Recent Advances in Near-Infrared Fluorescent Probes Based on Organic Small Molecules for the Detection and Imaging of Heavy Metal Ions

Abstract

Heavy metal pollution poses a severe threat to ecosystems and human health. Therefore, it is important to develop rapid, sensitive, and reliable detection methods. Fluorescent probes with near-infrared (NIR, 650–900 nm) emission have emerged as powerful analytical tools owing to their distinct advantages, including low background interference, deep-tissue penetration, minimal photodamage, and high signal-to-noise ratios, which make them particularly attractive for both environmental monitoring and bioimaging applications. This review systematically summarizes recent progress in NIR organic small-molecule fluorescent probes for the detection of heavy metal ions, covering the period from 2022. Representative examples are categorized and discussed, with a focus on key metal ions such as Cu²⁺, Hg²⁺, Cd²⁺, Zn²⁺, Pb²⁺, Fe²⁺ and Fe³⁺. The molecular design strategies, sensing performance, and practical applications in environmental samples, food samples and biological imaging are critically discussed. Furthermore, current challenges and future opportunities in this rapidly advancing field are highlighted.

1. Introduction

With the rapid development of modern industrialization and urbanization, heavy metal pollution has become one of the most severe global environmental problems. Heavy metal ions such as Hg²⁺, Pb²⁺, Cd²⁺, Fe²⁺/Fe³⁺, and Cu²⁺/Cu⁺ are released into the environment through various pathways, including mineral mining, metal smelting, electronic waste disposal, industrial wastewater discharge, and pesticide use, posing serious threats to ecosystems and human health. Owing to their characteristics of high toxicity, non-biodegradability and bioaccumulation, even at trace levels, heavy metal ions can accumulate in living organisms through the food chain, ultimately leading to various severe effects to human beings [1,2]. For example, excessive Cu²⁺ in the human body can induce oxidative stress, leading to gastrointestinal dysfunction and neurological damage. Chronic copper overload may affect the liver, brain, and kidneys, producing clinical manifestations resembling Wilson disease, including cirrhosis, tremor, and neuropsychiatric symptoms [3,4]. Hg²⁺ overload can cause irreversible damage to multiple organs and systems, including the kidneys, heart, central nervous system, cardiovascular system, and reproductive system. In particular, methylmercury (CH₃Hg⁺) readily crosses the blood–brain barrier and poses a particular threat to fetal and childhood neurodevelopment [5,6]. Zinc overload can result in multisystem toxicity, primarily manifesting as gastrointestinal disturbances, immune dysfunction, neurological damage, and dyslipidemia, thereby increasing the risk of atherosclerosis; it may also disrupt cardiac electrophysiological activity and contribute to arrhythmias [7]. Following entry into the human body, Cd²⁺ predominantly accumulates in the kidneys and bones. Excessive cadmium exposure can cause renal tubular injury, disturbances in calcium and phosphorus metabolism, osteoporosis, and osteomalacia, and in severe cases may progress to Itai-Itai disease. Chronic exposure is also associated with a significantly increased risk of malignancies, including lung and prostate cancers [8]. Lead poisoning can cause severe multisystem damage, primarily affecting the hematopoietic, nervous, digestive, urinary, cardiovascular, and immune systems [9]. Excessive iron intake can irritate the gastrointestinal tract, increase the risk of cardiovascular diseases, cause liver injury, and induce tumorigenesis [10,11]. Thus, the World Health Organization (WHO) and the United States Environmental Protection Agency (EPA) have set strict limits on the concentration of various heavy metal ions in drinking water. For example, the WHO has established a maximum allowable concentration of 0.5 μg/L for Hg²⁺ [12], 5 μg/L for Cd²⁺ [13], and 10–15 mg/L for Pb²⁺ [14]. Thus, the development of rapid, sensitive and reliable technology to detect heavy metal ions has significant social and economic significance for environmental monitoring, food safety and public health protection.

Traditional detection methods for monitoring heavy metal ions include atomic absorption spectrometry (AAS), atomic emission spectrometry (AES), inductively coupled plasma mass spectrometry (ICP-MS) and electrochemical analysis [15,16,17,18]. Although these methods have high sensitivity and accuracy, they have inherent limitations including high instrument costs, complex pre-processing procedures, time-consuming operations, and the inability to be utilized for in situ visualization of biological imaging, which limit their application in a wide range of scenarios [19,20].

The field of fluorescent probes began with the pioneering work of Minta et al., who reported the first Cu²⁺-selective fluorescent probe in 1989. Since then, thousands of fluorescent probes have been developed [21,22,23,24,25,26,27]. Compared with traditional methods, fluorescent probes are characterized by their high selectivity and sensitivity, real-time detection, low cost and non-invasive imaging, making them attractive tools in life sciences, environmental monitoring, and medical diagnostics, particularly for ion detection, bioimaging, and drug delivery [28,29,30,31]. Most fluorescent probes emit in the visible region (λ_em < 600 nm), making them susceptible to background fluorescence, tissue absorption and strong light scattering during application. In comparison, near-infrared fluorescent probes (NIR, 650–900 nm) offer the benefits of minimal background fluorescence, lower light scattering, deep-tissue imaging capability, and high imaging contrast [32,33,34,35,36]. Commonly used NIR fluorophores include BODIPYs [37,38,39,40], hemicyanines [41,42,43,44], cyanines [45], dicyanoisophorone analogues [46,47,48,49], DCM derivatives [50,51,52], rhodamines [53,54], etc. (Figure 1). These dyes generally exhibit large Stokes shifts, tunable structures, and pronounced spectral changes upon interaction with analytes. Owing to these advantages, NIR fluorescence imaging has gained considerable attention over the past decade, with numerous NIR-emitting materials being applied to detect intracellular active species, analyze tumor biomarkers, monitor disease progression, evaluate drug efficacy in vivo, and more recently, to address challenges in environmental and food safety monitoring.

Figure 1. General structures of commonly used NIR fluorophores.

In recent years, there have been several reviews covering the developments in NIR fluorescent probes, focusing primarily on the sensing and bioimaging of small molecules (e.g., biothiols, H₂S) [55,56] and biomacromolecules (e.g., proteins, enzymes, nucleic acids) [57,58,59], as well as their applications in cancer targeting and disease diagnosis (e.g., Alzheimer’s disease, kidney injury, urological disorders) [60,61,62,63]. For example, Yu, Cong and co-workers (2025) systematically outlined advances in NIR small-molecule probes for responsive biomarkers from 2018 to 2025, emphasizing their roles in cancer research and therapy [64]. Wang et al. (2025) comprehensively reviewed NIR fluorescent probes for detecting Alzheimer’s disease biomarkers, categorizing them by dye structure, for instance curcumin, BODIPY, quinoline, and donor acceptor systems, and discussing their spectral characteristics, in vitro detection, and in vivo imaging performance [65]. Additionally, many reviews have addressed design strategies based on specific fluorophore platforms (e.g., heptamethine cyanines, coumarins, rhodamines, silicon-substituted xanthenes) and their sensing and bioimaging applications [66,67,68,69,70]. Pu et al. (2021) systematically summarized recent advances in NIR probes built upon a hemicyanine skeleton for disease imaging and diagnosis [71], while Guo et al. (2025) highlighted hemicyanine-based NIR probes for detecting key disease-related biomarkers and their utility in diagnosis and treatment [72].

Despite these contributions, a comprehensive and dedicated review that systematically described the construction strategies, analytical performance, response mechanisms, and bioimaging utility of NIR fluorescent probes for heavy metal ions is still lacking. Over the past few years, remarkable progress has been made in the detection and bioimaging of ions such as Cu²⁺, Hg²⁺, Cd²⁺, Zn²⁺, Pb²⁺ and Fe²⁺/Fe³⁺ using NIR probes. Hence, a comprehensive and timely survey is urgently required to guide future developments in this vibrant research domain.

Here, we comprehensively reviewed the research progress on NIR fluorescent probes for heavy metal ions reported since 2022. This review introduces the common sensing mechanisms employed by these probes. Then, the probes are classified by the target metal ions, focusing on molecular design strategies, sensing behaviors, recognition mechanisms, and practical applications. Finally, based on an in-depth analysis of the sensing performance and application scope of the reported NIR probes, we identify the existing challenges and future directions in this rapidly evolving field. This review is expected to provide valuable guidance for the rational design of next-generation NIR fluorescent probes for the detection of heavy metal ions in complex environmental and biological systems.

2. Sensing Mechanism of Fluorescent Probes for Heavy Metal Ions

There are various mechanisms by which fluorescent probes detect heavy metal ions, including photo-induced electron transfer (PET), intramolecular charge transfer (ICT), twisted intramolecular charge transfer (TICT), fluorescence resonance energy transfer (FRET), excited-state intramolecular proton transfer (ESIPT), chelation-enhanced fluorescence (CHEF) or quenching (CHEQ), aggregation-induced emission (AIE), ligand-to-metal charge transfer (LMCT), metal-to-ligand charge transfer (MLCT), through-bond energy transfer (TBET), and so on. The following sections will detail the mechanisms most frequently employed in near-infrared fluorescent probes for heavy metal ion sensing, which are also extensively discussed in key review articles [73,74].

2.1. Photo-Induced Electron Transfer (PET)

Fluorescent probes based on the PET mechanism typically consist of three key components: a fluorophore, a linker, and a recognition group. In such a system, the fluorophore serves as an electron acceptor, while the recognition unit functions as an electron donor. A critical feature of this design is that the highest occupied molecular orbital (HOMO) energy level of the recognition unit (electron donor) lies between the HOMO and the lowest unoccupied molecular orbital (LUMO) energy levels of the fluorophore (electron acceptor) (Figure 2). Upon photoexcitation of the fluorophore, an electron is promoted from its HOMO to its LUMO, creating a vacancy in the HOMO of the fluorophore that is subsequently filled by an electron from the HOMO of the receptor. This inter-component electron transfer process, namely the PET process, effectively quenches the fluorescence of the fluorophore. When a target cation binds to the recognition unit, the HOMO energy level of the donor is lowered. It reaches a lower level than that of the fluorophore, thereby suppressing the PET pathway and restoring the fluorescence emission. This is a characteristic signal “turn-on” response.

Figure 2. Mechanism of photo-induced electron transfer (PET).

2.2. Intramolecular Charge Transfer (ICT)

Typically, ICT-based fluorescent probes are composed of an electron-donating moiety and an electron-withdrawing moiety linked via a conjugated bridge, thereby creating a “push–pull” electronic system (Figure 3). Upon binding of the recognition unit to the target analyte, the electron-donating or electron-withdrawing nature of the system is modulated, leading to readjustment of the charge distribution throughout the entire conjugated system. Consequently, the “push–pull” effect is either enhanced or weakened, producing a spectral shift (either red or blue) in absorption or fluorescence. This mechanism is often used to design ratiometric fluorescence probes.

Figure 3. The mechanism of intramolecular charge transfer (ICT).

2.3. Fluorescence Resonance Energy Transfer (FRET)

FRET is another frequently employed mechanism for constructing ratiometric fluorescent probes. Such probes typically consist of two fluorophores, serving as the energy donor–acceptor pair. The core principle of FRET relies on sufficient spectral overlap between donor emission and acceptor absorption, with their separation generally falling within 10–100 Å (Figure 4). In the absence of energy transfer, donor emission is predominantly observed. Upon target recognition, the FRET pathway is activated, causing an increase in acceptor emission and a corresponding decline in donor emission. Conversely, if the probe is initially in a FRET-active state, fluorescence from the acceptor predominates. Interaction with the analyte can inhibit the FRET progress, causing recovery of donor fluorescence and attenuation of acceptor emission.

Figure 4. The mechanism of fluorescence resonance energy transfer (FRET). D: donor; A: acceptor; hn: energy.

2.4. Chelation-Enhanced Fluorescence (CHEF)

CHEF serves as a crucial photophysical pathway for fluorescence sensing. In such probes, the fluorophore or receptor unit typically incorporates heteroatoms with lone-pair electrons (such as, N, O, and S) (Figure 5). Upon excitation, the probes undergo non-radiative decay through molecular vibrations, rotations, or electronic energy transfer, leading to fluorescence quenching or very weak emission. Upon specific coordination with the target metal ion (Mⁿ⁺), the lone-pair electrons become engaged in bond formation, which rigidifies the local conformation of the recognition site. Consequently, the radiative transition pathway is restored or significantly enhanced, generating an intense fluorescence signal (“turn-on” response).

Figure 5. The mechanism of chelation-enhanced fluorescence (CHEF).

3. NIR Fluorescent Probes for Heavy Metal Ions

3.1. NIR Fluorescent Probes for Cu²⁺

Copper (Cu²⁺) plays dual roles as both an essential trace element and a significant environmental pollutant. Anthropogenic activities, such as industrial manufacturing and agriculture, release substantial amounts of copper into the ecosystems. Cu²⁺ accumulates in water and soil and is subsequently enriched via the food chain, thereby constituting a significant hazard to public health [75]. Biologically, Cu²⁺ is an indispensable co-factor in numerous enzymes critical for hematopoiesis, neural function, development, and antioxidant defense [76]. Consequently, both deficiency and overload disrupt cellular homeostasis, leading to severe disorders, including Menkes and Wilson’s diseases, neurodegeneration, and oxidative damage, underscoring the critical importance of maintaining copper homeostasis [77]. Given its environmental prevalence and profound biological impact, developing reliable monitoring tools is essential. Leveraging their deep penetration into tissues and minimal background signals, NIR fluorescent probes offer a promising approach for the real-time detection of Cu²⁺ in complicated biological and environmental specimens. During the period of 2022–2026, diverse NIR fluorescent probes were successfully developed for Cu²⁺ sensing. Based on dyes including rhodamine, isophorone, heptamethine cyanine, and so on, these probes demonstrate outstanding sensing capabilities. The photophysical properties of representative probes are summarized in Table 1.

Table 1. The photophysical properties of NIR fluorescent probes for Cu²⁺.

Name	Solvent	Ex (nm)	Em (nm)	Sensing Mechanism	RESPONSE TIME	REVERSIBLE	Signal Transduction Mode	Detection Limit	pH Range	Application	Ref.
DDP-Cu	PBS/DMSO = 8/2, v/v	520	665	ICT	-	No	turn-off	36 nM	3–7.45	SH-SY5Y cells, HeLa cells, zebrafish, mice	[78]
DCM-CuCl	PBS /DMSO = 1/4, v/v	520	668	ICT	2.5 min	-	turn-on	0.13 μM	-	HeLa cells, tap water, Yangtze river water, Jialing river water	[79]
CL(Cu2+)	DMSO/H2O = 9/1, v/v	435	704/ 580	ICT	2.5 min	-	ratiometric	1.70 μM	4.0–11.0	L929 cells	[80]
QA	DMSO/PBS =3/7, v/v	425	688	CHEQ	-	Yes	turn-off	35.5 mM	-	HepG2 cells, C. elegans, mice, tap water, lake water	[81]
DCPO-Cu	DMF/HEPES = 7/3, v/v	420	655	ICT	<3 min	-	turn-on	0.018 mM	7.0–9.0	River water, tap water, mung bean sprouts, rice, onion roots, soil samples	[82]
VIC	DMF/HEPES = 1/9, v/v	425	675	CHEQ	30 s	-	turn-off	7 nM	5.0–12.0	Sandy soil, loess soil, forest soil, senna alexandrina mill, reed leaves, Chinese chives, garlic bolts, MCF-7 cells, mice	[83]
1(Cu2+)	CH3CN/H2O = 2/1, v/v	520/ 650	580/ 770	ICT	-	No	ratiometric	9 nM	4.0–8.0	HepG2 cells, paulownia sprout	[84]
NRh-NP	DMSO/ HEPES = 1/1, v/v	660	750	PET	20 s	Yes	turn-on	33 nM (fluorescence) 35 nM (UV)	6.5–8.0	MCF-7 cells, mice	[85]
NRh6G-FA	DMSO/ HEPES = 1/1, v/v	665	744	PET	60 s	Yes	turn-on	0.225 μM	7.0–9.5	MCF-7 cells, Astragalus Paeonia lactiflora	[86]
NRK-BD	MeCN/H2O = 7/3, v/v	560	730	PET	1.5 s	Yes	turn-on	81 nM	6.5–8.5	tap water, Tianyi Lake, Yellow river, samples, brassica chinensis, potatoes, carrots, 4 T1 cells, mice	[87]
NRTB	EtOH	600	765	PET	300 s	-	turn-on	22.6 nM	-	Cherry tomatoes, cucumbers, apples, cucumbers, 4T1 cells, mice	[88]
LDMD-N	PBS/CH3CN = 2/1, v/v	615	775	ICT	5 min	NO	turn-on	0.0543 µM	-	SH-SY5Ycells, zebrafish, mice	[89]
CySH	HEPES/EtOH = 1/1, v/v	694	784	PET	40 min	-	turn-on	124 nM	4.0–9.0	HepG2 cells	[90]
PhenIR	ACN/H2O = 1/1, v/v	750	808	CHEQ	-	Yes	turn-off	0.286 mM	4.0–8.2	WS1 fibroblast cells	[91]
CyCu1	HEPES/ DMSO = 99/1, v/v, 0.1% CrEL	710/ 523	810/ 600	ICT	40 min	NO	ratiometric	0.053 mM	-	SH-SY5Y cells, mice, Arabidopsis root	[92]
NUST-Cy-1	THF/H2O = 6/4, v/v	450/ 750	593/ 825	FRET	-	-	ratiometric	1.4 mM	-	Drinking water, HeLa cells, mice	[93]
HX	-	730	810	CHEF	-	Yes	turn-on	0.54 mM	-	Live cells, mice	[94]
HC-TPA	Tris	604	668	CHEQ	1 min	No	turn-off	48 nM	7.0–10.0	Apple, watermelon, grape Peach, lychee, tomato, HEK293T cells	[95]
DDAO-Cu	PBS	602	662	ICT	<3 min	NO	turn-on	1.8 nM	-	PC12 cells, zebrafish, mice	[96]
DCX-Cu	PBS/DMSO = 1/1, v/v	590	746	ICT	10 min	No	turn-on	19.47 nM	6–10	Fountain pool water, tap water, wastewater, HepG2 cells	[97]
NS-5	PBS/DMSO = 95/5, v/v	650	720	PET	20 min	-	turn-on	32 nM	-	HepG2 cells, LO2 cells, mice	[98]

Note: Probe names are shown in bold.

3.1.1. NIR Fluorescent Probes Derived from Dicyanoisohporone Derivatives

Dicyanoisophorone derivatives, owing to their unique D-π-A structure, exhibit excellent fluorescence properties in the NIR region, such as tunable emission wavelengths, large Stokes shifts, and good photostability. Thus, they have emerged as an ideal platform for designing Cu²⁺ fluorescent probes. To date, numerous NIR fluorescent probes (Figure 6) have been fabricated by introducing different specific recognition groups for Cu²⁺ to the dicyanoisophorone skeleton [78,79,80,81,82,83].

Figure 6. Structures of NIR fluorescent probes for Cu²⁺ based on dicyanoisohporone platform.

