Review of Mn-Doped Semiconductor Nanocrystals for Time-Resolved Luminescence Biosensing/Imaging

Abstract

Colloidal semiconductor nanocrystals (NCs) have been developed for decades and are widely applied in biosensing/imaging. However, their biosensing/imaging applications are mainly based on luminescence-intensity measurement, which suffers from autofluorescence in complex biological samples and thus limits the biosensing/imaging sensitivities. It is expected for these NCs to be further developed to gain luminescence features that can overcome sample autofluorescence. On the other hand, time-resolved luminescence measurement utilizing long-lived-luminescence probes is an efficient technique to eliminate short-lived autofluorescence of samples while recording time-resolved luminescence of the probes for signal measurement after pulsed excitation from a light source. Despite time-resolved measurement being very sensitive, the optical limitations of many of the current long-lived-luminescence probes cause time-resolved measurement to be generally performed in laboratories with bulky and costly instruments. In order to apply highly sensitive time-resolved measurement for in-field or point-of-care (POC) testing, it is essential to develop probes possessing high brightness, low-energy (visible-light) excitation, and long lifetimes of up to milliseconds. Such desired optical features can significantly simplify the design criteria of time-resolved measurement instruments and facilitate the development of low-cost, compact, sensitive instruments for in-field or POC testing. Mn-doped NCs have recently been in rapid development and provide a strategy to solve the challenges faced by both colloidal semiconductor NCs and time-resolved luminescence measurement. In this review, we outline the major achievements in the development of Mn-doped binary and multinary NCs, with emphasis on their synthesis approaches and luminescence mechanisms. Specifically, we demonstrate how researchers approached these obstacles to achieve the aforementioned desired optical properties on the basis of the progressive understanding of Mn emission mechanisms. Afterward, we review representative applications of Mn-doped NCs in time-resolved luminescence biosensing/imaging and present the potential of Mn-doped NCs in advancing time-resolved luminescence biosensing/imaging for in-field or POC testing.

Graphical Abstact

1. INTRODUCTION

Over the last few decades, colloidal semiconductor nanocrystals (NCs) or quantum dots such as representative binary CdS NCs have been extensively studied.^1–3 Due to their tunable luminescence emission at 400–700 nm, high brightness, stability against photobleaching, and small sizes of several nanometers, they have been widely applied in biosensing/imaging as nanoprobes and are already commercially available in the market.^4–7 In spite of their optical merits, these probes containing Cd have concerning properties, such as their high toxicity to biological organisms, as well as their need for extreme care in their synthesis and handling/disposal to avoid environmental contamination. In the last 10 years, I(II)-III-VI NCs such as Cu(Zn) InS₂/ZnS or Ag(Zn) InS₂/ZnS NCs were in rapid development.^8–11 Because I(II)-III-VI NCs do not contain Cd, they are environmentally benign and are also considered relatively safer as nanoprobes for biosensing/imaging.^8,12 For these I(II)-III-VI NCs, significant efforts have been made to demonstrate their wide applications in biosensing/imaging. However, for both binary and multinary NCs, most of these applications are based on their luminescence-intensity measurement when these NCs are under continuous light excitation. Such measurement suffers from serious autofluorescence when these NCs are applied to complex biological samples, which lowers their sensitivities in various applications. As such, a general question for these NCs is, how can they be further developed to gain new optical features, which can avoid autofluorescence and achieve higher sensitivities in their biological applications?

Time-resolved luminescence measurement utilizing long-lived-luminescence probes is an efficient technique to eliminate autofluorescence.¹³ In this technique, after the excitation light source is turned off, the autofluorescence of sample matrices decays completely in tens of nanoseconds, but such probes can keep emitting luminescence for signal recording. As a result, high signal-to-background ratios (or sensitivities) can be achieved through time-resolved luminescence measurement. Due to such a merit, long-lived-luminescence probes have recently been paid much attention and widely studied.^14–17 Typical probes include lanthanide-complex probes, transition-metal-complex probes, persistent luminescence nanoparticles, and so on. Generally, these probes have limited optical properties (high-energy UV excitation, low brightness, or short lifetimes, as explained in the next section). These limited optical properties complicate the development of time-resolved measurement instruments (using sophisticated optics and electronics) and usually make such instruments complex, costly, and bulky. For instance, UV excitation needs expensive UV-grade optics and UV excitation light sources, their low brightness requires high power excitation light sources, and their short lifetimes demand high-speed electronics. As a result, time-resolved measurement has to be performed in laboratories with bulky instruments and is therefore difficult to implement out of laboratories. In order to apply highly sensitive time-resolved measurement for in-field or point-of-care (POC) testing,¹⁸ it is essential to develop probes possessing high brightness, low-energy (visible-light) excitation, and long lifetimes of up to milliseconds. Such optical features can significantly simplify the design criteria of time-resolved measurement instruments and facilitate the development of low-cost, compact, but sensitive instruments for in-field or POC testing.

Considering the challenges faced by colloidal semiconductor NCs and time-resolved luminescence measurement in their biosensing/imaging applications, the question is whether or not new probes with high brightness, long fluorescence lifetimes, and low excitation energy can be further developed from these NCs to facilitate in-field or POC applications of time-resolved luminescence measurement. Recently, Mn-doped colloidal semiconductor NCs have been in rapid development.¹⁹ Due to spin relaxation and slow inversion between the ⁴T₁ and ⁶A₁ states of Mn, Mn-doped NCs ideally should present bright-yellow luminescence (~585 nm, corresponding to electron relaxation from ⁴T₁ to 6A₁) and prolonged luminescence lifetimes of up to milliseconds.²⁰ However, according to the literature, Mn-doped NCs present different optical properties from study to study (e.g., luminescence lifetimes from hundreds of nanoseconds to several milliseconds, luminescence wavelength from orange to red, and quantum yields (QYs) from several percentages to 90%).^20,21 Although the luminescence mechanisms for Mn-doped NCs, specifically Mn-doped multinary NCs, are still in discussion/debate, the recent literature also suggested that, for Mn-doped NCs, their optical properties could be significantly affected by not just the microscopic interactions between Mn and Mn but also the ones between Mn and host NCs.^21,22 Such microscopic interactions are complex, but they also provide opportunities to tune the optical properties of Mn-doped NCs and thus achieve the desired optical features (i.e., high brightness, long fluorescence lifetimes, and low excitation energy). Figure 1 illustrates Mn-doped NCs and their typical energy transfer for Mn emission from the ⁴T₁ state to the ⁶A₁ state of Mn dopants and the principle of time-resolved luminescence measurement using Mn-doped NCs. For the energy transfer for Mn emission, it is generally believed that the electrons in the valence band get excited by excitation light to form excitons (bound states of an electron and an electron hole), and the energy of excitons is then transferred to Mn dopants for Mn emission.

Figure 1.

(A) Illustration of Mn-doped NCs and energy transfer in NCs for Mn emission from the ⁴T₁ state to the ⁶A₁ state of Mn dopants. (B) Conceptual scheme of time-resolved luminescence measurement using Mn-doped NCs.

This review aims to present the development of Mn-doped NCs toward achieving the desired optical features and elucidate why they are promising in advancing time-resolved luminescence measurement. First, we briefly reviewed and summarized the limitations of the current probes for time-resolved luminescence measurement, which were used as references for further discussions/comparisons to the optical properties of Mn-doped NCs. Second, we outlined the major achievements in the development of Mn-doped binary and multinary NCs, with emphasis on their synthesis approaches and luminescence mechanisms. Specifically, we demonstrated how researchers approached the task of achieving the desired optical features on the basis of the progressive understanding of the Mn emission mechanisms. Third, we reviewed representative applications of Mn-doped NCs in time-resolved luminescence biosensing/imaging and presented the potential of Mn-doped NCs to advance time-resolved luminescence biosensing/imaging for in-field or POC testing. Although this review attempts to provide researchers with a clear comprehension of Mn-doped NCs and its potential, it should be noted that any papers related to Cd- or Pb-based NCs were not be reviewed even though they present significant discoveries in science or technology. We also acknowledge that many important papers may be overlooked and not cited in this review. We hope that this review will bring more attention to researchers about the next wave of highly bright colloidal semiconductor NCs with long lifetimes and low-energy excitation for time-resolved luminescence biosensing/imaging.

2. CURRENT TIME-RESOLVED LUMINESCENCE NANOPROBES AND THEIR LIMITATIONS

Many current probes with luminescence lifetimes in the tens to hundreds of nanoseconds are widely available, including endogenous fluorophores, transition-metal complexes (e.g., iridium- or ruthenium-based complexes),^16,23–25 and I(II)-III-VI NCs like CuInS₂/ZnS NCs.^26–36 Their time-resolved luminescence measurement typically adopts luminescence or fluorescence lifetime imaging microscopy (FLIM), which is based on the differences in the luminescence decay rate of probes in a sample. FLIM often uses a megahertz laser excitation source in order to achieve a high-quality image within a reasonably short time frame. FLIM also requires high-speed digitizers [e.g., 12.5 GS/s (PXIe-5185 National Instruments)] and highly sensitive charge-coupled-device (CCD) cameras or photodetectors.^26,37 Besides FLIM, researchers have also adopted a highly sensitive but expensive electron-multiplying CCD (EMCCD) to catch fast images within nanoseconds in a gated time frame.³⁸ Although both FLIM and custom-designed setups can be utilized readily for laboratory-based research, they cannot be used for in-field or POC testing in austere settings due to their large sizes and operation complexity.²⁶

Lanthanide-chelated complexes with long luminescence lifetimes of up to milliseconds are well-established probes for time-resolved luminescence measurement.^39–44 In these complexes, a lanthanide element (e.g., Tb or Eu) is chelated with an organic ligand through electrostatic interaction, and the organic ligand acts as a sensitizer, transferring energy with the lanthanide element and allowing for its spin-forbidden f–f electronic transition.^45,46 Commercial FLIM-based microscopes cannot be directly used for such lanthanide complexes due to the discrepancy between their megahertz excitation sources and the millisecond lifetime of such complexes. Researchers had to develop custom-designed instruments for these probes in their specific applications. For instance, Marriott et al. built a FLIM-like instrument using an expensive and complex optical design including two rotating choppers and a CCD camera.⁴⁷ Another example is the custom design of Grichine et al. capable of temporal sampling lifetime imaging microscopy (TSLIM) for long lifetime imaging and pinhole shifting lifetime microscopy (PSLM) for time-gated imaging via the use of commercial confocal microscopes. They made use of a LSM710 microscope and a photomultiplier tube (PMT) for their time-gated imaging of lanthanide probes in cultured cells and also displayed the potential of a LSM7 LIVE microscope equipped with a CCD for much faster scanning.⁴⁸ One other example is the instrument built by Jin et al., which used a highly sensitive time-resolved CCD camera (MEGA-10EX Stanford Photonics, >$50K) for cellular study using lanthanide complex probes.⁴⁹ Jin et al. also adopted a 365 nm UV light-emitting-diode (LED; >200 mW, Mic-LED from Prizmatix, >$1500) as the excitation light source, but the cost for these LEDs and their optics (an excitation filter to cut off the long spectral tail of the LED and a UV-grade lens) is high. Lanthanide complexes as probes have also been applied in microplate-based immunoassays (the microplate can handle small sample volumes and is broadly used in bioassays) to enhance the signal-to-background ratios and thus yield lower limits of detection (LODs).^50–52 The microplate readers with time-resolved luminescence measurement usually adopt high-power xenon lamps (e.g., 75 kW) as excitation sources and mechanical choppers for time-resolved measurement. However, these lamps and choppers are large in size and complex in design and operation. As a result, such microplate readers are complex, bulky, and expensive, and they are often used in laboratories.

