Glutathione-Responsive Disulfiram Delivery from Thiol-Functionalized N-Doped Carbon Nanodots: An in vitro Proof-of-Concept Study Authors: Mohadese Panaei, Mohamad Mahani, Faten Divsar, Faeze Khakbaz Published: 2026 ABSTRACT PurposeTo develop and evaluate a glutathione (GSH)-responsive “Off–On” drug delivery system based on thiol-functionalized N-doped carbon dots (SNCDs) for controlled intracellular release of disulfiram (DSF).Materials and MethodsSNCDs were synthesized using a solvothermal technique and subsequently characterized using dynamic laser scattering, high-resolution electron imaging, UV–Vis spectrophotometry, and Fourier-transform infrared spectroscopy (FTIR). DSF was immobilized onto SNCDs through redox-cleavable disulfide linkages, and loading efficiency was determined by UV–Vis analysis. The in vitro drug release kinetics were examined under simulated physiological (pH 7.4) and tumor-relevant acidic (pH 5.4) conditions, evaluating the impact of glutathione (GSH) by performing experiments with and without its inclusion. Cytotoxicity and compatibility studies in MCF-7 breast tumor cells, evaluated via MTT test, were used to investigate SNCD biocompatibility and the anticancer performance of DSF-SNCDs.ResultsThe FTIR spectra confirmed effective thiol functionalization and revealed oxygenated and nitrogenous groups on the particle surface suitable for disulfide linkage formation. DSF loading efficiency reached approximately 75%. Under non-reducing conditions, DSF-SNCDs exhibited minimal premature release (≤30% over 24 h at pH 7.4 and 5.4), demonstrating an “Off” state. In contrast, the presence of 5% (w/v) GSH at pH 5.4 triggered rapid DSF release, achieving approximately 80% cumulative release within 24 h, corresponding to the “On” state through reductive disulfide bond cleavage. Bare SNCDs showed high biocompatibility (>85% cell viability at 50 µg/mL), whereas DSF-SNCDs produced significantly enhanced cytotoxicity under GSH conditions, confirming redox-activated drug release.ConclusionSNCDs provide an effective GSH-responsive nanoplatform for DSF delivery, combining high drug loading, excellent stability under non-reducing conditions, and efficient intracellular drug release under reductive environments. This strategy minimizes premature drug leakage while enhancing anticancer activity, demonstrating strong potential for cancer therapy. Introduction Globally, cancer remains a critical public health crisis, consistently ranking as a primary cause of mortality despite significant advancements in modern medicine.1 Traditional chemotherapy is frequently constrained by a lack of selectivity, hindering its broader clinical utility. To overcome these limitations, targeting the tumor microenvironment (TME) offers a promising alternative. Unlike conventional methods, TME-responsive strategies enable more precise, localized drug delivery. This approach leverages the distinct physiological hallmarks of malignant cells—such as elevated concentrations of reactive oxygen species (ROS), increased glutathione (GSH) levels, acidic pH, and unique enzyme overexpression—to trigger drug release only within the pathological site, thereby distinguishing them from healthy tissues.2,3 For instance, nanocarriers engineered with dual-responsive motifs can facilitate the controlled release of drug in response to specific redox potential (glutathione overexpression) and localized acidic environments.4,5 In contrast to low-micromolar extracellular/plasma GSH in humans (typically about 1–6 μM, with a representative value of 2.8 ± 0.91 μM reported in healthy adults), intracellular GSH in tumor cells is generally in the millimolar range, commonly about 1–10 mM, ie, roughly 103-fold higher than plasma and redox-sensitive advanced drug delivery systems were formulated by taking advantage of the elevated GSH concentration in tumor cells.6–8 Clinicians have utilized disulfiram (DSF) to treat alcohol use disorder for more than half a century.9 However, it has also shown promise as an anticancer agent due to its ability to inhibit multiple signaling pathways involved in cancer progression and metastasis. Numerous studies have demonstrated the strong anticancer potential of this compound against a broad spectrum of malignancies.10 Despite these potential benefits, DSF’s clinical translation has been limited by poor solubility and low bioavailability. To overcome these challenges and enhance the therapeutic efficacy of DSF in cancer treatment, researchers have explored nanoparticle-based delivery systems.11 Although encapsulating or conjugating DSF onto nanoparticles like polymer nanoparticles, liposomes, and nano-emulsions includes several advantages such as enhancing both the stability and solubility of DSF in physiological conditions, drug’s effectiveness, greater control over the release profile of the drug, prolonged circulation time within the body and provide opportunities for active targeting strategies by functionalizing their surfaces (this approach enables accumulation of DSF-loaded nanoparticles at tumor sites while minimizing off-target effects on healthy tissues) but many still suffer from premature leakage, poor aqueous dispersibility and….12,13 DSF can