In 2022, Zhou et al. introduced a 2-pyridinecarbonyl group as a recognition unit and initial fluorescence quencher into the dicyanoisophorone platform to construct the probe DDP-Cu [78]. Owing to the interrupted ICT process, the probe was initially in a fluorescence-off state. Upon Cu²⁺-induced hydrolysis of the picolinate ester, DDP-Cu was converted to the NIR fluorophore with restored ICT, producing a strong emission at 665 nm (LOD = 36 nM). DDP-Cu enabled the imaging and quantification of Cu²⁺ in live cells, zebrafish, and brain tissues of AD model mice. Fluorescence imaging revealed that Cu²⁺ levels in AD brains increased 3.5-fold (mice) and 4-fold (Drosophila) and further uncovered the connections between elevated Cu²⁺, oxidative stress, and Aβ aggregation, making it a promising tool for studying Cu²⁺-related AD pathology.

In 2023, Zhang et al. incorporated a chlorine atom into the picolinate recognition moiety, obtaining the probe DCM-CuCl [79]. The introduction of a chloride substituent effectively accelerated the Cu²⁺-catalyzed hydrolysis reaction, shortening the response time to 2.5 min, while maintaining NIR emission at 668 nm. DCM-CuCl was effectively used for Cu²⁺ visualization in HeLa cells and for the assay of Cu²⁺ in real water samples, demonstrating the positive effect of chlorine modification on enhancing the reaction kinetics.

As research progressed, the strategic incorporation of appropriate recognition moieties onto the dicyanoisophorone framework led to the construction of multifunctional fluorescent probes that can simultaneously sense Cu²⁺ and other biologically or environmentally relevant analytes. In 2025, Li’s group created a coumarin–dicyanoisophorone-based probe CL(Cu²⁺), capable of simultaneously responding to lysine (Lys), Cu²⁺, Fe³⁺, and biothiols [80]. The coordination of Cu²⁺ involved the phenolic oxygen and imine nitrogen of CL(Cu²⁺), blocking ICT and causing fluorescence quenching at 704 nm (LOD = 1.70 mM). The resulting CL(Cu²⁺)-Cu²⁺ complex could further undergo competitive coordination with biothiols like glutathione (GSH), restoring fluorescence and enabling the secondary detection of GSH. CL(Cu²⁺) exhibited good biocompatibility in live-cell imaging.

Dong’s team reported a probe QA, featuring aggregation-induced emission (AIE) characteristics [81]. The oxime and phenolic hydroxyl groups in its structure coordinated with Cu²⁺ in a 2:1 stoichiometry, causing a fluorescence turn-off at 688 nm. Upon subsequent addition, GSH exhibited a stronger binding affinity for Cu²⁺ than QA, thereby displacing the metal ion from the QA-Cu²⁺ complex and liberating the free probe, which led to fluorescence recovery. The QA-Cu²⁺ complex demonstrated high selectivity for GSH, with an LOD of 0.347 μM. QA exhibited excellent selectivity, high sensitivity, strong anti-interference capability, and rapid response toward Cu²⁺ and GSH. The probe has been successfully applied to the quantitative determination of Cu²⁺ in tap water and lake water samples with satisfactory spiked recoveries. Furthermore, in combination with RGB color analysis and a smartphone platform, the probe enabled convenient on-site quantitative detection of Cu²⁺ with an LOD of 40.4 mM. In addition, QA was effectively applied to visualize Cu²⁺ and GSH in HepG-2 cells, C. elegans, and mice, indicating the advantages of AIE probes in biological research.

While the aforementioned isophorone-based probes have primarily been employed for Cu²⁺ detection under controlled laboratory conditions, recent efforts have shifted toward the development of practical tools that enable rapid, on-site monitoring of Cu²⁺ in real-world settings. In 2025, Li et al. synthesized the probe DCPO-Cu using 2-aminonicotinic acid as the responsive group for Cu²⁺ [82]. Cu²⁺ exclusively triggered fluorescence quenching at 655 nm, whereas other common metal ions and anions had negligible effects. Moreover, the coexistence of these ions did not compromise the probe’s response to Cu²⁺, highlighting its superior selectivity and anti-interference performance. DCPO-Cu not only enabled sensitive detection of Cu²⁺ in water samples, food (rice, mung beans) and plant tissues (onion roots), but was also further adapted into convenient test strips for field use. Additionally, combined with a smartphone color recognition APP, it achieved on-site quantitative analysis of Cu²⁺ in soil samples. However, the fluorescence quenching mechanism restricted its use for live-cell imaging, underscoring the need for further development of probes tailored to biomedical applications.

Zhang et al. constructed the probe VIC using diethanolamine as the recognition unit [83]. The N and O atoms in diethanolamine, along with the phenolic hydroxyl group, coordinated with Cu²⁺, blocking ICT and causing fluorescence quenching at 675 nm. VIC demonstrated excellent sensing performance for Cu²⁺, including fast response (30 s), ultra-high sensitivity (LOD = 7 nM), excellent selectivity, and wide pH adaptability (5.0–12.0), enabling the rapid assessment of Cu²⁺ in soil samples. Furthermore, the probe has been successfully fabricated into a smartphone-assisted portable test kit for rapid, on-site detection of Cu²⁺ residues on the surface of traditional Chinese medicines (senna alexandrina mill) and vegetables (reed leaves, Chinese chives and garlic bolts) with satisfactory recoveries (89.5–108.7%). With the advantages of low cytotoxicity and good biocompatibility, VIC can image exogenous Cu²⁺ in the endoplasmic reticulum of MCF-7 cells, as well as in live zebrafish and nude mice. Notably, VIC enabled three-dimensional fluorescence imaging of AD mouse brains, revealing that Cu²⁺ levels progressively increased with disease progression, highlighting its potential for studying Cu²⁺-related neurodegenerative diseases.

3.1.2. NIR Fluorescent Probes Derived from Rhodamine Derivatives

Owing to their unique spirolactam structure, high fluorescence quantum yields, excellent photostability and large molar extinction coefficients, rhodamine derivatives have become an ideal platform for designing “turn-on” fluorescent probes. Classical rhodamines are typically colorless and have a non-fluorescent spirocyclic form. Upon binding to specific analytes (such as metal ions), the spirolactam ring opens, leading to significant fluorescence enhancement and color change. This mechanism has been widely exploited for the highly selective detection of Cu²⁺. However, conventional rhodamine-based probes generally emit in the visible region (550–590 nm) and have small Stokes shifts (20–40 nm), which limits their utility in deep-tissue imaging. To overcome this limitation, a series of novel rhodamine-based Cu²⁺ probes (Figure 7) with emission wavelengths red-shifted to the NIR region and large Stokes shifts have been developed through strategies such as extending the π-conjugation system, introducing strong electron-donating groups, or constructing a hybrid scaffold [84,85,86,87,88,89].

Figure 7. Structures of NIR fluorescent probes for Cu²⁺ based on rhodamine skeleton.

For example, based on the mechanism mentioned above, Guo, Xie and co-workers strategically constructed an NIR colorimetric and fluorometric probe 1(Cu²⁺), featuring a bis-spiro rhodamine B framework, in 2023 [84]. The probe alone showed negligible color and fluorescence. Upon interaction with Cu²⁺, irreversible spirolactam ring opening was induced, generating a dominant absorption band at 700 nm with intensity linearly dependent on Cu²⁺ concentration. Concurrently, two emission bands emerged at 580 nm (λ_ex = 520 nm) and 770 nm (λ_ex = 650 nm). The probe showed high sensitivity for Cu²⁺ with LODs of 0.57 μM (colorimetric method), 9 nM (fluorometric at 580 nm), and 28 nM (fluorometric at 770 nm). Mechanistic studies revealed that the Cu²⁺-coordinated spiro-ring opening was followed by hydrolysis and acylation reactions, generating a new compound with NIR absorption/emission characteristics. The practical utility of 1(Cu²⁺) was demonstrated by the successful fluorescence imaging of Cu²⁺ in HepG2 cells and in the mitochondria of paulownia sprout root tip cells.

Later, Yuan’s group established a distinctive rhodamine–hemicyanine platform in which Cu²⁺ coordination opened the spirolactam ring and suppressed PET, thereby switching on both color and NIR fluorescence [85,86,87,88]. In 2024, the group combined a hemicyanine dye with a rhodamine framework to synthesize the probe NRh-NP [85]. Upon coordination with Cu²⁺ in a 1:1 stoichiometry, the probe produced a pronounced “turn-on” fluorescence response at 750 nm (λ_ex = 660 nm) with a yellow-to-green color transition. NRh-NP performed effectively at pH 6.5–8.0 and showed negligible cytotoxicity (cell viability >98%), enabling imaging of exogenous/endogenous Cu²⁺ in MCF-7 cells and in living mice. Later that year, they further optimized the structure to obtain NRh6G-FA [86]. It enabled highly selective recognition of Cu²⁺ with strong interference resistance, rapid response (<60 s) and an LOD of 0.225 μM. Beyond cellular imaging, NRh6G-FA was successfully used for spiked recovery experiments and naked-eye detection of Cu²⁺ in Chinese herbal medicines (Astragalus and Paeonia lactiflora). Subsequently, NRK-BD was constructed on an extended π-conjugated xanthene–quinoline core using 2-pyridinecarboxaldehyde as the recognition unit. Coordination through the carbonyl O, imine N, and pyridine N atoms produced a strong emission at 730 nm, an LOD of 81 nM, and an ultrafast response of 1.5 s, while retaining EDTA-reversible behavior. NRK-BD also enabled test strip detection of Cu²⁺ on vegetable surfaces and smartphone-assisted quantitative analysis in real water samples with B/G values correlating linearly with Cu²⁺ (0–60 μM, LOD = 1.55 μM). With negligible cytotoxicity (cell viability > 96%), NRK-BD also achieved quantitative detection of intracellular Cu²⁺ in 4T1 cells, and fluorescence imaging of Cu²⁺ in living mice, with no significant attenuation of the fluorescence signal over time.

More recently, the same group obtained the probe NRTB by introducing ethyl phenothiazine group as a strong electron donor into the rhodamine framework. This modification effectively disrupted the electronic symmetry and bathochromically shifted the emission to 765 nm (λ_ex = 610 nm) [88]. Cu²⁺ recognition triggered a 268-fold fluorescence enhancement and a new absorption band at 600 nm, accompanied by a colorless-to-blue transition. NRTB demonstrated high selectivity towards Cu²⁺, with LODs of 22.6 nM by fluorescence and 43.28 nM by UV-Vis absorption. NRTB was successfully applied to smartphone-assisted visual quantification (LOD = 1.52 μM) and test strip-based rapid screening of Cu²⁺ residues on the surfaces of fresh produce. NRTB also demonstrated negligible cytotoxicity, with cell viability greater than 90%, enabling the effective imaging of Cu²⁺ in 4T1 cells as well as in mice, with an 80-fold signal enhancement in vivo.

Among these four probes, NRh-NP, NRh6G-FA, and NRK-BD are reversible and applicable over a certain pH range, and they could rapidly respond to Cu²⁺ in aqueous media (<60 s). In contrast, the original report on NRTB did not explicitly address its reversibility or pH tolerance. NRTB was used for Cu²⁺ detection in organic solvents and exhibited slower response kinetics, with a response time of 300 s. This slower response may be associated with its extended π-conjugated system and strong electron-donating phenothiazine moiety. Despite this, NRTB exhibited higher sensitivity. Collectively, these four probes illustrate a clear evolution from reversible, fast-responsive rhodamine–hemicyanine probes to a π-extended phenothiazine–rhodamine analogue with red-shifted emission and improved sensitivity, while also revealing a trade-off between extended conjugation and response kinetics.

Given the complexity of the pathological mechanisms underlying AD, the design of bifunctional probes capable of sequentially monitoring multiple related biomarkers is crucial for improving diagnostic accuracy. In 2025, Yan and colleagues rationally designed a panel of NIR fluorescent probes with a rhodamine spirolactam structure for sequentially detecting Cu²⁺ and Aβ42 aggregates [89]. LDMD-N exhibited an ultra-low LOD of 0.0543 μM for Cu²⁺ with rapid response (5 min), significantly outperforming the other candidates in the LDM series. Upon reaction with Cu²⁺, the ring opened to generate the compound LDMD, which could subsequently bind to Aβ42 aggregates, inducing an enhanced emission at 750 nm. LDMD-N was effectively used to monitor Cu²⁺ fluctuations in living cells and revealed that Cu²⁺ promoted Aβ aggregation in a zebrafish model. In vitro brain section staining and in vivo imaging experiments confirmed that LDMD-N could clearly distinguish AD model mice from wild-type controls, providing a powerful and versatile tool for elucidating the complex interplay between Cu²⁺ and Aβ aggregation in the pathogenesis of AD with potential applications in early diagnosis and therapeutic development.

3.1.3. NIR Fluorescent Probes Derived from Hemicyanine/Cyanine Derivatives

Hemicyanine/cyanine dyes have emerged as outstanding scaffolds for the development of NIR fluorescent probes for Cu²⁺ due to their excellent near-infrared optical properties, including high molar extinction coefficients, high fluorescence quantum yields, emission tunability with 650–900 nm, and good biocompatibility. These probes primarily respond to Cu²⁺ through two mechanisms: (1) fluorescence quenching caused by the paramagnetic nature of Cu²⁺ (off–on response); and (2) Cu²⁺-triggered chemical transformations (e.g., hydrolysis) that alter the probe structure, yielding pronounced spectral shifts suitable for ratiometric detection. More recently, researchers have reported various high-performance Cu²⁺ probes (Figure 8) by introducing various recognition groups into the hemicyanine/cyanine framework, demonstrating broad application prospects in subcellular structure imaging, in vivo imaging, and disease model studies [90,91,92,93,94,95].

Figure 8. Structures of NIR fluorescent probes for Cu²⁺ based on rhodamine skeleton.

On the common NIR fluorophore platform of cyanine, replacing the Cu²⁺ recognition unit allows for systematic modulation of the probe’s response mechanism and imaging performance. For example, in 2022, Wang’s group introduced a sulfhydryl group (-SH) into a cyanine scaffold and exploited sulfur donor-triggered PET to construct a turn-on NIR probe, CySH (λ_ex/em = 694/784 nm) [90]. Coordination of Cu²⁺ with –SH suppressed PET, causing a significant concentration-dependent increase in fluorescence. The 2:1 binding mode was identified by Job’s plot and HR-MS. CySH demonstrated favorable cell permeability, low cytotoxicity and mitochondria-targeting ability, enabling the fluctuation of labile copper in the mitochondria of HepG-2 cells to be monitored. Under copper-overload conditions, HepG-2 cells pretreated with NaVO₃ and then exposed to CuCl₂ exhibited a higher fluorescence intensity than cells exposed to CuCl₂ alone, indicating that CySH enables monitoring of mitochondrial copper accumulation. Subsequently, Guo’s team reported a probe IRPhen, constructed by incorporating the classic Cu²⁺ ligand 1,10-phenanthroline into a cyanine framework via a Schiff base linker [91]. IRPhen absorbed strongly at 771 nm and emitted at 808 nm (λ_ex = 750 nm). Coordination with Cu²⁺ triggered rapid fluorescence quenching, driven by the paramagnetic nature of Cu²⁺ and the PET effect. IRPhen displayed an LOD of 0.286 μM, a 1:1 complexation stoichiometry with a high affinity (Kₐ = 1.3 × 10⁶ M⁻¹), excellent selectivity, and good stability at physiological pH. Importantly, the probe was cell-permeable and enabled confocal imaging of dynamic Cu²⁺ in WS1 fibroblasts. These studies demonstrated that coordination-based recognition could achieve either turn-on or reversible turn-off responses, although turn-off modes may still suffer from background fluorescence interference.

To circumvent paramagnetic quenching by Cu²⁺ and improve quantification, subsequent studies shifted toward reaction-based recognition and ratiometric signal output. New et al. [92] exploited the Cu²⁺-catalyzed hydrolysis of 2-picolinate esters to construct an NIR ratiometric probe, CyCu1, by modifying a cyanine scaffold (derived from IR-780). Upon reaction with Cu²⁺, CyCu1 underwent hydrolysis followed by tautomerization to afford the ketone product CyK1, causing a dramatic hypsochromic shift exceeding 200 nm in both absorption and fluorescence emission, enabling ratiometric detection (I₆₀₀/I₈₁₀) with high sensitivity (LOD = 53 nM). CyCu1 not only imaged exogenous Cu²⁺ in SH-SY5Y cells and neuroblastoma tumor tissues, but also effectively distinguished chelator-treated tumor tissues from controls, crucially offering the first detection of endogenous labile Cu²⁺ in neuroblastoma tumors. Most importantly, CyCu1, for the first time, revealed a significant decrease in Cu²⁺ levels in the cytoplasm, neurons, and extraneuronal space of the substantia nigra in patients with dementia with Lewy bodies (DLB) (Figure 9). By attaching a Schiff base structure bearing phenolic hydroxyl and C=N groups to the same cyanine fluorophore as that in CyCu1 via an ester bond, a FRET-based fluorescent probe, NUST-Cy-1, was crafted [93]. Upon excitation at 450 nm, NUST-Cy-1 displayed dual emission peaks at 593 nm and 825 nm, corresponding to the donor and the acceptor, respectively. However, Cu²⁺ induced the hydrolytic cleavage of the Schiff base unit and interrupted the FRET process, decreasing the 825 nm signal and generating a new peak at 790 nm, allowing for ratiometric detection (I₈₂₅/I₇₉₀) with an LOD of 1.4 μM. The bioimaging results demonstrated the successful application of NUST-Cy-1 for ratiometric Cu²⁺ imaging in HeLa cells and living mice.

Figure 9. Representative fluorescence images of CyCu1 for Cu²⁺ in the SN of human brain tissues from DLB patients and the control group. (a–d) Representative images of control SN. (e–h) Representative images of DLB SN. (i) Ratio of quantified fluorescence intensity in CH1 (580–650 nm) and CH2(750–800 nm) in nuclei, cytoplasm, whole cell, an extraneuronal space of control SN and DLB SN. The error bar represents SEM. * p < 0.05. Yellow arrow: nucleus; red arrow: cytoplasm. Reproduced with permission from Ref. [93]. Copyright (2024) American Chemical Society.

Overall, these four probes share the same cyanine fluorophore but employ different recognition units (SH, 1,10-phenanthroline, 2-picolinate ester, and Schiff base) to achieve PET turn-on, reversible turn-off, reaction-based ratiometric detection, and FRET ratiometric detection, respectively, extending Cu²⁺ imaging from the cellular level to mitochondria, tumor tissues, brain tissues, and living animals.