Lanthanide-doped polymer particles, silica nanoparticles, and fluoride nanoparticles with long luminescence lifetimes are also in development.^53–64 However, they also need high-energy UV light for excitation. Specifically, for lanthanide-doped fluoride nanoparticles, although they can be excited by near-infrared (NIR) light through upconversion for bioapplications, NIR excitation of these particles still requires high-power lasers and costly laser drivers.^65,66 These fluoride nanoparticles also suffer from low brightness or low QYs at 0.1–5%, which necessitates the need for more sensitive but expensive CCD cameras.^59,60

Ceramic luminescence materials, also referred to as afterglow materials, such as transition-metal or rare-earth-metal-doped ZnGa₂O₄, and SrAl₂O₄, have long luminescence lifetimes due to their unique electron trap states in their band gaps.^67–76 These materials are attractive for time-resolved luminescence biosensing/imaging. However, they are usually synthesized as bulk materials, and the mechanical breakdown of these materials to produce nanoparticles takes a long time (7–9 days for milling to get 200–500 nm sizes). The breakdown also introduces more surface defects, lowering their brightness. Some pioneers have reported ZnGa₂O₄ nanoparticles (~10 nm in size) using a hydrothermal process at ~220 °C for ~20 h (energy-consuming).^71–73 These particles have low QYs (<10%) after a post-thermal treatment at 1000 °C. These low-QY materials are wide-band-gap semiconductors and usually need high-powered UV light (370 nm Chauvet 18-in. 15-W UV light) in many instances.⁶⁷ A recent work reported Cr-doped ZnGa₂O₄ NCs that were ~30 nm with ~50% QY. However, they were synthesized in a solvothermal process with high temperature and high pressure, and UV excitation (~275 nm) was still required.⁷⁶ Other types of afterglow materials are metal oxide materials or their fluoride counterparts doped with Mn⁴⁺, which can produce very strong luminescence (>60% QY) in addition to a long lifetime (>8 ms).^77–82 Nevertheless, these wide-band-gap hosts still require UV excitation, and synthesis procedures to produce these Mn⁴⁺-doped materials usually involve toxic materials (e.g., hydrofluoric acid), long processes, or very high temperatures.⁸⁰ Similar to the previously mentioned oxides, these Mn⁴⁺-doped materials are usually synthesized in powder form, and further mechanical breakdown is required to mill these materials to nanoparticles, which again results in surface defects quenching the material luminescence and lifetime.⁷⁸

Among Cd- and Pb-free colloidal semiconductor NCs (~5 nm) with low toxicity, Si NCs have fluorescence lifetimes in microseconds due to their indirect band gap.^83–86 The production of Si NCs is complex and costly because it involves high temperature or highly toxic chemicals (48% HF). UV excitation is also still required.⁸⁷ Like ceramic luminescence materials, the defects on their surfaces due to a lack of a shell structure result in low brightness, causing the need of highly sensitive but expensive instruments like intensified CCDs (Andor iStar ICCDs).⁸⁷

All of the aforementioned probes are absolutely useful in specific applications (e.g., lanthanide complexes for cellular imaging and lanthanide-doped nanoparticles and persistent luminescence nanoparticles for in vivo imaging) but generally require complex and expensive instruments for time-resolved luminescence measurement. Such requirements of sophisticated instruments are rooted in the limited optical properties of these probes (short lifetime, low brightness, or high-energy UV excitation). The question is whether or not novel probes possessing long lifetime, high brightness, and low-energy excitation can be achieved to facilitate the development and use of low-cost, compact, but highly sensitive instruments for time-resolved luminescence biosensing/imaging. With recent advances in colloidal semiconductor NC research, Mn-doped NCs are promising to achieve all of the desired optical properties, and they should be the impetus to advance time-resolved luminescence biosensing/imaging to the next stage. To facilitate the comparison of all luminescence materials discussed in the paper, Table 1 lists key properties of these materials like the excitation wavelength (UV or visible light), brightness (QY), lifetime, emission wavelength, and particle size.

Table 1.

Optical Properties and Sizes of Different Probes with Time-Resolved Luminescence

probe type	excitation wavelength	emission wavelength (nm)	QY (%)	lifetime	size (nm)	photostability	ref
endogenous fluorophores	visible	400–850	<2	<16 ns	<1 (molecules)	no	26, 28–34
I(II)-III-VI NCs (with a ZnS shell)	visible	500–850	>50	100–500 ns	3–5	yes	35, 36
lanthanide complexes	UV	>450	~40	~1 ms	<1 (molecules)	no	39–46
lanthanide-doped polymer nanoparticles	UV	~620	NAa	<1 ms	100–400	no	53, 61–63
lanthanide-doped silica nanoparticles	UV	>450	<30	<1.5 ms	~65	yes	54–58, 65
lanthanide-doped fluoride nanoparticles (upconversion nanoparticles)	UV or NIR	>450	<5	<100 ms	<100	yes	59, 60, 65, 66
transition-metal complexes	visible	>450	<5	<1 μs	<1 (molecules)	no	16, 23–25
Si NCs	UV	~700	~26	<100 μs	~4	yes	83–87
persistent luminescence nanophosphors (breakdown by milling)	UV	~500	NAa	<10 min	~200	yes	67–70
persistent luminescence nanoparticles	UV	~700	<50	<5 h	<30	yes	71–76
Mn-doped ZnS/ZnS NCs	UV	~600	20–55	1–2 ms	3–4	yes	95, 98, 144, 147–151
Mn-doped ZnSe/ZnS NCs	visible	~600	3–70	<960 μs	~11	yes	96, 97, 99, 104, 105
Mn-doped I(II)-III-VI NCs (with a ZnS shell)	visible	~600	~44	>1 ms	3–5	yes	111–116, 119, 152–154

^aNA = not available.

3. SYNTHESIS AND LUMINESCENCE MECHANISM OF MN-DOPED NCS

3.1. Mn-Doped Binary NCs.

In the review of key achievements in Mn-doped binary host NCs, we aim to highlight Mn-doped ZnS and ZnSe NCs but avoid Cd- or Pb-based NCs due to their high biotoxicity, although the research of Mn-doped Cd-based NCs was also instrumental in achieving the high tunability of the NC optical properties and an in-depth physical understanding of the NC luminescence mechanisms.^19,88–94

The earliest study demonstrating Mn emission was reported by Bhargavi et al.⁹⁵ They utilized a low-temperature synthesis method by adding diethylmanganese in a tetrahydrofuran solvent to a mixture of diethylzinc and hydrogen sulfide in toluene to form Mn-doped ZnS NCs. Their NCs presented a quantum efficiency of 18% and a lifetime in the submicro-second range. A narrow emission peak at 590 nm was obtained under 300 nm UV excitation. This peak was attributed to the energy transfer from the ZnS host’s s–p electrons to the d orbital electrons of Mn dopants, which would further undergo a spin-forbidden ⁴T₁–⁶A₁ transition for Mn emission. Figure 2A illustrates the energy transition in the band gap of an excited NC without a dopant acting as an impurity center. τ_NR represents nonradiative surface recombination and τ_b–b represents a conduction band (CB)-to-valence band (VB) recombination. Figure 2B exhibits this energy transition but with an impurity center (Mn) in the band gap. τ_R represents the radiative pathway that directs energy to the ⁴T₁–⁶A₁ transition (τ_D). The luminescence excitation and emission spectra of Mn-doped ZnS NCs (dashed lines) were compared with the luminescence excitation and emission spectra of bulk Mn-doped ZnS (solid lines) in Figure 2C. In addition, it was noted that the Mn-doped NCs requires UV light for excitation. This need of UV excitation would become a great concern for biological applications of these doped NCs.

Figure 2.

(A and B) Illustration of energy transfer in excited ZnS NCs without and with Mn dopants, respectively. (C) Luminscence excitation and emission spectra of Mn-doped bulk ZnS represented as solid lines and ZnS NCs as dashed lines. All figures were reproduced with permission from ref 95. Copyright 1994 American Physical Society.

After this foundational achievement by Bhargavi et al., Suyver et al. developed Mn-doped ZnSe NCs through an organometallic synthesis.⁹⁶ Suyver et al. mixed manganese cyclohexanebutyrate dissolved in trioctylphosphine, Se dissolved in trioctylphosphine, and diethylzinc together. They then added the mixture to heated hexadecylamine through a syringe and continued the NC growth at 275 °C for 4 h to produce Mn-doped ZnSe NCs. Quantum efficiencies and luminescence lifetimes of Mn-doped ZnSe NCs with different Mn concentrations in NCs were measured. As the Mn concentration was increased from 0.2% to 0.9% in NCs, a maximum quantum efficiency of 2.7% was produced at 0.9% Mn in NCs, and the lifetime was shortened with increasing Mn concentrations (290 μs at 0.2% Mn and 190 μs at 0.9% Mn). In addition, a slight red shift in the emission spectra for increasing Mn concentrations was also observed. The NCs produced in this work were not bright (QY < 3%), the NC lifetimes were very short, and some precursors required a glovebox for handling. However, it was an important early work in disclosing the possible relationship between the Mn concentration and optical properties. It also demonstrated a new organometallic synthesis approach, which later became a general method.

At almost the same time, Norris et al. synthesized high-quality Mn-doped ZnSe NCs through an organometallic synthesis approach.⁹⁷ The researchers utilized dimethylmanganese (MnMe₂) as the Mn precursor to achieve high-quality Mn-doped ZnSe NCs with QYs of up to 22%. The researchers used electron paramagnetic resonance (EPR) and magnetic circular dichroism analysis to definitively show that Mn was not residing on the surface but was instead embedded within the NC, being substituted for Zn. In spite of the great strides made by this work, the lifetimes of these high-quality NCs were not measured, and this method still required a glovebox to handle toxic precursors in synthesis.

