0 V after varied cycles; comparison of different analytical methods for CCRF-CEM detection (PDF)

Author Contributions

J.L., X.H.X., and C.W. designed the experiments. C.W., X.P.Z., F.F.L., and Y. M. C. performed the experiments and wrote the paper. J.L. and X.H.X. revised the paper. All authors contributed to discussions.

The authors declare no competing financial interest.

Supplementary Material

nl9b05066_si_001.pdf (1.1MB, pdf)

References

  1. Smith P. L.; Baukrowitz T.; Yellen G. The inward rectification mechanism of the HERG cardiac potassium channel. Nature 1996, 379 (6568), 833–836. 10.1038/379833a0. [DOI] [PubMed] [Google Scholar]
  2. Siwy Z. S.; Howorka S. Engineered voltage-responsive nanopores. Chem. Soc. Rev. 2010, 39 (3), 1115–1132. 10.1039/B909105J. [DOI] [PubMed] [Google Scholar]
  3. Lan W. J.; Edwards M. A.; Luo L.; Perera R. T.; Wu X.; Martin C. R.; White H. S. Voltage-rectified current and fluid flow in conical nanopores. Acc. Chem. Res. 2016, 49 (11), 2605–2613. 10.1021/acs.accounts.6b00395. [DOI] [PubMed] [Google Scholar]
  4. Jiang Y. N.; Feng Y. P.; Su J. J.; Nie J. X.; Cao L. X.; Mao L. Q.; Jiang L.; Guo W. On the origin of ionic rectification in DNA-stuffed nanopores: the breaking and retrieving symmetry. J. Am. Chem. Soc. 2017, 139 (51), 18739–18746. 10.1021/jacs.7b11732. [DOI] [PubMed] [Google Scholar]
  5. Wang C.; Xu J. J.; Chen H. Y.; Xia X. H. Mass transport in nanofluidic devices. Sci. China: Chem. 2012, 55 (4), 453–468. 10.1007/s11426-012-4542-9. [DOI] [Google Scholar]
  6. Zhang Z.; Wen L.; Jiang L. Bioinspired smart asymmetric nanochannel membranes. Chem. Soc. Rev. 2018, 47 (2), 322–356. 10.1039/C7CS00688H. [DOI] [PubMed] [Google Scholar]
  7. Powell M. R.; Cleary L.; Davenport M.; Shea K. J.; Siwy Z. S. Electric-field induced wetting and dewetting in single hydrophobic nanopores. Nat. Nanotechnol. 2011, 6 (12), 798–802. 10.1038/nnano.2011.189. [DOI] [PubMed] [Google Scholar]
  8. Sun Y.; Chen S.; Chen X.; Xu Y.; Zhang S.; Ouyang Q.; Yang G.; Li H. A Highly selective and recyclable NO-responsive nanochannel based on a spiroring opening- closing reaction strategy. Nat. Commun. 2019, 10 (1), 1323. 10.1038/s41467-019-09163-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Perez-Mitta G.; Peinetti A. S.; Cortez M. L.; Toimil-Molares M. E.; Trautmann C.; Azzaroni O. Highly sensitive biosensing with solid-state nanopores displaying enzymatically reconfigurable rectification properties. Nano Lett. 2018, 18, 3303–3310. 10.1021/acs.nanolett.8b01281. [DOI] [PubMed] [Google Scholar]
  10. Li C. Y.; Ma F. X.; Wu Z. Q.; Gao H. L.; Shao W. T.; Wang K.; Xia X. H. Solution-pH-modulated rectification of ionic current in highly ordered nanochannel arrays patterned with chemical functional groups at designed positions. Adv. Funct. Mater. 2013, 23 (31), 3836–3844. 10.1002/adfm.201300315. [DOI] [Google Scholar]
