A novel fluorescent sensor based on CdTe quantum dots and DTAB‐modified Au nanoparticles through aggregation‐induced fluorescence recovery for the detection of l‐lysine

Abstract

BACKGROUND

l‐Lysine (l‐Lys), an essential amino acid, serves as a core indicator for evaluating food nutritional quality. Although naturally abundant in ingredients such as soybeans, its content must be precisely regulated during processing to meet nutritional standards. Moreover, improper processing and storage can induce l‐Lys degradation or conversion, compromising both nutritional value and food safety. Given its pivotal role in nutritional balance and quality assurance, there is an urgent need for simple, rapid, and highly sensitive analytical methods for l‐Lys detection.

RESULTS

An innovative ‘off–on’ fluorescent sensor tailored for food‐related applications was developed, utilizing CdTe quantum dots as the fluorophore and dodecyl trimethyl ammonium bromide self‐assembled citrate‐capped gold nanoparticles (DTAB‐AuNPs) as the quencher for the rapid detection of l‐lys. During the detection process, l‐lys served as a binding bridge, facilitating the aggregation of DTAB‐AuNPs. This reduction inhibited fluorescence resonance energy transfer between the fluorophore and the quencher, leading to fluorescence recovery. The degree of aggregation‐induced fluorescence recovery exhibited a linear correlation with the concentration of l‐lys within the range of 0.5–10 μm (R = 0.999). This newly developed sensing method demonstrated an impressive limit of detection of 3.22 nm. It was successfully applied to detect l‐lys in various food‐relevant samples, including milk, soybean, plasma and compound amino acid oral solution used in food‐related nutritional supplements.

CONCLUSION

The results highlight its great potential for practical applications in multiple fields such as quality control of dairy and agricultural products, as well as clinical diagnosis related to nutritional status assessment. © 2026 Society of Chemical Industry.

INTRODUCTION

Lysine is an indispensable amino acid pivotal to biological processes linked to food quality and human health. In human physiology, it regulates hormone synthesis, immune function, gene expression and material metabolism, which are all critical for maintaining physiological homeostasis. ¹ , ² , ³ , ⁴ Based on its stereochemical properties, lysine exists as two enantiomers – l‐lysine (l‐lys) and d‐lysine (d‐lys) – and one racemate (d/l‐lysine, a 1:1 mixture of the two enantiomers). Among these, l‐lys is the predominant isomer found in both biological samples and food matrices because it is the only form that can be efficiently absorbed and metabolized by organisms. This unique property underscores its critical role in protein synthesis, where it is often regarded as the ‘first limiting’ amino acid because of its indispensable nature in facilitating the utilization of other amino acids.

In the context of food science and nutrition, l‐lys holds profound implications for food quality and safety. As a nutritionally essential amino acid, l‐lys cannot be synthesized de novo by the human body in sufficient quantities, necessitating its intake through dietary sources. Foods rich in l‐lys, such as meat, fish, dairy products and certain legumes, are integral components of a balanced diet. A deficiency of l‐lys in the diet can disrupt protein metabolism, leading to a spectrum of clinical manifestations, including fatigue, muscle wasting, and impaired growth and development, particularly in pediatric populations. ⁵ , ⁶ , ⁷ Conversely, excessive intake of l‐lys, often resulting from over‐supplementation or consumption of highly fortified foods, may contribute to metabolic imbalances, such as obesity and related disorders, posing potential health risks. The detection and quantification of l‐lys in food matrices are of paramount importance for several reasons. First, monitoring l‐lys levels in food products is essential for ensuring nutritional adequacy and preventing deficiencies, which are critical for maintaining optimal health and preventing diet‐related diseases. Second, accurate determination of l‐lys content is vital for quality control in the food industry, particularly in the production of functional foods and dietary supplements, where precise labeling of nutritional information is mandated by regulatory authorities. This ensures that consumers receive accurate information about the nutritional value of the products they consume. Third understanding the distribution and concentration of l‐lys in different food matrices can facilitate the optimization of food formulations, enhancing the nutritional profile of food products and promoting healthier dietary choices. Moreover, the ability to detect l‐lys with high sensitivity and specificity has broader implications for food safety and public health. Abnormal fluctuations in l‐lys levels can serve as biomarkers for various metabolic disorders, neurological conditions and cardiovascular diseases. ⁸ , ⁹ Early detection of these abnormalities through reliable analytical methods can enable timely interventions, thereby mitigating the risk of disease progression and improving patient outcomes.