During the same period, exploiting the synthetic versatility of hemicyanine scaffolds, researchers conjugated hemicyanine to xanthene dyes via a C=C bridge to construct NIR fluorophores. The subsequent introduction of different Cu²⁺-specific recognition units afforded a series of high-performance NIR fluorescent probes. For instance, in 2025, a Schiff base moiety was introduced into a hemicyanine–xanthene scaffold as the Cu²⁺ recognition unit, yielding the Cu²⁺-responsive NIR fluorescent probe HX [94]. Among all the tested metal ions, only Cu²⁺ induced over 90% fluorescence quenching at 810 nm (λ_ex = 730 nm) within 60 s. HX showed a linear concentration-dependent fluorescence decrease upon Cu²⁺ binding, achieving an LOD of 0.54 μM. Taking advantage of its superior optical performance, HX was successfully used to visualize Cu²⁺ in H22 tumor-bearing mice. The results indicated that HX not only enabled visualization of Cu²⁺ accumulation in the liver for hepatotoxicity assessment but also allowed for the monitoring of tumor Cu²⁺ uptake to evaluate copper transporter (Ctr) function, offering a novel potential strategy for predicting tumor sensitivity to platinum-based drugs.

Recently, Li and colleagues incorporated a tris(2-pyridylmethyl)amine (TPA) unit into a hemicyanine–xanthene fluorophore analogous to that of HX and designed a reaction-based NIR fluorescent probe HC-TPA, achieving the cascade sensing of Cu²⁺ and GSH in complex food and biological systems, [95]. Upon rapid coordination with Cu²⁺, the fluorescence of HC-TPA at 668 nm was sharply quenched within 1 min. Subsequently, GSH triggered Cu²⁺-mediated oxidative cleavage, releasing the strongly NIR-emitting fluorophore HC, which restored the fluorescence at 696 nm within a response time of 20 min. DFT calculations revealed that HC-TPA exhibited weak intrinsic fluorescence due to the presence of a “dark” S₁ state; Cu²⁺ coordination further suppressed the fluorescence via a CHEQ mechanism, while GSH achieved a fluorescence “turn-on” response through reductive oxidative cleavage. The excellent analytical performance, including linear detection ranges of 100–800 nM for Cu²⁺ (LOD = 48 nM) and 6–30 μM for GSH (LOD = 0.56 μM), rapid response kinetics, and exceptional selectivity over competing metal ions, biogenic amines, and other thiol-containing biomolecules such as Cys and Hcy, enables spiked recovery tests of Cu²⁺ and GSH in food matrices (e.g., apples, grapes, and tomatoes), as well as fluorescence imaging of Cu²⁺ and GSH in HEK293T cells. Herein, the probe offers a promising platform for food safety monitoring and bioanalytical applications involving interconnected metal–thiol systems, leveraged by a mechanistically driven tandem sensing strategy.

3.1.4. NIR Fluorescent Probes Derived from Other NIR Fluorophores

In addition to the aforementioned common near-infrared (NIR) fluorophores, researchers have also explored other types of NIR dyes for the construction of fluorescent probes for Cu²⁺, as shown in Figure 10 [96,97,98]. Chen et al. introduced a picolinate moiety as the reaction site for Cu²⁺ into the NIR fluorophore 3-Dichloro-7-hydroxy-9,9-dimethylacridin-2(9H)-one (DDAO) to develop DDAO-Cu [96]. DDAO-Cu showed weak fluorescence owing to the inhibition of the ICT effect by the 2-picolinyl group. However, upon Cu²⁺-triggered hydrolysis of the picolinate ester, the hydroxyl group was released and the ICT effect was restored, causing a turn-on fluorescence response at 672 nm. DDAO-Cu exhibited ultra-high sensitivity (LOD = 1.8 nM) and selectivity towards Cu²⁺ with rapid response (<3 min) and favorable tissue penetration. DDAO-Cu was successfully used to image Cu²⁺ in cellular and zebrafish models of Parkinson’s disease (PD). Notably, DDAO-Cu demonstrated good blood–brain barrier permeability, achieving for the first time in situ imaging and differential analysis of Cu²⁺ levels in different brain regions (cortex, hippocampus, striatum, and substantia nigra) of Parkinson’s disease mice, revealing the complexity of copper distribution in brain regions under disease conditions.

Figure 10. Mechanisms of Cu²⁺ sensing by DDAO-Cu (A), DCX-Cu (B) and NS-5 (C).

Based on an identical recognition mechanism, Zhang’s group constructed DCX-Cu by introducing a 2-pyridinecarbonyl group into a benzopyranonitrile–xanthene scaffold [97]. Upon the addition of Cu²⁺, a strong fluorescence turn-on signal at 746 nm was observed for DCX-Cu, displaying a linear correlation with Cu²⁺ concentration (0–50 mM). Owing to its superior anti-interference performance and outstanding sensitivity (LOD = 19.47 nM), DCX-Cu enabled reliable Cu²⁺ detection in real water samples, with recoveries of 98–103%, and imaging of exogenous Cu²⁺ in HepG2 cells.

In 2025, Fang’s group introduced NS-5, an NIR probe for Cu⁺, featuring a tetradentate thioether ligand [98]. The PET-induced weak fluorescence was restored upon 1:1 Cu⁺ coordination, enabling rapid (>13-fold) enhancement at 720 nm with an LOD of 32 nM. NS-5 demonstrated good photostability, functioned effectively in aqueous media, and specifically localized to the mitochondria with strong colocalization with Mito-Tracker Green (PCC = 0.90). Cell imaging studies confirmed that the probe could monitor both endogenous and exogenous Cu⁺ fluctuations in HepG2 cells and differentiate Cu⁺ levels between hepatocellular carcinoma (HepG2) and normal liver (L02) cells. Moreover, NS-5 was effectively applied to in vivo imaging of labile Cu⁺ pools in living mice, successfully visualizing elevated Cu⁺ levels upon Cu(NO₃)₂ supplementation and reduced signals upon treatment with the copper chelator ATN-224.

3.2. NIR Fluorescent Probes for Hg²⁺

Among the most hazardous heavy metal pollutants, mercury ions (Hg²⁺) originate predominantly from anthropogenic sources, including fossil fuel combustion, mining, and industrial discharge [99]. Hg²⁺ exhibits strong bioaccumulation and persistence, with methylmercury (CH₃Hg⁺) capable of biomagnifying millions-fold through the food chain [100]. Owing to its strong affinity for thiol groups present in biomolecules, even exposure to trace amounts of Hg²⁺ are sufficient to cause irreversible damage to the neurological, renal, and immune systems [101]. Consequently, the development of highly sensitive and selective methods for detecting Hg²⁺ in environmental and biological samples is of critical importance. Recently, various NIR fluorescent probes based on rhodamine, isophorone, DCM, heptamethine cyanine scaffolds, etc., have been developed to detect Hg²⁺. The photophysical properties of representative probes are summarized in Table 2.

Table 2. The photophysical properties of NIR fluorescent probes for Hg²⁺.

Name	Solvent	Ex (nm)	Em (nm)	Sensing Mechanism	Response Time	Reversible	Signal Transduction Mode	Detection Limit	pH Range	Application	Ref.
HgS	PBS/DMSO = 7/3, v/v	430	645	ICT	3 min	No	turn-on	64 nM	6.5–8.5	Bottled water, river water, tap water, deionized water, HeLa cells, mice	[102]
CL(Hg2+)	EtOH/H2O = 1/1, v/v	379	695	ICT	<2 min	-	turn-on	0.83 mM	2.0–12.0	U87MG cells	[103]
N-FP	EtOH/H2O = 9/1, v/v	498	661	ICT	1 min	No	turn-on	0.72 mM	1.0–10.0	-	[104]
DSX	EtOH/PBS = 1/9, v/v	470	670	TICT	-	No	Turn-off	6.57 nM	-	Dawen river water, Yellow river water, Jinxiu river water, fish, apple, HeLa cells, zebrafish	[105]
NIRPC-Hg	HEPES/ACN = 9/1, v/v	420	666/535	ICT-ESIPT	120 min	No	ratiometric	9.21 nM	-	Tap water, Dawen river water, Yellow river water, HepG2 cells, zebrafish	[106]
YF-Hg	DMF/PBS = 5/5, v/v	468	659	ICT	12 min	No	turn-on	0.297 mM	7.0–10.0	Mouse fibroblasts	[107]
W-2a	DMSO/PBS = 1/1, v/v	500	656	ICT	25 min	No	turn-on	97 nM	-	EMT6 cells, mice	[108]
THI	MeCN/HEPES = 2/8, v/v	500	645	ICT	-	Yes	turn-on	7.22 mM	-	HepG2 cells	[109]
1(Hg2+)	MeCN/HEPES = 4/1, v/v	440	660	PET	30 s	Yes	turn-on	5.6 nM	7.5–9.5	Pond water, river water, HepG2 cells	[110]
RBLY	HEPES/EtOH = 5/1, v/v	620	695	Hg2+-induced spirolactam ring opening	120 s	Yes	turn-on	0.34 mM		Pure water, Mo Chou lake water, Yin Xing lake water, MCF-7 cells, zebrafish, soybean rhizoma	[111]
RBLS	EtOH/HEPES = 1/3, v/v	620	685	Hg2+-induced spirolactam ring opening	210 s	Yes	turn-on	0.26 mM	5.0–10.0	Xuanwu Lake water, Zixi Lake water, HeLa cells, zebrafish	[112]
CCS-Hg	PBS/EtOH = 3/2, v/v	680	780	Hg2+-induced spirolactam ring opening	<30 s	-	turn-on	73 nM	5.0–9.0	A549 cells	[113]
Rho-Hg	PBS/EtOH = 1/1, v/v	610	665	Hg2+-induced spirolactam ring opening	<1 min	-	turn-on	17.5 nM	2.0–10.0	Taozi lake water, Houhu lake water, tap Wate, HeLa cells, zebrafish	[114]
JZ-1	PBS	700	728	Hg2+-induced spirolactam ring opening	30 min	No	turn-on	33 nM	-	HeLa cells, zebrafish, mice	[115]
CSME	ACN/HEPES = 5/1, v/v	660	711	Hg2+-induced spirolactam ring opening	90 s	-	turn-on	4.93 nM (fluorescence) 58.9 nM (UV)	-	Tap water, drinking water, 3T3 cells	[116]
CIME	ACN/HEPES = 5/1, v/v	719	750	spirocyclic Hg2+-induced spirolactam ring opening	30 s	-	turn-on	0.65 nM (fluorescence) 37 nM (UV)	-	Tap water, drinking water, lake water, MCF-7 cells, 3T3 cells	[117]
NIR8	DMF/HEPES = 5/1, v/v	670	755	Hg2+-induced spirolactam ring opening	15 s	No	turn-on	11.3 nM (fluorescence) 49.3 nM (UV)	-	Tap water, drinking water, lake water, IH/3T3 cells, MCF-7 cells	[118]
NIR9	ACN/Tris = 1/1, v/v	715	750	Hg2+-induced spirolactam ring opening	10 s	-	turn-on	0.02 mM (fluorescence) 0.01 mM (UV)	-	Tap water, drinking water, lake water, cocoa, mushroom, sunflower seeds, A549 cells	[119]
NIR-2	Acetone/HEPES = 4/1, v/v	719	755	Hg2+-induced spirolactam ring opening, hydrolyzation	10 s	No	turn-on	0.377 nM (fluorescence) 4.54 nM (UV)	1.0–9.0	Tap water, drinking water, lake water, strawberry, watermelon, cherry, rice, lentil, green lentils, beans, corn, A549 cells, H1299 cells	[120]
QR	PBS/DMSO = 1/1, v/v	630/410	725/465	Hg2+-induced spirolactam ring opening	1 min	No	ratiometric	23 nM	5.0–8.0	Tap water, HeLa cells	[121]
Se-SiRH	MeOH/PBS= 1/1, v/v	616	672	Hg2+-induced spirolactam ring opening	40 min	-	turn-on	2.1 nM	3.0–9.0	Tap water, river water, HepG2 cells	[122]
CyP	CH3OH/PBS = 3/7, v/v	680	710	ICT	<10 min	-	turn-on	54.8 nM	7.0–8.0	SHSY-5Y cells, zebrafish, mice	[123]
CyHg-Cl	PBS/DMSO = 4/1, v/v	680	725	ICT	2.5 min	No	turn-on	0.59 µM	5.0–9.0	Tap water, Yangtze river water, distilled water, HeLa cells	[124]
HCYS	PBS	615	674	ICT	<1 min	Yes	turn-off	1.41 mM	-	HepG2 cells, HeLa cells, zebrafish, pea spouts, lake water, tap water, ultrapure water	[125]
HDS-Hg	DMSO/PBS = 1/4, v/v	700	765	ICT	3 min	No	turn-on	89 nM	-	Environmental water samples, paeonia lactiflora, panax quinquefolius glycyrrhiza uralensis, seafood samples, tobacco seedling leaves, HepG2 cells, THLE-2 cells, A549 cells, BRL-3A cells, mice	[126]
HTC-S	DMSO/PBS = 1/1, v/v	500/740	625/816	ICT	40 s	No	ratiometric	0.10 µM	-	Sandy soil, plant soil, carrot, radish, lettuce, rice, mung beans, shrimp, tap water, sewage water, distillate water, HeLa cells	[127]
CB-Hg	EtOH/HEPES = 1/1, v/v	670/430	720/495	CHEF	30 s	No	ratiometric	6.94 nM	2.0–10.0	Tap water, mineral water, rice, HeLa cells, zebrafish	[128]
DCM-Hg	DMSO/PBS= 5/5, v/v	550	695	ICT	25 min	No	turn-on	3.35 µM	5.0–9.0	HepG2 cells	[129]
TM-ND-Hg	DMSO/PBS = 7/3, v/v	611/436	794/516	ICT	<1 min	No	ratiometric	0.39 mM	6.0–9.0	A549 cells, zebrafish, mineral Water, tap water, lake water	[130]
TP-Hg	DMSO/PBS = 7/3, v/v	563	770	ICT	<1 min	No	turn-on	0.13 mM	6.0–9.0	A549 cells, zebrafish	[131]
BBN-Hg	DMSO:PBS = 3/7, v/v	560	650	ICT	-	No	turn-on	53 nM	5.8–8.0	Tap water, Chao lake water, tea water, scutellaria baicalensis water, loess soil, green soil, purple soil, sandy soil, MCF-7 cells, zebrafish, roots of rice and Salvia miltiorrhiza	[132]

Note: Probe names are shown in bold.

3.2.1. NIR Fluorescent Probes Derived from Dicyanoisophorone Derivatives

Leveraging the specific chemical reactivity of Hg²⁺, notably its ability to promote thioacetal/thioketal desulfurization and thiocarbonate hydrolysis, as well as principles of hard and soft acid–base (HSAB) theory, researchers have introduced various specific recognition groups (e.g., thioacetals/thioketals, thiocarbonates, and N/O/S-containing ligands) into the readily modifiable isophorone skeleton, which possesses desirable NIR luminescent properties. As illustrated in Figure 11, this strategy has enabled the creation of numerous superior fluorescent probes for Hg²⁺ detection [102,103,104,105,106,107,108,109,110].

Figure 11. Structures of NIR fluorescent probes for Hg²⁺ based on dicyanoisohporone platform.

Exploiting the high thiophilicity of Hg²⁺ to induce thiocarbonate hydrolysis, thereby triggering a fluorescence “turn-on” response, is a classic design strategy for Hg²⁺ probes. In 2022, Zhang and Zhou’s group presented the probe HgS by introducing a thiocarbonate moiety as the specific Hg²⁺ recognition unit into a dicyanoisophorone scaffold [102]. HgS exhibited fluorescence due to the PET effect. Upon specific reaction with Hg²⁺, the thiobenzoate moiety was cleaved and the ICT process was restored, triggering a pronounced emission increase at 645 nm (l_ex = 580 nm). HgS exhibited excellent selectivity and anti-interference, high sensitivity (LOD = 64 nM) towards Hg²⁺, and good biocompatibility, enabling successfully fluorescence imaging of exogenous Hg²⁺ in HeLa cells and living mice. Furthermore, HgS was effectively applied to the quantitative determination of Hg²⁺ in real water samples with satisfactory recoveries (83.4–108.6%).

Later, using the same mechanism mentioned above, Jiang et al. obtained the NIR fluorescent probe CL(Hg²⁺) by covalently conjugating a coumarin fluorophore with a dicyanoisophorone unit via a C=C bond to construct the NIR emissive skeleton, while introducing a thiocarbonate moiety as the responsive site for Hg²⁺/Ag⁺ [103]. Under excitation at 379 nm, CL(Hg²⁺) was weakly emissive at 695 nm. Upon addition of Hg²⁺ or Ag⁺, a strongly emissive coumarin–dicyanoisophorone fluorophore was released, which simultaneously triggered a remarkable emission enhancement and a visible color transition. CL(Hg²⁺) offers an exceptionally large Stokes shift (316 nm), outstanding selectivity, and good sensitivity, with LODs of 0.83 μM for Hg²⁺ and 8.8 μM for Ag⁺. Moreover, CL(Hg²⁺) features a rapid response to Hg²⁺ (<3 min), a wide pH tolerance (pH 2–12), and favorable biocompatibility (cell viability >90% even at 80 μM), enabling its successful application for in vitro fluorescence imaging of Hg²⁺ in U87MG cells, where the fluorescence intensity exhibits a concentration-dependent relationship with Hg²⁺.

Following a similar design strategy, in 2024, Li et al. constructed the probe N-FP [104]. Upon addition of Hg²⁺/Ag⁺ to N-FP solution, a distinct a fluorescence turn-on at 661 nm was observed, along with a bathochromic shift and a color change. N-FP exhibited LODs of 0.72 μM for Hg²⁺ and 1.1 μM for Ag⁺, rapid response (1–3 min), excellent selectivity, and efficient performance across an exceptionally wide pH range of 1–10. These features enabled its successful application for colorimetric and fluorimetric detection of Ag⁺ and Hg²⁺. N-FP was also investigated in parallel by another group [105]. Interestingly, in this report, the probe was developed for Hg²⁺-selective detection, featuring strong anti-interference performance, and an ultra-low LOD of 6.57 nM, enabling successful application in water and food sample analysis with satisfactory recoveries (93.79–101.80%). Benefiting from its low cytotoxicity and good photostability, the probe was further extended to visualize Hg²⁺ in HeLa cells and zebrafish via fluorescence imaging. After incubation with Hg²⁺, the intracellular red fluorescence was remarkably attenuated.

Based on a synergistic ICT-ESIPT mechanism, in 2025, Zhu and co-workers constructed NIRPC-Hg by integrating a benzothiazole ESIPT unit, a dicyanoisophorone ICT unit, and a thiocarbonate moiety as the Hg²⁺-specific recognition group for ratiometric fluorescence detection of Hg²⁺ [106]. The probe was initially emitted at 535 nm. However, upon Hg²⁺-promoted hydrolysis, the phenol was released, restoring the ICT-ESIPT pathway and shifting the emission to 665 nm, enabling ratiometric detection with an ultra-low LOD (9.21 nM). Using NIRPC-Hg, Hg²⁺ was successfully detected in environmental water samples and exogenous Hg²⁺ was imaged ratiometrically in HepG2 cells and zebrafish.

