Synthesis and Characterization of Au@MnO2 Nanoparticles as Plasmon Enhanced Spectroscopy Substrates

The plasmonic properties of Au nanoparticles (AuNP), which allow the observation of enhanced spectroscopic effects, are strongly affected by the aggregation and precipitation caused by the strong interactions between nanoparticles. To avoid AuNP aggregation and precipitation, the present study proposes coating with MnO2, forming AuNP@MnO2 core-shell structures. The MnO2 layers presented 1-10 nm thickness so that highly surface-enhanced fluorescence was obtained with maximum intensity given by 5 nm thick MnO2. The decrease in Raman intensity could be controlled, despite the inherent reduction in surface-enhanced Raman scattering (SERS) intensity with increasing adsorbate-surface distance. The decrease in Raman intensity was compensated by increasing AuNP stability caused by the MnO2 shell.


Introduction
The localized surface plasmon resonance (LSPR) may be characterized by the excitation of collective electron oscillation on metallic nanostructure surfaces by an exciting resonant radiation.LSPR occurs in the visible or nearinfrared region for specific metals, such as Ag, Au, Cu, and alkaline metals. 1 The frequency of the LSPR strongly depends on the size and shape of plasmonic nanostructures and the dielectric environment in which the plasmonic nanostructures are immersed. 2pecifically, when the refractive index increases in the vicinity of the nanoparticles, there is a displacement of the LSPR band to lower energies, usually referred to as a red shift; additionally, variations in the intensity and bandwidth may also occur. 3,4Following the changes in the LSPR band position during a deposition process on plasmonic nanostructure is, thus, an efficient way of evaluating surface changes in those nanostructures. 5he collective electron excitation in LSPR significantly increases the local electromagnetic field in the immediate surroundings of plasmonic metal nanostructures.This enhancement of the electric field mainly allows the occurrence of plasmon-enhanced spectroscopic phenomena such as surface-enhanced Raman scattering (SERS) and surface-enhanced fluorescence (SEF). 6,7][15] One major issue for plasmon-enhanced spectroscopies is the observation that nanoparticles of coinage metals (Au, Ag and Cu) interact strongly with each other and may form aggregates in suspension, which strongly affects both suspension stability and spectroscopic characteristics reproducibility.In this context, the shellisolated nanoparticle (SHIN) substrates emerged aiming to minimize the aggregation effects. 16SHIN substrates are comprised of nanoparticles covered by oxides, usually dielectric or semiconductors, such as silicon oxide; these core-shell nanostructures may preserve most of the plasmonic properties of the coinage metals and present the surface properties of the shell oxide.The initial proposition of SHIN resulted in increased interest in the importance of core-shell nanostructures for application in plasmon-enhanced spectroscopies in the last years due to the possibility of combining varied materials in core and shell and obtaining many unique properties. 17he derived techniques SHINERS (shell-isolated nanoparticles enhanced Raman spectroscopy) and SHINEF (shell-isolated nanoparticles enhanced fluorescence) thus, arose to obtain enhanced spectroscopy substrates with increased stability and presented the additional characteristics of avoiding direct contact between the adsorbate and the metallic surface. 18Both techniques have their performance modulated by the thickness of the dielectric layer, which separates the metallic surface of the molecular probe. 19For SHINERS, the increase in the thickness of the dielectric layer causes a decrease in Raman intensity of the molecular probe due to the increasing distance between the probe and the surface, so there is usually a trade-off between improved stability and SHINERS performance.
