2D Ti3C2Tx (MXene)-reinforced polyvinyl alcohol (PVA) nanofibers with enhanced mechanical and electrical properties

Novel 2D Ti3C2Tx (MXene)-reinforced polyvinyl alcohol (PVA) nanofibers have been successfully fabricated by an electrospinning technique. The high aspect ratio, hydrophilic surfaces, and metallic conductivity of delaminated MXene nanosheet render it promising nanofiller for high performance nanocomposites. Cellulose nanocrystals (CNC) were used to improve the mechanical properties of the nanofibers. The obtained electrospun nanofibers had diameter from 174 to 194 nm depending on ratio between PVA, CNC and MXene. Dynamic mechanical analysis demonstrated an increase in the elastic modulus from 392 MPa for neat PVA fibers to 855 MPa for fibers containing CNC and MXene at 25°C. Moreover, PVA nanofibers containing 0.14 wt. % Ti3C2Tx exhibited dc conductivity of 0.8 mS/cm conductivity which is superior compared to similar composites prepared using methods other than electrospinning. Improved mechanical and electrical characteristics of the Ti3C2Tx /CNC/PVA composites make them viable materials for high performance energy applications.


Introduction
The demand for highly conductive, lightweight and flexible electrodes for various energy applications has sparked interest in developing new methods to fabricate freestanding and flexible films, containing minimum inactive ingredients. Nanofibers prepared by electrospinning (ES) are of considerable interest for various applications due to their unique nanofibrous structure, large surface area, and high porosity [1][2][3].
Electrospun polyvinyl alcohol (PVA) fibers are of particular interest due to their biocompatibility and biodegradability [4,5]. PVA has good transparency and anti-electrostatic properties [6]. The hydrophilic nature of PVA-based films is useful for applications like desalination membranes due to reduced fouling by nonpolar molecules, microbes and fulvic PLOS

Preparation of carboxylated nanocellulose (CNC)
CNC was prepared from date palm leaves (Phoenix dactylifera) by slightly modified previously reported method [17]. Briefly, date palm leaves (10 g) were grinded to a fine powder and added to 1 L of 1 M ammonium persulfate solution. The mixture was heated to 60˚C and was vigorously stirred overnight to yield a white suspension of CNCs. The suspension was centrifuged (5000 rpm for 10 min. and washed with DI water). The centrifugation/washing cycles were repeated until the solution pH close to 4. The product was lyophilized to yield a white powder. CNCs in their sodium form were prepared by slow titration with 1 M NaOH until the suspension reaches pH % 7, followed by washing/centrifugation with deionized water. Purity and crystallinity were confirmed by TEM and XRD (S1A and S2A Figs, S1 File).

Preparation of delaminated Ti 3 C 2 T x sheets
Multi-layer (ML-Ti 3 C 2 T x ) powder was dispersed in deaerated 70% ethanol/water with a weight ratio of ML-Ti 3 C 2 T x : ethanol 250:1. The suspension was sonicated (20 minutes at 60% amplitude while 3 sec pulse on and 1 sec off) under flow in argon, and then centrifuged at 3000 rpm for 1 h to obtain the supernatant containing Ti 3 C 2 T x flakes (1.16 mg/ml). The electrical conductivity of the Ti 3 C 2 T x ink 19.4 μS×cm -1 was measured using a conductivity meter (Cond6 + meter). Purity and crystallinity were confirmed by TEM and XRD (S1 File, S1B and S2B Figs).

