Surface-enhanced Raman Scattering and Localized Surface Plasmon Resonance Detection of Aldehydes Using 4-ATP Functionalized Ag Nanorods

Formaldehyde, acetaldehyde, and benzaldehyde are well-known carcinogens affecting human health adversely. Thus, there is a need for efficient detection of these aldehydes. This work uses 4-aminothiophenol (4-ATP) functionalized silver nanorods (Ag NRs) to detect these three aldehydes. The detection mode includes localized surface plasmon resonance (LSPR) and surface-enhanced Raman scattering (SERS). The LSPR band of 4-ATP functionalized Ag NRs shows a linear decrease in absorbance with the increase in formaldehyde and acetaldehyde concentrations. A sensitivity of 0.96 and 0.79 ΔA/mM for formaldehyde and acetaldehyde were obtained. In the case of benzaldehyde, a nearly exponential decrease in absorbance with the increase in concentrations was observed. Above 98.4 μM concentration, the absorbance diminishes completely. The LoD for formaldehyde and acetaldehyde detection using LSPR is 33.8 and 24.6 μM, respectively. The SERS studies reveal that the 4-ATP binds to Ag NRs through both –SH and –NH2 groups and facilitates the inter-particle charge transfer process. The appearance of b2 modes of vibration for 4-ATP evidences this charge transfer process. In the presence of aldehydes, the change in the band shape, relative intensities, and band position were observed primarily in b2 modes of vibration, evidencing the modulation in the charge transfer process. These remarkable changes were seen in μM concentration of aldehydes. Therefore, detection of these aldehydes with 4-ATP functionalized Ag NRs using SERS is possible in concentrations as low as ~ 1 μM.


Introduction
Aldehydes such as formaldehyde, acetaldehyde, and benzaldehyde are volatile organic compounds present in industrial chemicals, the environment, and human metabolites [1,2]. Formaldehyde is classified as a human carcinogen. Formalin (37% aqueous formaldehyde solution) is a well-known food contaminant. It is widely used to increase the shelf-life of food products such as fish, meat, and milk. A higher dose of formaldehyde can cause eye and nose irritation, damage to the central nervous system, immune system disorders, nasopharyngeal cancer, and leukemia [3]. Acetaldehyde is frequently found in alcoholic drinks and fermented foods, where it is produced by the chemical oxidation of ethanol [4]. Acetaldehyde is also classified as a carcinogen to the esophagus. Long-term exposure to acetaldehyde may cause mutational damage to DNA and induce tumors [5]. It can also induce oxidative modification and morphological changes in human erythrocytes [6]. Benzaldehyde is commonly used as a raw material in industries and frequently used as a reagent in organic synthesis. Similar to formaldehyde and acetaldehyde, benzaldehyde is also a human carcinogen and can cause acute poisoning [7]. Several countries have put control on benzaldehyde due to its genotoxic and mutagenic effects [8]. Since these aldehydes have been categorized as human carcinogens and affect human health by causing damage to the central nervous system, respiratory dysfunction, nasopharyngeal carcinoma [3,[9][10][11], detection of them in small concentration with good selectivity and sensitivity is of utmost importance.
Apart from the abovementioned analytical methods, plasmonic sensing of these aldehydes has also been reported. Silver nanoparticles sensitized titanium dioxide [33] Ag nanoparticle decorated carbon nanotube [34] have been used for formaldehyde sensing. Martínez-Aquino et al. used resorcinol functionalized gold nanoparticles for the colorimetric detection of formaldehyde [35]. The refractive index sensitivity properties of gold spherical nanoparticles and nanorods were used to detect formaldehyde [36]. Among some of the recent work, acetaldehyde has been detected using luminescence properties of europium(III) [37] and also with a cataluminescence sensor based on NiO nanoparticles [38]. Plasmonic sensing was also performed for acetaldehyde detection [39]. Benzaldehyde has been mostly detected using a metal-organic framework [7,40,41]. In a recent review, a fluorescence-based sensor utilizing metal-organic frameworks for the detection of formaldehyde, acetaldehyde, and benzaldehyde has been reviewed [42].
