The detection of Fe³⁺ has garnered significant attention, underpinning pivotal advancements across diverse scientific and technological fields including synthetic chemistry^[1], biological imaging^{[2, 3]}, and food safety analytics^{[4, 5]}. There is a pressing demand for adept methodologies to monitor Fe³⁺ both in vivo and in vitro. Traditional methods like inductively coupled plasma-mass spectrometry^[6], voltammetry^[7], spectrophotometry^[8], and atomic absorption spectroscopy^[9] though can quantitatively detect Fe³⁺, they are usually complicated, time-consuming, and expensive^[2]. This motivates the development of rapid, streamlined modalities for Fe³⁺ quantification. Compared to conventional methods, fluorescent probes show advantages in faster responsiveness, greater practicability, and higher sensitivity^[10]. Among these probes, the ratiometric fluorescence is especially attractive for its self-calibrating measurements^[11]. For fluorescent materials, quantum dots (QDs) show particular promise due to readily tunable emissions^[12] and excellent photostability^[13]. Over the past decade, various QDs including carbon dots (CDs)^[14−17], CdTe QDs^{[18, 19]}, CuInS₂ QDs^[20], and graphene QDs (GQDs)^[21] have been explored to realize ratiometric fluorescent nanosensors for Fe³⁺. However, the existing ratiometric fluorescent nanosensors mainly exhibit dual emission in the visible region (380–700 nm)^[22]. This limits their use in in vivo biological imaging, where visible light is absorbed^[23], and in food safety detection, which often contends with the background fluorescence^[24]. Even if nanosensors satisfy ratiometric NIR fluorescence, they still have relatively high limits of detection (LODs) (larger than 0.1 μM)^{[25, 26]}, hindering the precise detection of Fe³⁺ in fields including biological toxicity evaluation and water quality assessment^[27]. Therefore, the development of innovative ratiometric fluorescent nanosensors that detect Fe³⁺ with dual NIR emission and a low LOD is essential, offering significant advantages for biomedical and environmental applications.

Compared with other QDs, silicon QDs (Si QDs) possess favorable biocompatibility^[13], eco-friendliness^[28], and earth-abundance^[29], enabling efficient NIR emission at wavelengths below 1100 nm^{[30, 31]}. Thus, Si QDs represent a promising fluorescent nanosensor for detecting Fe³⁺. However, existing Si QD-based Fe³⁺ nanosensors typically exhibit a single visible emission^[32−34]. To the best of our knowledge, the development of Si QDs as ratiometric NIR fluorescent nanosensors for Fe³⁺ detection remains unexplored. Incorporating erbium (Er) into Si QDs generates a secondary NIR peak around 1540 nm (⁴I_13/2 → ⁴I_15/2)^{[35, 36]}, corresponding to the 1500−1800 nm biological imaging window (BW-Ⅲ)^[37]. Moreover, Er possesses an [Xe]4f¹²6s² electronic configuration^[38], which becomes [Xe]4f¹¹ upon incorporation into a host material, shielded from external fields by 5s² and 5p⁶ electrons^[39]. This characteristic makes Er an ideal reference for ratiometric NIR fluorescent nanosensors. Therefore, if Er is doped into Si QDs, dual NIR emission can be obtained. Furthermore, hyperdoping Si QDs with Er (Si QDs:Er) effectively modulates their electron density given the low electronegativity (1.24)^[40] of Er, thereby enhancing surface coordination with analytes such as Fe^3+[33]. In addition, the electron transfer from Si QDs:Er to Fe³⁺ may be facilitated by the hyperdoping of Er atoms^[41]. Both enhanced surface coordination and electron transfer can potentially improve the sensitivity, further resulting in a lower LOD. We have previously synthesized Si QDs:Er using nonthermal plasma^[42], representing promising ratiometric NIR-fluorescent nanosensors^[43]. Therefore, developing water-dispersible Si QDs:Er merits investigation for detecting Fe³⁺.