Electrospinning of composite fibers
PVA was dissolved in distilled water at 80˚C for at least 2 hours to obtain 15 wt.% solution. Ti 3 C 2 T X and CNC were added to PVA solution followed by dispersion by ultra-sonication for 20 min in an ice bath. Compositions of prepared solutions, together with viscosity (measured at room temperature by vibro viscometer SV10) and conductivity (measured at room temperature by Thermo Scientific Orion 013005MD Conductivity Cell) are summarized in Table 1. All suspensions were visually inspected to confirm good dispersion of Ti 3 C 2 T X and CNC particles. Electrospun nanofibers were fabricated using NaBond electrospinning (Shenzhen, China). Typically, 5 mL PVA suspensions were loaded into a 10-mL syringe, with a blunt-end, stainless steel needle. An aluminium foil covered rotating drum collector was used as the collection screen, it was connected to the ground electrode of the power supply with 10 cm distance between the screen and the needle tip. The electrospinning process was carried out at room temperature at the voltage 17 kV, flow rate 0.3 mL/h and drum speed 200 RPM. The thickness of prepared mats was around 50 μm. Subsequently, electrospun mats were visually checked to confirm no defect or drops of the PVA solution that may potentially affect further characterization.

Scanning electron microscopy (SEM)
The surface morphology of the Ti 3 C 2 T X /CNC/PVA fibers was examined by a field emission scanning electron microscope (FE-SEM, Nova Nano SEM 650) equipped with Energy-dispersive X-ray spectroscopy (EDS) by Secondary electron images with 3 kV and different magnifications. All specimens were sputter-coated with 2 nm gold before the use of the SEM. The average thickness of the fibers was measured by image J software.

Atomic force microscopy (AFM)
AFM was used to further elucidate 2D and 3D surface morphology of electrospun nanofibers using MFP3D Asylum research (USA) equipped with a Silicon probe (Al reflex coated Veeco model-OLTESPA, Olympus; spring constant: 2 N×m -1 , resonant frequency: 70 kHz). Measurements were performed under ambient conditions using the Standard Topography AC air (tapping mode in air). An AFM head scanner applied with Si cantilever adjacent vertically in the sample resonant frequency of the free-oscillating cantilever set as the driving frequency.

Dynamic mechanical analysis (DMA)
Effect of Ti 3 C 2 T X and CNC loadings on the viscoelastic properties of PVA was studied using DMA (RSA-G2, TA Instruments, USA) in tensile mode at 25˚C and ambient atmosphere. Rectangular samples 15 mm×3 mm×0.05 mm were cut from the prepared samples. The linear viscoelastic range was determined from the constant value of the elastic modulus on 0.1-200 μm displacement and 30-100˚C temperature sweep at a heating rate of 5˚C min -1 and at 1 Hz frequency. The specimens were subjected to a tensile deformation of 0.1%.

Tensile measurement
Nanofibers mats were cut into narrow strips and twisted into short nanofiber yarns using the Cord maker & Fringe twister II (LACIS, CA, USA). In order to ensure uniform thickness, 5cm width strips were twisted for 20 seconds each. The twisted fibers were mounted on a C-shaped mechanical testing card with a 50 mm distance between C shape shoulders (See S1 File for more detail of sample preparation). The diameter of the twisted fibers yarns was measured using a SEM and nine readings, along the length of the yarn, were averaged to calculate the diameter of the yarn. Mechanical testing was done at room temperature using the LF LLOYD INSTRUMENTS (An Ameter Company) using a 10 g load cell (Transducer Techniques; Temecula, CA) with a pulling rate of 10 mm/min. Each sample was measured at least 10 times and average values with standard deviations were calculated to analyse mechanical properties of fibers.

Thermogravimetric measurement (TGA)
TGA measurements were performed using TGA 4000 (Perkin Elmer, USA) at temperature range from 30˚C to 600˚C at a heating rate of 20˚C/min under nitrogen atmosphere. Nitrogen gas was passed through the instrument at a flow rate of 20 ml×min −1 . The weight of the samples varied from 10 to 15 mg.