Various technologies have been developed for l‐lys detection, including electrochemical methods, ¹⁰ , ¹¹ HPLC, ¹² , ¹³ , ¹⁴ Fourier transform infrared (FTIR), ¹⁵ mass spectrometry ¹⁶ and colorimetry combined with spectroscopy. ¹⁷ They offer advantages such as high accuracy and sensitivity, but issues such as high cost, complex operating procedures, inconvenient portability and lengthy testing times limit their practical application in detection. By contrast, fluorescence sensors have become a hotspot in the development of new amino acid detection technologies because of their ease of operation, rapid response and high sensitivity. ¹⁸ , ¹⁹ , ²⁰ Recent research has witnessed significant progress in the development of nanofluorescent sensing probes for lysine detection, leveraging the unique optical properties of various nanomaterials. Initial studies focused on the design of metal–organic framework (MOF)‐based sensors, such as Eu‐MOF@CQDs, ²¹ which demonstrate high sensitivity and selectivity through ratiometric luminescence but suffered from complex synthesis and low water solubility. Subsequently, polyethyleneimine‐based nanosensors (PEINAC) ²² were developed, offering a simple and cost‐effective approach for l‐lys quantification in biological samples. however, matrix interference may occur in complex biological samples The advent of carbon dots (CDs) and carbonized polymer dots (CPDs) further expanded the repertoire of nanofluorescent probes, enabling simultaneous detection of l‐lys and pH with exceptional biocompatibility and imaging capabilities, as seen in ratiometric fluorescent CDs ²³ and dual‐emission CPDs. ²⁴ but they are primarily tailored for cell/zebrafish imaging rather than food sample analysis. Moreover, the integration of covalent organic frameworks (COFs) with lanthanide ions, such as Eu³⁺‐modified COFs, ²⁵ provided a new avenue for chiral recognition of l‐lys without the need for chiral modification. Additionally, innovative approaches using achiral UiO‐66‐NH₂ MOFs ²⁶ and portable smartphone platforms based on chiral fluorescent carbon dots ²⁷ have also demonstrated promising results in enantioselective detection and on‐site monitoring of l‐lys. Fluorescein‐modified polymer dots (F‐PDs) ²⁸ further enriched the toolbox for chiral discrimination of lysine enantiomers. These advancements collectively underscore the growing potential of nanofluorescent sensing probes, yet critical gaps remain – especially the lack of a sensor that integrates simplicity, high selectivity, and broad applicability to food‐related samples. Therefore, it is highly necessary to develop a novel fluorescent sensor that enables simple, rapid, and selective detection of l‐lys, with tailored suitability for food field applications.

In the present study, we introduce a novel approach leveraging a composite probe composed of two simple nanoparticles for the rapid and sensitive detection of l‐lys. By integrating CdTe quantum dots (CdTe QDs) as the fluorophore and dodecyl trimethylammonium bromide assembled citrate‐capped gold nanoparticles (DTAB‐AuNPs) as the quencher, we designed a fluorescent ‘turn‐off–on’ sensor. This innovative sensor exploits the unique ability of l‐lys to specifically bind to DTAB‐AuNPs, causing their aggregation and subsequently restoring the fluorescence of CdTe QDs that was initially quenched. The design of this sensor not only simplifies the detection process, but also enhances its sensitivity and selectivity. Moreover, its broad practical applicability has been demonstrated through successful detection of l‐lys in diverse food‐related samples such as milk, soybeans, plasma, and compound amino acid oral solution. This research paves the way for a convenient and efficient tool for quality control in the food industry, nutritional analysis, and monitoring of l‐lys levels in biological samples, thereby contributing to the advancement of food science and technology.

METHODS

Chemicals and reagents

Trisodium citrate dihydrate (Na₃C₆H₅O₇·2H₂O), sodium borohydride (NaBH₄), sodium tellurite (Na₂TeO₃), cadmium chloride (CdCl₂) and NaOH were purchased from Shanghai Source Biological Technology Co. (Shanghai, China); d‐lysine (d‐lys), l‐lysine (l‐lys), l‐serine (Ser), l‐histidine (His), d‐leucine (Leu), l‐leucine (Leu), l‐phenylalanine (Phe), l‐proline (Pro), l‐cysteine (Cys), d‐aspartic acid (Asp), l‐threonine (Thr), l‐arginine (Arg), l‐glutamate (Glu), l‐isoleucine (Ile) and l‐methionine (Met) were obtained from Tokyo Chemical Industry Co. Ltd (Tokyo, Japan); dodecyl trimethyl ammonium bromide (DTAB) and N‐acetyl‐l‐cysteine (NAC) were purchased from Shanghai Aladdin Biochemical Technology Co., Ltd (Shanghai, China). All chemicals were of analytical grade. Deionized water with a resistivity of 18.25 MΩ was prepared using a Mol General 220a water purification system (Shanghai Mole Scientific Instrument Co., Ltd., Shanghai, China).

