Benefit of porous silica nanoreactor in preparation of fluorescence carbon dots from citric acid

Citric acid (99.5%) was purchased from Proquimios (Brazil), sodium hydroxide (99%) was from Dinamica (Brazil), sodium bicarbonate (99%) was supplied by Synth (Brazil). Four sets of silica gel (all from Sigma–Aldrich) were used. Their different porous characteristics, particle size, and geometry are presented in table 1. Calcium hydride (95%), toluene(99.8%), and molecular sieves (3 Å, 4–8 mesh) were purchased from Sigma–Aldrich (USA). Toluene was dried by distillation over calcium hydride and stored in a dark bottle with molecular sieves. All aqueous solutions were prepared with ultra-pure water from a water purification system (The PURELAB Classic, Elga, UK).

The textural properties of silica gel samples were determined by means of the low-temperature nitrogen adsorption-desorption method at 77 K on an ASAP 2020 (Micromeritics, USA). The total pore volume was determined at the maximal adsorption point and specific surface area was calculated through the BET method. The pore size distribution was calculated through the Nguyen-Do approach modified by Gun'ko (MND) [52–55]. The pH values and the electrical conductivities of supernatant liquids were measured using a PHS-3E meter with a BioTrode lab glass membrane electrode (Hamilton, USA) and HI 8633 conductivity meter (HANNA instruments, UK), respectively. Fourier transform infrared spectroscopy (FT-IR) spectra, recorded within the 4000–400 cm⁻¹ range, were collected on FTLA-2000 spectrophotometer (Thermo Scientific Nicolet). Electronic (UV–vis) absorption spectra (in 200–450 nm range) were measured on a Cary 100 UV–vis spectrophotometer (Agilent, USA). X-ray photoelectron spectroscopy (XPS) was used to analyze the surface composition with the help of a K_α X-ray photoelectron spectrometer (Thermo Fisher Scientific, UK) equipped with a high-performance hemispherical electron analyzer with a 128-channel detector and an aluminum anode X-ray source (K_α = 1486.6 eV), providing an energy resolution of about 1 eV. Photoluminescence measurements were realized using LS 55 luminescence spectrometer (Perkin-Elmer, UK) with 1 cm optical path length quartz cuvettes and with a solid substrate measurement accessory.

The 3D photoluminescence emission spectra were recorded at room temperature and in ultra-high vacuum (10⁻⁷ mTorr) conditions at the Toroidal Grating Monochromator (TGM) beamline of the Brazilian Synchrotron Light Laboratory (LNLS). Photons within the energy range of 4.5–8 eV (276–155 nm) were used as a source of exciting light after filtering on a quartz window. Light emission was transferred by an optical fiber line (aperture: 600 μm) and measured on an R928 Hamamatsu photomultiplier (PMT). The spectra were corrected for the variation in the incident flux of the excitation beam using the excitation and emission spectra of sodium salicylate as standard.

The morphology of the samples was studied by means of a low-vacuum JSM-6490LV (JEOL, Japan) scanning electron microscope (SEM). Atomic force microscopy (AFM) images were obtained using a Bruker MultiMode 8 microscope. A Si tip with a force constant of 0.4 N m⁻¹ and a resonance frequency of 70 kHz were used in tapping the peak force method. The samples were diluted in deionized water, ultrasonicated and a single drop was dried on a clean mica surface in a nitrogen flux for 2 h. Statistical analysis of the thickness distribution was performed using Gwyddion 2.53 software.

The powder X-ray diffraction (XRD) pattern of the CDs product was collected at room temperature using a X'Pert PRO (Panalytical, UK) diffractometer equipped with a PIXcel one-dimensional hybrid pixel technology position-sensitive device detector and operated with Ni-filtered CuK_α radiation (λ = 1.54178 Å).

