Silica Shell Thickness-Dependent Fluorescence Properties of SiO2@Ag@SiO2@QDs Nanocomposites

Silica shell coatings, which constitute important technology for nanoparticle (NP) developments, are utilized in many applications. The silica shell’s thickness greatly affects distance-dependent optical properties, such as metal-enhanced fluorescence (MEF) and fluorescence quenching in plasmonic nanocomposites. However, the precise control of silica-shell thicknesses has been mainly conducted on single metal NPs, and rarely on complex nanocomposites. In this study, silica shell-coated Ag nanoparticle-assembled silica nanoparticles (SiO2@Ag@SiO2), with finely controlled silica shell thicknesses (4 nm to 38 nm), were prepared, and quantum dots (QDs) were introduced onto SiO2@Ag@SiO2. The dominant effect between plasmonic quenching and MEF was defined depending on the thickness of the silica shell between Ag and QDs. When the distance between Ag NPs to QDs was less than ~10 nm, SiO2@Ag@SiO2@QDs showed weaker fluorescence intensities than SiO2@QD (without metal) due to the quenching effect. On the other hand, when the distance between Ag NPs to QDs was from 10 nm to 14 nm, the fluorescence intensity of SiO2@Ag@SiO2@QD was stronger than SiO2@QDs due to MEF. The results provide background knowledge for controlling the thickness of silica shells in metal-containing nanocomposites and facilitate the development of potential applications utilizing the optimal plasmonic phenomenon.

According to the literature, fluorescent intensities increase dramatically between a metal surface and a fluorophore at a distance of 20 to 50 nm [22][23][24]. On the contrary, the quenching effect appears when the metal-fluorophore distance is below 20 nm because of the electron transfer from the excited fluorophore to the metal surface [22][23][24]. Therefore, it is necessary to develop appropriate synthesis methods, maintaining an optimum distance between the plasmonic and fluorescent components to maximize the fluorescent property 2 of 10 of hybrid plasmonic-enhanced nanomaterials and avoid metal-induced quenching. In addition, the methods should be simple and cost-effective with options for various functionalizations. Various methods have been developed via tuning the separation distance between the fluorophore and the metallic surface in controlling plasmonic-enhancement effects [25][26][27]. For example, Liang et al. used charged polyelectrolytes to control the distance between the polymeric fluorophore and Ag nanocubes [25]. In addition, various silica (SiO 2 ) shells (up to 30 nm) were coated on Ag NPs to investigate the interparticle distance-dependent fluorescence of Au nanoclusters [26]. Ke et al. reported the MEF and metal-enhanced singlet oxygen generation of SiO 2 -coated Au nanorod core-shell structures with aluminum phthalocyanine [28]. Among them, the core-shell structure provides spatial separations between plasmonic NPs and fluorophore for the MEF and also chemically stabilized metal NPs [26,29].
SiO 2 is the most widely used material for shell coatings [30][31][32][33] due to its costeffectiveness, chemical inertness, easy surface modification, biocompatibility, and optical transparency [34]. As a result, many studies on SiO 2 coatings of metal NP structures have been reported. For example, Tian et al. prepared Au@SiO 2 core-shells and etched its surface to create a pinhole structure [35,36]. Moreover, the presence of thin SiO 2 shells prevents the adsorption of molecules on the Au core and simultaneously improves thermal stability [37][38][39].
However, the precise control for SiO 2 shelling of nanocomposites remains a major challenge. In particular, the SiO 2 shell's thickness in nanocomposites, which include metallic components and fluorophores or Raman-labeling compounds, greatly affects their physicochemical properties such as plasmonic quenching, MEF, and SERS, which depend strongly on the distance between the metallic surfaces and the molecules with specific properties [24,[40][41][42]. Nevertheless, studies on SiO 2 shell's thickness control and optical properties of NPs have been mainly conducted on single metal NPs [35,[43][44][45]. Furthermore, only a few reports have been published on the fine-tunable control of the SiO 2 shell thicknesses of metallic nanocomposite and their effects from the metal surface to fluorophores on fluorescence characteristics.
