Nozzle-Jet-Printed Silver/Graphene Composite-Based Field-Effect Transistor Sensor for Phosphate Ion Detection.

High concentration of dissolved phosphate ions is the main responsible factor for eutrophication of natural water bodies. Therefore, detection of phosphate ions is essential for evaluating water eutrophication. There is a need at large-scale production of real-time monitoring technology to detect phosphorus accurately. In this study, facile enzymeless phosphate ion detection is reported using a nozzle-jet-printed silver/reduced graphene oxide (Ag/rGO) composite-based field-effect transistor sensor on flexible and disposable polymer substrates. The sensor exhibits promising results in low concentration as well as real-time phosphate ion detection. The sensor shows excellent performance with a wide linear range of 0.005-6.00 mM, high sensitivity of 62.2 μA/cm2/mM, and low detection limit of 0.2 μM. This facile combined technology readily facilitates the phosphate ion detection with high performance, long-term stability, excellent reproducibility, and good selectivity in the presence of other interfering anions. The sensor fabrication method and phosphate detection technique yield low-cost, user-friendly sensing devices with less analyte consumption, which are easy to fabricate on polymer substrates on a large scale. Besides, the sensor has the capability to sense phosphate ions in real water samples, which makes it applicable in environmental monitoring.

Introduction

Phosphorus is an essential macronutrient element, utilized by all living organisms for growth as well as for energy transport.[1] It often acts as a limiting nutrient in terrestrial and aquatic ecosystems for primary production.[2−4] Phosphorus, in the form of phosphate, plays an important role in photosynthesis, marine carbon cycle, and aerial carbon dioxide sequestration.[5] High levels of phosphorus in water systems can cause water eutrophication where growth of alga and planktons increases, causing algal bloom.[6] The eutrophication leads to the reduction or elimination of dissolved oxygen, which is extremely essential for aquatic living organisms. Some algal blooms release harmful toxins, which can badly affect human beings who consume this contaminated water or its products.[7−12] Hence, for the economical environmental protection, it is very crucial to use sensitive, selective, low-cost, online, and facile-technology-based phosphorus sensors featuring high accuracy to monitor phosphate levels in aquatic systems.

Significant efforts have been made to fabricate sensors for the development of phosphate-selective electrodes using several electroanalytical techniques.[13−20] Several optical methods including fluorescence and luminescence methods have been reported to detect orthophosphates with an accuracy of up to ∼0.01 mg/L only.[21−25] Although these methods express high limit of detection (LOD), they are still considered to be very tedious approaches. In the past, electrochemical amperometric sensors comprised of Ni, carbon black, and Co electrodes have been developed for phosphate detection.[26−30] Although the electrochemical method shows promising results, it suffers from some drawbacks such as low stability, electrode fabrication, and need of a reactive surface for phosphate ion detection. The ion selective electrode (ISE) method or the potentiometric sensor method is also reported for orthophosphate detection, and considerable efforts have been directed toward electrode membrane development.[31−34] The ISE method also suffers from limitations such as the limited lifetime of the membrane and high LOD. Recently, field-effect transistor (FET)-based sensors have attracted great interest in the sensing community due to their better performance and facile technology.[35−48] In a typical FET sensor, the source, drain, and channel electrodes work in combination to sense the target molecules by monitoring the electrical conductance change in the channel material. The channel material is so significant in FET sensor applications that its electronic properties and interaction with the adsorbed analyte strongly affect sensitivity. Graphene- and reduced graphene oxide (rGO)-based nanostructures exhibit excellent electronic properties and have been therefore used for several applications.[49−51] Recent graphene FET sensors for chemical sensing (including nitrate ions) have been explored with a better output.[52−54] Chen et al. reported a label-free micropatterned rGO film FET sensor that can detect selectively Cd2+ and Hg2+ ions in a real-time fashion.[55] Ikhsan et al. reported a facile synthesis of a graphene oxide–silver nanocomposite and its modified electrode for enhanced electrochemical detection of nitrite ions.[56] Recently, Mao et al. reported ultrasensitive rGO nanosheets and a ferritin-probe-based field-effect transistor sensor for the detection of orthophosphate ions with a detection limit of ∼0.026 μM.[57] However, most of the approaches use either very complicated techniques or polymeric additives to attach nanostructures and active probe materials on the electrode surfaces. Overall, a facile and cost-effective fabrication technique like nozzle-jet printing is required to fabricate highly reproducible sensors on large scale with excellent sensing performances.[58−60] From the last decade, the use of ink-jet printing technology has emerged as one of the powerful patterning tools for manufacturing electronic devices. The nozzle-jet printing method is just a simpler and modified form of ink-jet printing technique. This method can be used to print various kinds of materials such as Ag NPs, ZnO NPs, and NiO NPs.[60,61]

Herein, we reported nozzle-jet-printed Ag/rGO-composite-based enzymeless FET sensors on poly(ethylene terephthalate) (PET) substrates to detect phosphate ions.

