Fluorometric Detection of Thiamine Based on Hemoglobin–Cu3(PO4)2 Nanoflowers (NFs) with Peroxidase Mimetic Activity

Component analysis plays an important role in food production, pharmaceutics and agriculture. Nanozymes have attracted wide attention in analytical applications for their enzyme-like properties. In this work, a fluorometric method is described for the determination of thiamine (TH) (vitamin B1) based on hemoglobin–Cu3(PO4)2 nanoflowers (Hb–Cu3(PO4)2 NFs) with peroxidase-like properties. The Hb–Cu3(PO4)2 NFs catalyzed the decomposition of H2O2 into ·OH radicals in an alkaline solution that could efficiently react with nonfluorescent thiamine to fluoresce thiochrome. The fluorescence of thiochrome was further enhanced with a nonionic surfactant, Tween 80. Under optimal reaction conditions, the linear range for thiamine was from 5 × 10−8 to 5 × 10−5 mol/L. The correlation coefficient for the calibration curve and the limit of detection (LOD) were 0.9972 and 4.8 × 10−8 mol/L, respectively. The other vitamins did not bring about any obvious changes in fluorescence. The developed method based on hybrid nanoflowers is specific, pragmatically simple and sensitive, and has potential for application in thiamine detection.


Introduction
Thiamine (TH) (vitamin B 1 ) is an important water-soluble vitamin in the human body [1,2]. It is involved in various metabolic processes [3], and the phosphorylated derivative thiamine diphosphate is a transketolase for the substances necessary for the activity of mitochondrial pyruvate [4]. Humans and other mammals cannot synthesize thiamine by themselves, and instead, take in vitamin B 1 from the outside world [5]. The recommended daily intake of vitamin B 1 in adults is 0.5-1 mg [6]. A severe deficiency of vitamin B 1 can cause athlete's foot, neuralgia, optic neuropathy, anorexia and KORSAKOFF syndrome [7][8][9]. Between 3 and 5 mg/kg of vitamin B 1 are required as a food additive in rice, wheat and miscellaneous grains. The related detection methods for vitamin B 1 involve spectrophotometry [10,11], chemiluminescence [12], chromatography [13][14][15], fluorescence [7,16,17], electrochemistry [18] and capillary electrophoresis [19]. Notably, although some of them are capable of accurate or high-throughput monitoring of amounts of vitamin B 1 and have even been practically used in food quality control, these methods are restricted to expensive, low selectivity or complicated experimental conditions. Among these methods, the fluorometric method based on the oxidation of thiamine to thiochrome by potassium hexacyanoferrate (III) in alkaline solution is the most widely used. However, the thiochrome yield (about 67%) of the reaction using hexacyanoferrate as an oxidant is not high. Therefore, this method is often improved by substituting hexacyanoferrate with other materials, such as graphene quantum dot-capped gold nanoparticles [20] and oxygen

Instrument
The fluorescence measurements were performed on a Shimadzu RF-5301 PC fluorophotometer (Kyoto, Japan) and Microplate reader (BioTek, Winooski, VT, USA). A 1 cm path length quartz cuvette and 96-well microplates were used in the experiments. The widths of the excitation and emission slits of the fluorophotometer were set to 3.0 and 3.0 nm, respectively. The Fourier-transform infrared (FTIR) spectrum was recorded in the range of 400-4000 cm −1 on KBr (FTIR IRAffinity-1s, Shimadzu, Japan). The pH measurements were performed by a PHS-25 pH meter (Shanghai INESA Scientific Instrument Co. Ltd., Shanghai, China). The UV-vis spectrum was recorded in the range of 200-1100 nm on a UV-vis spectrophotometer (Puxi Inc., Beijing, China). Zeta potential was measured by photon correlation spectroscopy using a Zetasizer (Nano-ZS90). A scanning electron microscopy (SEM) image was characterized by a Zeiss Merlin at 1.0 kV. The purified water was obtained from a SMART-N Heal Force Water Purification System (Shanghai Canrex Analytic Instrument Co., Ltd., Pudong, Shanghai, China). The nanoparticle surface charge was determined with a Malvern Zetasizer. The P element was determined by an inductively coupled plasma mass spectrometer (ELAN DRC-e, PerkinElmer, Concord, Ontario, Canada).

