Laser-induced chemical transformation of free-standing graphene oxide membranes in liquid and gas ammonia environments

Abstract

Laser-induced chemical conversion of graphene oxide (GO) is an effective way to modify its properties and expand its potential use for numerous applications. In this work, a mechanically stable and flexible free-standing GO membrane is synthesized and further processed by ultraviolet laser radiation in gas and liquid ammonia-rich environments. Electron and atomic force microscopy, as well as X-ray photoelectron spectroscopy analysis, reveal that laser irradiation in gas leads to a large defect-induced morphology modification and high deoxygenation process, accompanied by the slight incorporation of nitrogen functionality to the reduced GO structure. Conversely, irradiation in liquid provokes significant integration of nitrogen groups, essentially amines, into a partially reduced GO structure, without evident modification of the morphology. Electrical measurements on the macro- and nano-scale point to a complex contribution of morphology and oxidized regions to the overall resistance of the rGO.

---

Introduction

Nowadays, the challenge regarding the fabrication of graphene-based materials is closely related to their industrial applications, involving preservation of the outstanding performances obtained in research laboratories through mass production.¹ Following this target, low-cost, eco-friendly, and more scalable protocols for the large quantity synthesis of high quality multifunctional graphene-based materials, and devices based on them, are still in demand. Recently, graphene oxide (GO) nanostructures have attracted great interest due to their exceptional physicochemical properties for many applications.^2–6 Besides, GO sheets are able to be converted into a graphene-like material, named reduced graphene oxide (rGO), by relatively easy, scalable, and cost-effective synthesis methods. GO has therefore arisen as a key element in obtaining graphene-like materials at an industrial scale.

GO exhibits very interesting functional properties: it is dispersible in water, is biocompatible, and its electrical conductivity and optical band gap can be tailored by just modifying its oxidation degree. Accordingly, GO is an electrical insulator whereas rGO is more conductive, depending on the deoxygenation level, and can act as a p-type semiconductor.^7–9 It has been recently reported that GO and partially reduced GO have tremendous potential to be used in varied electrochemical applications, such as photocatalysts, due to the complex interplay of the incorporated functional groups and graphitic regions with different external molecules through covalent and non-covalent interactions.^3,10,11 Furthermore, the functionalization of rGO with nitrogen species confers additional n-type nature, leading to the formation of a semiconductor with both p- and n-type characteristics.^12,13 The resulting synergic combination of mixed p–n heterojunctions in the same material, and the benefits of the relatively easy synthesis of N-doped graphene-like films from GO powder, significantly extend the room for the development of new efficient rGO-based devices, such as nonvolatile memory devices, transparent electrodes, supercapacitors and photocatalysts.^14–17

The structural modification of GO for the synthesis of rGO and N-doped rGO can be carried out by chemical and thermal treatments.¹⁸ However, these methods need either toxic reagents or high temperature processes, which hold up the large scale development of GO and rGO-based devices. Conversely, laser processing has arisen as a very promising tool for local, harmless and large-scale modification of the GO structure.¹⁹ For instance, laser direct writing has been demonstrated to be a fast and versatile method for the development of patterned rGO-based materials and devices composed of them.^20,21 Besides, thin films of functional rGO-based hybrid composites can be also processed through this technique onto almost any type of substrate.^22,23 Furthermore, laser irradiation can generate out-of-thermodynamic equilibrium photochemical and photothermal mechanisms,²⁴ leading to the generation of new different chemical pathways compared to conventional methods. Therefore, laser processing methods are certainly opening up very promising opportunities for the scalable fabrication of rGO-based devices.

In this work, a simple method is developed for the fabrication of mechanically stable, flexible, and free-standing membranes composed of GO. This method allows films to be obtained with areas up to hundreds of cm². After that, the membranes are treated with nanosecond pulsed ultraviolet laser radiation in gaseous and liquid ammonia-rich conditions. The structural and compositional characterization of the obtained materials reveals significant differences regarding the morphology and chemical composition of samples fabricated under analogous laser conditions but in different environments. These dissimilarities lead to distinct macroscopic and microscopic electrical properties.