In SEF, the direct contact between fluorophore and metal suppresses fluorescence due to coupling with surface plasmon, resulting in energy flow from fluorophore to metal competing with fluorescence. 20This occurrence results, for SHINEF and thinner layers, in an increase in fluorescence intensity with thickening of the dielectric layer, which is the result of decreasing the suppression caused by the metal surface conjugated with the LSPR-enhanced local electric field.The increase in fluorescence intensity tends to reach a maximum value for an optimal thickness because both the decrease in suppression and the LSPR enhanced field become less important for thicker layers. 21or both SHINERS and SHINEF methodologies using a suspension of core-shell nanoparticles and the molecular probe, the signal enhancement may be evaluated by the AEF (analytical enhancement factor), equation 1. AEF is the simplest way of calculating enhancement factors because it is a simple relationship between the intensities in the enhanced spectroscopies (both SERS/SHINERS and SEF/SHINEF) and the non-surface-enhanced spectra at some given concentration, with the experimental care of keeping the spectral conditions constant. 22AEF is well known to underestimate enhancement factors, as the relationship cannot account for the molecules that are not close enough to the surface to present an enhancement of spectroscopic events. 22,23Still, the method allows grasping the enhancement factor trend during materials optimization. (1) In equation 1, I SE is the intensity in the surface-enhanced (SE) spectra (SERS, SHINERS, SEF, or SHINEF), C SE is the concentration of the probe molecule used to obtain the SE spectra, I NE is the non-enhanced spectra (either Raman or fluorescence) of the molecular probe and C NE is the concentration of the molecular probe used to obtain the non-enhanced spectra.
Among the oxides that have been used to develop SHIN materials, manganese dioxide, MnO 2 , can be used as the shell layer of Au nanoparticles core plasmonic material, as it has been shown by Lin et al. 24 The results of the authors indicated that the MnO 2 thickness might be controlled by the initial concentration of potassium permanganate used as the manganese source.MnO 2 -covered Au nanoparticles (Au@MnO 2 ) kept their plasmonic absorption with a red shift proportional to MnO 2 layer thickness.MnO 2 is a semiconductor material that has low cost, low toxicity, and abundant storage, and is insoluble in strongly alkaline medium, differently from other usual SHIN layers such as SiO 2 and Al 2 O 3 . 24,25It has recently been shown that gold nanorods may be modified by manganese oxide, resulting in compact layers of the oxide, which could be used as SHINEF and SHINERS substrates. 26u@MnO 2 may present a relevant potential for applications, such as catalysis and fluorescence sensing.Recently, Zhou et al. 27 used the catalytic reduction of 4-nitrophenol to aniline for the degradation of methylene blue and fluorescence sensing of glutathione.Ag@MnO 2 was used by Abdulrahman et al. 28 to detect methyl parathion on the surface of an orange fruit, concluding that nanoparticles were stable in an alkaline medium and that they were efficient SHINERS substrates.
In the present study, MnO 2 -covered Au nanoparticles (Au@MnO 2 ) were obtained by controlling the oxide shell thicknesses over AuNPs, avoiding reagent excess; the manganese oxide coverage was characterized using LSPR, transmission electron microscopy (TEM), energydispersive X-ray spectroscopy (EDS), and scanning transmission electron microscopy high-angle annular darkfield (STEM-HAADF) chemical imaging.The resulting Au@MnO 2 nanoparticles were applied as substrates for two surface-enhanced spectroscopies, SHINERS (λ 0 = 633 nm) and SHINEF (λ 0 = 785 nm), by choosing the best layer thickness of each technique, using the same probe molecule, the cyanine dye IR-820.