Detection of Reactive Hydroxyl Radical (·OH) Production
The hydroxyl radical (·OH) production was measured using a fluorescence method. Terephthalic acid (TA) was used as a fluorescence probe for detection of ·OH in H 2 O 2 . First, 75 µL of 25 mM TA in the NaOH (pH 13) solution was added into the 3 mL of PBS (pH 7.4) containing 100 mM H 2 O 2 and 2 mg/mL Hb-Cu 3 (PO 4 ) 2 NFs. After 1 h incubation in the dark, the solution was examined for fluorometric measurements. Fluorescence spectra were obtained with an excitation wavelength of 315 nm and the emission spectra were recorded at a wavelength of 425 nm.

Enzyme-like Activity and Kinetic Parameter of Hb-Cu 3 (PO 4 ) 2 NFs
The enzyme-like activity of freshly synthesized Hb-Cu 3 (PO 4 ) 2 NFs was investigated spectrophotometrically by measuring the formation of DAP (2,3-diaminophenazine) from OPD at 450 nm (ε = 21,000 M −1 × cm −1 ) using a multiwell plate reader. Typically, 10 µL OPD (2 mM) was added into different 170 µL buffer solutions, followed by 10 µL Hb-Cu 3 (PO 4 ) 2 NFs (2 mg/mL) and 10 µL H 2 O 2 (100 mM) at 37 • C. The activity of the Hb-Cu 3 (PO 4 ) 2 NFs at different pH values (pH 3-13) was studied using similar conditions as those described above. The pH of the different solutions was measured using a pH meter. Additionally, the buffers with pH values from 9 to 13 were prepared from the mixture of 0.2 M disodium hydrogen phosphate and a sufficient quantity of 1 M sodium hydroxide. pH values from 5 to 8 were obtained from the mixture of 0.2 M disodium hydrogen phosphate and 0.2 M sodium dihydrogen phosphate. Solutions with pH values of 3 and 4 were obtained from the mixture of 0.1 M citric acid and sodium citrate.
The steady-state kinetics of Hb-Cu 3 (PO 4 ) 2 NFs were performed by varying the concentrations of H 2 O 2 (0-0.1 M) or OPD (0-2 mM) one at a time. The reaction was carried out in 160 µL phosphate buffers (pH 10) and monitored spectrophotometrically every 300 s using a multiwell plate reader. The kinetic curves were adjusted to the Michaelis-Menten model using the Origin (version 8.0, OriginLAB Corporation, Boston, MA, USA) software. The apparent kinetic parameters, Michaelis-Menten constant (K m ) and maximum rate of reaction (V max ) were calculated.

Fluorescent Detection of Thiamine
The quantitative determination of thiamine using a Hb-Cu 3 (PO 4 ) 2 NF-catalyzed fluorescent assay in the presence of H 2 O 2 was performed as follows. A 160 µL Na 2 HPO 4 -NaOH buffer (pH 10) with 0.4% Tween 80, 10 µL of 100 mM H 2 O 2 , 20 µL of different concentrations of thiamine (0.0001-10 mM) and 10 µL of 2 mg/mL Hb-Cu 3 (PO 4 ) 2 NFs were mixed at room temperature, sequentially. Then, the mixtures were placed in a constant temperature incubator at 37 • C for 5 min. The fluorescence values of the above-mentioned solutions were measured using a microplate reader with an excitation wavelength of 370 nm. The emission spectra of these reaction solutions were recorded at a wavelength of 441 nm.

