An Innovative Polyaniline/Platinum-Coated Fiber Optic – Surface Plasmon Resonance Sensor for Picomolar Detection of 4-Nitrophenol

The paper reports for the rst time an innovative polyaniline (PANI)/platinum (Pt)-coated ber optic – surface plasmon resonance (FO-SPR) sensor used for highly-sensitive 4-nitrophenol (4-NP) pollutant detection. The Pt thin lm was coated over an unclad core of an optical ber (FO) using a DC magnetron sputtering technique, while the 4-NP responsive PANI layer was synthetized using a cost-effective electroless polymerization method. The presence of the electrolessly-grown PANI on the Pt-coated FO was observed by eld-emission scanning electron microscopy (FE-SEM) and subsequently evidenced by energy dispersive X-ray analysis (EDX). These FO-SPR sensors with a demonstrated sensitivity of 1515 nm/RIU were then employed for 4-NP sensing, exhibiting am excellent limit of detection (LOD) in the low picomolar range (0.17 pM). The proposed sensor’s conguration has many other advantages, such as low-cost production, small size, immunity to electromagnetic interferences, remote sensing capability, and moreover, can be operated as a “stand-alone device”, making it thus well-suited for applications such as “on-site” screening of extremely low-level trace pollutants.


Introduction
The environmental pollution by phenol-based aromatic nitro compounds in water samples is a major concern worldwide [1]. These nitrophenols are mostly widespread within surrounding environment from industrial wastes, as they are extremely used in the production of pharmaceuticals, pesticides, insecticides, explosives and dyes [2]. According to USA Environmental Protection Agency (EPA), 4nitrophenol (4-NP) is the most toxic, hazardous and persistent organic pollutant, which can cause signi cant damages to the health and environment, even at low-level concentrations [3]. Hence, there is need for highly-stable, e cient, robust and reliable sensors that can detect traces of 4-NP, in a rapid and ultrasensitive manner [4]. Until now, several techniques such as capillary electrophoresis, uorescence, high-performance liquid chromatography (HPLC), mass spectrometry combined with liquid chromatography (LC-MS) or with gas chromatography (GC-MS), and surface enhanced Raman spectroscopy (SERS) have been widely employed for hazardous chemical pollutants sensing, including 4-NP [5,6]. However, these classical analytical techniques have limitations of being time-consuming, and they typically require sophisticated and expensive instrumentation, trained personnel, as well as multistep sample preparation protocols, being thus quite expensive techniques to be commonly used in daily life and industry [7]. In addition, electrochemical approaches such as cyclic voltammetry, linear sweep voltammetry, differential pulse voltammetry and chronoamperometry have similarly shown their potential for the detection of 4-NP [4,8]. Despite the fact that electrochemical methods are generally cost-effective, highly-sensitive and selective, their performance strongly depends on the electrode modi ers, and more important, they are not so stable at temperature and pH uctuations [9]. Different from the above-mentioned detection techniques, ber optic -surface plasmon resonance (FO-SPR) sensing is a relatively novel biochemical method with the advantages of featuring a compact footprint, label-free detection and real-time monitoring capabilities, as well as offering the possibility to perform rapid and non-invasive measurements [10,11,12]. Such a re ection-type FO-SPR sensor is commonly prepared by uncladding rst a small portion at one end of the FO, and then coating the exposed FO core by a plasmonic metal layer, typically gold (Au) or silver (Ag) [13,14]. The state of surface plasmons excited with light (guided by total internal re ection through the FO) at the metal/dielectric surface interface changes when the coated-FO core is immersed within the environment solution containing the target analyte. Thus, a SPR dip at a particular wavelength is then obtained in the re ection spectrum, which strongly depends on the refractive index (RI) of the sensing medium around the metallic layer [10,13]. Owing to that, FO-SPR sensors have been widely used in medical diagnostics and environmental monitoring applications, for studying molecular interactions and their binding speci city [15,16,17]. For example, Singh et al. reported the development of a transmission-type FO-SPR biosensor for the detection of phenolic compounds (catechol, m-cresol, phenol and 4-chlorophenol) in aqueous solutions [9]. The sensing probe was prepared by depositing Ag lm onto FO core via a thermal evaporation method followed by the immobilization of enzyme tyrosinase, using a gel entrapment technique. In this case, the authors claimed a limit of detection (LOD) for all analyzed phenolic compounds in the low µM concentrations range. Alternatively, Cennamo et al. presented a detection scheme of another nitrophenol compound (TNT -2,4,6-trinitrotoluene) based on the combined approach of FO-SPR and molecular imprinting technique [15]. The SPR device was obtained by coating a 60 nm thick Au lm over the FO core using a sputtering method. The sensing method demonstrated a detection limit of 51 µM with a sensitivity of 27 µm/M. However, one year later the authors have shown further improvements of the TNT sensor by designing a localized SPR (LSPR) device, through the incorporation of branched Au nanostars dispersed into a molecular imprinted polymer initially coated on the FO core. In this way, the authors obtained better LOD and sensitivity values, of 2.4 µM and 84 µm/M, respectively [16]. Noteworthy, to the best of our knowledge, yet there is no evidence in literature of employing a re ection-type FO-SPR sensor for 4-NP detection.
In this work, results on the fabrication and characterization of an innovative FO-SPR sensor, based on a polyaniline (PANI) / platinum (Pt) bilayer coated over an unclad FO core, and used for 4-NP detection, were reported for the rst time. The Pt thin lm was deposited by DC magnetron sputtering and it replaced the conventional Au layer commonly preferred with a traditional re ection-type FO-SPR sensor [13]. So far, only a limited number of theoretical studies were reported with Pt-coated FO-SPR sensors operating in transmission mode [17,18]. Herein, using Pt as a plasmonic material within re ection-type FO-SPR sensors is a novel approach and the excellent catalytic properties of Pt are essential for subsequent PANI synthesis steps [19,20]. Complementarily, PANI is an organic polymer with excellent stability and physico-chemical properties in terms of high electrical conductivity, large electro-active surface, and unique combination of RedOx states and proton doping pro les [4,21]. These particular features render PANI as an extremely responsive polymer to several molecular species, being so far successfully used in energy storage applications [22], pH monitoring [23], gas sensing [24] and pollutants detection including nitrophenol compounds [4]. In this work, PANI was synthetized using a cost-effective electroless polymerization approach, in an attempt to uniformly deposit thin sensitive PANI lms on the curved Pt-coated FO core three-dimensional (3D) geometry. The PANI/Pt-based FO-SPR sensor was then morphologically characterized and evaluated for highly-sensitive 4-NP pesticide detection in water samples, demonstrating a sensor's LOD in the low pM concentrations range. This work represents thus a step forward in the fabrication of reliable FO-SPR sensors, not only with improved performance, but also with extended functionality.