also encapsulated within carriers or nanoparticles through disulfide bridges. Hence, increased GSH levels inside cells can enhance the release of therapeutics from nanoparticles.14 Fluorescent carbon dots (CDs) loaded with therapeutic agents have been developed for various applications, including drug delivery and cellular imaging.15,16 This is mainly attributed to their biocompatibility and low toxicity, as well as distinctive optical properties, including remarkable photostability, and strong fluorescence emission that can be manipulated by adjusting particle size, surface chemistry, and excitation wavelength.17 In contrast to many conventional polymeric or inorganic nanocarriers, carbon dots can simultaneously serve as nanoscale drug-delivery platforms and intrinsically fluorescent components, enabling monitoring of their physicochemical behavior and cellular interactions. Their biological performance is strongly governed by surface chemistry; therefore, rational surface functionalization is essential to control drug loading, colloidal stability, cellular uptake, and stimulus-responsive release.18 Nitrogen doping was selected in the present study because incorporation of nitrogen-containing functionalities can modify the electronic structure, surface polarity, charge distribution, and density of reactive sites of carbon dots. These features may improve aqueous stability and provide additional hydrogen-bonding, electrostatic, and other noncovalent interactions with drug molecules. Nitrogen doping can also enhance the photoluminescence properties of carbon dots, thereby facilitating their characterization and potentially enabling optical tracking of the nanocarrier. Nevertheless, the contribution of nitrogen doping should be interpreted in the context of the whole surface structure, since the final properties of carbon dots depend on the introduction composition, reaction conditions, particle size, and distribution of surface groups.19,20 CDs coated with amino and carboxyl groups have been used as a carrier for doxorubicin (DOX) as a model drug.21 Pei et al22 designed a carbon dot-templated star-like copolymer prodrug system that achieves tumor-specific delivery via pH- and reduction-sensitive dual-release mechanisms. Nair et al23 reviewed the use of natural carbon-based quantum dots for drug delivery applications, highlighting their diverse functionality and biocompatibility. Wei et al24 reported the synthesis of green-emitting carbon quantum dots co-doped with nitrogen and sulfur, and showed their suitability for cellular imaging applications. Chung et al25 also designed nanoparticles composed of iron oxide and carbon dots that functioned as fluorescent agents for cancer imaging and drug delivery. Herein, the introduction of sulfur-containing or thiol-enriched surface functionalities was considered particularly relevant to the development of a redox-responsive disulfiram-delivery system. Thiol groups can alter the surface hydrophilicity, charge, and chemical reactivity of carbon dots and may provide additional interactions with disulfiram or with linker structures used for drug attachment. More importantly, thiol/disulfide chemistry used in intracellular redox-responsive delivery because the glutathione concentration is substantially higher in the cytosol than in the extracellular environment. In principle, GSH can participate in thiol–disulfide exchange reactions and destabilize disulfide-containing linkages, thereby promoting the release of a conjugated or associated payload. Thus, sulfur-containing surface modification was introduced to provide a chemically plausible basis for redox responsiveness, rather than merely to increase the loading capacity of the nanodots. Thus, thiol-functionalized nitrogen-doped carbon dots (SNCDs) were synthesized and conjugated with DSF. A GSH-triggered DSF-SNCD nano-delivery platform may improve the precision and release behavior of DSF in MCF-7 human tumor cells, ultimately enhancing its therapeutic potential. Under GSH-rich reducing conditions, the nanocarrier exhibited a stimulus-responsive release profile, with markedly accelerated DSF release compared with the GSH-free condition. To our knowledge, this study presents the first GSH-responsive DSF-SNCD nanocarrier for targeted delivery of disulfiram to cells with increased GSH content. In addition, the anticancer toxicity of DSF-SNCDs was compared against the effect of unencapsulated DSF. This study highlights the versatility and potential of carbon dots as fluorescent probes for cell imaging and drug delivery applications. Experimental Materials The citric acid (98% purity), urea, glutathione, and phosphate buffered saline (PBS) were obtained from Merck Company (Merck, Germany) and used without any further purification. Culture medium (DMEM), fetal bovine serum (FBS), and trypsin-ethylenediamine tetra acetic acid X10 were prepared from Biosera Company (Biosera, England). Breast cancer cells MCF-7 cell were purchased from the National Cell Bank of Iran (Pasteur Institute, Tehran, Iran). Quinine hemisulfate (95% purity), disulfiram, dithiothreitol (DTT), dimethyl sulfoxide (DMSO), ethylenediamine tetraacetic acid (EDTA) (98% purity), and Thiazolyl blue