**1. Introduction**

Anthrax is a well-known disease caused by *Bacillus anthracis*, which can a ffect almost all warm-blooded animals, including human beings, resulting in deadly infections after inhalation of over 10<sup>4</sup> spores in 36 h [1]. Since the spores of *Bacillus anthracis* are highly environmentally adaptive, they have been developed as a biological weapon, which makes them a biohazard threat [2]. As a main ingredient of the bacterial spores, 2,6-dipicolinic acid (DPA) represents 5−15% of the dry mass of the spores and can be served as a typical anthrax biomarker [3]. Thus, developing an e fficient and accurate method for DPA detection is very important in the fields of medical care and anti-terrorism.

Compared with traditional detection methods for DPA, fluorescence-based sensing methods have attracted plenty of interest owing to their real-time, economic, highly selective, and sensitive features [4,5]. Among these, sensing platforms based on rare earth ions (Ln<sup>3</sup>+) for DPA determination have received considerable attention due to their high coordination ability with DPA and their excellent spectroscopic properties, namely, large stokes shift, sharp emission bands, and long fluorescence (FL) lifetime [6]. When coordinated with DPA, the FL intensity of Ln3<sup>+</sup> becomes more intense via the absorbance-energy transfer-emission e ffect (AETE) [7,8]. However, most reported measurements rely on single fluorescent signal changes of Ln3+, which may be easily influenced by environmental or instrumental factors [3,6,7]. To conquer this limitation, ratiometric FL probes that contain another FL spectral peak as an internal reference would be an ideal choice to improve the accuracy of the

detection. Hitherto, various Ln3<sup>+</sup>-incorporated fluorescent ratiometric platforms for DPA detection have been exploited, such as silicon quantum dots [4], solid films [1], Micelle [9], and metal-organic framework [10].

As a new family member of carbon nanomaterials, carbon dots (CDs) have recently inspired substantial attention because of prominent properties such as facile preparation, low cost, high photostability, and nontoxicity [11,12]. Although a few CDs-based FL nanoprobes for DPA determination have been reported [13,14], further improvement is still needed for the DPA sensors, for example, more facile synthesis, lower cost, and higher photostability and sensitivity. Moreover, it is worth mentioning that using a more environmentally friendly approach to synthesize CDs with fine quality remains a pressing problem waiting to be resolved [15]. Using renewable and low cost green biomass as raw materials to synthesize CDs will inevitably promote the sustainable development of CDs and their applications.

Herein, CDs with bright blue FL were prepared by a simple and green method using schizochytrium (a kind of microalgae) as precursor. Subsequently, a new ratiometric FL nanoprobe (CDs-Tb) was prepared for the determination of DPA by grafting Tb3+ onto the surface of CDs (Scheme 1). Under optimal conditions, good linearity between the ratio FL intensity of *F*545/*F*445 and the DPA concentrations was observed within the experimental concentration range of 0.5–6 μM with the detection limit of 35.9 nM. Moreover, the CDs-Tb could realize sensitive detection of DPA in lake water samples. The comparison of several existing FL nanoprobes for DPA detection is listed in Table S1, indicating good sensitivity of our sensing system compared with previously reported ones [4,14,16–19].

To the best of our knowledge, this is the first example of CDs prepared from microalgae and pure water by using a hydrothermal method [15,20]. Although one case of hydrothermal synthesis of microalgae-based carbon dots has been reported, formaldehyde aqueous solution was added during the hydrothermal reaction, which was obviously not environmentally friendly [20]. Moreover, this work o ffers an e fficient self-calibrating and background-free method for the determination of DPA.

**Scheme 1.** Schematic diagram of the CDs-Tb nanoprobe for DPA recognition.

#### **2. Materials and Methods**

## *2.1. Materials*

Tris(hydroxymethyl)aminomethane, 2,6-dipicolinic acid (DPA), m-phthalic (mPA), o-phthalic (oPA), benzoic (BA), glutamic (Glu), D-aspartic (Asp) acid, Tb(NO3)3·6H2O and nitrate salts of metal ion of analytic grade were purchased from Shanghai Energy Chemical Corporation (Shanghai, China). Schizochytrium were purchased from Wudi Green Science Engineering Co., Ltd. (Shandong, China). Cellulose dialysis membrane were purchased from Jingke Hongda Biotechnology Co., Ltd. (Beijing, China).

#### *2.2. Preparation of CDs*

CDs with blue FL emissions were fabricated by a green and facile hydrothermal method (Scheme 2). Briefly, 2.0 g schizochytrium and 10 mL distilled water were placed in a Teflon-lined stainless steel vessel (23 mL) and heated at 200 ◦C for 4 h. The obtained mixture was centrifuged to remove large

particle residues. Subsequently, the redundant supernatant was dialyzed for two days via a cellulose dialysis membrane (MWCO 1000) in the pure water. After drying by lyophilization, CDs powders were collected.

**Scheme 2.** Schematic diagram of the preparation for CDs and CDs-Tb.

#### *2.3. Preparation of CDs-Tb*

Then, 0.1 mmol Tb(NO3)3 was added into 10 mL of aqueous CDs (1.0 mg·mL−1) solution. At room temperature, the mixtures were stirred for 2 h and then subjected to dialysis. "Free" Tb3+ ions were removed via a cellulose membrane (MWCO 500) in pure water for 2 days. After drying by lyophilization, CD-Tb powders were collected.