Various nanostructures of silver and gold have been utilized for SERS and LSPR studies [43][44][45][46]. The efficient use of these nanoparticles in SERS depends on the excitation wavelengths. It has been shown that the SERS signal is optimum when the excitation wavelength is close (within 150 nm) to the LSPR band maximum of the nanoparticle [47]. Therefore, a broad wavelength tunability of the LSPR band is advantageous for nanoparticles. Among silver nanostructures, triangular nanoplates and silver nanorods (Ag NRs) show sharp LSPR bands with broad tunability [48,49]. In triangular nanoplates, the in-plane dipole band can be tuned very easily from 530 to 780 nm in the visible region [48]. The higher resonance wavelength corresponds to the longer edge length of the nanoplate. Similar tunability is also observed for the silver nanorods where longitudinal band maximum can be tuned up to 680 nm for the Ag NRs of aspect ratio 15-16. Both these nanostructures have been used as efficient Raman substrates. For smaller edge-length nanoplates, the rounding of corners is more prominent. Also, in drop-casting, nanoplates tend to make self-assembly by staking. These two factors may lower the SERS efficiency of smaller triangular nanoplates. Such staking is not observed for nanorods, and the nanorods can be stabilized by suitable selection of capping molecules [49]. Overall the Ag NRs can be a better SERS substrate for the excitation wavelength below 650 nm.
4-aminothiophenol (4-ATP) is an excellent Raman reporter molecule with a -SH group and a -NH 2 group at the para position. This molecule has been used in several works involving SERS. It has been shown that the 4-ATP through its pi-conjugation ring can facilitate the charge transfer process between nanoparticles leading to the enhancement of the SERS signal [50]. Apart from signal enhancement, the forbidden Raman vibrational (b 2 ) modes also become active due to the inter-particle charge transfer. It is wellknown that the aldehydes interact and transform the primary amine group to the imine group [51,52]. In view of the Raman activity of 4-ATP and the possibility of interaction of the -NH 2 group of 4-ATP with aldehydes, this Raman reporter molecule can be useful for the sensitive detection of aldehydes.
This work presents the use of Ag NRs functionalized with 4-ATP for the detection of formaldehyde, acetaldehyde, and benzaldehyde. It is shown that 4-ATP interacts with Ag nanorods (Ag NRs) through both -SH and the -NH 2 functional groups and facilitates the inter-particle charge transfer. However, the relatively weaker interaction of Ag NRs with the -NH 2 group gets modulated in the presence of formaldehyde, acetaldehyde, or benzaldehyde. It appears that at higher concentrations of these aldehydes, the interaction of Ag NRs with the 4-ATP through -NH 2 group diminishes and leads to the annihilation of the charge transfer process. The suppression of the charge transfer process between Ag NRs bonded by 4-ATP in the presence of aldehydes can be utilized for the detection of formaldehyde, acetaldehyde, or benzaldehyde.

Synthesis and Purification of Ag Nanorods
Several approaches are available for the synthesis of Ag NRs [53][54][55][56][57][58]. The approach followed by the group of Murphy is among the most straightforward [54]. In their approach, different aspect ratio Ag NRs was prepared using two-step seed-mediated method. Later, Rekha et al. adopted the same synthesis procedure and showed excellent stability and SERS activity of prepared Ag NRs. The Ag NRs prepared in these works were well characterized by UV-Vis spectroscopy, FE-SEM, TEM, and XRD. So, the extinction spectrum of nanorods can be easily correlated with the shape and size of Ag NRs based on these works.
The synthesis of Ag NRs was performed using the seedmediated method reported earlier by Jana et al. [54]. To synthesize the seed nanoparticles, 0.25 mM silver nitrate and 0.25 mM trisodium citrate were mixed in water with a total volume of 20 mL. The mixture solution is stirred at 600 rpm and, while stirring 0.6 mL of 10 mM freshly prepared NaBH 4 was added. The stirring was stopped after 30 s, and the seed nanoparticles were aged for 2 h before its use.