Mn-doping strategies in the synthesis of Mn-doped ZnSe or ZnS NCs were further explored. Pradhan et al. synthesized MnSe/ZnSe NCs through nucleation doping.⁹⁸ Nucleation doping was realized by mixing Mn and host precursors during nucleation, followed by further growth of the ZnSe shell on core NCs, as shown by Figure 3A. In their synthesis, the luminescence of Mn-doped ZnSe NCs can be tuned from 575 to 595 nm by growing a thicker ZnSe shell. It was explained that, as the ZnSe outer shell grew thicker, the crystal field for each dopant ion became more symmetric in a long range. The NCs presented QYs of up to 30%; however, their lifetimes were not measured. Karan et al. reported a growth-doping-based synthesis of Mn-doped ZnSe_xS_1–x.⁹⁹ NCs. Their strategy consisted of host nucleation in the first step and then the introduction of manganese oxide to the surface of host nuclei, which was further followed by host growth to form a shell covering the surface-absorbed Mn dopants. A depiction of the generalized version of this growth-doping process is given in Figure 3B. This method achieved Mn-doped ZnSe_xS_1–x NCs with a ~40% QY and a lifetime at the millisecond level (1–2 ms).

Figure 3.

Illustration of Mn-doping strategies: (A) nucleation doping; (B) growth doping. All figures were reproduced with permission from ref 98. Copyright 2005 American Chemical Society.

Zheng et al. utilized nucleation doping to create MnS/ZnS core–shell NCs.^100,101 In their synthesis, they first grew MnS NCs and then coated them with a layer of ZnS shell. They further annealed their NCs under different temperatures (220–300 °C). It was found that the luminescence intensity and lifetime of the MnS/ZnS NCs with a thin ZnS shell significantly decreased as the annealing time or temperature increased, while the luminescence intensity and lifetime of the NCs with a thick ZnS shell remained almost constant under all tested annealing conditions. They attributed the decrease of the luminescence intensity and lifetime of the MnS/ZnS NCs with a thin ZnS shell to the leaching of Mn ions to the surface of the ZnS shell. With a small MnS core (~2 nm), MnS/ZnS NCs achieved QYs of 40–50%. The optimal NCs presented lifetimes between 1 and 2 milliseconds.¹⁰⁰ This study indicated that high-quality MnS/ZnS NCs could be achieved by adopting small-sized MnS cores and an appropriate thickness of the ZnS shell. Zhang et al. developed similar MnS/ZnS NCs with higher QYs of 65%.¹⁰² One of the key choices in their synthesis is the use of 1-dodecanethiol as a capping agent, which could control the growth/size of MnS cores and produce smaller MnS cores. Such smaller MnS cores are important to achieve high-quality NCs. However, in this study, the NC lifetime was not measured. In these studies, all precursors are relatively easy to prepare and handle, and no glovebox was needed.

Srivastava et al. developed Mn-doped ZnS NCs with a QY of 51.3% and a lifetime of 0.37 ms in a gram scale.¹⁰³ They put all precursors of Zn, Mn, and S along with a capping agent in an organic solvent and then heated the mixture to produce Mn-doped ZnS NCs. Afterward, excess Zn precursors were added to the reaction solution to grow a ZnS shell. Their work highlighted the importance of the amount of Mn and the ratio of the initial Zn and S precursors in achieving a high QY. Figure 4A shows the luminescence spectra (red line) of Mn-doped ZnS NCs with a ZnS shell, in comparison with the same NCs with no ZnS shell (black line). The enhanced luminescence intensity clearly indicates the importance of ZnS shell growth on the Mn-doped ZnS NC surfaces. Figure 4B presents the luminescence decay curve, indicating a submilli-second lifetime. EPR spectra of Mn-doped ZnS before overcoating (blue line) and after overcoating (black line) with a shell of ZnS are shown in Figure 4C. In these EPR spectra, for Mn-doped ZnS NCs without ZnS shells, additional peaks in the lower and higher magnetic fields can be observed, which may suggest the presence of Mn dopants on the ZnS NC surfaces. After overcoating a ZnS shell, such peaks disappear, confirming the encapsulation of Mn dopants inside the shell of Mn-doped ZnS NCs. The hyperfine structures in the EPR spectra of Mn-doped ZnS NCs clearly indicate the existence of Mn dopants in host ZnS NCs. A transmission electron microscopy (TEM) image of Mn-doped ZnS NCs with ZnS shells is shown in Figure 4D, indicating NC sizes of 3–4 nm. This study is significant because it produced high-quality NCs in gram levels. However, this study did not present the possible effects of the synthesis conditions (the amount of Mn and Zn/S molar ratio) on the luminescence lifetime of NCs. Moreover, the NCs still needed UV excitation.

Figure 4.

(A) PL spectra of Mn-doped ZnS with no ZnS shell growth represented as a black line and with ZnS shell growth as a red line. (B) Luminescence decay of Mn-doped ZnS with a ZnS shell. (C) EPR spectra of Mn-doped ZnS with no ZnS shell growth represented as a blue line and with ZnS shell growth as a black line. (D) TEM image of Mn-doped ZnS NCs with a ZnS shell. All figures were reproduced with permission from ref 103. Copyright 2010 American Chemical Society.

Pu et al. developed Mn-doped ZnSe/ZnS through nucleation doping.¹⁰⁴ Mn and Zn precursors were dissolved in a heated organic solvent, followed by the rapid injection of a Se precursor to form Mn-doped ZnSe core NCs. The core NCs were grown with a ZnS shell to form Mn-doped ZnSe/ZnS NCs. In their synthesis, they prepared a series of NCs with different Mn amounts in the synthesis. All of these NCs presented high QYs (>70%) and nearly single-exponential decay dynamics. Specifically, in the first step of their study, the researchers prepared Mn-doped ZnSe NCs (~6.5 nm in size) and found that the single-exponential decay lifetime can be tuned from ~100 to ~850 μs by increasing the number (32–431) of Mn dopants per NC, while the NC luminescence emission was maintained around 580 nm and a QY of >70%. Figure 5A shows the luminescence and absorption spectra of NCs with different Mn amounts per NC. It was observed that the number of Mn dopants per NC did not significantly affect the NC luminescence and absorption spectra. Figure 5B shows the luminescence decay curves of Mn-doped ZnSe/ZnS with different amounts of Mn per NC. These NCs with different Mn amounts present single-exponential luminescence decay dynamics, which should be associated with a monodisperse chemical environment of the Mn dopants within a NC, including a Mn–Mn distance, chemical coordination to Mn dopant, etc. Figure 5C presents the relationship between the NC luminescence lifetime and Mn/Zn molar ratio in the synthesis. The data in Figure 5A,B indicate the luminescence-lifetime tunability of these NCs by the number of Mn dopants per NC. However, the researchers further found that such Mn-doped ZnSe NCs were not photostable under UV irradiation. To overcome this issue, they grew a thick ZnS shell on Mn-doped ZnSe NCs to form Mn-doped ZnSe/ZnS NCs (~11 nm in size;Figure 5D). Although they found that there were some slight changes regarding the NC luminescence properties as shown in Figure 5E,F before and after ZnS shelling, the high QYs and the wide tunability of the luminescence lifetime remained. To demonstrate the possible application of these NCs for time-resolved imaging, the researchers prepared inks using three types of Mn-doped ZnSe/ZnS NCs with different PL lifetimes (86, 220, and 960 μs) and printed patterns of a circle and the alphabetical characters of “M” and “n” on a surface with three different inks, respectively. They then used a pulsed laser (Opolette 355 nm) to excite their sample and captured the lifetime decay of the emission with a high-speed camera (Andor iStar 334T ICCD camera) at 67 μs intervals. Figure 5G shows the captured images from the 1st, 5th, and 13th intervals, respectively, and the optical setup for the time-resolved imaging of the developed NCs. In this study, the researchers also performed bioimaging, which will be discussed in the applications section of this review.

Figure 5.

(A) Luminescence spectra and absorbance spectra of Mn-doped ZnSe NCs with different Mn amounts per NC. (B) Luminescence lifetime decays with different Mn amounts per NC. (C) Luminescence decay versus the Mn/Zn molar ratio. (D) TEM images of Mn-doped ZnSe without a ZnS shell (top) and with a ZnS shell grown on the surface of the NC (bottom). (E) XRD patterns of Mn-doped ZnSe without a ZnS shell (blue) and with a ZnS shell grown on the surface. (F) Luminescence decay of Mn-doped ZnSe without a ZnS shell (blue) and with a ZnS shell (red). (G) Images of patterns printed using three different types of Mn-doped ZnSe/ZnS NCs (with lifetimes at 86, 220, and 960 μs, respectively) at different time intervals after a single pulsed excitation and the instrument setup to obtain these images. All figures were reproduced with permission from ref 104. Copyright 2016 American Chemical Society.

Still based on the Mn-doped ZnSe/ZnS NCs developed by Pu et al., Yang et al. prepared NCs with different Mn amounts per NC (x% = Mn/Zn molar ratio per NC, and x = 0.1, 15, and 50). These NCs were around 10.5 nm in size and possessed 70–90% QYs. Moreover, X-ray diffraction (XRD) analysis on these NCs indicated that the Mn doping level did not affect the NC crystal structure. The researchers further analyzed these three types of NCs using temperature-dependent steady-state luminescence analysis, temperature-dependent EPR analysis, and temperature-dependent time-resolved luminescence analysis.¹⁰⁵ In these analyses, NCs with high Mn amounts (x% = 15% and 50%) presented much different characteristics compared to NCs with a very low Mn amount (x% = 0.1%). As the Mn amount increases, distinct shifts in the temperature-dependent luminescence peaks and a full-width at half-maximum (fwhm) can be seen in parts A and B of Figure 6, respectively. Parts C–E of Figure 6 show the temperature-dependent time-resolved luminescence of NCs with different Mn amounts. NCs with 0.1% Mn present a single-exponential decay for all measured temperatures. NCs present shorter lifetimes as the Mn amount increases to 15% and 50%. NCs with 15% or 50% Mn have a single-exponential decay at room temperature but show a biexponential decay as the temperature decreases. Parts F–H of Figure 6 show the temperature-dependent EPR spectral profiles of NCs with different Mn amounts; generally, a hyperfine structure in the EPR spectrum of NCs with a very low Mn amount but a broadened shape of the EPR spectrum is observed with increasing Mn amounts. The researchers concluded that the luminescence properties of NCs with high Mn concentrations (x% = 15% and 50%) are related to the Mn–Mn coupling effects. To demonstrate the concept of Mn–Mn coupling, the researchers illustrated the energy levels of Mn–Mn dimers and possible energy transition pathways, as shown in Figure 7.When magnetic coupling between Mn ions occurs in high Mn concentration samples, the ground state (⁶A₁) of these ions will interact antiferromagnetically and split into six different states (S₀–S₅), while their excited state (⁴T₁) will split into four different states (S₁–S₄). According to the researchers, the temperature would affect the population of the split energy levels, and thus the temperature would impact the energy transition pathways for luminescence. As a result, as the temperature varies (increases or decreases), the temperature-dependent luminescence spectra of NCs with high Mn amounts present significant shifts with respect to their luminescence peaks and fwhm’s compared to NCs with very low Mn amounts. The split energy levels due to Mn–Mn coupling would also further break the rules of the forbidden transition between ⁴T₁ and ⁶A₁ if the transition is between a pair of split energy levels: one, for instance, from ⁴T₁ and the other from ⁶A₁ with the same S. Consequentially, the Mn emission lifetime would become shorter depending on the Mn–Mn coupling strength (or the Mn amount in NCs). The biexponential decay dynamics of NCs with 15% or 50% Mn at low temperatures probably is due to the two temperature-dependent and competitive decay channels caused by Mn–Mn coupling. The Mn–Mn coupling theory also well explains the dependence of the EPR spectral profile on the Mn amount in NCs; a hyperfine structure in the EPR spectrum of NCs with a very low Mn amount presents the electron structure of isolated Mn dopants, but a broad shape of the EPR spectrum for NCs with a very high Mn amount is due to the further splitting of the split energy levels or heavy Mn–Mn coupling.^106–108

Figure 6.

