Semiconductor nanowires (NWs) have received great interest for their potential applications in various future nanoscale devices 1–5. Promising device performances in NW-based solar cells 6–14, light emitting diodes 15, lasers 16–20 and field effect transistors 21 have already been reported by utilizing the unique physical properties of NWs. In addition, the small cross-sectional area of the NWs has made it possible to overcome the challenge of epitaxial growth on lattice mismatched substrates 22, 23, as well as growth of axial and radial heterostructures in a single NW 24–29 even with atomically abrupt heterointerfaces 30, 31. While promising works related to the NW-based structures and device applications are on their high, graphene, a two-dimensional material has emerged and developed in parallel 32–35. Graphene, due to its intriguing material characteristics including high electrical 35 and thermal 36, 37 conductivities, optical transparency 38, and mechanical flexibility and strength 39, is a potential material for next-generation transparent, foldable and stretchable device applications 40–43. Therefore the longtime quest for creating a wide variety of flexible devices 44–47 could be enabled through graphene. Over the past years, there has been a tremendous improvement on the synthesis of high quality graphene using different techniques, such as graphene grown on Cu 48–51, Ni 52–54, and Pt 55, 56 foils by chemical vapor deposition (CVD), and high-temperature synthesis of epitaxial graphene on SiC substrates 57–60. Moreover, low-cost roll-to-roll 61, 62 and large-area 57 production of graphene with clean substrate transferable ability 63–66 have recently been successfully demonstrated. A new class of hybrid material systems can now be developed by growing semiconductor NWs directly on graphene, replacing the expensive single-crystalline semiconductor substrates (e.g. III–V and II–VI substrates) normally used for the epitaxial growth. In these hybrid systems, graphene can simultaneously act as a flexible, transparent electrode 67 and can produce various unconventional electronic and optoelectronic devices, including sensors 68–72, piezoelectric power sources 73–75, flexible displays 76–78 and solar cells 79–81.
In this review, we discuss the challenges for the epitaxial growth of various semiconductor NWs on graphene substrates. Recent research activities on the growth of group II–VI, IV and III–V NWs on graphene, their characterization and device applications will be presented. Recent reviews on the growth of inorganic nanostructures based on other material systems on graphene can be found in Refs. 82, 83.
Semiconductor NWs on graphene layers
Heteroepitaxial growth of semiconductor NWs on various semiconductor substrates has become an interesting research topic for monolithic integration of two or more semiconductors with diverse or complementary properties. In most of the cases, the Au-assisted vapor–liquid–solid (VLS) mechanism has been utilized for the epitaxial growth of the NWs using various techniques including metal-organic vapor phase epitaxy (MOVPE) 84-89, chemical vapor deposition (CVD) 90, 91, molecular beam epitaxy (MBE) 92–94, chemical beam epitaxy (CBE) 95, 96, laser ablation 97–99 and so on. However, partly due to the concern of Au incorporation in the grown material, other alternatives such as self-catalyzed 100–106 and catalyst-free 107–111 growth techniques have been extensively studied in recent years. The understanding of the NW growth mechanisms and the systematic tuning of the growth parameters have enabled the growth of NWs with precise control of their crystal structure 106, 112–121, position 122–125, axial and radial heterostructures 27, 31, 126, as well as formation of quantum dots (QDs) within the NW 127–130. All these advanced epitaxial synthesis techniques can potentially be also adapted for NWs grown on graphene substrates.
Graphene is a single-layer of carbon atoms bonded with sp2-hybridization and arranged in a regular hexagonal honeycomb lattice structure 131, 132. Due to the absence of dangling bonds, the graphene surface is chemically inert to any foreign atoms making it challenging to grow three-dimensional semiconductor materials on its surface by conventional epitaxy. One possible method to circumvent this problem is to use van der Waals epitaxy (vdWE) 133. Due to the different bonding mechanism in vdWE compared to conventional epitaxy with strong chemical bonding, the grown materials need not to be necessarily lattice matched with the graphitic substrate and no strain induced interfacial defects are to be expected. Due to the increasing interest of using graphene as a substrate material for epitaxial growth, recent density functional calculations have been employed to investigate the adsorption sites and adsorption energies of different atoms and molecules on the graphene surface 134, 135. These theoretical calculations predict the plausibility of also conventional epitaxy i.e. non-vdWE (based on a covalent or ionic binding mechanism) for growing different materials on graphitic surfaces. Based on this, in the following, we discuss different atomic arrangements of the semiconductor atoms on graphene for a successful non-vdWE growth to take place.
