Is graphene a good transparent electrode for photovoltaics and display applications?

: The current standard material used for transparent electrodes in displays, touch screens and solar cells is indium tin oxide (ITO) which has low sheet resistance (10 Ω / □ ), high optical transmission in the visible wavelength (85%) and does not suffer of optical haze. However, ITO is mechanically rigid and incompatible with future demands for flexible applications. Graphene materials share many of the properties desirable for flexible transparent conductors, including high optical transparency, high mechanical flexibility and strength. Whilst pristine graphene is not a good transparent conductor, functionalised graphene is at least 1000 times a better conductor than its pristine counterpart and it outperforms ITO. Here the authors review recent work on a novel graphene-based conductor with sheet resistance as low as 8.8 Ω / □ and 84% optical transmission. This material is obtained by ferric chloride (FeCl 3 ) intercalation into few-layer-graphene (FLG), giving rise to a new system which is the best known flexible and transparent electricity conductor. FeCl 3 -FLG shows no significant changes in the electrical and structural properties for a long exposure to air, to high levels of humidity and at temperatures of up to 150°C in atmosphere. These properties position FeCl 3 -FLG as a viable and attractive replacement to ITO.


Introduction
Over the past decades, the landscape of consumer electronics and optoelectronics has changed drastically. Technological advances have led to the development of portable media devices, such as tablets, smart phones and even wearable electronics. At the same time, the development of solar cells has enabled a viable alternative method of producing energy or revolutionary products such as prototypes of solar powered airplanes. However, with all these technological advances also the consumer demands are evolving and new products are under constant development. For example, flexible and more durable mobile phone screens are required as there is the chronic issue of broken displays due to dropping. The current technologies for displays, touch screens and solar cells use indium tin oxide (ITO) as a transparent electrode. However, this material is rigid and flexing or bending ITO causes a significant degradation in its conductive properties. Therefore, future flexible displays and photovoltaics, and the associated electronics require materials that are flexible, optically transparent and electrically conductive.
For a transparent conductor it is vital that the sheet resistance is low. High resistance electrodes have several major issues. First, the resistance of the electrode limits the size of the manufacturable electrode. When using the electrode in a display or solar cell the electrode should have a uniform potential, however, with high resistance electrodes there is a voltage drop as the distance from the contacts connecting the electrode to the rest of the circuit is increased. This leads to a non-uniform voltage potential across the electrode giving non-uniform light emission. To minimise this effect the size of the electrode must be limited. Secondly, the voltage drop across the electrode is dissipated as heat. Heating of the electrode can cause components such as LED pixels to emit more or less light and can reduce the lifetime of the pixels leading to an increase in device faultier rates. Finally, screens are becoming more portable and battery operated as part of high-tech electronics. The dissipation of energy across the electrode increases energy consumption and reduces the battery lifetime of the device. This reduction in the device efficiency limits the range of products that the electrodes can be used in. Similarly for photovoltaics, high resistance electrodes reduces the efficiency of the cells and causes unnecessary thermal heating which also acts to reduce the device lifetime.
ITO, which is the current standard material used for transparent electrodes in displays, touch screens and solar cells, has low sheet resistance (10 Ω/□) while having a high optical transmission (85%) with no optical hazing [1,2]. However, ITO is mechanically rigid and incompatible with flexible applications [1,2]. Monolayer and few-layer-graphene (FLG) share many of the properties desirable for flexible transparent conductors, including high optical transparency [3], high mechanical flexibility and strength [4,5]. However, for graphene materials to fulfil the requirements of transparent conductors they must have similar performance as ITO films. Currently, monolayer graphene has been shown to have high optical transmission (>97%), however, the sheet resistance is typically in excess of 1 kΩ/□. This leads to a higher dissipation of energy in the electrode material, hence reduced device efficiency. To make graphene viable as a transparent conductor it must be functionalised to reduce the sheet resistance while maintaining the high optical transmission. A reduction in the sheet resistance of monolayer and FLG has been achieved through chemical functionalisation, where the best results yield 30 Ω/□ with 90% optical transmission and 125 Ω/□ at 97.7% optical transmission, respectively [4,6]. However, this is still three times larger than values reported for 10 μm thick ITO [7].