Ag NP-assembled SiO 2 (SiO 2 @Ag) has been developed by our group as a basic platform for fluorescence, SERS, and magnetism [2,46,47]. The desired absorption wavelength of SiO 2 @Ag nanocomposite can be tuned efficiently by the presence of assembled Ag NPs on the SiO 2 surface [48,49].
In this study, the number of NPs and the amount of SiO 2 precursor were investigated to finely modulate the thickness of SiO 2 shell on the surface of SiO 2 @Ag without structural destruction. In addition, quantum dots were introduced into a SiO 2 shell-coated SiO 2 @Ag nanocomposite, and their optical properties according to the thickness of the shell are reported. This study presents a valuable approach that can provide optimal conditions for fluorescence enhancement through the fine control of the SiO 2 shell thickness of nanocomposites. Our results show that single metal nanoparticles as well as nanocomposites exhibit the presence of MEF phenomena according to the SiO 2 's shell thickness.

Results and Discussion
SiO 2 @Ag as nanocomposites were synthesized as pre-reported [50,51]. Briefly, SiO 2 NPs (153 ± 2.4 nm) were prepared by the modified Stöber method and were incubated with 3-mercaptopropyl trimethoxysilane (MPTS) to convert hydroxyl groups to thiol groups. Ag NPs were assembled on the SiO 2 -SH surface by reducing silver nitrate (AgNO 3 ) in ethylene glycol (EG) with octylamine (OA). The transmission electron microscope (TEM) images of the SiO 2 and SiO 2 @Ag NPs, shown in Figure S1, confirmed their uniform size and shape and excellent dispersion in ethanol without aggregation. The SiO 2 @Ag NPs exhibit their rough surfaces due to the assembly of Ag NPs. The average diameter of SiO 2 @Ag NPs, as measured by ImageJ software, is 188 ± 7.3 nm. The optical properties of the NPs were investigated by using UV-Vis spectrophotometry. Figure S1C shows the UV-Vis absorbance spectra of the SiO 2 and SiO 2 @Ag NPs. The UV-Vis absorbance of SiO 2 NPs decreased rapidly from 300 nm to 1100 nm, corresponding to the absorbance of typical SiO 2 NPs [52]. The UV-Vis absorbance of SiO 2 @Ag exhibits a broad localized surface plasma resonance (LSPR) band from 325 nm to 1100 nm and a maximum UV-Vis absorbance in the wavelength range from 400 to 500 nm, which corresponds to the absorption region of Ag NPs [53]. This result confirmed that Ag NPs were assembled on the SiO 2 surface, and the nanocomposite SiO 2 @Ag could absorb light from the visible to near-infrared region. According to the literature, LSPR from visible light to ultraviolet regions represents highly sensitive wavelengths depending on NP's component, shape, and ambient mediums [54]. Therefore, the optical property of SiO 2 @Ag can be controlled by experimental conditions in our research. The thickness of the SiO 2 shell is highly sensitive to experimental conditions. Therefore, it is difficult to evenly form a perfect silica shell with sub-nanometer thickness on the NPs because excessive reactions for the formation of non-core silica nanostructure such as silica NPs must be excluded. Therefore the SiO 2 coating on the SiO 2 @Ag surface was controlled and applied by two experimental parameters, the amount of SiO 2 @Ag NPs itself or the amount of SiO 2 @Ag NPs combined with the amount of SiO 2 precursor, according to a previously reported method [55]. Sodium silicate (Na 2 SiO 3 ) and tetraethyl orthosilicate (TEOS) were used as SiO 2 precursors. The morphology and structure of SiO 2 @Au@SiO 2 with different silica shell thicknesses are observed by TEM images (Scheme 1A, Figures 1 and S2). The observed silica shell layer on SiO 2 @Ag surface is quite homogeneous and ranges from 4 nm to 38 nm.
