**3. Results**

#### *3.1. Characterization of the CDs*

The schizochytrium-based CDs were facilely synthesized by a hydrothermal method (Scheme 2). TEM analysis was performed to characterize the microstructure of the CDs. As depicted in Figure 1a, the CDs were spherical shapes and showed monodispersity. The mean diameter of the CDs was 4.5 nm with size distribution from 2.8 to 6.3 nm (Figure 1b). As depicted in Figure 1c, clear lattice fringes of 0.33 nm and 0.20 nm could be observed, which were respectively close to the (002) and the (020) planes of graphitic carbon, implying its graphitic nature of CDs [21].

**Figure 1.** (**a**) TEM image; (**b**) the size distribution histogram; (**c**) HRTEM images; and (**d**) FTIR spectrum of CDs.

FTIR and XPS plots were determined to investigate the chemical bonds and the compositions of the CDs. Figure 1d shows the O–H and the N–H stretching vibrations in the 3000–3700 cm<sup>−</sup><sup>1</sup> regions as well as the vibrations of C=O, C–N, and C–O at 1641, 1396, and 1095 cm<sup>−</sup>1, indicating the existence of carboxyl, amine, and hydroxyl groups [21]. In the XPS plot of CDs (Figure 2a), peaks at 285.0, 398.9, and 532.1 eV were respectively corresponding to C1s, N1s, and O1s. Figure 2b,c further ensured the existence of C–N, N–H, C–O/C=O, and C–C/C=C bonds in CDs.

As shown in Figure 2d, the as-synthsized CDs displayed two absorption bands centered at 225 and 278 nm, as well as one weak band around 336 nm, which respectively corresponded to the Π–Π\* transition of the aromatic sp<sup>2</sup> structure and the n–Π\* transition of carbonyl [22,23].

**Figure 2.** (**a**) XPS spectrum; (**b**) High-resolution C1s XPS spectra; (**c**) High-resolution N1s XPS spectra and(d) UV-Vis absorption spectrum of CDs.

As portrayed in Figure S1, CDs demonstrated excitation wavelength-dependent property, which indicated the effects of different sized particles and various surface states distribution, as in most reported CDs [22–26].

The absolute quantum yield of CDs is determined to be 11% with the excitation wavelength of 330 nm, which is comparable with other reported CDs [13].

#### *3.2. Characterization of CDs-Tb*

In order to detect DPA, CDs-Tb has been synthesized as a ratiometric FL nanoprobe (Scheme 2). In this nanoprobe (Scheme 1), CDs are not only act as ligands to coordinate with Tb3+, but also serve as a fluorescence reference, while Tb3+ ions act as specific recognition unit and response signal.

XPS characteration was also carried out to determine the composition of CDs-Tb. Figure 3 clearly indicated the presence of carbon, nitrogen and oxygen in this nanoprobe. Additionally, a weak peak at 151.8 eV (Figure 3a,b) corresponding to Tb 4d is appeared, which confirmed the preparation of the CDs-Tb. Besides, two peaks at 1277.3 and 1242.5 eV corresponded to 3d3/2 and 3d5/2 of Tb3+ were observed in the HRXPS spectrum of Tb 3d (Figure 3c).

**Figure 3.** (**a**) XPS spectrum; (**b**) high-resolution Tb 4d XPS spectra; (**c–f**) high-resolution Tb3d, C1s, N1s and O1s XPS spectra of CDs-Tb.

Optical properties of CDs-Tb including FL emission and UV-vis spectra were investigated. As demonstrated in Figure S2, CDs-Tb showed one broad absorption band around 275 nm. Under the excitation of 265–375 nm, CDs-Tb exhibited excitation-dependent FL property with maximum emission and excitation peaks around 445 and 265 nm, respectively (Figure S3).

#### *3.3. Determination of the DPA*

The ability of this nanoprobe for detection of DPA was assessed systematically by FL titrations. As delineated in Figure 4a, without the addition of DPA, the emission spectrum of CDs-Tb was dominated by the blue FL emission of CDs centered at 445 nm. Upon addition of various amounts of DPA, the emission intensities of CDs remained nearly unchanged, while the emission intensities of peaks at 489, 545, 586, and 621 nm assigned to Tb3+ ions enhanced obviously, owing to effective energy transfer from DPA to Tb3+. Thus, the blue FL of the CDs centered at 445 nm served as the reference signal, while the green FL of DPA-sensitized Tb3+ served as the response signal in the CDs-Tb nanoprobe. Since the FL emission of Tb3+ at 545 nm was the most intense emission peak after addition of DPA, its FL intensity changes were set to quantitatively measure DPA.

Moreover, we investigated the influence of pH on the ratio FL intensity of F545/F445 in the nanoprobe. As portrayed in Figure S4, the ratio FL intensity remained nearly unchanged within pH range 4.0–8.0 because of the abundant oxygen-containing groups on the surface of CDs and the balance of their protonation and deprotonation. However, this ratio FL intensity remarkably reduced at pH > 9.0 on account of the generation of Tb3+ hydroxide [22].

