Influence of Structural Properties of Oleic Acid-Capped CdSe/ZnS Quantum Dots in the Detection of Hg²⁺ Ions

Abstract

Oleic acid-capped CdSe/ZnS quantum dots (QDs) were used to investigate their photoluminescence (PL) response to Hg²⁺ ions as a function of the surface properties of QDs. Three distinctly-size CdSe/ZnS QDs were obtained by varying the molar ratio of shell precursors, which were characterized by X-ray diffraction (XRD), high-resolution transmission electron microscopy (HR-TEM), Fourier-Transform infrared (FT-IR), X-ray photoelectron spectroscopy (XPS), absorption spectroscopy, and time-resolved fluorescence spectroscopy. Results revealed the obtention of zinc blende nanocrystals with sizes ranging from 2.7 to 3.2 nm (± 0.5) and ZnS thickness between 0.3 and 1.0 monolayer (ML). The variation of the [S]/[Zn] molar ratio introduced chemical species that act as traps, affecting the PL properties differently. Depending on the thickness of the shell and chemical speciation on surface, Hg²⁺ ions could induce quenching or enhancement of PL. Detection of mercury ions was evaluated in terms of Stern-Volmer equation, where the limit of detection (LOD) for the PL quenching system was 11.2 nM, while for the PL enhancing systems were 8.98 nM and 10.7 nM. Results demonstrate the performance of oleic acid-capped CdSe/ZnS QDs to detect Hg²⁺ and their capacity to turn the PL on/off depending on surface properties.

Supplementary Information

The online version contains supplementary material available at 10.1007/s10895-024-03828-0.

Introduction

Quantum dots (QDs) of semiconductors are nanocrystals with particle sizes between 1 and 10 nm, which have found applications in different fields, such as sensors, lasers, photovoltaics, and light-emitting diodes [1]. II-VI semiconductors like CdSe are mainly characterized by a broad absorption spectrum and a narrow and intense emission that could be multicolor since it depends on their size due to the quantum confinement effect [2]. These characteristics are very appreciated for metal ion sensing applications [3, 4].

Indeed, for QD-based metal ion detection, CdSe QDs have been modified with different ligands to avoid aggregation, enhance stability, and improve their selectivity and sensitivity [5]. With this goal, the CdSe QDs surface has been modified with hydrophilic ligands, resulting in a fluorescence signal intensity that could be quenched or enhanced as a response to metal ions [5, 6]. For instance, quenching of fluorescence has been reported with L-cysteine-capped CdSe QDs to detect Pb²⁺ions [7], mercapto acetic acid-capped CdSe QDs have been applied in the detection of Ag⁺ ions [8], and mercaptoethanesulfonate-capped CdSe QDs for Cu²⁺ determination [9]. Equally, enhancement of PL has been reported with L-cysteine-capped CdSe to detect Zn²⁺ ions [9] and L-carnitine-capped CdSe/ZnS QDs to detect Hg²⁺ [10]. Other QD types have also been reported considering the enhancement of PL intensity induced by cations. For example, mercaptopropionic acid-capped CdTe QDs in the presence of Cd²⁺ [11], mercaptoethanol-capped PbS QDs in the presence of mercapto acetic acid-capped InP QDs in the presence of Hg²⁺ ions [12] and L-cysteine capped CdS QDs in the presence of Ag⁺ ions [13].

Mercury ions, the most potent neurotoxin in human physiology, are selected as analyte in this investigation due to the severe environmental and public health problems in counties that use mercury in gold extraction mining. Many fluorescent-based optical sensors [14] have been applied to detect Hg²⁺ ions in real time with high sensitivity, selectivity, and signal-to-noise ratio [15].

For QDs-based mercury ion sensors, the interaction between analyte and nanocrystals leads to many pathways that affect photoluminescence intensity. According to the affinity of metal ions and QDs, cation exchange could be produced by replacing cations of starting QDs with cations of the analyte, preserving the original anion sublattice [16]. This cation exchange reaction forms small particles on the QDs surface, causing non-radiative recombination of excited electrons and holes that finally quenches fluorescence.

Conversely, the strategies reported by literature that led to increased fluorescence for determining cation ions are less frequent. Surface state passivation of QDs removes the traps that facilitate non-radiative recombination pathways [17]. Understanding the role of trap states is challenging. In its simplest form, a trap can be described as a stable nonbonding orbital of an undercoordinated atom. This orbital usually lies deep in the bandgap, where it can act as an electron or hole trap. Metal ions on QDs surface can produce non-coordinative atoms and dimers [18], indirectly affecting the number of dangling bonds and the presence of traps [19].