tetrazolium bromide (MTT, 98% purity) were obtained from Sigma-Aldrich, Germany. All chemicals were analytical grade and used as supplied. Deionized water was employed for preparing all aqueous solutions. Instrumentation HRTEM images of the prepared SNCDs were obtained using a JEOL JEM-2100 instrument operated at 200 kV for structural characterization. The optical absorption spectra in the ultraviolet–visible region was recorded spectrophotometrically using an Agilent Cary 60 instrument (Agilent Technologies, Santa Clara, USA). Fluorescence spectra were measured by a spectrofluorometer (Carry Eclipse, Varian, Australia). FT-IR analysis was carried out on a TENSOR 27 instrument (Bruker, Germany) employing KBr pellets. Dynamic light scattering (DLS) analysis by Zetasizer Nano S (Malvern Instruments Ltd., USA) was used to determine the average hydrodynamic particle size. Synthesis of Nitrogen-Doped Carbon Dots (NCDs) N-doped CDs were produced by dissolving citric acid (0.5 g) and urea (1.5 g) in 7 mL of dimethyl sulfoxide under constant stirring. The resulting uniform mixture was then poured into an autoclave lined with Teflon and subjected to heating at 160 °C for 4 h. The synthesis was accompanied by a pronounced color shift from colorless to dark brown, indicating the formation of nitrogen-doped carbon dots. Once cooling to ambient, the colloidal product was centrifuged at 12,000 rpm for 20 minutes to remove large aggregates. The supernatant was collected and dialyzed against distilled water for 24 hours using a dialysis membrane with a molecular weight cut-off (MWCO) of 3.5 kDa to remove unreacted precursors and small molecular impurities. The purified product was freeze-dried using a vacuum freeze-dryer, yielding a dark brown powder of NCDs, which was stored for further use. For characterization and subsequent functionalization, the NCDs were redispersed in distilled water at the desired concentration. Preparation of Sulfur-Functionalized NCDs (SNCDs) To sulfur functionalize the synthesized NCDs, 0.08 g of dithiothreitol (DTT) was added to 3 mL of an aqueous NCD colloid. The mixture was stirred continuously at room temperature for 24 hours to ensure thorough functionalization. After the reaction, the sulfur-functionalized NCDs (SNCDs) were purified to remove unreacted DTT and other byproducts using a dialysis membrane (MWCO 3.5 kDa) against distilled water for 24 hours. After preparation, the SNCD colloid (10 mg/mL) was maintained at 4 °C for further studies. Conjugation of Disulfiram (DSF) with SNCDs The DSF-SNCDs were prepared by exploiting the formation of disulfide bonds between DSF and the thiol groups on the SNCDs. Specifically, 210 µL of the SNCDs colloid (10 mg/mL) was added dropwise to 3 mL of a DSF solution (250 µg/mL in distilled water) under vigorous stirring. The mixture was allowed to react for 24 hours at room temperature to ensure maximum loading of DSF onto the SNCDs. The colloid was then transferred into a dialysis membrane (MWCO 3.5 kDa) and dialyzed against distilled water overnight to remove unbound DSF. The concentration of free DSF in the dialysate was determined by measuring the absorbance at 280 nm spectrophotometrically, using a calibration curve constructed from standard DSF solutions in deionized water (Figure S1). The drug loading efficiency (DLE) was calculated based on the initial and final concentrations of DSF. Fluorescence Efficiency of SNCDs The quantum yield of SNCD fluorescence was estimated by comparison with quinine hemisulfate dissolved in 0.10 M sulfuric acid, which has a reported reference value of 54%.26 The value was obtained from the standard comparative equation: \documentclass[12pt]{minimal} \usepackage{wasysym} \usepackage[substack]{amsmath} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage[mathscr]{eucal} \usepackage{mathrsfs} \DeclareFontFamily{T1}{linotext}{} \DeclareFontShape{T1}{linotext}{m}{n} {linotext }{} \DeclareSymbolFont{linotext}{T1}{linotext}{m}{n} \DeclareSymbolFontAlphabet{\mathLINOTEXT}{linotext} \begin{document}$${\varphi _s} = {\varphi _r} \left({{\raise0.7ex\hbox{${{m_s}}$} \!\mathord{\left/ {\vphantom {{{m_s}} {{m_r}}}}\right.\kern-0.5mu}\!\lower0.7ex\hbox{${{m_r}}$}}} \right)\left({{\raise0.7ex\hbox{${{\eta _s}}$} \!\mathord{\left/ {\vphantom {{{\eta _s}} {{\eta _r}}}}\right.\kern-0.5mu}\!\lower0.7ex\hbox{${{\eta _r}}$}}} \right)$$\end{document} In this expression, φ denotes quantum yield, m corresponds to the slope derived from the linear relationship between emission intensity and absorbance, and η represents the solvent refractive index. The subscripts s and r indicate the sample and the reference, respectively. Evaluation of DSF Dissolution and Release Kinetics A dialysis-based diffusion system was set up to track the release pattern of the DSF from the nanoparticles under simulated conditions. First, two milliliters of as prepared DSF-SNCDs colloid were transferred into a dialysis bag (MWCO 3.5 kDa), and then the dialysis bag was incubated in 20 mL of PBS buffer at pH 7.4 and 5.4 with or without GSH (5% w/v). Thermal and mechanical control was applied by keeping the release environment at 37 ± 1 °C under 100 rpm shaking for a 24 h period. To ensure the maintenance