Highly selective and ultra-low power consumption metal oxide based hydrogen gas sensor employing graphene oxide as molecular sieve

The excellent gas sensing performance of metal oxide based nano- and microstructures, including a fast response time and good sensitivity, is typically limited by their low selectivity. Therefore, novel approaches and strategies are required to gain a precise control of the selectivity. Here, we introduce a nanoporous few-layer graphene oxide (GO) membrane with permeability only to speci ﬁ c gas molecules to improve the selectivity of individual zinc oxide microwires (ZnO MWs) toward hydrogen (H 2 ) gas. The fabricated GO-covered ZnO MWs showed ultra-low power consumption (60 − 200 nW) and an excellent room temperature H 2 gas sensing properties with fast response (114 s) and recovery (30 s) times, and a low detection limit of ∼ 4 ppm, while no gas response was measured to all other tested gases. As proposed, the gas sensing mechanism is based on selective sieving of H 2 gas molecules through the GO membrane and further di ﬀ usion to the Schottky contacts, resulting in a decreased barrier height. Being based on a bottom-up fabrication approach, the presented results could have great potential for further technological applications such as high-performance and highly selective ultra-low power metal oxide-based gas sensors, opening new opportunities for the design of nanosensors and their integration in wireless and portable devices.


Introduction
Metal oxide based nano-and microstructures are one of the most prominent studied systems for the detection of gases owing to their high surface area, fast response as well as good sensitivity, with the possibility to integrate them into "electronic nose" systems [1][2][3][4][5]. Typically, commercial metal oxide based gas sensors require a power supply of tens to hundreds of mW in order to generate the relatively high operating temperatures of 200−400°C using micro-machined heating elements [6], which are too high for portable applications and remote systems such as Internet of Things (IoT), lab on a chip and electronic nose devices [6,7]. Therefore, it is preferable to develop such devices based on individual metal oxide microstructures with minimized size and the ability to operate at room temperature, as reported by many authors [1,3,8,9]. However, the fabrication of high performance metal oxide nano-and microstructures able to detect different gaseous species at room temperature is strongly limited by their low selectivity [10][11][12][13][14]. Many methods have already been reported for the increase of the selectivity of metal oxide based gas sensors, including (i) doping of the semiconducting oxides with metal ions [15][16][17], (ii) surface functionalization/decoration with metal (oxide) nanostructures and polymers [16,[18][19][20], (iii) formation of heterostructures, core-shell structures or nanocomposites by combination of n-and p-type metal oxides [4], (iv) the use of permeable membranes [14,[21][22][23], (v) the use sensor array, such as "electronic nose", and pattern recognition algorithms such as principal component analysis or artificial neutral networks [24]; (vi) combination of the already mentioned methods [4,17]. Membranes are key elements in industrial-scale chemical and gas purification [14], as well as in selective gas sensing [21]. For instance, Zhang et al. successfully prepared a new zeolitic imidazolate framework-8 membrane architecture supported on vertically aligned zinc oxide (ZnO) nanorods demonstrating a considerable increase of the ideal selectivity to H 2 /CO 2 [22]. Güntner et al. demonstrated that zeolite membranes are suitable filters for gas sensors, removing undesired species from mixtures like exhaled breath, exhibiting an exceptional selectivity (> 100) for formaldehyde (down to 30 ppb) at 90% relative humidity [23]. Recently, 2D nanomaterials such as graphene and graphene oxide have also emerged as promising candidates for the fabrication of selective gas membranes due to their high mechanical strength, relative inertness and impermeability to all standard gases [14]. Koenig et al. investigated the transport of a range of gases (H 2 , CO 2 , Ar, N 2 , CH 4 and SF 6 ) through suspended porous graphene demonstrating selective molecular sieving based on effusion through Ångstrom-sized pores [14]. Jiang et al. observed that subnanometer pores in graphene sheets are superior to traditional polymer and silica membranes, where bulk solubility and diffusivity dominate the transport of gas molecules through the material [25]. Formation of pores in graphene has been performed by different methods like ultraviolet-induced oxidative etching, oxygen plasma etching [14,26], and focused ion beam (FIB) [10]. Next to the use of graphene, graphene oxide (GO) has been shown to be another excellent candidate for the fabrication of selective membranes [27]. Compared to graphene sheets, GO sheets stack non-uniformly and can be permeable to small gases without the need for additional pore formation [11]. These results demonstrate that membranes based on graphene and related 2D nanomaterials in combination with the high sensitivity of metal oxide nano-and microstructures are promising candidates to increase the selectivity of the latter.