(A) Temperature dependence of the luminescence peaks with 0.1%, 15%, and 50% Mn. (B) Temperature dependence of the fwhm’s with 0.1%, 15%, and 50% Mn. (C–E) Mn-emission decay curves measured at the peak position between 15 and 350 K for NCs with 0.1%, 15%, and 50% Mn, respectively. (F–H) EPR spectra of NCs with 0.1%, 15%, and 50% Mn at 10, 50, and 100 K, respectively. All figures were reproduced with permission from ref 105. Copyright 2019 American Chemical Society.

Figure 7.

Illustration of the splitting of the energy levels of the Mn ⁴T¹ and ⁶A₁ states and the possible energy transition pathways in Mn–Mn coupling. The figure was reproduced from ref 105. Copyright 2019 American Chemical Society.

In spite of all of these impressive achievements and the new physical understandings of Mn-doped binary NCs, the host ZnS NCs generally still require high-energy UV-light excitation, which would burden the instrument development for time-resolved luminescence measurement. Mn-doped ZnSe NCs can be excited by visible light (405 nm); however, there are still some concerns about the toxicity of Se in synthesis and biological applications. New host NCs for Mn doping were explored and provided an alternative avenue to achieve low-energy excitation, high brightness, low toxicity, and long lifetimes.

3.2. Mn-Doped Multinary NCs.

Unlike binary NCs, multinary I(II)-III-VI NCs emit their luminescence based on the donor–acceptor-pair model.^{8–11,109,110} Acceptor states could be from the III element interstitials and I elements occupying the III sites. Meanwhile, donor states could be from S vacancies, I interstitials, or III elements occupying the I states. The energy transition between the donor and acceptor states is responsible for the luminescence of these multinary NCs. Composition tuning of these NCs is an important strategy to adjust the NC electronic structures and thus their optical properties. Different from Mn emission in binary NCs, the Mn-emission energy levels in multinary NCs are significantly impacted by the electronic structures of host NCs. As a result, the optical properties (optical spectra, brightness, and lifetimes) of Mn-doped I(II)-III-VI NCs would be significantly affected by composition tuning of host NCs.

In the early exploration of Mn-doped multinary NCs, Manna et al. developed Mn-doped CuInZnS and AgInZnS NCs and obtained pure Mn emission in both types of host NCs.¹¹¹ In their synthesis, they tried both growth and nucleation doping. In growth doping, AgInS or CuInS NCs with the Ag/In or Cu/In molar ratio at 1/10 were formed first, and Zn and Mn precursors were added to the NC surfaces to form Mn-doped AgInZnS or CuInZnS NCs. In nucleation doping, all metal precursors (Ag or Cu, In, Zn, and Mn with their own certain amounts) were mixed, followed by a quick S injection to form Mn-doped AgInZnS or CuInZnS NCs. Although the researchers did not clarify the exact positioning of the Mn dopants in NCs, they confirmed the presence of Mn dopants in NCs by inductively coupled plasma and EPR data. The Mn-doped NCs produced in both approaches presented high QYs (>40%) and long luminescence lifetimes at 1–2 ms. This study identified the importance of an appropriate amount of Zn in NCs for luminescence properties because Zn in NCs can ensure a wide energy band gap for the insertion of Mn emission-energy levels in the band gap and thus achieve pure Mn emission. However, this study did not investigate how the concentration levels of Mn and Ag (or Cu) in the NC synthesis would affect the luminescence properties (lifetimes and optical spectra).

Cao et al. developed Mn-doped CuInS/ZnS NCs with a photoluminescence (PL) QY of 66% and a lifetime of 3.78ms.¹¹² In their synthesis approach, CuInS NCs with a certain Cu/In ratio were formed first, and then a mixture of Zn and Mn precursors was coated onto the formed core NCs, followed by the sequential coating of the ZnS shell. During attempts to attain a high QY and a long lifetime, different Cu/In ratios (1/10–1/2) were evaluated, along with different Mn doping concentrations. In terms of the Cu/In ratios, lower ratios were shown to result in more narrow, refined emission peaks. It was explained that, as the ratio decreases, the band gap of the CuInS core NCs increases (InS has a larger band gap than CuS). As a result, the band gap became large enough to confine the Mn transition energy levels within this gap, leading to well-resolved or pure Mn emission. When varying the Mn concentration in the synthesis, they found that, as the Mn concentration increased, the NC luminescence spectrum presented a red shift from the original 550 to 610 nm, and the spectrum shape altered from a broad profile to a narrow one until a well-resolved Mn emission was achieved. The changes of the luminescence spectrum versus the Mn concentration were explained by involvement of the Mn emission centers in NCs quenching any defect based or donor–acceptor-based emission. The researchers also used EPR data to confirm the presence of Mn dopants in NCs.

Cao et al. later explored Mn-doped ZnInS NCs using nucleation doping with tunability of the Zn/In ratio, Mn concentration, ZnS shell, nucleation temperatures, the amount of ODE-S, etc.¹¹³ Under optimal synthesis conditions, Mn-doped ZnInS/NCs presented a QY of 56% and a lifetime of 4.2 ms. Like many other synthesis parameters, the Zn/In ratio significantly impacted the NC QY (Figure 8A). It was explained that Mn has a size similar to that of In but larger than that of Zn, and thus a low Zn/In ratio caused more Mn ions to diffuse into the host NCs so that NCs would have more Mn emission sites to enhance the QY. Similar to the aforementioned binary models such as Mn-doped ZnSe,⁹⁷ Mn replaces a host element in the lattice structure, which then embeds the Mn inside the core of the NC. As the Zn/In ratio was further reduced, more In ions would be replaced by Mn ions in the lattice structure due to their similar size. Because these two types of metal ions have different valences, a charge imbalance in NCs will be produced to quench NC luminescence, which would result in a lower QY. Note that the replacement of In ions by Mn dopants in the crystal lattice is a hypothesis, and it still needs to be further experimentally validated. Additionally, the researchers found that that the Mn concentration significantly impacted the NC QY, as shown in Figure 8B. The researchers explained that a high amount of Mn doping would allow for a high amount of Mn ions serving as acceptor sites that could produce a high QY. However, as Mn concentrations increased further, the NC QY dropped probably because the exciton energy is transferred from one Mn ion to one another through nonradiative transitions. This energy would eventually end up at a defect state, thereby lessening the QY. This study emphasizing optimization of the NC luminescence intensity presented excellent results, but it did not explore how the Zn/In ratio and Mn concentration affected the NC luminescence lifetime. Moreover, ZnInS/ZnS host NCs are also still wide-band-gap-semiconductor NCs, and the optimal Mn-doped ZnInS/ZnS NCs still need UV excitation.

Figure 8.

(A) Absorbance and luminescence spectra of Mn-doped ZnInS/ZnS NCs with different Zn/In molar ratios. (B) Luminescence spectra of Mn-doped ZnInS/ZnS NCs with different Mn concentrations. (C) Illustration of the possible effects of microenvironments surrounding the Mn dopant on Mn-emission-energy states in ZnS NCs and CuInZnS NCs as host NCs. Parts A and B were reproduced with permission from ref 113. Copyright 2019 Royal Society of Chemistry. Part C was reproduced with permission from ref 21. Copyright 2016 WILEY-VCH.

On the basis of reviewing the progress in Mn-doped NCs, Pradhan summarized and analyzed the luminescence properties of many Mn-doped binary and multinary NCs.²¹ Pradhan noted that although the ideal Mn emission should be fixed at 585 nm (yellow), Mn-doped NCs from many studies presented a wide luminescence range from yellow to red (up to 650 nm). It was also noted that some types of Mn-doped NCs presented shorter lifetimes as the Mn concentration increased in NCs and that the lifetime of Mn-doped NCs was also very dependent on the selection of host NCs (e.g., binary ZnS or multinary Cu–Zn–In–S). On the basis of such observations, Pradhan proposed that, besides the effects of Mn–Mn interaction (or magnetic coupling between Mn dopants), the luminescence spectrum or lifetime of Mn-doped NCs could be affected by the microenvironments surrounding the Mn dopants. Such microenvironments (i.e., neighboring elements or surface ligands) could interact with Mn dopants to cause the energy-level splitting of the ⁴T₁–⁶A₁ transition states. This splitting would result in a luminescence red shift, and it could also impact the NC lifetime. Figure 8C presents Pradhan’s proposed models of Mn dopants in different environments and the possible energy-level splitting of the Mn ⁴T₁ and ⁶A₁ states due to the interaction between Mn dopants and their surrounding environments.