Apparatus

The UV‐visible spectra were recorded on a UV‐visible spectrophotometer (UV‐1800PC; Mapada Instruments Co., Ltd., Shanghai, China). Transmission electron microscopy (TEM) images of the nanoparticles were carried on an accelerating voltage of 100 kV (H‐7000FA; Hitachi, Tokyo, Japan). Fluorometric measurement was carried out using a F‐7000FL spectrophotometer (Hitachi). All measurements were carried out at room temperature.

Synthesis of DTAB‐AuNPs and CdTe quantum dots

Preparation of citrate‐capped AuNPs

Citrate‐capped AuNPs were synthesized by using the standard citrate reduction method. ²⁹ In our experiment, 2 mL of 1% HAuCl₄ solution was added into 98 mL of boiling ultra‐pure water, then 12 mL of 50 mm trisodium citrate dihydrate (Na₃C₆H₅O₇·2H₂O) was added and the mixture solution was heated for approximately 3 min under microwave high fire until the color turned wine‐red. After that, the final solution was centrifuged at 12800 × g for 10 min and stored at 4 °C. The concentration of as‐prepared citrate‐capped AuNPs was estimated to be 8.04 nm via the extinction coefficient method ³⁰ (ε = 2.76 × 10⁸ m ⁻¹  cm ⁻¹ , A ₅₂₀ = 0.222, c = A/(εb). To improve interaction with DTAB, the AuNPs at 8.04 nm were concentrated by resuspending the centrifuged pellet in 68 mL deionized water, resulting in a concentrated concentration of 11.8 nm.

Synthesis of DTAB‐AuNPs

Here, 30 mL of 0.06 m DTAB and 20 mL of 11.8 nm citrate‐capped AuNPs solutions were incubated for 30 min at 37 °C. During incubation, the solution color changes from burgundy to grayish purple, indicating the formation of DTAB‐AuNPs. After incubation, centrifuge the mixture at 8950 × g for 15 min to separate DTAB‐AuNPs from unbound free DTAB. The supernatant was discarded. This centrifugation was repeated step twice to ensure complete removal of free DTAB. Finally, the washed DTAB‐AuNPs were dissolved in deionized water and stored at 4 °C for subsequent studies. The concentration of as‐prepared DTAB‐AuNPs was estimated to be 2.04 nm.

Synthesis of water‐soluble fluorescent CdTe quantum dots

Water‐soluble NAC‐capped CdTe quantum dots were synthesized via the hydrothermal method. Typically, 0.1 g of NAC and 0.1 g of CdCl₂ were added to 40 mL of ultrapure water and stirred magnetically for 15 min. The pH of the mixture was adjusted to 9, followed by purging the solution with nitrogen gas for 20 min. Next, 0.0055 g of Na₂TeO₃ and 0.0113 g of NaBH₄ were added to the mixture. After stirring for 15 min, the reaction mixture was transferred to a 50‐mL autoclave and heated at 200 °C for 50 min. The concentration of as‐prepared CdTe QDs was estimated to be 10.56 μm based on UV‐visible spectrophotometry and the Lambert–Beer Law. Briefly, the molar extinction coefficient (ε = 1.72 × 10 ⁴  m ⁻¹  cm ⁻¹ ) was calculated via the empirical formula ε = 10 ⁴  × (d/2.2) ² (d = 2.89 nm, TEM‐measured average diameter) and the concentration was 10.56 μm (A = 0.181 at 630 nm, 1 cm path length, c = A/(εb)). ³¹ CdTe QDs and DTAB‐AuNPs were stored at 4 °C in the dark with a shelf life of 2 months; the mixed probe was freshly prepared before use and stable for 24 h under the same storage conditions.

Detection of l‐lys

For the detection of l‐lys in solution, 100 μL of l‐lys with different concentrations was added to the solution containing 100 μL of 1.056 μm CdTe quantum dots (diluted 10 times from stock) and 90 μL of DTAB‐AuNPs. The final volume was adjusted to 1 mL with 10 mm Tris‐HCl buffer (pH 7.0). All fluorescence measurements were performed under the same buffer conditions to ensure consistent pH and ionic strength.

For selectivity experiments, other amino acids (100 μL each) were added instead of l‐lys under identical buffer conditions. Fluorescence spectra were recorded at an excitation wavelength of 360 nm.

To explore the feasibility of CdTe/DTAB‐AuNPs fluorescent sensor for l‐lys assay in real samples, milk, soybean, human plasma and compound amino acid oral solution samples were collected and subjected to a series of pre‐treatment steps to remove potential interferences, followed by analysis using the developed fluorescent sensor to determine the accuracy, precision and recovery rates of l‐lys detection in these complex matrices.

