Graphene nanosheets derived from plastic waste for the application of DSSCs and supercapacitors

The present study reports the upcycling process of waste plastics into value-added product graphene nanosheets (GNs) and their subsequent applications in dye sensitized solar cells (DSSCs) and supercapacitors. Bentonite nanoclay has been used as an agent for the degradation of waste plastics with two step pyrolysis processes at 450 °C and 945 °C in an inert atmosphere of N2 gas to obtain GNs. The GNs with few layers were confirmed by the RAMAN spectroscopy, XRD and HRTEM analyses. Further, FT-IR and EDX analyses also performed for the identification and quantitative analysis of functional groups in GNs. The GNs thus synthesized from plastic waste have been used for the fabrication of DSSCs and supercapacitors. The DSSC fabrication with GNs as part of photo-anode with polymeric electrolyte showed a high fill factor of 86.4% and high Voc of 0.77 V, which were also supported by the computational findings. On the other hand, the utilization of GNs as an active layer material of supercapacitor electrodes offered a high specific capacitance of 398 F/g with a scan rate of 0.005 V/s. The supercapacitor also exhibited significant energy density (Ed) and power density (Pd) of 38 Wh/kg and 1009.74 W/kg, respectively. Thus, the process illustrated the utility of waste plastics upcycling for conservation of EEE i.e., ecology, economy and energy for better tomorrow.

obtained from waste plastics have been utilized to design high performance Supercapacitors and DSSCs. Furthermore, the computational results suggest that reducing the GO to obtain RGO or defected GNs is the right approach to design high performance supercapacitors and DSSCs, which was followed in our experimentations. The cost benefit analysis shows that the upcyclization of waste plastics into GNs and their subsequent utilization in energy conversion and storage devices could lead to revolutionary changes for the sustainable management of ecology and economy.

Results and discussions
In order to demonstrate the usefulness of GNs prepared from the waste plastics through industrial symbiosis manner, the energy conversion and storage devices namely DSSCs and supercapacitors have been fabricated. The obtained GNs from our pilot scale plant as per the previously reported process 55 have been thermally reduced at higher temperature (945 °C) before utilizing for these applications. The reduced GNs were critically analyzed by Raman spectroscopy in order to confirm the uniformity of GNs, which is essential for the determination of their purity level before utilizing as photoanode layer in DSSCs and active layer material in supercapacitors. The GNs showed two prominent peaks at 1340/cm and 1591/cm corresponded to D and G bands, respectively (Fig. 2a). The D band demonstrated the stretching of sp 3 carbon atoms, while the G band showed the stretching of sp 2 carbon atoms. As reported earlier, during the synthesis process, some of the sp 2 carbon atoms get distorted and converted to sp 3 carbon atoms due to the presence of various functional groups as confirmed by the FT-IR spectroscopy. Therefore, the distorted geometry at some of the carbon atoms was confirmed by the presence of D band. The other carbon atoms which possess sp 2 hybridization, resulted a degree of graphitization in the GNs. Further, the intensity of the D band gives information about the domains of sp 2 carbons within the GNs, while the average size of the domains of sp 2 carbons is depicted by the I D /I G ratio [57][58][59][60] . The I D /I G ratio for the GNs was found to be 0.91, thereby showing more domains of sp 2 carbons within the reduced form of GNs. Additionally, the peaks at 2712/cm and 2896/cm depicted the presence of 2D and D + G bands due to process of double resonance. The 2D band demonstrated the presence of few layer graphene sheets within GNs, which was again confirmed by the I 2D /I G ratio (0. 