In this context, the growth of a shell of semiconducting material with a larger bandgap (3.7 > 1.7 eV) [2] than CdSe, such as ZnS, plays a determinant role since the shell passivates the traps and defects of core [20, 21], improving stability [1, 20], and photoluminescence [2, 22, 23]. This type of QDs is known as a type I core/shell system, characterized by the confinement of the exciton (an electron-hole pair) to the core [24]. Type I structures are the most studied in analytical applications since this configuration provides the best confinement of the exciton (an electron-hole pair) and the highest rates of radiative electron-hole recombination (i.e., the brightness of the photoluminescence) [25]. However, excess shell (greater than 2.2 monolayers [26]) can cause the opposite effect in the PL of core/shell QDs, because the lattice mismatch between CdSe and ZnS is 12%, which generates a strain at interface [27] that creates non-radiative recombination sites, impairing the properties of PL [21, 28, 29].

Our group has previously studied Hg²⁺ detection by PL measurements using CdSe/ZnS QDs modified with two ligands of differing water affinity, such as L-glutathione and oleic acid [30]. Both QDs produced PL quenching. Wang et at. [31] reported the fluorescence enhancement induced by Hg²⁺ using CdSe/ZnS QDs. The enhanced PL was attributed to Zn²⁺/Hg²⁺ cation exchange in the ZnS shell, producing Hg_xZn_1−xS/CdSe, increasing the separation of electrons and holes, and reducing the recombination rate.

Since oleic acid-capped CdSe/ZnS could detect metal ions, we used these nanomaterials to induce PL quenching and enhancement. For this, the present investigation used the variation of the molar ratio of shell precursors to obtain three nanocrystals with different shell thicknesses, forming traps that affected PL properties in different ways. We discuss the fluorescence response of QDs regarding the shell thickness, [S]/[Zn] molar relation, surface properties, and PL decay dynamics. Surface properties are characterized by FT-IR, XPS, TEM, and HR-TEM measurements. Results show that an adequate molar ratio of shell precursor could passivate the trap sites of surface QDs. In addition, a thicker ZnS shell (0.7-1.0 ML ranging) protects the core from exchange cation reactions, avoiding the quenching of PL in the presence of Hg²⁺ ions. Thus, we demonstrate the shell dependence to turn PL on/off induced by mercury ions and other metal cations.

Experimental Section

Materials

Cadmium oxide powder (99.9%), trioctylphosphine (TOP, 99%), selenium powder (99.9%), hexamethyldisilathiane ((TMS)₂S, 99%), diethylzinc (Et₂Zn, 1.0 M/hexane), oleic acid (OA), and 1-octadecene (ODE, tech. 99%) were purchased from Sigma-Aldrich. Metal chloride (HgCl₂, MnCl₂, CoCl₂, NiCl₂, ZnCl₂, CdCl₂, PbCl₂), ethanol (99.5%), acetone (99.8%), methanol (99.5%), and acetonitrile (CH₃CN, 99%) were obtained from Panreac. All reagents were analytical grade and used without further purification. Ultrapure water (18 MΩ) was used to prepare aqueous solutions.

Synthesis of Oleic Acid-Capped CdSe/ZnS QDs

The CdSe/ZnS QDs were prepared using a method previously [30], and the details of the synthesis process are provided in the supporting information. Three CdSe/ZnS samples with varied sizes were obtained by changing the amount of Et₂Zn to 31 µL (0.031 mmol), 140 µL (0.14 mmol), and 280 µL (0.28 mmol), while the amount of (TMS)₂S was kept constant (0.28 mmol).

Characterization

HR-TEM Analysis

The High-Resolution Transmission Electron Microscopy (HR-TEM) images of the QDs were taken using a Tecnai F20 Super Twin TMP Microscope, Field emission source, resolution of 0.1 nm at 200 kV, maximum magnification at TEM 1.0 MX, GATAN US 1000XP-P camera. EDX Oxford Instruments XMAX detector. STEM Analysis - FISCHIONE Instruments Model M3000 FP5360/22 HAADF Detector 120/200 kV. The samples were dispersed in chloroform using an ultrasonic bath. After, a drop was deposited on a carbon film grid; then, the sample was dried and analyzed in the transmission microscope using 450.000 x magnification. Image-J software was used to calculate the average diameter and d-spacing [111] of QDs with HR-TEM micrographs.

The shell thickness parameter is the difference between the size of QDs core-shell (CdSe/ZnS) and size core (CdSe) from HRTEM analysis (Eq. 1).

1

The monolayer number (ML) was calculated as the ratio value between the experimental shell thickness parameter and ZnS zinc blende theoric monolayer size (0.70 nm) (Eq. 2) [32].