  11. Vlassiouk I.; Siwy Z. S. Nanofluidic diode. Nano Lett. 2007, 7 (3), 552–556. 10.1021/nl062924b. [DOI] [PubMed] [Google Scholar]
  12. Zhang Z.; Li P.; Kong X.-Y.; Xie G.; Qian Y.; Wang Z.; Tian Y.; Wen L.; Jiang L. Bioinspired heterogeneous ion pump membranes: unidirectional selective pumping and controllable gating properties stemming from asymmetric ionic group distribution. J. Am. Chem. Soc. 2018, 140 (3), 1083–1090. 10.1021/jacs.7b11472. [DOI] [PubMed] [Google Scholar]
  13. Zhang Z.; Sui X.; Li P.; Xie G.; Kong X. Y.; Xiao K.; Gao L.; Wen L.; Jiang L. Ultrathin and ion-selective Janus membranes for high-performance osmotic energy conversion. J. Am. Chem. Soc. 2017, 139 (26), 8905–8914. 10.1021/jacs.7b02794. [DOI] [PubMed] [Google Scholar]
  14. Cheng L. J.; Guo L. J. Ionic current rectification, breakdown, and switching in heterogeneous oxide nanofluidic devices. ACS Nano 2009, 3 (3), 575–584. 10.1021/nn8007542. [DOI] [PubMed] [Google Scholar]
  15. Wang M.; Meng H. Q.; Wang D.; Yin Y. J.; Stroeve P.; Zhang Y. M.; Sheng Z. Z.; Chen B. Y.; Zhan K.; Hou X. Dynamic curvature nanochannel-based membrane with anomalous ionic transport behaviors and reversible rectification switch. Adv. Mater. 2019, 31 (11), 1805130. 10.1002/adma.201805130. [DOI] [PubMed] [Google Scholar]
  16. Zhang Z.; Kong X.-Y.; Xiao K.; Liu Q.; Xie G.; Li P.; Ma J.; Tian Y.; Wen L.; Jiang L. Engineered asymmetric heterogeneous membrane: a concentration-gradient-driven energy harvesting device. J. Am. Chem. Soc. 2015, 137 (46), 14765–14772. 10.1021/jacs.5b09918. [DOI] [PubMed] [Google Scholar]
  17. Coglitore D.; Merenda A.; Giamblanco N.; Dumee L. F.; Janot J.-M.; Balme S. Metal alloy solid-state nanopores for single nanoparticle detection. Phys. Chem. Chem. Phys. 2018, 20 (18), 12799–12807. 10.1039/C8CP01787E. [DOI] [PubMed] [Google Scholar]
  18. Balme S.; Lepoitevin M.; Dumee L. F.; Bechelany M.; Janot J.-M. Diffusion dynamics of latex nanoparticles coated with ssDNA across a single nanopore. Soft Matter 2017, 13 (2), 496–502. 10.1039/C6SM02461K. [DOI] [PubMed] [Google Scholar]
  19. Yanagi I.; Akahori R.; Takeda K. Stable fabrication of a large nanopore by controlled dielectric breakdown in a high-pH solution for the detection of various-sized molecules. Sci. Rep. 2019, 9, 13143. 10.1038/s41598-019-49622-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Plett T. S.; Cai W. J.; Thai M. L.; Vlassiouk I. V.; Penner R. M.; Siwy Z. S. Solid-state ionic diodes demonstrated in conical nanopores. J. Phys. Chem. C 2017, 121 (11), 6170–6176. 10.1021/acs.jpcc.7b00258. [DOI] [Google Scholar]
  21. Maglia G.; Heron A. J.; Hwang W. L.; Holden M. A.; Mikhailova E.; Li Q. H.; Cheley S.; Bayley H. Droplet networks with incorporated protein diodes show collective properties. Nat. Nanotechnol. 2009, 4 (7), 437–440. 10.1038/nnano.2009.121. [DOI] [PubMed] [Google Scholar]
