Photodetection properties of populated Fe₃O₄@TiO₂ core–shell/Si heterojunction prepared by laser ablation in water

Abstract

In this work, we demonstrated the first study on the preparation and characterization of the populated Fe₃O₄@TiO₂ core–shell/Si photodetector by laser ablation in liquid. The structural and optical properties of Fe₃O₄ nanoparticles and Fe₃O₄@TiO₂ core–shell nanoparticles were studied by X-ray diffraction, transmission electron microscope, and UV–Vis absorption. X-ray diffraction findings suggest the formation of a populated crystalline Fe₃O₄@TiO₂ core–shell through the existence of XRD peaks related to TiO₂ and Fe₃O₄. The optical properties revealed that the optical energy gap of Fe₃O₄@TiO₂ was 3 eV, while the optical energy gap of Fe₃O₄ was 2.8 eV. Raman studies reveal the presence of vibration modes centered at 91 cm⁻¹ (E_g), 144 cm⁻¹ (E_g), 396 cm⁻¹ (B_1g), 512 cm⁻¹ (B_1g), 541 cm⁻¹ (B_1g + A_1g), and 609 cm⁻¹ (E_g) which are belong to the TiO₂. The vibration modes related to the magnetite Fe₃O₄ are observed at 145^–1 (T_2g), 302 cm⁻¹ (T_2g), and 554 cm⁻¹ (T_1g). Transmission electron microscope results suggest the presence of a core–shell morphology with an average size of 60 nm. The current–voltage characteristics of Fe₃O₄/p-Si and Fe₃O₄@TiO₂ core–shell/p-Si photodetectors are measured in the dark and under illumination conditions. The maximum responsivity of the Fe₃O₄@TiO₂/Si photodetector was 0.5A/W at 400 nm, while the maximum responsivity of Fe₃O₄/p-Si photodetector was 0.4A/W at 500 nm. The specific detectivity and external quantum efficiency of the Fe₃O₄@TiO₂/p-Si photodetector are larger than those of Fe₃O₄/p-Si photodetector.

1 Introduction

Nanomaterials have drawn attention due to their unique optical and electrical characteristics compared to those of the bulk state [1]. These materials are considered promising and efficient for a variety of industrial and technological applications, for example, renewable power converters, sensing applications, corrosion resistance, catalysts, solar cells, photodetectors, and chemical and biochemical sensors [2,3,4,5,6]. The core–shell structure is a nanoparticle core covered by a certain thin layer of semiconducting material (shell) in order to decrease the reactivity and improve the dispersibility of the core nanoparticles. Additionally, the shell layer can supply surface chemistry for further modification and functionalization of the core nanoparticles. The core–shell structure was used for photodetectors to improve the performance of the photodetector figures of merit and manipulate the peak response. High-performance Ag@PbS-based core–shell photodetector, Ag@PbI₂/Si, and Au@CuO/Si core–shell structure photodetector have been reported [7,8,9,10]. The Fe₃O₄@TiO₂ core–shell structure is promising, and it has been subjected to extensive studies. Due to their tunable magnetic properties, iron oxide (Fe₃O₄) nanoparticles are very attractive materials. Titanium dioxide (TiO₂) nanoparticles have been used in the synthesis of core–shell structures as a photocatalytic agent. As reported, the Fe₃O₄@TiO₂ core–shell structure is stable, environmentally friendly, and biocompatible, and by controlling the core diameter and shell thickness; it can be used for important applications [11]. Fe₃O₄@TiO₂ core–shell structures have been used for many potential biomedical and biochemical applications [12, 13]. Pulsed laser ablation in liquid (PLAL) is a process in which the target is immersed in liquid and irradiated with high intensity laser pulses that result in the production of colloidal nanoparticles. PLAL exhibits many advantages, such as stability, inexpensive, simplicity, attractive size distribution, different particle morphologies, and does not need a vacuum [14,15,16,17,18]. Herein, a novel route to the synthesis of Fe₃O₄@TiO₂ core–shell structures was demonstrated. The PLAL technique was used to fabricate Fe₃O₄@TiO₂/Si photodetector. The optoelectronic properties of the Fe₃O₄@TiO₂ core–shell/Si photodetector were measured and compared with those of Fe₃O₄ NPs/Si photodetector.