2

XRD Patterns

Panalytical X’Pert3 Pro Multipurpose Diffractometer with CuK_α radiation source (45 kV and 40 mA) was employed to measure the powder XRD patterns. The diffraction dataset cards from the Joint Committee of Powders Diffraction Standards (JCPDS) were used to compare the obtained patterns.

FT-IR Analysis

Infrared spectra of quantum dots were recorded with a Thermo Nicolet Nexus 670 FTIR spectrometer at a resolution of 4 cm^–1 using a KBr pellet, summing 50 scans.

X-ray Photoelectron Spectroscopy (XPS)

The XPS experiments were recorded using the XPS/ISS/UPS- ACenteno surface characterization platform built by SPECS (Germany). The platform is equipped with a PHOIBOS 150 2D-DLD energy analyzer. A monochromatized Al Kα X-ray source (FOCUS 500) operated at 100 W was used for the measurements. The pass energy of the hemispherical analyzer was set at 100 eV for general spectra and 20 eV for high-resolution spectra. The samples were mounted on copper conductive tape in stainless steel metal sample holders for analysis and provided by the manufacturer SPECS. These sample holders are electrically connected to the spectrometer. CasaXPS program (Casa Software Ltd) for data analysis and the SPECS Prodigy library for RSF values were used.

UV-Vis Spectroscopy

Electronic absorption spectra were measured on a UV-Vis spectrophotometer EMC-11-UV. Samples were measured in dilute solution in 1 cm pathlength quartz cells.

Photoluminescence (PL) Measurements

Steady-state PL measurements were performed in an Agilent Varian Cary Eclipse Fluorescence spectrophotometer, using optically dilute solutions in 1 cm x 1 cm quartz cells. The PL quantum yield (PLQY, Φ) of CdSe/ZnS QDs (dissolved in chloroform) was calculated using rhodamine 6G in methanol solution as standard (Φ_s = 94%), with excitation wavelength (λ_exc) at 500 nm. Equation 3 was employed to calculate PLQY [33].

3

OD is the optical density (absorption) at the excitation wavelength, is the area under the emission band, is the refractive index of the solvent, s is the standard solution, and are values of the unknown solution.

Time-Resolved Photoluminescence (TRPL)

TRPL measurements were carried out using a Horiba Fluorolog Time-Correlated Single Photon Counter (TCSPC) system, using a 560 nm nano LED as the pulsed excitation source (1 MHz; FWHM = ca. 1 ns). Measurements were performed at room temperature.

Changes Optical Induced by Hg²⁺

Hg²⁺ detection was carried out as follows: 100 µL of Hg²⁺ ions in an aqueous solution (0.3 mM − 5.0 mM) were added to 3000 µL of a solution containing CdSe/ZnS QDs (240 ppm) dissolved in chloroform and ethanol (1:1). The mixtures containing Hg²⁺ and QDs systems were stirred at room temperature, and then PL spectra were measured using λ_exc = 380 nm, slit: 3 (excitation) and 2 (emission), integration time: 0.3 s.

Effect of Other Metal Ions

The changes in the fluorescence intensity originated from several metal ions (Mn²⁺, Co²⁺, Pb²⁺, Ni²⁺, Ba²⁺, Cd²⁺, and Zn²⁺) and were determined with experimental conditions similar to the detection of Hg²⁺. In a typical experiment, 100 µL of metal salt in an aqueous solution (5.0 µM) were added to 3000 µL of CdSe/ZnS QDs dissolved in a CHCl₃/EtOH 1:1 (240 ppm). The mixtures containing cations and QDs systems were stirred at room temperature, and then PL spectra were measured using λ_exc = 380 nm, slit: 3 (excitation) and 2 (emission), integration time: 0.3 s.

Results and Discussion

Structural Characterization

TEM and HR-TEM Analysis

CdSe QDs were synthesized by injecting Cd and Se precursors into a hot 1-octadecene and oleic acid solution. The growth of the ZnS shell around CdSe core was performed by a reaction of Et₂Zn and (TMS)₂S at 80 °C. The molar ratio of [S]/[Zn] precursors was 9.0, 2.0, and 1.0 to obtain particles with variations in the size and thickness of the shells (Fig. 1; Table 1). From TEM and HR-TEM images, the average diameters of QDs are 2.7 nm, 3.0 nm, and 3.2 nm (± 0.5 nm), corresponding to an amount of ZnS monolayer (ML) of 0.3 (QD-0.3 ML), 0.7 (QD-0.7 ML), and 1.0 nm (QD-1 ML), respectively. Figure S2 shows the size distribution obtained from TEM micrographs. ZnS monolayer around CdSe was calculated from the difference between the core diameter (2.5 nm) from HR-TEM images shown in Figure S1 and the theoretical value of one monolayer (Eqs. 1 and 2) [32]. TEM images also revealed the obtention of spherical structures with good crystallinity and reasonable, narrow size distributions. There is no evidence of an interface between the core and shell, meaning that the shell growth process occurs in a coherent epitaxial region, as described in previous reports [21].