For the synthesis of Ag NRs, three sets of 10 mL solutions were prepared with 0.25 mL of 10 mM AgNO 3 , 0.5 mL of 100 mM ascorbic acid, and 80 mM CTAB. The solution is mixed well, and 50, 80, and 120 μL volume of Ag seed nanoparticles were mixed in three sets of solutions, respectively. To each set, 100 μL of 1 M NaOH solution was mixed and shaken gently. Color of solutions changes to green, violet, and purple characteristics of Ag NRs. A FESEM image of Ag NRs prepared with 50 μL seed volume is shown in the supplementary Fig. S1. In the following discussion, identifiers 50-NR, 80-NR, and 120-NR are used for Ag NRs prepared using 50, 80, and 120 μL of Ag seed nanoparticles, respectively.
The sensing properties of prepared Ag NRs were investigated with aqueous solution of glucose. As discussed in the next section, the Ag NRs show spectral shift of the longitudinal LSPR band over time which could be due to the presence of growth medium. Therefore, to ascertain the sensing capability, the LSPR experiments were also performed after 20 h of synthesis, and after purification (removal of growth medium) of Ag NRs. For purification, nanorods (1 mL each in nine vials) were centrifuged at 3000 rpm for 10 min. and 50 μL residue was collected after discarding 950 μL supernatant. Vial (1) was used for LSPR experiments with glucose, whereas eight other vials were functionalized with 4-ATP for further LSPR and SERS experiments.

Functionalization of Ag NRs with 4-ATP
For each type of Ag NRs (50-NR, 80-NR, and 120-NR), eight vials after centrifugation were functionalized with 4-ATP. For this, 50 μL Ag NRs residue obtained after centrifugation was diluted by adding 0.5 mL of water. To this, 50 μL of 50 mM 4-ATP solution was mixed and sonicated for 2 min. The final concentration of 4-ATP in solution was 4.16 mM. These solutions were aged in the dark for 2 h before further experimentation. After 2 h, all eight vials were centrifuged again to remove excess of 4-ATP. Fifty microliters of residue was retained in each vial after centrifugation. The Ag NRs in vial (1) (for 50-NR, 80-NR, and 120-NR) were diluted with 1 mL of water and divided into ten vials with 100 μL each. The 4-ATP functionalized Ag NRs in these ten vials were used for LSPR experiments, whereas the other seven vials with concentrated Ag NRs were used for the SERS experiments.

Localized Surface Plasmon Resonance (LSPR) Experiments
LSPR experiments with glucose were performed with (i) as prepared Ag NRs, (ii) Ag NRs after 20-h aging, and (iii) Ag NRs after removal of the growth medium by centrifugation. For LSPR experiments with glucose before centrifugation, 300 μL of Ag NRs was diluted by adding 2 mL water, and absorption spectra were recorded by adding glucose (concentration: 200 mM) from 20 to 200 μL in the step of 20 μL and from 200 to 500 μL in the step of 50 μL. For Ag NRs after centrifugation, 50 μL residue nanoparticles were diluted by adding 250 μL water. Two hundred microliters of diluted AgNR solution was suspended into 2 mL of water, and spectra were recorded using the abovementioned volumes of glucose. For the LSPR experiments with formaldehyde, acetaldehyde, and benzaldehyde, each of ten vials with 100 μL 4-ATP functionalized Ag NRs solutions of 50-NR, 80-NR, and 120-NR was further diluted by adding 500 μL of water. Different volumes (from 0 to 180 μL in the step of 20 μL) of formaldehyde (1.27 mM) in 50-NR, acetaldehyde (1.22 mM) in 80-NR, and benzaldehyde (0.984 mM) in 120-NR were added to these vials. The final solutions were mixed well and incubated for 1 h before recording LSPR spectra.