In this work, a two-step surface functionalization method was implemented to render the Si QDs:Er water-dispersible. Initially, the Si QDs:Er underwent hydrosilylation with 1-dodecene, forming dodecyl-passivated Si QDs:Er (dodecyl-Si QDs:Er). Subsequently, these dodecyl-Si QDs:Er were encapsulated into micelles self-assembled from the amphiphilic F127 polymer. The F127-encapsulated dodecyl-Si QDs:Er (F127-dodecyl-Si QDs:Er) are capable of emitting NIR light at wavelengths of 810 and 1540 nm. Moreover, the F127-dodecyl-Si QDs:Er micelles demonstrated remarkable sensitivity in the ratiometric detection of aqueous Fe³⁺ ions, achieving an LOD as low as 0.06 μM. This performance surpasses that of other nanosensors that primarily emit in the visible range and exhibit higher LODs. Furthermore, the effects of pH, recyclability, and the fluorescent quenching mechanisms underlying the Fe³⁺ response were elucidated. Capitalizing on the Fe³⁺ detection capabilities, integration into a sequential logic circuit demonstrated a programmable memory function, expanding the utility in responsive photonics.

Silane (SiH₄, 80%) mixed with argon (Ar, 20%) was sourced from Linde Electronic & Specialty Gases Co., Ltd. (Suzhou, China). Erbium(Ⅲ) 2,2,6,6-tetramethyl-3,5-heptanedionate (Er(tmhd)₃, 99.999%) was procured from Nanjing ai mou yuan Scientific equipment Co., Ltd. (Nanjing, China). Sinopharm Chemical Reagent Co., Ltd. (Shanghai, China) provided hydrofluoric acid (HF, 40%), methanol (98.5%), and toluene (99.5%), CaCl₂ (96%), CoCl₂·6H₂O (99%), CuCl₂·2H₂O (99%), KCl (99.5%), MgCl₂·6H₂O (GR), MnCl₂·4H₂O (99%), NaCl (99.5%), SnCl₄ (99%), ZnCl₂ (99.95%), and FeCl₃·6H₂O. Glutathione (99%), cysteine (99%), homocysteine (99%), glycine (99%), alanine (99%), isoleucine (99%), phenylalanine (99%), tryptophan (99%), Na₂SO₄ (99%), CH₃COONa (99%), C₆H₅Na₃O₇ (99%), mesitylene (97%), and 1-dodecene (95%) were obtained from Aladdin (Shanghai, China). Pluronic F127 triblock copolymer was obtained from Sigma-Aldrich (St. Louis, MO, USA). L-Cysteine was purchased from Beijing Gypsy Biotechnology Co., Ltd. (Beijing, China). All experiments utilized ultrapure water (18.25 MΩ·cm).

Si QDs:Er were fabricated using a nonthermal plasma technique. SiH₄/Ar (20% SiH₄) and Er(tmhd)₃ vapor (500 sccm, carried in Ar) were fed at flow rates of 4.8 and 500 sccm, respectively. The Er(tmhd)₃ was heated to 160 °C for vaporization. Plasma was generated at 60 W employing a 13.56 MHz power source coupled with a matching network, with the pressure maintained at 3.3 mbar. The as-produced Si QDs:Er underwent hydrosilylation with 1-dodecene, resulting in the formation of dodecyl-Si QDs:Er. These were then diluted in toluene to achieve a concentration of 1 mg·mL⁻¹, creating the stock solution.

To form micelles, 2 mL dodecyl-Si QDs:Er stock was mixed into 5 mL F127 (2−20 mg·mL⁻¹) aqueous solution. The mixture was then stirred at 45 °C for 12 h, allowing the complete evaporation of organic solvents. Subsequently, 2 mL ultrapure water was introduced, followed by 20 minutes of sonication, resulting in a 1 mg·mL⁻¹ QD dispersion. This dispersion was subsequently filtered using a 0.45 μm membrane.

Photoluminescence (PL) emission spectra were captured using an FLS1000 system (Edinburgh Instruments). For transient analysis, we used a 405 nm, 100 Hz pulsed laser for excitation, quantifying PL decay through time-corrected single photon counting. Transmission electron microscopy (TEM) imaging was done on a Talos F200X G2 microscope (Thermo Fisher) at 200 kV. Fourier transform infrared (FTIR) spectra were captured with a Vertex 80v spectrometer (Bruker). For X-ray photoelectron spectroscopy (XPS), a Kratos AXIS Ultra DLD spectrometer was employed.