The nozzle-jet printing method was used in this work preferably due to its facile operation, environmental friendliness, and compatibility with inks of different viscosities. The printed active Ag/rGO channel materials were well characterized using field-emission scanning electron microscopy (FESEM, Hitachi S4700), atomic force microscopy (AFM), X-ray diffraction (XRD), elemental mapping, and energy-dispersive X-ray spectroscopy (EDX) for their various properties. The FET sensor shows remarkable results for phosphate ion detection in low concentration range and with high sensitivity in a short time. The sensor also exhibits high selectivity toward phosphate ions and relatively low responses to TTP, Ca2+, Mg2+, K+, HCO3–, and SO42– interfering ions. The reproducibility as well as the long-term stability of the sensor device was also confirmed. The outstanding performance of nozzle-jet-printed Ag/rGO-based enzymeless FET sensors promises a large-scale industrial fabrication technology for phosphate ion monitoring in water systems.

Results and Discussion

Figure shows the FESEM images of the interfacial area between the nozzle-jet-printed Ag electrode and the Ag/rGO channel, low- and high-resolution nozzle-jet-printed Ag/rGO channels having a surface area of 0.05 cm2 (width ∼0.1 cm and length ∼0.5 cm), and as-printed Ag source/drain electrodes. FESEM images clearly reveal the smooth and easily distinguishable as-printed Ag electrode and Ag/rGO composite channel separately. The two-dimensional (2D) and three-dimensional (3D) AFM images of the nozzle-jet-printed Ag and Ag/rGO channel are shown in Figure a–d. It can be observed from both 2D and 3D AFM images that the roughness and particles size of the Ag/rGO composite are greater than those of the as-printed Ag electrode. This observation is supported by the root-mean-square roughness values of the as-printed Ag electrode (∼35 nm) and the Ag/rGO printed composite channel (∼54 nm). Figure e exhibits the 2D surface profile analysis of both the nozzle-jet-printed Ag and Ag/rGO channel with the former showing a thickness of ∼1.2 μm and the later of ∼1 μm. Figure a shows XRD patterns of the pure rGO powder and nozzle-jet-printed Ag and Ag/rGO channel. A weak broad-band peak at 2θ = 24.05° is observed, which is attributed to the (002) diffraction plane of rGO. Its peak intensity indicates the degree of graphitization of the rGO.[62] In addition, the nozzle-jet-printed Ag/rGO composite shows five diffraction peaks, which are attributed to the (111), (200), (220), (311), and (331) planes of Ag NPs with a face-centered cubic structure. Figure b demonstrates EDX of the nozzle-jet-printed Ag/rGO channel, confirming the presence of all of the three elements, i.e., carbon, oxygen, and silver, of the composite material. The purity and distribution of the as-printed Ag/rGO channel are further supported by its elemental mapping, as represented in Figure c–f.

Figure 1

FESEM images of the nozzle-jet-printed (a) Ag/rGO interface, (b, c) low- and high-resolution Ag/rGO channels, and (d) as-printed Ag source/drain electrode.

Figure 2

AFM images of the nozzle-jet-printed (a, b) Ag electrode and (c, d) Ag/rGO channel, and (e) 2D surface profile analysis of both nozzle-jet-printed Ag and Ag/rGO channels.

Figure 3

(a) XRD, (b) EDX, and (c–f) EDS elemental mapping of the nozzle-jet-printed Ag/rGO channel.