In this work, we demonstrate that wet-chemical coating of individual ZnO microwires (MWs) with a thin (< 20 nm), nanoporous GO membrane leads to a strong reduction in permeability toward specific gas molecules, resulting in a highly selective and ultra-low power H 2 gas sensor (up to 200 nW). The gas response and its dependence on the gas concentration was studied in detail for different gases and volatile compounds, including methane, ammonia, ethanol, acetone, methanol and hydrogen. While the GO-covered ZnO MWs showed a high gas response toward H 2 with a detection limit of ∼4 ppm, no gas response was measured for all other gases. The proposed gas sensing mechanism is based on the effusion of gas molecules through nanopores (< 50 nm diameter) and the modification of the Schottky barrier at the electrical contacts, enabling a highly selective gas response. In addition to the study of GO/ZnO MWs, other devices based on electrochemically exfoliated graphene covered ZnO MWs (EG/ZnO), reduced graphene oxide covered ZnO MWs (rGO/ZnO), as well as individual EG, GO and rGO structures were fabricated and investigated in order to gain a deeper understanding of the gas sensing mechanism.

Fabrication of carbon 2D nanomaterial dispersions
Graphene oxide (GO) was prepared according to previously reported procedures [28,29]. Afterwards, the as produced GO powder was dispersed in deionized water by tip sonication to achieve a homogeneous and stable dispersion. Aqueous dispersions of electrochemically exfoliated graphene (EG) were kindly provided by Sixonia Tech GmbH and used without further chemical modification.

Fabrication of ZnO microstructures
First, highly porous macroscopic networks of interconnected, tetrapodal zinc oxide (t-ZnO) microparticles were prepared using the flame transport synthesis which is described in detail elsewhere [30][31][32]. In brief, a mixture of polyvinyl butyral and zinc powder with grain sizes of 1−5 μm (mass ratio 2:1) was prepared in a crucible and heated in a muffle furnace for 30 min at 900°C (heating rate 60°C min −1 ). The obtained loose t-ZnO powder was pressed into cylindrical tablets (diameter 6 mm, height 3 mm) and sintered at 1150°C for 5 h, resulting in a network of interconnected ZnO tetrapods.

Fabrication of macroscopic ZnO-EG/GO/rGO networks
To coat the t-ZnO network with carbon 2D nanomaterials, aqueous dispersions of GO or EG sheets (each with a concentration of 2 mg mL −1 ) were dropped on the highly porous template (94% porosity) to fill the free volume until no more dispersion was taken up. Afterwards, the infiltrated t-ZnO template was dried at 50°C for 4 h to promote evaporation of the solvent (water) and simultaneous deposition of the nanomaterials on the surface of the template. SEM images of t-ZnO templates coated with thin GO and EG layers are presented in Figs. S1 and S2, respectively. For preparation of a reduced GO (rGO)-covered t-ZnO network, the template was first coated with GO and subsequently reduced by immersion in diluted L-ascorbic acid (0.1 mg mL −1 ) for 24 h (at 50°C). Then, the template was washed thoroughly in deionized water.