For all the LSPR experiments in this work, the inexpensive lab-built UV-Vis experimental setup described elsewhere [48,59] is used. Briefly, the experimental setup uses an incandescent bulb and a UV lamp as a source in the 300-1000 nm spectral region. The light beam passes through the quartz cuvette of path length 10 mm. The transmitted light is collected using a fiber-optic cable and fed to the lab-built spectrometer. The processing of data to obtain absorption spectrum was performed externally. An acquisition time of 40 ms and 64 averaging was used in all the LSPR experiments.

Surface-enhanced Raman Scattering (SERS) Experiments
Seven vials with 4-ATP functionalized concentrated Ag NRs (see Section 2.3) were utilized for the SERS experiments. The Ag NRs in all vials were diluted by adding 50 μL water. Out of seven vials of 4-ATP functionalized Ag NRs (50-NR, 80-NR, and 120-NR), vial (1) was used as a control, whereas vials (2)-(6) were mixed with different concentrations of formaldehyde/acetaldehyde/benzaldehyde. The mixture solutions were incubated for 2 h to facilitate the reaction between 4-ATP functionalized Ag NRs and aldehydes. In the seventh vial of 50-NR, 80-NR, and 120-NR, Ag NRs were diluted by adding 150 μL of water followed by the addition of 10 μL of formaldehyde (12.7 M), acetaldehyde (12.2 M), and benzaldehyde (9.84 M), respectively. The resulting concentrations of formaldehyde, acetaldehyde, and benzaldehyde were 0.6, 0.58, and 0.47 M. After the incubation time, 2.5 μL of solutions from vials 2-6 was deposited on a hydrophobic slippery glass surface. The surface was prepared by coating the cleaned glass surface with commercially available hydrophobic chemical rainX. The advantage of a hydrophobic slippery surface is that as the deposited droplet dries, it shrinks with all its content because the border of the droplet is not pinned to the surface due to its slippery nature. In this way, a very small and sample enriched spot can be obtained for water-based solutions. For experiments, to obtain a higher concentration of Ag NRs, two drops of the same concentration were dried sequentially. For higher concentrations of aldehydes in the seventh vial, SERS experiments were performed in the solution phase.
All the surface-enhanced Raman spectroscopy (SERS) experiments were performed with our lab-built setup described elsewhere [60][61][62]. Briefly, in the setup, a laser diode of tunable wavelength (638-641 nm) was used as an excitation source. The light beam passes through a mirror with a hole (diameter: 3 mm) which also acts as a reflector of the SERS signal to the collimation optics. The SERS spectra were obtained using TE cooled spectrometer (Avantes: AvaSpec-ULS2048LTEC, wavelength range 520-1000 nm, resolution: 0.6 nm) using a fiber-optic patch cord. The recording of Raman spectra was performed using Avasoft software interfaced with the Avantes spectrometer.

Synthesis and UV-Vis Characterization of Silver Nanorods
The Ag nanorods (50-NR, 80-NR, and 120-NR) were synthesized by following the procedure reported by Jana et al. [54]. The synthesized Ag seed nanoparticles and nanorods were characterized by our lab-built UV-Vis spectroscopy setup. The spectrum of synthesized Ag seed nanoparticle and nanorods is shown in Fig. 1a. A single LSPR band with a maximum at 388 nm was observed for seed nanoparticles. The appearance of this band indicates the spherical shape of the seed nanoparticles. For Ag NRs, two LSPR bands, longitudinal towards higher wavelength and transverse Absorbance (a.u.) 638 Fig. 1 a UV-Vis absorption spectra of Ag seed and Ag NRs obtained with various volumes of seed; b absorption spectra of Ag NRs obtained after 20 h. The ratios of longitudinal and transverse band intensities and FWHM (in nm) are in parentheses towards lower wavelength, were observed. These two bands are consistent with the UV-Vis spectra of Ag NRs reported earlier [49,54], confirming the synthesis of Ag NRs. While the maxima of the transverse band in 50-NR, 80-NR, and 120-NR were nearly constant at 420 nm, the longitudinal band maxima were observed at 665, 608, and 571 nm for 50-NR, 80-NR, and 120-NR, respectively. As reported earlier, a lower volume of seed nanoparticles results in a higher aspect ratio of Ag NRs with the longitudinal band maximum towards a higher wavelength.