Beyond Hg²⁺-induced thiocarbonate hydrolysis, thioacetal deprotection is another key strategy for probe design. Guided by this mechanism, Chen’s group (2023) constructed a dual-channel colorimetric and NIR fluorescent probe YF-Hg by introducing a thioacetal moiety (a specific reaction site for Hg²⁺) into the malononitrile isophorone backbone [107]. After the addition of Hg²⁺, YF-Hg underwent a desulfurization reaction to generate the aldehyde precursor YF-CHO, causing a 12-fold fluorescence turn-on at 659 nm. Meanwhile, the probe exhibited a blue shift in absorption along with a distinct violet-to-red color variation. Selectivity experiments demonstrated that, apart from Hg²⁺, other common ions, bioactive molecules, H₂S, and H₂O₂ induced no significant spectral changes in YF-Hg. Moreover, their coexistence with Hg²⁺ did not interfere with Hg²⁺ detection by YF-Hg. In addition, YF-Hg exhibited a large Stokes shift (191 nm), favorable membrane permeability, and low cytotoxicity (mouse fibroblast viability >80% at 100 μM), enabling successful fluorescence imaging of Hg²⁺ in mouse fibroblasts, where intense red fluorescence signals were observed in the group treated with Hg²⁺. Furthermore, dual-colorimetric test strips were developed based on its color response, allowing for naked-eye, semi-quantitative visual detection of Hg²⁺.

Han and colleagues (2023) rationally designed W-2a by incorporating a phenylboronic acid pinacol ester as an ONOO⁻ recognition unit into a dithioacetal-based Hg²⁺ probe for the sequential and discriminative detection of Hg²⁺ and peroxynitrite (ONOO⁻) [108]. W-2a exhibited weak fluorescence due to hindered ICT. Hg²⁺ triggered dithioacetal deprotection, producing fluorescence at 653 nm, while subsequent ONOO⁻ cleavage of the phenylboronate ester further intensified the emission at 656 nm, facilitating stepwise detection at 656 nm of Hg²⁺ and ONOO⁻. The detection limits were 1.46 μM for Hg²⁺ and 97 nM for ONOO⁻ (Hg²⁺ first) or 7.25 nM and 0.12 μM (ONOO⁻ first). W-2a effectively enabled multi-channel fluorescence imaging of EMT6 cells, allowing us to monitor intracellular Hg²⁺ and ONOO⁻ levels. Furthermore, W-2a was utilized for imaging endogenous ONOO⁻ in an LPS-induced mouse model of inflammation. Compared with normal tissue in the right leg, the inflammatory region in the left leg exhibited significantly enhanced fluorescence signals. Coupled with a smartphone-based portable device and a WeChat mini-program (“Colorimetric acquisition”), the probe exhibited a good linear relationship between the R/B ratio and Hg²⁺ concentration (R² > 0.9) in both fluorescence and UV–vis dual modes, thereby providing an intelligent, on-site analytical platform for rapid qualitative and quantitative analysis of Hg²⁺ in the environment. Overall, W-2a serves as a versatile tool for studying Hg²⁺-induced oxidative stress and its downstream peroxynitrite signaling in living systems.

Apart from utilizing specific chemical reactions, Hg²⁺ fluorescent probes can also be developed through the coordination of Hg²⁺ with nitrogen, oxygen, and sulfur atoms, as guided by Pearson’s HSAB theory. Based on this approach, in 2022, Erdemir et al. synthesized an NIR probe THI derived from a thiazoline–isophorone framework [109]. A significant fluorescence turn-on response at 645 nm was elicited upon Hg²⁺ coordination with the thiazoline N/S atoms, which disrupted hydrogen bonding between water molecules and the diethylamino group, restored the ICT pathway, and resulted in an ~11-fold increase in quantum yield (from 0.0106 to 0.1177), a slight blue shift, and an LOD of 7.22 μM. The recognition process was reversible, as demonstrated by S²⁻-induced fluorescence quenching. Its practical utility has been validated through portable TLC test strip assays for naked-eye Hg²⁺ detection and successful application in real water sample analysis. Additionally, THI showed good permeability and low cytotoxicity. Cytotoxicity assays showed no obvious toxic effects on either HeLa or Hep2 cells, and the probe was applied to fluorescence imaging of Hg²⁺ in living HeLa cells, where bright red fluorescence signals were observed.

In 2024, Yan’s group constructed the fluorescent probe 1(Hg²⁺) via the condensation reaction of dicyanoisophorone aldehyde with 2-pyridinecarbohydrazide [110]. Upon coordination with Hg²⁺ in a 1:1 stoichiometry, the PET process was inhibited, generating a turn-on NIR signal at 660 nm together with a linear detection range of 0–7 μM. Mechanistic studies revealed that coordination occurred exclusively through the pyridine nitrogen, while the hydroxyl and amide groups were not involved. Within the physiological pH range (7.5–9.5), Probe 1 offered remarkable selectivity (Ka = 1.63 × 104 M⁻¹), excellent sensitivity (LOD = 5.6 nM), a 220 nm Stokes shift and a rapid equilibration time (30 s). The successful imaging of Hg²⁺ in HepG2 cells and reliable recovery tests in real water samples further confirmed its practical applicability.

3.2.2. NIR Fluorescent Probes Derived from Rhodamine Derivatives

Rhodamine dyes, characterized by their unique spirocyclic lactam structure, high fluorescence quantum yields, and excellent photostability, have emerged as an ideal platform for constructing “turn-on” fluorescent probes. In recent years, researchers have successfully extended the emission wavelengths of rhodamine derivatives into the NIR region by expanding the π-conjugation system or introducing heterocyclic structures, thereby significantly enhancing their application potential in bioimaging [111,112,113,114,115,116,117,118,119,120,121,122], as shown in Figure 12.

Figure 12. Structures of NIR fluorescent probes for Hg²⁺ based on rhodamine derivatives.

In 2022, utilizing a coumarin–rhodamine fused dye as the fluorophore, Kan’s group reported a novel fluorescent probe RBLY, via acyl chlorination, phenylhydrazine coupling, and Lawesson’s reagent sulfurization [111]. The probe exhibited negligible emission due to the dominant spirolactam conformation. Upon exposure to Hg²⁺, the spirocycle was selectively cleaved, triggering a pronounced turn-on response at 695 nm, with a detection limit of 0.34 μM. The 1:1 binding mode was confirmed by Job’s plot analysis and mass spectrometry and the binding constant was determined to be 1.63 × 10⁵ M⁻¹. The recognition of Hg²⁺ by RBLY was reversible (quenchable by S²⁻), and the system reached equilibrium within approximately 120 s. RBLY was capable of detecting Hg²⁺ in the actual environmental water samples. More importantly, the successful visualization of Hg²⁺ in MCF-7 cells, soybean root tissue, and zebrafish was achieved by using RBLY, underscoring its wide applicability in both plant and animal systems.

Subsequently, Kan’s group reported another structurally similar NIR fluorescent probe RBLS [112]. Complexation with Hg²⁺ in a 2:1 stoichiometric ratio triggered a prominent emission turn-on at 685 nm, yielding an LOD of 0.26 μM. Similar to RBLY, the recognition of Hg²⁺ by RBLS was reversible, and the probe proved effective for quantifying trace Hg²⁺ in environmental water samples and for fluorescence imaging in biological models. Notably, leveraging the distinct colorimetric response of RBLS, the research team further developed a test strip for naked-eye visualization of Hg²⁺, which exhibited a color change from slight yellow to blue as the Hg²⁺ concentration increased. More importantly, by integrating the test strip with smartphone photography and RGB color analysis, quantitative detection of Hg²⁺ was achieved with an LOD of 1.87 μM, thereby offering a practical tool for rapid on-site monitoring of Hg²⁺.

Liang et al. introduced a sulfur atom into chromenylium-cyanine dye via sodium sulfide treatment to yield CCS-Hg, which retained the classical “off–on” response characteristics of traditional rhodamines while offering the advantages of near-infrared excitation and emission (λ_ex/em = 680/780 nm) [113]. Upon coordination with Hg²⁺ in a 2:1 stoichiometric ratio, CCS-Hg displayed a strong NIR emission at 780 nm, along with a marked increase in absorbance at 680 nm and 740 nm, thereby achieving both fluorometric and colorimetric detection. Spectroscopic investigations revealed that CCS-Hg exhibited excellent selectivity and high sensitivity for Hg²⁺ detection, achieving detection limits of 73 nM (fluorometric mode) and 12 nM (colorimetric mode), and quick response (<30 s), under conditions at pH 5–9. Benefiting from these advantages, CCS-Hg was successfully employed for fluorescence imaging of Hg²⁺ distribution in living A549 cells.

Similar to CCS-Hg, Yuan’s group successfully engineered Rho-Hg by introducing a thiolactone moiety as the Hg²⁺-responsive switch into the YLF dye platform [114]. Rho-Hg showed a dramatic 14-fold fluorescence surge at 665 nm upon Hg²⁺ addition, with a low LOD of 17.5 nM, a rapid equilibration time (<1 min), and excellent pH tolerance spanning from 2 to 10. Colocalization experiments revealed that the red fluorescence signal was predominantly localized within mitochondria and exhibited substantial overlap with the green fluorescence of Mito-Tracker Green (Pearson’s correlation coefficient = 0.8), indicating that Rho-Hg possessed mitochondrial targeting capability. Owing to these merits, Rho-Hg was successfully utilized not only for imaging mitochondrial Hg²⁺ in HeLa cells but also for dynamically tracking Hg²⁺ accumulation and migration during zebrafish development. Fluorescence imaging revealed that as zebrafish grew to maturity, the red fluorescence signal gradually translocated from the embryonic chorion to the liver and intestine of the larvae, indicating the progressive transfer of Hg²⁺ to these organs.

More recently, Zhou’s team presented a fluorescent probe JZ-1, structurally analogous to CCS-Hg [115]. The probe JZ-1 featured dual NIR analytical wavelengths, with both excitation and emission maxima exceeding 700 nm (λ_ex = 700 nm, λ_em = 728 nm), which minimized background autofluorescence and enhances tissue penetration. Upon reaction with Hg²⁺, the solution of JZ-1 displayed a pronounced turn-on fluorescence signal at 780 nm (>200-fold), accompanied by significantly enhanced absorption bands at 663 nm and 722 nm. JZ-1 demonstrated high selectivity towards Hg²⁺, a low detection limit of 33 nM, and stable performance across a broad pH range (3.7–10.0) and temperature range (25–45 °C). Cell imaging studies showed that JZ-1 selectively accumulated in the mitochondria and effectively visualized exogenous Hg²⁺ in HeLa cells, while zebrafish experiments further validated its capacity for monitoring Hg²⁺ accumulation in vivo. Notably, the probe was effectively applied to image gastrointestinal damage in an acute mercury poisoning mouse model, intuitively revealing Hg²⁺ accumulation in the gastrointestinal tract. Additionally, H&E staining confirmed the mitigating effect of sodium selenite on mercury-induced toxicity.

Moreover, based on the chromenylium-cyanine platform, Ismail Yilmaz’s group systematically constructed a family of structurally analogous NIR fluorescent probes for sensing Hg²⁺, with progressive optimization through different recognition units (methionine, mercaptopropanehydrazide, thiourea, selenolactone), achieving improved sensitivity, response speed, and versatility for environmental and biological applications [116,117,118,119,120]. In 2023, they reported CSME, a methionine-modified probe that responded to Hg²⁺ via ratiometric and turn-on modes [116]. Upon Hg²⁺ addition, CSME exhibited fluorescence enhancement at 740 nm (linear range: 9.49 × 10⁻⁹–1.13 × 10⁻⁵ M) and a ratiometric absorbance change (A₇₁₁/A₃₈₅) over 6.29 × 10⁻⁸–1.86 × 10⁻⁴ M, together with a distinct colorimetric change from pale yellow to green. The sensing system reached equilibrium in just 90 s, offering LOD values of 58.9 nM (UV-Vis) and 4.93 nM (fluorescence). The practical applicability of CSME was further demonstrated through ratiometric/turn-on assays in real water, smartphone-assisted colorimetric analysis, and NIR fluorescence imaging of 3T3 cells.

In 2024, they developed CIME, a structurally similar probe with faster response (30 s) and higher sensitivity (LOD = 0.65 nM), enabling dual-mode detection in water samples, smartphone analysis, and imaging in MCF-7 and 3T3 cells [117]. In the same year, they reported NIR8, featuring a mercaptopropanehydrazide recognition group, which exhibited the fastest response (15 s) and detection limits of 49.3 nM (UV-Vis) and 11.3 nM (fluorescence) [118]. NIR8 was applied to water sample analysis, smartphone RGB detection, and imaging in MCF-7 and NIH/3T3 cells. In 2025, they introduced NIR9, a dual chemosensor capable of independently detecting Hg²⁺ (turn-on fluorescence at 750 nm, LOD = 20 nM) and Cu²⁺ (turn-off UV-Vis at 380 nm, LOD = 10 nM). The Hg²⁺ response involves desulfurization-induced spirolactam ring opening, while Cu²⁺ simply coordinates to the thiourea moiety without ring opening [119]. NIR9 has been successfully applied in environmental water, food samples, and A549 cell imaging. Most recently (2026), they designed NIR-2, a selenolactone-based probe that displayed strong turn-on fluorescence at 755 nm (LOD = 4.54 nM) and a ratiometric UV-Vis response (LOD = 0.377 nM) to Hg²⁺ [120]. NIR-2 featured an ultra-fast response (10 s), the widest linear range (1–10 nM), and the most sensitive smartphone-based detection (linear range 1.66 nM-7.50 μM). The researchers successfully employed NIR-2 for Hg²⁺ detection in water, food, and legume samples, and also utilized it for fluorescence imaging in A549 and H1299 cancer cells. All probes operated via the Hg²⁺-induced spiro ring opening mechanism, offering NIR emission, high sensitivity, fast response, dual-functional detection, and cell imaging capabilities, making them effective tools for convenient Hg²⁺ detection in environmental and biological samples.

A quinoline–rhodamine-based ratiometric NIR probe QR was engineered by Wang’s team, featuring a phenylthio ester group for selective Hg²⁺ recognition [121]. QR itself emitted visible fluorescence at 465 nm. Only Hg²⁺ promoted a spirolactam ring opening reaction, which induced significant expansion of molecular conjugation and generated a 1,3,4-oxadiazole moiety. This structural transformation triggered a ratiometric response (I₇₂₅/I₄₆₅) with a remarkable 354-fold enhancement. The probe possessed exceptional selectivity, fast response (<1 min), a remarkable Stokes shift, and a low LOD of 23 nM for sensing Hg²⁺. Leveraging its excellent optical properties, QR was effectively utilized to monitor Hg²⁺ in tap water with satisfactory recoveries. Moreover, QR demonstrated negligible cytotoxicity (cell viability >90% at 10 mM) and excellent cell-penetrating ability, facilitating the dual-channel ratiometric fluorescence visualization of Hg²⁺ in live HeLa cells, where the red channel fluorescence signal increased greatly while the green channel signal decreased upon the addition of Hg²⁺.

Recently, Wang and co-workers pioneered the development of an NIR probe Se-SiRH, featuring a selenophene-functionalized silicon–rhodamine B scaffold, leveraging the excellent chelation ability of the selenium atom toward Hg²⁺ [122]. The probe itself was colorless and non-fluorescent. Upon coordination with Hg²⁺ in a 2:1 stoichiometric ratio, Se-SiRH underwent spirocyclic opening, producing a >75-fold fluorescence turn-on at 672 nm and a colorless-to-pink color change. The sensing mechanism, involving Hg²⁺-triggered spirolactone opening facilitated by coordination via the carbonyl oxygen and selenophene selenium atom, was further validated by DFT calculations, which revealed a reduced HOMO-LUMO energy gap upon complexation. Owing to its superior selectivity and ultra-high sensitivity (LOD = 2.1 nM), the applicability of Se-SiRH was further extended to Hg²⁺ detection in environmental water matrices and confocal imaging in HepG2 cells.

3.2.3. NIR Fluorescent Probes Derived from Hemicyanine/Cyanine Dyes

NIR fluorescent probes constructed from a hemicyanine/cyanine scaffold have attracted considerable attention for detecting Hg²⁺ owing to superior photophysical properties. By exploiting the high thiophilicity of Hg²⁺, a series of novel NIR fluorescent probes (Figure 13) were developed by introducing specific recognition groups (such as carbonothioate, phosphinothioate, etc.) into hemicyanine or cyanine platforms to effectively modulate the ICT process, enabling turn-on, ratiometric, or quenching fluorescence responses toward Hg²⁺ [123,124,125,126,127].

Figure 13. Structures of NIR fluorescent probes for Hg²⁺ based on hemicyanine/cyanine dyes.

In 2022, Zhang’s group developed an “off–on” NIR probe CyP, built on a hemicyanine skeleton, for selective detection and imaging of Hg²⁺ in cells and animals [123]. The probe was composed of a xanthene unit, a trimethylindoline moiety, and a diphenylphosphinothioic chloride group. The fluorescence was initially quenched via covalent modification of the phenolic hydroxyl group with a diphenylphosphinothioate moiety, which suppressed the ICT process. Upon reaction with Hg²⁺, CyP underwent hydrolysis to product the fluorophore CyOH, restoring the ICT pathway and producing a strong NIR emission at 710 nm. The probe exhibited remarkable selectivity for Hg²⁺, accompanied by a binding constant of (1.45 ± 0.12) × 10⁵ M⁻¹, an LOD of 54.8 nM, and a rapid equilibration time (10 min). In addition, CyP worked well across a broad pH window (4.0–9.0) and showed low cytotoxicity, enabling successful fluorescence imaging of exogenous Hg²⁺ in SHSY-5Y neuroblastoma cells and zebrafish with good concentration-dependent signal enhancement (Figure 14). Furthermore, the probe demonstrated excellent tissue penetration (up to 800 μm) and was successfully applied for subcutaneous imaging in Balb/c mice, underscoring its potential as a versatile NIR tool for monitoring Hg²⁺-associated pathological conditions in complex biological environments.

Figure 14. (A) (a) Fluorescence imaging of different concentrations of Hg²⁺ in SHSY-5Y cells; (b) The corresponding flow cytometric analysis. Letters a–e represent Hg²⁺ concentrations of 0, 20, 40, 80, and 100 μM, respectively; (c) The corresponding mean fluorescence intensities in cell imaging. (d) The corresponding mean fluorescence intensities of SHSY-5Y cells determined by flow cytometry; (B) Fluorescence imaging of Hg²⁺ in zebrafish. (a) Diagram of the zebrafish imaging experiment process; (b) Fluorescence imaging of different concentrations of Hg²⁺ in zebrafish; (c) The corresponding mean fluorescence intensities in zebrafish imaging; (d) Plot of the linear relationship between Hg²⁺ concentration and fluorescence intensity in zebrafish. Reproduced with permission from Ref. [123]. Copyright (2022) Royal Society of Chemistry.