Synthesis of AuNPs
The gold nanoparticle suspensions (AuNP) have been synthesized from the precursor HAuCl 4 in an aqueous solution using Na 3 -cit as a reducing agent following the well-known procedure proposed by Frens. 29Briefly, 50 mL of an aqueous 0.01% m/v HAuCl 4 solution were brought to boiling under reflux, followed by the addition of 500 or 1000 μL of a 1.0% m/v Na 3 -cit solution.The boiling was maintained for about 5-10 min until the suspension acquired the characteristic red color.

Synthesis of Au@MnO 2
The AuNP presented 16 and 50 nm as average diameters.For AuNP (50 nm), the procedure was described by Lin et al. 24 for the MnO 2 thickness of 2.5 nm.For the 16 nm AuNP the procedure was adapted from Lin et al., 24 in which, initially, the amount of KMnO 4 was calculated to achieve the desired manganese dioxide layer thickness without precursor excess, considering the AuNP as spheres with an average diameter of 16 nm.The amount of KMnO 4 for each of the intended thicknesses is presented in Table 1.The values in Table 1 were obtained supposing that HAuCl 4 was converted entirely to AuNP, which is reasonable for the method described by Frens, 29 and that the AuNP were spherical with a diameter of 16 nm; from those considerations, and taking the gold density at room temperature (19.32 g cm -3 ), the mass of a 16 nm radius spherical AuNP was calculated.The numerical concentration of nanoparticles and the volume of one AuNP was estimated.Subsequently, the radius of the MnO 2 shell was added to that of AuNP and, considering the volume of MnO 2 to form the overlayer and the density of MnO 2 (5.0 g cm -3 ), the amount of MnO 2 necessary to cover each AuNP was determined from the numerical concentration of AuNP and, from that, the amount of KMnO 4 in the entries of Table 1.This procedure aimed to avoid that excess permanganate would still be present in the SHIN suspension.Oxalate ion reduced manganese(VII), so the concentration of oxalate was kept as a fifth of the permanganate concentration.To 10.0 mL of AuNP colloids, KOH (10.0 mmol L -1 ) was added to adjust the pH to 9.5-10.0.After the pH adjustment, in an ice bath, one of the volumes presented in Table 1 of a 10.0 mmol L -1 KMnO 4 solution was added, followed by the corresponding volume of 10.0 mmol L -1 Na 2 C 2 O 4 , to obtain a certain nominal MnO 2 thickness.After stirring for 10 min, the resulting suspension was introduced in a thermostatic bath at 60.0 °C for 2 h.After this time, the suspension changed from bright red to light purple in different degrees depending on the amount of KMnO 4 added.

Characterization of Au@MnO 2
The LSPR band of the Au@MnO 2 nanoparticles was measured in an Ocean Optics (Orlando, USA) USB2000 + XR1 + ES spectrometer with a light source in the 200-1800 nm range, using a 5 mm path-length quartz cell.The transmission electron microscopy (TEM) technique was used to characterize the size and shape of AuNP and Au@MnO 2 .Additionally, STEM-HAADF experiments were performed to obtain a qualitative analysis of the elemental composition of the nanostructures.The electron microscopy experiments with energy dispersive spectroscopy (EDS) chemical analyses were performed in a transmission electron microscope FEI (Hillsboro, USA) Titan 80-300 using a 300 kV acceleration voltage.

SHINERS and SHINEF spectroscopic measurements
The SHINEF and SHINERS spectra from the probe molecule IR-820 were recorded on a Bruker (Billerica, USA) SENTERRA Raman spectrometer coupled to an Olympus BX51 optic microscope.A 50× objective (numerical aperture = 0.50) was used to acquire the Raman spectra.The experimental conditions for the SHINEF spectra were as follows: λ 0 = 785 nm, laser power of 25 mW, and 60 s for acquisition; for the SHINERS spectra, λ 0 = 633 nm, laser power of 10 mW, and 60 s spectral acquisition.

Results and Discussion
Characterization of Au@MnO 2 The AuNP and Au@MnO 2 nanoparticles (using AuNPs 16 nm synthesized with 1000 μL of sodium citrate) were characterized by their LSPR position using UV-Vis spectroscopy.Figure 1 presents the extinction spectra of AuNP and the extinction spectra after reaction with different amounts of KMnO 4 , which were expected to result in the deposition of MnO 2 ultrathin shells.One can observe in Figure 1 a red shift of the AuNP LSPR band as the amount of permanganate added to the AuNP suspension increased.
The shift was of around 1 nm for the Au@MnO 2 nanoparticles with a ca.0.8 nm thick manganese oxide overlayer, and it was as high as 57 nm for ca. 10 nm oxide, which may also be due to the formation of clusters for higher permanganate concentrations.The shift of the LSPR band was an evidence of the formation of a MnO 2 overlayer on the AuNP, resulting in Au@MnO 2 and the increase in the MnO 2 shell thickness with the increase in MnO 4 available for reaction.Using UV-Vis spectroscopy as a diagnostic tool, it allowed quick monitoring of LSPR changes in Au@MnO 2 .Using careful experimental protocols, the technique revealed a quite reproducible tendency in the LSPR spectra with an increasing MnO 2 overlayer thickness.
TEM was used to ascertain the MnO 2 overlayer formation and morphology of Au@MnO 2 nanoparticles.TEM micrographs of Au@MnO 2 using AuNPs (50 nm) synthesized with 500 μL of sodium citrate are shown in Figure 2. It can be observed in Figure 2a that the AuNP, which was characterized by the sharp contrast to the background, is covered by a thin layer, indicated by a more diffuse contrast, assigned to the presence of manganese oxide.The high-resolution TEM image in Figure 2b presents the structure, where one may measure an interatomic distance of 2.3 Å (see zoom-in in Figure 2c), which is quite close to the interatomic distance in Au(111). 30n the other hand, the overlayer has no different interlayer distance having an amorphous structure, which may be a result of the deposition procedure.The observation of this overlayer is indicative of effective coverage of the Au nanoparticles by the oxide. 19EDS analysis gave further evidence that the overlayer on the AuNP is composed of manganese oxide, as shown in Figure 2d.
Finally, as additional evidence for the core-shell structure, STEM/HAADF images were acquired for the material and are presented in Figures 2e-2g.One may notice that the composite image in Figure 2e and chemical images for Au and Mn (based on the spectra in Figure 2d) are presented in Figures 2f and 2g, respectively.The lateral resolution of the image is ca.1.4 nm.The area occupied by the Au signal (Figure 2e) is smaller than the morphological image of the whole particle.On the other hand, the mapping for Mn (Figure 2g) indicates the presence of this element in an area similar to the composite image (Figure 2e), which suggests that there is an overlayer on the AuNP constituted of manganese oxide.It is also possible to estimate the thickness  of the manganese oxide layer based on the HAADF image as ca.2.8 nm, evidenced by the difference between the Au and Mn images.The above result adds evidence to the formation of a core-shell structure Au@MnO 2 .

