Polyvinylpyrrolidone-Capped CuInS₂ Colloidal Quantum Dots: Synthesis, Optical and Structural Assessment

Abstract

Ternary metal chalcogenide quantum dots (QDs), such as CuInS₂, have attracted significant attention due to their lower toxicity compared to binary counterparts containing cadmium or lead, making them promising candidates for biomedical imaging and solar energy applications. The surfactant choice is critical for controlling nanocrystal nucleation, growth kinetics, and functionalization. This directly affects the toxicity and applications of QDs. In this work, we report a synthesis protocol for PVP-capped CuInS₂ QDs in an aqueous solution. Using density functional theory (DFT) calculations, we predicted the coordination patterns of PVP on the CuInS₂ QDs surface, providing insights into the stabilization mechanism. The synthesized QDs were characterized using TEM, XRD, XPS, and FTIR to assess their morphology, chemical composition, and surface chemistry. The QDs exhibited dual photoluminescence (PL) maxima at 550 nm and 680 nm, attributed to defect-related emissions, making them suitable for cell imaging applications. Cytotoxicity studies and cell imaging experiments demonstrate the excellent biocompatibility and effective staining capabilities of the PVP-capped CuInS₂ QDs, highlighting their potential as fluorescent probes for long-term, multicolor cell imaging including two-photon microscopy.

1. Introduction

Colloidal quantum dots (QDs) of I–III–VI compounds are considered to be an alternative to the binary II–VI QDs such as CdS(Se) and PbS(Se) due to their relatively low toxicity owing to the absence of cadmium and lead in their composition. The large Stokes shift (200–800 meV) that reduces the reabsorption effect makes these materials superior to organic dyes and perovskite nanocrystals for some applications [1,2,3,4,5]. By controlling technological parameters, it is possible to change the photoluminescence (PL) peak position of I–III–VI QDs in the range from visible to near-infrared (450–1200 nm) both by varying the synthesis time and adjusting the molar ratio of the constituent I and III cations [6,7,8,9].

Compared to simpler and less toxic nanocrystal systems such as InP, CuInS₂ QDs offer several unique advantages, particularly their high tolerance to stoichiometric deviations and the ability to form various types of intrinsic point defects. These structural features allow for fine-tuning of the material’s optical properties through controlled defect engineering. The photoluminescence in CuInS₂ QDs is largely governed by emission from defect-related states, which can be modulated by adjusting the synthesis conditions and elemental composition within a broad homogeneity range. This flexibility enables precise control over emission wavelength, spectral width, and quantum yield—features that are more difficult to achieve in more stoichiometrically rigid systems like InP [10,11].

Colloidal QDs as functional materials find application in flexible electronics and solar energy [12,13,14,15,16]. Colloidal I–III–VI QDs are also attracting attention for biomedical research [17,18,19]. One of the promising areas is the use of QDs for the visualization of cells and tissues [20,21,22,23]. Usually, QDs must be conjugated with biomolecules to label specific cell proteins. Bioconjugation provides binding specificity and can be implemented in several ways. One of the perspectives for conjugating is to cap the QDs with a polymer surface.

Obtaining QDs in aqueous solutions using hydrophilic ligands such as mercaptopropionic acid, mercaptoacetic acid, L-glutathione, etc., [10,11] is an established direction in QDs synthesis. In [24], a synthesis protocol for polyvinylpyrrolidone-capped AgInS₂ QDs in aqueous solution was reported. Polyvinylpyrrolidone (PVP) is a synthetic polymer with valuable properties, including biocompatibility, non-toxicity, high chemical stability, and excellent solubility in water and most organic solvents. Due to these characteristics, PVP is widely used in various fields, such as pharmaceuticals, biomedicine, and environmental applications [25,26]. Coating the QDs with a non-toxic, water-soluble linear polymer composed of 1-vinyl-2-pyrrolidone monomers provides a promising opportunity to expand the use of QDs in biomedicine and biosensors.

There are two potential chemisorbing sites on PVP related to the amide group in the heterocyclic ring: carbonyl oxygen and nitrogen. Up to this point, the direct nanoparticle coordination with PVP was demonstrated predominately in the case of noble metal nanocrystals [27,28]. Different models of the interaction between PVP and metal and semiconductor nanoparticle surfaces were proposed, but definite understanding of the stabilization mechanism is lacking.