23). The presence of the high intensity D + G band showed low disorderness of GNs. Further, higher domains of sp 2 carbon atoms within the GNs are necessary to achieve high performance in DSSCs and supercapacitor devices. In this case, the thermal reduction of GNs enhanced the domains of sp 2 carbon atoms www.nature.com/scientificreports/ by reducing oxygen content within GNs as suggested by the EDX analysis. The EDX analysis showed that the GNs were constituted of only carbon and oxygen elements with the wt% of 98.24% and 1.76%, respectively, suggested the high level of purity after thermal reduction, whereas the same as per our previous report was 76.99% and 21.1% for unreduced GNs 53 . From this observation, it was concluded that the treatment of GNs at 945 °C reduced the GNs and hence enhanced the domains of sp 2 carbon atoms. However, due to the presence of some strongly bonded oxygen functionalities, the FT-IR spectra showed carbon-oxygen, carbon-oxygen-carbon in epoxy ring and combination of O-H deformation at 1065/cm, 1454/cm, and 3437/cm, respectively. The FT-IR spectra also depicted the presence of aromatic double bond and C-H in plane deformation vibrations at 1625/cm, 2853/cm, 2923/cm (Fig. 2b). Moreover, the thermal reduction also reduced the sheet resistance from 400 to 6 Ω/ sq, thereby enhancing their suitability as photoanode layer of DSSCs and active layer material of supercapacitors. The XRD analysis was also performed on the synthesized GNs (Fig. 2c). The XRD spectrum showed diffraction peaks at 2θ = 23.24 and 43.8° corresponded to the characteristic peak and graphitic peak of the reduced GNs. The thickness measurement and evaluation of number of layers within the GNs were done using the XRD graph, as reported previously 61 . The number of layers within GNs can be evaluated by dividing the crystal size (c) with the interlayer distance (d) followed by the addition of thickness of one graphene sheet (0.1 nm). The crystal size (C) was calculated by using Scherer equation 62 , while the Bragg's equation was used to calculate the interlayer distance (d) within the GNs 63 . The calculated crystal size of the GNs against the diffraction peak at 2θ = 23.24 was found to be 7.22 Å. The interlayer distance (d) was found to be 3.96 Å. Thus according to the XRD data, about 2-3 graphene layers were found within the waste plastic derived GNs. The calculated thickness of GNs from the XRD analysis shows resemblance with the thickness analyzed from the HRTEM images (Fig. 6), which corresponded to the functionalized graphene sheets with few oxygen-containing groups. Therefore, FT-IR investigation was carried out to reflect the presence of the various functional groups within the GNs (Fig. 2c). Further, the presence of these volatile groups and the purity of the graphitic nature of the GNs were evaluated by thermogravimetric analysis (TGA). TGA analysis is one of the few important tools to understand the skeleton properties of GNs as the FT-IR spectra indicated the presence of various groups (Fig. 2d). TGA graph depicted the weight profile of reduced GNs powder w.r.t the variation of temperature (heating rate 1°/min) in presence of N 2 flow. A relatively small amount of weight loss (~ 3.4 wt%) was observed in the temperature range 35-100 °C corresponded to the water molecules present within the sample, thereafter a clear single step degradation pattern can be seen due to the presence of very small amount of oxygen within the GNs (1.54%, EDX analysis), which shows high purity level of the thermally reduced GNs. Scheme 1. Schematic of the present process depicting the synthesis procedure of GNs from waste plastics and their applications for DSSC and supercapacitors.  55 , we reported that the defected graphene is more suitable than the pristine graphene for energy storage applications such as battery electrodes, supercapacitor electrodes, etc. In this work, DFT simulations have been performed to show how our thermally reduced defected GNs are a great choice over the unreduced GO and the pristine GNs for application as semiconducting layer in DSSCs and electrode active layer material in supercapacitors.