  22. Bolognesi G.; Friddin M. S.; Salehi-Reyhani A.; Barlow N. E.; Brooks N. J.; Ces O.; Elani Y. Sculpting and fusing biomimetic vesicle networks using optical tweezers. Nat. Commun. 2018, 9 (1), 1882. 10.1038/s41467-018-04282-w. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Ouyang W.; Han J.; Wang W. Nanofluidic Crystals: Nanofluidics in a close-packed nanoparticle array. Lab Chip 2017, 17 (18), 3006–3025. 10.1039/C7LC00588A. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Liu H.; Jiang Q.; Pang J.; Jiang Z.; Cao J.; Ji L.; Xia X.; Wang K. A multiparameter pH-sensitive nanodevice based on plasmonicnanopores. Adv. Funct. Mater. 2018, 28 (1), 1703847. 10.1002/adfm.201703847. [DOI] [Google Scholar]
  25. Yue W. Q.; Tan Z.; Li X. P.; Liu F. F.; Wang C. Micro/nanofluidic technologies for efficient isolation and detection of circulating tumor cells. TrAC, Trends Anal. Chem. 2019, 117, 101–115. 10.1016/j.trac.2019.06.009. [DOI] [Google Scholar]
  26. Cao J.; Zhao X. P.; Younis M. R.; Li Z. Q.; Xia X. H.; Wang C. Ultrasensitive capture, detection, and release of circulating tumor cells using a nanochannel-ionchannel hybrid coupled with electrochemical detection technique. Anal. Chem. 2017, 89 (20), 10957–10964. 10.1021/acs.analchem.7b02765. [DOI] [PubMed] [Google Scholar]
  27. Zong C.; Xu M. X.; Xu L. J.; Wei T.; Ma X.; Zheng X. S.; Hu R.; Ren B. Surface-enhanced raman spectroscopy for bioanalysis: reliability and challenge. Chem. Rev. 2018, 118 (10), 4946–4980. 10.1021/acs.chemrev.7b00668. [DOI] [PubMed] [Google Scholar]
  28. Wang C.; Nie X. G.; Shi Y.; Zhou Y.; Xu J. J.; Xia X. H.; Chen H. Y. Direct plasmon-accelerated electrochemical reaction on gold nanoparticles. ACS Nano 2017, 11 (6), 5897–5905. 10.1021/acsnano.7b01637. [DOI] [PubMed] [Google Scholar]
  29. Yuan J. H.; He F. Y.; Sun D. C.; Xia X. H. A Simple method for preparation of through-hole porous anodic alumina membrane. Chem. Mater. 2004, 16 (10), 1841–1844. 10.1021/cm049971u. [DOI] [Google Scholar]
  30. Gittins D. I.; Caruso F. Spontaneous phase transfer of nanoparticulate metals from organic to aqueous media. Angew. Chem., Int. Ed. 2001, 40 (16), 3001–3004. . [DOI] [PubMed] [Google Scholar]
  31. Gong Y.; Yan S. F.; Xia P. F.; Wang R. F.; Zhang K. X.; Cui L.; Yin J. B. Porous microspheres based on hydroxyapatite-graft-poly (c-benzyl-L-glutamate) with improving homogeneity of hydroxyapatite and osteogenesis. Mater. Lett. 2019, 250, 206–209. 10.1016/j.matlet.2019.05.015. [DOI] [Google Scholar]
  32. Park M. H.; Ofir Y.; Samanta B.; Rotello V. M. Robust and Responsive dendrimer-gold nanoparticle nanocomposites via dithiocarbamate crosslinking. Adv. Mater. 2009, 21 (22), 2323–2327. 10.1002/adma.200803368. [DOI] [Google Scholar]
  33. Wang C.; Liu F. F.; Tan Z.; Chen Y. M.; Hu W. C.; Xia X. H. Fabrication of bio-inspired 2D MOFs/PAA hybrid membrane for asymmetric ion transport. Adv. Funct. Mater. 2019, 1908804. 10.1002/adfm.201908804. [DOI] [Google Scholar]