Based on the similar mechanism mentioned above, Shang and Fang et al., in 2024, reported two hemicyanine-bassed probes CyHg and CyHg-Cl, employing a carbonothioate group as the Hg²⁺ recognition unit [124]. Notably, higher reactivity was observed for CyHgCl compared to CyHg, attributed to the electron-withdrawing effect of the chlorine substituent. The probe responded to Hg²⁺ via cleavage of the carbonothioate ester, releasing the chloro-substituted hemicyanine fluorophore that emitted strongly at 725 nm in the NIR region. Moreover, under identical experimental conditions, other common heavy metal ions such as Pb²⁺, Ag⁺, Cu²⁺, Cd²⁺, and Zn²⁺ caused no appreciable changes in the fluorescence spectrum of Cy-Hg, whereas Hg²⁺ did, demonstrating the good selectivity of Cy-Hg for Hg²⁺. CyHg-Cl was also capable of sensing Hg²⁺ in living HeLa cells and authentic water specimens. MTT assays demonstrated the low cytotoxicity of CyHg-Cl in HeLa cells (cell viability > 90% at 20 μM, IC₅₀ = 72 μM), and confocal imaging experiments confirmed its excellent membrane permeability and ability to visualize intracellular Hg²⁺. Furthermore, CyHg-Cl was successfully utilized to detect Hg²⁺ in real water samples.

Subsequently, Liu and colleagues fabricated HCYS, an NIR probe based on a hemicyanine framework featuring a thiomorpholine moiety as the Hg²⁺ recognition unit, for Hg²⁺ detection in environmental and biological systems via the ICT mechanism [125]. Among the common metal ions tested, only Hg²⁺ induced a rapid decrease in the fluorescence intensity of HCYS at 674 nm within 1 min, owing to its spin–orbit coupling and heavy atom effects. Furthermore, the fluorescence intensity exhibited a good linear relationship with Hg²⁺ concentration over the range of 0–25 μM, with a detection limit of 1.41 μM. The probe enabled imaging of Hg²⁺ in various cell lines, such as SH-SY5Y, HepG2 cells and HeLa cells, where the intracellular red fluorescence gradually diminished with prolonged incubation with Hg²⁺ and was recovered upon the addition of EDTA. It also allowed for dynamic monitoring of Hg²⁺ in living zebrafish. Moreover, HCYS can dynamically monitor Hg²⁺ in living zebrafish. Importantly, HCYS was capable of imaging Hg²⁺ in both longitudinal and transverse root slices of pea sprouts, confirming its ability to track mercury contamination in plant tissues.

In 2025, Liu et al. designed and synthesized HDS-Hg, a thiahemicyanine-based probe structurally similar to CyHg but featuring a monothiocarbonate group [126]. The probe itself exhibited weak fluorescence due to the electron-withdrawing thiocarbonate group that suppressed ICT [126]. Upon selective cleavage of the carbon–sulfur bond triggered by Hg²⁺, the NIR fluorophore HDS was released, restoring ICT and producing a pronounced turn-on signal at 765 nm. HDS-Hg demonstrated the advantages of a rapid response, high sensitivity (LOD = 89 nM), wide pH tolerance, superior selectivity, and favorable mitochondrial localization ability (Pearson’s coefficient = 0.94). The probe enabled not only the quantitative detection of Hg²⁺ in environmental water samples, herbal medicines, and seafood, but also fluorescence imaging of Hg²⁺ in HeLa cells, notably in tobacco seedling leaves, which was a rarely explored application for Hg²⁺ probes. Furthermore, owing to its NIR excitation and emission properties, HDS-Hg enabled in situ imaging in a mouse model of chronic mercury poisoning. The imaging results revealed that the red fluorescence signal was predominantly localized in the liver and exhibited a time-dependent increase with prolonged mercury exposure, demonstrating the potential of HDS-Hg for real-time monitoring and early diagnosis of Hg²⁺-induced liver injury.

Erdemir et al. constructed HTC-S for ratiometric and colorimetric sensing of Hg²⁺, by integrating a carbonothioate recognition group into a cyanine scaffold [127]. Upon specific reaction with Hg²⁺, HTC-S underwent hydrolysis, disrupting its extended π-conjugated system and causing a hypsochromic shift of fluorescence from 816 nm to 625 nm, a reduction in absorbance at 714 nm and 780 nm and the emergence of a new absorption band at 502 nm. HTC-S exhibited a wide linear range (0–30.0 μM), a low LOD of 0.10 μM, and a quick response (40 s). The probe was shown to be effective for monitoring Hg²⁺ residues on the surfaces of diverse food matrices and soil samples, as well as in water samples, through pronounced colorimetric and fluorescent changes. HTC-S was further loaded onto TLC plates and combined with a smartphone for quantitative analysis, revealing good linear relationships between the R/G and R/B ratios and Hg²⁺ concentrations in the range of 0–20 μM, thus providing a rapid on-site quantitative analysis method for Hg²⁺ detection. Cellular experiments confirmed that HTC-S exhibited low cytotoxicity (cell viability >90% at 10 mM) and mainly localized in mitochondria, as verified by the strong colocalization with Mito-Tracker Green (Pearson’s coefficient = 0.96). On this basis, HTC-S successfully achieved visual detection of Hg²⁺ in mitochondria of living cells: compared with the control group, after the addition of Hg²⁺, the red fluorescence signal in the NIR channel decreased, while the green fluorescence signal in channel 1 increased.

3.2.4. NIR Fluorescent Probes Derived from Other Organic Dyes

Apart from NIR probes constructed on rhodamine, cyanine, and isophorone frameworks, significant advances have also been made in the development of Hg²⁺ sensors using other NIR fluorophore platforms that operate via desulfurization-based reaction mechanisms, as illustrated in Figure 15 [128,129,130,131,132].

Figure 15. Structures and mechanisms of Hg²⁺ sensing by CB-Hg (A), DCM-Hg (B), TM-ND-Hg (C), TP-Hg (D) and BBN-Hg (E).

Lv’s group constructed a coumarin–benzopyrylium hybrid NIR probe CB-Hg for colorimetric and ratiometric Hg²⁺ sensing [128]. In aqueous solution, without Hg²⁺, the probe absorbed at 430 nm and emitted at 495 nm. Adding Hg²⁺ caused the absorption to red shift to 670 nm (Δλ = 240 nm), with a yellow-to-green color change, and produced a new NIR emission peak at 720 nm, while the 495 nm band diminished, enabling dual-channel ratiometric detection. CB-Hg showed extremely high sensitivity (LOD = 6.94 nM), excellent selectivity, a rapid response (30 s), and a 225 nm emission shift. The probe was further validated in tap water, mineral water, white rice, and live zebrafish, highlighting its practicality for Hg²⁺ monitoring in foodstuffs and biological specimens.

Zhou and colleagues constructed an NIR turn-on probe DCM-Hg, comprising a DCM fluorophore and a thiocarbonate recognition moiety [129]. Upon Hg²⁺-triggered desulfurization, the fluorophore DCM-OH was released, accompanied by prominent emission enhancement at 695 nm (LOD = 3.35 μM). With a Stokes shift of 145 nm that minimized self-absorption, DCM-Hg proved effective for fluorescence imaging of exogenous Hg²⁺ in HepG2 cells.

Utilizing a Hg²⁺-triggered thioacetal deprotection reaction, Zhang’s group designed an NIR probe TP-Hg, featuring a thiopyran–tetracarbonitrile core and a thioacetal-responsive unit [130]. Upon interaction with Hg²⁺, thioacetal deprotection triggered a marked turn-on NIR emission at 770 nm, with an LOD of 0.13 μM and an impressive 207 nm Stokes shift. Furthermore, TP-Hg possessed robust anti-interference capacity and naked-eye visual detection properties. Its practical versatility was demonstrated via successful fluorescence imaging of A549 cells and the fabrication of portable test strips for on-site Hg²⁺ monitoring, highlighting its potential for both biological imaging and environmental surveillance.

To further enhance detection accuracy, the group further designed TM-ND-Hg, a ratiometric fluorescent probe that integrated a benzothianone platform with a thioacetal group for detecting Hg²⁺ [131]. Interaction with Hg²⁺ caused a reduction in fluorescence at 516 nm and a concurrent enhancement at 794 nm, thereby achieving a ratiometric response with an LOD of 0.39 μM and an exceptionally large Stokes shift (358 nm). More importantly, TM-ND-Hg, in combination with smartphone-based RGB color recognition software, facilitated the portable on-site quantitative detection of Hg²⁺. Over the concentration range of 0–70 μM, both the R and B values demonstrated excellent linear relationships with Hg²⁺ concentration. In addition, the applicability of TM-ND-Hg was validated not only through the detection of Hg²⁺ in water samples, but also by fluorescence imaging of Hg²⁺ in living cells and organisms (zebrafish): as the Hg²⁺ concentration increased, the red-channel fluorescence signal in A549 cells and zebrafish was progressively enhanced, whereas the blue-channel fluorescence signal was gradually attenuated. These results suggest that TM-ND-Hg holds promise as an intelligent tool for rapid Hg²⁺ screening in complex matrices.

Employing a thiocarbonate recognition moiety similar to that of DCM-Hg, Fu’s group synthesized an NIR probe BBN-Hg, derived from an anthocyanin derivative, for visualizing Hg²⁺ [132]. Upon Hg²⁺-triggered cleavage, the probe released the fluorophore BBN-OH and the ICT effect was activated, producing a strong emission at 650 nm. BBN-Hg was employed for fluorescence imaging of Hg²⁺ in MCF-7 cells and zebrafish, with the red fluorescence signal displaying a concentration-dependent enhancement upon Hg²⁺ treatment. Remarkably, the probe was further applied to three-dimensional (3D) fluorescence imaging of zebrafish, thereby permitting direct visualization of the spatial distribution of Hg²⁺. Moreover, BBN-Hg proved useful not only for quantifying Hg²⁺ in real water samples (lake water, tea water, herbal extract) but also for the qualitative and quantitative analysis of Hg²⁺ in diverse soil types, while also enabling visualization of Hg²⁺ distribution in plant roots (Salvia miltiorrhiza and rice seedlings). Additionally, after immobilization on test strips, BBN-Hg allowed for qualitative recognition of Hg²⁺ through dual-modal colorimetric and fluorescent signals and could be coupled with smartphone-assisted RGB color analysis for on-site quantification. Taken together, BBN-Hg can provide a versatile tool for sensing and monitoring Hg²⁺ in both ecological and biological systems.

3.3. NIR Fluorescent Probes for Zn²⁺

As essential micronutrients, Zn²⁺ is critically involved in diverse biological processes, including enzymatic function, gene expression, and immune regulation [133]. However, imbalances in Zn²⁺ levels, whether deficiency or excess, are linked to serious health disorders, including developmental delays, neurodegenerative diseases, and gastrointestinal illness [134]. Beyond physiology, Zn²⁺ is widely used in industrial applications such as galvanization, battery production, and as a food additive. The environmental release of Zn²⁺ through industrial waste poses lasting ecological and health threats owing to its persistence and non-biodegradability [135]. Thus, the development of sensitive and selective monitoring methods for Zn²⁺ is crucial for understanding its biological roles and mitigating the related health and environmental impacts. Recently, large-scale NIR fluorescent probes with superior sensing performance have been developed to detect Zn²⁺. The photophysical properties of representative probes are summarized in Table 3.

Table 3. The photophysical properties of representative NIR fluorescent probes for Zn²⁺.

Name	Solvent	Ex (nm)	Em (nm)	Sensing Mechanism	Response Time	Reversible	Signal Transduction Mode	Detection Limit	pH Range	Application	Ref.
Probe 2	(EtOH + ACN)/ HEPES = 1/1, v/v	520	670	PET	1 min	Yes	turn-on	2.7 nM	6.0–7.0	Pond water	[136]
ISOHP	-	-	668	PET, isomerization	15 s	-	turn-on	0.21 mM	6.0–8.0	Drinking water, A549 cells	[137]
NIR-Zn	CH3CN/HEPES = 1/1, v/v	481	660	PET	30 s	Yes	turn-on	4.8 nM	-	Pond water, A549 cells	[138]
AXS	DMSO/H2O = 7/3, v/v	508	688.4	PET	-	Yes	turn-on	0.069 nM	-	Cauliflower, crispy pear, Chinese cabbage, garlic, Chinese date, lemon, Zinc gluconate oral liquid, C. elegans, mice	[139]
IQ	H2O/EtOH = 1/9, v/v	495	664	PET	30 s	Yes	turn-on	60.5 nM	6.0–8.0	Lake water, tap water, distilled water, rice, photo, sandy soil, plant soil, HeLa cells	[140]
BTDM	DMSO/H2O = 8/2, v/v	535	683	IET	250 s	Yes	turn-on	18 nM	-	Apple, orange, banana, garlic pear, C. elegans, mice,	[141]
Probe 1	HEPES/DMF = 1/1, v/v	482	653	PET	<20 s	-	turn-on	15.3 nM	5.0–11.0	Pond water, river water, HepG 2 cells	[142]
NIRF	DMSO/HEPES = 2/3, v/v	484	658	CHEF	20 s	Yes	turn-on	0.79 mM	4.0–11.0	Distilled water, tap water, lake water, sprouted potato, RAW264.7 cells, mice	[143]
DR6G	DMF/TRis-HCl = 1/1, v/v	483	646	FRET	within a few seconds,	Yes	turn-on	20.2 nM	7.0–11.0	Zinc gluconate, calcium zinc gluconate, HUVEC cells, mice	[144]
DHP	THF/HEPES = 1/1, v/v	428	660	CHEF, ESIPT, ICT	4.7 s	Yes	turn-on	0.13 mM	-	MCF-7 cells, HepG2 cells, mung bean sprout	[145]
TZn	HEPES/EtOH = 9/1, v/v	580	652	PET	70 s	Yes	turn-on	0.31 mM	-	Milk, water, red wine, apple juice, HeLa cells	[146]
Mito-Zn	HEPES/ACN = 99/1, v/v	684/502	807/653	PET	-	-	ratiometric	26.7 nM	-	HeLa cells, mice	[147]

Note: Probe names are shown in bold.

3.3.1. NIR Fluorescent Probes Derived from Dicyanoisophorone Derivatives

Dicyanoisophorone derivatives, which are characterized by their large planar conjugated structures, synthetic accessibility, and modifiable frameworks, have been widely exploited as NIR fluorophores. Based on the PET mechanism, numerous NIR fluorescent probes (Figure 16) have been developed for the selective detection of Zn²⁺ by introducing various recognition units into the dicyanoisophorone backbone, which exhibited superior sensing performance and hold significant promise for practical applications spanning environmental monitoring, food safety analysis and bioimaging [136,137,138,139,140,141,142,143,144,145].

Figure 16. Structures of NIR fluorescent probes for Zn²⁺ based on dicyanoisophorone derivatives.

In 2022, Yan, Wu and colleagues incorporated a furan-2-carbohydrazide moiety into the isophorone scaffold to afford Probe 2 [136]. The probe initially displayed negligible fluorescence, which was attributed to the PET process from the C=N bond to the isophorone unit. Upon coordination with Zn²⁺ via the hydroxyl oxygen, acylamino oxygen, and imine nitrogen, the PET effect was effectively suppressed, causing a fold increase in emission at 670 nm. A good linear correlation (R²= 0.9911) was established between fluorescence intensity and Zn²⁺ concentration, and the LOD was calculated to be 2.7 nM. Probe 2 featured high selectivity, reversibility, and successful application in real water sample analysis. Based on the same sensing mechanism, the group later reported ISOHP, which was synthesized via condensation of dicyanoisophorone with 2-hydrazinopyridine [137]. Structurally similar to Probe 2, ISOHP displayed a rapid turn-on response at 668 nm (within 15 s) upon 1:1 coordination with Zn²⁺ and Cd²⁺, with LODs of 0.21 μM and 0.31 μM, respectively. In contrast to Probe 2, ISOHP was further employed for fluorescence imaging in A549 cells and for visual detection using test strips.

To develop an NIR fluorescent probe with high selectivity for Zn²⁺ and suitability for bioimaging, in 2024, the team conjugated an o-phenylenediamine recognition group to a dicyanoisophorone scaffold, constructing the NIR probe NIR-Zn [138]. Binding Zn²⁺ in a 1:1 ratio suppressed the PET process, generating a sharp emission rise at 660 nm. NIR-Zn exhibited much higher selectivity for Zn²⁺ than for other metal ions, including Cd²⁺, along with a rapid response (<30 s), remarkable sensitivity (LOD = 4.8 nM), and a 179 nm Stokes shift. However, the coexistence of Cu²⁺, Co²⁺, and Ni²⁺ severely interfered with the detection signal of NIR-Zn toward Zn²⁺. Despite this interference, the probe was still effectively used for the quantitative detection of Zn²⁺ in pond water with excellent recovery and for confocal imaging of Zn²⁺ in A549 cells.

Similarly, in 2024, Miao’s team constructed AXS from the same dicyanoisophorone platform, functionalized with a furan-2-carbohydrazide moiety [139]. After specific binding to Zn²⁺, AXS displayed a prominent turn-on fluorescence response at 688.4 nm with outstanding sensing performance, including an exceptionally low detection limit (69.24 pM), a large Stokes shift (180 nm), good reversibility, high photostability, and pH insensitivity. AXS enabled both the qualitative and quantitative determination of Zn²⁺ in diverse real samples, including vegetables, fruits, and zinc gluconate oral solutions. Moreover, AXS was not only successfully used to fabricate portable test strips for visual on-site screening of Zn²⁺, but was also was coupled with a smartphone-assisted RGB analysis platform for accurate Zn²⁺ quantification with excellent sensitivity (LOD = 3.07 nM). Additionally, AXS was successfully employed for bioimaging of Zn²⁺ in Caenorhabditis elegans and mouse tissues (heart, liver, and kidney).

Erdemir et al. reported the NIR probe IQ for rapidly and selectively recognizing Zn²⁺ in environmental and biological specimens [140]. IQ exhibited weak intrinsic fluorescence, attributed to a PET state. Among various metal ions including Cd²⁺, Cu²⁺, and Hg²⁺, only Zn²⁺ significantly inhibited PET, increasing the fluorescence quantum yield of IQ from 0.056 to 0.40, which demonstrated good selectivity towards Zn²⁺. IQ was successfully loaded onto test strips, cotton swabs, and polysulfone capsules for convenient visual monitoring of Zn²⁺. With increasing Zn²⁺ concentration, a yellow-to-pink transition was observed under daylight, accompanied by a gradual enhancement in red fluorescence under UV light. Furthermore, when combined with smartphone-assisted RGB analysis, IQ enabled real-time quantitative detection in both colorimetric and fluorescent modes, with limits of detection of 0.30 μM and 0.21 μM, respectively. Its utility was further validated by Zn²⁺ analysis in water, food, and soil matrices. Finally, IQ successfully achieved the visual detection of Zn²⁺ in living cells, leveraging its NIR emission, large Stokes shift, and low cytotoxicity.