The simulations have been carried out using Synopsys-QuantumATK, a commercial DFT code 64 . The exchange and correlation interaction energy of electrons has been described using the revised Perdew-Burke-Ernzerhof (rPBE) functional within the Generalized Gradient Approximation (GGA) method 65 , while extracting the structural and electronic properties; whereas the optical absorption spectra have been extracted using semiempirical Meta-GGA (MGGA) functional of Tran and Blaha 66 . We have refrained from using MGGA to extract the structural and electronic properties of our graphene structures due to its inability in accurately estimating the total energies and band structures of metallic natured materials. The Kohn-Sham equations are solved by expanding the wave functions using localized Double-Zeta-Polarized (DZP) basis sets. A large density mesh cutoff of 130 Hartree has been used for accuracy of calculations. The Brillouin zone of 5 × 5 hexagonal supercell (consists 50 atoms) has been sampled using Monkhorst-Pack grid of 10 × 10 × 1 k-points, whereas the density of states (DOS) have been extracted using high k-points of 27 × 27 × 1 for accuracy. Computationally, both the defected GO and defected GNs are designed from the pristine GNs of 5 × 5 hexagonal supercell. The defected GNs of 2% and 6% concentration are obtained by subjecting the pristine GNs to single and triple vacancy defects, respectively; whereas the defected GO sheets are obtained by subjecting the pristine GNs to vacancy defects as well as subsequent oxidization with various functional groups viz. epoxy (R-O-R), carbonyl (R=O), hydroxyl (R-OH), and carboxyl (R-COOH). All these structures have been optimized with the help of limited-memory Broyden-Fletcher-Goldfarb-Shanno (L-BFGS) quasi-Newton method 67 , so that the forces on the atoms and stress on the supercell converge below the tolerance value of 0.05 eV/Å and 0.0006 eV/Å 3 , respectively. The structural properties of the graphene structures have been assessed with the help of C-C bond lengths and thermodynamic stability, as tabulated in Table 1. The pristine GNs has a C-C bond length of 1.42 Å, whereas the same for defected GO and defected GNs varied between 1.35 and 1.63 Å and 1.37-1.51 Å, respectively. The defected GO experienced large variation in C-C bond lengths than the defected GNs, and the maximum C-C bond length of the former reached upto 1.63 Å. This large increment in bond length and the slight out of plane alignment of the carbon atoms associated with the functional groups signify the degree of sp 3 hybridization present in the defected GO. On the other hand, the pristine and defected GNs remain purely sp 2 hybridized as here, E defected GNs/GO , E carbon , and E functional groups refer to the energy of defected GNs/GO, energy of single carbon atom of pristine graphene, and energy of functional groups (epoxy, carbonyl, hydroxy, and carboxy), respectively. L and x indicate length of the sheet and the number of carbon atoms present in the sheet, respectively. Table 1 shows the calculated formation energies for defected GNs/GO, where the lower formation energy indicates better thermodynamic stability. Moreover, the positive (negative) sign against the formation energy indicates endothermic (exothermic) process, where energy is consumed (released). It is worth noting that the formation of defected graphene from pristine graphene is an endothermic process, whereas the formation of defected GO from pristine graphene is an exothermic process. In general, the exothermic processes are energetically more favorable than the endothermic processes. This may be the reason that the bulk synthesis of graphene sheets yield GO first, rather than rGO or defected GNs directly, as witnessed in our previous study 55 . Furthermore, the formation energy increases with the defect size as seen from the Table 1, indicating the reduced stability of the defected GNs/GO with increasing defect size. In order to estimate the electronic nature of these sheets, the band structure and density of states (DOS) have been extracted and plotted in Fig. 4. From Fig. 4a, the pristine graphene exhibits a perfect zero band gap with the Dirac cones meeting at K-point of Brillouin zone. At 2% defect concentration, both the defected GO and defected GNs exhibited metallic nature, with the latter possessing more DOS near the Fermi level than the former (Fig. 4b,c). When the defect concentration is increased to 6%, defected GO exhibited a band gap of 0.15 eV, whereas the defected GNs retained its metallic nature (Fig. 4d,e). We have also checked the electronic nature by increasing the defect concentration. Our results suggest that the defected GNs mostly preserves its metallic nature and is thus preferable as electrode material over GO and pristine GNs. The experimental observation of enormous reduction in the sheet resistance after thermal reduction of oxidized GNs may be attributed to the variation in electronic nature, as observed computationally.
Further, the field emission scanning electron microscopy (FESEM) images show a clear rough and porous morphology of the GNs, which postulate the metallic nature of synthesized GNs. The FESEM images in Fig. 5 clearly show large clusters of the graphitic sheets with the presence of wrinkles at the surfaces. Figure 5a shows three consecutive stacks of few layer GNs designated as 1, 2 and 3, which indicate the surface morphology of GNs. The 3D surface morphology of these randomly selected stacked GNs indicate the varying area of individual sheets with corrugated edges (Fig. 5b). Further, these randomly selected GNs stacks are found to be linked with each other by weak Van der Waals forces. Another randomly selected stack of GNs designated by number 4 in Fig. 5c shows the individual sheets of GNs in close proximity, which are held by weak Van der Waals forces. The inflated form of the GNs in randomly selected stack may be a result of thermal treatment during the reduction process. This can be seen in the 3D surface morphology of the selected stack in Fig. 5d, where the sheets show significant separation among them with corrugated edges. www.nature.com/scientificreports/ The identification of number of layers in each cluster is found to be difficult from FESEM images, therefore HRTEM analysis was performed in order to get the average number of layers and crystalline behavior of the reduced GNs. Figure 6a,b shows the HRTEM image at a magnification of 5 nm and selected area electron diffraction (SAED) patterns for GNs, respectively. Figure 6a gives a clear view of different kinds of grain boundaries, thereby portraying poly crystalline properties within the GNs. It was again confirmed by selected area electron diffraction (SAED) patterns of GNs (Fig. 6b), where the diffraction patterns in the concentrated circle forms confirm the polycrystalline properties within the GNs. Further, the HRTEM image of the GNs clearly depicted the presence of the 3-5 layers, while the average number of sheets in the GNs were found to be below 5 as depicted by Fig. 6d. This shows the high quality of GNs and their suitability for DSSC and supercapacitor applications. Further, the 3D HRTEM image of the GNs show corrugated edges (Fig. 6c), which makes them favorable for the DSSCs and supercapacitors as per our previously reported work 53 .