The longitudinal band of synthesized Ag NRs shows a blue shift with time in the growth solution. Figure 1b shows the absorption spectra of synthesized nanorods after 20 h of synthesis. It is evident that 27, 26, and 15 nm blue shift occurs for 50-NR, 80-NR, and 120-NR, respectively. Apart from the blue shift, a decrease in FWHM of the longitudinal bands and an increase in the longitudinal to transverse band intensity ratio (L/T ratio) are seen. Despite the blue shift in the longitudinal band maxima, no wavelength shift occurs in the transverse band maxima. It is expected that in the growth solution, the reaction may continue and possibly lead to the increase in the length of nanorods resulting in a red shift of the longitudinal band. However, the observed blue shift is not consistent with this hypothesis. A similar blue shift of longitudinal bands with time has also been observed for small Au nanorods [63]. It was shown that in the presence of the growth medium, the nanorod length increases for a brief time, followed by increase in the width, leading to the decrease in the aspect ratio of nanorods. This decrease in the aspect ratio of the nanorods resulted in a blue shift of the longitudinal LSPR band. Apart from the explanation based only on the aspect ratio change, Recio et al. showed that under the non-dynamical condition, the interaction between collective longitudinal oscillation of electron with the surrounding medium could also cause the observed blue shift in Au nanorods of length < 100 nm [64]. It is also possible that depending on the thermodynamic environment of the growth medium, the surfactant molecule undergoes reorganization leading to the variation in the dielectric constant of the medium. This variation in the dielectric constant of the medium can result in the blue shift of the longitudinal plasmon band [65]. The observed blue shift in the longitudinal LSPR band in the synthesized Ag NRs could also be due to the dynamical processes in the presence of the growth medium.
While comparing the absorption spectra of synthesized nanorods, as prepared and after 20 h in Fig. 1a, b, the L/T ratio shows improvement along with the significant decrease in the FWHM. It is important to mention that the higher FWHM of absorption bands corresponds to lower lifetime or higher damping of excited plasmons [66]. The plasmon lifetime is directly related to the local field enhancement by nanorods. The lower plasmon lifetime leads to lower field enhancement which could be a critical factor for applications of nanorods.
The transverse LSPR band corresponds to the width of the nanorod and the contributions from the spherical nanoparticles in the colloidal solution. The spherical nanoparticles of larger size could have formed due to NaBH 4 present in the seed solution, which can reduce the AgNO 3 in the growth solution. However, negligible change in the position of the transverse band indicates that the width of nanorods did not change over 20 h. Because of the possible role of the growth medium for observed blue shift in the longitudinal LSPR bands, the synthesized Ag NRs were separated from the growth medium by centrifugation and re-dispersion in deionized water.

LSPR Properties of Silver Nanorods in the Presence of Glucose
Glucose, being a highly water-soluble small molecule, is useful for determining the sensing capability of nanostructures. By varying the glucose concentration minutely, the refractive index of the solution can be controlled very precisely. The synthesized Ag NR sensitivity was investigated with various concentrations of glucose. The sensing experiments were performed with as-prepared nanorods, nanorods after 20 h of synthesis, and nanorods after removal of the growth medium (24 h after synthesis). Figure 2a shows representative spectra obtained for 120-NR after 20 h of synthesis and with various glucose concentrations. Other results on glucose-sensing are shown in Supplementary Figs. S2 and S3. As it is evident in Fig. 2a, various glucose concentrations in the nanorod solution do not induce any red shift of the longitudinal band maximum. Instead, a change in the absorption was observed for both longitudinal and transverse plasmon bands. A decrease in absorbance was observed with a subsequent increase in glucose concentration. Similar to the phenomena observed in Au nanorods [67], it is likely that the glucose molecule binds to the nanorod surface and induces aggregation. This aggregation of Ag NRs may result in the