ance of SiO2 NPs decreased rapidly from 300 nm to 1100 nm, correspondin ance of typical SiO2 NPs [52]. The UV-Vis absorbance of SiO2@Ag exhibit ized surface plasma resonance (LSPR) band from 325 nm to 1100 nm and a Vis absorbance in the wavelength range from 400 to 500 nm, which corr absorption region of Ag NPs [53]. This result confirmed that Ag NPs wer the SiO2 surface, and the nanocomposite SiO2@Ag could absorb light from near-infrared region. According to the literature, LSPR from visible ligh regions represents highly sensitive wavelengths depending on NP's com and ambient mediums [54]. Therefore, the optical property of SiO2@Ag ca by experimental conditions in our research. The thickness of the SiO2 she sitive to experimental conditions. Therefore, it is difficult to evenly form shell with sub-nanometer thickness on the NPs because excessive reacti mation of non-core silica nanostructure such as silica NPs must be exclu the SiO2 coating on the SiO2@Ag surface was controlled and applied by tw parameters, the amount of SiO2@Ag NPs itself or the amount of SiO2@Ag with the amount of SiO2 precursor, according to a previously reported m dium silicate (Na2SiO3) and tetraethyl orthosilicate (TEOS) were used as S The morphology and structure of SiO2@Au@SiO2 with different silica shell observed by TEM images (Scheme 1A, Figure 1 and S2). The observed silic SiO2@Ag surface is quite homogeneous and ranges from 4 nm to 38 nm.
First, the formation of a silica shell from Na2SiO3 through solvent exchanges is a result of a sharp decrease in the solubility of Na2SiO3 in a mixture of water and ethanol, and controlling this process is difficult. Therefore, a small change of Na2SiO3 can also lead to a significant change in the formation of the silica-shell layer. Therefore, we fixed the amount of Na2SiO3 and changed the amount of SiO2@Ag to prevent the formation of non-core silica nanostructure, such as silica NPs. The amount of SiO2@Ag NPs was adjusted by changing the quantity of SiO2 from 5 mg ( Figures 1A and S1), 10 mg ( Figures 1A and S2), and 20 mg ( Figures 1A and S3), while 14.4 µ L Na2SiO3 was fixed. By controlling the amount of SiO2@Ag NPs, thin silica shells were obtained in the range of 4 to 13 nm (Figures 1A and S1: 4 ± 0.3 nm, Figures 1A and S2: 9 ± 1.1 nm, Figures 1A and S3: 13 ± 1.2 nm) without a leakage of Ag NPs from their surfaces. When the amount of SiO2@Ag decreases, the silica shell becomes is thicker. However, a decrease in SiO2@Ag NPs amounts leads to a low yield of SiO2@Ag@SiO2 products. Thickness of the silica shell layer on the surface of the SiO 2 @Ag NPs synthesized under various conditions. S1, S2, and S3 were synthesized using SiO 2 @Ag (5 mg, 10 mg, and 20 mg, respectively) and Na 2 SiO 3 (14.4 µL). S4, S5, and S6 were synthesized using 10 mg of SiO 2 @Ag (10 mg), Na 2 SiO 3 (14.4 µL), and TEOS (1.2 mM, 2.3 mM, and 4.7 mM, respectively).