Si QDs:Er were synthesized via nonthermal plasma with an Er concentration of 0.2%^[43], which exceeded the Er solubility limit in crystalline Si by two orders of magnitude^[44]. The as-synthesized hydrogen terminated Si QDs:Er (H-Si QDs:Er) underwent a two-step surface functionalization process to yield water-dispersible F127-dodecyl-Si QDs:Er micelles^[45], as illustrated in Fig. 1(a). Initially, H-Si QDs:Er were hydrosilylated with 1-dodecene to produce dodecyl-Si QDs:Er. These were then encapsulated in F127 to form water-dispersible F127-dodecyl-Si QDs:Er micelles, with diameters of the tens of nanometers. Each micelle contains about 5 to 20 dodecyl-Si QDs:Er. This two-step functionalization process was designed to realize the following key functionalities for Fe³⁺ detections: 1) The monodispersity of dodecyl-Si QDs:Er in toluene enables exquisite control over the micelle size^[45], which is crucial for in vivo detections^[46]. 2) The abundant hydroxyl (−OH) groups at the end of F127 can serve as coordination sites for Fe³⁺, enabling superior sensing capabilities for Si QDs:Er^{[47, 48]}.

Evidence for the successful synthesis of F127-dodecyl-Si QDs:Er was provided by the FTIR analysis, as detailed in Supplementary materials Fig. S1. In dodecyl-Si QDs:Er, the presence of the alkyl C−H_x-related signal at 2780−3026 cm⁻¹ and 1325−1516 cm⁻¹, together with the absence of the C=C (1641 cm⁻¹) and the =CH (3080 cm⁻¹) signals (Fig. S1, Supplementary materials), indicate the effective dodecyl-ligand attachments to the QD surfaces^[49]. For F127-dodecyl-Si QDs:Er, the FTIR spectrum further exhibits a prominent signal at 1115 cm⁻¹ which corresponds to C−C and C−O bonds within the F127 backbone (Fig. S1, Supplementary materials)^[50]. In addition, the signal at 3489 cm⁻¹ is responsible for the stretching vibration of −OH in F127^{[51, 52]}, validating the successful F127 encapsulation of dodecyl-Si QDs:Er. Fig. 1(b) displays the photo of a dodecyl-Si QDs:Er stock solution in toluene (left) and an aqueous F127-dodecyl-Si QDs:Er solution (right) under fluorescent light illumination. Both solutions exhibit good transparency, verifying uniform QD dispersion in both solvents. TEM and high-resolution TEM (HR-TEM) confirm the dodecyl-Si QDs:Er’s uniform size distribution and good crystallinity, with an interplanar spacing of the Si (111) plane at 0.317 nm (Fig. 1(c)). Additionally, the diameter of the Si QD in the stock solution (dodecyl-Si QDs:Er) is 4.1 ± 0.6 nm (Fig. 1(d)).

Then we move on to investigate the optical properties of F127-dodecyl-Si QDs:Er to determine the optimal synthetic conditions of these QDs for Fe³⁺ detection. Fig. 1(e) shows the PL spectra for the dodecyl-Si QDs:Er toluene stock solution and aqueous F127-dodecyl-Si QDs:Er solutions with different F127 concentrations. The PL peak at 810 nm for the stock solution is attributable to the band-to-band transition of Si QDs^{[53, 54]}, in alignment with the bandgap of Si QDs with a diameter of 4.1 ± 0.6 nm (Fig. 1(d)) due to the quantum confinement effect^[55]. Meanwhile, a 1540 nm PL peak appears in the spectrum, which originates from the ⁴I_13/2 → ⁴I_15/2 transition of Er^{3+[56, 57]}. Compared to the stock solution, the PL intensity of the 810 nm peak was markedly reduced in F127-dodecyl-Si QDs:Er solution at 2 mg·mL⁻¹ F127, accompanied by a blueshift of the PL peak from 810 to 787 nm. As the F127 concentration increased, the reduction of the PL intensity was mitigated (Fig. 1(f)) and the PL peak gradually redshifted to around 810 again (Fig. 1(g)). The initial PL intensity reduction stemmed from the loss of the unencapsulated dodecyl-Si QDs:Er during the filtration of F127-dodecyl-Si QDs:Er in water^{[45, 58]}. Increasing F127 concentrations led to more effective encapsulation of dodecyl-Si QDs:Er by F127 molecules. This reduced the filtering losses of Si QDs:Er in water and enhanced the PL intensity of the Si QDs. Additionally, higher F127 concentrations resulted in a decreased Si QDs:Er to F127 ratio. This led to a lower average number of Si QDs:Er being encapsulated within each micelle, which resulted in a reduction in the mean size of the F127-dodecyl-Si QDs:Er micelles in solution (Fig. S2(a), Supplementary materials).