hindered oscillations of plasmons along both long and short axes, leading to a decrease in absorbance. Figure 2b shows the plot of absorbance change with various glucose concentrations. The observed absorbance change in Fig. 2a was normalized with respect to the control sample with zero glucose concentration. A unit value is subtracted from the normalized absorbance to make the absorbance change relative to the control sample. The scattered plot in Fig. 2b represents the absorbance change, whereas the solid line is a linear fit to absorbance change data. The slope of the fitted line provides the sensitivity, which is 0.00649 ΔA/mM for 120-NR 20 h after the synthesis. Table 1 lists the sensitivities obtained with other nanorods with asprepared Ag NRs, Ag NRs after 20 h of synthesis, and Ag NRs after removal of the growth medium. As shown in the table, the sensitivity was maximum for 50 NR in all cases, and it decreased minutely towards 120-NR. Such behavior is expected as the aspect ratio of 50-NR is relatively higher than 80-NR and 120-NR. Therefore, a larger surface area is available for glucose molecules to bind on the surface of 50-NR, leading to higher sensitivity for 50-NR than 80-NR and 120-NR. The relative change in sensitivity is not very high as the aspect ratios of synthesized Ag NRs may not be significantly different. To obtain the precise sensitivity in the measurement, the limit of detection (LoD) is calculated from the linear fit in Fig. 2b. The calculation of LoD (3.3 × standard deviation of y-intercept/slope of regression line) is performed using the linear regression method. The obtained value of LoD for spectra shown in Fig. 2a

Functionalization of Ag NRs with 4-ATP and LSPR Detection of Aldehydes
For LSPR experiments with aldehydes, for 50-NR, 80-NR, and 120-NR, ten vials with 0.6 mL of 4-ATP functionalized Ag NRs in each were used. To 50-NR, in 10 vials, formaldehyde (concentration: 1.27 mM) volume from 0 to 180 μL in the step of 20 μL was added. Similarly, to 10 vials of 80-NR and 120-NR, the same volumes of acetaldehyde (concentration: 1.22 mM) and benzaldehyde (concentration: 0.984 mM) were added. Before LSPR spectra recording, all the solutions were incubated for 1 h. Figure 3a-c shows the absorption spectra for three aldehydes in three Ag NR solutions. A decrease in absorbance with an increase in concentrations of formaldehyde and acetaldehyde is evident in Fig. 3a, b. For benzaldehyde, a similar observation for the concentration below 89.5 μM was made. Above 89.5 μM concentration, the absorption vanishes abruptly. This absence of absorption for benzaldehyde above 89.5 μM was also evident through the color change of the solution. The characteristic color of 4-ATP functionalized Ag NRs disappeared, which indicates a stronger benzaldehyde interaction with 4-ATP functionalized Ag NRs. The decrease in the absorbance in the presence of three aldehydes indicates that the presence of aldehyde induces aggregation of NRs leading to micrometer size aggregate. In the case of benzaldehyde, this aggregation is stronger compared to formaldehyde and acetaldehyde. Figure 3d, e respectively shows the change in relative absorbance for 50-NR and 80-NR with different concentrations of formaldehyde and acetaldehyde. For both cases, a   Fig. 3f, absorbance decreases nearly exponentially. Although the absorbance decrease is evident for the lowest concentration of 31.7 μM, it is difficult to estimate the sensitivity for benzaldehyde using LSPR in the probed concentration range. Therefore, the LoD for benzaldehyde can be assumed as 31.7 μM.

SERS Detection of Aldehydes with 4-ATP Functionalized Ag NRs
Surface-enhanced Raman scattering (SERS) experiments on 4-ATP functionalized Ag NRs were performed using a lab-built setup. Figure 4a-c shows the SERS spectra of 120-NR, 80-NR, and 50-NR conjugated to 4-ATP. Figure 4a shows bands at 1077, 1140, 1175, 1387, 1436, 1540, and 1588 cm −1 . These bands are also observed in the spectrum of 80-NR and 50-NR. However, the band in region 1500-1600 cm −1 is not resolved and appears as a single broad feature with a center at 1572 cm −1 . In earlier work, SERS bands at the same positions for 4-ATP bound to Ag and Au nanoparticles were observed [50], which confirms the successful attachment of 4-ATP to Ag NRs in the present Absorption spectra of 120-NR with benzaldehyde  Table 2 lists the band positions and mode assignment of observed bands based on the earlier reported assignments [50,[68][69][70][71].