Recently, Qiao and Miao et al. constructed the probe BTDM, which possessed the same fluorophore scaffold as IQ, featuring a benzothiazole–hydrazine moiety as the Zn²⁺-specific recognition unit, for Zn²⁺ detection [141]. Only Zn²⁺ coordinated with BTDM in a 1:2 stoichiometric ratio via the imine nitrogen along with the benzothiazole nitrogen and sulfur atoms, triggering a pronounced turn-on emission at 683 nm. The probe was characterized by pH insensitivity (5–10), a remarkably large Stokes shift (248 nm) and detection limits of 34 nM (UV-Vis) and 18 nM (fluorescence). BTDM was subsequently fabricated into a multifunctional sensing toolkit, enabling Zn²⁺ detection through smartphones, test strips and electrospinning. These diverse platforms enabled BTDM to effectively quantify Zn²⁺ in actual foodstuff samples, such as apples, oranges, bananas, garlic and pears. Additionally, BTDM-functionalized electrospun fibers served as solid-state sensors for visual Zn²⁺ detection. More importantly, BTDM exhibited superior bioimaging capability and was employed for in vivo fluorescence imaging of Zn²⁺ in C. elegans and mice.

Beyond the pursuit of highly selective single-analyte probes, the integration of multifunctional recognition motifs into the isophorone platform has facilitated the development of NIR fluorescent sensors capable of simultaneous detection of Zn²⁺ and other anionic species. A representative example was reported in 2022 by Lei’s group, who constructed Probe 1 by introducing a methyl hydrazate group onto a dicyanoisophorone skeleton via Schiff base reaction [142]. The probe initially displayed weak emission, arising from the combined PET effect of the C=N bond and inefficient ICT. Coordination with Zn²⁺ via N and O atoms inhibited both quenching pathways, yielding a rapid fluorescence turn-on at 653 nm (within 20 s), with an LOD of 15.3 nM. In contrast, HClO reacted with the C=N bond via addition, inhibiting PET while allowing for ICT, which induced an emission enhancement with a bathochromic shift from 670 nm to 705 nm (response time: 4 min; LOD: 1.39 μM). Probe 1 was further validated by fluorescence imaging of Zn²⁺ and HClO in HepG2 cells and by recovery studies in environmental waters, highlighting its potential for multi-analyte applications in biological and environmental fields.

Yang’s group constructed an NIR fluorescent probe NIRF by incorporating semicarbazide hydrochloride into the dicyanoisophorone platform, enabling sequential detection of Zn²⁺ and CN^- [143]. Only Zn²⁺ could form a stable complex with NIRF (K_b = 2.06 × 10⁵), thereby causing significant fluorescence enhancement at 658 nm. Subsequent addition of CN⁻ triggered a ligand exchange reaction, displacing Zn²⁺ to form [Zn(CN)_x] complexes and releasing free NIRF, resulting in distinct fluorescence quenching with an LOD of 7.9 × 10⁻⁷ M for CN⁻. Benefiting from its excellent sensing performance and low cytotoxicity, NIRF was successfully applied to the visual detection of Zn²⁺ and CN⁻ in living macrophages and mice via fluorescence imaging, as well as for the detection of both ions in real samples, highlighting its potential for dual-analyte sensing in complex systems.

Based on the FRET mechanism, in 2024, Yang’s group presented an NIR fluorescent probe DR6G for sequentially monitoring of Zn²⁺ and pyrophosphate (PPi) [144]. The probe consisted of rhodamine 6G hydrazide (donor) and a dicyanoisophorone derivative (acceptor). Zn²⁺ induced spiro-ring opening, producing a remarkable fluorescence emission at 646 nm with high sensitivity (LOD = 20.2 nM). The resulting DR6G–Zn²⁺ complex enabled PPi detection via fluorescence quenching (LOD = 20.6 nM), attributed to the strong binding between Zn²⁺ and PPi. DR6G was successfully used to visualize Zn²⁺ and PPi in living cells, mice, and Staphylococcus aureus, and to quantify Zn²⁺ in oral liquids and PPi in toothpaste.

Recently, Yu’s group developed DHP, a dicyanoisophorone-based NIR probe bearing a di(thiophen-2-yl)ethyl amide [145]. Owing to the intrinsic PET effect, DHP initially exhibited weak fluorescence. Upon 1:1 coordination with Zn²⁺ via the phenolic oxygen, imine nitrogen, and carbonyl oxygen atoms, the PET pathway was effectively suppressed and the CHEF effect was activated, giving rise to a pronounced NIR turn-on response at 660 nm. The probe featured an ultrafast response (4.7 s), a large Stokes shift (160 nm), and a low LOD of 0.13 μM for Zn²⁺, along with excellent selectivity and robust anti-interference capability. Notably, the DHP-Zn²⁺ complex could be further employed for the sequential detection of pyrophosphate (PPi) via competitive displacement of Zn²⁺, achieving an LOD of 0.39 μM and rapid response (6.8 s), thereby endowing the probe with dual-analyte sensing functionality. Beyond solution-based assays, DHP was successfully integrated into smartphone-assisted RGB analysis and portable swab tests for visual and on-site Zn²⁺ detection. Bioimaging experiments demonstrated that, in addition to the sequential imaging of exogenous Zn²⁺ and PPi in living cells (MCF-7 and HepG2), DHP also enabled the detection of Zn²⁺ in plant tissue (mung bean sprouts), highlighting its low background interference in complex biological tissues.

3.3.2. NIR Fluorescent Probes Derived from Other Organic Dyes

DCM derivatives are commonly employed as NIR fluorophores due to their favorable photophysical properties. Zhang’s group reported TZn (Figure 17), constructed on a DCM scaffold, by using a dipicolylamine (DPA) moiety as a specific recognition motif for Zn²⁺ [146]. The PET pathway of the probe TZn was effectively inhibited upon 1:1 chelation with Zn²⁺ via the DPA nitrogen and the phenolic oxygen, resulting in marked NIR fluorescence enhancement at 652 nm and a concomitant 30 nm red shift in the absorption band. TZn demonstrated outstanding selectivity for Zn²⁺ against various competing metal ions, achieving a detection limit of 0.31 μM and a fast response time of 40 s. Furthermore, the probe possessed favorable photostability, low cytotoxicity (cell viability >95% at 10 mM) and good biocompatibility, allowing for successful fluorescence imaging of Zn²⁺ in living HeLa cells and reliable quantification of Zn²⁺ in diverse food matrices (milk, water, red wine, and apple juice) with recoveries in the range of 92–103%.

Figure 17. Structure and mechanism of TZn responses to Zn²⁺.

To quantitatively analyze Zn²⁺ in mitochondria, Huang and co-workers presented a ratiometric fluorescent probe Mito-Zn (Figure 18A) for mitochondrial Zn²⁺ detection [147]. The probe featured a cyanine-derived NIR fluorophore, a lipophilic cation for mitochondrial targeting, a pyridine-based Zn²⁺ recognition unit, and sulfo-groups for water solubility. Upon Zn²⁺ addition, Mito-Zn exhibited a distinct ratiometric response, characterized by a decrease in emission at 807 nm and a concomitant increase at 653 nm. The probe demonstrated high sensitivity (LOD = 26.7 nM), remarkable selectivity and low cytotoxicity (cell viability >92.4% at 60 mM). In addition, Mito-Zn mainly localized to mitochondria, as evidenced by strong colocalization with Mito-Tracker Green (PCC =0.88), enabling quantitative detection of exogenous Zn²⁺ in mitochondria. With increasing Zn²⁺ concentrations, the red-channel emission decreased gradually and the green-channel emission increased concomitantly, with the Fgreen/Fred ratio rising from 0.0425 to 0.8039. Additionally, it successfully visualized Zn²⁺ fluctuations in living cells under oxidative stress (Figure 18B) and notably revealed elevated Zn²⁺ levels in the brains of mice with depression.

Figure 18. (a) Structure and mechanism of Mito-Zn responses to Zn²⁺; (b) fluorescence imaging of endogenous Zn²⁺ in HeLa cells. (A–E) Fluorescence images from the green channel (λem = 610–690 nm). (F–J) Fluorescence images from the red channel (λem = 760–840 nm). (K–O) Pseudo images obtained from the green and red channels. (P) The corresponding Fgreen/Fred. Scale bar: 20 μm. Reproduced with permission from Ref. [147]. Copyright (2024) Royal Society of Chemistry.

3.4. NIR Fluorescent Probes for Pb²⁺

Lead ions (Pb²⁺), as a typical heavy metal pollutant, are extensively utilized in daily materials including gasoline, batteries, pigments, cosmetics, food additives, and toys, and can be accumulated in organisms through the environment and the food chain [148]. Given its high toxicity, persistence, and strong bioaccumulative nature, even trace Pb²⁺ exposure can cause irreversible damage to the nervous system, kidneys, and other vital organs [149]. In response, the WHO has set a maximum limit of 10 μM for Pb²⁺ in drinking water [14]. Therefore, developing simple, cost-effective, and rapid detection methods for Pb²⁺ in complex samples is of great significance. Recently, researchers have successively reported several Pb²⁺ detection systems based on NIR fluorescent probes (Figure 19), which demonstrate promising application potential in environmental monitoring and bioimaging owing to their superior photophysical properties and sensitive response mechanisms, as shown in Table 4 [150,151,152,153].

Figure 19. Structures of NIR fluorescent probes for Pb²⁺.

Table 4. The photophysical properties of representative NIR fluorescent probes for Pb²⁺.

Name	Solvent	Ex (nm)	Em (nm)	Sensing Mechanism	Response Time	Reversible	Signal Transduction Mode	Detection Limit	pH Range	Application	Ref.
NDHN	DMSO/H2O = 3/7, v/v	416 nm	667/584	ICT, PET, C=N isomerization	-	-	ratiometric	0.157 mM	-	Milk, milk powder, white wine, soybean, black bean, broad bean, HeLa cells, Caenorhabditis elegans, mice	[150]
DHB	DMF/HEPES = 4/6, v/v	430	650	ICT	-	Yes	turn-on	32.3 nM	6.0–7.0	Tap water, Songhua River water, soil extract, aqueous solution of licorice tablets extract, canned cola, HeLa cells, zebrafish, mice, rice roots	[151]
LCE1	Tris-HCl/DMF = 9/1, v/v	488	670	ICT	30 s	-	turn-off	0.34 mM	4.5–10.0	Tap water, river water, lake water, black tea, milk, Chrysopogon zizanioides (L.) Roberty, Medicago sativa L., MCF-7 cells, mice	[152]
YSQ	MeOH/H2O= 9/1, v/v	490	678	ICT, PET, C=N isomerization	<1 min	-	turn-on	18 nM	-	Lake water, tap water, PK-15 cells	[153]

Note: Probe names are shown in bold.

In 2024, Miao and colleagues reported the first fluorescent probe NDHN featuring both NIR emission and ratiometric fluorescence properties for sensing Pb²⁺ [150]. The probe consisted of a dicyanoisophorone scaffold, with a pyridine amide moiety serving as the Pb²⁺ recognition unit [150]. Coordination with Pb²⁺ inhibited the PET effect and C=N isomerization while enhancing the ICT effect, resulting in a red shift of the maximum absorption wavelength from 423 to 466 nm and a red shift of the fluorescence emission peak from 584 to 677 nm. NDHN exhibited high sensitivity and selectivity toward Pb²⁺, with minimal interference from competing ions. Leveraging these optical properties, a fluorescence/colorimetric/smartphone triple-mode sensing platform was constructed and successfully applied for Pb²⁺ detection in foods, including milk, milk powder and white wine. In smartphone mode, color intensity showed a good linear relationship with Pb²⁺ concentration over the range of 1–10 μM, with a detection limit of 130 nM. NDHN could be used for subcellular localization in HeLa cells and fluorescence imaging of exogenous Pb²⁺. The intracellular red fluorescence increased with increasing Pb²⁺ concentration and highly overlapped with the blue fluorescence of DAPI (PCC = 0.81). Notably, NDHN could be used for imaging Pb²⁺ in mouse organ cells, and NDHN-Pb²⁺ exhibited excellent tumor-targeting ability in tumor-bearing mice, with the fluorescence intensity in tumor tissue being 5.1 times higher than that in the liver. Thus, NDHN offers a promising molecular tool for diagnosing Pb²⁺-related diseases.

Similarly, Zhao’s group successfully constructed a multifunctional NIR probe DHB based on the isophorone fluorescent scaffold, using benzohydrazide as the Pb²⁺ recognition group [151]. DHB exhibited an “off–on–off” fluorescence response, enabling sequential detection of Pb²⁺ ions and glyphosate (Glyp). DHB initially showed weak orange–red emission. Upon binding with Pb²⁺, the ICT effect was significantly enhanced, leading to a substantial increase in fluorescence intensity and a 70 nm red shift of the emission wavelength. Concurrently, the maximum absorption increased, accompanied by a 47 nm red-shift. Subsequent addition of glyphosate (Glyp) competitively removed Pb²⁺ from the DHB-Pb²⁺ ensemble, restoring the initial weak fluorescence. This reversible response enabled at least five sequential cycles. The probe showed high selectivity, excellent anti-interference capability, and high sensitivity for Pb²⁺ and Gly, with reported LODs of 35.8 mM and 36.5 mM, respectively, and a large Stokes shift of 220 nm. DHB was further utilized to quantitatively determine Pb²⁺ and Glyp in diverse real samples, including environmental water, soil extracts, and food matrices, with acceptable recoveries of 97.2–113.2% for Pb²⁺ and 90.1–102.8% for Glyp. Both DHB and the DHB-Pb²⁺ complex showed low cytotoxicity, with >80% cell viability at 25 μM. In HeLa cells, DHB alone gave weak fluorescence, whereas Pb²⁺ induced concentration-dependent red fluorescence. Subsequent Glyp addition caused concentration-dependent fluorescence quenching. Similar phenomena were observed in rice root tips, zebrafish, and mouse models. Organ imaging suggested preferential hepatic accumulation of DHB-Pb²⁺, with stronger fluorescence in the liver than in other organs. These findings indicated that DHB may serve as a visual and potentially semi-quantitative probe for Pb²⁺ and Glyp in complex biological matrices.

In 2025, Meng’s group reported an NIR probe LCE1 derived from a TCF platform functionalized with an azacrown ether as the Pb²⁺ recognition group [152]. Pb²⁺ coordinated with the N/O donor atoms of the azacrown ether and the oxygen atom of the phenolic, forming a 1:1 complex with LCE1, thereby suppressing the ICT process and resulting in fluorescence quenching at 670 nm. Fluorescence titration showed good linearity over the Pb²⁺ concentration range of 0–6 μM (R² = 0.9917), with an LOD of 0.34 μM. LCE1 could stably respond to Pb²⁺ over the pH range of 4.5–10.0, with a response time of 30 s, and exhibited strong anti-interference capability in the presence of coexisting ions such as Cd²⁺, Ag⁺, Cu²⁺, Hg²⁺, and so on. The probe enabled the quantitative determination of Pb²⁺ in real samples, including tap water, river water, lake water, black tea, and milk, with favorable recoveries (95.2–108.2%). In addition, LCE1-based test strips were used not only for semi-quantitative detection of Pb²⁺, as indicated by a color change from deep pink to pale orange with increasing Pb²⁺ concentration, but also, in combination with smartphone-based RGB analysis, for on-site rapid relative quantification of Pb²⁺ in water samples. The R/(G + B) ratio showed a linear correlation with Pb²⁺ concentration over the range of 0–300 μM (R² = 0.9717). Furthermore, LCE1 exhibited low cytotoxicity (cell viability >85% at 30 μM) and was successfully applied for fluorescence imaging of Pb²⁺ in MCF-7 cells, zebrafish, nude mice, and Pb-accumulating plants (Chrysopogon zizanioides (L.) Roberty and Medicago sativa L.), where Pb²⁺-dependent fluorescence attenuation was observed. This study suggests that LCE1 holds promise for understanding the physiological roles of Pb²⁺, evaluating food safety, and screening plants for the phytoremediation of heavy metal-contaminated soils.

In addition, Liu and Xu et al. reported YSQ, an NIR probe built on an isophorone-isoquinoline scaffold [153]. In its free state, the probe showed weak emission owing to C=N isomerization and PET. However, coordination with Pb²⁺ effectively suppressed these quenching pathways and activated the CHEF effect, leading to a dramatic fluorescence enhancement at 678 nm. YSQ offered advantages of excellent selectivity, high sensitivity (LOD = 18 nM), fast response (1 min), and good reversibility. Based on these outstanding properties, YSQ was effectively employed to quantitatively detect Pb²⁺ in real water samples, imaging Pb²⁺ in living PK-15 cells, and visually sensing Pb²⁺ using test strips and SiO₂ powder. Furthermore, with the aid of smartphone-assisted RGB color analysis, the probe enabled on-site visual screening of Pb²⁺.

3.5. NIR Fluorescent Probes for Cd²⁺

Cadmium (Cd²⁺) is a widely used industrial metal, employed in electroplating, smelting, battery manufacturing, and phosphate fertilizers. Its extensive application has led to the release of cadmium-containing wastewater, exhaust gases, and waste residues into the environment, contaminating soil and water through atmospheric deposition and irrigation [154]. Unlike other heavy metals, Cd²⁺ is readily absorbed by crops, such as rice, and undergoes significant bioaccumulation in plants due to its non-biodegradability. Through contaminated food, water, and air, Cd²⁺ enters the human body, posing serious health risks, including cardiovascular diseases, Alzheimer’s disease, kidney dysfunction, calcium metabolism disorders and Parkinson’s disease [155]. A typical example is Itai-Itai disease, caused by long-term consumption of cadmium-contaminated rice, which remains the primary route of human Cd²⁺ exposure [156]. The WHO has set a stringent limit of 3 ppb for Cd²⁺ in drinking water. Therefore, developing portable, efficient, and sensitive methods for Cd²⁺ detection is of great importance.

However, there have been few reports on NIR fluorescence probes for Cd²⁺ detection in recent years, primarily due to the following reasons. First, the chemical similarity between Cd²⁺ and Zn²⁺ poses a fundamental obstacle to highly selective recognition. Moreover, as an essential trace element for the human body, Zn²⁺ is ubiquitous in biological systems and far more abundant than Cd²⁺, making it difficult for probes to accurately discriminate Cd²⁺ in the presence of Zn²⁺. Second, the heavy-atom effect of Cd²⁺ itself drives probe design toward “turn-off” modes with weak signal changes. This intrinsic tendency makes the design of probes based on conventional “turn-on” mechanisms extremely challenging. Existing strategies are largely limited to “turn-off” responses or require ratiometric probes with higher synthetic difficulty to compensate for insufficient signal changes, thereby significantly restricting detection sensitivity and practical value. Third, Cd²⁺ is primarily regarded as an environmental pollutant, and its direct pathophysiological relevance is less clear than that of ions such as Zn²⁺ or Cu²⁺, resulting in relatively low research attention. Fourth, red-shifting the probe emission wavelength into the NIR region (650–900 nm) is itself a complex molecular engineering task, and on this basis, designing a suitable recognition group for Cd²⁺ requires simultaneously balancing the electronic structure of the fluorophore and the coordination ability of the recognition unit, leaving a narrow design window.