RESULTS AND DISCUSSION

Characterization of the synthesized nanoparticles

The morphological properties of the as‐prepared CdTe quantum dots (QDs), gold nanoparticles (AuNPs), DTAB‐modified AuNPs (DTAB‐AuNPs) and DTAB‐AuNPs reacted with lysine (Lys) were investigated using TEM. As depicted in Fig. 1(a),(b), the CdTe QDs and AuNPs exhibited uniform dispersion and spherical shapes, having mean diameters of 2.89 nm and 12.71 nm, respectively. The average particle size of both DTAB‐AuNPs and DTAB‐AuNPs reacted with l‐lys was approximately 17 nm. However, a comparison of Fig. 1(c),(d) revealed that the DTAB‐AuNPs obviously aggregated into clusters after reacting with l‐lys. Furthermore, the optical properties of these nanoparticles were characterized using fluorescence (FL) spectrometry, UV absorption spectrometry and IR absorption spectrometry. As shown in Fig. 1(e), the synthesized orange CdTe QDs solution exhibited distinct fluorescence, with a maximum emission wavelength of approximately 635 nm at an excitation wavelength of 360 nm. The wine‐red AuNPs solution displayed a maximum UV absorption wavelength at 520 nm. By contrast, DTAB‐AuNPs exhibited a significant redshift (λ ≈ 580 nm) and broadening compared to AuNPs (λ = 520 nm), corresponding to the formation of clustered nanostructures, and the solution color turned grayish purple. ³² , ³³ Upon addition of l‐lys, the spectrum (black line) showed a marked decrease in absorbance in the longer‐wavelength region (≈ 580–700 nm), consistent with further aggregation and structural reorganization of the DTAB‐AuNP clusters. Moreover, the FTIR spectra of CdTe, CdTe/ DTAB‐AuNPs, CdTe QDs/DTAB‐AuNPs + Lys in Fig. 1(f) revealed that the surface of the quantum dots modified with NAC contained a large number of hydroxyl, carboxyl and amino groups. The peak at 3364 cm⁻¹ was mainly attributed to the stretching vibration of the ‐NH or ‐OH bond, and the peak at 1520 cm⁻¹ was assigned to N—H bending vibration (amide II band). After adding DTAB‐AuNPs, as a result of electrostatic interaction, the stretching vibration of ‐NH/‐OH shifted to 3620 cm⁻¹ (blueshift) and the C=O stretching vibration (1631 cm⁻¹, amide I band) showed weakened intensity. When l‐lys was introduced, the ‐NH/‐OH stretching vibration further blueshifted and the C=O vibration shifted to 1618 cm⁻¹, suggesting possible interactions between l‐lys functional groups and the sensor. As shown in Fig. 1(g), the X‐ray diffraction pattern of DTAB‐AuNPs exhibited four prominent peaks at 2θ values of 38.205°, 44.239°, 64.556° and 77.568°, corresponding to the (111), (200), (220), and (311) lattice planes of gold. The characterization results demonstrated that the two types of nanoparticles were successfully synthesized, and the sensor formed by the mixture of the two exhibited a response to l‐lys.

Figure 1.

(a–d) TEM images of CdTe, AuNPs, DTAB‐AuNPs, DTAB‐AuNPs + Lys. (e) The UV absorption spectrum of synthesized AuNPs (green line), DTAB‐AuNPs (red line), DTAB‐AuNPs + Lys (black line) and the FL emission spectrum of the CdTe quantum dots (blue line). (f) FTIR spectra of CdTe, CdTe/DTAB‐AuNPs and CdTe/DTAB‐AuNPs + Lys. (g) the X‐ray diffraction image of DTAB‐AuNPs.

