Stacked-overlapped graphdiyne nano-iontronics enabling enhanced monovalent/divalent cation selectivity for single-cell pH detection

TL;DR

A new nano-iontronic device using graphdiyne enables precise, real-time pH measurements at the single-cell level with high cation selectivity.

Contribution

First single-cell pH sensing with high monovalent/divalent cation selectivity using stacked graphdiyne's sub-nanometer pores and tunable chemistry.

Findings

so-GDY-based pH sensor shows linear ionic current decrease with pH reduction due to protonation of surface functional groups.

Device exhibits 5x faster transport of monovalent cations compared to divalent cations, with resistance to interference.

Sensor maintains stability, repeatability, and biocompatibility for single-cell and single-organelle pH detection.

Abstract

Developing nano-iontronic devices that minimize ionic interference is essential for precise measurements in complex physiological systems. Graphdiyne (GDY), a novel carbon allotrope featuring sub-nanometer pores, enables effective regulation of ionic transport and is therefore a promising material for high-performance iontronic applications. Here, we report a pH-responsive nano-iontronic device fabricated by stacking and overlapping graphdiyne (so-GDY) layers onto the tip of the nanopipette. This so-GDY-based pH nano-iontronic sensor exhibits a linear decrease in ionic current under negative potential as the pH decreases from 8.00 to 5.50. This response is attributed to protonation of the oxygen-containing functional groups on the so-GDY surface and edges, which diminishes the negative surface charge and thereby reduces ionic conductivity. A key advantage of this nano-iontronic device is its excellent selectivity, demonstrating robust resistance to interference from divalent cations (Mg2+, Ca2+) and small molecules within the pH range of 8.00–5.50, while maintaining stable detection currents. The so-GDY-based pH nano-iontronic device transports monovalent cations up to 5 times more rapidly than divalent cations, alongside excellent repeatability, reversibility, and stability. This combination of features yields a biocompatible, high-resolution tool for minimally invasive, real-time pH measurements at the single-cell and even at a single-organelle level, opening new avenues for investigating cellular dynamics and disease pathogenesis with enhanced clarity.

We have for the first time reported a single-cell pH iontronic sensing with high monovalent/divalent cation selectivity by using the sub-nanometer pores and tunable surface chemistry of stacked graphdiyne.

Figures

Fabrication and characterization of the so-GDY-based pH nano-iontronic device. (A) Schematic illustration of the in-situ fabrication procedure for the so-GDY-based pH nano-iontronic device; the right panel depicts the chemical structure of GDY and its sub-nanometer pore size. (B) Top-view SEM image of the so-GDY formed at the nanopipette orifice. (C) Side-view SEM image of the so-GDY formed at the nanopipette orifice. (D) Side-view SEM image of the so-GDY structure after axial cleavage by FIB. (E) XPS spectrum of the so-GDY powder synthesized in bulk with the same method as so-GDY at the tip. (F) FT-IR spectrum of the so-GDY powder synthesized in bulk with the same method as so-GDY at the tip. (G) Raman spectrum of the so-GDY powder synthesized in bulk with the same method as so-GDY at the tip of the nanopipette. Scale bars (B–D): 300 nm.

The performance of the so-GDY-based pH nano-iontronic device. (A) I-V curves recorded at different pH levels within the physiological range. Inset: pH calibration curve derived by plotting the current at −1 V obtained from I-V curves as a function of pH. The electrolyte was 100 mM KCl with 10 mM HEPES buffer (pH 5.00–8.62). All devices require calibration of the relationship between pH and current before testing. (B) The resistance to ionic interference for the so-GDY-based pH nano-iontronic device measured in 100 mM KCl with 10 mM HEPES buffer (pH 7.2–7.4) containing different metal cations (0.2 mM), CuCl2 (2 µM). (C) Normalized conductance gi for different cations. (D) The selectivity ratio Si (defined in Eq. 1) for different cations. The solution concentration of the light purple and light green areas in (C) and (D) is 10 mM and 5 mM, respectively (pH ∼6.50). Error bars in (C) and (D) represent the standard deviation SD (n = 3, n is derived from different experimental units) and data are presented as mean values ± SD.

Mechanism of pH sensitivity and monovalent/divalent selectivity in the so-GDY-based pH nano-iontronic device. (A) Schematic illustration of the structure of so-GDY and the corresponding ionic transport behavior within the nanochannels (upper panel; the three colors in the top view represent three stacked layers of so-GDY) and the protonation/deprotonation process (lower panel). (B) The zeta potential of the so-GDY powder measured in 100 mM KCl solution containing 10 mM HEPES buffer at various pH. (C) The hydration energy of different cations. Inset: schematic illustration of the dehydration of monovalent and divalent cations.