of sink conditions, 3 mL of the supernatant was periodically withdrawn and replaced with an identical volume of fresh PBS buffer over the duration of the experiment. The amount of DSF released from the nanocarrier was calculated using the calibration plot of DSF in PBS solution by measuring the absorbance at 280 nm. To monitor and eliminate any potential UV–Vis absorbance contribution arising from background S-NCD leaching, a parallel control experiment was conducted under identical conditions using 2.0 mL of pristine, drug-free S-NCDs inside a dialysis bag. To determine the profiles of DSF release from the SNCD nanoconjugates, the cumulative release fraction at successive intervals was calculated according to the following relation. \documentclass[12pt]{minimal} \usepackage{wasysym} \usepackage[substack]{amsmath} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage[mathscr]{eucal} \usepackage{mathrsfs} \DeclareFontFamily{T1}{linotext}{} \DeclareFontShape{T1}{linotext}{m}{n} {linotext }{} \DeclareSymbolFont{linotext}{T1}{linotext}{m}{n} \DeclareSymbolFontAlphabet{\mathLINOTEXT}{linotext} \begin{document}$${m_{DS{F_n}}} = \mathop \sum \limits_{i = 1}^n 20*{c_{DS{F_i}}}$$\end{document} \documentclass[12pt]{minimal} \usepackage{wasysym} \usepackage[substack]{amsmath} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage[mathscr]{eucal} \usepackage{mathrsfs} \DeclareFontFamily{T1}{linotext}{} \DeclareFontShape{T1}{linotext}{m}{n} {linotext }{} \DeclareSymbolFont{linotext}{T1}{linotext}{m}{n} \DeclareSymbolFontAlphabet{\mathLINOTEXT}{linotext} \begin{document}$$CD{R_n}\% = {{{m_{DS{F_n}}}} \over {{m_{DS{F_{loaded}}}}}} \times 100$$\end{document} Where \documentclass[12pt]{minimal} \usepackage{wasysym} \usepackage[substack]{amsmath} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage[mathscr]{eucal} \usepackage{mathrsfs} \DeclareFontFamily{T1}{linotext}{} \DeclareFontShape{T1}{linotext}{m}{n} {linotext }{} \DeclareSymbolFont{linotext}{T1}{linotext}{m}{n} \DeclareSymbolFontAlphabet{\mathLINOTEXT}{linotext} \begin{document}${m_{DS{F_n}}}$\end{document} is the cumulative released amount of DSF in μ g at nth sampling stage and \documentclass[12pt]{minimal} \usepackage{wasysym} \usepackage[substack]{amsmath} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage[mathscr]{eucal} \usepackage{mathrsfs} \DeclareFontFamily{T1}{linotext}{} \DeclareFontShape{T1}{linotext}{m}{n} {linotext }{} \DeclareSymbolFont{linotext}{T1}{linotext}{m}{n} \DeclareSymbolFontAlphabet{\mathLINOTEXT}{linotext} \begin{document}${c_{DS{F_i}}}$\end{document} is the calculated concentration of DSF in μ g/mL at the ith stage. The CDR is the abbreviation for cumulative drug released in percent and \documentclass[12pt]{minimal} \usepackage{wasysym} \usepackage[substack]{amsmath} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage[mathscr]{eucal} \usepackage{mathrsfs} \DeclareFontFamily{T1}{linotext}{} \DeclareFontShape{T1}{linotext}{m}{n} {linotext }{} \DeclareSymbolFont{linotext}{T1}{linotext}{m}{n} \DeclareSymbolFontAlphabet{\mathLINOTEXT}{linotext} \begin{document}${m_{DS{F_{loaded}}}}$\end{document} is the total amount of DSF in μ g conjugated to the SNCDs. MTT Test Cell viability was evaluated via the MTT reduction assay, following the established protocol.26 Initially, MCF-7 cell suspensions were aliquoted into 96-well culture plates at a seeding density of 5 × 103 cells/well, utilizing high-glucose DMEM fortified with 10% fetal bovine serum (FBS) and 1% antibiotic/antimycotic reagents. The cultures were allowed to adhere for 24 h at 37 °C within a 5% CO2 humidified incubator. Following this, the culture medium was exchanged for fresh medium incorporating varying concentrations of SNCDs, DSF-SNCDs, or pure DSF, and the plates were incubated for an additional 48 h. To quantify metabolic activity, the supernatant was removed, followed by a gentle PBS wash. Each well was then treated with 20 μL of MTT solution (5 mg/mL). Following a 4-hour incubation period at 37 °C, the MTT solution was aspirated, and 150 μL of DMSO was introduced to each well to facilitate the dissolution of the resulting formazan precipitates. After 10 minutes of mechanical agitation, the optical density was recorded at 570 nm utilizing a Wellscan Mk3 microplate reader (Labsystems Dragon Ltd, Finland). Results and Discussion The Mechanism of Redox-Responsive Drug Delivery System It is well known that GSH plays a crucial role in the human body, where it can be oxidized to glutathione disulfide (GSSG) in a normal tissue microenvironment. In the presence of NADPH-dependent glutathione reductase, GSSG is regenerated to GSH, thereby supporting redox equilibrium essential for cellular function and survival.27 By comparison, tumor cells generally maintain much higher intracellular GSH levels (approximately 2–20 mM) because of their altered microenvironment, growth-related nutrient demands, and genetic abnormalities.4,28 The key goal of this work is to construct a nanocarrier that remains stable in non-reducing environments (“Off” state) while rapidly releasing DSF in reducing cancer cells media (“On” state). Scheme 1 clearly depicts the release behavior of the redox-responsive drug delivery