First, the formation of a silica shell from Na 2 SiO 3 through solvent exchanges is a result of a sharp decrease in the solubility of Na 2 SiO 3 in a mixture of water and ethanol, and controlling this process is difficult. Therefore, a small change of Na 2 SiO 3 can also lead to a significant change in the formation of the silica-shell layer. Therefore, we fixed the amount of Na 2 SiO 3 and changed the amount of SiO 2 @Ag to prevent the formation of non-core silica nanostructure, such as silica NPs. The amount of SiO 2 @Ag NPs was adjusted by changing the quantity of SiO 2 from 5 mg ( Figures 1A and S1), 10 mg ( Figures 1A and S2), and 20 mg (Figures 1A and S3), while 14.4 µL Na 2 SiO 3 was fixed. By controlling the amount of SiO 2 @Ag NPs, thin silica shells were obtained in the range of 4 to 13 nm (Figures 1A and S1: 4 ± 0.3 nm, Figures 1A and S2: 9 ± 1.1 nm, Figures 1A and S3: 13 ± 1.2 nm) without a leakage of Ag NPs from their surfaces. When the amount of SiO 2 @Ag decreases, the silica shell becomes is thicker. However, a decrease in SiO 2 @Ag NPs amounts leads to a low yield of SiO 2 @Ag@SiO 2 products.
To generate a thicker silica shell on the SiO 2 @Ag surface with an expected product yield, TEOS was added into the S2 suspension to generate the samples Figure 1A, S4, S5, and S6. The formation of silica shell from TEOS occurs slowly and controllably through sol-gel processes. Therefore, we fixed the amount of SiO 2 @Ag and changed the amount of TEOS.
Smooth and thicker silica shells with variable thickness were grown on SiO 2 @Ag NPs by the addition of different concentrations of TEOS at 1.2 mM (Figure 1A S4), 2.3 mM (Figure 1A S5), and 4.7 mM (Figure 1A S6). The thicknesses of silica shells on the SiO 2 @Ag@SiO 2 are 16 ± 1.0 nm ( Figure 1A S4), 24 ± 1.3 nm (Figure 1A S5), and 38 nm ± 2.0 nm (Figure 1A S6). These results demonstrated that the increase in TEOS concentrations leads thicker silica shells on the SiO 2 @Ag surface ( Figure 1B). Moreover, the typical UV-Vis absorbance spectrum of SiO 2 @Ag@SiO 2 was observed in Figure S1C. Similarly to SiO 2 @Ag, the UV-Vis spectra of SiO 2 @Ag@SiO 2 also broaden from 325 to 1100 nm but its absorbance intensity slightly decreased.
The study on the effect of silica shell thickness on fluorescence properties was performed by introducing 7 mg of quantum dots (CdSe@ZnS, QDs, QY 96.2%) on the surface of SiO 2 @Ag@SiO 2 possessing different silica-shell thicknesses to generate various quantumdot-assembled SiO 2 @Ag@SiO 2 (SiO 2 @Ag@SiO 2 @QDs) (Scheme 1B). For the introduction of QDs, the surface of SiO 2 @Ag@SiO 2 was modified to a thiol group by MPTS. The morphology and size of SiO 2 @Ag@SiO 2 @QDs was confirmed by TEM images (Figure 2). As the silica shell became thinner, the introduced QDs were agglomerated and had a rough shape. In contrast, the aggregation decreased and distributed evenly as the silica shell was thicker. Moreover, the size of SiO 2 @Ag@SiO 2 @QDs increased with the introduction of QDs These results show that QDs can be introduced into SiO 2 @Ag@SiO 2 surfaces to generate SiO 2 @Ag@SiO 2 @QDs. When QDs were assembled on the surface of SiO 2 @Ag@SiO 2 , the maximum emission wavelength of QDs was blue shifted from 620 to 617 nm.
To generate a thicker silica shell on the SiO2@Ag surface with an expected product yield, TEOS was added into the S2 suspension to generate the samples Figure 1A, S4, S5, and S6. The formation of silica shell from TEOS occurs slowly and controllably through sol-gel processes. Therefore, we fixed the amount of SiO2@Ag and changed the amount of TEOS. Smooth and thicker silica shells with variable thickness were grown on SiO2@Ag NPs by the addition of different concentrations of TEOS at 1.2 mM (Figure 1A S4), 2.3 mM (Figure 1A S5), and 4.7 mM (Figure 1A S6). The thicknesses of silica shells on the SiO2@Ag@SiO2 are 16 ± 1.0 nm ( Figure 1A S4), 24 ± 1.3 nm (Figure 1A S5), and 38 nm ± 2.0 nm (Figure 1A S6). These results demonstrated that the increase in TEOS concentrations leads thicker silica shells on the SiO2@Ag surface ( Figure 1B). Moreover, the typical UV-Vis absorbance spectrum of SiO2@Ag@SiO2 was observed in Figure S1C. Similarly to SiO2@Ag, the UV-Vis spectra of SiO2@Ag@SiO2 also broaden from 325 to 1100 nm but its absorbance intensity slightly decreased.