Intracellular single-cell pH measurements. (A) Schematic illustration of the pH detection setup in a single cell. (B) Histogram of pHi for PC12 cells (n = 18), SH-SY5Y (n = 16), and PC12 mitochondria (n = 8). pHi was calculated from the current obtained at −1 V via linear sweep voltammetry (LSV) using the pre-calibration curve. Error bars represent standard deviation. Statistical comparisons were made using the Mann–Whitney rank-sum test: *p < 0.001 (pHi of PC12 cells vs SH-SY5Y cells), p < 0.0001 (pHi of PC12 cells vs pHM of PC12 cells). (C) Real-time pHi measurement in a single PC12 cell before, during, and after addition of cariporide to the extracellular medium using the so-GDY-based nano-iontronic device. Traces represent the average of three independent measurements; pHi was calculated from the current at −1 V extracted from the LSV. (D) Schematic illustration of the proposed mechanism by which cariporide modulates pHi in cells.

Equations

$$\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{upgreek} \usepackage{mathrsfs} \setlength{\oddsidemargin}{-69pt} \begin{document} \begin{eqnarray*} {{{S}}}_{{i}}\,{\mathrm{ = }}\,\frac{{{{{g}}}_{{i}}}}{{{{{g}}}_{{{\mathrm{K}}}^{\mathrm{ + }}}}}\,{\mathrm{ = }}\,\frac{{{{{c}}}_{{i}}{\mathrm{\ }} \cdot {\mathrm{|}}{{{Z}}}_{{i}}{\mathrm{|}}}}{{{{{c}}}_{{{\mathrm{K}}}^{\mathrm{ + }}} \cdot {{{Z}}}_{{{\mathrm{K}}}^{\mathrm{ + }}}}}, \end{eqnarray*}\end{document}$$

Full text

INTRODUCTION

Iontronics is an interdisciplinary field merging electronics and ionics [1], integrating aspects of electrochemistry [2], materials science [3], and biological sciences [4]. Unlike conventional electronic systems, iontronic devices use mobile ions as the primary signal mediators, which can markedly alter their circuit current output characteristics [5]. A major branch of iontronics is iontronic sensing [6,7], which offers distinct advantages over traditional sensing strategies, including excellent biocompatibility, enhanced sensitivity, ultralow detection limits, and label-free operation capabilities [8–10]. Nanofabrication techniques enable the precise engineering of nanopores and nanochannels for fabricating nano-iontronic devices. These engineered nanostructures provide effective means to regulate ion transport under nanoconfinement, thereby improving sensing performance and expanding the range of potential applications [11]. For example, Wang et al. demonstrated the genetic fusion of supercharged unstructured polypeptides (SUPs) with proteins of interest, using them as molecular carriers to enable more precise protein detection and detailed analysis of protein–protein interactions [12]. Although nano-iontronic devices offer numerous unique advantages, key challenges remain, particularly in minimizing the interference from both small and large ionic species in complex biological environments and improving the recovery time for reversible responses in real-time detection using current-based iontronic sensors [13]. These limitations hinder the accurate and continuous monitoring of dynamic and complex biological environments. Incorporating ion-interference-resistant materials with uniform sub-nanometer pores into solid-state nanopores may enhance the performance of iontronic devices by excluding interfering species while improving reversibility.

Graphdiyne (GDY) is an emerging two-dimensional carbon allotrope with a layered, porous structure composed of sp- and sp^2^-hybridized carbon atoms [14,15]. The incorporation of sp-hybridized carbons, in particular, impacts distinct physical and chemical properties compared to other carbon materials [16]. The extensive π-conjugated network, formed by benzene rings and acetylenic linkages, affords GDY uniformly distributed sub-nanometer-sized pores (5.5 Å) [17], a large surface area, wide interplanar spacing (3.6 Å), and heterogeneous charge distribution [18]. These attributes make GDY attractive for molecularly selective membranes [19], gas separation [17], and energy-related fields [20,21]. Furthermore, unique surface chemical properties endow GDY with stronger affinity to water than other carbon materials [22]. Upon exposure to air, the terminal alkyne groups of GDY readily oxidize, leading to the introduction of oxygen-containing functional groups [18]. This functionalization further enhances the water affinity of GDY and promotes proton conductivity [22,23]. Collectively, the well-defined porous structure and tunable surface properties highlight the strong potential of GDY for regulating ion transport and fabricating advanced nano-iontronic devices.