For fabrication of hollow microtubes or nanosheets comprised of GO, rGO and EG sheets, the wet-chemically coated t-ZnO networks were etched in diluted HCl (1 M) for 24 h, washed thoroughly in absolute ethanol, and dried by critical point drying using an EMS 3000. The resulting networks consisted of interconnected hollow microtubes comprised of GO, rGO or EG sheets. Further details regarding the fabrication as well as mechanical and electrical characterization of the t-ZnO hybrid networks and the aero-networks can be found in another work [33,34].

Fabrication of nano devices
Nano devices based on individual structures, i.e., ZnO MWs, exfoliated graphene-covered ZnO MWs (EG/ZnO, from samples with ZnO microstructures with a density of 300 mg cm −3 and EG with a density of 1.1 mg cm −3 ), graphene oxide-covered ZnO MWs (GO/ZnO, from samples with ZnO microstructures with a density of 300 mg cm −3 and GO with a density of 2 mg cm −3 ), reduced GO-covered ZnO MWs (rGO/ ZnO, from samples with ZnO microstructures with a density of 300 mg cm −3 and rGO with a density of 2 mg cm −3 ), EG, GO and rGO microtubes/microsheets were fabricated in a FIB/SEM system using the procedure reported by Lupan et al. [3,8,9,15,35,36].

Gas sensing measurements
For gas sensing measurements, nearly practical gaseous environments were intentionally used, i.e., the measurements were performed under atmospheric pressure [3,8,9], and dry air from the background (relative humidity of 30-40%) was used as a carrier gas to dilute the necessary concentrations of gases. All measurements with devices based on individual structures were performed at room temperature (20−24°C ) under a dc bias voltage of 1 V. The gas response was defined as the ratio of the sensor resistance upon exposure to air (R air ) and upon exposure to the sample gas (R gas ) for n-type response and as R gas /R air for ptype response [37]. More details can be found in previous works [37][38][39].

Results and discussion
The fabrication of the ZnO MWs covered with thin membranes of GO, rGO and EG is based on a simple wet-chemical coating technique which has recently been published and discussed in detail [33]. In brief, a highly porous (∼94%) 3D network of interconnected ZnO MWs was infiltrated with an aqueous dispersion of the carbon 2D nanomaterials, resulting in filling of the entire free volume of the network. During evaporation of the solvent, the 2D flakes assembled on the surface of the ZnO MWs to form a uniform nanoscopic layer. Figs. S1 and S2 show scanning electron microscopy (SEM) images of the ZnO MW networks coated with a thin GO and EG membrane, respectively. By variation of the number of infiltrations or the concentration of the nanomaterial dispersion, the coverage could be easily controlled from less than a  Fig. S1. Strikingly, the SEM images of GO-covered ZnO MWs (Figs. 1e and S3a-e) reveal a high density of wrinkles within the assembled GO layer. Wrinkling is known to be a ubiquitous phenomenon in large 2D membranes [40] and has already been reported for other graphene-based structures. For instance, Grosse et al. observed a small temperature rise at wrinkles in CVD grown hexagonal graphene grains using scanning Joule expansion microscopy [41], which is very important for gas sensing applications. Zhang et al. also observed a reversible wrinkling and unwrinkling of graphene sheets on the Pt (111) surface upon cycled heating and cooling treatments [42]. The high concentration of wrinkles can also be observed on the surface of rGO-covered ZnO MWs (Fig. S4). While the diameter of both the pristine and GO-covered ZnO MWs ( Fig. 1a and b) was 0.57 μm at one end and 1.2 μm at the other end, the diameter of the rGO-covered ZnO MW was 1.15 μm. Thus, all individual structures had practically the same diameter allowing the comparison of their gas sensing properties without considering the geometrical parameters. This is very important due to the high dependence of the gas response on the diameter in the case of individual structures, which was demonstrated experimentally and theoretically [39,43,44]. The structure of the fabricated devices based on pristine and GO-covered ZnO MWs is illustrated in Fig. 1c and f, respectively.
