Fiber optic magnetic field sensor using Co doped ZnO nanorods as cladding

A fiber optic magnetic field sensor is proposed and experimentally demonstrated. Pristine and Co doped ZnO nanorods of different Co concentrations (5, 10, 15 and 20 at%) were synthesized using a hydrothermal method. The synthesized nanorods were subjected to various characterization methods like X-ray diffraction (XRD), optical absorption, scanning electron microscopy, energy dispersive X-ray spectroscopy, Fourier transform infrared spectroscopy, vibrating sample magnetometry and X-ray photoelectron spectroscopy (XPS). XRD and XPS analysis confirms that the Co ions were successfully incorporated into the Zn site of the wurtzite ZnO lattice without altering the structure. The pristine and Co doped ZnO nanorods showed remarkable changes in the M–H loop where the diamagnetic behavior of ZnO changes to paramagnetic when doped with Co. The sensor structure is composed of cladding modified fiber coated with Co doped ZnO nanorods as a sensing material. The modified cladding is proportionally sensitive to the ambient magnetic field because of the magneto-optic effect. Experimental results revealed that the sensor has an operating magnetic field range from 17 mT to 180 mT and shows a maximum sensitivity of ∼18% for 15 at% Co doped ZnO nanorods. The proposed magnetic field sensor would be attractive due to its low cost fabrication, simplicity of the sensor head preparation, high sensitivity and reproducibility.


Introduction
Nanostructured metal oxides have acquired a signicant role in the scientic world due to their efficient technological applications in the eld of solar cells, optoelectronic devices, gas sensors and spintronics. [1][2][3][4][5] ZnO is a unique wide band gap (3.37 eV) semiconductor with a large exciton binding energy (60 meV) and good chemical stability. 6 The inception of impurities into the ZnO host is an efficient way to produce novel properties in ZnO. Diluting non-magnetic ZnO with transition metal dopants such as Mn 2+ , Co 2+ , Ni 2+ and Fe 2+ could deliver different magnetic properties. This new genre of semiconductor is entitled as dilute magnetic semiconductors. 7,8 In the last few decades, many researchers have curiously investigated the impact of doping ZnO nanostructures with TM ions. Among the various TM ions, Co has its own importance in that (i) it is more reconcilable with Zn while doping, (ii) it can amend the morphology and optical properties of ZnO nanostructures, (iii) easily soluble in ZnO nanostructures, (iv) it has a strong magnetic moment (m Co ¼ 1.8 m B ) compared with other transition metals and (v) the ionic radius of Zn 2+ (0.60Å) is nearly same as Co 2+ (0.58Å). 9 Many research groups have synthesized Co doped ZnO nanostructures and assessed their performance on doping. For instance, Sharma et al. have synthesized Co : ZnO nanoparticles using a co-precipitation method and showed their ferromagnetism at room temperature. 10 Xu et al. have synthesized Co doped ZnO nanoakes using a hydrothermal method and noticed their ferromagnetism at room temperature. However, with increasing Co 2+ dopant concentration, paramagnetism was exhibited. 11 Qiu et al. have also reported room temperature ferromagnetism (RTFM) in Co : ZnO by water-bubble template process. 12 In the past few years, several techniques have been ratied to synthesize pristine ZnO and TM doped ZnO nanostructures including co-precipitation, 13 hydrothermal, 14 sol-gel, 15 magnetron sputtering, 16 ball milling, 17 etc. Of these, the hydrothermal method offers various nanostructure morphologies, controlled particle size, high reaction rate, different phase formation, etc. 18 To explore the potential of nanoparticles in the eld of magnetic eld sensing, various types of sensor such as thin lms, 19 ber optics, 20 Hall effect based magnetic eld sensors and magnetic switches have been fabricated. Though, ber optic based magnetic sensors offer numerous advantages such as small size, on-line analysis, remote sensing, high sensitivity, immunity to electromagnetic interference and a capability of working in harsh environments. 21 In ber optic sensor, the principle of sensing is based on cladding modication technology, in which the middle part of the cladding was replaced with a magnetic sensing material. The magnetic eld sensing material causes the magneto-optic effect such as change in refractive index as a result of an applied external magnetic eld.
Hitherto, various kinds of ber optic based magnetic sensor have been proposed using magnetic uid (MF) as the cladding with different optical devices such as Fiber Bragg Grating (FBG), 22 Long Period Fiber Bragg Grating (LPFG), 23 singlemode-multi-mode-single-mode (SMS) structures, 24 multimode-single-mode-multi-mode (MSM) structures, 25 micro resonators, 26 and Fabry-Pérot devices. 27 However, minor shortcomings like the feeble response of MF to external magnetic elds, low interaction between core mode/lower order cladding mode and MF, structure instability and complicated technology mean that they fail to achieve higher magnetic sensitivity.
To overcome these shortcomings, a ber optic magnetic eld sensor based on Co doped ZnO nanorods is proposed and demonstrated. The impact of doping concentration on structural, optical and magnetic properties is also investigated.