Herein, to exploit the advantages of GDY and address limitations associated with nanoconfined iontronic devices, we develop a pH-responsive nano-iontronic device by assembling stacked and overlapped GDY (so-GDY) layers within the pore at the tip of a glass nanopipette. The inter-stacked structure, well-defined composition, and uniform sub-nanometer pores of so-GDY endow the pH nano-iontronic device with exceptional selectivity, enabling it to accurately distinguish between various pH levels while effectively rejecting interference from divalent cations like Mg^2+^ and Ca^2+^, as well as larger molecules. The design of the pH nano-iontronic device enables real-time, high-resolution monitoring of dynamic pH fluctuations within individual cells and intracellular organelles. By leveraging the structural and surface-chemical advantages of so-GDY, the proposed platform provides a practical route for probing cellular processes and may support the development of targeted therapeutic strategies.

RESULTS AND DISCUSSION

Fabrication and characterization of the so-GDY-based pH nano-iontronic device

The schematic diagram of the in-situ fabrication of the so-GDY-based pH nano-iontronic device is illustrated in Fig. 1A. Briefly, the nanopipette was filled with CuCl solution and then immersed in a suspension of hexakis-[(trimethylsilyl)ethynyl]benzene (HEB-TMS). This specific distribution of catalyst and HEB-TMS monomer promoted the growth of GDY at the nanopipette orifice (inner diameter 300 ± 50 nm, Fig. S1A), forming a compact and overlapping GDY structure that fully sealed the orifice, termed so-GDY. The morphology of the so-GDY was characterized by scanning electron microscopy (SEM), with top-view (Fig. 1B and Fig. S1B) and side-view (Fig. 1C) images revealing a robust, crack-free structure covering the nanopipette orifice. Additionally, side-view SEM images obtained via axial cleavage using a focused ion beam (FIB) (Fig. 1D and Fig. S1C) further confirmed that the so-GDY fabricated at the tip of the nanopipette formed a complete and continuous structure.

Fabrication and characterization of the so-GDY-based pH nano-iontronic device. (A) Schematic illustration of the in-situ fabrication procedure for the so-GDY-based pH nano-iontronic device; the right panel depicts the chemical structure of GDY and its sub-nanometer pore size. (B) Top-view SEM image of the so-GDY formed at the nanopipette orifice. (C) Side-view SEM image of the so-GDY formed at the nanopipette orifice. (D) Side-view SEM image of the so-GDY structure after axial cleavage by FIB. (E) XPS spectrum of the so-GDY powder synthesized in bulk with the same method as so-GDY at the tip. (F) FT-IR spectrum of the so-GDY powder synthesized in bulk with the same method as so-GDY at the tip. (G) Raman spectrum of the so-GDY powder synthesized in bulk with the same method as so-GDY at the tip of the nanopipette. Scale bars (B–D): 300 nm.

X-ray photoelectron spectroscopy (XPS) results show that the so-GDY is primarily composed of carbon and oxygen, with an O/C ratio of 22.8% (Fig. 1E). Compared with previously reported GDY structures [14,18], this higher O/C ratio indicates an increased content of oxygen-containing functional groups in so-GDY, which can be attributed to the synthetic strategy used in this work. To further characterize the composition of so-GDY, Fourier transform-infrared spectroscopy (FT-IR) and energy-dispersive X-ray spectroscopy coupled with scanning electron microscopy (SEM/EDS) were performed. The FT-IR spectrum (Fig. 1F) shows the band at 1708 cm^−1^ assigned to the stretching vibration of C=O, and the band at 2104 cm^−1^ attributed to the C≡C stretching vibration [22,23]. These FT-IR results demonstrate that the so-GDY contains abundant oxygen-containing functional groups, consistent with the XPS findings. SEM/EDS analysis further confirms the presence of C and O at the nanopipette orifice (Fig. S2). The oxygen content is attributed to oxidation of terminal alkynes and oxygen adsorption at the tip of the nanopipette, corroborating the FT-IR results. Raman spectroscopy (Fig. 1G) of the so-GDY powder synthesized in bulk under the same conditions as those used for in-situ growth at the nanopipette orifice, reveals four prominent peaks at ∼1367 cm^−1^, 1587 cm^−1^, 1916 cm^−1^, and 2182 cm^−1^, which is consistent with previous reports [18]. The D band at 1367 cm^−1^ is associated with the vibration of the bonds connecting two carbon triple bonds and the stretching of the C−C bonds. The G band at 1587 cm^−1^ indicates that the so-GDY contains abundant aromatic rings. The peak at 2182 cm^−1^ is attributed to conjugated diyne links (–C≡C–C≡C–) [22]. Collectively, these results demonstrate that the so-GDY retains the characteristic skeleton of GDY while incorporating oxygen-containing functional groups at its edges and defect sites.