Quantum Yield (QY) of CDs was determined from single-point methods [56]. Quinine sulfate in 0.1 M H₂SO₄ has been used as a standard. Before the measurements, CDs suspensions and quinine sulfate were diluted to the absorption level below 0.1 at the excitation wavelength (344–360 nm) in order to minimize reabsorption effects. Both absorbance and photoluminescence emission spectra were measured for these samples. QY of CDs was measured for two excitation light, at 340 and 350 nm and calculated using equation (1).

$$\begin{eqnarray}&&{Q}_{s}={Q}_{R}\displaystyle \frac{{I}_{S}}{{I}_{R}}\cdot \displaystyle \frac{{A}_{R}}{{A}_{S}}\cdot \displaystyle \frac{{\eta }_{S}^{2}}{{\eta }_{R}^{2}},\end{eqnarray} \tag{ 1 }$$

where ${Q}_{s}$ = quantum yield of CDs; ${Q}_{R}$ = quantum yield of reference; ${I}_{S}$ = integrated fluorescence intensity of CDs; ${I}_{R}$ = integrated fluorescence intensity of reference; ${A}_{S}$ = absorbance of CDs at 350 nm (340 nm); ${A}_{R}$ = absorbance of reference at 350 nm (340 nm); ${\eta }_{S}$ = refractive index of the solvent of CDs; ${\eta }_{R}$ = refractive index of the solvent of reference.

The integrated intensity of PL was determined using software SpectraGryph 1.2. The refractive index of water is 1.33.

Conventional synthesis of CDs was performed by the pyrolysis of bulk citric acid in the inert atmosphere [57]. Briefly, 10 g of CA was placed into 250 ml two-neck round-bottomed flask and heated to 170 °C under nitrogen atmosphere using graphite bath. The transparent liquid became orange after 30 min of heating, which implied the formation of CDs [57]. The resulted liquid was poured out into 130 ml of NaOH (10 mg ml⁻¹) solution under vigorous stirring. The obtained mixture was filtered through a 0.45 μm filter to remove large particles and stored for further analysis in the refrigerator (at 4 °C).

General scheme of SiO₂@CA preparation, thermal treatment, and CDs elution is presented in figure 1.

Zoom In Zoom Out Reset image size

A set of silica-based composites having CA adsorbed was prepared from bulk silica SiO₂@CA. The set consists of four silica samples with different porosity and was prepared using a similar procedure by adding 1.0 ml of CA aqueous solution (1.5 g ml⁻¹) to 1.0 g of silica. The mixture was sonicated for 10 min and the excess of liquid was removed by blotting it with filter paper until dry. Then 2 mL of diethyl ether was added to silica composite and shaken for 20 min Excess of ether was removed by decantation and the procedure was repeated twice. Finally, the samples were dried under vacuum (0.1 Torr) at 60 °C producing a loose powder.

An additional sample of SiO₂@CA was prepared as a reference (rSiO₂@CA) by following the experimental procedure, recommended in [41]. Briefly, 200 mg of silica gel SiO₂(11) and 100 mg of CA were mixed and sonicated in 0.5 ml of water. The mixture was dried (without filtration and washing) in an oven at 60 °C, producing an agglutinated powder.

All silica gels samples were heated at 170 °C and 200 °C under nitrogen atmosphere. Portions of the samples were collected at different treatment time (τ), from 15 to 720 min, to obtain SiO₂@CDs-τ.

The conditions of CDs releasing from SiO₂@CDs were studied in three consecutive steps: (1) the composites were suspended in 5 ml of water and shaken for 15 min. The solid phase was separated by filtration and the washing procedure was repeated till no fluorescence was observed for the filtrate. Alternatively, SiO₂@CDs were washed in Soxhlet apparatus with water. The supernatant was collected, filtrated through a 0.22 μm filter and stored at 0 °C. (2) The samples were washed with 0.1 mol l⁻¹ NaHCO₃ solution (portions of 5 ml) in the same manner as it is described for the first step. The supernatant was collected and stored equally. (3) Finally, the silica scaffold was dissolved in 3 ml of NaOH solution (3 mol l⁻¹) at room temperature.

For proper preparation of CDs under the nanoreactor approach, it is critical to use the hosts with suitable pore size. It is generally accepted that CDs have a particle size of less than 10 nm [5, 7, 58]. Therefore, in contrast to earlier publications where microporous zeolites with small particle size have been used, silica gels with 4–11 nm nominal pore diameter and large particles (40–200 μm) is used in the current research, table 1. The latter shall reduce CA adsorption on an external area of the particles and so prevent the formation of CDs outside of the pores.