Fig. 1.

TEM and HR-TEM (inset) micrographs of CdSe/ZnS QDs samples: a) QD-0.3 ML, b) QD-0.7 ML, c) QD-1.0 ML

Table 1.

Size and MLs of core and core/shell QDs

QDs	[S]/[Zn] molar ratio	Average diameter [nm]	ZnS shell thickness [nm]	ZnS MLs
CdSe	0	2.5 nm	0	0
QD-0.3 ML	9/1	2.7 ± 0.5	0.2	0.3
QD-0.7ML	2/1	3.0 ± 0.6	0.5	0.7
QD-1ML	1/1	3.2 ± 0.5	0.7	1.0

Crystal phases of QDs were determined with the interplanar distance spacing of selected areas in the HRTEM images, applying the Fast Fourier Transform (FFT) of ImageJ software (in Fig. 1, micrograph insets). The interplanar spacing is 0.31 nm for QD-0.3 ML, 0.31 nm for QD-0.7 ML, and 0.30 nm for QD-1 ML, which match the lattice spacing of (111) plane of cubic ZnS (d₁₁₁ = 0.31 nm for zinc blende ZnS) [32]. By comparing it with CdSe (Figure S1), the spacing value of the (111) plane is 0.34 nm, which matches the bulk cubic zinc blende of CdSe. These results are evidence of the growth of the ZnS shell on the CdSe core.

XRD Patterns

Figure 2 compares the X-ray powder diffraction (XRD) patterns of CdSe/ZnS samples with the CdSe pattern. For CdSe/ZnS samples, the diffraction peaks are centered around 25.5, 42.7, and 50.2 degrees, corresponding respectively to the (111), (220), and (311) lattice planes of cubic zinc blende CdSe (green peaks) [34]. Compared to CdSe, diffraction peaks of core/shell samples show more intense peaks and the characteristic shift to higher reflection angle positions due to the formation of ZnS shell [35, 36].

Fig. 2.

XRD patterns of CdSe (core) and CdSe/ZnS QDs with different shell thicknesses. The corresponding diffraction references of bulk CdSe (standard PDF card No. 03-065-2891) and bulk ZnS (standard PDF card No. 00-005-0566) are below and at the top, respectively

FT-IR Spectra

Figure 3 shows FTIR spectra in the region 2000 –1000 cm^− 1 of CdSe core and CdSe/ZnS QDs, which are compared with the oleic acid spectrum. The bands at 1460 cm^− 1, 1412 cm^− 1, and 1282 cm^− 1, respectively, due to O-H, CH_3, and C-O groups of the pure oleic acid [37] are observed in all samples of QDs. However, for CdSe and CdSe/ZnS samples, the C-O band is shifted to higher wavenumbers, and the intense band at 1707 cm^− 1, characteristic of C = O group of oleic acid, decreased significantly. FT-IR spectra of QDs samples show two bands around 1534 cm^− 1 and 1407 cm^− 1, which are respectively attributed to the asymmetric and symmetric stretching of the carboxylate anion (COO⁻), indicating a complexation between -COO⁻ group of oleic acid and a cation (Cd²⁺, Zn²⁺) of QDs surface. Separating asymmetric and symmetric carboxylate bands (Δν = ν_asymCOO⁻ – ν_symCOO⁻) reveals the coordination mode between QDs surface and ligand. In the present case, the separation range is 127 cm^− 1 and 136 cm^− 1, suggesting a bidentate coordination mode [37, 38].

Fig. 3.

FT-IR spectra of oleic acid pure and the synthesized CdSe core and oleic acid-capped CdSe/ZnS core/shell with different shell thicknesses

FT-IR spectra show a shift of the asymmetric band to higher wavenumbers as the shell thickness increases. This shift allows us to determine the cation affinity on QDs surface and -COO⁻ group. Since the amount of ZnS shell is different, the sites on QDs surface associated with the cation (Cd²⁺, Zn²⁺) must vary. CdSe is a reference sample with Cd²⁺ ions bound to oleic acid. QD-0.3 ML sample (with 0.3 monolayers) has Cd²⁺ and Zn²⁺ exposed on the QD surface, while QD-1 ML sample should have more Zn²⁺ ions exposed on surface to bind with oleic acid. According to the hard-soft acid-base (HSAB) theory [39], strong bonds are formed by electrostatic interaction between hard Lewis acid-base pairs and covalent interaction between soft pairs.