Synthesis of pristine and Co doped ZnO nanorods
Pristine ZnO and Zn 1Àx Co x O (x ¼ 5, 10, 15 and 20 at%) were synthesized using a hydrothermal method. All the chemicals used in this work were analytical reagent grade and used as received without further purication. Zinc acetate dihydrate (Zn(CH 3 COO) 2 $2H 2 O, 99.99%, Sigma-Aldrich) and cobalt(II) chloride, (99.99%, Sigma-Aldrich) in suitable weight percentages were used as starting materials. Additionally, sodium hydroxide (NaOH, 99.99%, Alfa Aesar) was used as the precipitating agent and cetyl trimethyl ammonium bromide (CTAB) (C 19 H 42 BrN) was used as a surfactant.
For the synthesis of pristine ZnO nanorods, 50 ml of 0.2 M zinc acetate dihydrate, 50 ml of 0.1 M CTAB and 100 ml of 2 M NaOH aqueous solution were prepared. Initially, zinc acetate dihydrate solution was stirred continuously for 3 h and aqueous NaOH solution was added dropwise to the mixture until the pH value of solution reaches 8.0. Later, the white precipitated solution was stirred for 1 h. Finally, CTAB solution was added into the mixture and stirred vigorously for another 30 min. Then, the nal solution was transferred into a 100 ml teonlined autoclave and maintained at 170 C for 72 h. Aer the reaction was completed, the resultant product was washed several times with distilled water and ethanol alternatively, and dried at 60 C overnight.
For the synthesis of Zn 1Àx Co x O (x ¼ 5, 10, 15 and 20 at%) nanorods, the calculated amount of cobalt(II) chloride and zinc acetate dihydrate were dissolved in 50 ml of distilled water and stirred for 3 h. Then, aqueous NaOH solution was added drop wise into the above mixture to form a white precipitate with a slight pale pink colour. Aer that the same procedure that was adopted for the synthesis of pristine ZnO nanorods was followed for the synthesis of Co doped ZnO nanorods. ber of about 42 cm length with 1000 mm diameter and 0.51 numerical aperture was used as a sensor head. A cladding region of about 3 cm in length was mechanically removed at the center portion of the ber using a chemical etching process followed by ne polishing. The ablated cladding region was chemically etched using acetone followed by polishing with a 1000 grid sheet. Then, the polished surface was cleaned and coated with Co (at% of 5, 10, 15 and 20) doped ZnO nanorods using a dip coating method.