In Table 2, it is interesting to note that the nontotally symmetric b 2 modes are visible in all the spectra. In earlier work [72], these b 2 symmetry modes appeared when an efficient inter-particle charge transfer occurred between nanostructures. The charge transfer process occurs when a single 4-ATP molecule interacts with two nanoparticles and provides a π-conjugation path between them. In case when 4-ATP interacts with nanoparticles only through -SH group or through -NH 2 group, these b 2 symmetry modes do not appear [72]. Therefore, the observed spectra in Fig. 4 indicate the conjugation of 4-ATP molecules to Ag NRs through both -SH and -NH 2 groups providing a π-conjugation bridge to facilitate the charge transfer process. Figure 5a-c shows the SERS spectra of 4-ATP functionalized 50-NR, 80-NR, and 120-NR with different concentrations of formaldehyde, acetaldehyde, and benzaldehyde. Noticeable changes in the SERS spectra are evident in Fig. 5a when formaldehyde was added to Ag NRs. An apparent band shape change in the broad feature at 1572 cm −1 occurs with an appearance of a clear band at 1577 cm −1 . Also, intensity increases for bands at 1140, 1387, and 1426 cm −1 . The band at 1186 cm −1 showed a decrease in intensity. With the increase in the concentrations, 1140, 1387, and 1426 cm −1 bands indicate intensity reduction relative to the 1077 cm −1 band. Simultaneously, intensity increase of band at 1577 cm −1 with the increase in formaldehyde concentration is observed; it becomes comparable to the intensity of 1077 cm −1 band. Figure 6a shows the trend in intensity of the 1140 cm −1 band with formaldehyde concentration. A linear fit to the experimental data is also shown.   Figure 6b shows a plot of intensity variation versus acetaldehyde concentrations. Similar to formaldehyde, the 1577 cm −1 band intensity increases and becomes comparable to that of 1077 cm −1 band at 122 mM concentration of acetaldehyde.
In the case of benzaldehyde (Fig. 5c), the intensity increase of bands was not so appreciable compared to that of formaldehyde and acetaldehyde. However, the 1140 cm −1 band shows an intensity decrease with increasing benzaldehyde concentration, and it nearly disappears above 98.4 μM concentration. Figure 6c shows the intensity variation of the 1140 cm −1 band. The relative intensity change of 1540 and 1588 cm −1 bands is also evident. For 984 μM benzaldehyde concentration, an evolution of two bands at 1567 and 1615 cm −1 was also observed along with another prominent band at 1163 cm −1 . Apart from these, a substantial decrease in intensity for the 1387 and 1436 cm −1 bands was observed. The change in relative intensities of bands, band shape, and spectral position of bands corresponds to the detection of three aldehydes with 4-ATP functionalized Ag NRs. These spectral changes are evident in concentrations as low as ~ 1 μM and therefore can be considered as the lowest detection limit of three aldehydes in this work.