system. Thiol-functionalized N-doped carbon quantum dots (SNCDs) provide a multifunctional surface suitable for DSF immobilization through disulfide chemistry. In healthy tissues media (low reducing power), disulfide linkages remain intact, limiting DSF leakage. The nanocarrier accumulates in tumor tissues through the enhanced permeability and retention effect during prolonged blood circulation in the presence of strong redox potential, the responsive linker decomposes, resulting in the collapse of the nanocarrier structure and prompt liberation of the encapsulated drugs. This mechanism allows for precise delivery, rapid drug release, and minimal side effects, leveraging the reduction potential gradient between cancerous and healthy cells. Scheme 1Redox-responsive system for drug delivery to cancer cells.A step-by-step schematic shows a nanoparticle labeled DSF-SNCD on the left with multiple disulfide (S to S) linkages highlighted in red and yellow, alongside scissors icons marking cleavage sites. A curved arrow labeled GSH points to the right toward a starburst reading High GSH, indicating an intracellular reducing environment. On the right, the particle is labeled SNCD, depicting the carrier after bond cleavage, with small molecular fragments shown as released cargo. A dotted blue bilayer labeled Tumor Cell represents the cell membrane and an oval labeled Nucleolus sits inside the cell to indicate the intracellular destination region.Schematic of disulfide-linked nanocarrier cleaved by glutathione inside a tumor cell. Redox-responsive system for drug delivery to cancer cells. Characterization of SNCDs and DSF-SNCDs We utilized a straightforward one-pot hydrothermal technique to produce NCDs, with citric acid serving as the primary carbon source, urea as the nitrogen source. For thiol functionalization of NCDs and facilitate their attachment to the drug via a covalent disulfide bond, DTT was utilized. NCDs typically exhibit significant ultraviolet absorption band with tails even reach into the visible spectrum.29 When the SNCDs are exposed to light from a 365 nm lamp, they exhibit a distinct and vibrant blue color. The particle size, surface engineering, and heteroatom doping of SNCDs strongly influence their optical properties, thereby expanding their applicability. Among these factors, nitrogen doping is particularly beneficial because it enhances aqueous solubility without the need for additional agents and simultaneously improves the photoluminescence efficiency of carbon dots.30 The UV-Vis spectrum of our SNCDs solution (Figure 1) exhibits absorption band at wavelength of 240 nm, which can be attributed to the π-π* transition of the aromatic C=C sp2 domains.20 The other strong absorption peak at 340 is assigned to the n–π* transition of possible C=O or C=N groups present in SNCDs structure, and the shoulder around 480 nm is given to multiple surface defect states.31 Similar to previous reports,32,33 the peak at 620 nm shows that our SNCDs are in an oxidized state. It was observed that the optical features of the carbon dots are significantly dictated by the functional groups present on their surface. Figure 1Electronic absorption spectrum of colloidal SNCDs.A line graph showing absorbance versus wavelength for a colloidal SNCDs. Electronic absorption spectrum of colloidal SNCDs. Figure 2A Shows that the SNCDs colloid can emit a bright blue fluorescence emission centered at 460 nm when excited at 365 nm. As shown in Figure 2B, by increasing the excitation wavelengths from 345 nm to 375 nm, the emission peaks showed red-shift (~ 20 nm) to longer wavelengths owing to the varied band gaps from different surface states of CDs.34 The SNCDs achieved their highest fluorescence intensity at an emission of 460 nm with 365 nm excitation. As documented in prior studies, the excitation-dependent emission observed in the synthesized SNCDs is primarily attributed to quantum confinement and surface-defect states.35 Moreover, Figure 2C illustrates that the fluorescence intensity increases with the concentration of the aqueous SNCD colloid. Notably, the inset demonstrates a linear relationship between the fluorescence intensity at 460 nm and the SNCD concentration. Figure 2(A) Excitation (black line) spectra (recorded with \documentclass[12pt]{minimal} \usepackage{wasysym} \usepackage[substack]{amsmath} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage[mathscr]{eucal} \usepackage{mathrsfs} \DeclareFontFamily{T1}{linotext}{} \DeclareFontShape{T1}{linotext}{m}{n} {linotext }{} \DeclareSymbolFont{linotext}{T1}{linotext}{m}{n} \DeclareSymbolFontAlphabet{\mathLINOTEXT}{linotext} \begin{document}${\lambda _{em}} = 460 nm$\end{document}) and emission (red line) spectra (recorded with \documentclass[12pt]{minimal} \usepackage{wasysym} \usepackage[substack]{amsmath} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage[mathscr]{eucal} \usepackage{mathrsfs} \DeclareFontFamily{T1}{linotext}{} \DeclareFontShape{T1}{linotext}{m}{n} {linotext }{} \DeclareSymbolFont{linotext}{T1}{linotext}{m}{n} \DeclareSymbolFontAlphabet{\mathLINOTEXT}{linotext} \begin{document}${\lambda _{ex}} = 365 nm$\end{document}); (B) Excitation-dependent fluorescence emission spectra of