After the successful assessment of the synthesis, thermal reduction, and structural as well as electronic aspects of the GNs derived from waste plastics, they have been used to fabricate the DSSCs and for energy conversion and storage purposes.
Assessment of GNs as a part of the photo-anode layer of DSSC. The suitability of the defected GNs as a part of the photo-anode layer and efficient photon absorber 69 for DSSC has been evaluated with the help of  Table 1. Carbon to carbon (C-C) bond length, formation energy per unit length (E F L ), electronic nature, peak quantum capacitance (C Q ), and the peak absorption coefficient (α) for various graphene/GO structures. Here, XX and YY indicate zig-zag and armchair directions of the sheets, respectively. UV, IR and Vis indicate the Ultra Violet, Infra Red and Visible regions of the electromagnetic spectrum, respectively, where the peaks have been witnessed in the absorption spectra. www.nature.com/scientificreports/ optical absorption spectra, and compared with the defected GO and pristine graphene. The optical absorption spectrum has been calculated from the frequency dependent complex dielectric constant, which has been computed through the Kubo-Greenwood formalism for calculating susceptibility tensor as 70 , and,

Structure type C-C bond length (Å) E F L (eV/Å) Electronic nature Peak C Q (μF/cm 2 ) Peak α (/nm)
here, ħ, V, Γ, and ƒ refer to reduced Planck constant, volume, broadening, and Fermi function, respectively. π i nm is the i-component of dipole matrix element between 'n' and 'm' states. The absorption coefficient (α) has been calculated using the real part (ε 1 ) and imaginary part (ε 2 ) of the complex dielectric function ε(ω) as, The optical absorption spectra calculated over the photon energy range of 0-5 eV are depicted in Fig. 7, where the infra red (IR), visible, and ultra violet (UV) regions constitute 0-1.78 eV, 1.78-3.1 eV, and 3.1-5 eV, respectively. XX, YY and ZZ indicate the zig-zag, armchair, and out of plane directions of the sheet, respectively.
The peak values of absorption coefficient offered by each of the sheets are listed in Table 1. It can be seen from Fig. 7a-e that only the defected GNs offer maximum absorption peaks in the visible region, whereas the pristine graphene has its maximum peak located in UV region, and the defected GOs have their maximum peaks in the IR region. The maximum peak offered by 2% defected GNs is 0.02/nm (in XX and YY) at 2.6 eV (visible region); while the 6% defected GNs offers 0.023/nm (in YY) at 1.78 eV (Red light of visible region) and 0.012/nm (in XX) at 2.09 eV (visible region). Though the pristine graphene has its maximum peak (0.035/nm) located at 4.45 eV (UV region), it also offers a good peak value (0.027/nm) at 2.9 eV in the visible region. All these absorption peaks of pristine graphene can be attributed to the π-π * transitions 71 . In general, high absorption in the visible region is required in order to attain high efficiency in DSSCs. Thus, the defected GNs are a better choice than the GO and pristine graphene to enhance the photon absorption in DSSCs, owing to its peaks in the visible region. Though the pristine graphene also offers high absorption in the visible region, we could not recommend it owing to the extreme complexity associated with its bulk synthesis. www.nature.com/scientificreports/ The defected GNs are used to fabricate DSSC, the schematic of which is depicted in Fig. 8a. An area of 0.45 cm 2 of the fabricated DSSC was exposed to a yellow light emitting halogen lamp (remaining portion was masked with black chart paper) kept at a sufficient height in order to simulate 1 sun condition (100 mW/cm 2 ). The Fig. 8b shows the obtained J-V characteristics of the DSSC, which shows almost negligible amount of current under dark conditions, whereas under illumination an open circuit voltage (V oc ) of 0.77 V with a short circuit current density (J sc ) of 0.302 mA/cm 2 has been observed. A gradual rise and fall in the current can be observed between 0.35 and 0.65 V with maxima at ~ 0.5 V, which is a result of the polarization effect of the polymer electrolyte as detailed in a previous study 72 . The obtained fill factor was 86.4% that lead to an overall efficiency (η) of 0.20%. The incorporation of GNs onto the working electrode led to an improvement in the band alignment. Thus, a metallic natured layer between the FTO and TiO 2 assisted in better collection and transfer of electrons, hence an enhancement in V oc .