In contrast, a weak association is observed between members of opposite groups. In the present case, oleic acid, with oxygen-containing headgroups, is a hard base with more affinity to Zn²⁺ because it has a greater hardness than Cd²⁺ [16, 39]. Therefore, the stronger affinity of -COO⁻ group and Zn²⁺ of QDs surface produced a shift to higher wavenumbers of the asymmetric band. The observed shift indicates that QD-0.7 ML and QD-1 ML samples have more species of Zn²⁺ exposed on surface compared with QD-0.3 ML sample, which has more Cd²⁺ species (with a lower affinity to -COO⁻ group of ligand).

XPS Analysis

Figure S3 shows the typical XPS survey spectra of core and core/shell samples. Both spectra show the presence of Cd, Se, C, and O. CdSe/ZnS samples additionally show the presence of Zn and S.

Figure 4 shows the high-resolution XPS spectrum of QDs samples. We compare the C1S and O1S spectra from oleic acid bound to CdSe and CdSe/ZnS surfaces. The C1s region is shown in Fig. 4A. The single peak of C1S of the CdSe core is deconvoluted into two components at 284.5 eV, associated with C-C and C-H bonds, and at 288.3 eV attributed to the bidentate carboxylate (–COO⁻) [40]. For core/shell samples, C1s peak is broader than the core’s spectrum, and the carboxylate peak around 288 eV is significantly reduced. QDs sample spectra are deconvoluted into two components: at 284.3 eV, attributed to C-C (or C-H) of oleic acid, and near 285.0 eV associated with the carbon adjacent to the carboxylic group, C-COO [41]. The O1s peak is shown in Fig. 4B, and a few changes between core and core/shell samples could be appreciated. Core/shell samples show a broader noisy peak, which is deconvoluted into two components at 530.7 eV, attributed to C–O and C = O of monodentate carboxylate, and at 532 eV assigned to bidentate carboxylate. Results demonstrate the chemical bonding between oleic acid and QDs surface.

Fig. 4.

XPS spectra of CdSe and CdSe/ZnS samples: A) C1s, B) O1s

Figure 5 shows Cd3d, Se3d, Zn2p, and S2p spectra of core and core/shell samples. Figure 5A shows the Cd3d region. For CdSe core, the characteristic doublet of Cd²⁺ is seen at 405.1 eV and 411.9 eV, which is attributed to Cd3d_5/2 and Cd3d_3/2, respectively. For the QD-0.3 ML sample, the Cd3d spectrum remained almost unchanged (compared with core). For samples with a thicker shell (QD-0.7 ML and QD-1 ML), Cd3d doublet is moved to slightly higher binding energies. Additionally, the doublet is deconvoluted into two components, indicating different environments of Cd²⁺ on surface, e.g., Cd²⁺ binding to Se²⁻ (or COO⁻) and underpassivated Cd²⁺ [42].

Fig. 5.

XPS spectra of CdSe and CdSe/ZnS samples: A) Cd3d, B) Se3d, C) Zn2p, and D) S2p

The Se3d peak for the core sample is observed at 54.1 eV (Fig. 5B), a characteristic of metal selenides [43]. The Se3d spectra for core/shell samples are very similar to those obtained for CdSe core; however, the signal is less intense and noisier due to the shell on core.

The Zn2p region of core/shell samples is shown in Fig. 5C. QD-0.3 ML sample shows the Zn2p peak resolved into two splitting peaks at 1045.7 eV and 1022.6 eV, corresponding to Zn2p_1/2 and Zn2p_3/2, respectively. The separation between them is 23.1 eV, characteristic of Zn²⁺ species [44]. QDs with a thicker shell (QD-0.7 ML and QD-1 ML) have similar profiles. The Zn2p doublet is shifted to higher binding energies. The peaks are asymmetrical and broader compared with QD-0.3 ML sample. These changes reveal the presence of various Zn²⁺ species, such as Zn²⁺ bound to S²⁻ and oleic acid and the under-passivated Zn²⁺. These Zn²⁺ species could be more evident in the samples with a thicker shell since the molar ratios of Zn and S precursors were more equivalent than QD-0.3 ML.

The S2p region is shown in Fig. 5D. For QD-0.3 ML, the broad peak centered at 161.0 eV could be due to different sulfur species (S²⁻ bonded to Zn²⁺ or underpassivated S²⁻) generated from S²⁻-rich solutions. In samples with the thicker shell, S2p peak is sharper and shifted to higher binding energies, indicating that when ZnS shell grows up from Zn and S concentrations with a molar equivalence, the under-passivated S²⁻ species are diminished.