The successful fabrication of so-GDY was further confirmed by comparing the current-voltage (I-V) curves of the so-GDY-covered nanopipette to those of a bare nanopipette. The experimental setup for ionic current measurements is shown in Fig. S3A. The I-V curves of both the bare and so-GDY-coated nanopipettes immersed in 100 mM KCl aqueous solution are presented in Fig. S3B. In contrast to the slightly negatively rectified I-V response of the bare nanopipette, the so-GDY-covered nanopipette exhibits pronounced ionic current rectification (ICR), with a much larger current at negative potential compared to positive potential. This negative ICR is attributed to the increased negative charge of the so-GDY, arising from the oxygen-containing species within the pores and at the terminal edges. These characteristics are consistent with the device’s cation selectivity and confirm the successful fabrication of the so-GDY-based pH nano-iontronic device.

The performance of the so-GDY-based pH nano-iontronic device

We investigated the influence of protons on the ionic transport in the so-GDY-based pH nano-iontronic device by recording I-V curves in 100 mM KCl and 10 mM 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES) buffer over a range of pH (Fig. 2A). The I-V curves measured across the common physiological pH range from 8.00 to 5.50 [24,25] exhibit stable responses at each pH, with the current at negative potential (−1 V) decreasing linearly as pH decreases. In contrast, the current in the positive potential range (0 to +1 V) remains nearly unchanged. These findings indicate that protons primarily modulate the ionic current under the negative potential range, with minimal influence under positive potential. As shown in the inset of Fig. 2A, the current response of the so-GDY-based pH nano-iontronic device exhibits a strong linear correlation with the pH (I (nA) = −21.7 pH + 90.6, R^2^ = 0.98), confirming the high sensitivity for pH detection within the physiological pH range. We further evaluated the reversibility and stability of the so-GDY-based pH nano-iontronic device. As shown in Fig. S4, the sensor reversibly switched between the current levels corresponding to pH 7.20 and pH 5.50 over 15 cycles, with no observable signal degradation or drift. The absolute current difference (ΔI) between pH 7.20 and pH 5.50 remained essentially constant throughout the test, confirming the robustness of the device for repeated measurements.

The performance of the so-GDY-based pH nano-iontronic device. (A) I-V curves recorded at different pH levels within the physiological range. Inset: pH calibration curve derived by plotting the current at −1 V obtained from I-V curves as a function of pH. The electrolyte was 100 mM KCl with 10 mM HEPES buffer (pH 5.00–8.62). All devices require calibration of the relationship between pH and current before testing. (B) The resistance to ionic interference for the so-GDY-based pH nano-iontronic device measured in 100 mM KCl with 10 mM HEPES buffer (pH 7.2–7.4) containing different metal cations (0.2 mM), CuCl2 (2 µM). (C) Normalized conductance gi for different cations. (D) The selectivity ratio Si (defined in Eq. 1) for different cations. The solution concentration of the light purple and light green areas in (C) and (D) is 10 mM and 5 mM, respectively (pH ∼6.50). Error bars in (C) and (D) represent the standard deviation SD (n = 3, n is derived from different experimental units) and data are presented as mean values ± SD.

To investigate ionic interference on device performance, we examined the effects of K⁺, Na⁺, Ca^2+^, Mg^2+^, and Cu^2+^. As shown in Fig. 2B, the presence of these common cations, particularly divalent cations, has minimal influence on pH sensing performance. Notably, a slight pH change (∆pH = 0.4) produced a larger change in the output current than did the presence of these cations. This high selectivity indicates that the so-GDY-based pH nano-iontronic device provides accurate pH measurements in the presence of other cations, supporting its applicability to complex biological and environmental samples (Fig. S5).

We also evaluated the response of the so-GDY-based pH nano-iontronic device to other ions and larger molecules, which is essential for accurate sensing in complex biological media. First, the device selectivity for cations and anions was examined under 10-fold and 50-fold concentration gradients. As shown in Fig. S6, the device was filled with 10 mM KCl and immersed in bulk solutions of 100 mM or 500 mM KCl. Under both concentration gradients, the I-V curves exhibit significant negative rectification, indicating cation-dominated transport. Driven by the concentration gradient, K⁺ and Cl⁻ diffuse from the high-concentration bulk into the low-concentration device, generating a diffusion potential. Using this diffusion potential, we calculated the ionic transference number (t) under 10-fold and 50-fold concentration gradients, respectively (Table S1). These results show that, for the concentration gradients up to 50-fold, tK^+^ exceeds tCl^−^, confirming that cations are the major carriers in the device. To evaluate cation selectivity, we recorded I-V curves in various cation chloride solutions at pH 6.50 (Fig. S7). Divalent cations exhibited substantially lower ionic conductance than monovalent cations. To account for the impact of different ionic electrophoretic mobilities (Table S2) on the conductivity in bulk solution, we defined the cation selectivity ratio (Si) relative to K^+^ according to previous report [26]:

\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{upgreek} \usepackage{mathrsfs} \setlength{\oddsidemargin}{-69pt} \begin{document} \begin{eqnarray*} {{{S}}}_{{i}}\,{\mathrm{ = }}\,\frac{{{{{g}}}_{{i}}}}{{{{{g}}}_{{{\mathrm{K}}}^{\mathrm{ + }}}}}\,{\mathrm{ = }}\,\frac{{{{{c}}}_{{i}}{\mathrm{\ }} \cdot {\mathrm{|}}{{{Z}}}_{{i}}{\mathrm{|}}}}{{{{{c}}}_{{{\mathrm{K}}}^{\mathrm{ + }}} \cdot {{{Z}}}_{{{\mathrm{K}}}^{\mathrm{ + }}}}}, \end{eqnarray*}\end{document}

where, g_i_ is a normalized conductance for each cation ( \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{upgreek} \usepackage{mathrsfs} \setlength{\oddsidemargin}{-69pt} \begin{document} {g}{\rm i}{\mathrm{ = }}\frac{{{G}{\rm i}}}{{{\mu }{\rm i}/{\mu }{{{\rm K}}^ + }}}\end{document} ), \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{upgreek} \usepackage{mathrsfs} \setlength{\oddsidemargin}{-69pt} \begin{document} {g}{{{\rm K}}^ + }\end{document} and Gi are the measured conductance of K^+^ and cation i in the solution, respectively; \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{upgreek} \usepackage{mathrsfs} \setlength{\oddsidemargin}{-69pt} \begin{document} {\mu }{{{\rm K}}^ + }\end{document} and μ_i_ are the bulk electrophoretic mobility of the K^+^ and cation i, respectively; \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{upgreek} \usepackage{mathrsfs} \setlength{\oddsidemargin}{-69pt} \begin{document} {c}{\mathrm{K}^ + }\end{document} and ci are the solution concentration of K^+^ and cation i, respectively; \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{upgreek} \usepackage{mathrsfs} \setlength{\oddsidemargin}{-69pt} \begin{document} {Z}{\mathrm{K}^ + }\end{document} and Zi are the charge numbers of K^+^ and cation i, respectively. This definition of selectivity implies that S_i_ = 1 represents no distinction between the cation i and K^+^.

The normalized conductance of representative monovalent and divalent cations commonly present in biological environments is shown in Fig. 2C. Divalent cations exhibit substantially lower normalized conductance than monovalent cations. As shown in Fig. 2D, the selectivity among monovalent cations is negligible, while for divalent cations S_i_ < 0.4 even at the elevated ionic strength. According to Eq. (1), the concentration of divalent ions accessing the device is only 0.2-fold that of K^+^, indicating a minor contribution of divalent species to the measured current. Collectively, these results demonstrate that the so-GDY-based pH nano-iontronic device strongly discriminates between monovalent and divalent cations, while exhibiting only weak differentiation among monovalent species.

To evaluate the transport of larger molecules through the so-GDY-based pH nano-iontronic device, we used Rhodamine 6G (R6G), a fluorescent molecule with dimensions of 1.2 × 1.3 nm (Fig. S8A and B), as a model probe. Because R6G is positively charged, a bias of −1 V was applied to drive its electrophoretic migration from the bulk solution into the nanopipette. R6G readily permeated the orifice of the bare nanopipette (Fig. S8C–E), as indicated by green fluorescence. In contrast, R6G was effectively excluded by the so-GDY nanochannels (Fig. S8F–H). No fluorescence was observed within the so-GDY-based device. These results confirm the size-exclusion capability of the so-GDY structure and indicate strong resistance to the ingress of larger biomolecules, thereby mitigating fouling.

The mechanism of the so-GDY-based pH nano-iontronic device

A proposed mechanism for the so-GDY-based pH nano-iontronic device is shown in Fig. 3A. The so-GDY structure contains a high density of oxygen species both within its nanochannels and at the terminal edges of its stacked architecture. Unlike the GDY synthesized under oxygen-free conditions, our in-situ fabrication method in air introduces additional oxygen-containing functional groups, and the unique overlapped stacking of the GDY layers creates nanochannels predominantly lined by these edge oxygen-containing functional groups. This combination supplies abundant protonation/deprotonation sites within the channels, enabling pH responsiveness in the physiological range. Based on a previous report [23], carboxyl groups protonate and deprotonate between pH 5.50 and 8.00, so we focus on the interactions between the carboxyl groups and different cations. At a low pH, the carboxyl groups are protonated, reducing the negative surface charge of so-GDY and thereby decreasing the ionic current at negative potential. In contrast, at a high pH, the carboxyl groups deprotonate, increasing the negative surface charge and consequently resulting in a higher ionic current magnitude at negative potential. The zeta potential results of so-GDY (Fig. 3B) confirm this hypothesis. The zeta potential becomes more negative with increasing pH, consistent with deprotonation of carboxyl groups and the proposed mechanism.