2 Experimental work

To synthesize Fe₃O₄@TiO₂ core–shell nanoparticles, Fe₃O₄ NPs were synthesized first by laser ablation of a high purity Fe₃O₄ pressed pellet (supplied by Alfa Aesar company) immersed in deionized water (DI). The Fe₃O₄ pellet was sintered at 1000 °C using a tube furnace. For ablation of iron oxide nanoparticles, a pulsed Nd: YAG laser with a wavelength of 1064 nm, pulse width of 7 ns, and laser fluence of 20 mJ/cm²/pulse was used. In the second step, a high purity TiO₂ powder (99.9%) (a mix of anatase and rutile tetragonal) purchased from BDH was pressed to prepare a pellet and then positioned inside the glass vessel filled with colloidal Fe₃O₄ nanoparticles and then irradiated with the same Nd:YAG laser with a laser fluence of 30 mJ/cm²/pulse. The focusing of the laser beam on the Fe₃O₄ and TiO₂ targets was performed using a converging lens with a focal length of 10 cm. The effective laser spot size was around 0.6 mm on the Fe₃O₄ and 0.7 mm on the TiO₂ target. Figure 1 shows the schematic diagram of the PLAL system used for the synthesis of Fe₃O₄@TiO₂ core–shell nanoparticles. The ablation time was adjusted to be 30 min for each step of ablation. The structures of Fe₃O₄ NPs and Fe₃O₄@TiO₂ core–shell NPs were studied using an X-ray diffractometer (XRD-6000, Shimadzu). The size and morphology of the nanoparticles were investigated using transmission electron microscopy TEM (EM208, Philips). An energy dispersive X-ray EDX equipped with a scanning electron microscope FE-SEM (Arya Electron Optic) was utilized to estimate the chemical composition of Fe₃O₄@TiO₂. The optical absorbance of the colloidal nanoparticles was measured using a spectrophotometer (Metertech, SP8001 Japan).

Fig. 1

Schematic illustration of laser ablation of Fe₃O₄@TiO₂ core–shell NPs experimental setup

To construct the photodetectors, a thin layer of Fe₃O₄ NPs and Fe₃O₄@TiO₂ core–shell NPs were deposited on mirror-like silicon substrates separately by the drop casting technique. The silicon substrate used was p-type with an orientation of (100) and had an electrical resistivity of 1–3 Ωcm. Cleaning the substrate was accomplished by the standard route. After the deposition, indium and aluminum were deposited on Fe₃O₄@TiO₂ and the backside of the silicon substrate, respectively, as ohmic contacts through a square mask of 1cm² area using a thermal evaporation system. The current–voltage characteristics of Fe₃O₄/Si and Fe₃O₄@TiO₂/Si heterojunction photodetectors were measured under dark ad illumination at room temperature. To investigate the spectral responsivity of the photodetectors, a monochromator (Jobin Yvon) was used in the spectral range of 350–800 nm after power calibration using a silicon power meter.

3 Results and discussion

The XRD patterns of TiO₂, Fe₃O_4, and Fe₃O₄@TiO₂ nanoparticles are shown in Fig. 2. Eight peaks for Fe₃O₄ located at 2θ = 17.5°, 28.4°, 30.1°, 33.4°,35.3°, 43.3°, 54.7°, 62.1°, and 68.2° corresponding to (111), (220), (102), (311), (222), (004), (224), (440), and (044) planes, respectively, were observed which were indexed to the cubic magnetite [19]. Some XRD peaks related to the Fe₂O₃ nanoparticles were also detected, which agrees with reported data [20]. The XRD pattern of TiO₂ powder confirmed the observed peaks belong to the tetragonal anatase phase (denoted by A) and rutile phase (denoted by R) TiO₂ according to JCPDs # 21–1272 and # 21–1276, respectively. As we can see, the XRD pattern of Fe₃O₄@TiO₂ core–shell nanoparticles exhibited eleven peaks. Four of them are located at 2θ = 25.8°, 27.4°, 40.3°, and 54.3° corresponding to (110), (100), (112), and (211) plane, respectively, which belong to TiO₂ NPs and the other peaks are indexed to Fe₃O₄ NPs [21]. Table 1 lists the crystallite size of Fe₃O₄ NPs and Fe₃O₄@TiO₂ nanoparticles for two dominant planes.