Considering the FT-IR and XPS results, the growth of a ZnS shell by varying [S]/[Zn] molar ratio in the precursor solutions generated underpassivated ionic species (Cd²⁺, Zn²⁺, and S²⁻), which could be evidenced by the broader and shifted peaks in the XPS spectra, particularly in the Cd3d, Zn2p, S2p regions. QD-0.3ML has more underpassivated S²⁻ species derived from S²⁻-rich solutions. Samples with a thicker shell ([S]/[Zn] molar ratio was 2/1 and 1/1) are characterized by having more underpassivated Zn²⁺ species. As was observed in the FT-IR spectra, for instance, QD-1 ML with more Zn²⁺ species in the QDs surface produced a shift that was more significant in the band related to bond -COO⁻-Zn²⁺.

Optical Characterization

Figure 6; Table 2 compare the main characteristics of the absorption and emission spectra of CdSe core and CdSe/ZnS samples. CdSe/ZnS samples show the typical red-shift in the first exciton absorption peak and the emission maximum regarding CdSe. The PLQY enhanced from 0.17 for CdSe to 0.54 for core/shell samples. This enhancement of PLQY is a typical result of the growth of ZnS shell on CdSe core [45]. Literature has demonstrated that ZnS shell passivates defects of core surface, increasing PLQY [21, 46]. However, PLQY is affected when many ZnS monolayers grow on the core [46]. In this work, a relationship between QY and the thickness of the ZnS shell is not observed since QD-0.7 ML has a higher PLQY (0.54), and the sample with one formed monolayer of ZnS shows a reduction of PLQY (0.47). We believe that the variation of the molar ratio of precursors could introduce trap states that affected the PLQY.

Fig. 6.

Absorption and PL spectra of CdSe and CdSe/ZnS nanocrystals at room temperature, λ_exc: 380 nm

Table 2.

Optical properties of CdSe and CdSe/ZnS

QDs	First absorption peak [nm]	Emission maximum [nm]	Stokes shift [eV]	Emission bandwidth (FWHM) [nm]	PL QY
CdSe	518	533	0.050	35 ± 0.1	0.17
QD-0.3 ML	563	576	0.040	31 ± 0.2	0.46
QD-0.7ML	564	578	0.047	34 ± 0.1	0.54
QD-1 ML	581	606	0.070	37 ± 0.1	0.47

Photoluminescence lifetime measurements were performed to understand the role of traps associated with [S]/[Zn] molar ratio and the exciton dynamics (Fig. 7). Nonexponential kinetics showed the PL decay curves of core and core/shell QDs. Still, they could be mathematically fit by multiple exponentials (Eq. 4), with relative amplitudes A_i and time constants τ_i. The average fluorescence lifetime (τ_FL) is calculated with Eq. 5.

4

5

Fig. 7.

Photoluminescence decay curves of CdSe/ZnS core/shell QDs with different shell thickness dissolved in chloroform, λ_em: 590 nm

Table 3 shows the average fluorescence lifetimes (τ_FL). A dependence of τ_FL values with shell thickness is observed. The ZnS shell passivates the non-radiative trap states on the CdSe core, making the core more photostable and less affected by non-radiative recombination pathways [22, 46–48]. No direct correlation was found between the average fluorescence lifetime and QY values, due probably to the trapping sites introduced in the nanocrystal structure during the synthesis process. We recall that QD-0.3 ML sample with the thinner shell was prepared from a solution rich in S²⁻, which generated underpassivated S²⁻ species on QDs surface. The anions (underpassivated S²⁻ species) are known for their role as hole traps [49, 50], affecting PLQY for QD-0.3 ML sample. By contrast, the sample with one monolayer of ZnS (QD-1 ML), obtained from an equal ratio of [S]/[Zn] precursors, is characterized by having underpassivated S²⁻ and Zn²⁺ species. According to previous investigations, PLQY increases up to a specific shell thickness (1.25 ML − 1.3 ML of ZnS [21]). After any further increase in the shell thickness, PL decreases due to the introduction of non-radiative recombination sites [46, 51, 52]. In the present investigation, the reduction of PLQY in QD-1 ML could be due to underpassivated Zn²⁺ and S²⁻ species acting as electron and hole traps [53], respectively decreasing PL.

Table 3.