Zaeimian et al. developed Mn-doped AgZnInS/ZnS NCs and investigated how the NC luminescence properties were affected by the Ag and Mn levels in NCs.¹¹⁴ The researchers used the nucleation-doping approach. In their synthesis, Mn, Zn, and In precursors were fixed at certain amounts, but the Ag concentration level was varied (i.e., from 0 to 0.1 mmol). After all metal precursors were mixed and dissolved in organic solvents, a slightly excessive amount of S was quickly injected to react with all metal precursors to produce Mn-doped AgZnInS core NCs. Then, a ZnS shell was grown/coated on the core NCs to form Mn-doped AgZnInS/ZnS NCs. In this study, it was found that, as the Ag concentration was increased, the NC absorption spectrum exhibited a red shift (Figure 9A), which rendered the NC excitable by visible light. The red shift of the absorption spectrum resulted from the narrowing-down of the gap between the donor states and the acceptor states; Ag contributes to the acceptor states, and more Ag elements in NCs would raise the acceptor states toward the donor states. The produced NCs also presented a luminescence red shift (from 600 to 625 nm) as the Ag concentration increased, as shown in Figure 9B. Additionally, when the Ag concentration was low (e.g., 0–0.05 mmol), the NC presented similar levels of lifetimes at around 1 ms and the highest QY (~44%) was achieved at Ag = 0.05 mmol. However, as the Ag concentration increased (e.g., Ag = 0.1 mmol), the NC lifetime and QY significantly decreased to 0.16 ms and 23%, respectively. This lifetime change can be seen in Figure 9C. An increasing amount of Ag in NCs would hypothetically interact with Mn, which could then cause the ⁴T₁ and ⁶A₁ energy levels to split, which could then cause the red shift and also a shorter lifetime. With the optimal host NCs, the researchers further doped the host NCs with Mn in different concentrations. It was observed that, as the Mn concentration increased to a high level, the NC lifetime again decreased, and a slight red shift was also observed. The red shift and the lifetime drop-off were attributed to the Mn–Mn interactions (magnetic coupling between Mn and Mn due to their short distances). Overall, this paper systematically investigated the effects of the Ag and Mn concentration levels on the NC optical properties. With the optimal synthetic conditions, the NCs present a QY of 44%, a lifetime of 1.2 ms, and low energy excitation. The researchers then constructed a compact apparatus (consisting of a pulsed 405 nm laser, a photosensor, and other necessary circuitry) and demonstrated the feasibility of optimized NCs for time-resolved luminescence measurement using such a compact instrument, indicating the application potential of the optimal NCs in time-resolved luminescence biosensing. It should also be noted that Mn-doped AgInZnS/ZnS NCs did not present well-resolved or pure Mn emission like the similar NCs prepared by Manna et al.¹¹¹ The overall luminescence spectra of Mn-doped AgInZnS/ZnS NCs by Zaeimian et al. presented a broader profile than that of the pure Mn emission. On the basis of this study, Lee et al. further explored the effect of Zn etching on Mn-doped AgInZnS NCs and found that Zn etching instead of ZnS shell coating can achieve a higher QY (~58%) and a longer lifetime (~1.4 ms) while still achieving low-energy excitation at 405 nm.¹¹⁵

Figure 9.

(A) Absorption spectra of Mn-doped AgZnInS/ZnS NCs with different Ag concentrations. (B) Luminescence spectra of Mn-doped AgZnInS/ZnS NCs with different Ag concentrations. (C) Lifetime decays of Mn-doped AgZnInS/ZnS NCs with different Mn concentrations. (D) Luminescence spectra of Mn-doped CuZnInS/ZnS NCs with different Cu concentrations (or Cu/Zn molar ratios). (E) Illustration of the energy-transfer pathways and the possible interaction between Cu and Mn as the Cu concentration increased in Mn-doped CuZnInS/ZnS NCs. (F) Illustration of the microenvironments surrounding the Mn dopant in Mn-doped AgZnInS/ZnS NCs or CuZnInS/ZnS NCs and the interaction between Ag (or Cu) and Mn. Parts A–C were reproduced with permission from ref 114. Copyright 2018 Elsevier. Parts D and E were reproduced from ref 116. Copyright 2019 Springer Nature.

Using a synthesis approach similar to that of Mn-doped AgZnInS/ZnS NCs, Harrison et al. developed Mn-doped CuZnInS/ZnS NCs and investigated how the NC luminescence properties were affected by the Cu and Mn levels in NCs.¹¹⁶ Compared to AgZnInS/ZnS NCs, similar results were obtained for Mn-doped CuZnInS/ZnS NCs, and the same mechanism explanations for Mn-doped AgZnInS/ZnS NCs are applicable to Mn-doped CuZnInS/ZnS NCs regarding the effects of the Cu and Mn concentrations on the NC luminescence properties. However, the red shift of the absorption and luminescence spectra of Mn-doped CuZnInS/ZnS NCs are relatively less significant, and also their lifetimes are relatively longer, even when the Cu tuning range is the same as the Ag tuning range. For instance, Ag tuning (0–0.1 mmol) caused a large continuous red shift (600–625 nm) in Mn-doped AgZnInS/ZnS NCs; however, Cu tuning (0–0.1 mmol) only resulted in a slight continuous shift (600–605 nm), as shown in Figure 9D. A significant red shift was observed for Mn-doped CuZnInS/ZnS NCs when the Cu concentration was set at 0.4 mmol. Figure 9E illustrates the possible role of Cu in the NC electronic structures and the possible interaction between Cu and Mn. According to the literature,^117,118 Cu in NCs would introduce isolated midgap d orbitals above the valence-band edge. Because the Cu amount is in a low-level range (e.g., 0–0.1 mmol), the midgap d orbitals could be isolated and localized above the valence band and may not impact the Mn emission. This would explain the slight red shift from 600 to 605 nm. Because the Cu amount significantly increases, the midgap d orbitals could further stack up to interfere with the ⁶A₁ state of Mn dopants. Such an interaction could cause a significant red shift as well as a drop in the lifetime. Additionally, it should be noted that, for Cu and Ag in the same mole range (0–0.1 mmol), Mn-doped CuZnInS/ZnS NCs presented a well-resolved Mn emission profile in their luminescence spectra, while Mn-doped AgZnInS/ZnS NCs possessed broad luminescence spectra with long tails toward the long-wavelength end. The researchers attributed all aforementioned differences between Mn-doped CuZnInS/ZnS NCs and Mn-doped AgZnInS/ZnS NCs to the atomic sizes of Cu and Ag. Because Ag is much larger in its size (including its nucleus and electron clouds) than Cu, the distance between Ag and Mn is shorter than that between Cu and Mn. As a result, the interaction between Ag and Mn through their orbital electron clouds is stronger than that between Cu and Mn, as shown in Figure 9F. Such a strong interaction between Ag and Mn probably is the reason that a much smaller amount of Ag would significantly impact the microenvironment of Mn, causing a significant red shift of the optical spectra, more shortened lifetimes, and a broadened spectral profile. However, the size effect of Ag and Cu is just a hypothesis, and it should be further verified experimentally. The optimal Mn-doped CuZnInS/ZnS NC obtained a QY of 33% and a lifetime ~3.18 ms, and they can easily be excited at 405 nm to emit luminescence. The studies by Zaeimian et al. and Harrison et al. investigated the roles of Mn, Ag, and Cu in affecting the luminescence properties of Mn-doped (Ag/Cu) ZnInS/ZnS NCs and answered some remaining questions from the work by Manna et al.¹¹¹

For Mn-doped multinary NCs, it is generally thought that it is beneficial to add Zn in the synthesis of core NCs, in order to form a stable ZnS crystal phase, augment the NC band gap, and thus facilitate the insertion of Mn emission-energy levels into the band gap. Lee et al. explored the synthesis of Mn-doped CuGaS/ZnS NCs without Zn addition in core NCs.¹¹⁹ They dissolved Mn, Cu, and Ga precursors to an organic solvent with high stoichiometric ratios among these precursors and then injected a slight excess of S to react with all of the metal precursors. Afterward, they added Zn and S precursors in a dropwise and alternative approach to coat ZnS on the core NCs. From the view of materials science and engineering, under a high stoichiometric ratio between Cu and Ga, the core NCs may not form a stable lattice phase but could be in amorphous phases (as implicated in the XRD pattern of the core NCs presented in this study), and the ZnS coating on the core NCs may not change their amorphous phases. However, the produced CuGaS/ZnS NCs showed a stable phase close to the ZnS crystal phase. The optimal NCs also presented well-resolved Mn emission with a QY of 47% and a lifetime of 3.67 ms, and they were excitable by 405 nm visible light. The researchers hypothesized that the core NCs probably underwent a phase change during the ZnS coating process; some Zn and S coated on the core NC surfaces could diffuse into the core NCs to restructure the core lattice to form a stable phase and also reconstitute the NC energy band gap for Mn emission. The researchers pointed out that such a process could happen in the synthesis of Mn-doped CuInS/ZnS NCs, as reported by Manna et al.¹¹¹ and Cao et al.,¹¹² where a three-step synthesis was reported: formation of the CuInS NCs with a Cu/In ratio at 1/10 or other values, then coating of the Zn and Mn precursors, and further sequential coating of the ZnS shell. CuInS NCs under such a high off-stoichiometry may not form a ZnS phase, but the final CuInS/ZnS NCs present a phase structure similar to those of ZnS crystals through XRD analysis. It is highly possible that a phase-change procedure occurred in the ZnS coating for Mn-doped CuInS/ZnS NCs.

For Mn-doped I(II)-III-VI NCs as discussed above,^{111,113–116,119} their luminescence decays usually present biexponential decay dynamics at room temperature, which include a fast decay component and a slow decay component. The decayed luminescence intensity I(t) of such NCs can be expressed as $I(t)=A_1{e}^{-t/{\tau }_1}+A_2{e}^{-t/{\tau }_2}$, where A₁ and A₂ are the amplitudes of the fast and long decay components at t = 0, respectively, and τ₁ and τ₂ are the short and long lifetime parameters, respectively. For such NCs, the origin of the short lifetime τ₁ of these NCs is still under debate. According to the literature,^{113–116,119} the short lifetime τ₁ could result from Mn–Mn coupling or the interaction of Mn with the host lattice field. Its related magnitude (A₁) should indicate the degree of the interaction of between Mn and Mn or between Mn and the host NC. Due to the nature of Mn emission for long luminescence lifetimes, the long lifetime (τ₂) may be related to the energy transition from Mn ⁴T₁ states to Mn ⁶A₁ states of isolated Mn dopants that are not be significantly affected by the NC lattice fields or adjacent Mn dopants. The magnitude A₂ corresponding to τ₂ should indicate the degree of such an isolation status of Mn dopants in NCs.

For Mn-doped I(II)-III-VI NCs, Figure 10 presents the general strategies and experimental methods to tune the luminescence properties of Mn-doped NCs. The general strategies include Mn–Mn and Mn–host NC interactions. To use Mn–Mn interaction to tune the NC luminescence properties, experimentally Mn concentrations in synthetic reactions can be adjusted to control the Mn amount per NC. A higher Mn amount per NC usually would cause more inner defects and increase the Mn–Mn coupling strength, which would result in a decrease of the NC brightness and lifetime and a red shift of the luminescence. Therefore, in the synthesis of Mn-doped NCs, the Mn concentration level in the synthesis needs to be optimized to achieve NCs with high-quality luminescence. Mn–host NC interaction is another strategy. In such a strategy, the composition of host NCs usually is tuned experimentally (e.g., changing the molar ratio of In/Zn, Ag/Zn, or Cu/Zn for host NCs). The composition tuning will change the host lattice field or the microenvironment surrounding Mn dopants, which would further affect Mn-related luminescence including its brightness, lifetime, excitation, and emission. Note that, for Mn-doped binary NCs, the composition of the host binary NCs is fixed (not tunable), but adjusting the Mn concentration per NC or doping Mn together with other dopants may be more feasible approaches in NC synthetic experiments.