Several types of dopants have been used for the chemical functionalisation of monolayer and FLG, such as gold chloride (AuCl 3 ) [ 6,8], ferric chloride (FeCl 3 ) [9][10][11], nitric acid (HNO 3 ) [4,8], thionyl chloride (SOCl 2 ) [ 12], bis(trifluoromethanesulfonyl) amide (TFSA) [13] and tetracyanoquinodimethane (TCNQ) [14]. Table 1 shows the sheet resistance and optical transmission for different functionalisation of graphene. Among these dopants, AuCl 3 was shown to decrease the sheet resistance of monolayer graphene to 150 Ω/□, HNO 3 down to 125 Ω/□ and FeCl 3 to 72 Ω/□. However, while decreasing the sheet resistance is one of the most significant aspects for transparent electrodes, other properties such as stability in air, at high temperature and humidity, and resistance to solvents may determine what type of doped graphene is suitable for applications in displays and photovoltaics. Thus, it has been reported that HNO 3 and AuCl 3 doping of monolayer and FLG is unstable in air, which makes graphene materials functionalised with these dopants not a viable ITO replacement in displays and photovoltaic applications. On the other hand, while FeCl 3 doped monolayer graphene is resistant to solvents such as anisole and isopropanol, it is very unstable to humidity due to the reaction of FeCl 3 with water [9], which is again undesirable for applications.
Another approach to produce low sheet resistance carbon-based electrodes is to use very thick graphitic layers (see Table 1). Examples include stacking multiple graphene layers [15]o r producing graphite films through laser conversion of polymers [16]. Thus the laser induced graphite yields a sheet resistance of 17 Ω/□ for a 25 µm thick film [16]. However, while this method improves the overall resistance of the electrode, it leads to a very low optical transparency. At the same time with such thick layers as electrodes, the device efficiency remains inferior to that of their ITO-based counterparts [15].
Here we review recent work on a novel graphene-based transparent conductor with a sheet resistance as low as 8.8 Ω/□ and 84% optical transmission in the visible wavelength. This material is obtained by intercalation of FeCl 3 into FLG. In this case, a strong charge transfer occurs between the graphene-and the intercalant-layers, resulting in heavy p-doping of graphene. This gives rise to a new system which is the best known flexible and transparent material able to conduct electricity. A complementary study of Raman spectroscopy, optical transmission spectroscopy and electrical transport techniques is used to characterise the intercalation of FeCl 3 and confirm the suitability of the resulting material as a transparent conductor. Furthermore, a stability study shows that FeCl 3 -FLG can withstand relative humidity of up to 100% at room temperature for 25 days, as well as temperatures of up to 150°C in airor as high as 620°C in vacuum as demonstrated by in-situ high resolution transmission electron microscopy measurements. This durability to extreme conditions position FeCl 3 -FLG as a viable and attractive replacement to ITO in electronics, such as 'smart' mirrors or windows, or even solar panels. Finally, we show that FeCl 3 -FLG can be employed effectively as a transparent electrode in atomically-thin photodetectors.

Intercalation of FLG
Intercalation is a process where molecules or atoms can enter at the edges of domains of graphite or FLG structures and then diffuse to form a continuous layer between the graphene sheets [10,11], see Fig. 1. This process is markedly different from doping of FLG [4,6,8,9,[12][13][14], where first the monolayers are doped and then sub-sequentially staked to produce the few-layer. Intercalation of various species, such as FeCl 3 [10,11,[19][20][21], Li [22], K [23,24], Rb [24], Br [25] and Ca [26], in FLG has been demonstrated so far. The addition of intercalants changes the electrical properties of the FLG. Moreover, there are many other properties that have been reported in intercalated graphite samples. For example, through the intercalation of magnetic molecules such as FeCl 3 into graphite inter and intra plane magnetism has been achieved [27], where the magnetic layers in the structure are as thin as 0.5 nm, many times thinner than magnetic films fabricated by thin film deposition techniques [28]. Furthermore, with the intercalation of metals such as lithium and calcium in graphite, there has been an observed superconducting phase transition [29,30].