The possible adsorption sites for the semiconductor atoms on graphene can be identified as 1) above the center of the hexagonal carbon rings (H-site), 2) above the bridge between the carbon atoms (B-site), and 3) above the top of a carbon atom (T-site) as indicated in the inset of Fig. 1(a). Already from the pioneering work done by Hiruma et al., it is known that semiconductor NWs mostly grow along the 111 direction in the case of a cubic semiconductor and along the  direction in the case of a hexagonal one 85, 136. Therefore, in the (111) plane of the cubic semiconductor (and (0001) for the hexagonal semiconductor), the atoms have a hexagonal symmetry as the carbon atoms in graphene. In a recent work by Munshi et al., it was shown that depending on the preferential adsorption sites and the type of semiconductor atoms, different lattice mismatch of the semiconductor with graphene will result in different atomic arrangements 137. Figure 1(a)–(d) illustrate such arrangements when the atoms are adsorbed on both H- and B-sites (Fig. 1(a), (b), (d)), and on either H- or B-sites (Fig. 1(c)). The adsorption energies for most of the semiconductor atoms are lower on the T-sites and hence excluded from the discussion here 135. The lattice constants, calculated for the four different atomic arrangements in Fig. 1(a)–(d), are plotted together with the band-gap vs. lattice constant diagram for some conventional semiconductors in Fig. 1(e). Therefore, Fig. 1(e) represents an overview of the lattice mismatch conditions to achieve non-vdWE growth in the (111) ((0001)) direction for cubic (hexagonal) semiconductors on graphene. It should be noted that the maximum lattice mismatch obtained here is within the lattice mismatch for which successful heteroepitaxial growth of semiconductor NWs has been achieved (e.g. InAs NWs on Si).
Group II–VI compound semiconductors
Among II–VI compound semiconductors, so far the epitaxial growth of ZnO NWs on graphene has been successfully demonstrated. ZnO being a wide bandgap oxide semiconductor is suitable for various device applications such as light emitting diodes 138–142, lasers 16, 143, 144, piezoelectric nanogenerators 145–147, sensors 148–151 as well as solar cells 152–154. Therefore, a hybrid structure between ZnO NWs and graphene would be interesting for designing a device where both of their unique properties can be exploited. Pioneering work has been done by Gyu-Chul Yi's group on the growth of ZnO NWs on graphene 82. They demonstrated the growth of vertically aligned ZnO NWs on few-layer-graphene by catalyst-free MOVPE using diethylzinc (DEZn) and oxygen as reactants 155. Before the NW growth, graphene sheets were transferred onto SiO2/Si substrates using a mechanical exfoliation technique. The NWs grew vertically oriented on the graphene layers, whereas they grew tilted with random orientation on the SiO2/Si surface as can be seen in the scanning electron microscopy (SEM) image in Fig. 2(a). In addition to the NWs, nanowalls were also formed at the step edges due to the increased nucleation and crystal growth process. The morphology and density of the NWs were also found to be dependent on the growth temperature. With increasing growth temperature, the nanostructures became longer and needle-like because of the increased surface diffusion of the adatoms, and their density reduced due to a reduction in the nucleation rate.
Recently, the growth of ZnO NWs on graphene layers was also successfully demonstrated using a hydrothermal process, as depicted in Fig. 2(b)156. The growth of ZnO NWs took place on the graphene covered surface, and no growth on SiO2 surface. The hydrothermal process being a low-temperature synthesis technique may thus be useful for fabricating NWs on graphene deposited on plastic substrates. Later, Kumar et al. used catalyst-assisted CVD to grow ZnO NWs and nanowalls on graphene layers 157. They showed that by controlling the Au layer thickness (used as the catalyst) the morphology of the ZnO nanostructure could be changed from NWs to nanowalls, through an intermediate stage where the NWs and nanowalls coexist and are interconnected. Figure 3(a)–(c) depict the growth of ZnO nanostructures for Au layer thicknesses of 0.5 nm, 1 nm, and 2 nm, respectively, showing the formation of NWs, intermediate structures and nanowalls. In a previous report by Ng et al., a similar dependence of the Au film thickness on the morphology of ZnO nanostructures grown on highly oriented pyrolytic graphite (HOPG) was observed 158. However, in none of these reports was the nature of the epitaxy discussed. In the case of a vdWE binding mechanism for these growths no lattice matching conditions would be required. However, using the atomic model described in Fig. 1, a non-vdWE with the H- and B-site configuration depicted in Fig. 1(a) is also possible. This atomic arrangement results in an almost coherent lattice-matching condition (mismatch ∼0.26%) for ZnO with graphene and this could be a key reason for achieving such high yield of vertical NWs on graphene.