The current-voltage (I-V) characteristics of devices based on individual ZnO, EG/ZnO, GO/ZnO and rGO/ZnO MWs (range of -5 to +5 V) are presented in Fig. 2a.
The graph directly demonstrates that the coating of carbon nanomaterials on the ZnO MWs has a strong influence on the I-V characteristics of the individual devices. The pristine ZnO MW had the highest resistance of all the fabricated devices, resulting in a rather low current at the same applied bias. By covering the ZnO MW with graphene-based nanomaterials (EG, GO, rGO), the resistance decreased, probably by adding their lower resistance to the higher resistance of the ZnO MW in parallel. The EG/ZnO MW showed the lowest resistance of all fabricated devices, which can be associated with the high conductivity of the graphene sheets [45,46]. The rGO-covered ZnO MW exhibited a slightly higher resistance compared to EG/ZnO, while the addition of GO to the ZnO MW decreased the resistivity only marginally. The latter can be attributed to the fact, that GO is insulating by nature [47], thereby only slightly contributing to the conductivity of the device. The room temperature gas sensing properties of all devices to different types of reducing gases and vapors with a concentration of 1000 ppm are presented in Fig. 2b. The error bars represent the deviation in the gas response R air /R gas (R air : sensor resistance in air, R gas : sensor resistance in sample gas) after several measurements at the same concentration of the sample gas. In the case of ZnO MWs and GO/ZnO MWs, an n-type gas response was observed, i.e., the decrease in electrical resistance upon exposure to reducing gases, which is typical for ZnO nano-or microwires [48,49]. This indicates that the gas response can be mainly attributed to the ZnO MW. The gas response of the pristine ZnO MW to 1000 ppm of H 2 , ethanol, CH 4 , NH 3 , methanol and acetone was 7.6, 4.4, 1.6, 2.9, 4 and 5.1, respectively, showing no selectivity toward any of the gases. In contrast to that, the addition of a GO membrane resulted in a drastic increase in selectivity toward H 2 gas (gas response 3.4), while there was no gas response to all other investigated gases. Compared to the pristine ZnO MW, however, the gas response of GO/ZnO was by a factor of 2.2 lower. The measurements with the EG/ZnO and rGO/ZnO MWs revealed a p-type response, i.e., the increase in electrical resistance upon exposure to reducing gases, which is typical for carbon based nanomaterials [46,50,51]. Therefore, the gas response can be mainly ascribed to the sorption and gas sensing properties of the EG and rGO membrane. For the rGO/ZnO MW, the gas response to 1000 ppm of H 2 , ethanol, CH 4 , NH 3 , methanol and acetone was 1.77, 1.26, 1, 2.3, 1.06 and 1, respectively, showing no selectivity. For the EG/ZnO MW only a low response to NH 3 (∼1.1) was observed, which is in accordance with the literature, but not of importance in the scope of this study. It can be assumed that the lower resistance and poor gas sensing properties of the EG membrane (compared to ZnO) lead to a considerably lower gas response, i.e., the change in resistance of the ZnO MW is shunted by a lower parallel resistance of the EG layer [52]. Since the GO/ZnO system represents the highest interest due to its high selectivity, only the gas sensing properties of ZnO (for comparison) and GO/ZnO MWs are investigated in detail in the following paragraphs. Fig. 2c shows the dependence of the room temperature gas response on the concentration of H 2 gas for pristine and GO-covered ZnO MWs. A power law relationship of gas response (S) and H 2 gas concentration (p H2 ) can be observed ( 2 ), where β is the slope of log S versus log p H2 [53]. The obtained β values for pristine and GO-covered ZnO MWs are 0.3 and 0.25, respectively. These values depend on the charge of the surface-adsorbed species and the type of chemical reaction on the