Given the complexity of environmental pollutants and the interconnectivity of metabolites within organisms, the development of multifunctional probes capable of the continuous detection of multiple analytes has become a research hotspot. In 2025, Zu and co-workers constructed the probe XL (Figure 20) by linking a dicyanodiphenyl isophorone derivative with a quinoline ring via a hydrazone bond, achieving continuous detection of Cd²⁺ and glutathione (GSH) [157]. The free probe emitted fluorescence at 565 nm. Upon binding with Cd²⁺ in a 1:2 molar ratio, the ICT process was activated, leading to a ratiometric fluorescence change, with the fluorescence intensity at 650 nm gradually increasing and progressively decreasing at 565 nm. When GSH was added to the XL-Cd²⁺ complex, the strong coordination ability of GSH competitively displaced Cd²⁺, releasing free XL and restoring fluorescence to 565 nm, while the solution color changed from purplish-red to yellow, enabling the detection of GSH with a detection limit of 0.784 μM. The probe demonstrates rapid response times, completing detection of both Cd²⁺ and GSH within 3 s. Based on its distinct color and fluorescence changes, the research team developed a smartphone-assisted visual detection platform, successfully applied for quantitative analysis of Cd²⁺ in water samples and GSH in garlic samples, providing a novel tool for rapid on-site detection.

Figure 20. Structure and mechanism of XL responses to Cd²⁺ and GSH.

3.6. NIR Fluorescent Probes for Fe²⁺/Fe³⁺

Iron (Fe²⁺/Fe³⁺) is an essential trace element involved in oxygen transport, DNA synthesis, metabolism, and immune defense. Intracellular iron exists mainly as redox-active Fe²⁺ in the labile iron pool [158]. Precise maintenance of iron homeostasis is critical for normal cellular function, and dysregulation of the Fe²⁺/Fe³⁺ balance can lead to serious health problems. Iron deficiency may cause iron-deficiency anemia and immune dysfunction, while iron overload induces oxidative damage via the Fenton reaction, leading to liver injury, neurodegeneration, cardiovascular disease, and cancer [159,160]. Moreover, excess Fe²⁺ drives ferroptosis, a key process in drug-induced liver injury and neurodegenerative diseases [161]. Thus, developing sensitive and selective methods for Fe²⁺/Fe³⁺ detection is crucial for investigating disease mechanisms and early diagnosis. NIR fluorescent probes, with their high spatiotemporal resolution and biocompatibility, have become powerful tools for real-time monitoring of iron dynamics in biological systems. The photophysical properties of representative probes are summarized in Table 5.

Table 5. The photophysical properties of representative NIR fluorescent probes for Fe²⁺/Fe³⁺.

Name	Solvent	Ex (nm)	Em (nm)	Sensing Mechanism	Response Time	Reversible	Signal Transduction Mode	Detection Limit	Ph Range	Application	Ref.
NRF-Fe2+	Tris-HCl/ DMSO = 7/3, v/v	430	460/675	FRET	<30 min	No	ratiometric	75 nM	2.0–8.0	A549 cells, mice	[162]
MDJ-O	PBS	535	720	PET	5 min	No	turn-on	3.09 mM	4.0–9.0	HepG2 cells, zebrafish, mice	[163]
DICO-Fe(II)	HEPES/EtOH = 9/1, v/v	530	666	ICT	<8 min	No	turn-on	47 nM	-	Spinach, rape, bean sprouts, chicken samples, HeLa cells, zebrafish, mice	[164]
DCA-MIn	MeOH/H2O = 5/5, v/v	570	670	ICT, PET	<15 s	-	turn-on	19.5 mM	4.0–8.0	Tap water, lake water, MCF-7 cells, anti-counterfeiting ink	[165]
DHX-Fe	PBS/EtOH = 7/3, v/v	610	690	ICT	<5 min	No	turn-on	115 nM	-	A2780 cells, mice	[166]
BDNO	DMSO/H2O = 1/1, v/v	610/ 540	715/570	ICT	<5 s	No	ratiometric	41 nM	4.0–10.0	Real water samples, apple juice, orange juice, wine, tap water, lake water, sea water, mung bean sprouts	[167]
FTP	EtOH/HEPES = 3/7, v/v	690	743	PET	<2 min	No	turn-on	24.0 nM (fluorescence) 79.7 nm (UV)	4.0–10.0	Tap water Wolong Lake water Tianyi Lake water, HepG2 cells, mice	[168]
CyLIP	HEPES	660	705	ICT	20 min	No	turn-on	51 nM	5.0–9.0	HepG2 cells, mice	[169]
QM-FV	HEPES/DMSO = 1/1, v/v	450	740	ICT	<150 s	No	turn-on	25 nM	-	HepG2 cells, HeLa cells, Huh7 cells, AML12 cells, HEK293T cells, Vero cells, mice	[170]

Note: Probe names are shown in bold.

3.6.1. NIR Fluorescent Probes Derived from Dicyanoisophorone Derivatives

In recent years, leveraging the readily modifiable nature of the isophorone skeleton, researchers have successfully developed a variety of high-performance NIR fluorescent probes for Fe²⁺ detection and imaging. These probes, which operate through distinct response mechanisms, were constructed by structurally tuning the isophorone scaffold and introducing specific recognition moieties for Fe²⁺ (Figure 21).

Figure 21. Structures and mechanisms of Fe²⁺ sensing by NRF-Fe2+ (A), MDJ-O (B) and DICO-Fe(II) (C).

To enable tracking of Fe²⁺ in iron-associated diseases in vivo, Feng’s group described a FRET-based ratiometric NIR probe NRF-Fe²⁺ [162]. It consisted of a coumarin moiety as the energy donor and a dicyanoisophorone derivative as the acceptor, with an N-oxide group functioning as the Fe²⁺ recognition unit. Reduction of N-oxide by Fe²⁺ induced a marked emission shift from 460 nm to 675 nm, enabling ratiometric response with an LOD of 72 nM. The probe exhibited outstanding selectivity for Fe²⁺ over metal ions, ROS, RSS, saccharides, and amino acids, with negligible interference even in 10 mM GSH and operated over pH 2–8. NRF-Fe²⁺ showed low cytotoxicity (A549 viability >82% at 0–50 μM) and enabled ratiometric imaging of exogenous and endogenous Fe²⁺ in living cells, where blue fluorescence gradually decreased with prolonged Fe²⁺ exposure, while red fluorescence was enhanced. DFO (deferoxamine mesylate, a Fe²⁺ chelator) reversed this signal. Moreover, similar phenomena were also observed when NRF-Fe2+ was employed to monitor Fe²⁺ dynamics during Erastin-induced ferroptosis. Notably, NRF-Fe²⁺ also detected elevated Fe²⁺ in acetaminophen (APAP)-induced liver injury, with fluorescence intensity approximately twice that of controls. As one of the few NIR ratiometric probes for tracking Fe²⁺ in ferroptosis and DILI, NRF-Fe²⁺ holds promise for Fe²⁺-related disease research, although it has a slow response (~30 min) and requires multi-step synthesis.

To address the slow response of many Fe²⁺ probes, in 2024, Xiao et al. introduced MDJ-O, an NIR probe featuring an isophorone-based fluorophore and an N-oxide unit on the julolidine ring as the Fe²⁺-responsive moiety [163]. Upon exposure to Fe²⁺, MDJ-O was converted to the strongly emissive MDJ, yielding a 49-fold increase in fluorescence intensity at 720 nm (l_ex = 535 nm) and a red shift of absorption peak from 390 nm to 550 nm within 5 min. The probe enabled both colorimetric and fluorometric detection of Fe²⁺ in nearly 100% aqueous medium. MDJ-O exhibited high selectivity for Fe²⁺ over other analytes, without interference except for Cu²⁺, and showed a pronounced response over pH 4–9. It had low cytotoxicity (cell viability > 90% at 10–30 μM), good intracellular photostability, and was mainly localized in lysosomes (PCC = 0.88, significantly higher than 0.56 for mitochondria). The successful fluorescence imaging of exogenous and endogenous Fe²⁺ in cells and imaging of Fe²⁺ in zebrafish in a concentration-dependent manner were achieved by using MDJ-O. More importantly, MDJ-O could detect both exogenous and endogenous Fe²⁺ in mice and visualize elevated Fe²⁺ levels in APAP-induced liver and kidney injury models (Figure 22), where fluorescence in the liver and kidneys was markedly stronger than in other organs. 2,2′-bipyridine (Bpy) pretreatment significantly suppressed this enhancement, confirming that the signal originated from Fe²⁺. Therefore, MDJ-O offers a promising tool for investigating drug-induced organ damage and iron-related diseases, although its capacity for Fe²⁺ sensing is still compromised by Cu²⁺ interference and a relatively high detection limit (3.09 mM).

Figure 22. (A) Fluorescence imaging of endogenous Fe²⁺ in zebrafish; (B) fluorescence imaging of exogenous Fe²⁺ in living mice; (C) fluorescence imaging of main organs of APAP-induced mice model. Reproduced with permission from Ref. [163]. Copyright (2024) Royal Society of Chemistry.

Later, to enable specific and sensitive monitoring of Fe²⁺ in food and biological systems, Zhu and Sheng et al. rationally engineered an NIR fluorescent probe DICO-Fe(II), incorporating a carbamoyl oxime recognition moiety and an isophorone-based fluorophore [164]. Unlike conventional N-oxide-based probes, the carbamoyl oxime group underwent Fe²⁺-mediated N–O bond cleavage, restoring the ICT effect and generating a turn-on emission at 666 nm within 8 min. DICO-Fe(II) exhibited high selectivity for Fe²⁺ over common cations, anions, reactive oxygen species, and biothiols, with negligible interference. Fluorescence titration showed excellent linearity over the Fe²⁺ concentration range of 0–50 μM (R² = 0.9917), with an LOD of 47 nM. Concurrently, the solution changed from colorless to tawny, enabling on-site visual detection via smartphone-assisted RGB analysis. Benefiting from its excellent sensitivity and selectivity, DICO-Fe(II) was successfully applied to the accurate determination of Fe²⁺ in food samples (spinach, rape, bean sprouts, and chicken) with acceptable recoveries of 89.87–110.22%. Moreover, the probe exhibited negligible cytotoxicity and excellent cell membrane permeability, enabling fluorescence imaging of Fe²⁺ in cells, where fluorescence intensity increased with Fe²⁺ concentration and was significantly suppressed by Bpy. DICO-Fe(II) also enabled visualization of Fe²⁺ in living zebrafish in a concentration-dependent manner. Notably, it allowed for fluorescent imaging of elevated Fe²⁺ levels in a mouse model of APAP-induced liver injury, whereas NAC treatment markedly reduced the signal, suggesting its potential as an effective tool for studying Fe²⁺ dynamics in drug-induced liver injury.

3.6.2. NIR Fluorescent Probes Derived from Other Organic Dyes

Taking advantage of the reducing property of Fe²⁺ and the strong coordination ability of Fe³⁺ with nitrogen and oxygen atoms, the researchers introduced specific recognition groups such as N-oxide and carbamoyl oxime into near-infrared fluorophores with large π-conjugated structures and successfully constructed a range of iron ion NIR fluorescent probes with excellent performance (Figure 23) [165,166,167,168,169,170].

Figure 23. Structures of NIR fluorescent probes for Fe²⁺/Fe³⁺.

In 2022, Tang’s group obtained DCA-Mln, a dual-channel colorimetric and NIR probe with a D-π-A framework, by coupling a dicyanovinyl group with a DCA fluorophore via the Knoevenagel condensation reaction [165]. DCA-MIn emitted strongly at 670 nm. While the fluorescence was efficiently quenched by Fe³⁺. Concurrently, a distinct color change from purple to peach pink was observed, allowing for naked-eye detection, and the solution color changed from purple to peach pink. The sensing mechanism was based on a 2:1 binding stoichiometry between DCA-Mln and Fe³⁺, which suppressed the ICT effect and triggered the PET process. DCA-Mln demonstrated excellent selectivity, a fast response (within 15 s) and a low LOD of 19.5 mM. Its applicability was further demonstrated in real water sample analysis, fluorescence imaging of Fe³⁺ in living MCF-7 cells, and anti-counterfeiting ink preparation.

To improve the sensitivity of NIR fluorescent probes for Fe²⁺ detection, He’s group designed an NIR probe DHX-Fe in 2025 [166]. The probe employed a quinoline–xanthene scaffold as the NIR emissive unit and a nitroxide radical as the responsive group for Fe²⁺. In the absence of Fe²⁺, the nitroxide radical quenched the fluorescence of DHX-Fe through electron transfer. Upon reaction with Fe²⁺, the nitroxide radical was reduced to hydroxylamine, triggering a 9.4-fold enhancement at 690 nm (λ_ex = 610 nm). Apart from Fe²⁺, DHX-Fe showed a negligible response to common metal ions, biological reductants, reactive oxygen species, and CYP450, and their coexistence does not interfere with Fe²⁺ detection. Furthermore, the probe displayed a fast response (<5 min) and high sensitivity (LOD = 115 nM) towards Fe²⁺ and operated effectively at a wide pH range (3–8). In addition, DHX-Fe exhibited low cytotoxicity (cell viability >80% at 50 μM) and excellent mitochondrial targeting capability (PCC = 0.89), enabling concentration-dependent fluorescence imaging of exogenous Fe²⁺ in A2780 and SKOV3 cells, with the fluorescence signal significantly reduced upon addition of Bpy or DFO. In cisplatin-resistant ovarian cancer cells (A2780/DPP), the probe was used to evaluate Fe²⁺ changes induced by various ferroptosis inducers, revealing that Brusatol markedly elevated mitochondrial Fe²⁺, which was reversed by Fer-1. Mechanistic studies showed that Brusatol induced ferroptosis by increasing Fe²⁺, ROS, and lipid peroxidation (LPO), decreasing GPX4, and increasing ACSL4. Furthermore, in A2780/DPP tumor-bearing nude mice, DHX-Fe enabled in vivo imaging of exogenous Fe²⁺ and dynamically monitored the sustained elevation of intratumoral Fe²⁺ during Brusatol treatment (days 7–14). Pharmacokinetic studies indicated hepatic and renal metabolism, and H&E staining confirmed no obvious toxicity in major organs. Therefore, as the first mitochondria-targeted NIR Fe²⁺ probe for evaluating ferroptosis inducers in chemoresistant ovarian cancer, DHX-Fe provides a powerful visual tool for overcoming tumor drug resistance.

To further improve the response speed and detection accuracy, based on the nitroxide radical reduction mechanism, Dutta et al. described BDNO, featuring a BODIPY scaffold modified with an N-oxide recognition unit for Fe²⁺ [167]. The ICT pathway of BDNO was suppressed by the N-oxide group, resulting in yellow fluorescence emission at 570 nm (λ_ex = 540 nm). Upon addition of Fe²⁺, the original emission decreased while new emission at 715 nm (λ_ex = 570 nm) increased, enabling self-calibrating ratiometric detection. BDNO offered an ultrafast response (<5 s), excellent selectivity, and high sensitivity (LOD = 41 nM), and responded effectively to Fe²⁺ over pH 4–10. Its practical utility was confirmed by the quantitative detection of Fe²⁺ in diverse food samples and environmental water samples with satisfactory recoveries. Furthermore, BDNO-functionalized paper strips and cotton buds allowed for on-site visual detection with the naked eye via both colorimetric and fluorescent modes without the need for specialized equipment. In addition, a smartphone-integrated test strip platform enabled quantitative Fe²⁺ detection with recoveries of 91.5–96.28%. These results suggest that BDNO holds promise for food safety monitoring and environmental water analysis. However, the reduced product BDN suffered from poor solubility in high-water-content systems, which may limit its NIR imaging applications in pure aqueous environments and biological systems.

Later, Jiang and colleagues reported a chromenylium–cyanine probe FTP for Fe³⁺ sensing via a spirolactam ring opening mechanism [168]. FTP was synthesized by conjugating a modified rhodamine scaffold with a diphenylphosphine moiety through a hydrazine bridge. In the solution of EtOH/PBS (v/v, 3/7), coordination with Fe³⁺ induced spirolactam ring opening, causing a sharp increase in fluorescence intensity at 743 nm and a distinct color change from pale yellow to green, enabling both fluorometric and colorimetric detection with LODs of 24.0 nM (fluorescence) and 79.7 nM (UV-Vis). In addition, FTP showed excellent selectivity and reversibility, a rapid response (<2 min), outstanding anti-interference, and favorable stability at pH of 4–10, enabling quantitative analysis of Fe³⁺ in water samples with good recoveries, as well as visual qualitative detection on filter paper and visual semi-quantitative detection on cotton swabs. Its applicability was further confirmed by fluorescence imaging of living cells and mice. Moreover, compared with the aforementioned probes, FTP exhibited improved aqueous solubility and sensitivity, but remains inadequate for bioimaging applications.

Recently, Fang’s group successfully constructed a water-soluble carbamoyl oxime-based NIR probe, Cy-LIP, utilizing an Fe²⁺-mediated reductive N–O bond cleavage mechanism [169]. In PBS solution, upon Fe²⁺ activation, Cy-LIP released a hemicyanine fluorophore, producing an 18-fold fluorescence increase at 705 nm. Cy-LIP showed high selectivity, rapid response kinetics (<20 min), high sensitivity (LOD = 51 nM), and a broad pH range (5–9). It successfully detected both exogenous and endogenous Fe²⁺ in cells and monitored Fe²⁺ dynamics during erastin/RSL3-induced ferroptosis, with malondialdehyde levels paralleling the Fe²⁺ signal, further confirming ferroptosis. Cy-LIP was mainly localized in mitochondria, showing a high PCC of 0.91 with Mito-Tracker Green, much higher than that with lysosomes (PCC = 0.43) or the endoplasmic reticulum (PCC = 0.51). Importantly, Cy-LIP detected elevated Fe²⁺ levels in murine models of rheumatoid arthritis (RA) and DILI. In the RA model, fluorescence increased progressively over time. In the DILI model, the APAP-treated group showed significantly higher fluorescence than the control, while NAC treatment markedly reduced the signal. Thus, Cy-LIP offers a versatile tool for studying iron-related disorders.

To simultaneously visualize changes in Fe²⁺ and intracellular viscosity during pathophysiological processes of hepatocellular carcinoma (HCC), in 2025, Hong, Qiu and Feng et al. synthesized a dual-detection probe QM-FV, utilizing an N-oxide moiety for Fe²⁺ recognition and a molecular rotor for viscosity sensing [170]. Upon Fe²⁺ addition, QM-FV exhibited marked fluorescence turn-on at 740 nm within 150 s, with a 30-fold fluorescence enhancement. The probe was effective over pH 3–8 and showed high selectivity toward Fe²⁺ over common metal ions, anions, amino acids, biothiols, and ROS/RNS. Meanwhile, elevated viscosity triggered a distinct orange–red emission enhancement at 587 nm, thereby enabling dual-channel independent responses to Fe²⁺ and viscosity. Owing to its high sensitivity (LOD = 25 nM), large stroke shift (290 nm), NIR emission, dual response and low cycotoxity, QM-FV enabled imaging of exogenous and endogenous Fe²⁺ in living cells and monitored the simultaneous elevation of intracellular Fe²⁺ and viscosity induced by monensin and nystatin. Moreover, the probe effectively discriminated cancer cells from normal counterparts, with both green and red channels showing markedly stronger fluorescence in cancer cells. Importantly, QM-FV was successfully applied to image Fe²⁺ and viscosity changes during erastin-induced ferroptosis and cisplatin-induced apoptosis, as well as in a mouse model of HCC, revealing a progressive increase in both parameters during disease progression. As the first dual-responsive probe capable of simultaneously monitoring Fe²⁺ and viscosity, QM-FV provides a powerful tool for investigating Fe²⁺- and viscosity-related pathological processes in HCC.