In this regard, the aim of our research is to develop an aqueous synthesis method for PVP-capped CuInS₂ QDs and investigate their stabilization mechanism using DFT. While similar synthetic approaches have been reported, the novelty of our work lies in combining experimental characterization with DFT modeling to elucidate the coordination chemistry between PVP and the QD surface. This theoretical insight provides a foundation for understanding the role of PVP in stabilizing CuInS₂ QDs, which could guide future optimization of synthesis protocols.

2. Materials and Methods

2.1. Chemicals

Indium(III) chloride (InCl₃, 99.9%) was purchased from Sigma Aldrich (St. Louis, MO, USA). Copper(II) acetate (Cu(CH₃COO)₂, 99%), sodium sulfide (Na₂S∙9H₂O, 98%), and Poly(N-vinyl-2-pyrrolidone) (av. M_w 12,600) were provided by Vekton (St. Petersburg, Russia). All chemicals were used without additional purification.

2.2. PVP-Capped CuInS₂ QDs Synthesis

PVP-capped CuInS₂ QDs were synthesized in aqueous solution as follows. Cationic precursor solution with molar ratio [Cu]:[In] = 1:4 was prepared by dissolving 2.0 mg (0.01 mmol) of copper(II) acetate and 8.0 mg of PVP in 5 mL of deionized water followed by addition of 8.8 mg (0.04 mmol) of indium (III) chloride. Quick injection of aqueous solution containing 48 mg (0.2 mmol) of sodium sulfide resulted in instantaneous nucleation and the obtained mixture of QD cores was heated at 95 °C for 1 h.

2.3. Methods

X-ray diffraction (XRD) and transmission electron microscopy (TEM) techniques were used for structural characterization. TEM images were obtained using a JEOL JEM-2100F microscope, JEOL, Akishima, Japan (accelerating voltage 200 kV, point resolution 0.19 nm). XRD patterns were taken on a Rigaku SmartLab diffractometer (Rigaku Corp., Tokyo, Japan) with a CuKα, 45 kW, 200 mA in the 2θ range from 10 to 60°. The surface chemical electronic state and composition of the PVP-capped CuInS₂ QDs were measured using a “K-Alpha” Thermo Scientific system (Waltham, MA, USA) X-ray photoelectron spectroscopy (XPS) instrument equipped with an Al-Kα (1486.6 eV) X-ray source. Infrared spectra of PVP-capped CuInS₂ QDs samples are taken using a Vertex 80 Fourier-transform infrared (FTIR) spectrometer (Bruker Optics, Ettlingen, Germany). A PL measurement setup based on the same FTIR was employed to obtain PL spectra. A 405 nm semiconductor laser diode was used as an excitation source. The UV–vis absorption spectra were recorded with a PE-5400UV UV–vis spectrophotometer (“Ekohim” LLC, St. Petersburg, Russia) in the 350–850 nm range and 1 nm resolution.

2.4. Computational Details

The possible coordination patterns between the PVP and CuInS₂ QDs’ surface were predicted by means of quantum mechanical modelling at the DFT level of theory. The geometry optimization procedure for all model structures was carried out at the B3LYP-D4/def2-TZVP level of theory by using the Orca 5.0.4 program package [29]. The following convergence tolerance criteria were used: energy change 5.0 × 10⁻⁶ Eh, maximum gradient 3.0 × 10⁻⁴ Eh/Bohr, RMS gradient 1.0 × 10⁻⁴ Eh/Bohr, maximum displacement 4.0 × 10⁻³ Bohr, and RMS displacement 2.0 × 10⁻³ Bohr. The RIJCOSX approximation [30,31] utilizing the appropriate auxiliary basis sets was used to reduce the computational costs. The couple perturbed self-consistent field equations were solved using the conjugated gradient method. The Hessian matrix was calculated for fully optimized model structure of PVP to confirm the location of a minima on the potential energy surface (no imaginary frequencies were found). The CuInS₂ crystal elementary cell in the chalcopyrite phase was utilized as model for the CuInS₂ QDs’ surface. Structural constraints were used for all atoms of CuInS₂ moieties in model structures to eliminate geometry distortion during the optimization process. The difference between the binding energies of the individual molecules of PVP and CuInS₂ in various associates of PVP–CuInS₂, in terms of total electronic energy values change (∆E), was used to assess the energetical favorability of the PVP and CuInS₂ interaction via different possible binding sites. The ChemCraft 1.8 program was used for the visualization of model structures of monomers and associates. Cartesian atomic coordinates of all model structures are given in the Supporting Information as xyz-files.

