High ambipolar mobility fluorinated MXene for ultrasensitive biosensing transistors.

As a two-dimensional transition metal carbide, MXene has long been regarded as an ideal material for biosensing owing to its metallic conductivity and solution processability ^[1], while its surface, rich in terminal groups such as -F, -OH, and -O, further endows it with designable electronic structures and physicochemical properties ^[2][3]. Field-effect transistor (FET) architectures can directly convert biological recognition events into electrical signals, enabling real-time label-free detection; however, the intrinsic zero band gap of pristine Ti₃C₂Tₓ results in an extremely low on/off ratio (only about 1.02), making effective signal modulation difficult. Meanwhile, in aqueous environments the -OH terminations readily trigger hydrolytic oxidation, destroying the conductive framework and introducing interfacial noise, severely limiting practical deployment in physiological settings. To address these two bottlenecks, this work proposes a synergistic strategy of in-situ fluorination and gate functionalization: increasing the LiF molar ratio during the etching synthesis stage exploits Le Chatelier's principle to drive the termination equilibrium toward Ti-F bonds, yielding highly fluorine-covered F-MXene (Ti₃C₂(OH)₀.₁F₁.₆₈O₀.₂₂) in a one-step process; fluorine termination confers a weakly hydrophilic, chemically inert interface that suppresses oxidation and reshapes the electronic states near the Fermi level, while the gate-functionalization architecture physically separates the recognition units from the transport channel, preserving the channel's electrical integrity. The high electronegativity of fluorine increases the electric double-layer thickness, lowers the interfacial capacitance, and amplifies the effective gate-voltage division, thereby significantly enhancing the biogating response.

Materials characterization shows that F-MXene retains the accordion-like layered structure with an interlayer spacing of 1.7 nm (TEM); the XRD (002) peak shifts to a lower angle of 7.8°, but the expansion is smaller than that of hydrophilic MXene, suggesting that fluorine termination suppresses water intercalation. Raman peak shifts corroborate the substitution of -OH by -F; XPS confirms a dominant F 1s peak and a marked attenuation of the O 1s signal; SR-GIWAXS yields a sharp (002) peak (3.09°, d≈10.9 Å), confirming compact layer ordering. In electrical tests, pristine MXene exhibits metallic transport (conductivity 1.362×10⁴ S cm⁻¹), whereas F-MXene displays clear V-shaped ambipolar transport with a charge neutrality point (CNP) of 0.02 V and an on/off ratio of 15.25; the further-optimized FF-MXene sets a new record for MXene transistors, with hole/electron mobilities reaching 10.42 and 30.23 cm² V⁻¹ s⁻¹, respectively, and good reproducibility across three device batches (μₚ 9.60–10.42, μₙ 18.10–30.23 cm² V⁻¹ s⁻¹). However, biosensing performance depends on the normalized transconductance efficiency |gₘ/I_ds|, where F-MXene (118.81 V⁻¹) far exceeds FF-MXene (1.59 V⁻¹), a 74.7-fold gain; combined with its higher on/off ratio and lower CNP, F-MXene was ultimately selected to construct the immunosensor. After oriented immobilization of anti-MPT64 antibodies on the gate via Protein A, the stepwise decrease in output current directly corroborates the biogating mechanism; blinded clinical serum validation confirmed ultrasensitive detection of the Mycobacterium tuberculosis antigen MPT64 with high diagnostic accuracy. Electrochemical aptasensors have previously achieved MPT64 detection ^[4][5], but this platform combines label-free, real-time operation, high mobility, and blinded clinical validation, advancing MXene FETs toward practical aqueous-phase biodiagnostics.

This work establishes a co-design paradigm coupling high-quality fluorinated MXene with high-performance electronic devices, resolving the core contradiction between the low on/off ratio and aqueous instability of metallic MXene, and achieving a balance between device-level mobility and biosensing sensitivity. Limitations remain: batch-to-batch CNP fluctuations suggest that random distribution of surface terminations leads to variations in local electronic structure; a trade-off exists between absolute mobility and transconductance efficiency, with FF-MXene exhibiting higher mobility but lower sensing performance; and long-term stability in complex physiological matrices, multichannel integration, and multiplexed detection capability still await systematic evaluation.