Broadband dielectric spectroscopy (BDS)
Dielectric measurements were performed using a Novocontrol GmbH Concept 40 broadband dielectric spectrometer (Montabaur, Germany), and data were collected over the frequency range 0.1Hz-3MHz at fixed temperatures in the range of -70 to 200˚C. Sample discs of 2 cm diameter were sandwiched between two gold-coated copper electrodes of 2 cm diameter and then transferred to the instrument for data collection. The AC conductivity was calculated, from the Novocontrol WinDeta software, by using the measured values of dielectric permittivity and the dielectric loss factor.

Morphological characterization
A mixed solution of PVA containing different wt % of Ti 3 C 2 T x and CNC was used to fabricate the electrospun nanofibers. Ti 3 C 2 T x CNC/PVA aqueous blends possess good physicochemical properties ( Table 1). As expected, viscosity and conductivity of the blends have increased with higher Ti 3 C 2 T x and CNC amounts, respectively. The well-dispersed aqueous solutions were directly electrospun to form the mixed ratio nanofibers described in Table 1.
The representative SEM micrographs of the four nanofibers composites are shown in Fig 1a  and described in Table 1.
The obtained nanofibers appear to be uniform and smooth with thickness between 174 to 194 nm with no beads and uniform distribution of fibres thickness. As the ratio of Ti 3 C 2 T X increases in the composite the nanofibers diameter decreased as clearly seen for blends C 1 M 1 , C 0 M 2 as compared with those containing only PVA (C 0 M 0 ). The average diameters of the nanofibers measured from SEM were 182 (±53), 195 (±33), 174 (±53), 179 (±36) nm for C 0 M 0 , C 2 M 0 , C 0 M 2 and C 1 M 1 respectively. The presence of Ti 3 C 2 T x within C 0 M 2 and C 1 M 1 electrospun nanofibers was confirmed by tracing Ti element in the EDS spectra. (S3 Fig; S1 File).
Fig 1b shows EDS mapping of C 0 M 2 containing Ti 3 C 2 T X (green color) where good dispersion of Ti 3 C 2 T X particles within PVA matrix were confirmed.
AFM was used to further elucidate the surface morphology of the fibers as described in Fig  2. The 2D AFM-phase and 3D AFM-height images were obtained using intermittent-contact (tapping) mode. The C 2 M 0 , C 0 M 2 and C 1 M 1 samples exceled axially oriented structures of fibers with relatively smooth surfaces. No visible aggregation was observed in the presence of CNC or Ti 3 C 2 T X which were sufficiently located inside the fibers. Moreover, no significant differences of roughness have been found in longitudinal as well as lateral directions ensuring the well dispersion of them within the PVA fibers. Some of the fibers were bundled together leading to the formation of multi-fiber structures. In all cases, the prepared fibers led to the formation of interlaced structures responsible for the creation of relative regular networks.