2.5. Cell Culture and Processing

Mouse C2C12 myoblasts (ATCC CRL-1772) were purchased from Sigma Aldrich (St. Louis, MO, USA). The cells were cultured in growth medium, consisting of Dulbecco’s modified Eagle’s medium enhanced with 10% fetal bovine serum (BioloT, St. Petersburg, Russia), and 1% penicillin/streptomycin (BioloT, St. Petersburg, Russia) at 37 °C and 5% CO₂.

The PVP-capped CuInS₂ QDs at a concentration of 1 ÷ 70 nmol/L were added in 12-well plates (BioloT, St. Petersburg, Russia) with growing C2C12 cells for 2–3 days. The number and morphology of cells in each well were monitored using bright field microscopy with a LEICA DMI8 microscope (Leica Microsystems GmbH, Wetzlar, Germany). The difference between cells was determined by visual analysis. Total cell numbers were calculated by manual counting of cells per image section. Approximately 10 image sections were analyzed for every sample in each experiment.

The viability of cells was determined by MTT-assay. The cells were primary incubated with PVP-capped CuInS₂ particles added at a concentration of 1 ÷ 70 nmol/L for 3 days. Then the medium was removed, the cells were washed up three times and 10 µL of MTT solution (1 mg/mL) was placed in each well plate. The MTT solution was prepared in phosphate-buffered saline at a pH of 7.4. The cells were incubated for 4–6 h at 37 °C in a CO₂ incubator at a constant shaking. The formazan crystals were dissolved in 300 µL of dimethyl sulfoxide). The optical density was assessed at 540 nm. The optical density at 720 nm was recorded to correct a background noise. The optical density ratio between the samples’ values and the optical density of the control cells was used to determine the viability of the cells.

2.6. Imaging and Signal Analysis

For microscopy, the cells were fixed with 70% ethanol for 15 min at 4 °C. Then, cells were washed in phosphate-buffered saline 3 times. Slides with fixed cells were put at inverted position on a cover glass. The samples were examined via multiphoton microscopy (Bergamo II multiphoton microscope, Thorlabs, Newton, MA, USA) [32]. A tunable femtosecond Ti:Sapphire laser (Tiberius, Thorlabs, Newton, MA, USA) was used with the following pulse parameters: duration of 140 fs, a repetition rate of 77 MHz, a wavelength of 780 nm, and an average laser power of 75 mW. Photoluminescence was collected with a photomultiplier tube in the wavelength range of 300 to 705 nm.

3. Results and Discussion

3.1. Structure Characterization

In order to establish morphology and structure of the PVP-capped CuInS₂ QDs, TEM and XRD studies were performed.

TEM measurements (Figure 1a) reveal an average particle diameter of approximately 2 nm, smaller than the exciton Bohr radius in CuInS₂, confirming strong quantum confinement. The XRD pattern (Figure 1b) had a complicated form with broadening of XRD peaks characteristic of nano-sized crystalline materials. It pointed to the presence of a CuInS₂ phase (patterns at 2θ values of 27.9, 46.4, and 55.0°). The observed Bragg peaks agreed well with (112), (220), and (312) planes of chalcopyrite CuInS₂ (JCPDS 65-2732) indicating that obtained nanoparticles were nuclei of a tetragonal crystalline phase of CuInS₂.

Figure 2 shows the XPS spectra of PVP-capped CuInS₂ QDs.

From Figure 2b, it can be seen that Cu 2p peaks were found at 931.3 eV (Cu 2p_3/2) and 951.1 eV (Cu 2p_1/2), In 3d peaks were found at 444.3 eV (In 3d_5/2) and 451.8 eV (In 3d_3/2), and S 2p peaks were found at 168.4 eV and 161.4 eV. The sulfur spectrum is represented by a superposition of two chemical states with various binding energy shifts of the S 2p doublet. The main peaks of the survey spectrum belonged to Cu, In, S, C, N, Na, and O elements.