Characterization of Hb-Cu 3 (PO 4 ) 2 NFs
The Hb-Cu 3 (PO 4 ) 2 NFs were characterized using IR, UV-vis, SEM and Zetasizer methods. The IR spectra of CuSO 4 , hemoglobin and Hb-Cu 3 (PO 4 ) 2 NFs are shown in Figure 1a. The main characteristic vibration peaks of Hb were found at 1134, 1317 and 1679 cm −1 and attributed to v(C-N) and v(N-H), respectively. The peak found at 3306 cm −1 was assigned to the O-H of water. The spectrum of Hb-Cu 3 (PO 4 ) 2 NFs maintained the characteristic peaks of Hb at 1049, 1652 and 1734 cm −1 assigned to νas(N-H) and νas(C-N). The bands at 1041 and 1162 cm −1 were attributed to P-O and P=O vibrations. The existence of a phosphate element was also proved using an inductively coupled plasma mass spectrometer. The result was similar to that found in the literature [24]. As shown in Figure 1b, the UV-vis spectrum of Hb exhibited an absorbance at 254 nm. Cu 2+ exhibited an absorbance at 800 nm. The characteristic peaks of Cu 2+ and Hb were observed in the spectrum of the Hb-Cu 3 (PO 4 ) 2 NFs. The morphological structure of Hb-Cu 3 (PO 4 ) 2 NFs obtained by SEM is shown in Figure 1c. The SEM images indicate that Hb-Cu 3 (PO 4 ) 2 NFs have a flower-like structure with a size of about 15 µm. The surface zeta potential measurements show that the ζ of CuSO 4 , hemoglobin and Hb-Cu 3 (PO 4 ) 2 NFs were approximately −2.1 mV, 10.7 mV and −4.5 mV, respectively, as shown in Figure 1d. The ζ and the SEM consistently proved that the Hb-Cu 3 (PO 4 ) 2 NFs were successfully synthesized.

Characterization of Hb-Cu3(PO4)2 NFs
The Hb-Cu3(PO4)2 NFs were characterized using IR, UV-vis, SEM and Zetasizer methods. The IR spectra of CuSO4, hemoglobin and Hb-Cu3(PO4)2 NFs are shown in Figure 1a. The main characteristic vibration peaks of Hb were found at 1134, 1317 and 1679 cm −1 and attributed to v(C-N) and v(N-H), respectively. The peak found at 3306 cm −1 was assigned to the O-H of water. The spectrum of Hb-Cu3(PO4)2 NFs maintained the characteristic peaks of Hb at 1049, 1652 and 1734 cm −1 assigned to νas(N-H) and νas(C-N). The bands at 1041 and 1162 cm −1 were attributed to P-O and P=O vibrations. The existence of a phosphate element was also proved using an inductively coupled plasma mass spectrometer. The result was similar to that found in the literature [24]. As shown in Figure 1b, the UV-vis spectrum of Hb exhibited an absorbance at 254 nm. Cu 2+ exhibited an absorbance at 800 nm. The characteristic peaks of Cu 2+ and Hb were observed in the spectrum of the Hb-Cu3(PO4)2 NFs. The morphological structure of Hb-Cu3(PO4)2 NFs obtained by SEM is shown in Figure 1c. The SEM images indicate that Hb-Cu3(PO4)2 NFs have a flower-like structure with a size of about 15μm. The surface zeta potential measurements show that the ζ of CuSO4, hemoglobin and Hb-Cu3(PO4)2 NFs were approximately -2.1 mV, 10.7 mV and -4.5 mV, respectively, as shown in Figure 1d. The ζ and the SEM consistently proved that the Hb-Cu3(PO4)2 NFs were successfully synthesized.