Mechanical characterization
Mechanical characterization of electrospun nanofibers was performed by dynamic mechanical measurement and tensile measurement.
The dynamic mechanical behaviour of the electrospun mats containing CNC or/and Ti 3 C 2 T X fillers was measured in the broad temperature range (Fig 3A) to provide a complex investigation of the prepared electrospun nanofibers. Storage modulus increases in order: neat PVA, CNC/PVA and Ti 3 C 2 T X /PVA what indicated the reinforcement effect of CNC and  (Table 1), B) dispersion pattern of Ti 3 C 2 T X shown by EDS mapping of sample C 0 M 2 where; C (yellow), O (oxygen) and Ti (green).
Ti 3 C 2 T X within PVA fibers. The comparison of storage modulus of samples containing 0.14 wt % either CNC or Ti 3 C 2 T X with sample containing both fillers with concentration 0.07 wt. % each indicated synergic effect of CNC and Ti 3 C 2 T X fillers.
Indeed, the nanofibers containing 0.14 wt % CNC exhibited around 20% increasing of storage modulus compare to neat PVA fibers. In addition, nanofibers containing 0.07 wt.% of both, CNC and Ti 3 C 2 T X fillers exhibited over 100% increase in of storage modulus compare to PVA nanofibers at 25˚C, what is superior compared to previously reported study [35], where 74% increase in storage modulus of PVA film by addition of 3 wt.% nanocellulose at 25˚C have been reported. Our hypothesis is that the self-orientation of CNC nanofibers during electrospinning within PVA matrix caused by the electrical field has remarkable influence on the mechanical properties of the nanofibers composites.
Furthermore, Fig 3B is the tan δ curve, which is the ratio of loss modulus to storage modulus and is a measure of the damping behaviour of a material [11]. Fig 3Ba showed a shift in the glass transition temperature with addition of both CNC and Ti 3 C 2 T X fillers.
Glass transition temperature (T g ) of neat PVA nanofibers was 70.2˚C. The addition of CNC resulted in an increase of the T g to 79.1˚C indicating significant improvement of mechanical response for this type of polymer. The addition of Ti 3 C 2 T X to PVA matrix led to increasing of T g up to 80.5˚C. Furthermore, PVA nanofibers containing CNC and Ti 3 C 2 T X (C 1 M 1 ) assigned T g 82.3˚C which also support hypothesis about synergic effect imparted by both fillers.
As a second approach to study mechanical properties of prepared nanofibers, tensile test has been performed. Obtained Young's modulus, tensile strength and elongation at break for PVA electrospun composites are listed in Fig 4. A slight increase of Young's modulus of PVA electrospun mats with addition CNC fillers, from 221±51 MPa to 241±51 MPa was observed. Sample containing Ti 3 C 2 T X particles showed additional increase of Young's modulus 283±60 MPa and sample containing both fillers showed Young's modulus of 293±59 MPa. Young's modulus obtained from tensile measurement is in line with storage modulus obtained from DMA study. However, absolute values cannot be compared due to nature of the experiment and sample preparation. Tensile measurements were performed for twisted samples to eliminate effect of mostly randomly orientated nanofibers within electrospun mat, whereas DMA measurement were done for specimens cutted from electrospun mats without twisting.
On the other hand, only slight dependence of tensile strength on the filler content 19.5±2.4 to 22.01.3±1.6 MPa was observed.
The elongation at break decreased with addition of the fillers, a common behaviour for polymeric composites. Elongation at break for neat PVA was 179 ±23% whereas elongation for samples containing fillers ranging from 98±10% to 125 ±21%.
It can be concluded that addition of CNC improves mechanical performance of composites probably due to its highly crystalline structure and high aspect ratio [35], whereas Ti 3 C 2 T X also tend to have high mechanical performance and contributes to increase in the of mechanical properties of the nanofibers [33].

Thermogravimetric measurement
Thermal gravimetric analysis (TGA) is commonly used technique to investigate thermal decomposition behaviours of polymers at different temperatures, and also demonstrates thermal-resistant of various materials.  TGA measurement of all prepared electrospun composites is shown in Fig 5. First weight loss, at the temperature up to 100˚C, occurred due to evaporation of moisture since PVA is known as a hydrophilic polymer. The most expressive weight-loss (more than 70%) occurs from 245˚C to 380˚C, due the breakdown of the polymer The third weight-loss step, of about 20%, is observed between 380˚C to 470˚C, which can be associated mainly with the PVA degradation. The onset temperature of PVA degradation is 270.2˚C with peak maximum at the temperature 287˚C. The addition of CNC caused the most significant enhancement of thermal properties of PVA, where the onset temperature shifts 343.5˚C and peak maximum to 367.7˚C. Similar results, increasing of thermal properties of PVA by adding nanocellulose were reported previously [36,37].
MXene particles also caused increasing of thermal stability of the electrospun PVA nanofibers compared to neat PVA with onset temperature 290.9˚C and peak maximum at the temperature 354.9˚C. In addition, PVA containing both fillers, MXene and CNC assigned onset temperature 282.5 and maximum peak degradation at 345.6˚C. Herein, it can be stated that MXene as well as CNC caused remarkable enhancement of thermal properties of electrospun PVA nanofibers.