The intercalation of graphene structure increases the charge carrier concentration and hence the Fermi energy. This populates higher energy states with charge carriers leading to a higher total number of conducting state, hence to a higher conduction and lower sheet resistance [10]. For example, FeCl 3 intercalation has been shown to reduce the sheet resistance of FLG down to 8 Ω/□. More recently, low values of resistivity have also been reported in graphitic lithium intercalated films (sheet resistance of 3 Ω/□ [22]). While the conductivities of the FLG and ultrathin graphite are significantly reduced, these systems have a high optical transmission (above 80%) [10,22]. In this case, a reduction of the graphite film thickness would improve the optical transmission of the material at the expenses of higher electrical resistivity as expected by Ohm's law. These studies have shown that intercalation is a promising route for tailoring the properties of graphene for the application of transparent conductive electrodes. However, the stability of some intercalated compounds in air, specifically in the presence of moisture and oxygen, is generally poor. For example, it has been reported that lithium intercalated ultrathin graphite is not stable in air and rapidly de-intercalates quickly making the material highly resistive [22]. On the other hand, there are studies demonstrating the stability of FeCl 3 doped graphene to a range of solvent environments [9] as well as the extreme stability of FeCl 3 intercalated few graphene in high humidity and high temperature in air [31].
The intercalation of few layer graphene with FeCl 3 is performed using the two zone vapour transport method [10]. The FLG samples are loaded into a glass tube with anhydrous FeCl 3 powder shown in Fig. 1. The tube is evacuated to 1 × 10 −5 mbar using a turbo molecular pump and sealed using an isolation valve. The FLG sample and the FeCl 3 powder are placed in two different zones, denoted as T Sample and T Powder which are heated to 360 and 315°C , respectively. The FeCl 3 powder sublimes while it is heated, releasing FeCl 3 vapour along with a small amount of Cl 2 gas and solid FeCl 2 . As the FeCl 3 vapour passes over the sample it intercalates the FLG. In the final stage of the process T Powder is reduced first to stop the sublimation of FeCl 3 and then the rest of the system is cooled.
To make FeCl 3 intercalated graphene viable for next generation electrical applications, there are several approaches for producing large area FLG samples (from 1 cm 2 up to few m 2 ). One approach is to use FLG grown on silicon carbide (4H-SiC) [32]. This method of growth produces high quality films, of size up to 4-inch wafers, with little variation in the number of layers grown and is combined with a substrate which is compatible with the FeCl 3 intercalation procedure. Other methods of growing large area FLG to be used for intercalation with FeCl 3 involve chemical vapour deposition techniques [33] such as multi-layered graphene grown on nickel and transferred onto a silicon or glass wafer, or multi-layered graphene formed by positioning upon one another a series of monolayer graphene grown on copper. For the experiments discussed in this review, 1 cm 2 epitaxial FLG was obtained through the thermal decomposition of Si face of a 4H-SiC substrate. The C face was cleaned of any graphene present using plasma etching. The sample was then intercalated with FeCl 3 for 24 h using the procedure outlined above. Fig. 2a shows the Raman spectra for FLG, obtained by mechanical exfoliation on SiO 2 /Si, before the intercalation procedure. For the pristine graphene, with increasing the number of layers there is an increase in the G-peak intensity and an evolution of the multi peak structure of the two-dimensional (2D) peak. For the few-layer epitaxial graphene there is a large contribution of the SiC substrate in the spectral region of the G-peak. Fig. 2c shows the G and 2D peaks, after subtracting the SiC Raman background. The multi peak structure of the 2D peak indicates that the sample is not monolayer [34]. It has been shown that for FLG grown on SiC the full width at half maximum (FWHM) of the 2D peak exhibits a linear dependence on the inverse number of layers (N): FWHM (2D) = [−45(1/N ) + 88] cm −1 [17]. The 2D peak of the samples discussed in this review has a FWHM of 72 cm −1 , from which a layer number of 3 is estimated.

Raman spectroscopy of FeCl intercalated FLG
After the intercalation, for both mechanically exfoliated and epitaxially grown FLG, changes in the Raman spectra are observed due to the charge transfer from the FeCl 3 molecules [10,11,20,21]. First an upshift in the position of the G-peak occurs, which is related to the configuration of FeCl 3 and the graphene layers in the structure [10,20,21], shown in Fig. 3. Specifically, for each layer of graphene which has one adjacent layer of FeCl 3 the G-peak is upshifted to 1612 cm −1 , known as the G 1 peak. Similarly, for each layer of graphene sandwiched by two FeCl 3 layers the G-peak is upshifted to 1624 cm −1 , i.e. the G 2 peak. The frequencies, line widths and line shape of the G 1 and G 2 peaks do not depend on the number of graphene layers. The second change in the Raman spectra observed upon FeCl 3 intercalation is that the multi-peak structure of the 2D-peak becomes single peaked, as demonstrated in Fig. 3. In addition, in the case of intercalated FLG on SiC the 2D peak position downshifts to that of monolayer graphene on 4H-SiC (2715 cm −1 ) [17]. The observed changes in the Raman spectra upon intercalation with FeCl 3 are consistent with observations from studies where FLG was intercalated with potassium and rubidium [23,24]. However, in the case of a partial intercalation, identification of the structural composition cannot rely uniquely on the Raman spectra. Further confirmation is required using complementary electrical transport measurements [10].

