Magnetic field-induced self-assembly of chemically modified graphene oxide on cellulose fabrics for the fabrication of flexible conductive devices

Graphite was purchased from Sigma-Aldrich. In addition, Iron(III) chloride hexahydrate, Iron(II) chloride tetrahydrate, Silane coupling agent KH-560, sodium hydroxide, hydrochloric acid, methanol, toluene, HI acid, sulfuric acid, hydrogen peroxide, 1-ethyl-3-(3-dimethylaminopropyl) carbodiimide hydrochloride (EDC), sodium nitrate and potassium permanganate were supplied by Sinopharm chemical reagent Co., Ltd. Cellulose fabric was purchased from Suzhou Shuishan Industrial Co., Ltd, Jiangsu, China and was viscose woven structure with 141 g/m².

The solid sample of NaOH (5.0 g) was added into the deionized water (250 mL), heated to 80 °C. The samples, including FeCl₃·6H₂O (6.75 g), FeCl₂·4H₂O (2.48 g), the deionized water (25 mL) and HCl (0.83 mL) were mixed respectively into the heated NaOH solution and loaded into the three-necked flask. The reaction was continued for 1 h under the protection of nitrogen (Han et al. 2007; Kim et al. 2011). Then, the solution was vacuum-filtered, resulting in black slurry. The obtained slurry was washed with water and methyl alcohol many times until the pH of the slurry was 7. The magnetic Fe₃O₄ nanoparticle was finally obtained. To endow Fe₃O₄ with amino groups, the obtained magnetic Fe₃O₄ nanoparticle and silane coupling agent KH-560 (named as APTES) were successively dispersed in the mixed solution consist of methyl alcohol and methylbenzene at 110 °C for 10 h (Wang et al. 2009). The volume ratio of methyl alcohol to methylbenzene is 1.0 (70 mL:70 mL). Then, the APTES/Fe₃O₄ nanoparticle was obtained.

The graphene oxide (GO) sheets were synthesized from graphite flakes using a modified Hummers’ method, as reported previously (Hummers;Offeman 1958). The weight ratio of graphite, potassium permanganate (KMnO₄) and sulfuric acid (H₂SO₄, 98%), was 1:3:14. Then the mixture was stirred at 0 °C for 2 h and heated to 45 °C for 30 min. Then the deionized water (280 mL) and H₂O₂ (10 mL 30%) were added and stirred for 15 min. After the reaction completely static, the supernatant liquid was removed and the residues was washed with hydrochloric acid (HCl) and water, then, centrifuged and washed with water many times to neutral. The GO solution was finally obtained.

To produce the GOF solution, the obtained GO (2 g) was diluted with water (100 mL) and then, the APTES/Fe₃O₄ nanoparticle was added according to the ratio of GO to APTES/Fe₃O₄ nanoparticle. The weight ratio of GO and APTES/Fe₃O₄ was respectively 95:5, 90:10, 85:15 and 80:20.

GO-coated and GOF-coated fabrics were prepared via the multi-dipping-drying treatment of GO and GOF dispersions on the cellulose fabrics (warp: 25 cm; filling: 2 cm), as shown in Fig. 1. Firstly, the GO solution (250 mL) with mass fraction of 0.61% was diluted in water. Then, the cellulose fabrics were dipped into the prepared dispersion, sonicated for 30 min at room temperature, without the external magnetic field and dried at 50 °C for 12 h. This process was known as the assembly of the GO layers onto cellulose fabrics. According to this method, the GO layers were assembled layer-by-layer onto the surface of cellulose fabrics for five times. In a similar way, for the GOF-coated cellulose fabrics, the layer-by-layer assembly of GOF layers onto the surface of cellulose fabrics was carried out in an external magnetic field.