To understand how varying F127 concentrations affect the PL peak wavelength, we used HR-TEM to check the mean size of the Si QDs within the F127-dodecyl-Si QDs:Er micelles with different F127 concentrations (Fig. S2(b), Supplementary materials). It is found that the QD size was reduced upon the encapsulation of F127 first (i.e., from 4.1 to 3.2 nm), and became larger with the increase of the F127 concentration (i.e., from 3.2 to 3.7 nm). Further insights from XPS indicate that all the Si QDs (with and without the F127) have been partially oxidized during synthesis (Fig. S2(c), Supplementary materials), to different degrees. Detailed XPS analysis^[43] reveals a ~1.07 nm thick oxide layer on the stock Si QDs (Table S1, Supplementary materials). After F127 encapsulation, thicker oxide layers were formed on the aqueous micelle-embedded Si QDs due to the OH⁻ diffusion^[45] in water, of which the oxide thickness varied from ~1.43 nm to ~1.18 nm when the F127 concentration increased from 2 to 10 mg·mL⁻¹. The increased F127 concentration might lead to a more complete encapsulation of the Si QDs:Er, thus preventing the further oxidation of the QDs by OH⁻ and stopping the QD size from reducing^[59]. These results suggest that the PL peak of the Si QDs is primarily influenced by the quantum confinement effect. This effect is linked to the change in the size of the Si QDs, which are adjusted by the altered oxidation process caused by F127 incorporation in aqueous solutions.

For the Er³⁺-induced 1540 nm PL peak, its PL intensity also reduced with the incorporation of low concentration F127 (2 mg·mL⁻¹), due to the filtration loss of F127-dodecyl-Si QDs:Er in water. Unlike the 810 nm PL peak, the PL intensity of the 1540 nm peak kept decreasing with the continuous increase of the F127 concentration (Fig. 1(f)). The cause of this phenomenon is likely the vibrational quenching of electronically excited Er³⁺ by −OH and C−H_x oscillators, whose signal intensity exhibits a continuous increase concomitant with rising F127 concentration (Figs. S3(a) and S3(b), Supplementary materials)^{[60, 61]}. In the meantime, the 1540 nm Er³⁺ peak remains unshifted with the encapsulation of F127 (Fig. 1(g)). This is attributed to the shielding effect of the outer 5s² and 5p⁶ shells on the 4f electrons of Er^{3+[42, 62]}. Given the opposite trends in PL intensity changes between the Si QD and Er³⁺ PL peaks with varying F127 concentrations (Fig. 1(f)), we selected F127-dodecyl-Si QDs:Er at a concentration of 10 mg·mL⁻¹. This concentration offers balanced signal-to-noise ratios for both PL peaks, making it suitable for further studies.

The dual NIR emission and the presence of abundant −OH groups suggest F127-dodecyl-Si QDs:Er may serve as a good ratiometric fluorescent nanosensor for Fe³⁺ detection^{[11, 63]}. As a proof-of-concept, a nanosensor for Fe³⁺ was demonstrated herein. It is known that the real-world application environment is complex and unpredictable. Here we first explore how pH levels ranging from 2 to 13 affect the sensor performance of F127-dodecyl-Si QDs:Er. Supplementary materials Fig. S4(a) illustrates how the pH impacts the PL intensity at 810 nm for F127-dodecyl-Si QDs:Er, both before and after combining with Fe³⁺. The PL intensity of the Si QDs gradually decreased with the increase in pH value, with a drastic reduction at pH above 7. Meanwhile, the introduction of Fe³⁺ significantly quenched the PL emitted by Si QDs, demonstrating the capability of F127-dodecyl-Si QDs:Er for Fe³⁺ detections. To evaluate the sensitivity of the PL-intensity change upon the attachment of Fe³⁺, we define the quenching efficiency (ηQ.E.) by

1ηQ.E.=I810nm−I810nm(Fe)I810nm,

where I810nm and I810nm(Fe) are the PL intensities of F127-dodecyl-Si QDs:Er at 810 nm without and with the introduction of Fe³⁺, respectively. The pH-dependent ηQ.E. is also shown in Supplementary materials Fig. S4(a), which gradually increases with the escalation of the pH. As a result, the neutral pH condition (pH = 7) guarantees both the high PL intensity and ηQ.E., which is optimal for the utilization of F127-dodecyl-Si QDs:Er for Fe³⁺ detections.