Experimental

GO films of about 20 μm in thickness were prepared by a synthesis pathway consisting of (i) a three-step graphite oxidation-exfoliation procedure using the main precursors of Marcano-Tour’s improved method,²⁵ (ii) obtaining the GO solid-precursor by separation in a water–ethanol GO suspension, and (iii) drop-casting of the prepared aqueous GO suspension on a glass wafer, followed by drying and peeling of the film. In the first step of the graphite oxidation-exfoliation method, 270 mL of H₂SO₄ (95–97% purity), 30 mL of H₃PO₄ (85% purity) and 2.7 g of graphite (purum powder, ≤0.1 mm, Fulka) were mixed under stirring in an 800 mL glass jar placed in ice bath. After about 20 min, 12 g of KMnO₄ (99% purity, Merck) was gradually added, maintaining the stirring conditions at about 0 °C for 2 h. Afterwards, the mixture was held without stirring at room temperature for 4 days. In the second step, the mixture was stirred again in an ice bath and after 20 min, 200 mL of H₂O₂ (3% vol.) was gradually added (over a period of about 20 min). After 1 h, the obtained mixture was centrifuged (5000 rpm for 15 min), and the supernatant was decanted away. The remaining solid material was then washed successively with 200 mL of H₂O (bidistilled), 100 mL of HCl (37% vol.) and 100 mL of absolute EtOH. The last two washing steps were repeated twice. After each washing step, the suspension was dispersed by magnetically stirring (5 min), sonicated (15 min) and centrifuged (5000 rpm for 15 min) and the supernatant was decanted away. In the third step, the resulting solid was dispersed and sonicated (15 min) again in 200 mL of a 50% vol. aqueous ethanol solution. The resulting GO suspension was held in a closed jar for 7 days in order to separate by sedimentation: the expanded graphite remaining after oxidation as well as the macroscopic precipitated graphite oxide. It is known that GO sheets form very stable dispersions in water because of their strong hydrophilic character and sediments in pure ethanol.²⁶ Thus, on the top side of the jar, only a water–ethanol suspension with floating GO nanosheets was formed. Afterwards, about 200 mL of this initial GO suspension was harvested from the toppart of the jar and dried in an open atmosphere for 3 days. A dark-brown solid GO was obtained. Then, a 2.5 mg mL⁻¹ GO suspension was obtained by dispersing solid GO in bidistilled water under magnetic stirring. This suspension was then poured onto a smooth glass wafer and dried for 3 days in ambient conditions. The resulting GO film was peeled off with the aid of a wide blade. This method allows free-standing, mechanically stable and flexible GO membranes of hundreds of cm² to be obtained in a straightforward way.

GO membranes were cut into pieces of few cm² (Fig. 1a) and submitted to laser treatment at room temperature. Irradiation experiments were made using a 266 nm wavelength Quantel Brilliant Nd:YAG laser system, emitting pulses of 3 ns duration and at 10 Hz repetition rate. Irradiations were performed by means of a squared laser spot, 1 × 1 mm² in size, and applying subsequent laser pulses in the range of 1–1000. The incident laser fluence values were fixed at 50 and 100 mJ cm⁻². Patterns of several mm² dimensions were obtained by irradiation of adjacent spots with 50% overlap (an example of a pattern is shown in the inset of Fig. 1a). Two sets of experiments were performed in ammonia (NH₃)-rich reactive environments. In the first set of experiments, GO membranes were irradiated inside a reaction chamber previously evacuated down to a residual pressure of 10⁻⁴ Pa. The laser treatments were performed in gaseous conditions by flowing N₂ gas through a NH₃–H₂O liquid solution (30% v/v) and introducing the gas mixture into the reaction chamber at nearly atmospheric pressure. In the second set of experiments, GO membranes were directly immersed in a NH₃–H₂O liquid solution (30% v/v), inside a quartz vessel. The laser treatments were performed using the same irradiation conditions in gas and liquid conditions. For convenient comparisons, the raw GO membrane and the most significantly laser treated samples were coded and are summarized in Table 1.