Discussion of the detection mechanism

The primary mechanism underlying the sensor's ‘off’ response can be attributed to fluorescence resonance energy transfer (FRET) occurring between CdTe quantum dots (donor) and DTAB‐AuNPs (acceptor). This conclusion is supported by three key experimental evidences: First, as shown in Fig. 1(e), significant overlap exists between the emission spectrum of CdTe quantum dots and the absorption spectrum of DTAB‐AuNPs, a necessary prerequisite for FRET occurrence. Second, upon formation of the CdTe/DTAB‐AuNPs complex, the average fluorescence lifetime of CdTe quantum dots shortens markedly from 68.99 ns to 37.89 ns (Fig. 2(c)). This lifetime reduction is a hallmark of FRET‐induced dynamic quenching, distinguishing it from static quenching that typically does not affect excited‐state lifetime; Finally, the reversible ‘switch‐on’ response induced by l‐lys further supports the FRET mechanism, as quenching processes based on permanent quenching or internal filtering effects typically cannot achieve such highly efficient reversible recovery through simple analyte addition. Having established FRET as the dominant quenching pathway, we proceeded to investigate the detailed mechanism. To elucidate the response mechanism of the CdTe QDs/DTAB‐AuNPs sensor towards l‐lys, we first investigated the ‘turn‐off’ process of the fluorescent sensor. From Fig. 2(a), it was evident that the fluorescence of QDs could be significantly quenched by DTAB‐modified AuNPs rather than AuNPs. The possible reasons can be explained by comparing the particle sizes, the differences in UV absorption spectra and the overlap of UV absorption spectra with the fluorescence spectra of CdTe QDs, as shown in Fig. 1(e). When only citric trisodium was used to modify AuNPs, the particle size increased and the maximum UV absorption peak was at approximately 520 nm, which had little overlap with the fluorescence emission peak of the fluorophore CdTe QDs. When DTAB was further used to modify the synthesized citrate‐capped AuNPs, the particle size was further enlarged. Because DTAB could form a hydrophobic layer that encapsulated a few citrate‐capped AuNPs together to form larger particles, the UV absorption peak was redshifted and significantly overlapped with the fluorescence emission peak of the CdTe QDs. This suggests that FRET or inner‐filter effect (IFE) may occur between the fluorophore CdTe QDs and DTAB‐AuNPs, leading to fluorescence ‘turn‐off’. ³⁴ Zeta potential measurements were performed in 10 mm Tris–HCl buffer (pH 7.0) at room temperature using nanoparticle dispersions with concentrations of approximately 0.1 μm for CdTe QDs and 0.2 nm for DTAB‐AuNPs. Zeta potential studies (Fig. 2(b)) showed that CdTe QDs had a negative zeta potential because their surface was modified with NAC, which contains a large number of negatively charged carboxyl groups. DTAB‐AuNPs had a positive zeta potential. The two nanoparticles could be attracted to each other by electrostatic interaction, getting close enough to create a FRET effect and eventually forming the CdTe QDs/DTAB‐AuNPs sensor, which exhibited a positive zeta potential. Fluorescence lifetime tests were also used to verify the reaction mechanism, and the results are shown in Fig. 2(c). The mean fluorescence lifetimes of CdTe and CdTe‐DTAB‐AuNPs were calculated to be 68.99 ns and 37.89 ns, respectively. This result implied that the fluorescence lifetime of CdTe decayed after reaction with DTAB‐AuNPs, indicating that the mechanism of fluorescence quenching was attributed to FRET rather than IFE.

Figure 2.

(a) The FL spectra of CdTe QDs, CdTe/DTAB‐AuNPs sensor and CdTe/DTAB‐AuNPs + l‐lys. (b) The zeta potential of CdTe QDs, CdTe/DTAB‐AuNPs sensor and CdTe/DTAB‐AuNPs + l‐lys. (c) The fluorescence lifetime test of CdTe QDs and CdTe/DTAB‐AuNPs. (d) The FL intensities of CdTe QDs reacted with AuNPs, DTAB, DTAB‐AuNPs and l‐lys. (e) The FL intensities of CdTe/DTAB‐AuNPs sensor after reacting with l‐lys, d‐lys, d/l‐leucine, 1,5‐diaminopentane and 6‐aminocaproic acid.

In the second step, the ‘turn‐on’ process of the sensor by l‐lys was investigated. After the CdTe QDs and DTAB‐AuNPs constituted the ‘turn‐off’ sensor, a significant recovery of fluorescence was observed when l‐lys was added to the sensor solution. The addition of l‐lys promoted the aggregation of DTAB‐AuNPs, as directly observed by TEM (Fig. 1(d)). This aggregation was accompanied by a distinct reduction in the longer‐wavelength absorption of DTAB‐AuNPs (Fig. 1(e), black line). Consequently, the spectral overlap between the absorption of the aggregated DTAB‐AuNPs and the emission of CdTe QDs decreased, which suppressed the FRET efficiency and led to fluorescence recovery. With the addition of l‐lys, a large number of functional groups (amino, carboxyl, etc.) in the l‐lys structure formed many hydrogen bonds between amino acids and continuously bound to the surrounding DTAB‐AuNPs, resulting in an obvious aggregation behavior of DTAB‐AuNPs. Simultaneously, CdTe QDs could be observed to be re‐dispersed into the solution. As a result, the overlap between the absorption peak and the fluorescence emission peak of CdTe QDs became smaller, which suppressed the FRET and eventually led to fluorescence ‘turn‐on’.