Group IV semiconductors
The most important semiconductor in electronics as well as in solar cells today is silicon. A hybrid system between Si and graphene would open up opportunities to fabricate new types of Si-based functional devices 159, 160. Growth of Si NWs has recently been demonstrated on graphene by metal-assisted VLS mechanism using atmospheric pressure CVD 161. CVD grown graphene on Cu foil was transferred onto a SiO2 covered Si substrate using chemical exfoliation method. The Si NWs were grown on the graphene/SiO2/Si substrate using Au as a catalyst. After growth, the NW/graphene hybrid structure was transferred onto an arbitrary substrate. However, the Si NWs were randomly oriented with no preferred growth direction, indicating the absence of an epitaxial relationship with the underlying graphene layers. The lack of epitaxial relation between the Si NWs and graphene could be attributed to the transfer process. The atomic arrangement of Fig. 1(b) would result in a lattice mismatch of Si with graphene (∼3.9%) showing that non-vdWE growth is likely to be successful for Si NWs if the graphene substrate preparation and the Si NW nucleation and growth conditions are properly optimized. In fact, epitaxial growth of Si nanoislands on graphene sheet was demonstrated by Lee at al. using a vapor transport technique 162, however in this case the epitaxial nature was attributed to vdWE. Interestingly, the Si nanoislands were found to produce a small bandgap opening in graphene. Therefore future fundamental studies on how the properties of graphene are modified due to the epitaxial growth of NWs will become interesting topics for future studies and very important in order to optimize these hybrids systems for various device applications.
Group III–V compound semiconductors
III–V semiconductors are very useful for various optoelectronic applications including photovoltaics 163, as well as for high-power and high-speed electronics 164. However, the high cost of III–V semiconductor substrates and the difficulty in achieving hetero-epitaxial growth with high quality on cheaper substrates like Si still hinder their full potential in applications. Alternatives, such as epitaxial lift-off, and peel-and-stamp techniques to transfer the grown structures from their original substrates to an arbitrary target substrate have been successfully demonstrated. Such techniques would not only allow reusing the expensive III–V substrates but also create the possibility for fabricating flexible, low-cost electronic and optoelectronic devices 165–167. Another alternative would be to grow these semiconductor materials directly on a flexible and cheap substrate such as graphene, which could then also act as a transparent electrode 168 in optoelectronic device applications 169–173. Therefore, the rationale for synthesizing III–V NWs on graphene in a controllable way has to be studied extensively in order to use the III–V NW/graphene hybrid systems for various functional device applications.
GaAs being one of the most important III–V semiconductors, we first discuss studies made on the growth of GaAs NWs on graphitic substrates. Recently, the Ga self-catalyzed VLS technique was successfully utilized to synthesize the GaAs NWs on different graphitic substrates using molecular beam epitaxy (MBE) 137. In this study, prior to the growth, a Ga pre-deposition step for 20 s was applied to form liquid Ga droplets which facilitate the NW growth. The NW growth was initiated by opening the As flux. The temperature of the Ga effusion cell was pre-set to yield a nominal planar GaAs growth rate of 2 Å/s, and the valve position of the As cracker cell was adjusted to yield the required fluxes. Figure 4(a) shows such GaAs NWs grown for 10 min with an As flux 6 × 10–6 Torr at a substrate temperature 610 °C. As can be seen, the NWs are vertically aligned with the substrate and have uniform hexagonal cross-sectional shape, indicating that the growth is epitaxial. Although, NWs with good morphology were obtained, the average density of the NWs across the entire sample was low. The low density of NWs was attributed to the reduced nucleation rate at high temperature. To increase the NW density, two series of growth experiments were performed the 1st series by varying the growth temperature and the 2nd series by varying the As flux. It was found that the NW density was improved by using a low growth temperature (540 °C) with an As flux 3 × 10–6 Torr. However, due to the low growth temperature the diffusion length of Ga adatoms reduced and formed two-dimensional parasitic crystals which eventually covered the entire graphitic surface for longer growth durations. Based on this observation, a two-temperature growth strategy was adapted where the NW nucleation process was induced at a low temperature for a short duration (10 s) and the NW growth was continued (5 min) at higher temperature. Figure 4(b) represents a SEM image of the GaAs NWs grown using the two-temperature growth strategy. Figure 4(c) depicts the GaAs NWs grown with the same condition as for Fig. 4(b) but grown on few-layer epitaxial graphene synthesized on a SiC substrate. Similar results were also obtained when the growth was carried out on exfoliated graphene on SiO2/Si substrates. The results obtained here on the GaAs NW growth on different graphitic substrates ensure the robustness and reproducibility of the adapted method, and confirm the epitaxial relationship of the NWs with the substrates.