Reagents and materials
All the reagents used in this work were of analytical grade (99.99% purity, unless otherwise speci ed). Ultra-clean deionized water (DIW), puri ed by a TKA Milli-Q 50 system, was consistently used throughout the experiments. Acetone, sulfuric acid (97% H 2 SO 4 ), D(+)-sucrose, ethanol and aniline (99% C 6 H 5 NH 2 ) were supplied by Merk. The nitrogen and oxygen 5.0 purity gas bottles were acquired from Messer. The TEQS multimode FO of 400 µm diameter was provided by Thorlabs. The additional tools used in the aniline polymerization protocol, such as the Pasteur glass pipettes (2 mL capacity, 230 mm length), the double-wall glass (diameter 55 mm, capacity 150 mL, height 85 mm) and the magnetic stir plate were obtained from VWR. The TC120 heated circulating bath was purchased from Grant Instruments.

FO-SPR setup and sensors fabrication
The "in-house" developed FO-SPR sensing platform consists of several components, as illustrated in Fig. 1A: a polychromatic tungsten halogen light sources (AvaLight, Avantes), an UV-VIS spectrophotometer (AvaSpec 2048, Avantes), an interchangeable FO-SPR sensor ( Fig. 1B) inserted into a SMA (SubMiniature version A) connector (Avantes) and mounted in a bifurcated FO (Avantes), as well as an automated computer-controlled robotic arm programmed using the ColiDrive software (Colinbus). The light passing through the SPR sensitive zone is re ected back at the FO sensing tip and measured using the spectrometer. Any change in the surrounding environment occurring at the Pt surface results in a shift of the typical SPR spectral resonance dip, subsequently monitored in real-time and processed using an "in-house" developed LabVIEW script (National Instruments). The interchangeable FO-SPR sensors were prepared using a previously described protocol [13,25]. In brief, the multimode FO with a diameter of 400 µm was rst split into 3.6 cm long segments. Then, a sensitive SPR zone of 0.6 cm was constructed at one side by mechanically removing the jacket and subsequently uncladding the FO in acetone. The exposed FO silica core was then carefully dried with dust-free tissues and under N 2 gas ow. Next, the sensor tips were isotropically coated by a thin Pt layer (40 nm) using a sputter coater (Quorum Q150R ES, UK). The DC plasma was engaged for 15 min at 54 mA in an Ar atmosphere kept at 2.5 Pa. The FO-SPR sensor tips were installed on a rotating stage (100 rpm) to improve the Pt FO coverage during the sputtering process, while the deposited thickness was monitored using the built-in quartz crystal oscillator (QCM). The reliability of the Pt layer thickness covering the FO-SPR sensor tips and its evenness were thus assured by the well-known isotropic nature of the sputtering process, coupled with the accuracy of the QCM real-time measurement. Ultimately, the Pt coated-FO sensor tips were used as catalysts in the electroless polymerization process of aniline.