It has been previously reported that adding an extra layer of GNs in between FTO and TiO 2 layers can reduce the tendency of charge carriers to recombine by enhancing the charge transport rate and lowering the possibility of recombination, thereby improving the photoelectrical conversion efficiency. Accordingly, under the condition of illumination, the conduction band (CB) of the TiO 2 received the photoinduced electrons from the photoexcited dye molecules, now as the layer of TiO 2 is anchored with GNs, the photoexcited electrons were captured rapidly by GNs without any obstacles. Now the collected electrons can transport from TiO 2 to the conductive FTO substrate more effectively through the bridges of the GNs and hence slowed down the rate of recombination and back reaction 73 . Due to the 2D geometry and porous structure of the GNs, the TiO 2 nanoparticles anchor well into the GNs, increasing the probability of capturing the photo induced electrons. This results in rapid transportation of the photo induced electrons and suppression of the charge recombination process at the interface. Furthermore, the introduction of GNs into the interface of FTO and TiO 2 increases the porosity within the system, which enhances the possibility of light scattering during the transmission of incident photons in the device. In other words, incorporation of GNs enhanced the roughness of the interface, which allows more light to scatter and thereby maximizing the device efficiency. The utilization of solid polymer electrolyte enhances the stability of the device in comparison to other reported electrolytes. Further, we have observed that higher concentrations of graphene lowers the efficiency of the system, as the grapheme-TiO 2 interface that acted as  www.nature.com/scientificreports/ charge collection centre may offer short circuit within the system. In addition to it, the excessive presence of the graphene will also lower the photon absorption characteristics of the dye, thereby reducing the device efficiency.
In the present case, we have observed 1.5 wt% of GNs as the optimized parameter for the fabrication of DSSCs. Furthermore, electrochemical impedance spectroscopy (EIS) was performed using CHI 604D electrochemical workstation for measuring the impedance between the junctions of the cell in both dark and illuminated states. The characteristics obtained after fitting the curve using inbuilt software of the machine is shown in Fig. 8c, and an expanded portion of the squared region is shown in the inset. The equivalent circuit used for the fitting is depicted in Fig. 8d, where R s , R t , and R c represent the series resistance, the impedance between working electrode and redox electrolyte, and the impedance between counter electrode and electrolyte, respectively. A remarkable difference in the impedance offered by the dark and illuminated conditions is observed. The impedance under illumination is barely visible in comparison to that of dark condition. A much clear observation is made when the graph is expanded to 600 Ω. The value of R s remains 24 Ω for both the conditions as it is the impedance occurring due to the external connections, FTO, etc. The junction impedance between the counter electrode and electrolyte (R C ) also stays almost the same for both conditions, i.e. 98.33 Ω. The major change is observed in the value of R t , which seems to cross 30 kΩ under dark condition, and limited to only 564.83 Ω under illumination.

Assessment of GNs as active layer material of supercapacitors.
