Femtomolar hydrogen sulfide detection via hybrid small-molecule nano-arrays

Abstract

Early disease diagnosis hinges on the sensitive detection of signaling molecules. Among these, hydrogen sulfide (H₂S) has emerged as a critical player in cardiovascular and nervous system signaling. On-chip immunoassays, particularly nanoarray-based interfacial detection, offer promising avenues for ultra-sensitive analysis due to their confined reaction volumes and precise signal localization. Beyond the DNA or protein biomolecules array, this work presents a promising hybrid small molecule nano-array for H₂S detection, using the power of dual molecules: a dye for fluorescence emission and a quencher with specific H₂S reactivity. Upon H₂S interaction, the quenched fluorescence reignites, creating an easily detectable array of bright spots. The molecule nano-array sensor shows exceptional responses to H₂S over 8 magnitudes of dynamic range from 1 fM to 0.1 μM, with a remarkable detection limit of 1 fM, just using a 10 μL solution. This H₂S detection method has the potential to significantly improve bioassay platforms, and the hybrid small-molecule nano-arrays we developed could be a valuable tool for advancing signaling molecule detection.

Fig. 1. Illustration of highly sensitive H₂S detection via a hybrid molecule nano-array.

a The design of a hybrid molecule nano-array for H₂S detection, involving F1²⁺, RB, and cyclodextrin (CD) molecules. During the process of H₂S detection, the spots within the molecule array transition from the fluorescence “off” state to an “on” state. By integrating the fluorescence intensity (INTEN method) at high concentrations or counting the number of the “on” spots (NUMB method), the correlation with H₂S concentration can be established. b The chemical equation of F1²⁺ to F2 reduced by H₂S. c The typical fluorescent image of the array and the corresponding correlation spots obtained by INTEN and NUMB methods at high H₂S concentration and low concentration respectively.

Fig. 2. Fabrication and characterization of hybrid molecule nano-arrays.

a Schematic illustration of fabricating hybrid molecule nano-array using modified nanoxerography method, which involves surface charge writing, reversed-micelle preparation, and site-specific assembly by electrical trapping. b Electric potential map after writing a 5 × 5 charge array on the substrate. Scale bar, 5 μm. c AFM height map of the hybrid molecule nano-array, with an average height of 50 nm. Scale bar, 5 μm. d Dynamic light scattering (DLS) result of reversed-micelles, with the statistic size in dispersion around 3 nm. Inset is a photograph of the micelle dispersion in a test tube. e Scheme of the fluorescence array imaging reader. A homemade inverted widefield microscope is equipped with a fiber-coupled 532 nm continuous-wave laser diode and a sensitive Electron Multiplying Charge-Coupled Device (EMCCD) camera. f Schematic illustration of H₂S detection, and corresponding PL images of the hybrid molecule array after incubation with water (10 μL) and H₂S solution (10 μL, 0.1 μM). Scale bar, 5 μm.

Fig. 3. H₂S detection at high concentration using INTEN method.

a PL images of serial NaHS concentration in 10 pM, 100 pM, 1 nM, 10 nM, 100 nM, and 1 μM in the 5 × 5 nano-assay. The images of other concentrations (from 0.1 fM to 1 pM) are shown in Supplementary Fig. S12. Scale bar, 5 μm. b The working curve of NaHS tested on the nanoarray using the INTEN method. The black line shows the fitting result by the 4-parameter logistic regression model. The blue dotted line represents the background levels, which are defined as the mean luminescence intensity of negative controls with a three-fold standard deviation. The fitting parameters are listed in Supplementary Table S1. The blue dash line shows the linear regression after logit transformation of the method. Error bars indicate the standard deviation from three replicate experiments.

Fig. 4. H₂S detection at low concentration using NUMB method.

a PL images of serial NaHS concentration in 1 fM, 10 fM, 100 fM, 1 pM, 10 pM and 100 pM in the assay of 20 × 20 molecules nano-array. Scale bar, 9 μm. b The fitting results of the NUMB and INTEN methods. The 4-parameter logistic regression model yields an LOD of 1 fM in the NUMB method and 70 fM in the INTEN method. The dotted line represents the background levels, which are defined as the mean luminescence intensity of negative controls with a three-fold standard deviation. The fitting parameters are listed in Supplementary Table S1. Error bars indicate the standard deviation from three replicate experiments.