As delineated in Figure 4a, the FL intensity of Tb3+ at 545 nm was obviously enhanced upon the addition of DPA. Nearly nine-fold increase was observed when DPA concentration was 9 μM. The ratio FL intensity of F545/F445 and the DPA concentrations showed a linear relationship with R<sup>2</sup> = 0.985 in the experimental concentration range 0.5–6 μM (Figure 4b). The following equation was utilized to calculate the detection limit of CDs-Tb toward DPA: the detection limit = 3SB/S, where SB was standard error for the blank test, which was determined by 10 continuous scannings of the blank sample, and where S was the slope of the calibration curve. The obtained detection limit was 35.9 nM, which was superior compared to values in previously published literature and significantly lower than the infectious dosage of the spores (60 μM) [4,13,14,16–19,22].

**Figure 4.** (**a**) Fluorescence (FL) response of CDs-Tb upon the addition of DPA (λex = 270 nm); (**b**) ratiometric calibration plot of CDs-Tb (F545 /F445) and DPA concentration.

#### *3.4. Mechanism for DPA Detection Using CDs-Tb*

We further explored the mechanism for DPA detection. As depicted in Figure 5a,b, the FL lifetime of CDs in the CDs-Tb solution (0.1 mg·mL−1) slightly decreased from 30.23 ns to 27.46 ns upon the addition of DPA (80 μM), while that of the green FL of Tb3+ in the CDs-Tb significantly increased from 11.90 μs to 822.41 μs. Such phenomenon indicated that DPA could absorb the excitation light and transfer its energy to Tb3+ in CDs-Tb effectively, implying the mechanism for DPA detection was attributed to the AETE from DPA to Tb3+ [7,8]. As for CDs in CDs-Tb, the reduction of excitation light absorption may have been responsible for its slightly decreased lifetime, which was nearly negligible [13].

Moreover, as demonstrated in Figure 5c,d, the FL intensity of free CDs remained almost unchanged after adding DPA, while the signal of free Tb3+ was dramatically enhanced, implying the vital role of Tb3+ in DPA detection. As portrayed in Figure 5d, new bands located at 489, 545, and 586 nm were clearly observed, which corresponded to the characteristic emission of Tb3+ induced by the effective energy transfer from DPA to Tb3+ [21].

**Figure 5.** FL decay profiles of emissions at (**a**) 445 nm and (**b**) 545 nm of CDs-Tb and CDs-Tb+DPA (λex = 270 nm); FL intensity of (**c**) the CDs (0.5 mg·mL−1) and (**d**) the Tb3+ ions (1 mM) with and without DPA (60 μM).

#### *3.5. Selectivity of DPA Detection*

As another vital parameter for a nanoprobe, selectivity of CDs-Tb was evaluated. The effects of other structurally related species, such as m-phthalic (mPA), o-phthalic (oPA), benzoic (BA), glutamic (Glu), and D-aspartic (Asp) acid, were studied under similar conditions. The effects of various common cellular ions (Ca<sup>2</sup>+, Mg<sup>2</sup>+, Na<sup>+</sup>) were also investigated. Notably, no obvious intensity changes occurred when adding the above interfering species compared to DPA (Figure 6a). Furthermore, the coexistence of the above species did not cause obvious interference for CDs-Tb when sensing DPA (Figure 6b) and allowed for selective determination of DPA.

**Figure 6.** (**a**) Influences of DPA and interfering species on the FL intensity of CDs-Tb. (**b**) Blue bars indicate influences of interfering species (10 μM) on the FL intensity of CDs-Tb and red bars are influences of interfering species and DPA (4 μM) on the FL intensity of CDs-Tb (10 <sup>μ</sup>g·mL−1), λex = 270 nm.

#### *3.6. Analysis in Real Samples*

To further evaluate the applicability of this ratiometric nanoprobe toward DPA, some studies were carried out in the lake water sample. Because no DPA was detected in the lake sample by using the CDs-Tb, a recovery measurement was performed. DPA solutions (1, 2, 5 μM) were added into the lake water sample and determined by the CDs-Tb nanoprobe, respectively. With the obtained ratiometric calibration plot of CDs-Tb (Figure 4b), DPA concentrations were determined. Table 1 demonstrated that the relative standard deviation (RSD) of these tests did not exceed 3.79%, while recoveries were within the range from 96.5% to 105.01%. This ratiometric sensing system showed high accuracy and precision toward DPA determination in real samples.


**Table 1.** Detection of DPA in the lake water sample. pH: 7.6; λex: 270 nm; *c*CDs-Tb: 10 <sup>μ</sup>g·mL−1.

1 Average of three repeated detections. 2 Recovery (%) = (*<sup>c</sup>*found/*c*added) × 100. RSD: relative standard deviation.