Conductivity characterization
The dc conductivity (σ 0 ) of the Ti 3 C 2 T x /CNC/PVA nanofiber composites is given in Fig 6. DC conductivity exhibited the typical behaviour for polymers and disordered solids as it has frequency independence at low frequencies then rises monotonically with the s-th power of frequency, where s ranges between 0.7 and 1.0 [13,38,39].
The conductivity shows a plateau at low frequency which corresponds to the dc conductivity and the frequency at which it becomes frequency dependent is called the relaxation frequency, ω s [40].
The dc conductivity increases and the plateau widen to include the entire frequency range for high MXene concentration sample C 0 M 2 . The results revealed a highly conductive composite film with excellent mechanical characteristics. In addition to being conductive filler, Ti 3 C 2 T x acted as a reinforcing component, along with the CNC, causing enhancements in the mechanical properties of the PVA fibers. Similarly, Zhang et al. used Ti 3 Si 0.75 Al 0.25 C 2 and Ti 3 Al 0.9 Si 0.1 C 2 nanosheets to prepare poly(methyl methacrylate) (PMMA) based composites with great enhancements in the mechanical and thermal properties [41]. Their composites showed fivefold increase in the Young modulus and twofold increase in the tensile strength 2D Ti 3 C 2 T x (MXene) based nanofibers comparing to the control PMMA. They revealed that the obtained mechanical characteristics enhancements were much higher than those obtained with the boron nitride/methyl methacrylate and the graphene/PMMA composites having the same weight percent of nanosheets [41]. They also revealed that the nanosheets improved the thermal conductivity of the composites but no reference to their electrical conductivity.
In PVA, electron conduction occurred by hopping through the film due to polymer polarization to transport electrons [42]. The presence of CNC would cause large resistivity to these hopping and thus result in conductivity reduction observed in Fig 5 for sample C 2 M 0 . It is also worth noting that the conduction behaviour exhibited by sample C 1 M 1 is almost the same as that of the control PVA although it contains the conductive MXene nanosheets. However, small amount of Ti 3 C 2 T x does not seem to be enough to achieve the connected conductive pathway or it is below the percolation threshold for the conductivity in these composites.
Ti 3 C 2 T x was used recently by Ling et al. to prepare flexible/mechanically strong PVA composites films by simple mixing and solvent casting technique. They were investigated as electrodes for supercapacitors and demonstrated a notable volumetric capacitance in KOH electrolyte [30]. Although these Ti 3 C 2 T x /PVA films showed excellent conductivity; the authors used very high amounts of Ti 3 C 2 T x in preparing the composites [30]. For example, a sample containing 40 wt% of Ti 3 C 2 T x revealed a conductivity of 40 mS/cm. This amount of MXene is much higher comparing to the amounts used in the work presented herein. Sample C 0 M 2 in our study exhibits a conductivity of ffi 0.08 mS/cm, with only 0.14 wt% of Ti 3 C 2 T x within PVA fibers and it has great flexibility and mechanical strength. This reveals the great potential of the Ti 3 C 2 T x /CNC/PVA electrospun nanofiber composites for usage as flexible and wearable energy storage devices [30].
Polarity in PVA arises from the electronegativity difference among its atoms, for example, oxygen vs. carbon atoms. Conducting and nonconducting nanoparticles are typically used to obtain nanocomposites with high values of dielectric constant. Examples conducting fillers include graphene, carbon black, and carbon nanotubes; while the nonconducting ones include barium titanate, strontium titanate, and calcium titanate [43][44][45][46]. At certain concentration, percolation threshold, filler nanoparticles start to generate an interconnected network leads to an abrupt increase in the conduction. Percolation occurs at low concentrations of conducting fillers, like the Ti 3 C 2 T x used herein, in comparison to non-conducting fillers. Fillers shape (zero-, one-, and two-dimensional), size, and orientation are among the parameters that determine the percolation threshold [43][44][45][46]. Ti 3 C 2 T x as a two-dimensional material can easily achieve the percolation limits comparing to lower-dimensional fillers. In addition, orientation of the Ti 3 C 2 T x nanosheets due to electrospinning could facilitate easier connection to form a continuum at lower concentration. This would explain the high conductivity obtained for sample C 0 M 2 , Fig 5, which contains only 0.14 wt% of Ti 3 C 2 T x .