The study on the effect of silica shell thickness on fluorescence properties was performed by introducing 7 mg of quantum dots (CdSe@ZnS, QDs, QY 96.2%) on the surface of SiO2@Ag@SiO2 possessing different silica-shell thicknesses to generate various quantum-dot-assembled SiO2@Ag@SiO2 (SiO2@Ag@SiO2@QDs) (Scheme 1B). For the introduction of QDs, the surface of SiO2@Ag@SiO2 was modified to a thiol group by MPTS. The morphology and size of SiO2@Ag@SiO2@QDs was confirmed by TEM images (Figure 2). As the silica shell became thinner, the introduced QDs were agglomerated and had a rough shape. In contrast, the aggregation decreased and distributed evenly as the silica shell was thicker. Moreover, the size of SiO2@Ag@SiO2@QDs increased with the introduction of QDs  . TEM images of SiO2@Ag@SiO2@QDs using SiO2@Ag@SiO2 with various silica shells thicknesses (S1 to S6) to introduce 7 mg QDs on their surfaces to generate the SiO2@Ag@SiO2@QDs corresponding to (QS1-QS6). Figure 3 shows the change of fluorescence intensity by metals according to shell thicknesses. The emission spectra in Figure 3 showed that the fluorescent intensities at 620 nm of QDs significantly increased with silica-shell thicknesses and reached a maximum value at the 14 nm silica shell. Fluorescence intensities tended to decrease as shell thicknesses exceeded 14 nm. To compare the effect of metals on the fluorescence of QDs, SiO2 NPs with similar sizes were introduced by QDs ( Figure S3). When the distance between Figure 2. TEM images of SiO 2 @Ag@SiO 2 @QDs using SiO 2 @Ag@SiO 2 with various silica shells thicknesses (S1 to S6) to introduce 7 mg QDs on their surfaces to generate the SiO 2 @Ag@SiO 2 @QDs corresponding to (QS1-QS6). Figure 3 shows the change of fluorescence intensity by metals according to shell thicknesses. The emission spectra in Figure 3 showed that the fluorescent intensities at 620 nm of QDs significantly increased with silica-shell thicknesses and reached a maximum value at the 14 nm silica shell. Fluorescence intensities tended to decrease as shell thicknesses exceeded 14 nm. To compare the effect of metals on the fluorescence of QDs, SiO 2 NPs with similar sizes were introduced by QDs ( Figure S3). When the distance between the QDs to Ag NPs was less than~9 nm, the fluorescence intensities of SiO 2 @Ag@SiO 2 @QDs were weaker than those of SiO 2 @QDs. However, it was observed that the fluorescence intensities of SiO 2 @Ag@SiO 2 @QDs were stronger when the distance between QDs and Ag NPs was between about 9 nm and 14 nm, and then it weakened again as this distance increased. The increase in fluorescence intensity with the silica shell thickness can be explained by two effects that affect the emission spectrum, which varies with distance [56]: the emission quenching [57] of metal NPs for the photoexcited QDs due to resonant energy transfers and the fluorescence enhancement of NPs promoted by the excitation of their localized surface plasmon resonances [58]. These phenomena, quenching and MEF, occur when the distance between the metal NPs to the fluorophores is within 20 nm [59], but details about the distance have not been revealed. It is important to note that small changes between metal NPs and fluorophores particularly affect metal emission properties. The only variable influencing is the plasmonic electromagnetic field decay exponential with the distance to the metal surface. Quenching is a phenomenon in which Förster resonance energy transfer (FRET) occurs when the distance between the metal NPs and the fluorophore is less than 10 nm. Therefore, it is considered that the fluorescence intensity of SiO 2 @Ag@SiO 2 @QD is weaker than that of SiO 2 @QD due to the quenching process caused by a distance between Ag NPs and QDs that is less than 10 nm [60]. Moreover, it is believed that the resonance excitation of LSPR on Ag NPs on the SiO 2 surface generates an enhanced local field when the distance between Ag NPs and QDs is 9-14 nm, which greatly increases the fluorescence intensity of QDs. Therefore, when the distance between the QDs and the Ag NPs increases within 14 nm, it is estimated that the fluorescence signal is strengthened due to the increase in the MEF and a decrease in the quenching effect.