Sensor head preparation and setup
The scheme of the experimental setup used for measuring the magnetic eld sensing characteristics of the fabricated sensor is shown in Fig. 1(b). In the sensor setup, a broadband light source (halogen lamp-SLS201/M) with the wavelength ranging from 300 to 2600 nm is coupled at one end of the ber and the intensity spectrum was recorded with a ber optic spectrometer (CCS200/M) having a spectral range of 200 to 1000 nm at the other end of the ber. The sensor head was inserted between the two poles of an electromagnet which generates a static magnetic eld around the sensing head. The experiment was conducted at ambient temperature (28 C).  are very well matched with standard JCPDS card no: . No other secondary phases of Co clusters were found. The nonexistence of impurity peaks reveal that Co ions were successfully incorporated into the ZnO lattice. The decrease in diffraction peak intensity is due to an increase in the concentration of impurities. 28 The average crystallite size and strain of the nanorods are calculated from the Williamson and Hall (W-H) plot. 29 The average crystallite size and strain of the nanoparticles are shown in Table 1. It was found that due to the incorporation of impurities, the crystallite size decreases. Fig. 3(a-j) shows the morphologies of pristine and Zn 1Àx Co x O nanoparticles at different magnications. It is seen that the synthesized nanopowder shows a rod like morphology. In order to analyse the complete morphology information of the ZnO nanorods, a plausible formation mechanism for the ZnO nanorods was proposed, as schematically illustrated in Fig. 4.

Morphological and elemental analysis
The formation mechanism of Co doped ZnO nanorods can be invoked via the following chemical reactions: Decomposition of Zn(Ac) 2 : Production of pristine ZnO nanoparticles: Decomposition of CoCl 2 : Production of Co doped ZnO nanoparticles:  Reaction (1) and (5) shows the decomposition of Zn(Ac) 2 and CoCl 2 in an ambient environment producing Zn(OH) 2 and Co(OH) 2 , respectively. If the pH value in aqueous solution is about 11, where Zn(OH) 2 is the main chemical compound, during the hydrothermal method, a part of the Zn(OH) 2 colloid species dissolves into Zn 2+ and OH À according to reaction (2). When the concentration of Zn 2+ and OH À reaches the super saturation degree of ZnO, then ZnO nuclei will form according to reaction (3) and (4). Reaction (6) indicates the dissolution of Co(OH) 2 in water to produce Co 2+ ions that will be incorporated into the ZnO lattice. Reaction (7) is the nal step of the growth process to accomplish Co doped ZnO nanoparticles. 30 To obtain the consistent nanoparticle size, the headway approach to synthesize Co doped ZnO nanoparticles using a hydrothermal method is closely related to a new understanding of the formation mechanism by the introduction of the surfactant CTAB. CTAB is a cationic surfactant having small hydrophilic head and a hydrophobic tail. The element Zn was obtained in [Zn(OH) 4 ] 2À as a negatively charged tetrahedrons that were formed according to reaction (3), whereas CTA + was positively charged with a tetrahedral head. At a higher reaction temperature ($170 C), the cationic charge of CTA + and anionic charge of [Zn(OH) 4 ] 2À were formed primarily by electrostatic interactions (reaction (4)). The CTAB could accelerate the ionization of [Zn(OH) 4 ] 2À as it is a strong acid-weak-base salt. 31 It was assumed that the CTAB was aggregated in between the ZnO crystallites during hydrothermal crystallization and aer washing thrice with ethanol, the rod like structure of ZnO was formed.
The elemental composition of pristine and Zn 1Àx Co x O nanorods were investigated using EDS spectra and shown in Fig. 5(a-e). The spectra affirm the presence of Zn and O in ZnO, likewise, Zn, O and Co in Co doped ZnO nanorods. Fig. 6 shows the elemental mapping of Zn, O and Co on a single Co doped ZnO nanorod. The mapping results indicated that Zn, O and Co atoms are uniformly distributed on the single nanorod. Fig. 7 shows the vibrational frequencies of pristine and Co doped ZnO nanorods. The peak obtained between 3420-3650 cm À1 corresponds to the stretching vibration of O-H. 32 The stretching vibrations of Zn-O are observed at 414, 422, 428, 457 and 466 cm À1 for pristine and Co doped ZnO nanorods. 33 Upon Co doping, the IR peak shis consistently from 414 cm À1 to 466 cm À1 and this shi is attributed to the incorporation of Co 2+ ions in the ZnO lattice. 34     35 This affirms the incorporation of Co 2+ ions into the ZnO lattice, rather than forming as cobalt oxide (CoO) or Co metal. Fig. 9 shows magnetization versus magnetic eld (M-H) measurements with maximum applied eld AE10 kOe at room temperature (300 K). All the Co doped ZnO nanorods revealed paramagnetic behaviour excluding pristine ZnO (diamagnetic behaviour). The diamagnetic behaviour of pristine ZnO is ascribed to the presence of paired electrons in its d orbital. The 5% and 10% Co doped ZnO nanorods (ZC1 and ZC2) deduces weak ferromagnetism (Fig. 9(b)). It is suggested that a few of the   Paper doped Co 2+ cations occupy the next nearest lattice sites. The nearest Co-Co pairs couple in an antiferromagnetic way and suppress the magnetization. [36][37][38][39] Thus, weak ferromagnetism was observed in the ZC1 and ZC2 samples. When the level of Co doping increases to 20%, the sample demonstrates linear magnetization curves with the absence of a hysteresis loop within the applied eld (ZC3 and ZC4), which can be concluded as good paramagnetism. As Co dopant concentration increases, more and more nearest Co-Co pairs exhibit larger antiferromagnetic interactions, which leads to the vanishing of hysteresis in the ZC3 and ZC4 samples. This implies that higher Co doping concentrations in ZnO lead to paramagnetism. 40 Similar paramagnetic behaviour with the same composition has been reported by many research groups. [41][42][43][44][45] However, for higher magnetizing elds the magnetization is found to increase with increasing Co concentration (Fig. 9(c)).