In a separate report by Tateno et al., the growth of different III–V semiconductor NWs was demonstrated on graphitic substrates using Au-assisted MOVPE 174. Here, GaP NWs were grown on both graphene/SiC and HOPG substrates, whereas GaAs and InP NWs were grown on HOPG substrate. It was observed that in addition to the vertically oriented NWs, tilted and surface grown NWs were also formed. The NW growth was also performed on Si(111) substrates for comparison. In all cases, however, the yield of vertical NWs was found to be low, indicating a need for further optimization of the growth conditions. Nucleation at step edges was argued to be a necessary condition to produce vertically aligned NWs. Earlier, Mohseni et al. employed Au-assisted MBE to grow GaAs NWs on carbon nanotube composite films 175. However, the NWs were randomly oriented and did not grow epitaxially from the carbon nanotube which seems to be due to that the NW nucleation took place at parasitic GaAs crystals and not directly on the carbon nanotubes.
Epitaxial growth of InAs NWs on graphitic substrates has also been reported recently. Hong et al. used catalyst-free MOVPE to grow the InAs NWs 176, 177. They employed two different substrate preparation methods: 1) pristine graphitic substrates and 2) O2 plasma treated graphitic substrates. Both InAs NWs and nanoislands formed on both types of substrates. It was observed that the O2 plasma treated graphitic substrates could produce a higher density of NWs and a lower density of islands when the treatment time was well optimized. At longer O2 plasma treatment times the NW density reduced and the island density increased. Figure 5(a) and (b) show such InAs NWs grown on an O2 plasma treated graphitic substrate. It can be seen that the NWs are vertically aligned and both NWs and islands have the same hexagonal in-plane orientation suggesting an epitaxial growth mechanism, which was attributed to vdWE. Later, the epitaxial growth of InAs NWs on pristine graphitic substrates was also successfully demonstrated by Munshi et al. using catalyst-free MBE as depicted in Fig. 5(c)137. In addition to the binary III–V semiconductors, the growth of a ternary InGaAs NWs on graphene layers was recently demonstrated by Mohseni et al. using catalyst-free MOVPE as shown in Fig. 5(d)178. The general observation from Fig. 5 is that the catalyst-free InAs (and InGaAs) NWs have an asymmetric NW cross-section and higher NW density as compared to the self-catalyzed GaAs NWs in Fig. 4. The asymmetric cross-section could be attributed to the catalyst-free growth mode for the InAs and InGaAs NWs, whereas the higher InAs NW density to the nearly coherent lattice matching condition with graphene (cf. Fig. 1(e)).
The epitaxial relationship of the grown NWs was further confirmed by investigating the GaAs and InAs NW/graphene interfaces using cross-sectional transmission electron microscopy (TEM) (Fig. 6) 137, 176. The graphene layer underneath the GaAs NW was found to be flat in Fig. 6(a). The GaAs NWs mainly have a cubic crystal structure (zinc blende) and grow in the 111B direction with side-facets. In contrast to the GaAs NW/graphene interface, the InAs NW/graphite interface shows the presence of 1–2 monolayer ledges of graphitic layers (Fig. 6(b, c)). On the other hand, the InAs island/graphite interface was found to be rough. In the former case, it was concluded that the 1–2 monolayer ledges (using a short O2 plasma treatment) act as preferential nucleation sites for the InAs NWs whereas the rough graphitic surface (using a long O2 plasma treatment) for the InAs islands. Although there are several challenges for growing vertically aligned III–V NWs on graphene with high yield, the above mentioned results are already very interesting and will probably encourage many new studies in order to optimize the process as well as explorations of other III–V materials.