FO-SPR refractometric measurements
The sensitivity of the Pt-coated FO-SPR sensors was evaluated by performing refractive index (RI) measurements in sucrose dilutions (0, 2, 4, 8, 12% w/w). The Brix values of the prepared sucrose dilutions were checked with a digital refractometer (Atago Palette PR-32) and their corresponding RI values are presented in Table 1. . After a given immersion time, the sensors were taken out of the electroless reactor and thoroughly washed with DIW. Before employing the as-prepared sensors for 4-NP sensing, the obtained PANI lms were undoped with a 1 M NH 4 OH solution for 10 min to induce an initial well-known Emeraldine base state of PANI, marked by a change in the lm color from light green to dark blue.

Observations of the FO-SPR surfaces and PANI thickness measurement
A eld-emission scanning electron microscope (FE-SEM, JEOL7600F) equipped with an energy dispersive X-ray (EDX) analyzer was used to investigate the surface morphology and structural properties of the PANI lm deposited on the Pt-coated FO-SPR sensors. A low accelerating voltage (2 kV) was constantly applied to reduce the charging effects and to extract more information close to the sample surface. The EDX spectroscopy was effectively used to qualitatively and quantitatively con rm the elemental composition of the fabricated FO-SPR sensors. Noteworthy, SEM and EDX analysis were carried out on Pt-coated FO-SPR sensors covered by PANI lms intentionally doped in a 1M HCl solution for 10 min to enhance their conductivity and hence to further reduce specimens charging during the SEM observations.
In a second stage, the PANI lm thickness was determined by pro lometry (Stylus Pro ler XP-2, Ambios Technology), providing precision surface topography measurements with 1.5 Å vertical resolution. In this case, the Pt-coated FO-SPR sensor tips were half-covered with an adhesive tape before their immersion into the electroless reactor, in order to generate a Pt-PANI height-step pro le after the tape subsequent removal. The FO-SPR sensors were horizontally positioned on the pro lometer specimen holder and the pro lometer tip was moved on top of the cylindrical FO side about 400 µm across the Pt-PANI height-step pro le.

Detection of 4-nitrophenol in water samples
The as-prepared PANI/Pt-coated FO-SPR sensors were further used for direct and subsequent detection of different concentrations of 4-NP (0, 1, 100, 10 3 , 10 5 and 10 6 pM) in DIW water. Each 4-NP concentration was measured three times independently, using freshly prepared PANI/Pt-coated FO-SPR sensors.

Sensitivity of the platinum coated FO-SPR sensors
The sensitivity performance of the Pt-coated FO-SPR sensors was evaluated by performing RI measurements in serial sucrose dilutions. As aforementioned already, despite few theoretical attempts [17,18], this work reports for the rst time on the fabrication of a re ection-type FO-SPR sensor based on a Pt plasmonic layer and the determination of its performance indicators (i.e. sensitivity -S and gure of merit -FOM). Figure 2A shows the SPR spectral dips obtained at 0 (red curve) and 12% (blue curve) sucrose concentrations. The obtained SPR shifts were plotted as a function of RI values for generating the calibration curve presented in Fig. 2B. The sensitivity values were then extracted from the slope of these calibration curves. In this way, the sensitivity was determined to be around 1515 nm/RIU. In the case of FO-SPR sensors, the sensitivity (S) is expressed as the ratio between the wavelength shift (Δλ SPR ) and the RI change (Δn) in the analyzing medium: S = Δλ SPR /Δn [nm/RIU] [10]. Furthermore, the FOM of the Ptcoated FO-SPR sensors was also calculated to be around 7 RIU − 1 . Table 2 gives a brief performance comparison among various types of FO-SPR sensors reported in literature.