Enzyme-Like Activities and Kinetic Parameters of Hb-Cu3(PO4)2 NFs
In order to clarify the mechanism of Hb-Cu3(PO4)2 NFs, terephthalic acid (TA) was used as a fluorescence probe for the tracking of ·OH because it can capture ·OH and generate a fluorescent product, 2-hydroxyterephthalic acid (HTA), at 425 nm. As shown in Figure S1, the control groups, TA, TA with H2O2 and TA with Hb-Cu3(PO4)2 NFs, did not show significant intensity for HTA. Only in the presence of Hb-Cu3(PO4)2 NFs and H2O2 could fluorescence be found. This result supports the hypothesis that Hb-Cu3(PO4)2 NFs can catalyze H2O2 to generate ·OH, thus demonstrating the peroxidase-like activities of Hb-Cu3(PO4)2 NFs. Therefore, eight systems were established to prove that Hb-Cu3(PO4)2 NFs are peroxidases. Figure 2 clearly shows that fluorescence intensity increased with the presence of Hb-Cu3(PO4)2 NFs and H2O2. No fluorescence intensity was observed in the absence of H2O2 or Hb-Cu3(PO4)2 NFs. In addition, an equal amount of Hb or copper ion (1 mg/mL) in the presence of both TH and H2O2 showed weak fluorescence ( Figure S2). These results indicate

Enzyme-Like Activities and Kinetic Parameters of Hb-Cu 3 (PO 4 ) 2 NFs
In order to clarify the mechanism of Hb-Cu 3 (PO 4 ) 2 NFs, terephthalic acid (TA) was used as a fluorescence probe for the tracking of ·OH because it can capture ·OH and generate a fluorescent product, 2-hydroxyterephthalic acid (HTA), at 425 nm. As shown in Figure S1, the control groups, TA, TA with H 2 O 2 and TA with Hb-Cu 3 (PO 4 ) 2 NFs, did not show significant intensity for HTA. Only in the presence of Hb-Cu 3 (PO 4 ) 2 NFs and H 2 O 2 could fluorescence be found. This result supports the hypothesis that Hb-Cu 3 (PO 4 ) 2 NFs can catalyze H 2 O 2 to generate ·OH, thus demonstrating the peroxidase-like activities of Hb-Cu 3 (PO 4 ) 2 NFs. Therefore, eight systems were established to prove that Hb-Cu 3 (PO 4 ) 2 NFs are peroxidases. that Hb-Cu3(PO4)2 NFs can catalytically activate H2O2 decomposition to generate·OH radicals and lead to more efficient oxidation of thiamine. According to a previous report [24], Hb-Cu3(PO4)2 NFs showed peroxidase activity under acidic pH conditions. However, the above fluorescence experiment results indicate that Hb-Cu3(PO4)2 NFs were highly active even at pH 10. Therefore, the effect of pH on the enzyme activity of Hb-Cu3(PO4)2 NFs was throughout measured by varying the pH and keeping the OPD and/or H2O2 concentration constant. As shown in Figure S3, the peroxidase-like activity of Hb-Cu3(PO4)2 NFs was present at pH 3 to 11. Interestingly, as shown in Figure S4, it was found that Hb-Cu3(PO4)2 NFs could catalyze oxidase reactions with OPD in the absence of H2O2 and showed oxidase-like activity at pH 12 and 13. In contrast, no obvious absorbance was found for other pH solutions The results indicate that the Hb-Cu3(PO4)2 NFs actualized dual enzyme-like activity. In Figure S5, it can be observed that the OPD oxidation rate catalyzed by the Hb-Cu3(PO4)2 NFs was dependent on the concentration of nanoflowers and period of time at pH 10. The absorbance values were increased with an increasing amount of NFs and length of time. According to a method described previously [26][27][28][29][30], the mechanism of peroxidase-like catalytic activities of Hb-Cu3(PO4)2 NFs can be further investigated using steady-state kinetic assays at pH 10. As shown in Figure S6  According to a previous report [24], Hb-Cu 3 (PO 4 ) 2 NFs showed peroxidase activity under acidic pH conditions. However, the above fluorescence experiment results indicate that Hb-Cu 3 (PO 4 ) 2 NFs were highly active even at pH 10. Therefore, the effect of pH on the enzyme activity of Hb-Cu 3 (PO 4 ) 2 NFs was throughout measured by varying the pH and keeping the OPD and/or H 2 O 2 concentration constant. As shown in Figure S3, the peroxidase-like activity of Hb-Cu 3 (PO 4 ) 2 NFs was present at pH 3 to 11. Interestingly, as shown in Figure S4, it was found that Hb-Cu 3 (PO 4 ) 2 NFs could catalyze oxidase reactions with OPD in the absence of H 2 O 2 and showed oxidase-like activity at pH 12 and 13. In contrast, no obvious absorbance was found for other pH solutions The results indicate that the Hb-Cu 3 (PO 4 ) 2 NFs actualized dual enzyme-like activity. In Figure S5, it can be observed that the OPD oxidation rate catalyzed by the Hb-Cu 3 (PO 4 ) 2 NFs was dependent on the concentration of nanoflowers and period of time at pH 10. The absorbance values were increased with an increasing amount of NFs and length of time. According to a method described previously [26][27][28][29][30], the mechanism of peroxidase-like catalytic activities of Hb-Cu 3 (PO 4 ) 2 NFs can be further investigated using steady-state kinetic assays at pH 10. As shown in Figure S6