The electrical properties (Ids–Vds) of the Ag/rGO-based FET sensor were investigated before performing the actual phosphate ion detection as shown in Figure S1 (Supporting Information). The results clearly show an Ohmic contact between Ag electrodes and the Ag/rGO channel. Therefore, it can be considered that Ag/rGO is effectively printed between Ag electrodes without any deterioration in electrical contact and conductivity. The electrocatalytic properties of the nozzle-jet-printed Ag/rGO FET sensor were determined in 0.1 M N-(2-hydroxyethyl)piperazine-N′-ethanesulfonic acid (HEPES) buffer (pH 7.5) without and with 1 mM phosphate (Figure a). There was a substantial increase in the current response, with Ag/rGO showing a higher response than bare rGO, in the presence of phosphate ions. It has been reported earlier that the Ag/rGO composite demonstrates a synergetic catalytic property.[62] The change in drain current is entirely due to the conductivity and catalytic property of the Ag/rGO composite material in the presence of phosphate ions. The Ag NPs contribute to enhancing the drain current of the Ag/rGO film, which could be attributed to the lower work function of Ag NPs (4.2 eV) than that of rGO (4.65 eV). This leads to the formation of Ag+ ions, which, in turn, act as an active site for PO4– ion adsorption. The increased drain current can therefore be due to the redox reaction occurring at the surface of the sensor. Different concentrations of phosphate ion solutions were prepared in 0.1 M HEPES buffer (pH 7.5) to study the working performance of Ag/rGO FET sensor devices. Figure b shows the I–V response of the Ag/rGO FET sensor device toward different concentrations of phosphate ions. There was no current leakage observed when FET was exposed to phosphate solution. It was observed that the drain current increased gradually with phosphate ion concentration. The current changes noticeably for phosphate ions at a higher potential range, i.e., 1.3–3.0 V. The calibration plot was drawn after each experiment was performed thrice, and average current responses were taken at a gate voltage of 1.7 V (Figure c). Finally, linear range, sensitivity, and limit of detection were determined from the current–concentration-calibrated plot. From the slope of the calibrated plot, the sensitivity was found to be 62.2 μA/cm2/mM. The printed Ag/rGO FET sensor device showed linear range from 0.005 to 6.00 mM with a high regression coefficient (R2 = 0.99751) and low limit of detection (i.e., LOD ∼ 1.20 μM). The LOD was calculated by dividing the slope of the current–concentration plot (S) with the standard deviation (SD) of response.[63]

Figure 4

I–V response of the Ag/rGO FET sensor device toward solutions (a) without and with 1 mM phosphate and (b) different phosphate concentrations in 0.1 M HEPES buffer solution. (c) Calibration curve of the graph (b). The error bars show the mean value of three current responses.

The sensing mechanism of the phosphate ion detection sensor was elucidated and is represented in the below chemical equations.The sensing mechanism was further confirmed by performing the X-ray photoelectron spectroscopy (XPS) analysis of the Ag/rGO FET sensor, before and after treatment with phosphate ions as shown in Figure S2 (Supporting Information). The peak at 367.5 eV in Figure S2a (black spectra) indicates the availability of Ag+ ions in the Ag/rGO composite material. After treatment with phosphate ion solution, the silver peak shifts slightly to a higher binding energy value (369.5 eV). This indicates that Ag+ ions interact with the PO4– ions already available in the phosphate ion solution, resulting in the increase in drain current. The major shift in the peak position of silver (Ag 3d) and slight shifts in the peak positions of oxygen (O 1s) and carbon (C 1s) indicate the major role of silver in the entire electrocatalytic performance. Besides, the presence of the phosphorous (P 2p) peak at 135.6 eV (Figure S2d) further reveals the adsorption and redox reactions at the sensor surface. Comparison of the nozzle-jet-printed Ag/rGO FET phosphate ion sensor with the previously reported ones is summarized in Table . The overall performance of the Ag/rGO FET sensors reveals better capability than that of the previously reported ones.

Table 1

Comparison of the Ag/rGO FET Sensor Performance with That of the Previously Reported Phosphate Ion Sensors

electrode and material	detection method	linear range (μM)	sensitivity	detection limit (μM)	refs
screen-printed electrode	flow amperometry			0.60	(29)
cobalt (Co) microelectrodes	stirred batch amperometry	0.00001–0.001	33 mV/dec	1.0 × 10–5	(30)
reduced graphene oxide/ferritin field-effect transistor	I–V measurement			0.026	(57)
CB-SPE	cyclic voltammetry	0.5–100	0.12 μA/cm2/mM	0.10	(64)
CA/CN-PyOd-CoPC-SPCE	stirred batch amperometry	0.0000025–0.00013		2.00	(65)
GC/hybrid films containing molybdate anions in chitosan and ionic liquid	amperometry	0.79–31.5		0.79	(66)
GC/molybdovanadate in solution	differential pulse voltammetry	0.5–40		0.50	(67)
ZnO nanoflake–cobalt	cyclic voltammetry	1–100	13.2 μA/dec	1.00	(68)
fully printed Ag–rGO/Ag FET on PET	I–V measurement	5–6000	62.2 μA/cm2/mM	1.20	this work