Electrical properties of FeCl 3 intercalated FLG
The electrical properties of pristine and FeCl 3 intercalated graphene mechanically exfoliated on SiO 2 /Si were measured in devices which  [10] b Schematic crystal structure for a graphene sheet with one or two adjacent FeCl 3 layers. Reproduced with permission in [10] c Raman spectra for epitaxial pristine FLG (top) and FeCl 3 intercalated FLG (bottom). White curve is a fit of the 2D peak to a multiple Lorentzian function (grey curves). Upon intercalation with FeCl 3 a single Lorentzian is the best fit of the 2D peak (white). Reproduced with permission in [11] were patterned into a Hall bar geometry and electrically contacted with Au/Cr electrodes using electron beam lithography, Ar plasma etching (for patterning), thermal evaporation and lift-off (for Au/ Cr). The sheet resistance of the pristine and FeCl 3 intercalated epitaxial FLG was measured in a four terminal configuration in macroscopic samples, with a conductive channel of fixed width (0.7 cm) and channel length ranging from 0.7 to 4.2 mm using Ti/ Au (5/50 nm) contacts.
The sheet resistance of pristine and FeCl 3 intercalated FLG with 2 up to 5 layers was measured as a function of temperature, as shown in Fig. 3a. The room temperature sheet resistance of bilayer graphene is as high as 2.5 kΩ/□ and is monotonously decreasing to 120 Ω/□ for five-layer-graphene. Furthermore, a semi-metallic temperature dependence is observed, where the sheet resistance of all of the pristine graphene samples increases with lowering the temperature. On the contrary, the sheet resistance of FeCl 3 intercalated FLG, Fig. 3b, shows a significant reduction in the room temperature value for all thicknesses. The sheet resistance decreases as the layer number increases reaching a value as low as 8 Ω/□ for five layers FeCl 3 -FLG. Furthermore, FeCl 3 -FLG shows the temperature dependence of a metallic conductor, where the resistivity decreases with lowering the temperature. Fig. 4a shows that the pristine epitaxial graphene has an average sheet resistance of 174 ± 9 Ω/□, estimated from the linear fit. The corresponding values of resistance in the same devices after intercalation with FeCl 3 are systematically lower with an average value of 16.6 ± 0.6 Ω/□. The distribution of sheet resistance of pristine and FeCl 3 intercalated epitaxial graphene is presented as two histograms, as shown in Figs. 4b and c. In both cases, there is a single peaked distribution with narrow spreading of the sheet resistivity demonstrating the homogeneous and reproducible electrical properties in large area FeCl 3 intercalated graphene.
To reveal the origin of the low sheet resistance of FeCl 3 -FLG, the Hall resistance (R xy ) was measured as a function of applied perpendicular magnetic field (B), as shown in Fig. 5a. Fitting the linear Hall resistance allows for the estimation of the charge carrier concentration (n) using the equation R xy = V xy /I = −(1/ne)B with V xy is the measured Hall voltage, I is the current though the channel and e is the electronic charge. All samples thicker than bilayer have a high charge carrier concentration between 4 × 10 14 and 9 × 10 14 cm −2 , as shown in Fig. 5b. For large area epitaxial samples, a concentration of n > 1.5 × 10 14 cm −2 was achieved. The charge carrier concentration far exceeds the values achievable with top-gates on SiC [18] and is comparable to that achieved in electric double layer FLG transistors [35]. Furthermore, the sign of the fitted gradient shows the dominant charge carriers are holes. This record high charge carrier concentration is three orders of magnitude larger than pristine graphene. The increased charge carrier concentration increases the Fermi energy, hence the increase in the conduction of the material. To demonstrate that there is no degradation of the electrical quality of the intercalated graphene, the charge carrier mobility is estimated, μ H = 1/(n H eρ xx ) with ρ xx is the longitudinal resistivity and e is the electron charge, which is shown in Fig. 5c. Charge carrier mobilities similar to pristine graphene are observed in the intercalated samples, with a maximum value of 3650 cm 2 /(V s). Furthermore, the room temperature mean free path for five-layer-intercalated graphene is 0.6 μm, suggesting low amounts of disorder or structural damage present after the intercalation process.