RGO-coated cellulose fabrics were prepared by subjecting GO coated cellulose fabrics to a reductive treatment. Based on the number of the repeated times of dipping-drying treatments, the five GO coated fabrics (GO-100/0–1, GO-100/0–2, GO-100/0–3, GO-100/0–4 and GO-100/0–5) were respectively immersed in the HI solutions (47%, 100 mL) and was kept at 80 °C for 3 min under constant stirring. Then, the fabrics were washed five times with distilled water in order to remove the excessive reducing agents. Finally, the fabrics were dried at 50 °C and different RGO-100/0 samples were collected (RGO-100/0–1, RGO-100/0–2, RGO-100/0–3, RGO-100/0–4 and RGO-100/0–5). The different proportions of RGOF-coated cellulose fabrics (RGOF-95/5, RGOF-90/10, RGOF-85/15 and RGOF-80/20) were prepared by the similar methods. According to the number of the repeated times of dipping-drying treatments, twenty RGOF-coated cellulose fabrics were obtained.

In this work, we fabricated flexible conductive fabrics via the layer-by-layer assembly of grapheme oxide (GO) gafted with ferroderric oxide on cellulose fabrics in an external magnetic field and followed by chemical reduction. The structure of the untreated cellulose fabrics, GO-coated fabrics and their reducing products was confirmed via x-ray photoelectron spectroscopy (XPS) with an XSAM800 (Kratos, UK) by using monochromatized Al Ka (1486.6 eV) at 180 W (15 mA, 12 kV). The morphology of the different fabrics was examined using field emission scanning electron microscopy (FESEM, JEOL JSM-7100F) at an acceleration voltage 5 kV and transmission electron microscopy (TEM, JEM-2100 Electron Microscope in Japan) using an acceleration voltage of 200 kV. Polarized optical microscopy (POM) measurements of the samples were conducted with a DM2700P Leica Microsystems (German). Moreover, the flexible electrical conductivity of the different fabrics was measured and the relationship between the assembled structure and properties was further investigated. The electrical conductivities of the RGO- and RGOF-coated fabrics were tested according to the International electro technical commission 93–1980 standard (Zhao et al. 2015; Zhang et al. 2016). In detail, two copper electrodes with the spacing of 5 cm were placed on the specimens (width: 1.5 cm; thickness: 0.036 cm) and the electrical resistivity could be calculated by the equation:where R is the electrical resistivity of the fabric (Ω), L is the distance between the two electrodes (cm) and S is the cross-sectional area of the fabric (cm²).

To investigate the chemical changes during the reaction of APTES/Fe₃O₄ and GO nanosheets and the reduction of the resulting cellulose fabrics, the XPS measurements were performed and the results are shown in Fig. 2. Figure 2a showed the XPS survey spectra of pure cellulose fabrics (CF), GO-100/0–3, GOF-80/20–3, RGO-100/0–3, RGOF-80/20–3 and these samples were mainly composed of C, O, N and Fe elements. Figure 2b, c showed the C1s peaks of the five samples. For GO-100/0–3 sample, the peak position of C1s was observed at 284.6 eV, 286.0 eV, 287.9 eV. Compared to C1s of GO-100/0–3, the peaks at 284.6 eV, 285.9 eV, 287.5 eV, 288.6 eV for GOF-80/20–3 was corresponding to the formation of chemical bonds between APTES/Fe₃O₄ nanoparticle and GO, which was agreement with the experimental results (Ramasundaram et al. 2014). It can be seen from Table 1 that fractions of carbon and oxygen in GO-100/0–3 were 74.1 and 24.8% by calculating the area of the corresponding peak in the deconvoluted C1s XPS spectra. After reduction, the C/O atomic ratio was significantly increased in RGO-100/0–3 sample (C: 76.7%, O: 21.5%), which suggested that the remove of some functional groups and supported our previous results (Zhang et al. 2019). In addition, the iron content after reduction was decreased from 0.5% in GOF-80/20–3 to 0 in RGOF-80/20–3, which indicated that the iron ion was also reduced. The decreased iron content would have a great influence on the structural change from the magnetic GOF-80/20–3 fabrics to non-magnetic RGOF-80/20–3, which could be demonstrated by our previous results (Zhang; Wu et al. 2019).