Fig. 2

XRD patterns of Fe₃O₄, TiO₂, and populated Fe₃O₄@TiO₂ core–shell NPs

Table 1 XRD analysis of Fe₃O₄ NPs and Fe₃O₄@TiO₂ core–shell NPs

The appearance of XRD peaks of Fe₂O₃ can be ascribed as follows: The repeating ablating of the target and creating fresh O atoms disrupts the O₂ partial pressure balance in the laser plume. When the partial pressure of oxygen in the plume changes, different iron oxides, such as Fe₂O₃, develop. Figure 3-a shows the optical absorption spectra of colloidal Fe₃O₄ and Fe₃O₄@TiO₂ core–shell NPs. The absorption in the UV region of Fe₃O₄@TiO₂ is higher than that of Fe₃O₄ due to the presence of a nanocomposite structure consisting of TiO₂ and Fe₃O₄. The absorption decreases sharply after 200 nm and saturates after 400 nm for two samples. A small absorption peak was observed at 250 nm due to the quantum size effect of nanosized Fe₃O₄ and Fe₃O₄@TiO₂ comes from the high energy of exciton. The freshly prepared colloidal Fe₃O₄ and Fe₃O₄@TiO₂ core–shell suspension is shown in the inset of Fig. 3-a. It is clear that the color of the colloidal nanoparticles was changed after ablation, from transparent to light pink. It is obvious that the color of the colloidal is mostly dependent upon the size and concentration of the nanoparticles in colloids. As shown in Fig. 3-b, the optical energy gap of the nanoparticles was determined from Tauc plot via plotting (αhν)² (α is absorption coefficient) against photon energy (hν) and extrapolation of the linear part to the x-axis gives the energy gap [22, 23]. The optical energy gap of Fe₃O₄ and Fe₃O₄@TiO₂ NPs was 2.8 and 3 eV, respectively, as shown in Fig. 3-b. The energy gap of the Fe₃O₄ increased after core–shell formation, indicating the presence of Fe₃O₄@TiO₂ core–shell structure.

Fig. 3

a Optical absorption of Fe₃O₄ NPs and Fe₃O₄@TiO₂ core–shell NPs. Inset is the photograph colloids and (b) is (αhν)² versus photon energy plot

The TEM images of Fe₃O₄ and Fe₃O₄@TiO₂ core–shell NPs are shown in Fig. 4. The TEM image of Fe₃O₄ shows the formation of small, aggregated, and agglomerated spherical nanoparticles due to the high surface energy and van der Waals attraction force [24, 25].

Fig. 4

a TEM image of Fe₃O₄ NPs, b TEM image Fe₃O₄@TiO₂ core–shell NPs. and c TEM image of mono-dispersed core–shell structure

The formation of particles of different sizes can be attributed as follows: The small particles produced inside the laser-induced plasma try to escape from the laser plume into liquid media and are cooled without any grain growth. While the particles remain inside the plasma, they grow and form large particles due to the high temperature [26]. The average particle size of Fe₃O₄ and populated Fe₃O₄@TiO₂ was determined using Image J software and found to be 22 nm and 60 nm, respectively. The TEM image shown in Fig. 4-b suggests the presence of the populated core–shell and free nanoparticles. Figure 4-c shows a TEM image of a mono-dispersed core–shell, it revealed that the TiO₂ shell surrounds several Fe₃O₄ cores nanoparticles. As shown in Fig. 4, not all the synthesized nanoparticles have a core–shell morphology, and the estimation of the yield of core–shell production is very important for device application, and this needs deep study and new characterization [27, 28]. Figure 5 shows the EDX spectra of Fe₃O₄ and populated Fe₃O₄@TiO₂ core–shell NPs.