The temperature dependence of the dc conductivity (σ dc ) is presented in Fig 6. Values of σ dc were calculated from the plateau values of the dc conductivity at low frequency, S6 Fig. DC conductivity accounts for the long-range hopping of the charge carriers and it involves a finite and reversible storage of charge in the material or at the interface [47].
At low f, time scale or half period of oscillation (2f) -1 , increases, charge carriers and ionic moieties can execute more elementary hops before the applied field reverses [48]. The monotonic upright raise of the plateau values, S6 Fig, with increasing temperature accounts for the dc conductivity increase especially above the polymer T g . The chain segmental mobility increases in the rubbery state and the charge carriers become more mobile as their motion is coupled to that of the chain segment. The temperature dependence of σ dc in Fig 7 demonstrates strong Arrhenius behaviour according to the Eq 1: Where σ o is the pre-exponent factor and k B is Boltzmann constant. The conductivity activation energy (E a ) values extracted from Fig 7 are presented in Table 2. The activation energy decreased as the content of the Ti 3 C 2 T x in the composites was increased. It is interesting to note that for a PVA sample containing 14 wt% Ti 3 C 2 T X , the electronic conduction only required 0.02 eV of energy to launch however we were not able to obtain uniform fibers (data not shown). This low energy barrier reflects metallic or narrow  band-gap semiconducting characteristics. Therefore, the electrospun composites presented herein could be a good candidate for electrode materials in Li-ion batteries [49]. In their theoretical study on Ti 3 C 2 monolayer, Tang et al. predicted low diffusion barrier (0.07 eV) for an isolated Li atom on the Ti 3 C 2 when compared to that for anatase TiO 2 (0.35-0.65 eV) or graphite (~0.3 eV) [49]. This means that Li can exhibit faster transport and higher charge/discharge rates through the Ti 3 C 2 monolayer comparing to TiO 2 and graphite anodes. Our results for 14 wt% Ti 3 C 2 T X electrospun composite confirm the theoretical prediction of low conduction barrier for this type of materials. However, Tang et al. also reported that surface functionalization of the Ti 3 C 2 nanosheets by F or OH groups degrades the Li diffusion and decreases their Li storage capacity [50]. Our samples showed low conduction barrier although the Ti 3 C 2 surface is functionalized with F or OH groups. Therefore, the conductivity enhancement obtained for our samples could be related to the electrospinning process used during their preparation. MXene based anode material in Li-ion batteries was reported to exhibit good capacity retention during 1000 galvanostatic charge/discharge cycles at rates up to 10˚C [51].

Conclusions
In conclusion, this work reports the preparation of Ti 3 C 2 T x , MXene and CNC reinforced electrospun PVA nanofibers for the first time. Diameter of obtained nanofibers ranged from 174 to 195 nm and varied depending on the ratio between PVA, Ti 3 C 2 T x and CNC as measured by SEM and AFM microscopy. DMA showed an increase in the elastic modulus from 392 MPa for neat PVA fibres up to 855 MPa for fibres contain both fillers at temperature 25˚C, which indicates a synergic reinforcing effect of both fillers. More particularly the electrical conductivity of the nanofiber composites increased with increasing content of Ti 3 C 2 T x , particularly for C 0 M 2 sample (0.14 wt.% Ti 3 C 2 T x ) where conductivity 0.8 mS×cm -1 was measured and showed narrow band-gap semiconducting characteristics, which makes them good candidates for electrode materials in Li