To assess the performance of F127-dodecyl-Si QDs:Er as a ratiometric NIR fluorescent nanosensor, the PL spectra were recorded using the titration method across a range of Fe³⁺ concentrations, from 0 to 100 mM, detailed in Fig. 2(a). It is evident that with the increase in the Fe³⁺ concentration (C_Fe), the PL intensity at 810 nm gradually decreases, while the PL intensity at 1540 nm remains nearly constant (Fig. S4(b), Supplementary materials). This result validates that the PL intensity at 1540 nm can serve as a reference signal, enabling the use of the fluorescence ratio Q = I₈₁₀/I₁₅₄₀ for Fe³⁺ detection. Fig. 2(b) exhibits the relationship between Q and C_Fe, manifesting a well-defined linear correlation within the ranges of 0−0.01 mM (Region Ⅰ) and 10−100 mM (Region Ⅲ). The adoption of a linear range for detection is favored owing to its superior calibration simplicity and enhanced accuracy relative to nonlinear ranges^[64]. The linear relationship within the 0−0.01 mM (Region Ⅰ) range can be well fitted by

2Q=329.3−5.8CFe,

with a correlation coefficient (R²) of 0.97, as shown in Fig. 2(c). And the linear relationship within the range of 10−100 mM (Region Ⅲ) can be well fitted by

3Q=149.6−1.3CFe,

with a R² of 0.96, as depicted in Fig. 2(d). The LOD of the nanosensor was then calculated via^[65]

4LOD=3σK,

where σ represents the standard deviation of five blank samples, and K is the slope of the linear range. Using this approach, LODs of 0.06 and 0.267 μM were obtained for Regions Ⅰ and Ⅲ, respectively. These values notably fall below the maximum allowed contaminant level for Fe³⁺ in drinking water mandated by the US Environmental Protection Agency at 5.37 μM (0.3 mg·L⁻¹)^[66], highlighting the superb sensitivity of the F127-dodecyl-Si QDs:Er nanosensors. Fig. 2(e) compares the LOD and the linear range of our F127-dodecyl-Si QDs:Er nanosensor to prior ratiometric Fe³⁺ sensors (Supplementary materials Table S2)^{[19, 21, 67−76]}, validating the superior performances of our Si QD nanosensors.

In addition to its ultrahigh sensitivity, we further checked the selectivity and anti-interference capabilities of our F127-dodecyl-Si QDs:Er nanosensors for Fe³⁺ detections. We employed the nanosensor to monitor various samples with different background cations including Ca²⁺, Co²⁺, Cu²⁺, K⁺, Mg²⁺, Mn²⁺, Na⁺, Sn⁴⁺, or Zn²⁺, both individually and in their mixtures with Fe³⁺ in water. Fig. 2(f) displays the results, indicating that the F127-dodecyl-Si QDs:Er nanosensor has negligible responses to cations other than Fe³⁺, as the Q value remains relatively unchanged upon the addition of other cations. Notably, the Q value only decreased dramatically with the introduction of Fe³⁺, irrespective of the presence of other cations in the solution. Furthermore, Supplementary materials Fig. S4(c) demonstrates that the F127-dodecyl-Si QDs:Er nanosensor maintains its performance robustly against various biological analytes. The obtained results highlight the outstanding selectivity and anti-interference capabilities of our F127-dodecyl-Si QDs:Er nanosensors. We attribute these properties to the rapid chelation between Fe³⁺ and the oxygen functional groups in F127-dodecyl-Si QDs:Er^{[77, 78]}.