Mechanism of pH sensitivity and monovalent/divalent selectivity in the so-GDY-based pH nano-iontronic device. (A) Schematic illustration of the structure of so-GDY and the corresponding ionic transport behavior within the nanochannels (upper panel; the three colors in the top view represent three stacked layers of so-GDY) and the protonation/deprotonation process (lower panel). (B) The zeta potential of the so-GDY powder measured in 100 mM KCl solution containing 10 mM HEPES buffer at various pH. (C) The hydration energy of different cations. Inset: schematic illustration of the dehydration of monovalent and divalent cations.

Monovalent/divalent cation selectivity of the so-GDY-based pH nano-iontronic device is primarily governed by (i) the size-sieving effect of the so-GDY nanochannels and (ii) the interaction between cations and carboxyl groups. Based on our aforementioned experimental results, the larger hydrated cations and molecules experience greater transport resistance as they pass through the device, and the sufficiently large species are effectively rejected. This size-based sieving effect explains the strong exclusion observed for divalent cations. We hypothesize that the observed blocking is attributed to their hydration radii and energy of the divalent cations. As shown in Fig. 3C and Table S2, the ions in aqueous solution exhibit increased effective radii because of the solvation effects. These hydrated ions require partial dehydration to pass through the confined nanochannels, which poses a higher energy barrier for divalent cations due to their larger hydration energies. As a result, the so-GDY structure effectively blocks the transport of divalent cations.

Another major determinant of ion selectivity is the interaction between the cations and oxygen-containing functional groups. The excessively strong interactions between cations and functional groups distributed on the nanochannels can hinder their transport [27,28]. The zeta potential measurements demonstrate that our so-GDY is negatively charged at the pH range of 5−8. As a result, the divalent cations (Ca^2+^ and Mg^2+^) exhibit stronger electrostatic interactions with carboxyl groups than the monovalent cations (Li^+^, Na⁺, and K⁺), thereby obstructing their transport. In addition, the I-V curves show that when using the 50 mM divalent cation solution, currents at both −1 V and +1 V increase, and the negative rectification disappears (Fig. S9A and B). These results indicate that a sufficient amount of adsorbed divalent cations changes the surface charge of so-GDY. In contrast, the I-V curves of the 100 mM monovalent cation solution show obvious negative rectification, and dramatic increase of the currents only between −1 V and 0 V (Fig. S9C and D). These results demonstrate that the divalent metal cations have a stronger binding affinity for carboxyl groups than the monovalent cations under the same cation strength. This stronger interaction contributes more significantly to preventing the transport of divalent ions.

In summary, the so-GDY-based pH nano-iontronic device effectively blocks both larger biomolecules and divalent cations (Ca^2+^ and Mg^2+^), while permitting the permeation of monovalent cations. The minor perturbations in ion concentration have a negligible impact on pH sensing at physiological concentrations (Fig. 2B), thereby ensuring accurate pH measurements.

Intracellular single-cell pH measurements

Before the intracellular pH measurements, we evaluated the potential interference from common physiological cations on the so-GDY-based pH nano-iontronic device. The typical intracellular cation concentrations and their variations are listed in Table S3. As shown in Figs S10 and S11, a 10 mM change in KCl concentration induces a 7.32% current shift, substantially smaller than the 30.0% signal change caused by a 0.8 pH change. Moreover, the simultaneous addition of 2 mM CaCl_2_ and MgCl_2_ at pH 7.99 and 5.02 produced a negligible change in the measured current. These results confirm that at this physiologically relevant concentration, the common cations (K^+^, Ca^2+^, and Mg^2+^) have negligible impact on the pH-sensitive signal, demonstrating the excellent resistance to divalent cation interference within the defined detection range.

Intracellular pH (pH_i_) is vital for cellular function, influencing apoptosis, enzyme activity, ion transport, and metabolism [29]. Cancer cells exhibit unique pH distributions with increased pH_i_ and decreased extracellular pH (pH_e_), promoting tumor growth, invasion, and dissemination [30,31]. Similarly, neurodegenerative diseases such as Alzheimer’s and Parkinson’s are also associated with abnormal pH, contributing to protein aggregation, mitochondrial dysfunction, and neuronal death [32,33]. Conventional techniques for measuring pH_i_, including nuclear magnetic resonance (NMR) [34], fluorescent microscopy [35], surface-enhanced Raman scattering (SERS) [36], and microelectrodes [37], suffer from limitations including insufficient sensitivity and susceptibility to background interference. To address these limitations, we employed the so-GDY-based pH nano-iontronic device as a biocompatible and high-resolution platform for continuous pH_i_ monitoring. This device minimizes cellular disruption and enables precise pH measurements, providing a useful tool for interrogating complex biological processes and disease mechanisms.