the QDs to Ag NPs was less than ~9 nm, the fluorescence intensities of SiO2@Ag@SiO2@QDs were weaker than those of SiO2@QDs. However, it was observed that the fluorescence intensities of SiO2@Ag@SiO2@QDs were stronger when the distance between QDs and Ag NPs was between about 9 nm and 14 nm, and then it weakened again as this distance increased. The increase in fluorescence intensity with the silica shell thickness can be explained by two effects that affect the emission spectrum, which varies with distance [56]: the emission quenching [57] of metal NPs for the photoexcited QDs due to resonant energy transfers and the fluorescence enhancement of NPs promoted by the excitation of their localized surface plasmon resonances [58]. These phenomena, quenching and MEF, occur when the distance between the metal NPs to the fluorophores is within 20 nm [59], but details about the distance have not been revealed. It is important to note that small changes between metal NPs and fluorophores particularly affect metal emission properties. The only variable influencing is the plasmonic electromagnetic field decay exponential with the distance to the metal surface. Quenching is a phenomenon in which Förster resonance energy transfer (FRET) occurs when the distance between the metal NPs and the fluorophore is less than 10 nm. Therefore, it is considered that the fluorescence intensity of SiO2@Ag@SiO2@QD is weaker than that of SiO2@QD due to the quenching process caused by a distance between Ag NPs and QDs that is less than 10 nm [60]. Moreover, it is believed that the resonance excitation of LSPR on Ag NPs on the SiO2 surface generates an enhanced local field when the distance between Ag NPs and QDs is 9-14 nm, which greatly increases the fluorescence intensity of QDs. Therefore, when the distance between the QDs and the Ag NPs increases within 14 nm, it is estimated that the fluorescence signal is strengthened due to the increase in the MEF and a decrease in the quenching effect.

Preparation of Ag-Assembled Silica NPs (SiO 2 @Ag)
SiO 2 NPs (150 nm) were prepared using the modified Stöber method [61]. The surface of SiO 2 NPs was converted to thiol groups by incubating SiO 2 NPs (4 mL, 50 mg·mL −1 suspension in EtOH), MPTS (200 µL) and NH 4 OH solutions (27%, 40 µL). The suspension was stirred at 700 rpm at 25 • C for 12 h. Next, the suspension was centrifuged at 8500 rpm and washed several times with EtOH. The thiolated SiO 2 NPs (SiO 2 -SH) was dispersed in EtOH and the final concentration was adjusted to 50 mg·mL −1 . Ag NPs were added on the surface of the thiolated SiO 2 NPs by a reduction of AgNO3 in the presence of PVP. The SiO 2 -SH NPs (0.6 mL, 50 mg·mL −1 in EtOH) were added in an EG solution containing PVP (5 mg), AgNO 3 (26 mg), and octylamine (41.4 µL). The suspension was stirred at 700 rpm at 25 • C for 1 h. Then, the suspension was centrifuged at 8500 rpm and washed several times with EtOH. SiO 2 @Ag was dispersed in EtOH.