Condition Optimization
A series of experiments were conducted to establish the optimum analytical conditions for the oxidation of thiamine by H2O2 with Hb-Cu3(PO4)2 NFs as catalyst. The parameters investigated were reaction time; the concentrations of pH, H2O2 and Hb-Cu3(PO4)2 NFs; and the fluorescence sensitizer, Tween 80.

Effect of pH and Reaction Time
The solution pH was an important factor which influenced the peroxidase-like activity of the Hb-Cu3(PO4)2 NFs. As shown in Figure 4a, when using different phosphate buffers ranging from pH 8 to 13, the peak of the maximum fluorescence intensity was obtained with a phosphate buffer of pH 10. The fluorescence intensity of thiochrome rapidly increased with reaction time, initially from 0 to 300 s, and then the values slowly reached a plateau, indicating that the H2O2 was completely consumed by oxidation of the substrate TH. Thus, 300 s was chosen as the optimal reaction time for a phosphate buffer of pH 10.

Effect of H2O2 Concentration
To achieve a maximum oxidation rate of the thiamine, optimization of hydrogen peroxide concentration was necessary. As shown in Figure 4b, the fluorescence intensity increased quickly with the concentration of H2O2 up to 100 mM and then decreased. Therefore, 100 mM H2O2 was selected as the optimum concentration.

Effect of Hb-Cu3(PO4)2 NF Concentration
In order to achieve the optimal oxidation rate of the thiamine catalyzed by Hb-Cu3(PO4)2 NFs, the effect of Hb-Cu3(PO4)2 NF concentrations (0-4 mg/mL) on the fluorescence of thiochrome was investigated. As shown in Figure 4c, the fluorescence intensity first increased and then slowly decreased with an increase in Hb-Cu3(PO4)2 NF concentration. The maximum fluorescence intensity

Effect of pH and Reaction Time
The solution pH was an important factor which influenced the peroxidase-like activity of the Hb-Cu 3 (PO 4 ) 2 NFs. As shown in Figure 4a, when using different phosphate buffers ranging from pH 8 to 13, the peak of the maximum fluorescence intensity was obtained with a phosphate buffer of pH 10. The fluorescence intensity of thiochrome rapidly increased with reaction time, initially from 0 to 300 s, and then the values slowly reached a plateau, indicating that the H 2 O 2 was completely consumed by oxidation of the substrate TH. Thus, 300 s was chosen as the optimal reaction time for a phosphate buffer of pH 10.
Sensors 2020, 20, x FOR PEER REVIEW 7 of 12 in the system was obtained at 2 mg/mL of Hb-Cu3(PO4)2 NF solution. Thus, 2 mg/mL was selected for the Hb-Cu3(PO4)2 NFs as the optimal concentration.