PANI deposition on the Pt-coated FO-SPR sensors
PANI thin lms were synthesized on the Pt-coated FO-SPR substrates using a relatively novel electroless deposition method well described in literature [26,27], where PANI is simply obtained through the polymerization of aniline on the Pt surface acting as a catalyst. The process is based on spontaneous chemical reactions in acidic medium, involving reduction of dissolved oxygen as cathodic half-reaction and oxidation of aniline as anodic half-reaction at the metal/solution interface [26]. The polymerization reaction is thus initiated on the Pt surface by a catalytic oxygen reduction, and then the primary formed PANI layer takes over the autocatalytic polymerization of aniline. Consequently, when the Pt-coated FO-SPR sensors were immersed in the electroless reactor kept under oxygen saturation, a light greenish color gradually appeared on their surface. The greenish color appearance is a characteristic of the acidi ed Emeraldine mid RedOx state of PANI. In this work, the thickness of the electrolessly-grown PANI lm on the Pt-coated FO-SPR sensor was also studied as a function of the reaction time. The PANI thickness was accurately evaluated by pro lometry.
Several Pt-coated FO-SPR sensors were immersed in the equimolar (0.4 M) aqueous solution aniline and H 2 SO 4 kept at 25°C under continuous oxygen bubbling, and gradually removed after 2, 4 and 6 h, respectively. The thickness of the grown PANI lm after each immersion duration was measured using surface pro lometry. As can be observed in Fig. 3B a linear time-dependence of the PANI thickness was found, as previously reported [27]. This signi es that the polymerization rate is constant, PANI growth occurring at a rate of ~ 17 nm/h. A typical example of pro lometric measurement after 6 h PANI growth is shown in Fig. 3A, where the height-step of ~ 95 nm between the Pt-coated FO-SPR surface and the PANI lm denotes the thickness of the latter.

Morphological and structural characterization of the FO-SPR surface
FE-SEM micrographs acquired from the surface of a Pt-coated FO-SPR sensor before and after PANI electroless deposition for 6 h at 25°C are shown in Fig. 4. As can be observed, the Pt-coated FO-SPR sensor has a homogenous and smooth surface (Fig. 4A). Noteworthy, the FO-SPR sensors (inset of Fig. 4A) were investigated in several places, with similar results obtained for both the FO tip and its circular sides, as expected due to the FO-SPR sensors rotation during the deposition step and isotropic nature of the DC magnetron sputtering process [13,25]. Similarly, the PANI lm on the Pt-coated FO-SPR sensor (Fig. 4B) is evidenced by the roughened curly aspect of the surface (magni ed in the inset of Fig. 4B), typical for a thin PANI lm grown on a Pt substrate through an electroless synthesis procedure [26,29]. As can be noticed, the obtained PANI lm was homogeneous, dense, with good conformality and well adhered to the Pt-coated FO-SPR surface. This wavy aspect of the PANI surface was found to be more pronounced for the 6 h electroless synthesis duration and it is believed that it plays an important role in the FO-SPR sensing performance [25], justifying the reason for selecting the 95 nm PANI thickness as optimal for further 4-NP detection studies.
Corresponding EDX patterns of the FO-SPR surface presented in Fig. 4 are shown in Fig. 5. The EDX analysis performed on the Pt-coated FO-SPR surface (Fig. 4B) well indicates the presence of the constitutive elements, i.e. oxygen (O), silica (Si) and Pt. The small intensity of the carbon (C) peak in the Pt-coated FO-SPR surface can be attributed to the adhesive carbonic conductive tape used to x the sensors during the EDX measurements. In addition, the PANI presence on the Pt-coated FO-SPR surface is suggested by a nitrogen (N) peak and a higher intensity carbon (C) peak, respectively (Fig. 4A). However, small chlorine (Cl) traces could be also observed in Fig. 4A due to the PANI protonation step in HCl. The insets of Figs. 4A and 4B show associated quantitative elemental calculation charts, where the results described well the coexistence of both, supporting Pt-coated FO silica core (i.e. Pt/SiO 2 ) and PANI lm.