In order to investigate the morphology of RGO- and RGOF-coated fabrics with number of the deposition cycles in an external magnetic field, the evolution of texture of these fabrics were observed by FESEM, as shown in Fig. 3. With the same number of deposition cycles, the pristine RGO coated-fabrics were smoothly covered the entire fabric surface. It is observed that the greater the weight assigned to the APTES/Fe₃O₄ nanoparticles; the more increase of surface coarseness will be increased to that coated fabrics. Additionally, the surface coarseness increased with the increasing number of deposition cycles in same mass ratios. To be specific, with the deposition cycles increasing from 1 to 3, it is displayed that the morphologies of RGOF sheets gradually change into the well-aligned alternating ordered structures along the external-field direction. The red dotted box shown in Fig. 3 exhibited the well arrayed coated-fabrics with GO and APTES/Fe₃O₄ nanoparticles in different mixture ratios. However, with further increases deposition cycles, i.e., from 4 to 5 times, the multilayer fabrics appeared to be sloped downwards to the right, especially the fabric located at bent locations, which suggested that the higher loading of APTES/Fe₃O₄ nanoparticles was unfavorable for the periodic structural rearrangements of RGOF layers.

In order to confirm that the change of macroscopic alignment of RGOF in FESEM could be controlled by applying a magnetic field, polarized optical microscopy (POM) measurements were performed for GOF dispersion liquid (Fig. 4). It can be seen from Fig. 4 that the texture of GOF dispersion liquid was separated by disclinations without magnetic field (see Fig. 4a), however, it was appeared to be directional arrangement and eventually converged to form a large domain (see Fig. 4b) in the present of magnetic field. It was indicated that the macroscopic orientation of GOF dispersion liquid could be tuned by the magnetic field.

The morphological changes of GOF dispersion was monitored by TEM measurements to further understand the magnetic alignment displayed in Fig. 2 after applying an external magnetic field and the results are shown in Fig. 5. In the absence of the magnetic field, as shown in Fig. 5a, the GOF nanosheets exhibited the irregular dispersion morphology with the agglomerates of APTES/Fe₃O₄ nanoparticles, which resulted from the interaction between GOF nanosheets and APTES/Fe₃O₄ nanoparticles and the intra-molecular hydrogen bonding between GOF sheets(Chen et al. 2008; Tong et al. 2019). In the presence of a magnetic field, the magnetic force of magnetic field to APTES/Fe₃O₄ nanoparticles would compress the GOF nanosheets into a parallel band along the direction of the magnetic field (Fig. 5(b)), which would contribute to control the microscopic structures of GOF sheets.

According to the self-assembly theory of magnetic-filed induction, there are five kinds of dominant interactions in magnetic-field self-assembly system, including the interaction force between external magnetic field and APTES/Fe₃O₄ nanoparticles, mutual attraction between magnetic dipole, van der Waals interaction between magnetic APTES/Fe₃O₄ nanoparticles, steric hindrance of APTES/Fe₃O₄ nanoparticles, the resistance between APTES/Fe₃O₄ nanoparticles during motion (Helgesen et al. 1988; Morimoto;Maekawa 1999; Butter et al. 2003; Lalatonne et al. 2004; Pileni; Ngo 2005). The main driving force generated by self-assembly of GOF coated cellulose fabrics in an external magnetic field are the magnetic force of magnetic field to APTES/Fe₃O₄ nanoparticles and the interaction between magnetic doublets. To be specific, the magnetic easy axis of APTES/Fe₃O₄ nanoparticles with magnetic anisotropy was oriented to the magnetic line of force when the APTES/Fe₃O₄ nanoparticles in GOF coated cellulose fabrics were magnetized with the magnetic field turned on. The magnetized magnetic APTES/Fe₃O₄ nanoparticles become a magnetic dipole and correspondingly, the magnetic dipole moment between the magnetic dipoles could be generated. The mutual repulsion or attraction between the magnetic dipole moments could cause the magnetic APTES/Fe₃O₄ nanoparticles to be oriented, which determined the self-assembly behavior of RGOF coated cellulose fabrics (Helgesen, Skjeltorp et al. 1988; Butter, Bomans et al. 2003). The magnetic force acting on the APTES/Fe₃O₄ nanoparticles can be written aswhere \({V}_{p}\) is the volume of APTES/Fe₃O₄ nanoparticle; \({\mu }_{0}\) is the vacuum permeability; \(\chi\) is the magnetic susceptibility; \(N\) is the geometry factor and the value is 0.33 for the spherical particle; \(H\) is the magnetic field strength and \(x\) is the distance between APTES/Fe₃O₄ nanoparticle and magnetic iron. It is noticed from Eq. (2) that the smaller the value for \(x\) can change to a larger magnetic force, which would speed up the movement of APTES/Fe₃O₄ nanoparticles, leading to greater macroscopic alignment of GOF nanosheets.

