Au Nanoparticles Coated ZnO Film for Chemical Sensing by PIERS Coupled to SERS

Abstract

In this work, we present a study on the sensing of chemical molecules with Au nanoparticles coated ZnO film by employing the photo-induced enhanced Raman scattering (PIERS) coupled to the surface-enhanced Raman scattering (SERS). Moreover, the interest of coupling the PIERS to classical SERS (electromagnetic contribution) is to achieve a high sensitivity of detection. In our work, we employed the thiophenol molecule for its quality of adsorption on metals and its small size. For this thiophenol detection, we found a limit concentration of 100 pM, an adsorption capacity ($K_{Ads}$) of 3.2 × 10${}^6$ M${}^{-1}$ and an analytical enhancement factor of 1.9 × 10${}^{10}$.

1. Introduction

In the last 10 years, the realization of surface-enhanced Raman scattering (SERS) substrates, having considerable factors of Raman enhancement and a weak cost of production, are highly fundamental points for the detection of biological and chemical molecules [1,2,3]. Nowadays, several fabrication techniques (colloidal synthesis [4,5,6] and lithographies [7,8,9]) are usable for the realization of any type of plasmonic nanosystems having various geometries and shapes. These techniques have allowed to obtain plasmonic nanosystems inducing huge Raman enhancements, nevertheless, a couple of these techniques are too costly to be potentially transferable to industry as Deep UV and UV lithographies [10,11,12], and Electron Beam Lithography (EBL) [13,14,15]. Other techniques are employed for their low-cost of fabrication: colloidal synthesis [16,17,18], nanoimprint lithography [19,20,21] (NIL), and nanosphere lithography [22,23,24] (NSL). Concerning to NIL and NSL, certain limits subsist to achieve an excellent definition of plasmonic nanosystems on huge surfaces requested for practical applications of SERS substrates. In addition, several plasmonic nanosystems have been already considered in order to obtain large enhancement factors (EF) for SERS [25,26,27,28,29]. In addition, significant EFs can be also realized by inserting a metallic layer under the plasmonic nanosystems [30,31,32,33]. Moreover, another way is to use semiconducting nanostructures (e.g., silicon, zinc oxide) combined to plasmonic nanoparticles being able, thus, to improve the detection limit [34,35,36,37,38,39,40]. Furthermore, a supplementary way for improving the Raman enhancement has arisen which is the photo-induced enhanced Raman scattering (PIERS). The PIERS effect enables an additional chemical enhancement of the Raman signal compared to standard SERS by the intermediate of an UV-irradiated semiconducting substrate [41,42,43,44,45,46]. This is explained by the generation of surface oxygen vacancies during the UV-irradiation of the semiconducting substrate which allow improving the Raman signal of studied molecules by the intermediate of charge transfer processes between the UV-irradiated substrate and the molecules via plasmonic (metallic) nanoparticles dropped on this same substrate irradiated by UV [41,42,43,44,45,46], or by promoting back-transfer charges to metallic nanoparticles [47]. A few groups have already employed the PIERS effect for sensitive detection of various molecules [45,46,48,49]. However, in all these works [45,48,49], the excitation wavelength for their Raman measurements is relatively far from the plasmon resonance of their nanostructures benefiting no significant Raman enhancement due to an electromagnetic contribution.

Our investigation is aimed to reveal a highly sensitive detection of thiophenol by using photo-activated ZnO film coated with gold nanoparticles. Thanks to these substrates, we will thus produce photo-activated plasmonic nanosystems at a low-cost of production, and a sensitive detection of thiophenol molecules by using the PIERS effect combined to SERS (electromagnetic contribution).

2. Experimental Details

2.1. Production and Structural Characterization of ZnO Film and Au Nanoparticles

The zinc oxide films were produced on c-plane sapphire (c-Al${}_2$O${}_3$) substrates by using the technique of pulsed laser deposition with a 99.9995% pure sintered ZnO target and a KrF excimer laser. Next, an annealing at 600 °C in O${}_2$ was performed. Regarding to the production of Au nanoparticles (AuNPs), they were synthesized with the Turkevich method. We added a volume of 1 mL of trisodium citrate (Na${}_3$C${}_6$H${}_5$O${}_7$ with a concentration of 8.5 × 10${}^{-4}$ M) to a boiling aqueous solution of 20 mL of HAuCl${}_4$ under vigorous stirring over 30 min. The average size of AuNPs is 30 nm for the diameter. Regarding to the structural characterization of the samples, we employed a Siemens D5000 XRD system (Siemens, Erlangen, Germany) in reflection mode and a quartz monochromator (CuK${\alpha }_1$ = 1.54056 Å) in order to acquire a X-ray diffraction (XRD) pattern.