SNCDs; (C) Concentration-dependent fluorescence spectra of SNCDs (\documentclass[12pt]{minimal} \usepackage{wasysym} \usepackage[substack]{amsmath} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage[mathscr]{eucal} \usepackage{mathrsfs} \DeclareFontFamily{T1}{linotext}{} \DeclareFontShape{T1}{linotext}{m}{n} {linotext }{} \DeclareSymbolFont{linotext}{T1}{linotext}{m}{n} \DeclareSymbolFontAlphabet{\mathLINOTEXT}{linotext} \begin{document}${\lambda _{ex}} = 365 nm$\end{document}). The inset shows the linear plot of fluorescence intensity at 460 nm versus SNCD concentration.Image A: Multi-line graph of Excitation and Emission. X-axis: Wavelength (nm) 200-700, Y-axis: Fl. intensity (a.u.) 0-250. Excitation peaks at 230 nm and 360 nm, trough at 280 nm. Emission peaks at 460 nm, decreases to 700 nm. Image B: Excitation-dependent fluorescence emission spectra of SNCDs. X-axis: Wavelength (nm) 350-600, Y-axis: Fl. intensity (a.u.) 0-300. Peaks: 345nm (450, 230), 350nm (455, 260), 360nm (460, 290), 365nm (460, 310), 375nm (470, 230). Highest peak at 365nm, 310 a.u.; peak shifts from 450 nm to 470 nm. Image C: Concentration-dependent fluorescence spectra of SNCDs. X-axis: Wavelength (nm) 400-650, Y-axis: Fl. intensity (a.u.) 0-300. Peaks: 110 ul (460, 95), 210 ul (465, 170), 310 ul (465, 250), 410 ul (465, 290). Intensity increases with concentration. Inset: Line plot with markers. X-axis: Volume of SNCD (uL) 100-400, Y-axis: Fl. intensity (a.u.) 50-300. Points: (110, 100), (210, 170), (310, 250), (410, 290), showing upward trend.A multi line graph set showing SNCDs excitation, emission and fluorescence spectra with an inset plot. (A) Excitation (black line) spectra (recorded with \documentclass[12pt]{minimal} \usepackage{wasysym} \usepackage[substack]{amsmath} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage[mathscr]{eucal} \usepackage{mathrsfs} \DeclareFontFamily{T1}{linotext}{} \DeclareFontShape{T1}{linotext}{m}{n} {linotext }{} \DeclareSymbolFont{linotext}{T1}{linotext}{m}{n} \DeclareSymbolFontAlphabet{\mathLINOTEXT}{linotext} \begin{document}${\lambda _{em}} = 460 nm$\end{document}) and emission (red line) spectra (recorded with \documentclass[12pt]{minimal} \usepackage{wasysym} \usepackage[substack]{amsmath} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage[mathscr]{eucal} \usepackage{mathrsfs} \DeclareFontFamily{T1}{linotext}{} \DeclareFontShape{T1}{linotext}{m}{n} {linotext }{} \DeclareSymbolFont{linotext}{T1}{linotext}{m}{n} \DeclareSymbolFontAlphabet{\mathLINOTEXT}{linotext} \begin{document}${\lambda _{ex}} = 365 nm$\end{document}); (B) Excitation-dependent fluorescence emission spectra of SNCDs; (C) Concentration-dependent fluorescence spectra of SNCDs (\documentclass[12pt]{minimal} \usepackage{wasysym} \usepackage[substack]{amsmath} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage[mathscr]{eucal} \usepackage{mathrsfs} \DeclareFontFamily{T1}{linotext}{} \DeclareFontShape{T1}{linotext}{m}{n} {linotext }{} \DeclareSymbolFont{linotext}{T1}{linotext}{m}{n} \DeclareSymbolFontAlphabet{\mathLINOTEXT}{linotext} \begin{document}${\lambda _{ex}} = 365 nm$\end{document}). The inset shows the linear plot of fluorescence intensity at 460 nm versus SNCD concentration. DLS measurements (Figure 3A) revealed a hydrodynamic size of 2 ± 1 nm. Furthermore, TEM imaging (Figure 3B) revealed spherical, uniformly dispersed SNCDs with an average diameter of approximately 3 nm. Figure 3DLS analysis (A) and TEM image (scale bar: 5 nm) (B) of SNCDs.A line graph showing DLS analysis and a transmission electron microscopy micrograph showing SNCDs size characteristics. DLS analysis (A) and TEM image (scale bar: 5 nm) (B) of SNCDs. The functional groups of the SNCDs were detected using FTIR analysis and presented in Figure 4. The FT-IR spectrum exhibits a prominent broad absorption band between 3000 and 3500 cm−1, indicative of N–H and O–H stretching, which substantiates the incorporation of nitrogen within the SNCDs. Spectral features at 1673 cm−1, 1622 cm−1, and 1436 cm−1 are assigned to C=O stretching (carboxyl groups), N–H bending, and C–H bending modes, respectively.36 Furthermore, the absorption at 1027 cm−1 represents C–O stretching, while the peak appearing at 1314 cm−1 is characteristic of C–N stretching. The signals at 2913 cm−1 and 2999 cm−1 arise from the symmetric and asymmetric stretching of CH2 units.37 C–S stretching vibrations are identified within the 703–785 cm−1 region,38 and the bands noted between 2550 and 2600 cm−1 are attributed to the S–H stretching of thiol moieties.39 Figure 4FTIR spectrum of the synthesized SNCDs.A line graph of transmittance versus wavenumber for SNCDs. The x-axis is labeled, Wavenumber centimeter superscript minus 1, decreasing from 4000 to 500. The y-axis is labeled, Transmittance percent, ranging from 20 to 100 with 10 percent increments. The curve starts near 100 percent at about 4000, drops to a broad minimum near 3412.46 at about 30 percent, then rises toward about 95 to 100 percent by about 2500. Narrow dips appear near 2999.06 and 2913.18. A deep dip occurs near 1673.84, followed by dips near 1622.04, 1436.11, 1417.99, 1314.75, 1184.15 and a very deep trough near 1027.00 reaching about 20 percent. Additional