Optical properties of FeCl 3 intercalated FLG
The optical properties of FeCl 3 -FLG were studied in samples with the different number of layers deposited onto the transparent glass substrates. To monitor the change in the optical transmission the measurements were done before and after the intercalation of FLG. These devices are fabricated on SiO 2 /Si substrates and the highly-doped Si substrate is used as a gate to adjust the Fermi level to the charge neutrality of the system. Reproduced with permission in [10] b Temperature dependence of the square resistance for FeCl 3 -FLG of different thicknesses. Reproduced with permission in [10] Fig. 5 Hall coefficient measurements for the estimation of carrier density and mobility of FeCl 3 -FLG a Hall resistance of FeCl 3 -FLG as a function of magnetic field. Inset shows the data for the bilayer sample on a smaller B scale. Reproduced with permission in [10] b Carrier density for FeCl 3 -FLG as a function of the number of graphene layers. Reproduced with permission in [10] c Carrier mobility for FeCl 3 -FLG as a function of the number of graphene layers. Reproduced with permission in [10] Fig. 4 Electrical properties of large area pristine and FeCl 3 intercalated graphene epitaxially grown on SiC a Plot of the measured four terminal resistance (R) as a function of the ratio between the conductive channel length (L) and width (W). Top data points correspond to pristine FLG on 4HSiC while the bottom data points correspond to FeCl 3 -FLG on 4H-SiC. Inset shows a scheme of the configuration used to measure the resistivity. Reproduced with permission in [11] b, c Histograms showing the distribution of resistivity for pristine FLG (b) and FeCl 3 -FLG (c), respectively. Reproduced with, permission in [11] Fig. 6a shows the optical transmission of pristine FLG with the different number of layers while Fig. 6b shows the optical transmission after their intercalation with FeCl 3 , both for the visible wavelength range of light (400-850 nm). Upon intercalation, the transmittance slightly decreases at low wavelengths, but it is still above 80%. In particular for 550 nm, the transmission of pristine FLG ranges from 97.7% for monolayer graphene to 88.5% for five-layer-graphene. The FeCl 3 intercalated bilayer graphene has a transmission of 94.5%, which reduces monotonously to 84.1% for five layers. For wavelengths longer than 550 nm we observe an increase in the optical transparency of FeCl 3 -FLG. This is a significant advantage of FeCl 3 -FLG material compared to ITO whose transparency decreases for wavelengths longer than 600 nm [2]. This property will provide useful applications that require conductive electrodes which are transparent both in visible and near infrared ranges. For instance, FeCl 3 -FLG transparent electrodes could be used for solar cells to harvest energy over an extended wavelength range as compared to ITO-based devices.

Stability of FeCl 3 intercalated FLG
The FLG intercalated with different species such as Li [22], K [23,24], Rb [24], Br [25] and Ca [26], are not stable upon exposure to air, resulting in their de-intercalation. On the other hand, FeCl 3 is highly soluble in water, therefore it would be intuitive to suggest that FeCl 3 intercalated graphene would not be stable in air, due to the ambient humidity, quickly reducing the material to pristine graphene. However, it was found that FeCl 3 -FLG is stable in a variety of conditions and shows no significant de-intercalation [10,31]. This stability was demonstrated with four experiments: (i) material processing using standard electron beam lithography; (ii) leaving the sample exposed to atmosphere for 1 year and comparing the Raman spectra before and after; (iii) heating the sample in atmosphere up to 150°C and measuring the room temperature resistivity; and (iv) exposing a sample to high levels of humidity (>95%) and measuring the resistivity.