2.2. Functionalization of Au Nanoparticles Coated ZnO Films with Thiophenol Molecules

To investigate the PIERS chemical sensing coupled to SERS using the Au nanoparticles coated ZnO films, we employed the thiophenol molecules for its quality of adsorption on the metal surface [50,51]. The first step was to realize thiophenol solutions (thiophenol + ethanol) with concentrations from 100 pM to 1 mM. Then, the substrates were immersed in the thiophenol solution for a duration of 24 h. Hereafter, the substrates were thoroughly dried by employing a nitrogen gun. In addition, we used a thiophenol concentration of 1 M and the same protocol as used previously for the reference experiment.

2.3. Extinction/Raman Measurements and Irradiation with UV Light

For the extinction measurements, a spectrometer (Labram, Horiba Scientific, Kyoto, Japan) in standard transmission geometry with unpolarized light was used. The transmitted light was collected by using a microscope objective (×10, N.A. = 0.25). For the Raman spectroscopy, the same spectrometer was employed whose the spectral resolution is 1 cm${}^{-1}$ and with an excitation wavelength of 532 nm (power = 2.4 mW). We fixed the acquisition time to 5 s for these Raman measurements. We focused the excitation laser on the samples by using a microscope objective (×100, N.A. = 0.9). Hereafter, this microscope objective has collected the Raman signal in a backscattering configuration. Additionally, for the reference experiment (the ZnO film without any Au nanoparticles), the same conditions of excitation were employed. Regarding to the UV-irradiation of the sample, an UV pencil lamp (Edmund Optics, $\lambda$ = 254 nm with a nominal output of 4.5 mW/cm${}^2$) was used and irradiated the sample under normal incidence. The distance was fixed at 2.4 cm between the sample and the lamp.

3. Results and Discussion

Regarding our investigation, we produced a ZnO film whose the thickness was 200 nm on sapphire (c-Al${}_2$O${}_3$) substrate by pulsed laser deposition. Afterwards, we drop-casted on the ZnO film Au nanoparticles (AuNPs) whose the diameter was 30 nm, and they were air dried (see Figure 1). Then, the composition of substrates was confirmed by recording an X-ray diffraction (XRD) pattern (see Figure 1) in which two characteristic peaks of ZnO (wurtzite; planes (002) and (110)) and also two peaks for gold (planes (111) and (220)) were pointed. A last diffraction peak (plane (006)) was also pointed corresponding to the sapphire substrate.

Next, we recorded extinction spectra of Au nanoparticles on ZnO film without UV-irradiation and with UV-irradiation times of 15 min and 30 min in order to determine the position of their plasmon resonance (see Figure 2a). We observed a blueshift of 20 nm and 36 nm for the plasmonic resonance of Au nanoparticles when the samples were irradiated with UV during 15 min and 30 min, respectively. This blueshift in the plasmonic resonance upon UV-irradiation is explained by an increase in electron density due to a charge transfer mechanism induced by light from ZnO film to gold nanoparticles. This increase in electron density can be evaluated as follows [45,46,52]:

$$\begin{array}{c}\hfill \frac{\Delta N}{N}=\frac{-2\Delta \lambda }{{\lambda }_{plasmon}}\end{array}$$

(1)

where $\Delta \lambda$ and ${\lambda }_{plasmon}$ represent the shift of the plasmonic resonance and the position of the plasmonic resonance before UV-irradiation, respectively. Thereby, the increase in electron density ($\Delta N/N$) is 12.9% for the UV-irradiation time of 30 min.

For the assessment of the detection performances of the Au nanoparticles coated ZnO films, thiophenol molecules were grafted on these plasmonic ZnO films by employing the procedure depicted in the Section 2.2. Subsequently, we realized Raman spectra of thiophenol on plasmonic ZnO films for the excitation wavelength of 532 nm (see Figure 2b and Figure 3a,b) for each concentration of thiophenol. From the Figure 2b and Figure 3a,b, four Raman peaks of thiophenol molecules [53,54] are well-present (see Table 1 below):

Table 1. Vibration modes of thiophenol.