Figure 10.

General strategies and experimental methods used in Mn-doped NCs for tuning of the NC luminescence properties.

Although all aforementioned papers focused on the development of Mn-doped NCs toward achieving the desired optical features as discussed in the Introduction, it should be noted that there are many studies using Mn doping to tune the NC luminescence in a wide-wavelength range (from yellow to red to NIR)^120–122 or to achieve NCs with double emission peaks covering green to red for the application of photovoltaic devices (e.g., LEDs).^123–130 Although the exact luminescence mechanisms still need to be investigated further, the Mn–Mn coupling and interaction between Mn and host NCs still can be applied to explain unique luminescence properties in these studies, as discussed by many researchers in their own studies. Generally, it was observed that Mn-doped NCs with emission color tuning presented short lifetimes in the range of several microseconds to tens or hundreds of microseconds. Mn-doped NCs with dual emission peaks showed a lifetime of hundreds of nanoseconds for the blue peak and a lifetime of microseconds or up to milliseconds for the yellow or orange peak. These Mn-doped NCs may still be used as time-resolved luminescence probes for multiplexing measurement due to their different lifetimes at different wavelengths, if appropriate instruments are available for specific biosensing/imaging applications.

4. APPLICATIONS OF MN-DOPED NCS IN TIME-RESOLVED LUMINESCENCE BIOSENSING/IMAGING

The literature has reported many biosensing/imaging applications of Mn-doped NCs but mainly with luminescence-intensity-based measurement (not time-resolved luminescence measurement). These types of work are not the target of this review. In this section, only the application of these Mn-doped NCs in time-resolved luminescence biosensing/imaging will be reviewed. For each representative application, we mainly looked into the synthesis, the optical properties, the surface modification of Mn-doped NCs, what analyte(s) these NCs were used to detect, what instrument was used in time-resolved luminescence measurement, and what sensitivities or LODs were achieved in the detection. For the selectivity of biosensing/imaging, one should refer to the details included in the cited papers. It should also be noted that time-resolved luminescence measurement can also be achieved with commercially available equipment, such as homogeneous time-resolved fluorescence (HTRF) assay technology and related instruments (e.g., Perkin EnVision Multilabel Micro-plate Reader Pred 2105-AV), which are commonly employed by researchers.^131–134 However, this technology is usually developed for lanthanide-based probes, which require UV excitation. As a result, the measurement instruments for HTRF assays usually are supplied by high-power Xe lamps. These Xe lamps and their necessary optical accessories cause these commercial instruments to be very complex and bulky in size and thereby make them difficult for in-field or POC testing. Focusing on the Mn-doped NCs as time-resolved luminescence probes, this section aims to show the progress from their bioapplications using very bulky laboratory-based instruments to those using compact but sensitive custom-designed instruments, which may be more suitable for in-field or POC testing.

4.1. Surface Modification.

In many cases, due to nonpolar-solvent-based syntheses, colloidal semiconductor NCs naturally possess hydrophobic ligands such as trioctylphosphine, trioctylphosphine, oleic acid, or 1-dodecanethiol on their surfaces. These NCs are not soluble in aqueous solutions. Surface modification of such NCs allowing them to be dissolved in a water phase is an important step toward their biological application. Ligand exchange and amphiphilic polymer encapsulation are two well-established methods for NC surface modification.^135,136

As illustrated in Figure 11A, ligand exchange is a process in which the surface hydrophobic ligands of the NCs are replaced by other water-soluble organic compounds (typically thiol-possessing hydrophilic ligands).^137–139 Thiol-possessing hydrophilic ligands, such as 3-mercaptopropionic acid (MPA), dihydrolipoic acid (DHLA), thiol–poly(ethylene glycol) (PEG) with functional groups on the other end of PEG (HSPEG functional groups), or thiol–zwitterions (like HS-sulfobetaine and HS-carboxybetaine), can be applied to replace original hydrophobic ligands on NC surfaces via strong binding to the NCs due to the high chemical affinity of thiols to NC surface metals.^135,136 Once the NC surfaces have been successfully modified with hydrophilic ligands, the NCs with hydrophilic ligands will be soluble in the water phase. Polymer encapsulation, as shown in Figure 11B, utilizes amphiphilic polymers (one long C backbone with multiple short hydrophobic chains on one side and multiple hydrophilic groups on the other side) to encapsulate NCs.^140–143 In the encapsulation process, the short hydrophobic chains of the polymers will interact with the hydrophobic ligands on NC surfaces through hydrophobic–hydrophobic interaction. As a result, the polymers encapsulate NCs with their short hydrophobic chains, while the hydrophilic groups of the polymers are exposed to water, which allows the polymer-modified NCs to be soluble in the water phase. The hydrophilic groups on the polymer backbone can be simple carboxyl groups (–COOH), PEG with various functional groups, or zwitterions. To link biomolecules on the surface-modified NCs, generally appropriate cross-linking chemistry is used for bioconjugation between biomolecules and the surface-modified NCs. For instance, the cross-linking chemistry using 1-ethyl-3-[3-(dimethylamino)propyl]carbodiimide (EDC) and N-hydroxysuccinimide (NHS) can link the amine group of biomolecules to the carboxyl groups (−COOH) on surface-modified NCs.^135,136 Appropriate cross-linking chemistry allows for the addition of a vast variety of biomolecules, such as complex proteins, primary/secondary antibodies, DNA, RNA, and so on, to the hydrophilic ligands on the NC surfaces. The bioconjugated NCs may then serve as high-affinity probes to bind with analytes. For Mn-doped NCs, researchers adopt either ligand exchange or polymer encapsulation to modify such NCs for their further biological applications.

Figure 11.

(A) Illustration of ligand exchange on the surface of colloidal semiconductor NCs. (B) Depiction of a NC and its hydrophobic ligands on its surface that are conjugated with amphiphilic polymers and a PEG layer to enable binding of target ligands. Part A was reproduced with permission from ref 138. Copyright 2013 Royal Society of Chemistry. Part B was reproduced with permission from ref 142. Copyright 2011 Nature Publishing Group.

4.2. Applications of Mn-Doped Binary NCs.

Zhu et al. utilized their Mn-doped ZnSe NCs in a time-gated approach in order to determine the presence of 5-fluorouracil (5-FU), a drug used in cancer treatment.¹⁴⁴ Their NCs were synthesized via a microwave-based hydrothermal synthesis technique. These NCs possessed QYs of up to 16% and a lifetime of 400 μs. The NCs originally possessed alkyl ligands on their surfaces. These alkyl ligands were replaced with MPA through ligand exchange. The carboxyl groups present on MPA acted as receptor sites for binding of the amine group of 5-FU. The presence of 5-FU on NC surfaces further quenched luminescence emission of the NCs. A laboratory-based LS-55 fluorometer was used in order to measure the time-resolved luminescence of the samples. Excitation occurred at UV of 350 nm with a delay of 0.2 ms and a gating time of 0.4 ms. This assay allowed for a LOD of 128 nM in human serum and also exhibited great accuracy (2–4%) when tested with excess amounts of ions, glucose, or amino acids present in the serum sample. At a later stage, Zhu et al. also used the same MPA-coated NCs to detect tiopronin, a heart-disease biomarker.¹⁴⁵ These MPA-coated NCs were first mixed with methylviologen (1,1-dimethyl-4,4-bipyridinium chloride; MV²⁺), an electron-transfer agent with two quaternary ammonium groups, which allowed for electrostatic linkage of these NCs. Tiopronin could bond with MV²⁺, thereby breaking apart this aggregate of NCs, reducing the fluorescent quenching caused by MV²⁺, and allowing detection to occur, as shown in Figure 12A. This assay was measured using the same instrumentation with the same timing settings as those utilized in the detection of 5-FU, and a LOD of 0.1 μM for tiopronin in human serum was achieved. Figure 12B illustrates the time-resolved measurement principle for this bioassay. Figure 12C displays the quenching ability of different concentrations of methylviologen when conjugated with NCs. Figure 12D shows reduced quenching of the NCMV²⁺ complex caused by the presence of tiopronin during time-resolved detection, and a calibration curve for the detection of tiopronin. The delay time allows for autofluorescence to decay before Mn emission from the NC bioconjugates is measured. Zhu et al. also applied such NCs with similar luminescence quenching or turn-on approaches to detect avidin and vascular endothelial growth factor-165 (VEGF165, a cancer biomarker).^146,147 In the development of these bioassays, the same instrument was used in time-resolved measurement.

Figure 12.

(A) Illustrations of the bioassay principle for tiopronin detection using Mn-doped ZnSe NCs. (B) Illustration of the time-resolved measurement principle. (C) Luminescence spectra of the NC–MV²⁺ bioconjugates with different concentrations of MV²⁺: (a) 0 μM; (b) 2 μM; (c) 4 μM; (d) 6 μM; (e) 8 μM). The inset is the decay curve of fluorescence emission of the Mn-doped NCs. (D) Time-resolved detection of different concentrations of tiopronin via decreased luminescent quenching of the NC–MV²⁺ conjugates. The inset is the calibration curve for tiopronin. All figures were reproduced with permission from ref 145. Copyright 2013 Royal Society of Chemistry.

Chang et al. used their Mn-doped ZnS NCs for the detection of different proteins in a time-resolved mode.¹⁴⁸ This study devised four different NCs capped with different ligands in order to detect nine different target proteins. The four types of NCs are MPA-capped NCs, N-acetyl-l-cysteine-capped NCs, glutathione NCs, and α-thiogylcerol-capped NCs. These NCs all had lifetimes of about 1.4 ms and QYs that ranged from 4 to 18% depending on the capping ligands. The surface-modified NCs were applied to a solution mixing human urine with nine proteins at a concentration of 500 nM. The differential luminescence responses resulted from the combination of various interactions between proteins and the NCs, such as hydrogen bonding, electrostatic interaction, as well as the interaction/coordination of metal atoms in protein. The distinct fingerprint can be quantitatively analyzed to obtain protein identification and differentiation more clearly by using linear discriminate analysis. Because the lifetime of human urine was ~43 ns, a decay time of 0.2 ms was allowed for the efficient detection of proteins in the sample. When compared with the steady-state luminescence measurement, the time-resolved measurement outperformed their sensing counterpart. Specifically, in an analysis of the nine proteins present in human urine, six out of nine proteins could not be distinguished with above 95% confidence through steady-state luminescence measurement, while nine proteins could be detected with 100% confidence through time-resolved measurement. In this assay, a time-correlated-single-photon-counting-based FLS-920 laboratory spectrofluorometer was used for time-resolved signal measurement with UV excitation at 300 nm for each sample.