To confirm the compatibility with living cells, we evaluated cell viability prior to pH_i_ measurements using sequential Hoechst 33342 and propidium iodide (PI) staining. Hoechst 33342 can easily penetrate the cell membrane and the membranes of the internal cellular organelles, emitting blue fluorescence upon excitation. Dead cells usually show brighter fluorescence because of higher cell permeability [38,39]. As shown in Fig. S12, during several insertions and up to 60 min after withdrawal of the device, the fluorescence intensity of the cells shows minimal change. This result indicates that the cells remain viable and the device had a negligible impact on cellular integrity. PI, a dye that cannot permeate cell membranes, is crucial in verifying the secure seal between the cell membrane and the so-GDY-based pH nano-iontronic device, as well as assessing cell membrane integrity. PI is excluded from living cells but emits red fluorescence in dead cells. Therefore, any breaches or damage caused by the device insertion would result in red fluorescence within the cells [38,40]. In our work, during several insertions and up to 60 m after the withdrawal of the so-GDY-based pH nano-iontronic device, no red fluorescence was observed, indicating the cell membrane integrity remained intact (Fig. S13).

We then applied the so-GDY-based pH nano-iontronic device to quantify pH_i_ in PC12 and SH-SY5Y cell lines. A schematic of the experimental setup is presented in Fig. 4A. The statistical analysis of pH_i_ for PC12 cells, SH-SY5Y cells, and PC12 mitochondria is shown in Fig. 4B. The pH results were calculated based on the pre-calibration curve. PC12 cells exhibited a lower pH_i_ (6.96 ± 0.26) than SH-SY5Y (7.23 ± 0.19, p < 0.001). The measured pH_i_ is coincident with the result in a previous report [41]. Representative images of PC12 and SH-SY5Y cells during pH measurements are shown in Fig. S14. The relatively acidic intracellular environment in some cancer cells typically arises from increased glycolysis (lactate production) and tumor hypoxia, conferring advantages such as enhanced proliferation, invasion, and apoptosis resistance [42]. We further utilized the device to investigate the pH of PC12 mitochondria. To access mitochondria, PC12 cells were stained with rhodamine 123. Bright-field and fluorescent images of PC12 mitochondria during pH measurements are presented in Fig. S15. The measured mitochondrial pH (pH_M_) was 7.7 ± 0.3, significantly higher than the pH_i_ (p < 0.0001). The higher pH_M_ compared to the pH_i_ is primarily due to the active pumping of protons across the inner mitochondrial membrane during oxidative phosphorylation. This process is essential for ATP synthesis and creates a proton gradient that drives the ATP synthase. As a result, the mitochondrial matrix maintains a more alkaline pH compared to the pH_i_ [43,44]. In conclusion, the so-GDY-based pH nano-iontronic device offers a precise and non-invasive method for measuring intracellular and organellar pH, providing valuable insights into complex cellular processes.

*Intracellular single-cell pH measurements. (A) Schematic illustration of the pH detection setup in a single cell. (B) Histogram of pHi for PC12 cells (n = 18), SH-SY5Y (n = 16), and PC12 mitochondria (n = 8). pHi was calculated from the current obtained at −1 V via linear sweep voltammetry (LSV) using the pre-calibration curve. Error bars represent standard deviation. Statistical comparisons were made using the Mann–Whitney rank-sum test: ***p < 0.001 (pHi of PC12 cells vs SH-SY5Y cells), ***p < 0.0001 (pHi of PC12 cells vs pHM of PC12 cells). (C) Real-time pHi measurement in a single PC12 cell before, during, and after addition of cariporide to the extracellular medium using the so-GDY-based nano-iontronic device. Traces represent the average of three independent measurements; pHi was calculated from the current at −1 V extracted from the LSV. (D) Schematic illustration of the proposed mechanism by which cariporide modulates pHi in cells.