Vibration Modes	Name	Values (cm−1)
C–H out-of-plane bending, ring out-of-plane deformation	$\gamma$(CH), r–o–d	999
Ring in-plane deformation, C–C symmetric stretching	r–i–d, $\nu$(CC)	1022
C–C symmetric stretching, C–S stretching	$\nu$(CC), $\nu$(CS)	1073
C–C symmetric stretching mode	$\nu$(CC)	1573

Table 1. Vibration modes of thiophenol.

Vibration Modes	Name	Values (cm−1)
C–H out-of-plane bending, ring out-of-plane deformation	$\gamma$(CH), r–o–d	999
Ring in-plane deformation, C–C symmetric stretching	r–i–d, $\nu$(CC)	1022
C–C symmetric stretching, C–S stretching	$\nu$(CC), $\nu$(CS)	1073
C–C symmetric stretching mode	$\nu$(CC)	1573

The vibration mode at 1073 cm${}^{-1}$ and the UV-irradiation time of 30 min were chosen for its strongest SERS + PIERS intensity for demonstrating the sensitivity of Au/ZnO films (see Figure 2b). In addition, the time of 30 min for the UV-irradiation was chosen in order to obtain a good compromise between the experiment duration and the PIERS enhancement for optimizing cost and speed of these experiments. Figure 2b displays the SERS spectrum (in black color) and SERS + PIERS spectra (in red and blue color with the UV-irradiation time of 15 min and 30 min, respectively). The first observation is that the SERS intensity (black spectrum without UV-irradiation) is higher than in our previous study [45] due to the fact that the excitation wavelength used here (532 nm) is close to the plasmon resonance of our Au nanoparticles on ZnO film. Indeed, the main rule for enhancing (electromagnetic contribution) the Raman intensity is that the plasmon resonance of these metallic nanoparticles must be located between the excitation wavelength and the wavelength of Raman emission for the vibration mode chosen for the investigation [55]. Here, it is well the case due to our choice of 30 nm for the gold nanoparticle diameter. Indeed, the plasmonic resonance of these Au nanoparticles is, thus, nicely located between ${\lambda }_{exc}$ = 532 nm (green dashed line) and ${\lambda }_{Raman}$ = 564 nm (mode at 1073 cm${}^{-1}$, black dashed line), as shown in Figure 2a (see the purple extinction spectrum), compared to our previous work where the excitation wavelength was far from the plasmon resonance of gold nanoparticles [45]. Moreover, when the substrate is irradiated with UV, the plasmonic resonance of our Au nanoparticles on ZnO film stays or is very close from this range comprised between the excitation wavelength and the wavelength of Raman emission, as displayed in Figure 2a (see the blue and dark yellow spectra of extinction), and thus allowing to keep an electromagnetic contribution for the SERS effect in addition to the PIERS effect. Thereby, the Raman intensity is higher than for the only SERS effect or the only PIERS effect, thus permitting a better sensitivity of detection in terms of concentration of molecules (see Figure 2b). In addition, by employing the E${}^4$ model, the enhancement factor (EF) can be expressed as follows [55,56]: EF ∝ $Q_{ext}\left({\lambda }_{Exc}\right)\times Q_{ext}\left({\lambda }_{Ram}\right)$ where $Q_{ext}$ are extinction intensities. From

Table 2. Enhancement factor obtained with the E${}^4$ model (in Arb. Unit). The values of extinction intensities are obtained from the Figure 2a.

Extinction Spectrum	${\mathit{Q}}_{\mathit{ext}}\left({\mathit{\lambda }}_{\mathit{Exc}}\right)$ in Arb. Unit	${\mathit{Q}}_{\mathit{ext}}\left({\mathit{\lambda }}_{\mathit{Ram}}\right)$ in Arb. Unit	EF (Arb. Unit)
Blue	1.788	1.359	2.429
Dark yellow	1.728	1.496	2.585
Purple	1.444	1.664	2.403

, we observe that EF keeps a relatively constant value in or close to the domain bounded by the excitation wavelength and the chosen wavelength of Raman emission, indicating the presence of the electromagnetic contribution.