Furthermore, as shown in Fig. 2(d), DTAB could also quench the fluorescence of CdTe QDs, but AuNPs could not. However, Lys could not recover the fluorescence quenched by DTAB. This suggested that the fluorescence quenching mechanism between DTAB and CdTe QDs was different from that between DTAB‐AuNPs. When DTAB was directly mixed with CdTe QDs, the positively charged DTAB would fully combine with CdTe QDs, neutralize the negative charge on the surface of CdTe QDs, reduce the electrostatic repulsion between CdTe QDs particles and produce an aggregation‐induced fluorescence quenching effect. The added l‐lys had no way to fully compete DTAB back from CdTe QDs. When DTAB was modified on the surface of AuNPs, although its positive charge would also quench the fluorescence of CdTe QDs, the binding between DTAB and CdTe QDs would definitely be blocked by larger gold nanoparticles. At this time, it was possible for l‐lys to compete DTAB from the CdTe surface to produce a fluorescence recovery effect. ³⁵ , ³⁶ , ³⁷

However, unlike L‐lys, arginine had a weak effect on the fluorescent sensor when the pH was lower than its isoelectric point (pI) of 9.91. This indicated that l‐lys‐induced aggregation of DTAB‐AuNPs was not only a result of simple electrostatic attraction, but also its specific amino and carboxyl structure may play a key role. To analyze whether the selective detection of l‐lys by the CdTe QDs/DTAB‐AuNPs sensor was derived from the amino or carboxyl group of l‐lys, the fluorescent sensor was tested for its fluorescence response to 1,5‐diaminopentane (lacking a carboxyl group), 6‐ aminocaproic acid (lacking an α‐NH₂ group) and d/l‐leucine (lacking an ε‐NH₂ group) instead of l‐lys. As shown in Fig. 2(e), the emission peak at 635 nm showed no response to any of these structurally different chemicals, indicating that all amino and carboxyl functional groups of l‐lys play an important role in the detection process.

The mechanism of lysine‐induced DTAB‐AuNPs aggregation is speculated to be that the ε‐NH₂ of lysine may interact with reactive groups (e.g. from the NAC capping layer or residual citrate) associated with the DTAB‐AuNPs, whereas the α‐NH₂ and carboxyl groups at the other end could interact with another l‐lys molecule or the nanoparticle surface, potentially bridging multiple DTAB‐AuNPs and leading to aggregation. More interestingly, when d‐lys was used to replace l‐lys in the experiment, although d‐lys had a certain effect on fluorescence recovery, its recovery effect was less than half that of l‐lys, indicating that the spatial position of amino and carbonyl groups also contributed to the bridge connection between l‐lys and the sensor.

Briefly, at the stage of sensor formation, DTAB can wrap up a couple of citrate‐capped AuNPs together and introduce hydrophobic dodecane groups on the surface, resulting in larger particle size and enabling them to interact with CdTe QDs to generate FRET for fluorescence quenching. During the detection, l‐lys acts as a binding bridge and enables multiple DTAB‐AuNPs to aggregate, inhibiting the FRET effect with CdTe QDs and finally recovering fluorescence. According to the above research results and inferences, the detection mechanism of l‐lys based on the CdTe QDs/DTAB‐AuNPs fluorescent sensor is shown in Scheme 1.

Scheme 1.

The detection mechanism of CdTe QDs/DTAB‐AuNPs fluorescent sensor for l‐lys. This schematic illustrates two key states: (i) the formation of cadmium telluride quantum dot/DTAB‐gold nanoparticle fluorescence resonance energy transfer pairs and the resulting fluorescence quenching and (ii) the spatial separation of donor‐acceptor pairs due to l‐lys‐bridged DTAB‐gold nanoparticle aggregation, thereby enabling fluorescence recovery. This schematic integrates key findings from the aforementioned spectral, lifetime and zeta potential data.

Quantitative assay of l‐lys using the CdTe QDs/DTAB‐AuNPs fluorescent sensor

To enhance the sensitivity and accuracy of quantitative analysis, we optimized the main factors that affect the quenching and recovery of fluorescence intensity, including the volume of the quencher DTAB‐AuNPs, the reaction time of the sensor mixed with 10 μm l‐lys and the pH values of the detection system. Figure 3(a) demonstrated that the (F ₁ – F ₀)/F ₀ at 635 nm increased significantly with the increase in the DTAB‐AuNPs volume within the range of 70–90 μL. When the DTAB‐AuNPs volume exceeded 90 μL, there were obvious changes between the fluorescence intensities in the absence and presence of l‐lys. As shown in Fig. 3(b), when the reaction time increased from 10 to 15 min, the fluorescence recovery degree also increased, but further increasing the reaction time decreased the fluorescence recovery. As shown in Fig. 3(c), the (F ₁ – F ₀)/F ₀ at 635 nm increased significantly with increasing pH. It reached a maximum at pH 7.0 and then gradually decreased. Finally, 90 μL was chosen as the optimized added volume of DTAB‐AuNPs to form the sensor, 15 min was chosen as the reaction time after adding l‐lys to the sensor, and the pH of the detection system was selected to be 7.0.