Preparation of Silica-Shell-Coated SiO 2 @Ag with Various Thicknesses (SiO 2 @Ag@SiO 2 )
Various amounts of the SiO 2 @Ag (5, 10, and 20 mg) were separately dispersed in EtOH (1 mL) to prepare samples S1, S2, and S3. Distilled water (15 mL) containing 14.4 µL Na 2 SiO 3 was mixed with the above SiO 2 @Ag suspension. The prepared suspension was stirred at 700 rpm for 1 h. A 60 mL aliquot of EtOH was then added to the resulting suspensions for solvent exchange. After 3h, parts of sample S2 containing 10 mg of SiO 2 @Ag were mixed with different concentrations of TEOS (1.2 mM, 2.3 mM, and 4.7 mM) under stirring at RT for 24 h to prepare samples S4, S5, and S6. After stirring, the above suspension was centrifuged at 8500 rpm and washed with EtOH to remove excess reagents.

Physical Property Analysis of NPs
The size and morphology of prepared NPs were measured by transmission electron microscope (Libra 120, Carl Zeiss, Jena, Germany). The prepared NPs were well dispersed in EtOH (1 mg·mL −1 ). Then, 10 µL of the sample was dropped and dried at 25 • C on a 400-mesh copper grid (Pelco, Presno, CA, USA). The thicknesses of the SiO 2 shell were analyzed by digitalized measurements using Image J software (v.1.53k, Bethesda, MD, USA). The average size of the NPs and the thickness of the SiO 2 shell were calculated after analyzing at least 50 NPs.

Measurement of UV-Vis Absorption Spectra
The particles were well dispersed in EtOH to obtain a suspension of 2 mg·mL −1 nanoparticle and transferred to a cuvette. UV-Vis absorption of the sample was performed in the wavelength from 300 to 1100 nm at the scanning speed of 1 or 5 nm/s by using a UV-Vis spectrophotometer (Mecasys OPTIZEN POP, Daejeon, Korea) 3.2.6. Fluorescence Analysis of SiO 2 @Ag@SiO 2 @QDs The fluorescence emission spectrum was analyzed by using a Cary Eclipse Fluorescence Spectrophotometer (Agilent Technologies, Santa Clara, CA, USA). An NP suspension (1 mg·mL −1 ) measuring 300 µL was added in a 96-well plate. The excitation wavelength was set at 385 nm. The sample was excited for 10 s and the fluorescence of the sample was collected in the range from 550 to 700 nm for 10 s.

Conclusions
A fascinating approach to silica-shell coatings on SiO 2 @Ag, one of the nanocomposites for multilayer synthesis, allowed us to control silica-shell thicknesses in a wide range from 4 nm to 38 nm. The thickness of the silica shell was finely controlled with two methods: controlling the amount of SiO 2 @Ag itself or combining the amount of SiO 2 @Ag and the silica precursor. Silica-shell thicknesses on the SiO 2 @Ag surface were measured to be 4 ± 0.3, 9 ± 1.1, 13 ± 1.2, 16 ± 1.0, 24 ± 1.3, and 38 ± 2.0 nm. As a result of testing, the effects of metal on the fluorescence according to the distance using the difference in the silica-shell thickness, the quenching effect, and MEF were observed with at a distance between Ag NPs to QDs within about 14 nm. When the distance between Ag NPs to QDs was less than~10 nm, SiO 2 @Ag@SiO 2 @QDs showed weaker fluorescence intensities than SiO 2 @QD (without metal) due to the quenching effect. On the other hand, when the distance between Ag NPs to QDs was from 10 nm to 14 nm, the fluorescence intensity of SiO 2 @Ag@SiO 2 @QD was stronger than SiO 2 @QDs due to MEF. These results are expected to be useful for synthesizing multilayer nanocomposites with optimized SERS and MEF effects by the fine control technology of silica shells.