Detection of 4-Nitrophenol in water samples
The as-prepared PANI/Pt-coated FO-SPR sensors (with ~ 95 nm PANI thickness) were used to detect ve concentrations of 4-NP (0-106 pM range) in DIW samples. First, the FO-SPR sensor speci city was tested by employing a Pt-coated sensor (without PANI layer) for detecting the highest 4-NP concentration in DIW (10 6 pM). As shown in Fig. 6A, although insigni cant, a SPR wavelength shift of less than 1 nm can be observed, possibly due a very slight change in the RI of the 4-NP-reach analyzing medium, and/or due to absorption effects of the 4-NP molecules on the Pt surface in the absence of the sensitive PANI lm. Figure 6B shows the linear calibration curve acquired after 1 min detection of each 4-NP concentration with the PANI/Pt-coated FO-SPR sensor. The linear relationship was given by a regression equation with a coe cient of determination (R 2 ) of 0.987 and a slope (sensor's sensitivity for 4-NP detection) of 8.56 nm/pM. Furthermore, the LOD was estimated from equation: LOD = 3SE/S, where SE is the standard error of the regression line and S is the sensor' sensitivity [10,25], leading to a promising LOD value of 0.17 pM (or equivalently, 2.36⋅10 − 11 µg/mL). The sensing mechanism can be generally explained in terms of a PANI-mediated process in which H + -terminated sites of PANI trigger the reduction of 4-NP to 4-hydroxylaminophenol, followed by subsequent oxidation of the latter to yield 4-nitrosophenol (see Fig. 7) [7,30]. Consequently, the catalytic properties of PANI/Pt bilayer take over a RedOx reaction, generating important changes within medium's RI through 4-NP conversion into 4-nitrosophenol, and causing further sensitive shifts within SPR spectral dips wavelength position, as observed within the calibration curve presented in Fig. 6B. These excellent performance indicators are also a consequence of an optimal PANI thickness (i.e. 95 nm) and of its particular roughened curly-like super cial morphology (Fig. 4B) that generates an overall increase of the FO-SPR sensor active surface [25], thus potentially providing a more e cient catalytic surface reaction between 4-NP and PANI lm.

Conclusions
In this work, an innovative PANI/Pt-coated FO-SPR sensor was used to determine the amount of 4-NP in DIW samples. The sensing area was fabricated by coating the uncladded FO core with Pt and subsequently depositing a thin sensitive PANI lm. The Pt plasmonic layer was evenly coated on the cylindrical FO silica core by DC magnetron sputtering and afterwards used as catalyst for the uniform and conformal polymerization process of PANI via a simple cost-effective electroless procedure. The sensitivity of the as-prepared Pt-coated FO-SPR sensors was rst evaluated in serial sucrose dilutions (of different RI units), owning to a value of 1515 nm/RIU, comparable with the traditionally-reported Aucoated FO-SPR sensors. Second, the PANI/Pt-coated FO-SPR sensors unveiled encouraging results when employed for 4-NP detection, as the LOD was estimated to 0.17 pM, with a sensitivity of 8.56 nm/pM, excellent performances obtained so far in respect with previous literature reports. These PANI/Pt-coated FO-SPR sensors may provide a broad interest for applications, especially in highly-sensitive real-time detection of extremely low trace-level pollutants.  Evaluation of the Pt-coated FO-SPR sensors performance. (A) SPR spectral dips obtained at 0 (red) and 12% (blue) sucrose concentrations; (B) Corresponding calibration curve measured at sucrose concentrations (0, 2, 4, 8 and 12% w/w) with the Pt-coated FO-SPR sensors. The error bars represent standard deviation (n=5) and R2 denotes the coe cient of determination.  FE-SEM micrographs of the Pt-coated FO-SPR surface before (A) and after (B) PANI electroless deposition for 6 h at 25°C. The inset of (A): low-magni cation SEM image of the FO-SPR sensor tip where the red mark denotes the area where the higher-magni cation SEM images (A and B) were captured; The inset of (B): corresponding closed-up magni cation SEM image of the PANI/Pt-coated FO-SPR sensor surface.

Declarations
For SEM observations, the conductivity of the PANI lm was slightly enhanced by protonating it for 10 min in a 1 M HCl solution.   Proposed sensing mechanism of the 4-NP with the PANI/Pt-coated FO-SPR sensor, based on the mediated PANI/Pt catalytic activity of 4-NP reduction.