The chemical composition is complex in the obtained sample. According to deconvolution presented in Figure 2d, 83.3 at.% of sulfur is in the three-component CuInS₂ system (S 2p_3/2 peak at 161.4 eV) [33]. Another chemical state of sulfur was ascribed to sodium sulphate [34].

In Figure 3, FTIR spectra of PVP and PVP-capped CuInS₂ QDs are presented.

The FTIR spectrum of PVP-capped CuInS₂ QDs reveal characteristic absorption bands related to PVP heterocycle (1291, 1319, 1374, 1423, 1438, 1463, and 1495 cm⁻¹) [35], C–N stretching vibration (1074 [36] and 1018 cm⁻¹), the out-of-plane C–H bending, rocking, and OH wagging (845, 736, and 650 cm⁻¹). The absorption bands at 2955 and 2922 cm⁻¹ are due to C–H bond symmetric and asymmetric stretching, respectively, and the broad line at around 3500–3400 cm⁻¹ is characteristic for the –OH group of H₂O [35].

The results of FTIR spectroscopy did not allow us to experimentally determine the coordination mechanism in PVP-capped CuInS₂ QDs in this work.

3.2. Theoretical Calculations

Bulk-scale DFT modeling of nanocrystals [37] demands substantial computational resources, rendering such calculations costly. To theoretically study the association of PVP with CuInS₂ on the molecular level and quantitatively estimate the relative energetical favorability of hypothetically formed associates, we carried out quantum mechanical calculations at the B3LYP-D4/def2-TZVP level of theory for model structures. To calculate the binding energy between PVP and CuInS₂, we used the unit cell of CuInS₂ as model. As noted previously [38], the cluster approach combined with hybrid functionals effectively reduces computational costs while maintaining accuracy in electronic property descriptions. This method was widely employed in catalytic reaction studies, where DFT errors arising from model size constraints often have minimal impact on relative energy comparisons. Thus, our approach focuses on elucidating fundamental electronic interactions between PVP and the CuInS₂ QDs surface, circumventing the need to explicitly model the entire QD. The full geometry optimization of model CuInS₂ QDs resulted in a significant distortion of the initial unit cell structure (Figure 4, process 1′–1). Therefore, we used structural constraints for model CuInS₂ QDs to preserve the correct crystal geometry. Positional constraints on the unit cell prevent relaxation artifacts that could distort critical parameters such as adsorption energy. This strategy aligns with established practices in nanomaterial simulations, where constraints are routinely used to maintain structural representativeness, as demonstrated in electrocatalyst modeling studies [39]. We considered different initial locations of the PVP relative to CuInS₂ QDs near the copper and indium atoms (Figure 4). It was shown that the Cu–O binding energy could be up to –36 kcal/mol depending on the conformation of the associate (Figure 4, 2–5). The location of PVP near the indium atom results in the migration of PVP to the copper atom (Figure 4, processes 3′–3, 4′–4, and 5′–5). The binding In–O is only possible with the participation of two lone electron pairs of the PVP oxygen and the two closely spaced indium atoms (Figure 4, process 6′–6) with binding energy equal to –33 kcal/mol. Thus, dominant coordination of PVP to CuInS₂ via the formation of Cu–O bonds was collaterally confirmed by DFT calculations and highly expected in real samples of QDs. While these findings have not affected current synthesis protocol directly, they may provide a theoretical foundation for future optimization of PVP-capped QDs by elucidating the coordination interactions between the oxygen atom of the PVP fragments and the Cu(I) and In(III) ions at the molecular level.

3.3. Optical Properties of PVP-Capped CuInS₂ QDs

The amount of surfactant was chosen according to previous work [24], where we showed that the maximization of PL intensity of AgInS₂ QDs was achieved at an amount of PVP equal to 8 mg per 0.004 mmol [In³⁺]. PL spectra of all luminescent CuInS₂ samples exhibit two maxima, with peaks at approximately 550 nm and 680 nm (Figure 5).