Fig. 1 (a) Image of a GO membrane piece. Inset: interdigital pattern processed on GO membrane in an ammonia-rich gaseous environment. SEM images of (b) raw GO membrane, (c) G100 and (d) L100 samples.

Table 1 Sample identification as a function of experimental conditions

Code	Sample
Ref	As synthesized GO film
G50	GO irradiated in gaseous NH3-rich atmosphere. 50 mJ cm^−2, 500 pulses per zone
G100	GO irradiated in gaseous NH3-rich atmosphere. 100 mJ cm^−2, 500 pulses per zone
LRef	GO membrane exposed to liquid NH3-rich ambient for ca. 1 h (not irradiated)
L100	GO irradiated in liquid NH3-rich ambient. 100 mJ cm^−2, 500 pulses per zone

---

The morphologies of the as-synthesized and irradiated layers were studied by field emission scanning electron microscopy (SEM) with a FEI QUANTA 200 FEG-ESEM equipment, and atomic force microscopy (AFM) through an Agilent 5100 system, working at intermittent contact mode. X-ray photoelectron spectroscopy (XPS) was used for the surface compositional study of the obtained materials. The measurements were done in a 10⁻⁷ Pa ultra-high vacuum environment by means of a SPECS XPS spectrometer based on a Phoibos 150 electron energy analyzer. The system operated in constant energy mode. A monochromatic Al Kα line (1486.61 eV) X-ray source was used for excitation. High resolution spectra were recorded over a 20 eV range with a 10 eV pass energy and 0.7 eV energy resolution. The electrical nature of the processed layers was studied by means of van der Pauw resistance measurements in 5 × 5 mm² samples using a Keithley 2612A source-meter system. Furthermore, resistance maps at the microscale, as well as current–voltage measurements at nanometric locations, were acquired with an Agilent 5500LS scanning probe microscope equipped with a Resiscope II module (CSI). The measurements were obtained using platinum-coated silicon tips (RMN-25-PT300, Rocky Mountain Nanotechnology) and in less than 10% relative humidity to avoid undesired electrochemical reactions. The bias voltage is applied to the sample surface, where silver paste contacts are connected. The AFM and resistance maps data were processed with Mountains 7.2 software from Digital Surf. Finally, approximate numerical calculations were made in order to get an understanding of the thermophysical mechanisms appearing during the irradiation of the GO material. For the sake of simplicity, photochemical mechanisms were not considered and only photothermal processes were simulated in 2D models by means of COMSOL 5.2 multiphysics software. A description of the model can be found in ref. 22 and 27. The time evolution of the temperature of 20 μm thick GO membranes in contact with air (gas) and water (liquid) during irradiation with one laser pulse at 50 and 100 mJ cm⁻² was assessed. The simulation included the coupling of heat transfer and turbulent fluid dynamics models. Graphene oxide thermal and optical parameters were identical to those used in ref. 22, whereas the thermophysical and mechanical-dynamical properties of the fluids were provided by the COMSOL materials library.