Luminescence lifetimes measured by time-correlated single-photon counting (TCSPC) of CdSe/ZnS QDs

QDs	τFL(ns)	χ2
QD-0.3 ML	3.29	1.49
QD-0.7 ML	4.34	1.14
QD-1 ML	4.82	1.05

Effect of Hg²⁺ Ions on Luminescence Properties

The changes in the PL spectra of CdSe/ZnS QDs samples induced by Hg²⁺ ions were studied in a homogeneous mixture containing QDs dissolved in chloroform and Hg²⁺ ions dissolved in water. Results show that the PL intensity could be quenched or enhanced by Hg²⁺ ions. The two opposite effects are shown in Fig. 8 and depend on CdSe/ZnS surface properties (thickness shell and surface traps). First, we presented the progressive fluorescence quenching by increasing the Hg²⁺ concentration for QD-0.3 ML sample (Fig. 8A), accompanied by a shift of 3 nm of maximum emission wavelength. FL quenching by Hg²⁺ could be explained by the metal cation exchange reaction that produces the displacement of Cd²⁺ or Zn²⁺ in the lattice by Hg²⁺ ions generating HgSe (or HgS) particles on surface. This cation exchange is due to the lower solubility constant (K_SP) of HgSe or HgS compared with K_SP CdSe or ZnS being favored thermodynamically, as discussed previously in several reports [54].

Fig. 8.

Effect of Hg²⁺ on PL spectra of oleic acid-capped CdSe/ZnS QDs. (A) PL quenching using QD-0.3 ML sample, (B) PL enhancing using QD-0.7 ML sample, and (C) PL enhancing using QD-1 ML sample. At the bottom the Stern-Volmer relationship between fluorescence intensity of CdSe/ZnS QDs and Hg²⁺ concentration

The PL quenching by increasing the Hg²⁺ concentration in the range of 0.3-3.0 µM is evaluated in terms of Stern-Volmer equation (Eq. 6):

6

I_o and I are the fluorescence intensities at a constant wavelength in a mercury ion-free solution and a given mercury ion concentration, respectively. [Hg²⁺] is the concentration of Hg²⁺, and K_sv is the Stern-Volmer FL quenching constant. A good lineality is obtained with a correlation coefficient (R²) equal to 0.9976 (bottom of Fig. 8A). The limit of detection (LOD), calculated with equation LOD = 3σ/k, where σ is the standard deviation of the y-intercept of the regression line, and k is the slope of the calibration graph, is found to be 0.0112 µM (11.2 nM). This value is lower than previously reported using oleic acid-capped CdSe/ZnS in a heterogeneous mixture that produces PL quenching induced by Hg²⁺ [30].

Figure 8B and C show the enhancement of PL intensity of the QD-0.7 ML and QD-1 ML samples in the presence of Hg²⁺ in the concentration range from 0.3 µM to 5.0 µM. The enhancement of PL (I/I_o) of these two QDs is evaluated by the equation: I/I_o = 1+K_sv[Hg²⁺], where K_sv is the Stern-Volmer constant for the enhanced PL. A good linearity is obtained for both samples (see bottom of Fig. 7B and C): R² is 0.9982, and LOD is 0.00898 µM (8.98 nM) for QD-0.7 ML sample. R² is 0.9975, and LOD is 0.0107 µM (10.7 nM) for QD-1 ML sample. The enhancement of PL by CdSe/ZnS with oleic acid as a capping agent is observed for the first time. The enhanced PL of QDs induced by Hg²⁺ [31] and other metal ions [55] has been reported in other investigations. Results reveal the shell’s role in detecting Hg²⁺. A thicker shell (than 0.3 ML) protects the core from the exchange reaction with Hg²⁺, providing more stability and brightness. Furthermore, thicker shell samples (QD-0.7 ML and QD-1ML) have more underpassivated Zn²⁺ species on the surface, which are more exposed for the exchange with Hg²⁺ than Cd²⁺ species from core. These results show a significant improvement in the emission properties of oleic acid-capped CdSe/ZnS and an improved capacity to detect Hg²⁺ with lower LOD values.

The influence of Hg²⁺ on the PL lifetimes was investigated. Figure 9 shows the PL decay curves for CdSe/ZnS samples mixed with an Hg²⁺ aqueous solution. Table 4 gives the average fluorescence lifetimes, showing that the decays are very sensitive to the surface states of QDs. When Hg²⁺ is present, longer lifetimes occur for samples with a thicker shell (0.7 and 1.0 ML). The increased PL of CdSe/ZnS QDs by mercury ions could be attributed to the Zn²⁺-to-Hg²⁺ cation exchange in the ZnS shell, which favors the separation of electrons and holes, reducing the recombination rate [31].

Fig. 9.

Photoluminiscence decay curves of CdSe/ZnS QDs with different shell thicknesses before and after adding Hg²⁺ ions (5 μM). λem: 590 nm. Solvent: chloroform/ethanol (1:1)

Table 4.