Figure 5 shows the absorption and PL spectra of CuInS₂ QDs in aqueous solution. A broad and featureless absorption curve with a long absorption tail in the low-energy area without clearly defined excitonic maxima was observed. The excitation of Cu¹⁺-related states forming sub-bandgap levels is manifested by an absorption feature around 550–650 nm, which results in a broad, featureless absorption shoulder that stretches into the visible region [40,41].

Unlike other QDs of this type, the PL spectra display two bands with maxima at approximately 550 nm and 680 nm. In the work [42], AgInS₂ QDs were synthesized with PL curves similarly featuring two maxima. The authors proposed the formation of In₂S₃ and AgInS₂ phases. Their explanation of PL spectra was presented in terms of reverse Type I core/shell structure. The PL mechanism was introduced as two parallel processes: direct radiative recombination in the In₂S₃ core and further trapping of a fraction of excitons by the narrower bandgap Ag-In-S shell phase, followed by their recombination. This model does not appear applicable to our experimental results due to the small average size of the QDs (approximately 2 nm). Size distribution effects may cause observation of similar multiple band PL spectra [43]. Although TEM (Figure 1a) analysis indicates a relatively narrow size distribution (2.0 ± 0.5 nm), the resolution limits of our measurements prevent definitive exclusion of minor subpopulations that could contribute to the observed PL.

CuInS₂ QDs are known to exhibit a high tolerance for off-stoichiometry, leading to the formation of intrinsic point defects that can significantly influence their optical properties [44]. In this regard, various types of intragap point defect states were proposed to participate in emission process. Structurally, the point defects could take the form of an ordered neutral defect complexes of the two I group element vacancies and one antisite defect, which is the element of the III group in a position of the element of the I group (2V_Cu + In_Cu) [34,45]. Typically, PL characteristics of I–III–VI are explained either in terms of donor–acceptor pair (DAP) recombination mechanism or recombination of localized hole and delocalized electron. As was demonstrated in [46], defect-mediated recombination via DAP mechanism may account for the broad PL spectrum, predominantly resulting from heterogeneity in defect positioning rather than intrinsic line broadening. The transition energy varies depending on the radial coordinate of Cu-related defects (e.g., Cu²⁺ or V_Cu), and the higher-energy emission (550 nm in our case) likely corresponds to surface-proximal defects with weaker electron–hole Coulomb coupling, while the lower-energy peak (680 nm) reflects deeper defects with stronger localization effects. The similar observation of two distinct PL peaks in the visible region (around 650 nm) and in the near-infrared region (around 900 nm) was observed in CuInS₂ QDs obtained through partial cation exchange in Cu_2−xS nanocrystals [47].

As was demonstrated in [48], for CuInS₂ QDs, the observation of two distinct defect-mediated recombination pathways involving copper-related centers is possible. In this case, the higher-energy emission originates from Cu²⁺ defects, where the inherent 3d⁹ configuration facilitates direct radiative recombination with conduction band electrons, while the lower-energy emission arises from Cu⁺ defects (Cu_In antisites) that require prior hole trapping for activation. These defect states create characteristic intragap energy levels, with Cu⁺ centers contributing to sub-bandgap absorption feature around 550–650 nm and influencing the broad absorption tail. The relative intensities of these emission bands reflect the complex interplay between stoichiometry (Cu/In ratio) and defect formation during synthesis, and the trapped-hole PL pathway itself was shown to be consistent with results of transient absorption spectroscopy measurements [49]. Nevertheless, in our case the energy gap between PL bands is probably too wide to be explained exclusively by this mechanism. While the defect-mediated mechanism provides the most consistent interpretation of our optical and structural data, future time-resolved studies could further elucidate the relative contributions of different recombination pathways.

3.4. Live Cell Imaging Using PVP-Capped CuInS₂ QDs

Most conventional fluorophores have an excitation spectrum in the 400–500 nm range, whereas the excitation laser operates in the 700–1000 nm range [50]. Thus, the search for suitable fluorophores for this application is of interest. QDs have the potential to become a new class of fluorescent probes for many biological and biomedical applications [22,51,52,53], especially cellular imaging. It should be noted that QDs can be promising for two-photon fluorescence microscopy due to their unique properties such as photostability, and the ability to penetrate biological tissues with reduced photodamage [54]. Size-controlled QDs allow for the customization of two-photon fluorescence wavelengths, facilitating applications in multicolor imaging and optoelectronics. In addition, QDs can also provide information about temperature in tissues [55].