Fig. 5

a EDX spectrum of Fe₃O₄ NPs and b EDX spectrum of Fe₃O₄@TiO₂ core- shell NPs

The EDX spectrum of Fe₃O₄ NPs showed the presence of only iron and oxygen elements, while the EDX spectrum of Fe₃O₄@TiO₂ core–shell NPs reveals the existence of iron, oxygen, and titanium elements. The origin of the Ti element comes from the presence of TiO₂ NPs in solution and from the shell, depending on the core–shell production yield. Figure 6 shows Raman spectra of Fe₃O₄ NPs and Fe₃O₄@TiO₂ core–shell NPs. The Raman spectra of populated Fe₃O₄@TiO₂ NPs showed the vibration modes centered at 144 cm⁻¹ (E_g), 396 cm⁻¹ (B_1g), 512 cm⁻¹ (B_1g), 541 cm⁻¹ (B_1g + A_1g), and 609 cm⁻¹ (E_g). These vibration modes are indexed to the anatase and rutile phases of TiO₂, which is in good agreement with the results of XRD. Three vibration modes belonging to magnetite Fe₃O₄ were observed at 94 cm⁻¹ (T_2g), 145^–1 (T_2g) and 554 cm⁻¹ (T_1g). Furthermore, Raman peaks related to the Fe₂O₃ phase are also observed in Fig. 6 [26, 29,30,31].

Fig. 6

Raman spectra of Fe₃O₄ NPs and Fe₃O₄@TiO₂ core–shell NPs

Figure 7 shows the dark and illuminated I-V characteristics of Fe₃O₄/Si and populated Fe₃O₄@TiO₂/Si heterojunction photodetectors measured at room temperature. The photodetectors exhibit rectification behavior, indicating the formation of a junction between the Fe₃O₄@TiO₂ layer and the silicon substrate. The forward current increases as the bias voltage increases due to the decrease in the depletion layer width [32], and the turn-on voltage of the Fe₃O₄/Si and Fe₃O₄@TiO₂/Si heterojunctions was 1.3 and 1 V, respectively. The forward current of the Fe₃O₄@TiO₂/Si photodetector is larger than that of the Fe₃O₄/Si due to the decrease in the electrical resistivity of Fe₃O₄ after making a nanocomposite with TiO₂ NPs.

Fig. 7

Dark and illuminated I-V characteristics of a populated Fe₃O₄ NPs/Si photodetector and b Fe₃O₄@TiO₂ core–shell NPs/Si photodetector. The upper inset is the semi-logarithmic relationship of ln I_f-Vs and lower inset is the cross sectional view of Fe₃O₄@TiO₂ core–shell NPs/Si photodetector

The reverse current of the two heterojunctions was slightly increased with reverse bias voltage and no breakdown was observed up to a bias voltage of 5 V. By using the following diode equation, the ideality factor (β) of the heterojunction was calculated by

$$\beta =\frac{q\Delta V}{kT \mathrm{ln}\frac{\Delta I}{{I}_{\mathrm{s}}}}$$

(1)

where I_s is the saturation current, which is determined from the semi-logarithmic I_f -V plot (inset of Fig. 7), k is the Boltzman coefficient, and T is the operating temperature. The value of β for Fe₃O₄/Si and Fe₃O₄@TiO₂ core–shell/Si heterojunctions were 7 and 3.5, respectively, indicating the presence of structural defects [33, 34]. The reverse current of the photodetector was increased after being illuminated with white light due to the generation of e–h pairs as a result of photon absorption in the sensitive area. The photocurrent of the Fe₃O₄@TiO₂ core–shell/Si is larger than that of Fe₃O₄ /Si by a factor of 1.6 at a bias voltage of 5 V and a light intensity of 30mW/cm² due to the increase in the light absorption as well as the decrease in the structural defects [35, 36]. On the other hand, the TiO₂ NPs act as a buffer layer that contributes to the reduction in the lattice constant mismatch between the Fe₃O₄ and the silicon substrate. As we can see from Fig. 7, increasing the light intensity causes an increase in the photocurrent of the photodetector as a result of the production of more e–h pairs in the depletion region. No saturation in the photocurrent was noticed after increasing the light intensity to 30mW/cm². This result indicates that the fabricated photodetectors have good linearity characteristics. The ON/OFF ratio is estimated as shown in Fig. 8, and it was found to be 65.5 and 105 for Fe₃O₄/Si and Fe₃O₄@TiO₂ core–shell/Si photodetectors, respectively. This result indicates that the presence of the TiO₂ shell significantly improved the photodetection ability and figures of merit of the Fe₃O₄/Si photodetector.