dips are labeled near 910.51, 861.15, 785.74, 703.36, 628.40, 582.84, 549.29 and 418.02.A line graph of SNCDs Fourier transform infrared spectrum with multiple absorption bands and sharp troughs. FTIR spectrum of the synthesized SNCDs. To calculate the SNCD quantum yield, the slope of the integrated emission (λem = 460 nm) versus absorbance (λex = 365 nm) plot (Figure 5A) was calculated and compared to the reference data for quinine hemisulfate (Figure 5B) using the slope method. A quantum yield of 5.2% was obtained for the SNCDs. This level of photoluminescence, resulting from the urea/citric acid hydrothermal route and thiol modification, reflects an enhancement relative to various other carbon quantum dot systems documented in the literature.26 Figure 5Fluorescence intensity–absorbance relationships at different concentrations of quinine hemisulfate solutions and SNCDs colloids. (A) Linear relationship between fluorescence intensity at 450 nm (λex = 340 nm) and absorbance at 340 nm for quinine hemisulfate solutions, (B) Linear relationship between fluorescence intensity at 460 nm (λex = 365 nm) and absorbance at 365 nm for SNCD colloids. (Data are presented as mean ± SD (n = 3)).Image A shows a scatter plot with a linear fit and error bars. X-axis: Absorbance (a.u.), 0.012 to 0.028; Y-axis: Fl. intensity (a.u.), 60 to 180. Points: (0.013, 70), (0.018, 108), (0.025, 148), (0.029, 170). Line equation: y = 6325.7x - 12.56, R superscript 2 = 0.9917. Image B shows a scatter plot with a linear fit and error bars. X-axis: Wavelength (nm), 0.02 to 0.12; Y-axis: Fl. intensity (a.u.), 0 to 45. Points: (0.02, 6), (0.03, 10), (0.05, 19), (0.06, 24), (0.11, 39). Line equation: y = 342.98x + 1.315, R superscript 2 = 0.9965.Two scatter plots with fitted lines showing fluorescence intensity versus absorbance. Fluorescence intensity–absorbance relationships at different concentrations of quinine hemisulfate solutions and SNCDs colloids. (A) Linear relationship between fluorescence intensity at 450 nm (λex = 340 nm) and absorbance at 340 nm for quinine hemisulfate solutions, (B) Linear relationship between fluorescence intensity at 460 nm (λex = 365 nm) and absorbance at 365 nm for SNCD colloids. (Data are presented as mean ± SD (n = 3)). Drug Loading Performance: Loading Efficiency (LE) Encouraged by the unique nanoscale structure and high specific surface area of SNCDs, DSF was chosen as a model anticancer agent and loaded onto SNCDs. To establish a reliable DSF quantification method for drug loading calculations, UV–Vis spectroscopy was used. DSF shows a strong characteristic absorption peak at 280 nm, which was selected as the analytical wavelength for concentration determination (Figure S1A). A linear relationship between DSF concentration and absorbance was obtained (\documentclass[12pt]{minimal} \usepackage{wasysym} \usepackage[substack]{amsmath} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage[mathscr]{eucal} \usepackage{mathrsfs} \DeclareFontFamily{T1}{linotext}{} \DeclareFontShape{T1}{linotext}{m}{n} {linotext }{} \DeclareSymbolFont{linotext}{T1}{linotext}{m}{n} \DeclareSymbolFontAlphabet{\mathLINOTEXT}{linotext} \begin{document}${R^2} = 0.9967$\end{document}, Figure S1B). Using the UV–Vis calibration curve at 280 nm (Figure S1B), the amount of DSF loaded was quantified and the loading efficiency (LE) was determined as: \documentclass[12pt]{minimal} \usepackage{wasysym} \usepackage[substack]{amsmath} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage[mathscr]{eucal} \usepackage{mathrsfs} \DeclareFontFamily{T1}{linotext}{} \DeclareFontShape{T1}{linotext}{m}{n} {linotext }{} \DeclareSymbolFont{linotext}{T1}{linotext}{m}{n} \DeclareSymbolFontAlphabet{\mathLINOTEXT}{linotext} \begin{document}$$LE \left(\% \right) = {{{m_{DSF,loaded}}} \over {{m_{DSF, initial}}}}*100$$\end{document} The LE was ~75%, indicating that SNCDs possess excellent drug loading capacity. This high LE is consistent with the presence of abundant surface functional groups and strong drug–carrier interactions, and it also supports the low premature release observed under non-reducing conditions (Figure 6). Figure 6DSF release from the DSF-SNCDs particles over time at pH 5.4 and 7.4, with and without 5% GSH. (Data are presented as mean ± SD (n = 3)).A graph shows DSF release from DSF-SNCDs particles over time at pH 5.4 and 7.4, with and without 5% GSH. Time ranges from 0 to 25 hours and release percentage from 0 to 100. Markers: square for pH 5.4(GSH5%), circle for pH 5.4, upward triangle for pH 7.4, downward triangle for DSF. Error bars included. pH 5.4(GSH5%) starts at 18% at 0.5h, peaks at 80% at 24h. pH 5.4 begins at 10% at 0.5h, peaks at 34% at 5h, drops to 24% at 12h, ends at 29% at 24h. pH 7.4 starts at 5% at 0.5h, peaks at 29% at 10h, stabilizes around 28% from 12h to 24h. DSF starts at 18% at 0.5h, peaks at 98% at 24h, with a significant rise from 50% at 5h to 72% at 6h. DSF shows the fastest increase and highest release, while pH 5.4(GSH5%) steadily rises. pH 7.4 increases to 29% at 10h, stabilizing thereafter. pH 5.4 peaks early, decreases and ends at 29% at 24h.A multi-line graph showing DSF release percent over time at pH 5.4 and pH 7.4 with and without