The suitability of defected GNs as the active layer material of supercapacitor electrodes has been estimated with the help of quantum capacitance (C Q ) calculations, and compared with GO and pristine GNs. In general, C Q is the intrinsic capacitance offered by graphene or any other 2D/1D material, and is prominent in those with limited DOS. It plays an important role in determining the total capacitance (C T ) offered by the graphene supercapacitor, by appearing in series with the double layer or Helmholtz capacitance (C DL ) that forms between electrode and the electrolytic ions therefore, 1 C T = 1 C Q + 1 C DL . Thus, in order to achieve a high total capacitance, it is important to use electrodes that offer high quantum capacitance. Mathematically, C Q can be expressed as the variation caused in excess charge density (Q) of the graphene electrode with respect to the applied potential ( ϕ ), and can be extracted from the DOS profile. The excess charge density on graphene, when the chemical potential is shifted by applying ϕ/e , is given by 74 here, e, E, D(E), and f (E) indicate the elementary charge, energy with respect to the Fermi level, DOS, and Fermi-Dirac distribution function, respectively.
Differentiating Q with respect to ϕ yields the differential quantum capacitance, here, the thermal broadening function F T (E) is given by, The C Q diff expression that maintains the charge/discharge synchronization between quantum capacitance and DOS profile, is given by, The integrated quantum capacitance (C Q ) reported in this work is given by, The computed C Q of the defected GNs along with the defected GO and pristine GNs is depicted in Fig. 9, and the peak C Q values are tabulated in Table 1. The C Q is plotted over a potential range of − 0.6 to 0.6 V only, considering the typical electrochemical window of aqueous electrolyte of the supercapacitor. From Fig. 9, pristine GNs posses a very low C Q with the peak value of just 10.73 μF/cm 2 at − 0.6 V, which is in good agreement with the experimental report by Ponomarenko et al. 75 . At a defect concentration of 2%, the defected GNs and the defected GO exhibited peak quantum capacitances of 185.25 μF/cm 2 and 100.23 μF/cm 2 , respectively. At a defect concentration of 6%, the defected GNs and the defected GO exhibited peak quantum capacitance of 252.87 μF/ cm 2 and 116.01 μF/cm 2 , respectively. Thus, the defected GNs clearly provide superior quantum capacitance over the defected GO and pristine GNs, which makes it an ideal choice for electrode active layer material of supercapacitor devices.
Thus, in order to make a good agreement with computational analysis, waste plastic derived GNs was used as active layer materials for graphite based supercapacitors with a geometry of Graphite sheet|GNs|1 M www.nature.com/scientificreports/ H 3 PO 4 |GNs|Graphite sheet. The defected GNs are mixed with the binder PVDF-HFP (Poly (vinylidene fluoride-co-hexafluoropropylene)) in 10% acetone solution for 5-6 h over magnetic stirrer to make sticky paints for the fabrication of the graphite based electrodes. The cyclic voltammetry of fabricated cell (Fig. 10a) showed excellent reversibility over the performed voltage range at the scan rate of 5 mV/s, 10 mV/s, 50 mV/s and 100 mV/s. The CV curve showed nearly rectangular shape due to lower contact resistance and formation of electrical double layers which showed the one of the features for a good supercapacitor. The specific capacitance via CV shown in Table 2 for the fabricated cell is given by where i is the current, s is the rate of the scan. From the CV curve and using the above equation, the fabricated cell showed the highest specific capacitance of 398 F/g at the scan rate of 5 mV/s, while it is found to be 340 F/g, 254 F/g and 215 F/g for the scan rate of 10 mV/s, 50 mV/s and 100 mV/s, respectively. Further, EIS of the fabricated cell (Fig. 10b) shown by Nyquist plot were performed to investigate the equivalent sheet resistance (ESR) and specific capacitance at 10 MHz. Figure 10b showed the steeply rising incident toward the lower frequency region which indicates the capacitance behavior shown by the GNs with the electrolytes due to the formation of the electrical double layer at the interface of electrode and electrolyte. The device showed lower ESR value of 4.78 Ω/cm 2 , which showed the sophisticated feasibility of the ion transportation within the GNs. In addition to it, the small semicircle shown by the EIS curve in the higher frequency region indicated the smaller charge transfer resistance, which arises due to the bulk properties of both electrolyte and the contact of GNs and graphite sheet for the fabricated cell. Further, including all the parameter of the EIS study, the specific capacitance for the same cell via EIS curve were calculated by using the following equation where f is the frequency of applied AC signal, Z″ is the imaginary part and Z′ is the real part of impedance at 10 MHz. The specific capacitance via this technique and using above equation is found to be 221.77 F/g. Figure 10c shows the galvanostatic charging discharging (GCD) behavior of the cell at 1 A/g, 2 A/g, 5 A/g and 7 A/g of current densities over the potential window of 0-1 V. The cell showed