surface [54,55]. Since it is nearly the same for both investigated systems, the same type of surface chemical reaction between chemisorbed oxygen species and the H 2 gas molecules can be expected [54,55]. The detection limits were measured using the signal/noise ratio 2 , as reported by Dua et al. [50], and yielded values of ∼2 ppm for the pristine  ZnO MW and ∼4 ppm for the GO/ZnO MW. The calculated power consumption of pristine and GO-covered ZnO MWs at 1 V applied bias voltage versus the concentration of H 2 gas is also presented in Fig. 2c. In a passive state, i.e., when no gas is applied, the power consumption of pristine and GO-covered ZnO MWs was ∼4 nW and ∼60 nW, respectively. By increasing the H 2 gas concentration up to 1000 ppm, the power consumption increased up to ∼28 nW and ∼200 nW, respectively. This demonstrates that the fabricated devices can operate at extremely low power levels in the nW region, even at application of relatively high gas concentrations of 1000 ppm (0.1%). The dynamic gas response of individual pristine and GO-covered ZnO MWs to different concentrations of H 2 at room temperature is presented in Fig. 2e and f (dynamic resistance measurement in Figs. S5a and S5b), respectively, while the calculated response and recovery times are shown in Fig. 2d Fig. 2c and f). In the case of the pristine ZnO MW, the signal recovered completely after evacuating the H 2 gas from the test chamber (Fig. 2e), while it did not fully recover for GO/ZnO (Fig. 2f) (at least not in the investigated time intervals). Furthermore, the response and recovery times for the pristine ZnO MW decreased from 200 s to 43 s and from 68 s to 31 s, respectively, by increasing the concentration of H 2 gas from 60 to 2000 ppm (Fig. 2d). In contrast, it can be seen that the response and recovery times of the GO-covered ZnO MW to H 2 were significantly reduced (Fig. 2d), i.e., from 211 s to 114 s (response) and from 106 s to 30 s (recovery) upon increasing the H 2 concentration from 125 ppm to 1000 ppm. Thus, the GO membrane did not only strongly enhance the selectivity of the ZnO MW, it also improved the response and recovery times. The dynamic gas response to 1000 ppm of different gases and vapors (hydrogen, ethanol, methane, ammonia, methanol and acetone) of pristine and GO-covered ZnO MWs is presented in Fig. 3a and b (dynamic resistance measurement in Fig. S6a and b), respectively.
The graphs confirm the high selectivity of the GO-covered ZnO MW toward H 2 gas compared to the pristine ZnO MW, showing no gas response to all tested gases except H 2 with a significant increase in signal. In order to investigate the stability of the GO/ZnO sensors, the gas response of the device was also measured under cyclic H 2 gas exposure  with the same concentration (1000 ppm, Fig. 3c and Fig. S6c). The graph demonstrates the repeatability of the gas response with a residual standard deviation smaller than 5%. The dynamic response of the individual rGO/ZnO MW to 1000 ppm of H 2 , ethanol, NH 3 and methanol is presented in Fig. S7. In order to exclude the potential influence of the GO membrane on the gas response of GO-covered ZnO MWs, gas sensing measurements were also performed on individual GO membranes. These were prepared by wet-chemical etching of the GO/ZnO MWs, resulting in nanoscopic thin GO microsheets (inset in Fig. 3d). The gas response of the GO membrane was measured toward 1000 ppm of different gases at room temperature. As revealed from the results in Fig. 3d, the membrane did not show any response to all of the tested gases. This confirms that the GO membrane itself does not contribute to the gas response of the GO/ZnO system. The gas response and current-voltage characteristics of devices based on individual EG and rGO membranes are presented in Fig. S8, showing minor gas responses to NH 3 typical for carbon-based nanomaterials with enhanced conductivity [56].