Figure 3.

(a) Influence of the added volume of DTAB‐AuNPs to construct the sensor. (b) Influence of reaction time between CdTe QDs/DTAB‐AuNPs sensor and l‐lys. (c) Influence of pH values to construct the sensor.

Under the optimized experimental conditions, the CdTe QDs/DTAB‐AuNPs sensor was utilized to detect l‐lys. All following optimization and calibration data are presented as the mean ± SD from three independent experiments (n = 3). Figure 4(a) shows that the recovered fluorescence intensity responses of the sensor gradually increased with increasing concentration of l‐lys. Figure 4(b) also shows that the (F₁ ‐ F₀)/F₀ at 635 nm was highly correlated with the concentration of l‐lys and the linear range was determined to be 0.5–10 μm with a coefficient of determination (R ²) of 0.999. The limit of detection (LOD) was calculated using the formula LOD = 3σ/S, where σ represents the SD of 10 replicate measurements of the blank signal (sensor without analyte) and S denotes the slope of the calibration curve shown in Fig. 4(b). The results indicated an LOD value of 3.22 nm. Table 1 summarizes the comparative analysis with several recently reported l‐lys fluorescent sensors. This sensor exhibits an extremely low LOD of 3.22 nm, comparable to or superior than most existing fluorescent detection methods. Its linear range spans 0.5–10 μm, which, although narrower than some chromatographic or electrochemical methods, is highly suitable for trace detection in food and biological samples. Regarding selectivity, this sensor exhibits strong specificity toward l‐lys, significantly outperforming other amino acids (including d‐lys and structural analogues) and surpassing multiple existing nanosensors prone to interference from complex matrices. Furthermore, its successful application in diverse real‐world samples, including milk, soybeans, plasma and compound amino acid oral solutions, fully validates its robustness and practicality. However, certain limitations remain. The sensor relies on the synthesis and stability of two distinct nanomaterials, which may impact its reproducibility and ease of operation compared to single‐component probes or commercial kits. Its narrow linear range necessitates sample dilution procedures, Based on the same dataset, the limit of quantification (LOQ) for this method was calculated as 10.7 nm (LOQ = 10σ/S). As shown in Table 2, the relative standard deviations (RSDs) obtained from spiked recovery tests in different real samples (ranging from 1.9% to 3.8%) demonstrate the method's excellent intra‐day precision. In summary, the sensor developed in this study exhibits excellent linear response and extremely high sensitivity (LOD 3.22 nm) for l‐lys within the range 0.5–10 μm. Notably, the method's advantages lie in its exceptionally low detection limit and rapid response time, making it highly suitable for the rapid screening of trace amounts of l‐lys. However, its linear range is relatively narrow, requiring appropriate dilution for samples at higher concentrations. Furthermore, the sensor's performance depends on the synthesis and reproducibility of the nanomaterials, leaving room for improvement in operational simplicity compared to some commercially available single‐component detection kits. Future work will focus on optimizing the operational workflow through probe immobilization or reagent integration.

Figure 4.

(a) Fluorescence responses of CdTe QDs/DTAB‐AuNPs sensor with different concentrations of lysine (0–100 μm). (b) Calibration curve for l‐lys.

Table 1.

Comparison with other methods for detection of l‐lysine

Methods	Probes	Linear range (μm)	LOD (μm)	Reference
Ratiometric fluorescence	Eu‐BTB@CQDs	0–20	3.68	Yiang et al. 21
Fluorescence	PEINAC	1–1000	0.13	Tong et al. 22
Fluorescence	Carbon dots	0.1–700	1.9 × 10−2	Chang et al.23
Ratiometric fluorescence	Eu3+@TAB‐DFP‐COF	0.1–130	1.75 × 10−2	MA et al. 25
Fluorescence	Carbon dots	10–100	0.56	Wei et al. 27
Fluorescence	N‐GQDs	2–40	1.06	Li et al. 39
Fluorescence	DFC	0–37.5	0.14	Gao et al. 40
Fluorescence	CdTe quantum dots and DTAB‐AuNPs	0.5 ~ 10	3.22 × 10−3	Present study

Table 2.