SHINEF
The Au@MnO 2 materials prepared are potential substrates for SHINERS and SHINEF applications.In the present study, the indocyanine dye IR-820 was employed as a probe molecule because it presents pre-resonance Raman at 633 nm, which allows evaluating the SHINERS performance and using it as a fluorescence probe when excited at 785 nm. 26,31The performance of the Au@MnO 2 nanoparticles (using AuNPs 16 nm synthesized with 1000 μL of sodium citrate) was evaluated for both SHINERS and SHINEF by the analytical enhancement factor AEF proposed by Le Ru et al. 22

(equation 1).
Fluorescence is expected to be suppressed in metallic surfaces such as the AuNP, which is an opposite behavior compared to the Raman scattering of molecules adsorbed on plasmonic nanostructures. 18The fluorescence suppression in the presence of metallic surfaces (like AuNPs) is due to the energy transfer from the excited states of the fluorophore to the metallic structure so the radiative decay strongly loses intensity. 32,33This way, for the occurrence of plasmonic enhancement of fluorescence, resulting in SEF, it is essential that there is a spatial gap between the metal and the fluorophore to decrease the probability of non-radiative energy transfer.On the other hand, the gap thickness has to be carefully controlled, so that there is a balance between the necessary distance fluorophore-surface but there still local field enhancement by the plasmonic excitation of the AuNP to increase the fluorescence intensity.The distance range most usually verified as the best for SEF in the literature is 5 to 10 nm. 18,20,21n the present study, the fluorescence spectra of IR-820 1.0 × 10 -5 mol L -1 excited at 785 nm were obtained from the dye in contact with either AuNP or Au@MnO 2 with nominal thicknesses of 2.5, 5.0 and 10 nm, in the pursue of the best thickness for optimizing the SHINEF performance.Figure 3 presents the fluorescence spectrum of IR-820 and SEF/SHINEF spectra of IR-820 in the presence of AuNPs and AuNPs@MnO 2 .
The fluorescence spectrum of IR-820 in direct contact with AuNPs in Figure 3 presented a sharp decrease in fluorescence intensity compared to the IR-820 solution.As can be observed in Table 2, the calculated AEF indicates 10 times decrease in intensity.Additionally, one may observe an increase in the IR-820 fluorescence intensity as the MnO 2 layer thickness increases from 2.5 to 5 nm, followed by an intensity decrease for the 10 nm thick MnO 2 layer; this result is translated into AEF in the entries of Table 2.
In addition to the intensity variation, differences were also observed in the fluorescence spectral profile.The observed changes, which were both in maximum position and line width, may be result of the changes in the medium around the IR-820 fluorophore.The medium changes could interfere with those spectral characteristics due to the complexity of the plasmon-enhancement phenomenon. 20However, some factors are usually the most important in the spectral profile of cyanine dyes like IR-820 and can strongly interfere in the spectral profile, such as the formation of dimers in higher concentrations, the adsorption geometry, and intensity of interaction between the dye and the MnO 2 surface. 34o the above three factors may be attributed most of the variations observed in IR-820 SHINEF spectra, as it has been previously observed for IR-820 adsorbed on Ag@SiO 2 . 31