Moreover, the recyclability of the F127-dodecyl-Si QDs:Er nanosensor was evaluated by alternatingly exposing the QDs to 20 mM Fe³⁺ and 20 mM L-cysteine solutions. The fluorescence of Si QDs undergoes rapid quenching upon the introduction of Fe³⁺ and is subsequently restored to its initial value after the addition of L-cysteine, as shown in Supplementary materials Fig. S4(d). The reversibility of the quenching-recovery mechanism is due to the strong reducibility of L-cysteine, which contains three active functional groups (−COOH, −NH₂, and −SH). These groups strongly chelate Fe³⁺, effectively liberating the Si QDs and restoring their PL intensity^{[79, 80]}. This reversible quenching-recovery cycle could be repeated at least 11 times. This cycle number surpasses that of the majority of extant Fe³⁺ nanosensors^[81−86], thereby evidencing the excellent recyclability of the F127-dodecyl-Si QDs:Er nanosensor. Such renewable detection capabilities are imperative for sustainable environmental monitoring and protection, contributing to the advancement of sustainable development initiatives^[87].

To elucidate the underlying fluorescence quenching mechanism, we carried out FTIR, time-resolved PL (TRPL), and XPS measurements on F127-dodecyl-Si QDs:Er without and with Fe³⁺. Supplementary materials Fig. S5(a) presents the FTIR spectra for F127-dodecyl-Si QDs:Er in the presence of varying concentrations of Fe³⁺. It is found that the characteristic −OH peak at 3300−3600 cm⁻¹ broadens and redshifts with rising Fe³⁺ concentrations, indicating the coordination of Fe³⁺ with −OH moieties on the micelles to form Fe-related complex chelates^{[2, 48]}. The dependence of the wavenumber of the −OH peak on the Fe³⁺ concentration is plotted in Fig. 3(a). The wavenumber of the −OH peak manifests a pronounced initial reduction alongside rising Fe³⁺ concentrations in Region Ⅰ (0−0.01 mM), ascribed to the abundant −OH binding sites that could interact with Fe³⁺. When the Fe³⁺ concentration increases to Region Ⅱ (0.01−10 mM), the decline of the wavenumber is retarded and gradually saturates, as the most −OH binding sites have been occupied. Then a plateau emerges at the further higher concentration of Region Ⅲ (10−100 mM), given that no more −OH binding sites are available for Fe³⁺ chelation. The TRPL results also reveal a Fe³⁺-concentration-dependent three-region feature (Fig. S5(b), Supplementary materials, and Fig. 3(b)). Notably, in Region Ⅰ, the PL lifetime of the F127-dodecyl-Si QDs:Er shows minimal variation despite increasing concentrations of Fe³⁺ (Fig. 3(b)). This stability suggests minimal charge transfer between the Si QDs and Fe³⁺, which could otherwise quench the PL from the Si QDs^[88]. However, the PL lifetime starts to decrease as the Fe³⁺ concentration is increased to Region Ⅱ and rapidly drops in Region Ⅲ. The above findings suggest that the fluorescence quenching in different Fe³⁺ concentration regions is attributed to different mechanisms.

In Region Ⅰ, the fluorescence quenching does not arise from the charge transfer between Si QDs and Fe³⁺. Rather, it results from the chelation of Fe³⁺ with the −OH functional groups on F127-dodecyl-Si QDs:Er. This interaction leads to the formation of a nonfluorescent ground-state complex, a phenomenon known as static quenching^{[48, 88, 89]}. The static quenching process in Region Ⅰ was further evidenced by XPS results, as shown in Fig. 3(c). The O 1s spectrum for the pristine F127-dodecyl-Si QDs:Er exhibits a peak at 534.2 eV, associated with C−O. Upon Fe³⁺ additions, the C−O peak redshifts to 533.8 eV, accompanied by the appearance of an O−Fe signal. This result further proves that the Fe³⁺ ions coordinate with −OH moieties on the micelle surface. Additionally, Fe 2p XPS signals that emerged at 724.5 eV (Fe 2p_1/2) and 711.0 eV (Fe 2p_3/2) following the introduction of Fe³⁺ into the F127-dodecyl-Si QDs:Er solution (Fig. 3(d)), exhibit redshifts relative to those in FeCl₃ (Fe 2p_1/2: 725.0 eV; Fe 2p_3/2: 711.5 eV). This change in binding energy corroborates an altered chemical environment for Fe^3+[87]. Taken together, the FTIR and XPS data indicate static quenching via ground state complex formation that contributes to the fluorescence quenching upon Fe³⁺ addition^[88].