Since porous characteristics of the host are essential for the nanoreactor approach, they were justified form nitrogen adsorption experiment. It was confirmed that the selected silica samples have the desirable pore sizes from 5.5 to 10.2 nm, with 0.7–1.25 cm³ g^-1 pore volume, table 2. Accordingly, the mesopores contribute more than 95% (commonly 99.7 ± 0.3) to the overall porosity of the silica gels, table 2. Selected samples of the host have a unimodal and narrow distribution of the pores (figure 2), which allows expecting the narrow distribution of hosted CDs particles.

Zoom In Zoom Out Reset image size

A crucial aspect of the nanoreactor synthesis of CDs is verification that a major part of the precursor is localized in the host pores. To ensure this, the volume of CA solution used for the preparation of SiO₂@CA composites was limited by pore volumes of the silica gels. Also, high loading of hydrophilic silanol groups on the silica surface ensures good affinity of CA to the host. Taking into account the pore volume of silica gel samples and solubility of CA in water (1.5 g ml⁻¹) maximum loading of the precursor was estimated, table 2. For example, 1 g of silica SiO₂(11) can hold up to 1.9 g of CA inside of the pores. Nevertheless, after the treatment of silica gel even with a small volume of liquid, wet material is always obtained due adsorption of some part of the water solution on inter-particle space. In order to obtain SiO₂@CA having the precursor in the pores only, this interparticle liquid having dissolved compounds must be removed before the pyrolysis. To prove the importance of such pre-treatments, a sample labeled as rSiO₂(11)@CA was prepared according to the procedures recommended in [41, 42, 59]. By comparison of mass the silica sample before and after treatment with CA, it was found that loading of CA on rSiO₂(11)@CA constitute of 112% of maximum pore capacities for this material. Therefore, after drying rSiO₂(11)@CA is a compact glued agglomerate. After two consecutive rinsings of rSiO₂(11)@CA with ether, CA loading is reduced to 60% of maximum pore capacity. Such samples of SiO₂@CA present itself as a loose dry powder. The SEM images confirm the difference between rSiO₂@CA particles and SiO₂(11)@CA particles (figure 3), demonstrating a lot of CA microcrystals located on the external surface of rSiO₂@CA and absence of those on SiO₂(11)@CA.

Zoom In Zoom Out Reset image size

Thus, the importance of the procedure for removing CA excess from an outer surface of silica composite was considered. In order to do so, wet samples of as-prepared SiO₂@CA were blotted with filter paper until dry and then were washed out with two portions of diethyl ether. The ether is not miscible with water, therefore it was expected that ether will not penetrate the pores of wet SiO₂@CA filled by water and so will remove the excess of CA form the external surface of the particles only. It could get samples of SiO₂@CA with high loading of CA in the poses and clean outer surface as it is demonstrated in figure 3.

3.3.1. Synthesis of SiO₂@CDs

Four samples of silica-based material were used as a host for CDs precursors. After impregnation with the precursor and pre-treatment, samples of SiO₂@CA were thermally treated in an inert atmosphere at 170 °C for 5–600 min and then subsequently washed with water, NaHCO₃, and NaOH to release CDs, table 3.

Table 3.  Overview of the host samples used for the preparation of CDs from citric acid precursor.

Host	Pore diameter, nm	Time of pyrolysis, min	Post-treatment
SiO2	5.5, 6.0, 6.9, 10.2	5, 15, 30, 60, 120, 240, 360, 600	H2O, NaHCO3, NaOH
—	n.a.	180	H2O, NaHCO3

Earlier it was shown that appropriate temperature for conventional CDs synthesis from CA is 150 °C–200 °C [34, 60]. The temperature in 170 °C was chosen as the middle of the optimal interval and maintained constant in order to compare the results of pyrolysis for different samples. Conventional CDs were also obtained by pyrolysis of CA at 170 °C for 180 min After the thermal treatment SiO₂@CDs had light brown, and conventional CDs black color¹, figures S1–S2 in the ESM. Under UV light, SiO₂@CDs demonstrated bright blue photoluminescence, figure 4. The reference sample rSiO₂@CDs has a brown color with randomly-distributed photoluminescence, figure S2 in the ESM.