Afterward, Jing et al. applied Mn-doped ZnS for the detection of Fe(II) ions using the same instrument.¹⁴⁹ In order to detect Fe(II) ions, their MPA-capped NCs were mixed with a solution of cysteine. Cysteine bridges enhanced the luminescence of the NCs and still allowed the NCs to possess a lifetime of 505.42 μs. These cysteine-bridged NCs were then utilized to measure Fe(II) ions in an aqueous solution. Fe(II) ions significantly quenched the phosphorescence of these cysteine-bridged NCs, which allowed for detection. In the aqueous solution, a LOD of 19 nM was achieved. Then they tested Fe(II)-ion-spiked urine samples and Fe-ion-spiked human serum samples [the Fe(II) ion concentrations were at 150, 500, and 800 nM for each type of sample matrix]. In the time-resolved mode, the recovery rates for all concentrations were above 92%, but the recovery rates were lower in the range from 50 to 81% in the non-time-resolved mode (steady measurement). This study did prove the effectiveness of time-resolved measurement.

Zou et al. employed Mn-doped ZnS NCs for the detection of Cu(II).¹⁵⁰ They capped their NCs with alginate, in order to make use of the carboxyl present to coordinate with Cu ions. Their NCs had a lifetime of 2.1 ms and a QY of 4.2%. Cu ions would quench the emission of the NCs. The assay was measured with a Hitachi F-4600 fluorescence spectrophotometer with a delay time of 0.1 ms and an excitation at 316 nm. In an aqueous solution, a LOD of 6 nM was reported. They spiked Cu ions into tail water to achieve samples with different Cu ion concentrations (0, 0.1, 1, and 5 μM), and the recovery rates for these concentrations were all above 96%.

Garcia-Cortés et al. applied Mn-doped ZnS NCs in a time-resolved luminescence immunoassay.¹⁵¹ They prepared Mn-doped ZnS NCs with a QY of 3.1% and a lifetime of 1.9 ms. Their NCs were capped with DHLA. EDC/NHS reagents were then used to modify the carboxyl group on the surface of the NCs with covalent binding to the amino group of the secondary antibodies. In their sandwich-type immunoassay for prostate-specific antigen (PSA, a prostate cancer biomarker), they coated the primary antibody in microwells of a microplate to capture PSA in human serum. Secondary antibody–NC conjugates were then applied to label the captured PSA. After washing, the signal of the labeling NCs was read in the time-resolved mode. Figure 13A shows the procedure of this immunoassay. A laboratory-based spectrometer [Varian Cary Eclipse fluorescence spectrometer, with a Xe discharge lamp (75 kW), a PMT (model R-298), and a monochromator (Czerny-Turner)] was used for signal measurement. In its time-resolved signal measurement, an excitation at 300 nm, a 0.2 ms delay time, and a 5 ms gating time were used. This assay achieved a LOD at 17 pg/mL PSA in human serum. Figure 13B displays the whole calibration curve of the immunoassay in a wide PSA concentration range. Figure 13C presents the calibration curve at the low concentration end of PSA.

Figure 13.

(A) Diagram of the step-by-step processes involved in the time-resolved luminescence sandwich immunoassay for PSA, a prostate cancer biomarker. (B) Immunoassay calibration curve (signal versus PSA concentration) using Mn-doped ZnS NCs as time-resolved luminescence probes. (C) Immunoassay calibration curve at the low concentration end of PSA. All figures are reproduced with permission from ref 151. Copyright 2017 Elsevier.

Pu et al. developed Mn-doped ZnSe/ZnS NCs and modified their surfaces with MPA for time-resolved cellular imaging.¹⁰⁴ To verify that their NCs could be distinguished from autofluorescence, the researchers performed a time-resolved emission spectrum of a mixture of bovine serum albumin (BSA) with two different Mn-doped ZnSe/ZnS NCs (lifetimes at 110 and 840 μs, respectively). BSA has a broad luminescence with a peak at 475 nm and a lifetime of <1 μs. As shown in Figure 14A, as time evolved, the emission contour changed significantly; the emission peak position moved from the luminescence peak (475 nm) of BSA to the luminescence peak (~600 nm) of Mn-doped ZnSe/ZnS NCs with a clear signal strength. On the basis of this test, the researchers then performed cellular imaging of RAW264.7 mouse macrophage cells. The researchers seeded these cells into 24-well plates. They then added their NCs with different lifetimes (200, 550, and 840 μs) separately to these cells and allowed for a day-long incubation period. These NCs were also highly bright, possessing QYs > 70% for all lifetimes. After removal of the labeled cells from the culture plates, they mixed these cells and imaged them using a pulsed laser (Opolette 355 nm pulsed laser) coupled with a microscope (Olympus IX83) and an Andor iXon DU-897 EMCCD camera with a 100 ns delay time and a 20 μs gate time. A false color image representing three different NCs is shown in Figure 14B. It is clear via this study that Mn-doped ZnSe/ZnS NCs can be used in time-resolved luminescence bioimaging.

Figure 14.

(A) Time-resolved emission spectrum of the solution composed of BSA and two types of Mn-doped ZnSe/ZnS NCs (lifetimes at 110 and 840 μs, respectively). Colors from red to blue represent the logarithmic PL intensity from high to low. (B) Time-resolved bioimaging of RAW264.7 cells, a false color imaging of a mixture of three groups of cells labeled with Mn-doped ZnSe/ZnS NCs with three different luminescence lifetimes (200, 550, and 840 μs). All figures were reproduced from ref 104. Copyright 2016 American Chemical Society.

It can be seen that Mn-doped ZnS or ZnSe NCs were used in the time-resolved mode to sensitively detect analytes or image cells in biological samples. In general, however, laboratory-based instruments were adopted or developed for signal measurement, and all NCs were excited at UV wavelengths. In this review, we also note that in the literature, the biosensing/imaging applications of Mn-doped NCs utilizing luminescence-intensity measurement are more numerous than those using time-resolved luminescence measurement. This is because the instruments used in the time-resolved mode are more expensive and not widely available to many research laboratories.

4.3. Applications of Mn-Doped Multinary NCs.

Gallian et al. utilized Mn-doped AgZnInS/ZnS NCs with a lifetime of ~1.3 ms and a QY of >40% to detect Cu ions in an alcohol solution.¹⁵² They modified the NC surfaces with polymer encapsulation to achieve NC probes with carboxyl groups, which allowed for the binding of the NC probes with Cu ions. The presence of Cu ions on NC-probe surfaces would quench the luminescence of the NC probes. In the assay, the Cu-ion-spiked alcohol sample and NC probes in water were mixed and placed in a quartz cuvette unit (~4 mL). This cuvette unit was subsequently excited with a laser (L405P20, 405 nm, 20 mW, ~$40) accompanied by proper filters and lens. Once the laser was off, after a 50 μs delay to avoid autofluorescence, a PMT (H11706–20 Hamamatsu Photonics) was activated to record the time-resolved luminescence of the NC probes in a 100 μs gating window. The PMT then sent this electric signal to a microcontroller unit for signal processing.

Parts A and B of Figure 15 present the principal scheme of the instrument and the developed instrument in a small size, respectively. The total cost of such a small instrument is around $4,500, which is around one-tenth of a well-equipped laboratory microplate reader. The cost of the developed instrument can be reduced further through a cost-reduction phase in its continued development. Before performing the assay for Cu-ion detection, using 5% BSA or 10 μg/mL NC probes in 5% BSA in the cuvette unit, the researchers prolonged the PMT gate signal to observe the signal change before and after the laser was turned off. As shown in Figure 8C,D, the PMT was saturated by the autofluorescence from BSA during the laser-ON time frame, no matter whether the cuvette was loaded with 5% BSA or 10 μg/mL NC probes in 5% BSA. However, after the laser was turned off, the autofluorescence of 5% BSA would fall to zero due to the short lifetime of BSA (as proven in Figure 15C), but 10 μg/mL NC probes in 5% BSA gave a decayed but still strong luminescence signal (Figure 15D). The luminescence of 10 μg/mL NC probes could not be distinguished from the autofluorescence of BSA during the laser-ON time frame but could be clearly read in the time-resolved mode (after the laser was turned off). In the assay for Cu-ion detection (as shown in Figure 15E), a series of Cu-ion solutions with different Cu-ion concentrations were prepared and mixed with a certain amount of NC probes, and the luminescence of the NC probes was measured in a time-resolved mode. Figure 15F presents a calibration curve using the time-resolved measurement, alongside a calibration curve where the NC-probe luminescence was collected during the laser-ON time frame. When the two calibration curves are compared, it is clear that the time-resolved measurement is more sensitive in detecting lower Cu-ion concentrations. As shown in this study, the time-resolved measurement yields a LOD at ~31 nM, while the measurement during the laser pulse produces a LOD at ~466 nM. It can be seen that, under the same optic setting, the time-resolved mode did achieve a LOD of more than 10 times lower than the regular measurement in which the excitation source is on.

Figure 15.

(A) Principal scheme of time-resolved luminescence instrument. (B) Developed small instruments. (C) PMT response with a prolonged gate signal to observe the luminescence change before and after the laser-OFF time frame for 5% BSA. (D) PMT response with a prolonged gate signal to observe the luminescence change before and after the laser-OFF time frame for 10 μg/mL NC probes in 5% BSA. (E) Illustration of the assay for Cu-ion detection using NC luminescence quenching. (F) Calibration curves of the Cu-ion assay under two different luminescence measurement modes: time-resolved and regular measurement during the laser-ON time frame. All figures were reproduced with permission from ref 152. Copyright 2019 American Institute of Physics.

Following this work, Gallian et al. modified the optical path of their instrument, as shown in Figure 16A,B, and made the instrument able to read signals from the microwell of a microplate.¹⁵³ They then performed a micromagnetic bead-based immunoassay to detect capsular polysaccharide (CPS) of Burkholderia pseudomallei, which is a biomarker for pathogen infection. In the immunoassay, they used monoclonal antibody (mAb 4C4)-coated magnetic beads to capture CPS in PBS with 5% milk or human serum, and after washing, they used 4C4–NC conjugates as probes to label the captured CPS. Afterward, they used an immunoaffinity separation buffer to release the NC probes from the surfaces of the magnetic beads into the buffer for signal reading. They read the same microplate by performing immunoassay under a non-time-resolved microplate reader (PerkinElmer 2030) and also under their developed time-resolved luminescence instrument. Parts C and D of Figure 9 present the calibration curves under the two measurement modes, respectively. In a comparison of parts C and D of Figure 16, it can be seen that CPS at 0.1 and 1 ng/mL cannot be distinguished from the background using the microplate reader, but these are clearly measurable using the time-resolved luminescence instrument. In the non-time-resolved mode, the LOD for CPS in PBS with 5% milk was at ~9 ng/mL, and the LOD for CPS in human serum was at ~10 ng/mL. In the time-resolved mode, the assay presented a LOD for CPS in PBS with 5% milk at ~64 pg/mL, and a LOD for CPS in human serum at ~23 pg/mL. It is clear that the time-resolved mode produced a LOD 2 orders lower compared to the non-time-resolved (or steady) mode. On the basis of Gallian’s work, Hegseth et al. further redesigned the time-resolved instrument and packaged its electronics and optics into a portable small box.¹⁵⁴ The instrument can read microplates and was further optimized to achieve a higher sensitivity.