The gas sensing mechanism of individual ZnO structures has already been proposed and described in detail [15]. In this case, the surfacedependent electrical properties of the ZnO MWs are dominated by the formed Schottky contacts at the Pt/ZnO interface [49,52,57]. As confirmed by the current-voltage characteristics of the sensing device based on the ZnO MW (Fig. 2a), the higher work function of Pt (ϕ M = 6.1 eV) compared to the electron affinity of ZnO (χ = 4.5 eV) [58] leads to the formation of double Schottky contacts. Owing to the strong catalytic effect of noble metals like Pt or Pd, the formation of Schottky contacts with ZnO are known to contribute significantly to the H 2 gas response [43,52,57,59,60]. This can explain the higher gas response to H 2 gas of device based on individual ZnO MW compared to other tested gases, observed in our case (see Fig. 2b) and reported previously [3,8,9]. Similarly, this effect has even been reported for sensing devices based on carbon nanotubes and silicon, which are usually not sensitive to H 2 gas at room temperature [43,57,61,62]. This can explain the observed higher gas response of the ZnO MWs toward H 2 gas compared to the other sample gases, which has already been reported in previous experimental studies [3,8,9]. The high selectivity of the GO-covered ZnO MWs, however, has not been investigated yet and may be explained as follows. A perfect monolayer graphene sheet is impermeable to gases as small as He [63]. Therefore, to use graphene or GO as a membrane material, it is necessary to form nanosized pores in the GO layer in order to achieve a certain gas permeability [25]. In this work, the formation of these pores might be related to the fabrication of the GO membrane on the ZnO MW. Since the GO membrane is produced by wet-chemical assembly of individual GO sheets on the surface of a ZnO MW [33], the assembled layer constitutes nanosized pores between the interconnections of the GO sheets which originate from partial overlapping and non-uniform assembly. Additionally, during the deposition of Pt for the formation of electrical contacts between the GO/ZnO MW and Au pads in the FIB/SEM system, nanopores could be formed within the GO membrane near the Pt contacts [10]. For example, Celebi et al. used Ga-based FIB to perforate apertures between 14 nm and 1 mm in diameter and He-based FIB for < 10 nm-pore drilling [10]. In addition to the pore formation within the GO membrane, the irregular stacking in few-layer GO causes an interplanar spacing larger than that of graphitic layers (0.335 nm) [11]. As a result, both the nanopores and increased interplanar spacing in the GO membrane may play a central role for size-selective molecular sieving of the tested gas molecules. In particular, smaller molecules with weaker interaction with GO (i.e., H 2 ) diffuse faster than larger molecules (i.e., all other tested gas molecules, see molecular diameters in Table S1), giving rise to the observed response of the GO-covered ZnO MW only to H 2 gas among all tested gases [13,63]. As schematically shown in Fig. 4a, the diffusion of the H 2 molecules can occur through the nanopores and the interlayer region of the GO sheets [13].
The interlayer diffusion could be the rate-limiting process compared to the cross diffusion through nanopores in GO layers, because the lateral size of the 2D diffusion channel is the order of micrometers, while the interlayer distance is the order of sub-nanometers [13,64]. In the case of nanopores (< 50 nm), the transport of molecules occurs via effusion, because the mean free path (λ) becomes larger than the nanopore diameter (d) [10]. The effusion mechanism is known to be where n is the gas number density, u is the mean molecular speed, P is the pressure, k B is the Boltzmann constant, T is the temperature, and m is the molecular weight. Consequently, only H 2 molecules will diffuse at the interface of the Pt electrode and the ZnO MW, resulting in a decreasing height of the Schottky barrier (Δϕ SB = ϕ SB -Δ′) due to the formation of a dipole layer at the Pt/ZnO interface (Fig. 4) [60]. In this case, ϕ SB is determined by [48,49]: where E C and E F are the minimum energy of the conduction band and the Fermi level energy, respectively, and V bi is the built-in potential at the Pt/ZnO interface. The width of the depletion layer (L SB ) is given by [49,52]: where ε s is the permittivity of ZnO, N D is the donor impurity density in the ZnO MW, q is the unit electronic charge, V is the applied voltage, and T is the absolute temperature. Therefore, the current of the device is governed mainly by V bi and L SB . The change in current of the device related to the variation in ϕ SB due to the formation of a dipole layer can be expressed as [18,49,52]: where A** is the Richardson constant, and I air and I gas are the currents under exposure to air and gas, respectively. The influence of water vapors on the gas response is an important factor. As demonstrated by Nair et al. [12], submicrometer-thick graphene oxide membranes allow unimpeded permeation of water. Therefore, it is important to investigate the influence of water vapors on the gas sensing properties of GO-covered ZnO MWs. Fig. S9 shows the gas response of pristine and GO-covered ZnO MWs to 50 and 1000 ppm of H 2 gas at 30% and 70% relative humidity, revealing a decreasing gas response for both types of structures with increasing relative humidity. This is typical for metal oxide structures and might arise from adsorption of water molecules on the surface of the ZnO MW (hydroxyl poisoning), which leads to lowering in gas response.