XPS analysis
Furthermore, to conrm the presence of elements and its chemical bonding states in synthesized Co doped ZnO nanorods (x ¼ 0.15), XPS analysis has been carried out. Fig. 10(a) depicts the full range survey spectrum of Co doped ZnO nanorods (x ¼ 0.15), which reveals the presence of characteristic peaks of Zn, O and Co in the synthesized Co doped ZnO nanorods. The selected data was corrected with C 1s carbon contamination peak (284.6 eV). Fig. 10(b) shows the high resolution spectra of the Zn 2p energy state. The core level binding energy of Zn 2p 3/2 and Zn 2p 1/2 was observed at 1021.9 eV and 1044.9 eV, respectively. The energy difference between these two peaks ($23 eV) conrms that Zn exists primarily in the Zn 2+ chemical state. 46 Fig. 10(c) depicts three distinct characteristic peaks of the O 1s energy state observed at 528.3 eV, 530.1 eV and 531.8 eV, which are ascribed to the formation of three different O species in the synthesized nanorods. The lower binding energy implies that lattice oxygen in hexagonal wurtzite structure is surrounded by zinc and cobalt ions. The medium binding energy affirms the presence of oxygen vacancies in the ZnO matrix. The higher binding energy attributes the formation of adsorbed oxygen (O 2À ) on the surface of the ZnO nanorods. 47 Fig . 10(d) shows the Co 2p energy state high resolution spectra. The two characteristic peaks of Co 2p 3/2 and Co 2p 1/2 , located at 780.4 eV and 796.4 eV, respectively, have been observed. In the high resolution spectra of Co 2p, the energy difference between Co 2p 3/2 and Co 2p 1/2 energy states is $16 eV, clearly showing that Co was successfully incorporated as a divalent ion. This is very well matched with previous literature reports. [48][49][50] In addition to this, two shake-up satellite peaks (S1 and S2) were observed at 785 eV and 796.4 eV along with the main characteristic peaks of the Co 2p energy state. Thus, XPS analysis clearly reveals that Co 2+ was successfully incorporated into the ZnO lattice by substituting Zn 2+ without any additional impurities or phases.