We use Raman spectroscopy to characterise the homogeneity of the FeCl 3 intercalation in FLGs after the fabrication of electrical contacts using standard electron beam lithography. During this process, the samples are coated with a Poly(methyl methacrylate) (PMMA) bilayer (495 K PMMA and the second layer is a 950 K PMMA), both are baked at 180°C for 3 min, followed by electron beam writing, development in a solution of Methyl isobutyl ketoneisopropyl alcohol (MIBK-IPA), thermal evaporation for the metallic contacts, lift-off in acetone and rinsing in IPA. Fig. 7a shows a micrograph picture of the 5 L intercalated device. A direct comparison of the Raman spectra measured at ten different locations on the device (Fig. 7b), as highlighted in the picture, shows no appreciable variation in either intensity or position of the Raman peaks. This demonstrates that the intercalated flakes have a high homogeneity, i.e. the structure is the same at any point of the flake. Fig. 7c shows a comparison between Raman spectra collected at different positions on a FeCl 3 -FLG sample, after keeping the samples for 3 months and one year in air. The peak positions shape and relative intensities remain the same. If de-intercalation had occurred, an increase in the G peak and subsequent reduction in the G 1 and G 2 peaks should have been occurred. This is not observed allowing to conclude that FeCl 3 intercalated graphene is stable in atmosphere at room temperature.  [10] b Raman spectra of a FeCl 3 -FLG devices taken at different locations as indicated in (a). Reproduced with permission in [10] c Raman spectra of the same FeCl 3 -FLG device as in b taken at different locations after 3 months and after one year. Reproduced with permission in [10] IET Circuits Devices Syst., 2015, Vol. 9, Iss. 6 A third experiment was conducted to determine the stability of FeCl 3 intercalated graphene up to temperatures of 150°C in atmosphere. The Raman spectra before and after heating are shown in Fig. 8a. No changes are observed in the positions of the G 1 and G 2 peaks before and after heating the sample to 150°C for 30 min. This demonstrates that there is no de-intercalation occurring in the sample. Furthermore, the stability at temperatures up to 150°C was tested using electrical transport measurements. For these measurements, the resistivity was first measured at room temperature, the sample was heated to 50°C for 30 min, the sample was cooled to room temperature and the room temperature resistivity was measured. The temperatures the sample was exposed to where incrementally increased from 50 to 150°C, shown in Fig. 8b, with the temperature profile is shown in Fig. 8c. As the resistivity of the sample does not change, there is no de-intercalation while the sample is heated in an atmosphere, reinforcing the results from the Raman maps shown in Fig. 8a.
Finally, the stability of FeCl 3 -FLG was demonstrated in high humidity conditions. FeCl 3 -FLG was loaded into a chamber where the humidity can be controlled from 0 to 95%, and the resistivity was measured in-situ. The resistivity of the sample was measured before humidity exposure, then the sample was exposed to >95% humidity for 1 day. The sample was dried and the resistivity was re-measured. This process was repeated until the cumulative exposure was 24 days. Fig. 9a shows the resistivity upon drying the samples against the exposure time to humidity. There was no significant change in the resistivity after the prolonged exposure to high levels of humidity. To complement the electrical transport study for high humidity exposure a Raman study was performed for the same sample. Fig. 9b shows the Raman spectra of the same region on the sample after exposure to high levels of humidity. No change is visible in the relative intensities of the G 1 and G 2 peaks after an exposure of 25 days.
Combining the information from the above experiments demonstrates the stability of FeCl 3 -FLG for long atmospheric exposure, temperatures of up to 150°C and 24 days at >95% humidity.

FeCl 3 intercalated FLG as a transparent conductor
The high optical transparency observed in FeCl 3 -FLG is complemented by the remarkable electrical conductivity and outstanding stability to a variety of conditions such as air, humidity and high temperature. However, to replace materials such as ITO as a transparent electrode, the properties should exceed or Fig. 8 Stability of FeCl 3 -FLG to heating in atmosphere a Raman maps showing the G1 and G2 peaks before and after heating to 150°C. Reproduced with permission in [31] b Sheet resistance measured at room temperature after exposure to heating detailed in (c). Reproduced with permission in [31] Fig. 9 Stability of FeCl 3 -FLG to humidity a Sheet resistance of FeCl 3 -FLG measured after exposure to >95% humidity from 0 to 24 days. Inset picture shows the measured device. Graphs in the inset show the in situ Rsq against time and the relative levels of humidity. Reproduced with permission in [31] b Raman spectra for FeCl 3 -FLG after exposure to humidity (>95%) for 0 to 25 days. Reproduced with permission in [31] match that of commercially available ITO (R s =10Ω/□ with a 550 nm transmission of 85%) [7]. Fig. 10 shows the 550 nm transmission and conductivity of ITO and the leading air-stable carbon-based materials such as doped carbon nanotube arrays [4,36], doped monolayer graphene and FeCl 3 -FLG. The multiple points represent different thicknesses of the materials, for example ITO has a resistivity of 7 Ω/□ and 550 nm transmission of 75% for a thick film, where reducing the thickness increases the resistivity to 100 Ω/□ with a 550 nm transmission of 90%. Reducing the material thickness increases the transmission, but reduces the parallel conductance in the film, hence increasing the resistivity, this leads to the bounding grey area highlighting the empirical dependence of resistivity on the 550 nm transmission. In the comparison, the ideal transparent conductor would have a high optical transmission and low resistivity would tend to the bottom right corner of the graph. It is apparent that the resistivity and optical transmission of the five-layer-intercalated graphene outperform the commercially available ITO films. Furthermore, the doped carbon nanotubes and doped monolayer graphene are poor performers, when compared to FeCl 3 -FLG.