Luminescence lifetime measured by TCSPC of CdSe/ZnS QDs in presence of Hg²⁺ solution

QDs + [Hg2+]	τFL(ns)	χ2
QD-0.30 ML	2.81	1.10
QD-0.7 ML	4.43	1.12
QD-1.0 ML	5.00	1.06

QD-0.3 ML mixed with Hg²⁺ produces shorter average fluorescence lifetimes, corroborating the effect of these cations to PL detriment. Similar sensing systems have also reported shorter FL lifetimes, accompanied by FL quenching in presence of metal ions [56, 57]. As we explained, FL quenching by Hg²⁺ is derived by the metal cation exchange reaction with Cd²⁺ (from core) or Zn²⁺ (from shell), generating HgSe and HgS particles that quench FL. In addition, the underpassivated S²⁻ species in the QD-0.3 ML sample bind to Hg²⁺, producing HgS and HgSe.

Effect of Other Metal Ions

The effect of other metal transition cations on the fluorescence of CdSe/ZnS quantum dots (QD-0.3 ML, QD-0.7 ML, and QD-1 ML) was performed in homogeneous mixtures. A solution of QDs in chloroform/ethanol (1/1 v/v) was mixed with aqueous solutions of metal chloride salts (Co²⁺, Mn²⁺, Ni²⁺, Zn²⁺, Pb²⁺, Cd²⁺ ) (Fig. 10). For QD-0.3 ML sample, the reduction of fluorescence intensity was observed with all the evaluated cations; however, Hg²⁺ reduced 100% fluorescence quenching. This result could be explained by the cation exchange reaction between Hg²⁺ ions and Se²⁻ and S²⁻ anions. Based on the K_sp values of HgSe (4.0 × 10^− 59) and HgS (1.6 × 10^− 52), which are lower than other ions [16, 39], a better ability of Hg²⁺ to PL quench is expected.

Fig. 10.

Relative photoluminescence response of CdSe/ZnS QDs with different shell thicknesses in the presence of Hg²⁺ and other metal transition ions

In the QD-0.7 ML and QD-1 ML samples, all the evaluated ions produce fluorescence turn-on (Fig. 10). However, the turn-on effect of Hg²⁺ induced is higher [16, 39]. The role of the shell is essential to protect the core and avoid cation exchange reactions with foreign metal ions. We demonstrated that cations can passivate the surface of the QDs, eliminating some trapping sites and thus increasing the photoluminescence.

Conclusions

Three samples of oleic acid-capped CdSe/ZnS QDs with different shell thicknesses (0.3 ML, 0.7 ML, and 1.0 ML) were obtained by varying the [S]/[Zn] molar ratio of shell precursors. Characterization measurements of their optical and structural properties revealed differences in chemical speciation. Excess of S²⁻ precursor generated underpassivated S²⁻ species in the sample with the thinner shell. In contrast, in samples with a thicker shell (obtained with a [S]/[Zn] molar ratio very close), the underpassivated Zn²⁺ species were primarily formed. Underpassivated S²⁻ and Zn²⁺ species were related to hole and electron trapping sites, respectively, which affect PL.

The ability of CdSe/ZnS QDs samples to detect Hg²⁺ by PL quenching and enhancing was studied. These oppositive mechanisms could be tuned due to changes in the thickness and chemical speciation in core/shell QDs. PL quenching is produced for the sample with the thinner shell due probably to the cation exchange of Hg²⁺ by Cd²⁺ (from core) affecting PL. Samples with a thicker shell (with 0.7 and 1.0 ML) protect the core of the unwanted cation exchange reaction, producing longer fluorescence lifetimes and better PL. In addition, Hg²⁺ ions contribute to the passivation of traps. In detecting Hg²⁺, the lower LOD value (8.98 nM) was obtained with the CdSe/ZnS sample with 0.7 ML. However, other foreign cations also achieved increased PL; therefore, the selectivity is affected. Results show that core/shell QDs are promising for detecting Hg²⁺ ions, and tuning shell properties is a determinant factor in the sensitivity and selectivity of the application.

Electronic Supplementary Material

Below is the link to the electronic supplementary material.

Author Contributions

F.O. and B.G. carried out experimental work and prepared all figures. N.M. and G.G. supervised the experimental work and wrote the manuscript.

Funding

Universidad Nacional de Colombia, Sede Bogotá, funded this research.

Open Access funding provided by Colombia Consortium

Data Availability

All data generated or analyzed during this study are included in this manuscript and its supplementary information file.

Declarations

Ethical Approval

This article does not contain any studies with human or animal subjects.

Competing Interests

The authors declare no competing interests.