From Figure 3c, we obtained a limit of detection (LOD) of 100 pM for thiophenol molecules. This LOD is more sensitive than that achieved with Klarite SERS substrates composed of gold inverted pyramids (LOD = 10${}^{-8}$ M, see reference [57]). Furthermore, the experimental data of the Figure 3c were fitted with a Langmuir model defined as follows [57,58]:

$$\begin{array}{c}\hfill I=I_{max}\frac{K_{Ads}{C}_{Thiophenol}}{1+K_{Ads}{C}_{Thiophenol}}\end{array}$$

(2)

where I is the SERS + PIERS intensity at the concentration C${}_{Thiophenol}$. $I_{max}$ is the maximum value of the SERS + PIERS intensity when a self-assembled monolayer of thiophenol is constituted. C${}_{Thiophenol}$ is the employed concentration of thiophenol, and $K_{Ads}$ represents the adsorption constant. The experimental data and the Langmuir model are in good accordance. From this model, we have determined the value of $K_{Ads}$ which is 3.2 × 10${}^6$ M${}^{-1}$, and this latter was bigger than that determined with Klarite SERS substrates ($K_{Ads}$ = 1.1 × 10${}^6$ M${}^{-1}$, see reference [57]). This result has indicated that thiophenol molecules were well-adsorbed on the gold nanoparticles. In addition, we also noted that the Raman (SERS + PIERS) intensity varied linearly with the number of detected thiophenol molecules (see Figure S1 in the Supplementary Materials). Lastly, the analytical enhancement factor (AEF) of the plasmonic ZnO film at LOD was calculated, with the following formula [59]:

$$\begin{array}{c}\hfill AEF=\frac{I_{SERS+PIERS}}{I_{Raman}}\times \frac{C_{Raman}}{C_{SERS+PIERS}}\end{array}$$

(3)

where $C_{SERS+PIERS}$ (100 pM) corresponds to the thiophenol concentration for SERS+PIERS experiment, and $C_{Raman}$ (1 M) is the thiophenol concentration used for the experiment of reference. $I_{SERS+PIERS}$ (see the black SERS+PIERS spectrum in Figure 3b) is the SERS + PIERS intensity, and $I_{Raman}$ (see the blue Raman spectrum in Figure 3d) is the Raman intensity. Figure 3d corresponds to the Raman spectra of the ZnO film without AuNPs (reference experiment). We obtained an AEF value of 1.9 × 10${}^{10}$. To conclude this work, we studied the uniformity of the Raman (SERS + PIERS) signal for our plasmonic substrates. To do that, Raman intensities of the vibration mode at 1073 cm${}^{-1}$ were acquired on 10 positions taken distinctly on the substrate with same excitation conditions for assessing the relative standard deviation (RSD) (see error bars of the Figure 3c). We achieved a nice uniformity the average RSD of which was <20% for all the experimental points of the Figure 3c. To complete that, the reproducibility of samples was also investigated by registering the Raman intensity of the same vibration mode (1073 cm${}^{-1}$) on three samples. Figure 4a,b display the Raman intensity from these samples (for UV-irradiation of 30 min) for the thiophenol concentrations of 100 pM and 1 mM whose the average RSD is <20%, respectively. Thus, we registered a nice uniformity and reproducibility of the Raman signal for these plasmonic substrates whose the average RSD was <20%.

4. Conclusions

In this work, we have proven that Au nanoparticles coated ZnO films could be excellent sensors of thiophenol molecules by using the combination of SERS and PIERS effects. We have chosen the thiophenol molecule because their Raman peaks are well-reported and for its quality of adsorption on metals. Indeed, we found an adsorption capacity $K_{Ads}$ of 3.2 × 10${}^6$ M${}^{-1}$ for thiophenol molecules on gold nanoparticles, which is bigger than those determined in other works mentioned above. For this sensing, a LOD of 100 pM, and an AEF of 1.9 × 10${}^{10}$ were achieved. This LOD of 100 pM was more sensitive than those achieved with other plasmonic nanosystems cited in this work. In addition, the SERS+PIERS signal for our Au nanoparticles coated ZnO film was uniform and reproducible the average RSD of which was <20%. Henceforth, this work can be further optimized by varying the size, the shape and the arrangement of plasmonic nanostructures allowing potentially a better enhancement of the Raman signal than nanospheres used in this study.