3.7.
Magnetic eld sensing analysis 3.7.1. Sensing mechanism. Fig. 11 shows the magnetic eld sensing mechanism in the proposed ber optic sensor. In the clad modied ber optic sensor, the output light spectral variation with applied magnetic eld leads the evanescent wave absorption in the modied cladding due to the change in the refractive index. When light travels with total internal reection at the interface of the core and modied cladding, not all of the light intensity is reected back but a part of it penetrates into the cladding material and its intensity decays exponentially away from the interface. This phenomenon is called an evanescent eld. 51 When an external magnetic eld was subjected transversely to the direction of incident light and interacts with the evanescent eld, the transmitted light spectrum and the output light intensity of the sensor varies. The reason for the change in intensity variation for the cladding modied ber in an external magnetic eld is based on the following relation, [52][53][54]   From the equation, 3 r is the dielectric constant and c is the electric susceptibility. When an external magnetic eld is applied perpendicularly to the direction of the propagated light, then vc vH . 0 So the refractive index of Co doped ZnO nanorods will increase with increasing in magnetic eld strength. In the present work, a part of the cladding (n clad ¼ 1.402) was replaced with the synthesized pristine and Co doped ZnO nanorods (n ZnO ¼ 1.91) which contributes to a certain amount of decreased attenuation in the guided signal depending on the absorbance of the cladding.
Hence, the change in refractive index and the absorbance of the cladding should affect the total internal reection and evanescent eld respectively. This is reected in the change in intensity of the signal which is guided along the ber. 55 The proposed ber optic sensor works in a leaky mode condition as the refractive index of the modied cladding (n ZnO ¼ 1.91) is higher than that of the core (n core ¼ 1.492).
3.7.2. Spectral analysis. Fig. 12(a-e) shows the magnetic eld sensing characteristics of pristine and Co doped ZnO nanorods with the magnetic eld ranging from 17.2 mT to 190.6 mT. The spectra exhibit three peaks around 693, 772 and 946 nm which are characteristic of the optical ber used. These spectra suggest that the spectra only undergo intensity variation for different magnetic eld strengths. It is clearly seen from the gure that the spectral intensity increases monotonically with increasing applied magnetic eld strength. This is due to the decrease in evanescent wave absorption with increasing magnetic eld. The light intensity in the absence of a magnetic eld is taken as a reference.
3.7.3. Sensitivity analysis. The sensitivity of the proposed magnetic eld sensor is calculated using the following relation, 56 where I a is the intensity in the absence of a magnetic eld and I m is the intensity in the presence of a magnetic eld. Fig. 13 shows the magnetic eld sensitivity plot of pristine and Co doped ZnO nanorods in different magnetic eld ranging from 17.2 mT to 190.6 mT at an ambient temperature of 28 C. From Fig. 13, it can be seen that Co doped ZnO nanorods (x ¼ 0.05, 0.1, 0.15 and 0.2) show enhanced sensitivity compared to pristine ZnO. Particularly, Co doped ZnO nanorods (x ¼ 0.15) exhibit a maximum sensitivity of $18% compared with that of other Co doped ZnO nanorods (x ¼ 0.05, 0.1 and 0.2). On the basis of the data plotted in Fig. 13, we can summarize that the sensor response increases fairly up to a Co doping of x ¼ 0.15, then decreases for higher doping concentrations. It is seen that the materials with higher magnetization possess high sensitivity.
Further, beyond the magnetic eld of 180.8 mT, the sensor showed a saturated response. Therefore, the operating range for the proposed magnetic sensor is in the range of 17.2 mT to 180.8 mT.
In order to study the effect of reproducibility, the sensor was characterized over ten cycles with a magnetic eld strength of 190 mT and the experimental results are shown in Fig. 14. It is seen that the sensing response for all cycles are nearly identical with insubstantial sensitivity uctuations. This clearly depicts that the sensor has good reproducibility.

Conclusions
A ber optic magnetic eld sensor using Co doped ZnO nanorods (5, 10, 15 and 20 at%) has been proposed and experimentally demonstrated. The analysis conrms that the nanorods are in hexagonal wurtzite structure. Furthermore, the VSM analysis indicates that the Co doped ZnO nanorods exhibit weak ferromagnetism at lower doping levels (ZC1 and ZC2) and paramagnetism at higher doping levels (ZC3 and ZC4). Experimental results show that the sensor has an operating magnetic eld range from 17.2 mT to 180.8 mT and revealed the maximum sensitivity of $18% for Co doped ZnO nanorods (x ¼ 0.15) which shows higher value of magnetization. The proposed magnetic eld sensor will serve as a better platform for the development of ber optic magnetic eld sensors as better replacements for current Hall effect based sensors.

Conflicts of interest
There are no conicts to declare.