In Region Ⅱ, the reduction in uncoordinated −OH groups in F127-dodecyl-Si QDs:Er initiated a dynamic quenching process involving charge transfers between Si QDs and Fe³⁺. This process eventually became the dominant quenching mechanism in Region Ⅲ^[90]. Fig. 3(e) illustrates the Fe³⁺-concentration-dependent quenching mechanisms in our ratiometric NIR fluorescent Fe³⁺ nanosensor. In the low concentration range (Region Ⅰ), static quenching is dominant due to Fe³⁺ coordination to the −OH binding sites of F127-dodecyl-Si QDs:Er micelles (①). As the Fe³⁺ concentration increases to the intermediate range (Region Ⅱ), dynamic quenching by electron transfer becomes increasingly competitive (②). Concurrently, the static pathway progresses toward saturation as binding sites fill, while incremental electron transfer from the conducting band of the photoexcited QDs to empty d-orbitals on the Fe³⁺ enhances nonradiative relaxation^{[91, 92]}. Finally, at higher Fe³⁺ concentrations (Region Ⅲ), the static quenching sites reach saturation, leaving electron transfer as the primary quenching mechanism. This interplay between the static and dynamic quenching pathways rationalizes the unique ratiometric sensing behavior across the three concentration regions.

Chemical-to-fluorescent signal conversion enables intriguing molecular logic devices and reprogrammable memories^{[93, 94]}. Here, a sequential logic circuit with memory function was constructed using the titration controller (Δc) and a predefined concentration threshold (c_T) as the inputs, and the Q as the output. We define 'logical 0' as Δc = 0 μM representing the titration controller being switched off, along with the c_T input in the c_T < 5 μM range. 'Logical 1' occurs when the titration controller is switched on (Δc > 0 μM) and c_T = 5 μM. Additionally, we establish a fluorescence ratio threshold (Q_c = 297.9) to determine the logical output, where 'logical 1' corresponds to Q < Q_c and 'logical 0' to Q ≥ Q_c. This setup allows for eight distinct two-digit input combinations, as depicted in Fig. 4(a). In this system, when both inputs are at 'logical 0' (i.e., Δc = 0 μM and c_T < 5 μM), the output remains at 'logical 0' (i.e., Q > Q_c). The output is also 'logical 0' (i.e., Q > Q_c) regardless of the sequence in which the titration controller is switched on (Δc > 0 μM) while maintaining the second input at c_T < 5 μM. Conversely, when the Fe³⁺ concentration is set to c_T = 5 μM and there is no titration controller active (and vice versa), the output is still 'logical 0' (i.e., Q = Q_c). When the titration controller is switched on, and then the Fe³⁺ concentration is set to c_T = 5 μM, the output is 'logical 0' (i.e., Q = Q_c). The sequential logic circuit is only activated ('logical 1') by first setting the system to c_T = 5 μM, followed by engaging the titration controller (Δc > 0 μM) that results in a further Fe³⁺ concentration increment, which decreases the Q below the threshold of Q_c (i.e., Q < Q_c). The corresponding truth tables are provided in Fig. 4(b). Uniquely, this system relies solely on Fe³⁺ unlike existing sequential logic circuits that rely on multiple metal ions^{[95, 96]}, preventing interference and improving reliability. This sequence-dependent behavior is visualized through a crossword puzzle analogy (Fig. 4(c)). Inputs Δc and c_T are designated as characters "A" and "C", respectively. Adding "A" first generated the string "ACT", representing insufficient quenching (Q ≥ Q_c, OFF). Reversed input order produced "CAR" due to adequate quenching (Q < Q_c, ON). Varying the sequence produced distinct outputs, verifying the memory capability.

In summary, an amphiphilic polymer encapsulation strategy was implemented to synthesize water-dispersible Si QDs:Er exhibiting dual NIR emission at 810 and 1540 nm. The aqueous nanosensors enabled ratiometric fluorescent detection of Fe³⁺ with an exceptional 0.06 μM LOD, outperforming prior sensors. Systematic analyses revealed a pH-dependent response and recyclable operation via reversible quenching interactions. Furthermore, static and dynamic quenching mechanisms were proven through detailed FTIR, XPS, and time-resolved PL analyses. Moreover, integration into a sequential logic circuit demonstrated programmable memory functioning. Overall, this work establishes a versatile Si QDs:Er platform to realize multifunctional ratiometric NIR nanosensors through deliberate compositional engineering. The sensitive Fe³⁺ quantification and programmable photonics open intriguing possibilities for responsive diagnostic and therapeutic platforms.