The selectivity and interference test of the nozzle-jet-printed Ag/rGO FET sensors was carried out by separately detecting 1 mM each of phosphate and other individual interfering ions such as, SO42–, Mg2+, K+, HCO3–, Ca2+, and thiamine pyrophosphate chloride (TPP) in 0.1 M HEPES buffer (pH 7.4), as shown in Figure a. Their calibrated responses were plotted in the form of a histogram, at a gate voltage of 1.7 V, as presented in Figure b. It can be observed from the calibrated histogram that our fabricated sensors are selective for PO4– ion measurements in the presence of other interfering ions.

Figure 5

(a) I–V responses of the Ag/rGO FET sensor showing the response of phosphate ions and individual interfering ions (each 1 mM) in 0.1 M HEPES buffer solution, (b) calibrated histogram plot, where the response of phosphate ion was set as 100%. The error bars show the mean value of three current responses.

The reproducibility test was carried out by preparing five nozzle-jet-printed Ag/rGO FET sensors in similar conditions for the detection of 1.0 mM phosphate ions in 0.1 M HEPES (pH 7.4) solution by I–V method, as shown in Figure a. The calibrated histogram in the inset of Figure a reveals a relative standard deviation (RSD) of ∼2.5%. The stability of a single Ag/rGO FET sensor, stored under normal conditions, was checked for four consecutive weeks as shown in Figure b. The corresponding calibrated plot in the inset of Figure b confirmed excellent stability with a current loss of ∼2.0% during 1 mM phosphate ion detection. Again, all calibrated drain current responses were recorded at a gate voltage of 1.7 V. In general, the present nozzle-jet-printed Ag/rGO FET sensor device exhibited excellent reproducibility and stability, which prompts the use of nozzle-jet-printed nonenzymatic FET sensor for real-time detection of phosphate.

Figure 6

(a) I–V response of five Ag/rGO FET sensor devices and (b) I–V response of a single Ag/rGO FET sensor device in 0.1 M HEPES buffer solution containing 1 mM phosphate solution. The error bars show the mean value of three current responses.

The standard addition-recovery-based method was utilized for practical application to detect the phosphate ions in a real water sample using the Ag/rGO FET sensor, and the results are shown in Figure . The water sample collected from a local lake revealed 0.05 ± 0.015 mM phosphate ions with the Ag/rGO FET sensor. Practically, known concentrations of phosphate salt were added to four samples of lake water, and the phosphate ion concentrations were subsequently measured. The results showed satisfactory recoveries of about 100% for the added known phosphate ion concentrations (i.e., 0.07, 0.1, 0.3, and 0.5 mM), as represented in Table . Thus, it can be concluded that the developed nozzle-jet-printed Ag/rGO FET nonenzymatic sensor displays a unique and potential reliability for the detection of phosphate ions in real water samples.

Figure 7

I–V responses of the standard addition detection method for phosphate ions in real water samples by the Ag/rGO FET sensor device.

Table 2

Phosphate Ion Detection in Real Water Samples

real sample	added concentration (mM)	measured concentration (mM)	recovery (%)	RSD% (n = 3)
sample 1 (lake water)		∼0.05		2.21
sample 2	0.02	∼0.07	100.0	3.14
sample 3	0.05	∼0.10	99.8	4.22
sample 4	0.25	∼0.30	99.9	3.25
sample 5	0.45	∼0.50	99.9	3.30

Conclusions

We have successfully fabricated a nozzle-jet-printed Ag/rGO FET enzymeless phosphate ion sensor on flexible PET substrates. The Ag precursor ink and Ag/rGO hybrid ink were prepared to print FET devices with good adhesion. The Ag/rGO FET sensor demonstrated excellent performance, showing high sensitivity (62.2 μA/cm2/mM), wide linear range (0.005–6.00 mM), and a very low detection limit (∼0.20 μM). The Ag/rGO FET sensor exhibited long-term stability, good selectivity, high reproducibility, and applicability in real water samples with high accuracy. Additionally, the proposed method can be employed for developing various mineral ion sensors for large-scale production. Hence, the nozzle-jet printing technique can provide an efficient strategic way for the development of cost-effective, large-scale production, highly efficient, and environmentally friendly sensing electrodes for various electroanalytic applications.

