Dual Molecularly Imprinted Mesoporous Organosilica Nanoprobes for AI-Assisted Glutathione Redox Ratio Sensing

Abstract

Molecular recognition nanomaterials that directly decode mixture composition remain underdeveloped because chemically related analytes often produce nonlinear and partially overlapping optical responses. Here, dual molecularly imprinted mesoporous organosilica nanoprobes are developed for AI-assisted determination of the glutathione redox ratio. Zwitterion-embedded mesoporous organosilica nanocavities are independently imprinted with reduced glutathione (GSH) or glutathione disulfide (GSSG), followed by pore confinement of blue-emissive graphene quantum dots as fluorescence transducers. The resulting MIMO_GSH-Q and MIMO_GSSG-Q probes generate complementary raw fluorescence fingerprints from mixed GSH/GSSG samples. A deep neural network trained on paired microplate fluorescence intensities performs inverse prediction of τ = [GSH]/[GSSG]. A single test mixture at τ = 0.60, absent from the training ratios, yields a mean predicted ratio of 0.575. These results provide a proof-of-concept demonstration of two-channel ratio inference under the tested conditions, while broader external validation is required to establish its range of applicability. This work demonstrates a trainable nanoporous recognition platform that transforms selective nanoconfined fluorescence responses into quantitative redox-ratio information, providing a general route toward AI-decodable sensing nanomaterials.

Surface-localized thermoresponsive molecularly imprinted polymer for gentamicin recognition and aqueous regeneration
Dominika Rapacz-Kinas, Mikołaj Mroczek, Joanna Wolska +1
RSC advances24 Oct 2026Peer-reviewedCC BYFull textGeneral chemistryEurope PMC
Ethylene-bridged Periodic Mesoporous Organosilica Nanoparticles: controlling architecture through mechanistic investigations
Sireen Al Khawand, Stéphanie Kodjikian, Philippe Trens +5
Materials Today Communications30 Sep 2026Peer-reviewedCC BY 4.0Full textMaterials & polymersHAL
Development and Optimization of a Molecularly Imprinted Polymer for Electrochemical Detection of Escherichia coli
Chuchart Pintavirooj, Naphatsawan Vongmanee, Sarinporn Visitsattapongse
Polymers14 Oct 2026Peer-reviewedCC BY 4.0Full textAnalytical chemistryCrossrefEurope PMC
Stimulus-responsive molecularly imprinted polymers for bisphenol A and dibutyl phthalate: Analytical trends, challenges, and future perspectives
Ike Susanti, Aliya Nur Hasanah, Rimadani Pratiwi +1
Sensors and Actuators Reports30 Oct 2026Peer-reviewedCC BYFull textAnalytical chemistryDOAJ
Electropolymerized Molecularly Imprinted Polymers Supported on Carbon-Based Materials for (Bio)sensing: Direct and Indirect Detection Strategies
Sergio Espinoza Torres, Astrid Choquehuanca-Azaña, Nathália F. B. Azeredo +2
Biosensors22 Jun 2026Peer-reviewedCC BY5 citedAnalytical chemistryOpenAlex

Cite

BibTeX RIS

BibTeX

@article{kim2026dual,
  title = {{Dual Molecularly Imprinted Mesoporous Organosilica Nanoprobes for AI-Assisted Glutathione Redox Ratio Sensing}},
  author = {Youngdo Kim and Jun Jo},
  journal = {ACS Omega},
  year = {2026},
  month = oct,
  doi = {10.1021/acsomega.6c08928},
  url = {https://pubs.acs.org/acsodf/article/doi/10.1021/acsomega.6c08928/5461664/Dual-Molecularly-Imprinted-Mesoporous-Organosilica},
  note = {CC BY-NC-ND 4.0},
}

RIS

TY  - JOUR
TI  - Dual Molecularly Imprinted Mesoporous Organosilica Nanoprobes for AI-Assisted Glutathione Redox Ratio Sensing
AU  - Youngdo Kim
AU  - Jun Jo
PY  - 2026
DA  - 2026/10/30
JO  - ACS Omega
DO  - 10.1021/acsomega.6c08928
UR  - https://pubs.acs.org/acsodf/article/doi/10.1021/acsomega.6c08928/5461664/Dual-Molecularly-Imprinted-Mesoporous-Organosilica
AB  - Molecular recognition nanomaterials that directly decode mixture composition remain underdeveloped because chemically related analytes often produce nonlinear and partially overlapping optical responses. Here, dual molecularly imprinted mesoporous organosilica nanoprobes are developed for AI-assisted determination of the glutathione redox ratio. Zwitterion-embedded mesoporous organosilica nanocavities are independently imprinted with reduced glutathione (GSH) or glutathione disulfide (GSSG), followed by pore confinement of blue-emissive graphene quantum dots as fluorescence transducers. The resulting MIMO_GSH-Q and MIMO_GSSG-Q probes generate complementary raw fluorescence fingerprints from mixed GSH/GSSG samples. A deep neural network trained on paired microplate fluorescence intensities performs inverse prediction of τ = [GSH]/[GSSG]. A single test mixture at τ = 0.60, absent from the training ratios, yields a mean predicted ratio of 0.575. These results provide a proof-of-concept demonstration of two-channel ratio inference under the tested conditions, while broader external validation is required to establish its range of applicability. This work demonstrates a trainable nanoporous recognition platform that transforms selective nanoconfined fluorescence responses into quantitative redox-ratio information, providing a general route toward AI-decodable sensing nanomaterials.
SN  - 2470-1343
ER  -

Abstract

Related

Cite