Zoom In Zoom Out Reset image size

3.3.2. FT-IR spectra

The FTIR spectrum of silica gel host exhibits several peaks related to silanol groups: at 3750 cm⁻¹ for isolated Si-OH; 3670–3220 cm⁻¹ for hydrogen-bonding Si-OH; at 1640 cm⁻¹ deformation vibration Si-OH. Also, FTIR spectra of silica host demonstrate two peaks at 1800–2000 cm⁻¹ that corresponds to overtones from Si–O–Si stretching vibrations, figure 5. Spectrum of SiO₂(6)@CDs-180 reveal itself as a superposition of the host and the guest (CDs) and in addition to the bands of silica, SiO₂(6)@CDs-180 sample demonstrates well-defined peaks at 1775 cm⁻¹ and 1702 cm⁻¹ that can be attributed to stretching vibration of C=O fragments in –C(O)(OR) and –COOH functional groups of silica-adsorbed CDs [51, 59].

Zoom In Zoom Out Reset image size

3.3.3. XPS spectra

The XPS survey spectra for the SiO₂(6)@CDs-180 (figure 6(a)) indicate predominantly O1s, C1s, Si2s and Si2p peaks centered, respectively, at 532, 286, 155 and 103 eV.

Zoom In Zoom Out Reset image size

The high-resolution C1s XPS spectra of samples revealed different types of carbon atoms in CDs. Curve-fitting analysis of high-resolution XPS of C1s reveals bands at 284.5 eV, 286.3, 287.5 and 289 eV that can be assigned to C–C (44%), C–OH (24%), carbonyl (26%) and carboxyl (6%) groups, correspondingly [24], figure 6(c). This king of XPS is very typical for CDs and demonstrates the formation of partly decarboxylated CDs. A weak band at 551,5 eV attributed to carbonyl fragments in CDs can be seen in high-resolution O1s XPS spectrum of SiO₂(6)@CDs-180, figure 6(b). This kind of group is common for CDs obtained earlier [1].

3.3.4. Photoluminescence of solid composites SiO₂@CDs-τ.

The incipience of CDs has been monitored by measuring the photoluminescence from solid samples of SiO₂@CDs obtained for different time of thermal treatment, figure 7. Evidently, the CDs are formed in silica pores very quickly. At least, luminescence of SiO₂@CDs can already be detected after 5 min of SiO₂@CA treatment at 170 °C. However, the position of PL maximum (420 nm) for SiO₂@CDs obtained after such a short time (5–60 min) does not match to PL of CDs obtained from CA by conventional methods (445 nm) [28]. Together with the peak at 420 nm, PL spectra of all SiO₂@CDs exhibit a peak at 450 nm, figure 7. Thus, PL spectra of SiO₂@CA commonly show a wide band, which consists of two peaks at 420 and 450 nm.

Zoom In Zoom Out Reset image size

While the peak positions do not substantially depend on the time of SiO₂@CDs thermal treatment and the size of the host pores, the relative intensity of long-wave emission rises with time for all studied SiO₂@CDs that can manifest the development of CDs with enhanced decarboxylation degree.

The absolute intensity of SiO₂@CDs reaches its maximum for 15–30 min of SiO₂@CA thermal treatment and then sharply declines for all studied samples, figure 8(a). After 100 min of the treatment, no further depression of PL was observed, figure 8(b).

Zoom In Zoom Out Reset image size

The photoluminescence spectra of SiO₂@CDs clearly exhibit the formation of CDs captured in silica pores. Apparently, those CDs also have different sizes and shapes since they are characterized by multiple maxima in PL spectra of SiO₂@CDs. The nature of the PL band at 420 nm calls for further investigations.

The SiO₂@CDs composites also demonstrate photoluminescence that occurs under excitation with high energy radiation (165–276 nm, 7.5–4.5 eV). For example, SiO₂(6)@CDs generates wide emission with a maximum at 525 nm, figure 9(b). From two-dimensional mapping, as a function of the excitation energy, it can be seen that the peak position of photoluminescence emission almost does not change in the function of excitation energy, figure 9(a). The most intense photoluminescence for SiO₂(6)@CDs is observed at 4.0–4.2 eV (310–295 nm). Presumably, SiO₂(6)@CDs can be useful for adsorption and transformation of short wave UV radiation.