Recovery results of l‐lysine in real samples

Sample	Detected/μm	Added/μm	Found/μm	Recovery/%	RSD/% (n = 3)
Cow milk	2.00	1.00	3.02	102.00	3.40
	2.00	5.00	7.19	104.00	2.70
	2.00	10.0	11.85	98.50	2.90
Soybean	5.00	1.00	5.99	99.00	3.80
	5.00	5.00	10.18	103.00	2.50
	5.00	10.0	15.6	106.00	2.10
Compound amino acid oral solution	8.30	1.00	9.35	105.00	3.30
	8.30	5.00	13.27	99.40	2.60
	8.30	10.0	18.50	103.00	2.70
Human plasma	6.25	1.00	7.27	102.00	1.90
	6.25	5.00	11.16	98.20	2.50
	6.25	10.0	16.30	103.00	2.20

Selectivity of the detection by CdTe QDs/DTAB‐AuNPs sensor

To investigate the selectivity of this detection system, we studied the response of the system to 12 types of other essential amino acids. Figure 5 shows the fluorescence response of l‐lys (100 μm) and other 12 essential amino acids (1000 μm) under optimal conditions. The l‐lys caused a significant increase in (F ₁ – F ₀)/F ₀ at 635 nm. Although the d‐Lys and arginine has a certain response, it is much lower than that produced by l‐lys. The other amino acids basically have no fluorescence recovery effect on the sensor. The results indicate that the novel sensor has good specificity for l‐lys.

Figure 5.

The selectivity of the CdTe/DTAB‐AuNPs for analyzing lysine (100 μm) over the other 12 amino acids (1000 μm).

Analysis l‐lys in real samples

To further evaluate the potential application of the proposed sensor in real samples, the sensor was used to detect l‐lys in cow milk, and soybean, compound amino acid oral solution, and human plasma. All samples were subjected to the following pretreatment steps to minimize matrix interference and maintain consistent detection conditions. Cow milk was centrifuged at 1430 × g for 10 min to remove the fat layer, after which the skimmed milk was diluted 10‐fold with 10 mm Tris‐HCl buffer (pH 7.0). Soybean powder was extracted with hot water (80 °C) for 30 min, centrifuged and the supernatant was diluted 20‐fold with the same buffer. The compound amino acid oral solution was filtered through a 0.22‐μm membrane and then diluted 10‐fold with the buffer to bring the l‐lys concentration within the linear calibration range. Considering that the concentration level of l‐lys in the blank plasma is between 150–250 μm, ³⁸ the plasma sample was diluted 30 times to fit the linear range of the established standard curve when performing the recovery experiment aminocaproic. In all cases, the final detection system was maintained at pH 7.0 with a consistent ionic strength provided by the Tris‐HCl buffer. In addition, d‐lys is rarely present in plasma, so it does not interfere with detection. Known amounts of l‐lys standard solutions were added to the blank plasma matrix and detected using the established new method.

The recovery rate data in Table 2 revealed that the CdTe QDs/DTAB‐AuNPs fluorescent sensor demonstrates good performance in detecting L‐lys across diverse real samples. In cow milk, soybean, as well as plasma containing various proteins, fats and other components, and compound amino acid oral solution with a complex composition, the sensor achieves average recovery rates ranging from 98.2% to 106% and RSDs between 1.9% and 3.8%. This indicates the sensor's high accuracy, precision, and robustness in accurately measuring l‐lys at the same time as effectively overcoming interference from complex matrices in different sample types. This indicates that the sensor has great potential for practical applications in various fields, such as monitoring of nutritional components in dairy and agricultural products, clinical diagnosis, and quality control in the pharmaceutical industry.

CONCLUSIONS

In the present study, an economical, simple and efficient fluorescent ‘off–on’ sensor was constructed for the detection of l‐lys. The sensor is a mixture of CdTe QDs and DTAB‐AuNPs, in which the CdTe QDs are fluorophore and DTAB‐AuNPs are quenching agents. l‐lys in aqueous solution acts as a bridge to connect multiple DTAB‐AuNPs because of its specific amino and carboxyl functional groups, reducing the fluorescence quenching effect and resulting in the fluorescence recovery. The CdTe QDs/DTAB‐AuNPs sensor can be used for rapid detection of l‐lys in the linear range of 0.5–10 μm with a detection limit of 3.22 nm. In addition, the method achieved the accurate quantitative detection of l‐lys in four types of real samples with recoveries in the range 98.3–106%, demonstrating the promising applicability for biosensing, quality control and nutritional monitoring in the future. Although this method demonstrates numerous advantages, certain limitations warrant attention. First, its linear detection range (0.5–10 μm) is relatively narrow, potentially requiring dilution of samples with high endogenous l‐lys concentrations. Additionally, sensor performance depends on the synthesis and reproducibility of nanomaterials, and its ease of operation still has room for improvement compared to some commercially available single‐component detection kits. Future work will focus on optimizing the operational workflow through probe immobilization or reagent integration.

CONFLICTS OF INTEREST

The authors declare that they have no conflicts of interest.

DATA AVAILABILITY STATEMENT

The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.