Effect of Tween 80 Concentration
Tween 80, a hydrophilic nonionic surfactant, had been utilized in the drug loading and the biosensor fields for its hydrophilicity and low toxicity [31][32][33]. Previous reports have shown that Tween 80 can enhance the fluorescence of a system. For example, it has been used in fluorescence detections of nicotinamide [34] and nifanib [35]. In this work, Tween 80 was added in the fluorescence determination for thiamine. The effect of the amount of Tween 80 (0.1-0.5%) was investigated with the other parameters kept constant. As shown in Figure 4d, the fluorescence intensity first increased and then slowly decreased along with an increase in the amount of Tween 80. The maximum fluorescence intensity in the system was obtained at a Tween 80 solution of 0.4%. Therefore, a Tween 80 solution of 0.4% was chosen as the optimal concentration.

Calibration Curve for Thiamine Detection
Under the optimized reaction conditions, the relationship between the fluorescence intensity and thiamine concentration catalyzed by Hb-Cu3(PO4)2 NFs was investigated. As shown in Figure 5, the fluorescence intensity increased with increasing concentration of thiamine. The linear regression equation was F = 19.247 + 2.452 × 10 8 × Cthiamine with a correlation coefficient of 0.9984. There was a

Effect of H 2 O 2 Concentration
To achieve a maximum oxidation rate of the thiamine, optimization of hydrogen peroxide concentration was necessary. As shown in Figure 4b, the fluorescence intensity increased quickly with the concentration of H 2 O 2 up to 100 mM and then decreased. Therefore, 100 mM H 2 O 2 was selected as the optimum concentration.

Effect of Hb-Cu 3 (PO 4 ) 2 NF Concentration
In order to achieve the optimal oxidation rate of the thiamine catalyzed by Hb-Cu 3 (PO 4 ) 2 NFs, the effect of Hb-Cu 3 (PO 4 ) 2 NF concentrations (0-4 mg/mL) on the fluorescence of thiochrome was investigated. As shown in Figure 4c, the fluorescence intensity first increased and then slowly decreased with an increase in Hb-Cu 3 (PO 4 ) 2 NF concentration. The maximum fluorescence intensity in the system was obtained at 2 mg/mL of Hb-Cu 3 (PO 4 ) 2 NF solution. Thus, 2 mg/mL was selected for the Hb-Cu 3 (PO 4 ) 2 NFs as the optimal concentration.

Effect of Tween 80 Concentration
Tween 80, a hydrophilic nonionic surfactant, had been utilized in the drug loading and the biosensor fields for its hydrophilicity and low toxicity [31][32][33]. Previous reports have shown that Tween 80 can enhance the fluorescence of a system. For example, it has been used in fluorescence detections of nicotinamide [34] and nifanib [35]. In this work, Tween 80 was added in the fluorescence determination for thiamine. The effect of the amount of Tween 80 (0.1-0.5%) was investigated with the other parameters kept constant. As shown in Figure 4d, the fluorescence intensity first increased and then slowly decreased along with an increase in the amount of Tween 80. The maximum fluorescence intensity in the system was obtained at a Tween 80 solution of 0.4%. Therefore, a Tween 80 solution of 0.4% was chosen as the optimal concentration.

Calibration Curve for Thiamine Detection
Under the optimized reaction conditions, the relationship between the fluorescence intensity and thiamine concentration catalyzed by Hb-Cu 3 (PO 4 ) 2 NFs was investigated. As shown in Figure 5, the fluorescence intensity increased with increasing concentration of thiamine. The linear regression equation was F = 19.247 + 2.452 × 10 8 × C thiamine with a correlation coefficient of 0.9984. There was a good linear correlation between the fluorescence intensity and the thiamine concentration in the range of 5 × 10 −8 to 5 × 10 −5 M. The lower limit of detection of Hb-Cu 3 (PO 4 ) 2 NFs for thiamine was found to be 4.8 × 10 −8 M, which is similar to the HRP-based fluorescent method for the detection of thiamine [36]. It is almost believable that its ability to detect thiamine is similar to the HRP-based fluorescent method. Compared with HRP, nanozymes have the advantages of being highly stable against denaturing, low-cost, easy to store and suitable for treatment. The mimic enzyme activity of Hb-Cu 3 (PO 4 ) 2 NFs was firstly utilized in fluorometric sensing of thiamine under basic conditions. Compared with other systems for the determination of thiamine, as shown in Table 1, the proposed method based on Hb-Cu 3 (PO 4 ) 2 NFs is on a par with other fluorescence methods. Additionally, the synthesis of the Hb-Cu 3 (PO 4 ) 2 NFs was convenient and simple. Therefore, the method should be a potential candidate for fluorescence sensors of thiamine.
Compared with other systems for the determination of thiamine, as shown in Table 1, the proposed method based on Hb-Cu3(PO4)2 NFs is on a par with other fluorescence methods. Additionally, the synthesis of the Hb-Cu3(PO4)2 NFs was convenient and simple. Therefore, the method should be a potential candidate for fluorescence sensors of thiamine.