Figure 16.

(A) Developed compact time-resolved luminescence instrument that can read a microplate. (B) Principal illustration of the optical path in the time-resolved luminescence instrument. (C) Calibration curves of immunoassay measured using a nontime-resolved microplate reader. (D) Calibration curves of immunoassay measured using the time-resolved luminescence instrument. All figures were reproduced from ref 153 and are available under the terms of the Creative Commons Attribution License (CC BY).

These studies clearly indicated that Mn-doped multinary NCs with low-energy excitation, long lifetimes, and high brightness can simplify and facilitate the development of compact, low-cost, but highly sensitive time-resolved luminescence instruments. Such NC probes and time-resolved luminescence instruments can be further developed for broader biosensing/imaging applications including in-field or POC testing.

5. PERSPECTIVE

5.1. Understanding Photophysics.

With the current understanding of the synthesis procedures and the luminescence mechanisms of Mn-doped NCs, it can be expected that further investigations could be more focused on the physics behind the interaction between multinary host NCs and Mn dopants. Appropriate characterization methods or tools should be adopted or developed to study how the lattice fields or the lattice vibrations of host NCs interfere with the energy levels of Mn dopants and how such interactions could further affect the luminescence properties (lifetime, brightness, and emission/excitation wavelengths) of Mn-doped NCs. Temperature-dependent luminescence emission and lifetime measurement or temperature-dependent EPR could be effective approaches to probe into the interaction between host NCs and Mn dopants. Additionally, ultrafast spectroscopy, which makes use of femto- to attosecond pulses of light to study the transient absorption spectra photodynamics of various materials, has also become an emerging technique that is capable of further analyzing the host–Mn interaction.¹⁵⁵ Recent studies have reported ultrafast data about the charge-transfer dynamics of Mn-doped NCs containing Cd or Pb.^156–158 The current understanding of the physics behind the energy transfer in Mn-doped multinary NCs as discussed in this paper is very limited. However, ultrafast spectroscopy may allow researchers to investigate the electronic transitions occurring in the band gap and the subsequent Mn ion electron spin-flip transition in-depth. This sort of data may also provide clues into how to improve the PL properties of these Mn-doped multinary NCs.

Recent studies on Mn-doped NCs containing Cd or Pb have also shown that hot electrons can be generated from Mn-doped NCs using weak continuous waveform excitation equivalent to the concentrated solar radiation without requiring intense or high-energy photons. The generated hot electrons can be further utilized in photoinduced chemical processes requiring efficient electron transfer such as in photocatalysis. Specifically, Son et al. investigated the possible photophysical pathways generating hot electrons in Mn-doped NCs (specifically Mn-doped CdSe/ZnS NCs and Mn-doped CsPbBr₃ NCs), disclosed two possible pathways (consecutive exciton-to-Mn energy transfer and exciton-to-Mn energy transfer, followed by Auger back energy transfer from Mn dopant to electron), and demonstrated the possible photocatalysis applications.^159–161 As a possible future research topic, the experimental approaches from the work by Son et al. could be applied to study whether or not hot electrons can be generated in Mn-doped NCs not containing Cd or Pb elements. If yes, these NCs can be applied in photocatalysis that may be involved in biosensing/imaging procedures.

Additionally, research into the interaction between multiple types of dopants [Mn and other doping element(s)] in host NCs could also be a focus. Other doping element(s) in NCs could minimize NC defects to enhance NC brightness, provide energy levels in the NC band gap for energy transfer between the NC conduction band and Mn emission-energy levels to tune NC luminescence lifetimes, or facilitate visible-light absorption for NC excitation. Temperature-dependent analytics (the data on luminescence emission and lifetime, or EPR) and ultrafast spectroscopy can also be applied to investigate the interaction of multiple dopants. For the doping of multiple types of dopants, the selection of appropriate or novel host NCs that can provide occupation sites for both Mn and other doping elements is critical. Notably, novel host NCs are expected to be environmentally stable and contain no Cd or Pb. Such studies would facilitate us to develop higher-quality NCs for time-resolved luminescence measurement.

5.2. Engineering New Nanocomposites.

Although researchers achieved Mn-doped NCs with the desired optical properties, the shortcomings of the NCs themselves are still clear: their luminescence emission is relatively fixed in a small-wavelength range, and their lifetimes are still limited to milliseconds. It is of particular interest to increase their lifetimes to tens of milliseconds (or even longer up to seconds) or to tune their emission to NIR or longer wavelengths, while retaining high brightness. It may be difficult to directly engineer Mn-doped NCs to achieve such properties. A possible approach is to develop nanocomposites integrating multiphase materials in a nanoscale domain to gain collective or novel material characteristics from individual components. Recently, organic after-glowing materials with long luminescence lifetimes (up to seconds) are emerging for biomedical applications.^162–168 These materials usually emit the blue-to-green color range.^167,169 There may be a good opportunity to couple these organic after-glowing materials with Mn-doped NCs to achieve new nanocomposites that possess orange or red Mn emission but the longer lifetime of organic after-glowing materials. Additionally, Mn-doped NCs can be conjugated with organic NIR dyes to achieve NIR emission with a millisecond lifetime through Förster resonance energy transfer. We believe a substantial amount of additional research is needed in the fields merging luminescence organic materials with Mn-doped inorganic luminescent probes.

5.3. Broadening Biological Applications.

Researchers may find great promise in further developing portable and highly sensitive time-resolved instruments that make use of these unique NCs for broader in-field or POC testing for infectious pathogen detection, food safety analysis, environmental monitoring, etc. With the advancement of such portable time-resolved instruments, many conventional (non-time-resolved) NC-based luminescence bioassay studies can be converted to time-resolved luminescence ones for higher sensitivities or lower LODs and further facilitate their in-field or POC applications. For instance, many commercial instruments capable of HTRF may be renovated via the use of these Mn-doped NCs. With the introduction of visible-light-excitable, long-lifetime, highly bright probes, large and expensive instruments that were traditionally used for HTRF may be substantially reduced in size and price via simplifications in the optical accessories needed to use such equipment. These simplifications could potentially produce new variations of these commercial time-resolved readers that are capable of POC testing.

Further development in the field of time-resolved bioimaging with the use of Mn-doped NCs as imaging probes is needed. Although it has been generally established that time-resolved bioimaging can greatly enhance the signal-to-noise ratio in complex biological samples due to the reduction of autofluorescence, the use of Mn-doped NCs in this field has been extremely limited. This can be explained by the fact that the vast majority of commercially available time-resolved imaging instruments, such as those used in FLIM and phosphorescence lifetime imaging microscopy, are equipped with the optical equipment and electronics necessary for probes with very short lifetimes only. Due to this limitation in availability, millisecond lifetime probes usually require their own custom-made instrumentation for bioimaging. Although some researchers have already proven the feasibility of such self-made instruments for bioimaging with imaging probes with lifetimes at the millisecond levels, these reported instruments are still very bulky and expensive.^48,49,104 Small and sensitive time-resolved imaging instruments using Mn-doped NCs are still in demand.¹⁷⁰ In the development of such imaging instruments, it should be noted that Mn-doped binary NCs (like Mn-ZnSe/ZnS NCs) usually present single-exponential decays, while Mn-doped multinary NCs generally possess biexponential decays. Such decay characteristics should be considered if a series of images at different delayed times are acquired in imaging, and then the lifetime of each image pixel is estimated by fitting the measured luminescence decay data of each pixel to a certain exponential decay function.

Researchers may also find great promise in utilizing these Mn-doped NCs for time-resolved flow cytometry, which has become a highly sensitive detection method for the multiplexed analysis of biological media, specific metabolic pathways, molecular kinetics, and efficacy of different pharmacological treatments.¹⁷¹ In time-resolved flow cytometry, cells flow down a channel to a hydrodynamically focused region where they are concentrated into a single-file line for subsequent excitation with a pulsed-laser light source and detection via a photodetector. Conventionally, time-resolved flow cytometry can distinguish probes with differing lifetimes for multiplexing detection (lifetimes at around tens of nanoseconds matching with high throughput or high flow velocity) and also avoid the autofluorescence in biological samples to achieve high sensitivity. However, conventional time-resolved flow cytometry requires expensive and laboratory-based instruments. Recently, time-resolved flow cytometry on a microfluidics chip has been reported.¹⁷² The fluid flow velocity in microfluidics is more than 1000 times slower than that in conventional time-resolved flow cytometry. Such a reduced throughput will allow the use of cellular probes with a lifetime at milliseconds for bioanalysis. Highly bright Mn-doped NCs could be good candidates for time-resolved microfluidics flow cytometry. The low throughput and the optical properties of Mn-doped NCs may synergistically facilitate the miniaturization of time-resolved microfluidics flow cytometers (a microfluidics chip plus a compact, low-cost, but sensitive optical reader). Such miniaturized cytometers could be used in resource-limited environments for POC applications.

Additionally, in this review, we observed that the luminescence lifetime of Mn-doped NCs can be tuned in a wide range (e.g., from hundreds of microseconds to a few milliseconds). Such lifetime-tuning capabilities will render these NCs suitable for multiplexing biosensing in time-resolved microfluidics flow cytometry. If these NCs with different lifetimes can be engineered to different emission wavelengths through coupling with organic dyes, multiplexing biosensing/imaging can be achieved in both time and spectral domains with appropriate time-resolved luminescence measurement instruments.

6. CONCLUSION

Overall, in this review, we began with outlining the current luminescent probes to understand their limitations for in-field or POC testing, while also recognizing the potential of Mn-doped NCs in overcoming such limitations. On the basis of this recognition, we then reviewed the past and current work on Mn-doped binary and multinary NCs (not containing Cd or Pb) with increased attention on their synthesis and luminescent mechanisms. We showed how researchers achieved the desired optical features with an improved understanding of Mn emission. Afterward, we summarized representative applications of these NCs in time-resolved luminescence biosensing/imaging. In the review of these applications, we brought attention to what analytes these NCs were utilized to detect, how their surfaces were modified for this detection, what instruments were used in time-resolved luminescence measurement, and what sensitivities or LODs were achieved in the detection. Through reviewing these applications, we attempted to present the potential of Mn-doped NCs in advancing time-resolved luminescence biosensing/imaging toward in-field or POC testing. Additionally, we discussed the perspectives of future research related to Mn-doped NCs. We expect that this review will help garner more attention on highly bright Mn-doped NCs with long lifetimes and low-energy excitation for time-resolved luminescence biosensing/imaging.