Atomically thin photo-detectors using FeCl 3 intercalated FLG as transparent electrode
The conversion of photons into electrical signals is of tremendous societal importance since it is used in a multitude of technologies. Photodetectors based on graphene-metal interface [37][38][39][40][41] are at the core of a new generation of ultrafast devices with a remarkable high bandwidth (500 GHz), zero source-drain bias (hence zero dark current) operation, and good internal quantum efficiency. In these devices, photocurrent is generated by illumination of one of the metal/graphene interfaces of a graphene field-effect transistor. However, these devices employ opaque metallic nanostructures which would introduce significant haze caused by light scattering when used in smart windows and mirrors. Therefore, metal electrodes were replaced by FeCl 3 -FLG [42] to realise an all-graphene-based photodetector. The leap to all-graphene structures which have a high transparency would enable the development of a new generation of transparent photovoltaic devices which do not suffer from haze. In this section, we review the optoelectronic properties of these novel graphene/FeCl 3 -FLG heterostructures which were investigated using photovoltage spectroscopy.
The graphene/FeCl 3 -FLG photodetector was fabricated as follows: pristine FLG was first deposited by mechanical exfoliation onto heavily doped Si/SiO 2 substrate and intercalated with FeCl 3 . Subsequently, a pristine FLG flake was transferred over the FeCl 3 -FLG flake. Independent multiple electric contacts were made by Cr/Au (5 nm/70 nm) to the bottom FeCl 3 -FLG and top FLG allowing the characterisation of these heterostructures. Fig. 11 shows the schematic of the final device along with the measurement setup. The FeCl 3 -FLG flake was kept on the ground, while a small dc bias of 0.1 mV was applied to the pristine FLG flake. The graphene photodetector was then illuminated by a 532 nm HeNe laser focused by using a 100 × objective to 1.5 μm spot size at a power of 8.2 μW. The beam was chopped at 370 Hz, and the chopper was used as reference to a lock-in amplifier which measures the photovoltage. The heavily doped Si substrate acts as a global back-gate which was used to tune the chemical potential of graphene, whereas the resistivity of the FeCl 3 -FLG flake was unaffected by the typical values of used gate voltage due to the high doping level. The devices were mounted on a scanning stage, which allows the mapping of the photoresponse of these graphene-based heterointerfaces in the x-y directions with a spatial resolution of 1 μm. The optoelectronic properties of these graphene-based hybrid structures are characterised by measuring the photovoltage generated across the pristine FLG/FeCl 3 -FLG interface while rastering the laser spot over the active device area.   [42] a Schematic of the device design and the experimental setup for photovoltage spectroscopy b Crystal structure for fully intercalated FeCl 3 -FLG and for FLG flakes Fig. 12 shows the photovoltage generated in a typical device as a function of position of the laser beam. It is apparent that there is a strong photovoltage at the Au/FLG (blue) and FLG/FeCl 3 -FLG (red) interfaces, while the photovoltage at the FeCl 3 -FLG/Au is nearly 0.
To understand the origin of the generated photovoltage, the position of the laser beam was fixed on a specific location of the interfaces, and by changing the back-gate voltage, the chemical potential was modulated from holes to electrons in the pristine FLG. Figs. 13a and b show the gate dependence of the resistance for the different interfaces found in the device as indicated in the graph. In particular, for the pristine flake, the charge neutrality point occurs at 20 V and the crossover from hole transport to electron transport can be studied.