Zoom In Zoom Out Reset image size

To elute the resulted CDs from the silica host three consecutive steps were applied. First, they were thoroughly washed with water until the conductivity of the washing solution droped down essentially and absorption at 300 nm was about zero, figures 10(a), (b).

Zoom In Zoom Out Reset image size

Then, 0.1 mol l⁻¹ of NaHCO₃ water solution was exploited for the same purposes. Finally, SiO₂ scaffold was dissolved in NaOH. It has been found that water elutes strongly luminescent CDs from all samples of SiO₂@CDs-τ. Four washing cycles were enough to remove most of the trapped CDs, figure 10(a). Further washing of SiO₂@CDs-τ with NaHCO₃ and even solubilization of the host in NaOH do not reveal an essential quantity of the CDs, except for the samples treated for 600 min, figures 10(c), (d). The fact of a weak affinity of the CDs to the SiO₂ host can be concluded from this experiment. It also can be assumed that silica gel is more appropriate host for confining of CDs particles then earlier used zeolites because it releases the CDs through treatment with water and does not require solubilization of the host in alkali solution.

During the thermal treatment of SiO₂@CA, most of CDs are growing inside of the host pores without essential interaction with the surface of silica. Such a type of CDs particles can be eluted with water from the hosts. Immersing of SiO₂@CDs into NaHCO₃ solution promotes ionization of silanol groups of silica surface and carboxylic fragments on CDs surface. Next, negatively charged silica surface stimulates elution of the negatively charged CDs.

3.5.1. Microscopy

AFM technique was utilized to probe the sample topography and measure particle height. AFM image of CDs prepared from rSiO₂@CDs according to the procedure [41] shows the wide distribution of the particle's height mainly from 5 to 8 nm, figure 11(a). The CDs obtained from SiO₂(4)@CDs-180 have a smaller height (2–4 nm) and narrower distribution profile, figure 11(b).

Zoom In Zoom Out Reset image size

These data support the results reflected on the SEM images of rSiO₂@CA particles (figure 3), suggesting the possible formation of CDs on the outer surface of the silica template. If so, the growth of such particles is not confined by the template, and CA pyrolysis can lead to formation CDs with size, larger than the size of silica pores. On the contrary, the experimental procedure proposed in current research demonstrates a narrow distribution of the CDs particle height. Thus, the importance of the pre-treatment of SiO₂@CA nanocomposites for removing CA excess that can be adsorbed on the outer surface of silica is feasible.

3.5.2. The x-ray diffraction of CDs.

3.5.3. Photoluminescence of CDs obtained in nanoreactors

All CDs eluted from SiO₂ hosts exhibit emission band centered at 447 nm, figure 12. The band intensity essentially depends on the time of SiO₂@CA thermal treatment and reaches its maximum for CDs obtained from SiO₂@CDs-60 (figures 12(a)–(c)). The PL band symmetry of CDs deteriorates with increasing of the maternal host pore size, reviling short-wavelength shoulder, figure 12(d).

Zoom In Zoom Out Reset image size

This observation is consistent with those obtained for SiO₂@CDs (figure 7) and can indicate the formation of less decarboxylated CDs for short treatment time (less than 60 min) in silica gel with larger pores.

Quantum yield was determined for CDs from SiO₂(d)@CDs-r thermally treated for 60–600 min. The results presented in table 4 demonstrate that the CDs obtained in silica host with narrow pores have higher QY (up to 10.1%) than those obtained in larger porous (8.0%). It seems that prolonged time of the thermal treatment increases QY of CDs, at least for those obtained in silicas with narrower pores.

Table 4.  Quantum yield of water suspension of CDs, obtained from SiO₂(d)@CDs-r.

Sample	QY (exc. 340 nm), %	QY (exc. 350 nm), %
SiO2(4)@CDs-170–60	7.3	5.5
SiO2(4)@CDs-170–600	10.1	7.9
SiO2(6)@CDs-170–180	6.3	4.7
SiO2(11)@CDs-170–600	8.0	5.2