Figure 6 showed the obtained cycles-volume resistive curves of the fabric specimens in warp directions after induction by the magnetic field. With an increase of the deposition cycles from 1 to 3, the value of different coated fabrics decreased gradually under induction of magnetic field, resulting from the morphologies of RGO and RGOF Nano sheets shown in Fig. 3. Particularly, the value of RGOF-85/15–3 fabrics exhibited a low volume resistivity of 64.8 Ω·cm, it's four times smaller than the Tian’s work (Tian et al. 2016) (shown in Table 2) and about 2 times that of pure RGO-coated viscose fabrics of 5-RGO-c-VF with 5 number of deposition cycles (137.94 Ω·cm) in the previous work (Zhang;Cao et al. 2016). At the same time, the surface resistivity is about the same as the Zhao et al.’s work it handles five times (Zhao; Wang et al. 2020), indicating the increased electrical conductivity of the coated RGOF-85/15–3 fabrics, which revealed that a certain amount of intervention of APTES/Fe₃O₄ nanoparticles under the external magnetic field could improve the mobility of charge carriers effectively among the sheets of fabrics (Shenoy;Reddy et al. 2008). It was also demonstrated that the well-aligned alternating ordered structures of RGOF-85/15–3 fabrics shown in Fig. 3 could improve the electrical conductivity of coated cellulose fabrics. Further, compared with those of coated cellulose fabrics of pure RGO, the results suggested that the magnetic induced coated cellulose fabrics of RGOF sheets grafted with Ferro ferric oxide could not only to retain the high electrical conductivity, but also to decrease the number of the deposition cycles and the graphene oxide consumption. The samples with more coating of RGOF and RGOF showed a slightly increase in volume resistivity due to the increased surface coarseness. Additionally, with the same number of cycles of the layer-by-layer assembly, an increase in the weight ratio of APTES/Fe₃O₄ nanoparticles would be beneficial for increasing the electrical conductivity of the coated fabrics.

To further confirm the electrical conductivity and durability of the GO and GOF deposited fabrics, the cyclic resistance change response of the deposited fabrics was tested after being washed with deionized water twelve times and the washing step kept 30 min in room temperature each time, as shown in Fig. 7. It can be found from Fig. 7 that after twelve times water laundering, the resistance of the RGOF-80/20–3 fabric slightly increased from 0.10 kΩ cm (before washing) to 1.31 kΩ cm (after washing) compared with that of the RGO-100/0–3 fabric (before washing: 1.25 kΩ cm; after washing: 19.47 kΩ cm), indicating that the degradation of its electrical conductivity performance is not obvious after washing. It was also demonstrated that the coated-fabrics with GO and APTES/Fe₃O₄ nanoparticles under an external magnetic field possessed excellent water laundering durability.

In order to investigate the change in microstructure to the influence of thermal conductivity property under induction of external magnetic field, the temperature response time of the different RGO- and RGOF-coated fabrics to 55 °C was observed by the change of label color, shown in Fig. 8, Table 3 and Movie S1-5. The minimal response time was respectively 4.998 s, 4.898 s, 4.781 s, 4.374 s and 3.963 s in different weight ratios of RGO-100/0, RGOF-95/5, RGOF-90/10, RGOF-85/15 and RGOF-80/20 when the deposition cycle was 3. It was demonstrated that the well-aligned alternating ordered structures shown in the red dotted box of Fig. 3 could improve the thermal conductivity of coated cellulose fabrics.