A comparison of the gate dependence (in the range of −50 V < V g < 70 V) of the photovoltage for all interfaces shows striking differences in the measured signal depending on the interface that is measured (see Figs. 13c and d). More specifically, no detectable photovoltage was generated at the Au/FeCl 3 -FLG (black). This is in contrast to the photovoltage generated at the Au/FLG (blue) and FLG/FeCl 3 -FLG (red) which is non-zero [up to (30 mV/W], it is non-monotonous, it switches sign when the gate voltage drives the Fermi energy across the charge neutrality point, and it decreases monotonously for very high doping levels. The origin of the photovoltage was attributed to photothermoelectric effects. Furthermore, the photovoltage generated at the Au/FLG is of comparable magnitude to that measured at the FLG/FeCl 3 -FLG interface, but it has opposite sign, that is, negative in the hole side and positive in the electron side. These experiments show that FeCl 3 -FLG can be used as a replacement for metal contacts in a new generation of all-graphene-based photodetectors. . On the other hand, several methods developed for the dry transfer of graphene on arbitrary substrates [4,[44][45][46][47] can be employed to transfer FeCl 3 -FLG from glass, Si/SiO 2 or SiC onto flexible substrates that do not tolerate the conditions of the intercalation process, such as polyethylene terephthalate, polyethylene naphthalate (PEN) and polymers. In particular, Caldwell et al. [44] demonstrated the dry transfer of epitaxial graphene from SiC onto arbitrary substrates using thermal release tape. Using the thermal release tape method we have transferred FeCl 3 -FLG from a glass to a flexible PEN substrate, see Fig. 14a. To demonstrate the conduction of the transferred film the sample was tested as a component in a simple circuit. Contacts were attached to the sample using silver paint and put in series with a battery and an LED. Fig. 14b demonstrates the transferred film conducting while the substrate is deformed by flexing. Fig. 12 Photovoltage spectroscopy for the interfaces of the device highlighted in the optical micrograph picture These measurements are taken for the V g =0V,1×10 −4 V source-drain bias and with a laser power of 8.2 µW. The region highlighted in white in the photovoltage maps corresponds to the FLG/FeCl3-FLG interface and has positive sign, whereas the region highlighted in black is the Au/FLG interface and is negative. In these measurements, the Au contact connected to the FeCl3-FLG flake is grounded while the Au contact connected to the FLG is the source. Reproduced with permission in [42] Fig. 13 Gate tunable photovoltage in graphene/FeCl 3 -FLG photodetectors. Reproduced with permission in [42] a Gate dependence of the resistance for interface A b Gate dependence of the resistance for interface B c Gate dependence of the photovoltage when the laser position is located at the Au/ Graphene materials have great potential for future electronic devices with novel functionalities due to their tuneable properties [48][49][50][51][52][53], as well as for emerging flexible and wearable technologies [54] such as healthcare electronics and energy-harvesting devices. Recent advances in chemical functionalisation have shown that the properties of graphene materials can even be enhanced to unprecedented levels by chemical bonding of a molecule or a chemical element to the pristine material [55][56][57][58]. The most recent example of the potential of chemical functionalisation is FLG intercalated with FeCl 3 dubbed GraphExeter [59]. In this case, functionalisation with FeCl 3 of FLG results in the best transparent electrical conductor which outperforms indium-tin oxide used currently in displays.
In this paper, we have reviewed recent work on the successful functionalisation of FLG through FeCl 3 intercalation [10,11,33,60,61] which is confirmed though Raman spectroscopy and electrical transport measurements. Through intercalation, a significant reduction of the electrical resistivity of up to three orders of magnitude is achieved when FeCl 3 -FLG is compared to the pristine analogue. Only small changes occur in the high optical transmission spectra of FLG after the intercalation procedure, where the optical transmission is still high enough to compete with commercially used ITO films. A systematic study of the stability of FeCl 3 -FLG shows no significant changes in the electrical and structural properties for a long exposures to air, to high levels of humidity and at temperatures of up to 150°C in atmosphere. Finally, the investigation of novel graphene/FeCl 3 -FLG heterostructures demonstrated that FeCl 3 -FLG can replace metals in a new generation of all-graphene photodetectors [42,62]. The technological advantages combined with the unique electro-optical properties found in FeCl 3 -FLG make these materials a valuable alternative to ITO for photovoltaics and display applications.