GSH. DSF release from the DSF-SNCDs particles over time at pH 5.4 and 7.4, with and without 5% GSH. (Data are presented as mean ± SD (n = 3)). In vitro Release: PH Stability vs GSH-Triggered “Off–On” Switching To evaluate controlled release behavior and stimulus responsiveness, cumulative DSF release from DSF‑SNCDs was investigated (Figure 6). Free DSF exhibited a rapid burst release, with >90% diffusion into the medium within ~15 h, reflecting weak retention in the absence of a carrier. In contrast, DSF‑SNCDs displayed substantially suppressed release at both pH 7.4 and pH 5.4 in the absence of reductant, with cumulative release remaining ≤~30% after 24 h. The relatively small difference between pH 7.4 and 5.4 indicates that pH alone does not govern drug liberation, confirming that DSF is largely retained on the carrier (“Off” state) under non-reducing conditions. Upon introduction of GSH (5% w/v in outside solution of dialysis bag) at pH 5.4, a pronounced increase in DSF release was observed, reaching ~80% at 24 h. This “On” state is attributed to GSH-mediated cleavage of disulfide linkages (reductive/thiol–disulfide exchange processes), enabling accelerated liberation of DSF in reducing environments. Collectively, these profiles demonstrate that DSF‑SNCDs behave primarily as a redox-responsive system rather than a purely pH-responsive carrier, thereby minimizing premature leakage while enabling intracellular activation. In vitro Cytotoxicity and Biocompatibility The biocompatibility of the carrier and the redox-triggered therapeutic action were assessed by MTT assay on MCF‑7 cells (Figure 7). Previous studies presented that nitrogen doped carbon dots were not cytotoxicity to MCF-7 cells.20 As shown in Figure 7, Bare SNCDs (CQD) showed excellent cytocompatibility, maintaining near-normal cell viability across the tested concentrations and remaining >~85% viable at 50 µg/mL, indicating low inherent toxicity of the nanocarrier. Figure 7Cytotoxicity of SNCDs with different concentrations for MCF-7 cells. (Data are presented as mean ± SD (n = 3)).A grouped bar graph showing cell viability for CQD, CQD and Drug, CQD and Drug in the presence of glutathione and Drug across concentration. Cytotoxicity of SNCDs with different concentrations for MCF-7 cells. (Data are presented as mean ± SD (n = 3)). Free DSF exhibited dose-dependent cytotoxicity, as expected. Notably, DSF‑loaded SNCDs (CQD‑D) in the absence of GSH caused lower cytotoxicity than free DSF, consistent with restricted DSF availability in the “Off” state. However, in the presence of GSH (CQD‑D + G), cell viability decreased sharply in a concentration-dependent manner; importantly, viability dropped to ~50% at 6.25 µg/mL, confirming that reducing conditions unlock the drug activity. This behavior directly supports the “Off–On” release mechanism observed in the release studies (Figure 6) and indicates that SNCDs can effectively deliver DSF in a GSH-rich environment, potentially enhancing therapeutic selectivity. Study Limitations In the in vitro release experiments, a reduced glutathione (GSH) concentration of 5% w/v (~163 mM) was introduced into the external release medium (outside the dialysis bag). While the effective local concentration experienced by the nanocarrier is modulated by the dialysis membrane molecular weight cut-off (MWCO), diffusion kinetics, and compartment volume ratios, we acknowledge that this represents a supranormal, accelerated reducing challenge rather than a direct simulation of the tumor microenvironment (TME). The release data presented herein should be interpreted as proof-of-concept evidence demonstrating the intrinsic redox responsiveness and cleavage kinetics of the nanocarrier under strong reducing stress, rather than as a direct quantitative prediction of in vivo drug release behavior. Future studies will systematically evaluate drug release and stability across a physiologically relevant GSH gradient to comprehensively elucidate the in vivo-mimicking release kinetics. Conclusion In conclusion, thiol-functionalized nitrogen-doped carbon quantum dots (SNCDs) were synthesized and evaluated as a glutathione (GSH)-responsive nanocarrier for disulfiram (DSF) delivery based on redox-cleavable linkages. Spectrophotometric analysis demonstrated an efficient drug loading capacity of ~75%. In vitro release profiles exhibited comparatively low drug liberation under non-reducing physiological and mildly acidic conditions (pH 7.4 and 5.4; ≤30% at 24 h), whereas the presence of GSH markedly triggered DSF release (~80% at 24 h). Cell viability assays indicated the cytocompatibility of blank SNCDs and demonstrated enhanced cytotoxicity of DSF-SNCDs against MCF-7 cells under redox-active conditions, which is consistent with the trigger-dependent release behavior observed in cell-free assays. Overall, these preliminary in vitro findings demonstrate the proof-of-concept potential of DSF-SNCDs for redox-responsive drug delivery with suppressed premature leakage. However, comprehensive cellular uptake studies, intracellular mechanistic analyses, and in vivo evaluations remain essential to fully determine their systemic biosafety, biodistribution, and therapeutic applicability.