  34. Ilaiyaraja P.; Deb A. K. S.; Ponraju D. Removal of uranium and thorium from aqueous solution by ultrafiltration (UF) and PAMAM dendrimer assisted ultrafiltration (DAUF). J. Radioanal. Nucl. Chem. 2015, 303 (1), 441–450. 10.1007/s10967-014-3462-x. [DOI] [Google Scholar]
  35. Yang Z.; Shi D.; Zhang X.; Liu H.; Chen M.; Liu S. Co-deposition motif for constructing inverse opal photonic crystals with pH sensing. RSC Adv. 2015, 5 (85), 69263–69267. 10.1039/C5RA08046K. [DOI] [Google Scholar]
  36. Zhang Z.; Xie G. H.; Xiao K.; Kong X. Y.; Li P.; Tian Y.; Wen L. P.; Jiang L. Asymmetric multifunctional heterogeneous membranes for pH- and temperature-cooperative smart ion transport modulation. Adv. Mater. 2016, 28 (43), 9613–9619. 10.1002/adma.201602758. [DOI] [PubMed] [Google Scholar]
  37. Hsu J. P.; Chen Y. M.; Yang S. T.; Lin C. Y.; Tseng S. Influence of salt valence on the rectification behavior of nanochannels. J. Colloid Interface Sci. 2018, 531, 483–492. 10.1016/j.jcis.2018.07.012. [DOI] [PubMed] [Google Scholar]
  38. Daiguji H. Ion transport in nanofluidic channels. Chem. Soc. Rev. 2010, 39 (3), 901–911. 10.1039/B820556F. [DOI] [PubMed] [Google Scholar]
  39. Rao S.; Si K. J.; Yap L. W.; Xiang Y.; Cheng W. Free-standing bilayered nanoparticle superlatticenanosheets with asymmetric ionic transport behaviors. ACS Nano 2015, 9 (11), 11218–11224. 10.1021/acsnano.5b04784. [DOI] [PubMed] [Google Scholar]
  40. Pérez-Mitta G.; Albesa A. G.; ToimilMolares M. E.; Trautmann C.; Azzaroni O. The influence of divalent anions on the rectification properties of nanofluidic diodes: insights from experiments and theoretical simulations. ChemPhysChem 2016, 17 (17), 2718–2725. 10.1002/cphc.201600370. [DOI] [PubMed] [Google Scholar]
  41. Wei J.; Du G.; Guo J.; Li Y.; Liu W.; Yao H.; Zhao J.; Wu R.; Chen H.; Ponomarov A. The rectification of mono-and bivalent ions in single conical nanopores. Nucl. Instrum. Methods Phys. Res., Sect. B 2017, 404, 219–223. 10.1016/j.nimb.2016.12.015. [DOI] [Google Scholar]
  42. Jani A. M. M.; Losic D.; Voelcker N. H. Nanoporous anodic aluminium oxide: Advances in surface engineering and emerging applications. Prog. Mater. Sci. 2013, 58 (5), 636–704. 10.1016/j.pmatsci.2013.01.002. [DOI] [Google Scholar]
  43. Wang S. S.; Zhao X. P.; Liu F. F.; Younis M. R.; Xia X. H.; Wang C. Direct plasmon-enhanced electrochemistry for enabling ultrasensitive and label-free detection of circulating tumor cells in blood. Anal. Chem. 2019, 91, 4413–4420. 10.1021/acs.analchem.8b04908. [DOI] [PubMed] [Google Scholar]
  44. Pang X. H.; Cui C.; Su M. H.; Wang Y. G.; Wei Q.; Tan W. H. Construction of self-powered cytosensing device based on ZnOnanodisks@ g-C3N4 quantum dots and application in the detection of CCRF-CEM cells. Nano Energy 2018, 46, 101–109. 10.1016/j.nanoen.2018.01.018. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

nl9b05066_si_001.pdf (1.1MB, pdf)

Articles from Nano Letters are provided here courtesy of American Chemical Society

RESOURCES