There are several approaches for QDs staining. Usually, the main approach for cell or tissue labelling employs either secondary antibody (Ab)-conjugated QDs to target primary Ab, or direct conjugation between primary Ab and QDs [22,51]. Direct QD conjugation methods may suffer from several major problems, which limit their application in cancer molecular profiling.

The QDs are found to be stable in aqueous solutions and could be used for cell staining. The toxicity of CuInS₂ QDs in C2C12 cells has been determined by MTT assays. With a wide concentration range of 6.7−333.3 nmol mL⁻¹, CuInS₂ QDs were incubated with C2C12 cells for 24–48 h at 37 °C. As illustrated in Figure 6d, the cell viabilities are 2.3, 32.5, 54.9, 57.5, and 81.8% for QDs in ethylene glycol and 3.1, 34.7, 55.7, 67.7, and 84.6% for QDs in water. The results show that CuInS₂ QDs possess low toxicity and excellent biocompatibility.

As seen in Figure 6a, purple formazan crystals are observed in the cells’ cytoplasm and around the cell nucleus of the control samples. It should be noted that the color is very intense, which correlates to high cell viability (almost 95%) in the control. Increasing the concentration of QDs to 6.7 leads to a subsequent cell decrease of up to 75%. It should be noted that cell viability is higher with QDs obtained in water. Using QDs at a concentration of 33 nmol/L results in a reduction in cellular life capacity by up to 54–57%. Further increase in QDs’ concentration leads to gradual cell lysis, and we can see the destroyed cells in optical microscope images.

Additional proof of less mitochondrial activity in cells could be the smaller amount of formed formazan crystals in cells (Figure 6b). At the concentration of QDs higher than 33 nmol/L, C2C12 cells are dying. In Figure 6c, we can see transparent cells without formazan crystals, which proves viability inhibition. The use of high concentrations of QDs (above 50 nmol/L) leads to the formation of cellular conglomerates that detach from the surface (Supplementary Materials, Figure S1i,h).

To further evaluate the availability of CuInS₂ QDs in cell imaging, C2C12 cells were used. After being incubated with CuInS₂ for 24 h, C2C12 cells show a strong fluorescence emission. The CuInS₂ QDs were found to enter the cells (Figure 6e,h). It should be noted that the nuclei of viable cells generally do not fluoresce (Figure 6h) (Supplementary Materials, Figure S1a). This may serve as an additional marker for cell viability. Strong fluorescence was achieved at relatively low concentrations of QDs (up to 6.7 nmol/L), which is notably lower than the concentrations required in existing methods [54,56]. As the concentration of toxic QDs increases, cellular viability decreases, and the destruction of cell membranes results in uniform cell fluorescence. Thus, CuInS₂ QDs can be used not only for staining cells, but also for assessing their viability. These advancements highlight the potential of QDs in biomedical research, diagnostics, and nanoscale optoelectronics.

4. Conclusions

In this study, we developed a method for synthesizing PVP-capped CuInS₂ QDs in aqueous solution, highlighting their potential for biomedical imaging and optical applications. DFT calculations provided valuable insights into the coordination mechanism of PVP with CuInS₂ nanocrystals, revealing that PVP predominantly binds via Cu–O bonds. While our experimental synthesis was empirically optimized, the DFT modeling offers valuable insights for rational design of polymer-capped QDs in subsequent studies. This understanding offers insights into the synthesis optimization process and could potentially guide the development of other polymer-capped QDs with customized optical and surface properties. The QDs exhibited dual photoluminescence peaks at 550 nm and 680 nm, attributed to emissions consistent with defect-associated recombination pathways, making them highly suitable for cell imaging and other applications requiring tunable optical properties. Furthermore, the PVP-capped CuInS₂ QDs demonstrated excellent biocompatibility and effective cell staining capabilities, highlighting their potential as fluorescent probes for long-term cell imaging assessment using two-photon microscopy. The unique combination of favorable optical properties and polymer protection makes these QDs promising candidates for a wide range of applications, including biomedical imaging, solar energy conversion, and environmental sensing. The successful synthesis and characterization of PVP-capped CuInS₂ QDs pave the way for their use as versatile fluorescent probes in biomedical research and beyond.