3.7. Cross-Comparison Among Different Metal Ions

In the above discussion of research progress on heavy metal ions, we first classified the probes by metal ion species and then summarized the research progress on target ions based on different NIR fluorescent platforms. To help readers better understand the latest advances and remaining challenges in this field, we cross-compared different metal ions from the following aspects and discussed the challenges they share.

In terms of fluorophore selection, NIR fluorescent probes for Cu²⁺, Hg²⁺, Zn²⁺, Pb²⁺, Cd²⁺ and Fe²⁺/Fe³⁺ show a certain preference. NIR fluorescent probes for Cu²⁺ and Hg²⁺ are often based on rhodamine, dicyanoisophorone, and hemicyanine/cyanine scaffolds; Zn²⁺ probes commonly employ isophorone and cyanine scaffolds. Pb²⁺ probes are frequently constructed on a dicyanoisophorone scaffold. Fe²⁺/Fe³⁺ probes often combine reactive recognition groups such as N-oxide or carbamoyl oxide with isophorone or xanthene scaffolds. Overall, rhodamine, dicyanoisophorone, hemicyanine and cyanine are the four most commonly used scaffolds for constructing NIR heavy metal ion probes.

From the perspective of recognition mechanisms, reversible coordination-based probes and irreversible reaction-based probes each have advantages and limitations. Zn²⁺, Pb²⁺, and some Cu²⁺ and Hg²⁺ probes mostly adopt reversible coordination mechanisms, which offer the advantages of reusability, a relatively fast response, and suitability for real-time imaging, but they may suffer from interference by metal ions with similar chemical properties. For example, the probe CL(Hg²⁺) can respond to both Hg²⁺ and Ag⁺. Fe²⁺ probes, some Cu²⁺ probes, and some Hg²⁺ probes employ irreversible reaction-based mechanisms. For instance, Cu²⁺ can induce picolinate ester hydrolysis or spirolactam ring opening of rhodamine. Hg²⁺ can induce thioacetal desulfurization, spirolactam ring opening of rhodamine, or thioester hydrolysis. Fe²⁺ can reduce N-oxide or hydrolyze carbamoyl oxide. Irreversible reaction-based probes offer advantages such as high selectivity and irreversible signal accumulation, but they may suffer from a relatively slow response, lack of reusability, and more complex evaluation of in vivo metabolism and toxicity.

In terms of interfering ions, Cd²⁺/Zn²⁺ and Hg²⁺/Ag⁺ are two of the most challenging ion pairs. Zn²⁺ and Cd²⁺ have similar ionic radii and coordination chemistry, and some Zn²⁺ probes are susceptible to Cd²⁺ interference, and vice versa. Hg²⁺ and Ag⁺ also share similar thiophilicity, and some Hg²⁺ probes produce false positive signals in the presence of Ag⁺. In addition, Fe²⁺ probes are sometimes interfered with by Cu²⁺ (e.g., probe MDJ-O). Therefore, high selectivity remains one of the core challenges in the design of NIR heavy metal ion probes.

For practical applications, the development of probes for different metal ions is uneven. NIR probes for Cu²⁺, Hg²⁺, and Zn²⁺ have been reported more extensively and have been widely used in water samples, food, cells, and in vivo imaging, whereas NIR probes for Cd²⁺, Pb²⁺, and Fe²⁺/Fe³⁺ are relatively scarce. Most probes still rely on organic cosolvents (e.g., DMSO, DMF, EtOH and acetonitrile) and have limited water solubility, which restricts their application in environmental water samples and on-site detection.

In bioimaging, most probes remain at the level of cytotoxicity evaluation, and systematic studies on biocompatibility, biodistribution, metabolism, long-term toxicity, clearance pathways, and subcellular targeting are lacking. Although some probes have been applied to cell, zebrafish, mouse, and plant imaging, cases that truly achieve clinical translation or environmental on-site application remain rare. In addition, the emission wavelengths of most probes are confined to the NIR-I region (650–900 nm), while NIR-II (1000–1700 nm) probes remain very rare, limiting the resolution and penetration depth of deep-tissue imaging.

4. Comparative Analysis of Fluorophore Platforms

Although the preceding sections have systematically reviewed the research progress of NIR fluorescent probes categorized by metal ions, it is evident that the same fluorophore platforms (e.g., isophorone, rhodamine, hemicyanie and cyanine) have been repeatedly employed for the detection of different ions. Therefore, an in-depth comparison of the intrinsic merits and limitations of these platforms is crucial for guiding the rational design of future probes.

Dicyanoisophorone (DCI)-based fluorophores are highly favored owing to their well-defined D–π–A architecture, which confers excellent ICT characteristics. Consequently, DCI-based probes generally possess large Stokes shifts (typically exceeding 150 nm, with some exceeding 300 nm), effectively avoiding excitation light scattering and self-absorption interference while enhancing detection sensitivity. Moreover, their core structure is readily amenable to modification through π-conjugation extension or heteroatom substitution, enabling precise tuning of the emission wavelength into the NIR region. Despite these advantages, DCI-based fluorophores also have limitations. The most prominent issue is their poor water solubility, which to some extent restricts their direct application in pure aqueous systems and compromises detection efficiency in biological environments. In addition, DCI-based probes often exhibit relatively low fluorescence quantum yields, which adversely affects the signal-to-noise ratio and imaging contrast. In terms of sensing mechanisms, many DCI-based probes rely on irreversible chemical reactions (e.g., Cu²⁺-induced hydrolysis or Hg²⁺-triggered desulfurization), thereby limiting their ability to monitor dynamic fluctuations of target ions in real time. Furthermore, although the emission wavelengths of DCI platforms have entered the NIR region, the emission peaks of most probes remain below 700 nm. Thus, compared with the NIR-II window (1000–1700 nm), there is still room for improvement in tissue penetration depth.

Rhodamine derivatives, owing to their unique spirolactam structure, serve as an ideal platform for constructing “turn-on” fluorescent probes. In the free state, rhodamine-based probes exist in a colorless and non-fluorescent spirocyclic form with extremely low background signals. Upon binding to specific analytes (such as metal ions), the spirolactam ring opens, accompanied by pronounced fluorescence enhancement and naked-eye color changes, thereby providing a natural “switch-on” signal for highly selective detection. Rhodamine dyes also have high fluorescence quantum yields, large molar extinction coefficients, and excellent photostability, endowing them with high sensitivity and good reproducibility in analytical detection. In recent years, by extending the π-conjugated system, introducing strong electron-donating groups, or constructing hybrid scaffolds (for example replacing the oxygen atom in the xanthene core with silicon or carbon), researchers have successfully red-shifted the emission wavelengths of rhodamine derivatives into the NIR region (>650 nm) while retaining their spirocycle ring opening response mechanism, significantly expanding their potential for bioimaging. However, rhodamine dyes generally suffer from small Stokes shifts, leading to severe self-absorption, low signal-to-noise ratios, and insufficient deep-tissue penetration, which seriously hampers their application in complex systems such as food detection, environmental monitoring, and bioimaging. In addition, the spirocycle ring opening process of many rhodamine-based probes is irreversible, preventing probe reuse and making it difficult to continuously monitor dynamic changes in target ion concentrations. Finally, rhodamine-based probes in complex biological matrices may also be constrained by environmental factors (e.g., acidic conditions) and background interference.

Cyanine dyes are among the most widely studied and applied NIR fluorophores. Their principal advantages include extremely high molar extinction coefficients (typically >10⁵ L·mol⁻¹·cm⁻¹), good fluorescence quantum yields, and facile tuning of absorption and emission into the NIR region (650–900 nm) through structural modification of terminal heterocycles and meso-substituents on the methine chain. They also generally show good biocompatibility, and certain dyes, such as IR-780, even exhibit intrinsic tumor-targeting ability. However, cyanine dyes have inherent drawbacks. Insufficient photostability is the most critical bottleneck. Under illumination, they readily undergo photooxidative degradation, mechanistically involving regioselective C–C cleavage of the methine chain by singlet oxygen to form a dioxetane intermediate that subsequently decomposes, leading to fluorescence signal decay and deteriorated probe performance, particularly in long-term imaging experiments. Poor water solubility is another key issue limiting their biological applications. Basic cyanine dyes are highly lipophilic and require modifications such as sulfonation to prepare water-soluble derivatives (e.g., Sulfo-Cy7) suitable for aqueous systems. At high concentrations, cyanine dyes also tend to self-aggregate, causing aggregation-caused quenching (ACQ) and restricting their use in scenarios requiring high probe concentrations. In terms of chemical stability, cyanine dyes are sensitive to reactive oxygen and nitrogen species and may undergo nonspecific degradation in complex biological environments. Many cyanine-based probes also rely on irreversible chemical reactions induced by target ions (e.g., carbothioate hydrolysis or desulfurization), which similarly limits reversibility.

Hemicyanine scaffolds, as asymmetric derivatives of cyanine dyes, combine NIR emission with greater molecular design flexibility. Their core advantage also stems from a unique D–π–A structure. By introducing different recognition groups, the ICT process can be effectively modulated to achieve “turn-on,” ratiometric, or quenching-type fluorescence responses toward target ions. Hemicyanines generally exhibit larger Stokes shifts (some exceeding 200 nm), effectively avoiding self-absorption and excitation light scattering interference and thereby improving detection sensitivity. In addition, hemicyanine dyes generally show good mitochondrial targeting ability, conferring unique advantages for organelle-level imaging. In terms of reversibility, some hemicyanine probes have demonstrated good reversible response capability, enabling multiple detection cycles while maintaining stable performance. However, the emission wavelengths of hemicyanine-based fluorophores are generally shorter than those of cyanines, with maximum absorption mostly limited to 650–720 nm and emission to 680–750 nm, resulting in relatively limited tissue penetration depth. Additionally, although the D–π–A structure of hemicyanines offers flexible design space, some probes still suffer from relatively low fluorescence quantum yields, especially in the long-wavelength emission region, which may affect imaging contrast. Traditional hemicyanine dyes also exhibit relatively insufficient photostability and pronounced solvatochromic effects, which limit their application in cellular microenvironment detection to some extent.

Overall, these four representative fluorophore platforms each have distinct characteristics, and their performance features show a clear correspondence with suitable application scenarios. DCI-based probes, with their large Stokes shifts and tunable emission wavelengths, are suitable for applications extremely sensitive to self-absorption and background interference, but water solubility and quantum yield remain their shortcomings. Rhodamine-based probes offer low-background advantages unmatched by other platforms through their spirocycle switch mechanism, performing excellently in high-SNR detection. However, irreversibility and relatively small Stokes shifts limit their dynamic monitoring capability and deep-tissue imaging applications. Cyanine-based probes, with long excitation/emission wavelengths and high molar extinction coefficients, occupy an irreplaceable position in deep-tissue imaging and photodynamic therapy, but their photostability and water solubility require systematic optimization. Hemicyanine platforms are molecularly flexible, and their D–π–A structure makes them ideal templates for constructing activatable probes, with intrinsic mitochondrial targeting ability, although their wavelength and quantum yield still need improvement.

Rather than pursuing a single “perfect” platform, future development of fluorophore platforms can be advanced in four complementary directions. First, hybrid fluorophores should be designed. This can be achieved, for example, by combining the D–π–A architecture of DCI with the long-wavelength emission of cyanines while exploiting the spirocycle ring opening mechanism of rhodamines to achieve highly selective responses. Second, water solubility should be optimized. The aqueous compatibility of DCI and cyanine platforms can be improved by introducing hydrophilic groups such as sulfonate, quaternary ammonium, or PEG chains. Third, reversibly responsive NIR fluorescent probes should be developed. For instance, probes based on competitive coordination or redox-reversible reactions can be designed to enable real-time tracking of dynamic target ion fluctuations. Fourth, the emission wavelength should be extended into the NIR-II region. Emission beyond 1000 nm can be achieved by further extending π-conjugation and introducing strong electron-donating groups, thereby meeting the demands of deeper tissue imaging. The synergistic implementation of these strategies is expected to advance NIR fluorescent probes for heavy metal ion detection from “capable of detection” to “precise, dynamic, and field-deployable detection.”

5. Summary and Outlook

Excessive accumulation of heavy metal ions in the environment poses a critical hazard to both ecological balance and public health. Consequently, achieving accurate quantification of these pollutants and elucidating their correlation with disease mechanisms carries considerable scientific significance and societal benefits. In recent years, NIR small-molecule fluorescent probes have become powerful tools for heavy metal ion monitoring, with notable strides made since 2022. This review provides a comprehensive and updated overview of NIR organic fluorogenic probes developed for the detection of major heavy metals, specifically Cu²⁺, Hg²⁺, Zn²⁺, Pb²⁺, Cd²⁺, Fe²⁺, and Fe³⁺. We focus particularly on the molecular design strategies, response modes, analytical performance, and biological imaging utilities of representative sensors for these seven ions. A range of high-performance NIR probes have been successfully engineered by leveraging fluorophore cores such as isophorone, rhodamine, heptamethine cyanine, and DCM derivatives, coupled with the incorporation of target-specific recognition units. The sophistication of probe design continues to advance, with sensing mechanisms encompassing a variety of photophysical pathways, including PET, ICT, FRET, CHEF, and ESIPT. The emission wavelengths of these probes generally exceed 650 nm, and in certain cases, Stokes shifts exceeding 200 nm have been achieved, substantially mitigating autofluorescence interference and markedly improving the signal-to-noise ratio in imaging applications.

From the current perspective, the development of NIR fluorescent probes exhibits the following trends. (1) The function of the probe has evolved from detecting a single target to simultaneously identifying multiple analytes. Researchers have developed multifunctional probes capable of sequentially detecting multiple targets, providing powerful tools for multi-factor correlation analysis in complex systems. For example, the NIR probe QA presented by Dong’s group enabled the sequential detection of Cu²⁺ and GSH; W-2a developed by Wang et al. responded to Hg²⁺ and ONOO⁻ with dual fluorescence signals, serving as an effective tool for studying the correlation mechanism between intracellular Hg²⁺ and ONOO⁻; DNB constructed by Zhao and colleagues achieved the sequential detection of Pb²⁺ and the pesticide glyphosate both in living cells and in vivo. (2) The application of probes has expanded from laboratory analyses to on-site rapid detection. By integrating probes with portable technologies such as smartphone-based RGB recognition, visual test strips, and SiO₂ powder-based detection platforms, NIR fluorescent probes have been successfully applied for on-site quantitative screening of heavy metal ions in complex matrices like real water samples, food, soil, and residues on the surface of traditional Chinese medicines, significantly enhancing detection convenience and practicality. (3) The application of fluorescent probes in imaging has progressively advanced from the cellular level to multi-level in vivo imaging, including plant tissues (e.g., roots of soybean, rice, vetiver, alfalfa), model organisms (e.g., zebrafish, Caenorhabditis elegans), and mammalian in vivo imaging (e.g., mice, nude mice). Some probes have even achieved 3D fluorescence imaging and tumor-targeted imaging. For instance, Zhang’s team utilized VIC to perform, for the first time, 3D fluorescence imaging analysis of the relationship between Cu²⁺ accumulation and disease progression in the brain tissue of an AD mouse model, providing a visual tool for studying the distribution, migration, and toxicity mechanisms of heavy metal ions in living organisms. (4) The practical application of probes has progressively extended from fundamental research to the exploration of disease mechanisms. Some probes have been successfully applied for the dynamic monitoring of relevant metal ions (Cu²⁺, Zn²⁺) in disease models such as Alzheimer’s disease, Parkinson’s disease, and depression, revealing the correlation between ion homeostasis imbalance and disease progression, thus opening new avenues for early disease diagnosis and mechanistic studies.

Despite the fruitful achievements in NIR fluorescent probes for heavy metal ions, several challenges in their development and translation into practical applications remain. First, the ability to achieve high-specificity recognition and anti-interference towards chemically similar ion pairs, such as Cd²⁺/Zn²⁺ and Hg²⁺/Ag⁺, remains insufficient. Second, most probes remain limited to cytotoxicity evaluation, and systematic studies on biocompatibility, biodistribution, metabolism, long-term toxicity, and subcellular targeting are still lacking. Third, matrix effects in complex samples, such as soil extracts, food matrices, and traditional Chinese medicine surfaces, can compromise detection accuracy, and robust signal correction strategies are urgently needed. Fourth, photostability, storage stability, water solubility, reproducibility, and cost remain important factors limiting on-site detection and commercialization. Fifth, emission wavelengths of most probes are confined to the NIR-I region, while NIR-II region (1000–1700 nm) probes remain scarce. Sixth, since 2022, no NIR fluorescent probes for As detection have been reported, and only one NIR probe for Cr³⁺ detection has been documented, indicating that the coverage of target analytes remains imbalanced. Finally, integrating probes into test strips or smartphone devices for portable on-site detection also requires further optimization of probe loading, signal stability, and readout methods.

Therefore, the following issues should be given more considerations in future: (1) Designing and synthesizing new fluorophores with emission wavelengths exceeding 1000 nm using strategies like expanding conjugated systems and introducing strong electron-donating groups to achieve high-resolution deep-tissue imaging. (2) Constructing smart responsive probes integrated machine learning-assisted design that can respond to specific ion and possess signal self-calibration or logic operations capabilities. (3) Developing multi-mode imaging probes that combine fluorescence imaging with photoacoustic imaging, magnetic resonance imaging and other modalities to achieve comprehensive visualization from the molecular to the whole-body level. (4) Promoting the development of standardized detection kits and portable detection devices for practical needs such as environmental pollution monitoring, food safety screening, and occupational disease prevention. (5) Deepening the application of probes in mechanism studies of complex diseases like neurodegenerative and metabolic disorders to systematically investigate the dynamic changes and regulatory mechanisms of heavy metal ions, providing new targets and strategies for disease diagnosis and treatment. (6) Expanding the scope of target metal ions by developing highly selective and sensitive NIR fluorescent probes for highly toxic metal ions such as arsenic and chromium. It is also necessary to simultaneously improve probe stability, water solubility, biocompatibility, and targeting ability and to construct standardized portable detection platforms.

In summary, NIR fluorescent probes demonstrate immense application potential and promising development prospects in the field of heavy metal ion detection. With the interdisciplinary integration of materials science, chemical biology, artificial intelligence, and other fields, it is believed that in the future, an increasing number of NIR fluorescent probes for heavy metal ions with superior performance, enhanced functionality, and broader applications will continue to emerge, providing powerful molecular tools for environmental monitoring, food safety, disease diagnosis, and life science research.