Determination of Thiamine in Functional Food Tablet Samples
In order to evaluate the feasibility of the proposed method based on Hb-Cu3(PO4)2 NFs, the concentrations of thiamine in multivitamin functional food tablets samples were analyzed under the optimal conditions. As shown in Table 2, the recovery level of the developed approach was between 103.20 and 115.20%. The result suggests that the developed method is accurate and reliable for thiamine detection in real samples.

Determination of Thiamine in Functional Food Tablet Samples
In order to evaluate the feasibility of the proposed method based on Hb-Cu 3 (PO 4 ) 2 NFs, the concentrations of thiamine in multivitamin functional food tablets samples were analyzed under the optimal conditions. As shown in Table 2, the recovery level of the developed approach was between 103.20 and 115.20%. The result suggests that the developed method is accurate and reliable for thiamine detection in real samples.

Interference Study
The selectivity of the proposed method was examined by comparing the fluorescence change of the solution with thiamine (0.3 mg/mL) with other coexistence substrates (30 mg/mL) including K + , Sensors 2020, 20, 6359 9 of 11 Na + , Cl -, HCO 3-, glucose, starch, VB 2 , VB 7 , VB 9 and Vc. These control compounds are usually found in foods. As shown in Figure 6, the fluorescence of the solution changed only when thiamine was added to the mixture. The addition of the other molecules, in contrast, had no obvious effect on the fluorescence of the solutions. Thus, the proposed fluorometric method displayed a high selectivity for the determination of thiamine.

Interference Study
The selectivity of the proposed method was examined by comparing the fluorescence change of the solution with thiamine (0.3 mg/mL) with other coexistence substrates (30 mg/mL) including K + , Na + , Cl -, HCO 3-, glucose, starch, VB2, VB7, VB9 and Vc. These control compounds are usually found in foods. As shown in Figure 6, the fluorescence of the solution changed only when thiamine was added to the mixture. The addition of the other molecules, in contrast, had no obvious effect on the fluorescence of the solutions. Thus, the proposed fluorometric method displayed a high selectivity for the determination of thiamine.

Conclusions
In summary, a new fluorometric enhancement method for the detection of thiamine has been demonstrated. The peroxidase-like activity of Hb-Cu3(PO4)2 NFs was first found in a basic solution. Under the optimal conditions, H2O2 could be decomposed into·OH radicals by Hb-Cu3(PO4)2 NFs that oxidize thiamine rapidly and efficiently. A linear correlation was established between fluorescence intensity and the concentration of thiamine from 0.05-50 μM with a detection limit of 48 nM. Future research may focus on attempts to explore the new protein-inorganic hybrid nanoflowers or analytes in the detecting systems.

Conclusions
In summary, a new fluorometric enhancement method for the detection of thiamine has been demonstrated. The peroxidase-like activity of Hb-Cu 3 (PO 4 ) 2 NFs was first found in a basic solution. Under the optimal conditions, H 2 O 2 could be decomposed into·OH radicals by Hb-Cu 3 (PO 4 ) 2 NFs that oxidize thiamine rapidly and efficiently. A linear correlation was established between fluorescence intensity and the concentration of thiamine from 0.05-50 µM with a detection limit of 48 nM. Future research may focus on attempts to explore the new protein-inorganic hybrid nanoflowers or analytes in the detecting systems.