Nature of the semiconductor/graphene epitaxial binding
As has been described above, the epitaxial nature of the grown semiconductor NWs on graphene and graphene-like substrates have been debated to be vdWE or non-vdWE in the literature. Since the absence of dangling bonds at the surface of a substrate is a pre-requisite for vdWE, the growth of NWs on graphene has mostly been ascribed to vdWE [176, 178, 179] and it is only recently that a non-vdWE involving covalent/ionic bonds has been suggested 134, 135, 137, 174, 180. These types of chemical bonds are likely to be originated from the preferential adsorption sites of the semiconductor atoms or due to the presence of defect related dangling bonds on graphene. Therefore in the case of non-vdWE, a lattice-matching condition would be preferred for the epitaxial growth. Due to the presence of these chemical bonds, the latter type of epitaxy would result in a mechanically stronger adhesion between the grown NWs and graphene. In fact, ultrasonication treatment and bending experiments performed on ZnO 180 and GaAs 137 NWs, respectively, qualitatively indicated such mechanical stability. A quantitative measure of the bonding type and bonding strength will require further detailed studies of the NW/graphene hybrid system in the future.
Nanowire material quality
The quality of the grown semiconductor nanowires has to be considered for any successful device applications. Contamination from the graphitic substrates, especially carbon, during the high temperature growth might raise a concern on the material quality. In order to address this issue, the grown NWs have been characterized. Optical characterization of ZnO NWs grown on the graphitic substrates by catalyst-free MOVPE showed no evidence of any carbon related defect peak, indicating no carbon contamination 155. Similarly, photodetectors made of single GaAs NWs grown by self-catalyzed MBE showed a high photoresponsivity comparable to that when the NWs were grown on a GaAs substrate, again suggesting a high purity material 137. These results demonstrate that high quality materials can be grown on the graphitic substrates. However, a graphene substrate if “not clean” could be a potential source of contamination of the grown materials particularly in the case of an ultra-clean system like MBE. For example, the chemical exfoliation process of CVD grown graphene usually uses polymethyl methacrylate (PMMA) for support which leaves residues on the transferred graphene and may not only degrade the quality of the materials grown on it but also could affect the intrinsic properties of graphene 181. The exfoliation of graphene without using PMMA usually results in a wrinkled rather than a smooth graphene surface, affecting the vertical alignment of the grown NWs. Therefore, a clean graphene transfer technique should be adapted in order to ensure a good NW morphology and high material quality.
Nanowire/graphene hybrid applications
The NW/graphene hybrid system has the potential to be utilized in a number of functional electronic and optoelectronic device applications. Here we outline some of the reported potential applications.
Due to the high electrical conductivity and optical transparency, graphene is an ideal candidate as an electrode material in solar cells. Using graphene as a replacement of indium tin oxide (ITO) as a transparent electrode has already been reported for organic solar cells 182, 183. In these reports graphene has been deposited onto a pre-existing structure. However if semiconductor NW solar cells are grown on graphene it can act as a substrate as well as a transparent, conducting electrode. Such idea has recently been attempted by Park et al. who have fabricated a ZnO NW/graphene hybrid solar cell 179. Figure 7(a) illustrates the scheme used for making this NW/graphene hybrid structure. Arrays of ZnO NWs were grown on graphene by a hydrothermal method after treating it with two different conducting polymers, poly(3,4-ethylenedioxythiophene)-block-poly(ethylene glycol) (PEDOT:PEG) and poly(thiophene-3-2-(2-methoxyethoxy)-ethoxy-2,5-diyl) (Plexcore OC RG-1200) to increase the NW yield. PbS QDs and poly(3-hexylthiophene) (P3HT) were used as a hole-transporting donor materials, whereas the ZnO NWs acts as an electron transporting channel to the graphene cathode. The flat-band energy diagram of the device is depicted in Fig. 7(b). Figure 7(c) and (d) illustrate the comparison of the solar cell device performances using a standard ITO electrode and a graphene electrode. Results show that the power conversion efficiencies for the graphene-based devices approach those of the ITO-based devices, demonstrating the promising application of graphene as a combined substrate and transparent electrode in a solar cell device.