nearly triangular shape of the curve over the performed current densities which arise due to capacitive behavior of the cell. Further, these curves used for the estimation of specific capacitance behavior of the both cell by using the following equation where i is the scan rate, m represent the mass loading over the electrode and Δt/ΔV is the linear part slop in the discharge curve. The evaluated specific capacitance has shown in Table 3 for the fabricated cells by using the above equation. The highest specific capacitance value shown by the cell is found to be 273.69 F/g at 1 A/g of current density. Additionally, the fabricated cell shows the very appreciable energy density (E d ) and power density (P d ) of 38 Wh/kg and 1009.74 W/kg, respectively which were calculated by using the following equation   www.nature.com/scientificreports/ 15% increase in the emission of green house gases (GHGs) 76 . However, utilizing this waste as a source material to synthesize graphene can establish an uninterrupted supply chain of raw material, and also save the environment from the exponential increment of carbon footprint. It has been estimated based on the previous reports that fabrication of highly efficient graphene based DSSC and supercapacitor of 1 cm 2 area requires only 1-10 mg of GNs. Therefore, if the waste plastic is efficiently collected and upcycled through the proposed method then the resulting GNs will be sufficient to fulfill the energy conversion, storage, and other applications as per the contemporary market demand.
Cost benefit assessment. The cost associated with the production of GNs from waste plastic through the proposed industrial symbiosis show a significant 100-fold benefit in comparison to the cost of commercially available GNs. The production cost has been estimated by considering all possible costs which include labor, transportation, electricity, chemicals, sealing, packing and management costs. The cost of producing 10 g of the graphene though our industrial symbiosis process is estimated to be at max 1 USD, whereas the cost of purchasing 10 g of commercially available graphene ranges between 100 and 200 USD. The production cost may further be reduced through bulk synthesis of the GNs and automation of the process. Thus, our proposed method is not only viable but also quite cheaper. To conclude, the conversion of waste plastics into GNs though the proposed method could have dramatic positive impact on both ecology and economy.

Conclusion
The sustainable environmental protection from the plastic waste could be achieved by diverting this waste to the upcycling processes rather than the traditional recycling processes, as the former offers greater benefit from both the economic and ecologic perspectives. This work presents an industrial symbiosis process for conversion of the waste plastic into grapehene nanosheets (GNs) and their subsequent utilization as a case study for the fabrication of low cost energy conversion and storage devices viz. DSSCs and supercapacitors, respectively. Besides the experimental synthesis and device fabrication, we have also performed the first principle computations to understand and support the experimental results. The DSSC fabricated with GNs as semiconducting layer with polymeric electrolyte exhibited an open circuit voltage (V oc ) of 0.77 V with a short circuit current density (J sc ) of 0.302 mA/cm 2 . The fabricated supercapacitor with GNs as active layer of electrode material offered an impressive specific capacitance of 398 F/g with a scan rate of 0.005 V/s. The cost benefit analysis revealed that the proposed upcycling process has the potential to revolutionize waste plastic management and GNs production segments, as the process offers high value added recovery in comparison to the traditional recycling methods, and also produces the GNs about 100 fold cheaper than the cost of acquiring the commercially available GNs.

Methods
Synthesis of graphene nanosheets (GNs). Synthesis of GNs from waste plastics has been performed as per our previously reported procedure 53 . However, in the present work, we have modified our method by changing the temperature range to demonstrate the application of the GNs for DSSCs and supercapacitors. Briefly, waste plastic were collected from local market and categorized into polypropylene (PP), polyethylene (PE) and polyethylene tetrathalate (PET), as these plastic contains maximum amount of carbon in comparison to other kind of plastics. After the categorization, the real world mix plastics (PP, PE and PET) were shredded into the small flakes (average size of 10 mm × 5 mm) with the help of shredder unit, which were later on washed with sodium hydroxide and glass washing soap solution (laboratory grade) to remove all the muddy and oily impurities from the mixed flakes of the plastics. Then the flakes were dried inside the drier chamber connected with hot air gun. Once the mixed flakes of the plastics were dried, mixing of the bentonite nanoclay as degradation agent was done inside the mixing chamber. Inside the mixing chamber, washed and dried mix plastics were mixed with bentonite nanoclay in a ratio of 1000:3 and mixed for 30 min in order to assure uniform mixing of the nanoclay.