SHINERRS
Figure 4 displays the SHINERS spectra of IR-820 1.0 × 10 -5 mol L -1 on Au@MnO 2 (using AuNPs 16 nm  synthesized with 1000 μL of sodium citrate) with several thicknesses of the manganese oxide overlayer and the Raman spectrum of an IR-820 solution in the same concentration, excited at 633 nm.It would be interesting to reinforce that the IR-820 dye is in pre-resonance with the 633 nm laser line so that the spectra obtained could be better described as 'shell-isolated nanoparticles enhanced resonance Raman scattering' (SHINERRS).The SHINERRS spectra obtained in the present study agree very well with the spectra of previous studies on the IR-820 dye, 26,31,35 which were interpreted as an indication of adsorption of the dye with the aromatic ring perpendicular to the surface and low interaction of the sulfonate groups with the metallic surface.
It is also worth mentioning that Au@MnO 2 presented higher stability, both chemical and temporal when compared to non-protected AuNP.However, a trend that may be noticed in Figure 4 is that the Raman intensity of the IR-820 dye decreased with the MnO 2 overlayer thickness, which is a consequence of the increasing distance between the plasmonic nanoparticles and probe molecule.One may notice that there is no substantial decrease in the SHINERS intensity as the oxide layer thickens, as seen in Table 3.The SHINERS intensity decreases more markedly from 1.2 to 2.0 nm, but thicker layers present a fluctuation in SHINERS intensity.That may be attributed to the pre-resonant properties of the probe dye IR-820 excited at 633 nm.
The AEF for the SHINERRS spectra were calculated according to equation 1 and are presented in Table 3, using the band at 1520 cm -1 .The decrease in the Raman intensity may be explained by the decrease of the electromagnetic field by the fourth power of the distance between the Raman molecular probe and the plasmonic surface. 36However, it should be noticed that the advantages of using a surface with different properties compared to gold may turn out to be advantageous depending on the application, even with the decrease in performance of the Raman enhancement.That would be the case in samples with a complex environment, such as dissolved salts or molecules, which could cause aggregation or coagulation of unprotected nanoparticles.
The SHINERRS results evidence that the increasing distance between the adsorbate and the Au core strongly affects the scattering intensity.However, one may still observe a considerable enhancement even for the 10 nm MnO 2 layer.It should be reinforced that this study used moderated concentration, which increases the underestimation of the enhancement factor by the AEF method.

Conclusions
Gold nanoparticles coated with different manganese dioxide ultrathin overlayers were synthesized, varying the starting amount of manganese precursor permanganate to ensure that no excess of the precursors was present in the final product.The Au@MnO 2 materials were characterized by UV-Vis, high-resolution TEM, EDS, and STEM-HAADF techniques.The red shift of the AuNP LSPR band with the increasing amount of MnO 2 was a strong indication of the oxide coating and a qualitative evidence of the growing thickness of the shell.The layer with MnO 2 was confirmed by high-resolution TEM, EDS, and STEM-HAADF techniques.
Au@MnO 2 were used as substrates in SHINERS and SHINEF spectroscopies, using the probe molecule IR-820 dye, and calculating the AEF to compare different shell thicknesses.For SHINERS, the nanoparticles with a 1.2 nm layer presented the best performance, while for SHINEF nanoparticles with a 5.0 nm layer presented the best performance.

Figure 1 .
Figure 1.UV-Vis spectra of Au@MnO 2 core-shell nanoparticles for different shell thicknesses.All spectra have been normalized for the extinction at the top of the LSPR band.

Table 2 .
SHINEF AEF for four different bands of the IR-820 spectra on Au@MnO 2 as a function of the nominal thickness of the manganese oxide overlayer, calculated according to equation 1, for the spectra excited at shell-isolated nanoparticles enhanced fluorescence analytical enhancement factor.

Table 3 .
SHINERS AEF for four different bands of the IR-820 spectra on Au@MnO 2 as a function of the thickness of the manganese oxide overlayer, calculated according to equation 1, for the spectra excited at 633 nm υ / cm -AEF: shell-isolated nanoparticles enhanced Raman spectroscopy analytical enhancement factor.

Table 1 .
Volumes of 10 mmol L -1 KMnO 4 and Na 2 C 2 O 4 solutions used for achieving different nominal thicknesses of MnO 2 over the Au nanoparticles(16 nm)