Fig. 8

ON/OFF ratio of Fe₃O₄/Si and Fe₃O₄@TiO₂ core–shell/Si photodetectors

The responsivity R_λ of the photodetector was determined using the following equation:

$${R}_{\lambda }=\frac{{I}_{\mathrm{ph}}}{P}$$

(2)

where I_ph is the photocurrent and P is the power of light at a certain wavelength. The variation of the responsivity with wavelength for the Fe₃O₄/Si and Fe₃O₄@TiO₂ core–shell/Si photodetectors at a bias voltage of 5 V is shown in Fig. 9. The responsivity plot of Fe₃O₄@TiO₂ core–shell /Si showed the presence of a peak response at 400 nm with a responsivity of 0.5A/W, while the Fe₃O₄/Si photodetector has a peak response at 475 nm with a responsivity of 0.4A/W. This blue shift can be explained on the basis that the energy gap of Fe₃O₄@TiO₂ core–shell is larger than that of Fe₃O₄ NPs. As is obvious from Fig. 9, the Fe₃O₄@TiO₂ core–shell/Si photodetector has a quantum efficiency greater than 100% in the spectral region (400-550 nm), while the Fe₃O₄/Si photodetector has a quantum efficiency of less than 100%.

Fig. 9

Spectral responsivity plots of Fe₃O₄NPs/Si photodetector and Fe₃O₄@TiO₂ core–shell NPs/Si photodetectors at 5 V bias

This improvement can be ascribed to the presence of core–shell structure, which contributes significantly to the decrease in the e–h recombination as a result of the presence of a high electric field. Table 2 lists the specific detectivity (D*) and the noise equivalent power (NEP) of Fe₃O₄/Si and Fe₃O₄@TiO₂/Si photodetectors. The specific detectivity of the photodetector can be estimated by

$${D}^{*}=\frac{{I}_{\mathrm{ph}}({A)}^{1/2}}{P (2e{I}_{\mathrm{d}}{)}^{0.5}}$$

(3)

where A is the sensitive area of the photodetector, e is the electron charge, and I_d is the dark current at a certain bias voltage. As shown in Table 2, the detectivity of the Fe₃O₄@TiO₂/Si was higher than that of the Fe₃O₄/Si photodetector, which means that the Fe₃O₄@TiO₂/Si can be used effectively to detect weak light signals. A comparison of the figures of merit of the fabricated Fe₃O₄@TiO₂/Si photodetector with other heterojunction-based silicon photodetectors of core–shell structure is found in Table 3.

Table 2 D* and NEP of the photodetectors

Table 3 A Comparison between the Figs. of merit of various Si-based core–shell photodetectors with Fe₃O₄@TiO₂/Si

4 Conclusion

We have demonstrated a novel route to synthesizing populated Fe₃O₄@TiO₂ core–shell/Si photodetector with a quantum efficiency higher than 100% at 400 nm without using a catalyst. Laser ablation of Fe₃O₄ in a solution of TiO₂ nanoparticles resulted in the production of a populated Fe₃O₄@TiO₂ core–shell structure. The structural and optical properties of Fe₃O₄ nanoparticles and Fe₃O₄@TiO₂ core–shell nanoparticles were investigated. XRD data confirmed the formation of crystalline Fe₃O₄@TiO₂. TEM studies suggest the presence of Fe₃O₄@TiO₂ core–shell nanoparticles with an average particle size of 60 nm. The optical energy gap of core–shell was larger than that of magnetite. The junction characteristics of the Fe₃O₄@TiO₂ core–shell/Si are better than those of Fe₃O₄/Si. The responsivity of the Fe₃O₄@TiO₂ core–shell/Si photodetector was 0.5A/W at 400 nm, while the responsivity of Fe₃O₄/Si was 0.4A/W at 500 nm. The external quantum efficiency and specific detectivity of the Fe₃O₄@TiO₂ core–shell/Si were higher than those of Fe₃O₄/Si. The technique used for core–shell is competitive, inexpensive, and suitable for the fabrication of high-performance photodetectors.