Conclusions
In summary, we have demonstrated a simple concept for the fabrication of a highly selective and ultra-low power (60−200 nW) H 2 gas sensor, which combines the excellent sensitivity of metal oxide microstructures with the selectivity of a few-layer GO membrane based on molecular sieving. The devices were prepared by a scalable wet-chemical synthesis strategy in which a self-assembled layer of 2D nanomaterial sheets, such as GO, formed a homogeneous coating on the metal oxide surface. We propose that the induced nanopores in the GO membrane act as a size-selective sieve, which only allows permeation of hydrogen molecules among the tested gases, while other gases (e.g., ethanol, methane, ammonia, acetone and methanol) cannot pass the membrane. The obtained experimental results support this hypothesis and are consistent with theoretical models developed in the literature based on effusion through nanometer-sized (< 50 nm) pores. The gas sensing mechanism is explained based on the modulation of the Schottky barrier height due to dissociation of H 2 gas molecules at the Pt/ZnO interface and subsequent diffusion, leading to the formation of a dipole layer. The obtained results demonstrate the great potential to use GO with engineered nanopores for the development of highly selective hydrogen gas sensors. Combined with the small size, light weight and possibility to integrate such MWs with a high density, such devices can become the essential building blocks for mobile breath analysis systems, air quality monitoring systems, and sensor networks technology for nanosensing applications.

Declaration of competing interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to Fabian Schütt is currently working as a research group leader at the Chair of Functional Nanomaterials at the Institute for Materials Science, Faculty of Engineering, Kiel University, Germany. He has a strong background in material science and engineering. He finished his PhD in 2018 (highest honors). During his PhD and in the scope of the "Graphene Flagship" he developed new methods for the synthesis of highly porous and lightweight foam structures from 1D and 2D nanomaterials, such as graphene and carbon nanotubes. His PhD thesis was awarded by the priority research area "Kiel Nano, Surface and Interface Science" of Kiel University. He has a broad experience in applied-Graphene research and is involved in several international research collaborations and projects. E-mail: fas@tf.uni-kiel.de Yogendra Kumar Mishra is Professor MSO at Mads Clausen Institute, NanoSYD, University of Southern Denmark (SDU), Denmark. Prior joining to SDU, he was leading a scientifically independent group at Functional Nanomaterials Chair, Institute for Material Science, Kiel University, Germany. He did Habilitation in Materials Science from Kiel University in 2015 and Ph.D. in Physics in 2008 from Jawaharlal Nehru University, New Delhi, India. In Kiel, he introduced a new flame-based process for versatile nanostructuring of metal oxides and their 3D interconnected networks. These zinc oxide tetrapods found many applications in engineering and biomedical fields and additionally, they can be used as sacrificial templates to create hybrid and new materials in tetrapodal forms. At NanoSYD, the main research focus of his research group is in the direction of 'Smart 3D Materials for Advanced Technologies'. E-mail: mishra@mci.sdu.dk