Experimental Section

Chemical Reagents

Silver acetate [Ag(OCOCH3), 99%], ethylamine (C2H5NH2, 2.0 M in methanol), ammonium hydroxide (NH4OH, 28.0–30.0% NH3 basis), formic acid (HCOOH, 98.0%), graphite (Sigma-Aldrich, Asbury Carbon, 3061), potassium permanganate (KMnO4, 99.0%), H2O2 (30 wt % in water), potassium phosphate monobasic (KH2PO4, 99.0%), MgCl2, KCl, CaCl2, Na2SO4, thiamine pyrophosphate chloride (TPP), 0.1 M N-2-hydroxyethylpiperazine-N-2-ethanesulfonic acid (HEPES), ethanol, ethanolamine (99.0%), and hydroxyethyl cellulose (HEC, 2% in water–methanol solution) were purchased from Sigma-Aldrich. All aqueous solutions were prepared using Milli-Q water purified deionized water.

Formulation of Hybrid Ag/rGO Ink

Ag/rGO hybrid ink was formulated from silver and rGO inks as shown in Scheme a. Prior to the formulation of Ag/rGO composite ink, silver ink and rGO powder were synthesized separately as discussed in our earlier reports.[61,69] Primarily, rGO ink was synthesized by mixing 10 mg of rGO fine powder in 4.5 mL of deionized water and 4.5 mL of ethanol followed by sonication for 30 min. Subsequently, ethanolamine (0.2 mL) and 2% HEC (1 mL) were added separately as dispersing agents. Additionally, the additives (ethanolamine and 2% HEC) maintain the viscosity of the ink and avoid clogging problems of the ink in the nozzle during the printing process. The whole solution was sonicated for around 10 h and then kept standing for 30 min. The supernatant of well-dispersed rGO solution was transferred into another glass vial, and finally, 2 mL of silver precursor ink was thoroughly mixed in it. The obtained black-colored solution represents the hybrid Ag/rGO ink.

Scheme 1

(a) Schematic Representation of Ag/rGO Hybrid Ink Formulation, (b) Nozzle-Jet-Printed Ag/rGO-Based FET Sensor, and (c) Optical Image of Printed Ag/rGO-Based FET Sensors on the PET Substrate

Fabrication of Ag/rGO FET Sensors

In the FET sensor (Scheme b), source and drain electrodes were fabricated by nozzle-jet printing technology using Ag precursor ink on the PET substrate at a nozzle speed of 4000 mm/s and a nozzle pressure of ∼40 kPa. The printed electrodes were annealed at 90 °C for 1 h. Thereafter, a channel of Ag/rGO active material with a width of 0.1 cm and a length of 0.5 cm was twice overprinted between source and drain Ag electrodes at a nozzle speed of 4000 mm/s and a nozzle pressure of ∼50 kPa. The thickness of the Ag/rGO channel can be controlled either by varying nozzle moving speeds (1000–8000 mm/s) or using nozzles of different diameters. However, in the latter case, the viscosity of solution needs proper optimization. The fully printed devices were annealed at 130 °C for 1 h on a hot plate. After annealing, the Ag NPs embedded within rGO sheets provide excellent adhesion on the PET substrate. An optical image of nozzle-jet-printed Ag/rGO-based FET sensor devices has been shown in Scheme c.

Characterization

The surface morphology of electrodes was characterized with field-emission scanning electron microscopy (FESEM, Hitachi S4700) and atomic force microscopy (AFM). X-ray diffraction (XRD), elemental mapping, and energy-dispersive X-ray (EDX) spectroscopy were used to analyze the high crystallinity, element distribution, and element composition of the printed Ag/rGO, respectively. The thickness of the printed features was measured by a 2D surface profiler with a scan length 1.2 μm, scan speed 100 μm/s, and sampling rate of 50 Hz. All of the electrical characterizations were performed using a probe-station-attached semiconductor parameter analyzer (model: HP 4155A) in ambient conditions. For analyte detection, a floating gate (Ag/AgCl) was utilized to measure the drain current with a fixed drain–source voltage of 0.1 V, whereas the gate–source voltage was increased in the range of 0–3.0 V. The different concentrations of analyte solutions (each 10 μL) were pipetted into the active area/channel area between source–drain electrodes, and electrical measurements were immediately performed. For each measurement, the electrodes were rinsed in a buffer prior to reuse. All experiments were performed in 0.1 M HEPES buffer (pH = 7.5) solution containing different phosphate concentrations.