Light emitting diodes
Light emitting diodes using semiconductor NWs in foldable and stretchable forms would add new functionalities in next generation displays and solid state lighting 184. One way of fabricating such devices is to grow the semiconductor NWs, although non-epitaxially, on conventional plastic substrates using low temperature techniques 185. However, the high temperature required for the epitaxial growth of the semiconductors NWs using MOVPE or MBE techniques makes plastic substrates unsuitable. Other viable route could instead be to use graphene as the substrate due to its inertness and stability with temperature. Lee et al. grew ZnO/GaN core-shell NWs on a graphene/SiO2/Si substrate using CVD 186. After making the metal top contact, SiO2 was etched by chemical etching, and the core–shell NWs with graphene were transferred onto a Cu-coated polyethylene terephthalate (Cu/PET) substrate. Figure 8 depicts the performance of the device for different degrees of bending of the Cu/PET substrate. By comparing the light emission from the device (Fig. 8(a)), electroluminescence (EL) (Fig. 8(b)), and electrical characteristics (Fig. 8(c)) for bending radii: ∞, 5.5 mm and 3.9 mm, it was observed that the device performance does not change significantly. This results show that light emitting diodes based on a NW/ graphene hybrid can be reliable and sustainable to mechanical strain and suitable for flexible display applications.
Other device applications based on the NW/graphene hybrid system have also been successfully demonstrated. Choi et al. have fabricated a fully rollable and transparent piezoelectric nanogenerator based on a ZnO NW/graphene hybrid architecture where graphene has been used both as the top and bottom electrode 187. The excellent device characteristics under different rolling and bending conditions demonstrate the stability and reliability of the hybrid device. Yi et al. have demonstrated a flexible gas sensor made of vertically aligned ZnO NWs on graphene 188. This device show stable mechanical and electrical characteristics with high sensitivity under multiple bending and releasing cycles. Therefore NW/graphene based hybrid structures are promising candidates for various rollable and stretchable devices.
Conclusions and outlook
In conclusion, the semiconductor NW/graphene hybrid system holds great potential for a future generation of flexible and stretchable electronic and optoelectronic devices. Recent advances on the growth of NWs on graphene, the device fabrication of the hybrid structure and their device performance are already intriguing. However, there are several challenges both on growth aspects and on the device fabrication which have to become understood and overcome. The inertness of the two-dimensional graphene surface makes the growth of the NWs much more difficult and challenging as compared to when they are grown on other three-dimensional semiconductor substrates, e.g. Si. Therefore, proper strategies have to be developed to modify the graphene surface which favor the nucleation and can enhance the vertical yield of the NWs with desired density. Moreover, a position controlled growth using a proper mask, like what has been achieved on various semiconductor substrates, would be desired for a better understanding of the growth mechanisms, control of the NW morphology and a better tuning of other physical parameters. Due to the rapid improvements in wafer-scale and roll-to-roll production of CVD graphene, the semiconductor device cost could soon drastically reduce if graphene is used to replace the semiconductor substrate for the epitaxial growth. Once a controlled NW growth and device fabrication have been achieved, new types of functional devices may become realized based on the NW/graphene hybrid structure.
We would like to thank our colleagues at NTNU, D. L. Dheeraj, V. T. Fauske, D. C. Kim, A. T. J. van Helvoort, and B. O. Fimland for their support. We acknowledge the financial support from NTNU and the Research Council of Norway.
Abdul Mazid Munshi received his M.Sc. degree in Physics in 2007 and M.Tech. degree in Solid State Materials in 2009, both from Indian Institute of Technology Delhi, India. He is currently working towards his Ph.D. degree in Prof. Helge Weman's group at Norwegian University of Science and Technology, Norway. Some of his areas of research interests include semiconductor nanostructures, graphene, epitaxial growth, structural and optical characterizations, and photovoltaics.
Helge Weman is a full professor in nanoelectronics in the Department of Electronics and Telecommunications at the Norwegian University of Science and Technology (NTNU) in Trondheim, Norway. He received his Ph.D. in Semiconductor Physics in 1988 from Linköping University, Sweden. During his career he has since then held various positions at UCSB (USA), NTT Optoelectronics Lab (Japan), EPFL (Switzerland), and IBM Res. Lab (Switzerland). Since 2005, Weman is leading a research group at NTNU on III–V semiconductor nanowires and graphene for use in solar cells and various photonic applications. During 2010–2013 he is the director of a Nordic consortium (NANORDSUN) on nanowire based solar cells. Weman has authored more than 100 refereed journal papers, some 200+ conference papers and is the inventor of six patent applications. Since 2010 he is a member of the Norwegian Academy of Technological Sciences (NTVA). In June 2012 he co-founded the company CrayoNano AS, where he is currently the Chief Technology Officer and member of the Board.