Here we take 20 kg of the waste plastics and 60 g of the bentonite nanoclay according to the mentioned ratio.
After the proper mixing of the bentonite nanoclay with mixed plastics, process of the primary pyrrolysis was performed inside the indigenously developed pyrrolysis chamber in the inert atmosphere of N 2 with the flow rate of 10 ml/min at the temperature of 450 °C (heating rate 9 °C/min) for 50 min. This step of the pyrolysis removes all the petroleum products while leaving back the carbon skeleton in the form of black charred residue. The black charred residue thus obtained then undergo for the secondary stage pyrolysis process. In our previous work, the synthesis of the graphene nanosheets was depicted at the temperature of 750 °C. However, in the present work we raise the temperature by 945 °C for the reduction of the graphene nanosheets to increase the number of sp 2 hybridized carbon atoms within the graphene nanosheets, so that these GNs were implemented as a part of photoanode material in DSSC and active layer material in supercapacitor devices, respectively.
Fabrication of DSSC. The fabrication of DSSC was carried out by following a previously reported method 76,77 . 8 mg of the prepared GNs was added to 4 ml polyethylene glycol (PEG, Sigma Aldrich, Mw = 200) in an eppendorf vial and sonicated for 4-5 days in order to obtain a uniform solution. 50 μl of the prepared solution was spin-coated at ~ 1200 rpm for 60 s on fluorine doped tin oxide (FTO) glass. This was followed by the calcination of FTO glass at 350 °C for 1 h in order to remove the polymer. Two scotch tapes were pasted on both sides of the 'almost transparent' layer of GNs for controlling the thickness of the TiO 2 paste (Alibaba, China). The viscous paste was spread uniformly by doctor blade method followed by calcination at 500 °C for 30 min. The electrode after bringing down to room ambient temperature was immersed in an ethanolic solution of Ruthenium dye (Solaronix, Switzerland) for 6 h. The standard PEO:PEG electrolyte with Iodine redox couple and Platinum coated counter electrode was used for completing the sandwich structure as detailed in previous reports 78 . www.nature.com/scientificreports/ Fabrication of supercapacitors. To fabricate an Electric Double Layer Capacitor (EDLC), we have used 1 cm × 1 cm graphite sheet (current collector) and the GNs as an active layer material with binder PVDF-HFP (Poly (vinylidene fluoride-co-hexafluoropropylene)) (Sigma Aldrich, USA) in the ratio (90:10) dissolved in acetone. The GNs was well coated on current collector and kept in laboratory oven at 100 °C overnight. 1 M H 3 PO 4 with a millipore filter paper (0.45 micron) as separator was sandwiched between the electrodes to perform the Cyclic Voltammetry (CV), galvanostatic charge-discharging (GCD) and Electrochemical Impedance Spectroscopy (EIS) measurements.
Characterizations. GNs synthesized from waste plastics were characterized by various spectroscopic and imaging techniques as previously reported 53 . However, after thermal reduction of the GNs Raman spectroscopy was performed by using Renishaw Raman Microscope to understand the effect of thermal reduction on the GNs. Furthermore, X-ray diffraction (XRD) analysis was performed by using Rigaku diffractometer with Cu-Kα radiation. FT-IR spectroscopy was performed to analyze the functionality of the GNs by using PerkinElmer spectrum-2. TGA analysis was performed by TGA 4000, Perkin Elmer. Additionally, the surface morphology analysis of the GNs was analyzed through FESEM using NOVA NanoSEM 450. High resolution transmission electron microscopy (HRTEM) was performed by using JEOL, JEM 2100 microscope to get analyze the thickness and crystalline behaviour of the GNs. Sheet resistance of the GNs was analyzed by four probe setup using Keithley 2400 SourceMeter. Current-Voltage characteristics of the fabricated GNs based DSSC was carried out by using Keithley 2400 SourceMeter. Cyclic Voltammetry (CV) analysis, galvanostatic charge-discharging (GCD) and EIS spectroscopy was performed for measuring the impedance between the junctions of the supercapacitor cell by using CHI 604D electrochemical workstation.

Data availability
The data that support the plots within this paper and other findings of this study are available from the corresponding author upon reasonable request.