Building on the single-cell and single-organelle measurements above, we applied the so-GDY-based pH nano-iontronic device for real-time, in-situ monitoring of individual cancer cells during drug treatment. Cariporide, an inhibitor of the Na^+^/H^+^ exchanger NHE1, was used to evaluate drug-induced pH_i_ changes in PC12 cells. NHE1 is critical for maintaining intracellular pH in cancer cells [45], and its inhibition by cariporide can induce cellular acidification and apoptosis [46]. Real-time pH_i_ measurements using the so-GDY-based pH nano-iontronic device (Fig. 4C) revealed a rapid decline in pH_i_ from 7.04 to 5.53 upon cariporide administration (10 µM). This effect was consistent across multiple cells (average pH_i_: 7.06 ± 0.04 before, 5.9 ± 0.3 after cariporide). Subsequently, a gradual pH recovery was observed, likely due to membrane disruption caused by cell death exposing the device to the buffer. Cariporide toxicity in PC12 cells was conducted by confocal microscopy as shown in Fig. S16. A schematic of the proposed mechanism, including NHE1 inhibition and downstream cellular effects, is shown in Fig. 4D. These measurements are in close quantitative agreement with values reported for established functional nano-sensors (Table S4). To further confirm the results of the so-GDY-based pH nano-iontronic device, we measured the pH_i_ of PC12 cells using a pH-sensitive dye 2′,7′-bis-(2-carboxyethyl)-5-(and-6)-carboxyfluorescein, acetoxymethyl ester (BCECF-AM) and flow cytometry (Fig. S17). The results show that pH_i_ of normal PC12 cells was 7.16 ± 0.03 and confirmed a pH_i_ decrease upon cariporide stimulation, which is consistent with the typical response magnitude reported for NHE1 inhibition in various cell types [47–49]. Although all methods confirmed cariporide-induced acidification, absolute pH values differed because of methodological factors (see Supporting Information). Overall, these results establish the so-GDY-based pH nano-iontronic device as a powerful platform for real-time and in-situ monitoring of pH_i_ during drug treatment.

CONCLUSION

In conclusion, by leveraging the unique structural properties of GDY, we have, for the first time, successfully fabricated so-GDY at the tip of a nanopipette, to design a so-GDY-based pH nano-iontronic device. Using this device, we investigated the ion transport through the so-GDY confined at the nanopipette tip, and found that transport is strongly modulated by the proton concentration in the bulk solution. Specifically, the ionic current under negative potential decreases linearly as pH decreases from 8.00 to 5.50, which is attributed to the protonation of oxygen-containing functional groups on the so-GDY surface and edges, which reduces the negative surface charge thereby lowering the ionic current at −1 V. The so-GDY-based pH nano-iontronic device exhibits high monovalent/divalent cation selectivity, reversibility, and stability, offering a biocompatible, high-resolution platform for continuous pH monitoring. By minimizing cellular disruption while enabling precise, real-time pH measurements, this device offers a useful tool for studying complex biological processes and disease mechanisms.

MATERIALS AND METHODS

Fabrication of the so-GDY-based pH nano-iontronic device

The single-channel nanopipettes were fabricated using a CO_2_-laser-based pipette puller (P-2000, Sutter Instrument Co.) with glass capillaries (O.D.: 1.5 mm and I.D.: 1.1 mm) from the same company. A 10 μL solution of 12.1 mM CuCl in DMF was backfilled into the nanopipette. The backfilled nanopipette was then immersed in 12.8 mM HEB-TMS for 24 h at 60°C. This allowed the two solutions to meet at the pipette orifice, forming a compact structure of so-GDY at the tips of the glass nanopipettes. After the reaction, the nanopipettes were washed with DMF, ethanol, water, 1 M HCl and water sequentially to remove the unreacted HEB-TMS and CuO nanoparticles from the so-GDY structure. Finally, the so-GDY nanopipette was backfilled with appropriate electrolyte for further measurements.

Electrochemical measurements

A potentiostat (CHI 660E, CH Instruments) in a two-electrode mode was used to monitor the current flow. The so-GDY-based pH nano-iontronic device was backfilled and immersed in electrolyte solutions with two Ag/AgCl wires placed in the nanopipette and the external bath solution, respectively, to connect the circuit. The bias potentials shown in the main text were Vinterval vs. Vexternal. All devices required calibration of the relationship between pH and current before testing.

Cell culture

Both SH-SY5Y and PC12 cells were obtained from the National Infrastructure of Cell Line Resource (Beijing, China) and kept in a 100% humidified incubator at 37°C with 95% air and 5% CO₂. PC12 cells were cultured in RPMI 1640 medium supplemented with 10% horse serum (HS), 5% fetal bovine serum (FBS), and a 1% penicillin-streptomycin mixture. SH-SY5Y cells were maintained in Dulbecco's modified Eagle's medium (DMEM) with 10% FBS and a 1% penicillin-streptomycin mixture. The culture medium was refreshed every 2 days. To assess the effects of the so-GDY-based pH nano-iontronic device and cariporide on cells, we performed assays for cellular viability, mitochondrial membrane potential, and cellular apoptosis (detailed protocols are provided in the Supporting Information).

Single-cell experiment

The ionic current measurements were performed on an EPC-10 patch clamp system. The so-GDY-based pH nano-iontronic device was fixed to the holder and I-V was recorded from +1 V to −1 V against a Ag/AgCl reference electrode. Before experiments, the cells were washed three times using a mixture of 100 mM KCl, 10 mM HEPES, pH 7.2–7.4.

Supplementary Material

nwag050_Supplemental_File