The SERS spectra in the solution were also recorded at higher concentrations of formaldehyde (0.6 M), acetaldehyde (0.58 M), and benzaldehyde (0.47 M). Figure 7 shows the corresponding spectra. It is interesting to note that b2 symmetry bands completely disappear from the spectrum at higher concentration of formaldehyde. Prominent bands at 1078 and 1595 cm −1 and low-intensity features at 1135 and 1167 cm −1 were observed. For acetaldehyde, the most intense band appears at 1572 cm −1 , which is ~ 23 cm −1 shifted compared to the band at 1595 cm −1 in formaldehyde. Apart from this, a low-intensity band appears at 1620 cm −1 , which was nearly absent (tentative band at 1632 cm −1 in the formaldehyde spectrum). The band at 1077 cm −1 stays intact, while a new band arises at 1169 cm −1 with characteristic bands at 1134 and 1197 cm −1 with lower intensity. These two low-intensity bands could be the same bands observed at 1135 and 1187 cm −1 in formaldehyde. In the case of benzaldehyde, two prominent bands appeared at 1568 and 1170 cm −1 , along with characteristic 4-ATP bands at 1078, 1141, and 1192 cm −1 . Also, a small band appears at 1617 cm −1 . The bands observed at 1568 and 1617 cm −1 were identified using harmonic frequency calculation with the G09 program suite [73]. The calculation details are given in the supplementary material along with optimized structures (in Fig. S4) and calculated Raman spectra in Fig. S5. The 1568 cm −1 band is due to the C-C stretch of the ring frame, whereas the small band corresponds to the C = N stretching mode. In benzaldehyde, the 1170 cm −1 band is much more pronounced than acetaldehyde and has relative intensity two times compared to the band at 1078 cm −1 . The band shape of the 1572 cm −1 band in acetaldehyde is entirely different from the 1568 and 1595 cm −1 bands in benzaldehyde and formaldehyde. The appearance of C = N bands confirms the reaction between primary amine of 4-ATP with aldehydes followed by the formation of imine. It is essential to mention that the SERS signal strength was maximum for benzaldehyde, followed by formaldehyde and acetaldehyde. The same trend was also observed in calculated Raman spectra ( Supplementary Fig. S5). The higher SERS strength for benzaldehyde could be due to the delocalized π-ring system compared to other aldehydes. The LSPR detection of three selected aldehyde shows decrease in the absorbance with increase in the aldehyde concentration which is possibly due to the aggregation of Ag NRs. In the SERS spectra, modulation of band intensity, band position, and band shape were observed primarily in the non-totally symmetric b 2 modes of vibration. In addition, at higher concentration of aldehyde, vibration corresponding to C = N stretching mode appears in the SERS spectrum. The appearance on this mode clearly indicates the nucleophilic addition of the aldehyde to the -NH 2 group of 4-ATP leading to disruption of the inter-particle charge transfer. Therefore, both LSPR and SERS techniques could be useful for the detection of three aldehydes.

Conclusions
Silver nanorods (Ag NRs) were synthesized using the seedmediated method by utilizing 50 (50-NR), 80 (80-NR), and 120μL (120-NR) seed volumes. The synthesized Ag NRs were characterized using UV-Vis spectroscopy, and the expected aspect ratio of Ag NRs is ~ 12.5-15.5. Spectral shifts towards blue were observed in the longitudinal plasmon resonance bands with time which could be due to the involved dynamical processes in the presence of the growth medium. The synthesized Ag NRs were investigated for their refractive index sensitivity against various glucose concentrations. The obtained sensitivity for glucose detection was better than 0.005 ΔA/ mM. The 4-ATP functionalized Ag NRs were investigated for formaldehyde, acetaldehyde, and benzaldehyde sensing using LSPR. The obtained sensitivities for formaldehyde and acetaldehyde were 0.96 and 0.79 ΔA/mM, respectively. For benzaldehyde, the change in absorption was nearly exponential and disappeared abruptly above 100 μM concentration. The 4-ATP functionalized Ag NRs were also utilized for the detection of formaldehyde, acetaldehyde, and benzaldehyde using surface-enhanced Raman scattering (SERS). In the absence of aldehydes, 4-ATP conjugated Ag NRs show b 2 symmetry bands, characteristics of the inter-particle charge transfer process. With aldehydes, the disruption in the chargetransfer process was observed, and changes in the band shape, relative band intensities, and shift/evolution of bands were observed. These changes were evident in aldehyde concentrations ~ 1 μM. The observation establishes that the 4-ATP bound Ag NRs and possibly other Ag nanostructures can be efficiently utilized for the detection of formaldehyde, acetaldehyde, and benzaldehyde by investigating the inter-particle charge transfer processes.
Author Contribution RKS has conceptualized the work, performed the experiments, analyzed the data, wrote the manuscript, and generated the funding for the work. Data Availability All data generated or utilized during this study are included in this article and its supplementary information files.

Competing Interests
The author declares no competing interests.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/.