Results and discussion

Raw GO membranes show a dark brown-grey color and smooth surfaces under visual inspection. Nevertheless, SEM inspection reveals an irregular topography with hundreds of nm-sized protrusions and ripples, caused by bending of the GO sheets during their stacking-drying process (Fig. 1b). The laser irradiation of GO membranes in a gaseous ammonia-rich environment leads to a clear modification of the surface morphology after the accumulation of 100 laser pulses. In Fig. 1c, mountain-like structures of a few microns in size can be appreciated. These features show a round aspect, pointing to melting and merging mechanisms of the irradiated material being the mean cause of their formation. Furthermore, high resolution AFM images show that most of the mountain-shaped structures obtained with 100 mJ cm⁻² laser fluence present terrace- and small filament-like features around 50–100 nm in height-width (Fig. 2a and b). These features tend to align, forming parallel bundles, and can reach areas of over a few microns. The extent and size of the bundles increase with the number of accumulated pulses. However, they are less developed at a laser fluence of 50 mJ cm⁻². Besides, no filamentous features are observed in samples obtained with this laser fluence. Laser irradiation of a GO membrane immersed in a liquid ammonia solution leads to a small modification of the film color, being somewhat pale grey. However, SEM and AFM studies reveal that even after the accumulation of 500 laser pulses per zone at 100 mJ cm⁻² fluence, no significant alteration of the surface morphology is provoked (Fig. 1d and 2c). Thus, samples irradiated in liquid media are similar to a raw GO membrane.

Fig. 2 AFM topographic maps of (a and b) G100, and (c) L100 samples. Inset in (b) shows the topographic profile along the dashed line.

According to our photothermal numerical simulations, the absorbed laser energy of each pulse induces fast thermal cycles up to about 1 μs in duration in the GO material (Fig. 3). Heating times are very short, in the ns range, with temperature increase rates reaching up to 10¹² K s⁻¹, while cooling times are orders of magnitude longer. As can be observed, an increase of the applied laser fluence leads to augmentation of the maximum temperature, as well as the thermal cycle duration. Moreover, the high temperature region is confined to just a few hundreds of nanometers in depth (inset of Fig. 3). Interestingly, and in accordance with the experimental results, the effect of laser radiation on temperature evolution is larger in samples irradiated in a gaseous environment than in those irradiated in liquids. Thus, while the maximum temperature reaches up to 3000 K in GO irradiated in gas at 100 mJ cm⁻², it only reaches 1400 K when irradiated in similar laser conditions, but in a liquid medium. This difference in thermal response is due to the larger thermal energy transferred from the heated GO membrane to the liquid surrounding medium, compared to the gaseous one. It should be mentioned that, according to our numerical simulations, the velocity magnitude of the liquid in contact with the GO membrane is negligible up to several microseconds after the start of the heating process. Thus, convection movements of the liquid would not significantly influence the GO thermal response.

Fig. 3 Simulated temperature evolution with time caused by one laser pulse. Square and circle symbols refer to irradiation in gaseous and liquid environments, respectively. Open and solid symbols respectively represent irradiation with 50 and 100 mJ cm⁻² laser fluence. Laser pulse intensity time evolution, in arbitrary units, is also plotted for reference. Inset: temperature in GO membrane and velocity magnitude of liquid water distributions at 6 ns, irradiated with a 100 mJ cm⁻² laser pulse.

Ultraviolet radiation pulses are expected to cause photochemical transformation of the structures of materials, as well as rapid temperature variations. These mechanisms, with their respective roles, would be at the origin of the observed morphology evolution in samples irradiated in gaseous conditions. Our numerical calculations show that the irradiated material does not reach its melting temperature (if considered to be about 4800 K, the same as the graphene backbone) under single pulse irradiation conditions. Indeed, SEM micrographs of the irradiated samples do not show considerable morphological changes till 100 accumulated pulses. Conversely, at this number of pulses, molten-like structures can be distinguished (Fig. 1c). Filament-like structures point to the laser-induced creation of structural defects in the GO material, which provoke a steady stress-induced bending of the GO sheets.^28–30 The observed terrace-like features would denote the edges of the stacked GO sheets. Upon the accumulation of pulses, the proliferation and propagation of crystalline defects would take place. Then, the sites with a sufficient density of crystalline defects, exhibiting chemical bonds with a lower dissociation energy, would behave as nucleation sites for local premelting and even vaporization processes at temperatures well below the melting point of graphene.³¹ Therefore, the accumulation of laser pulses would promote concomitant melting to some depth of extended regions of the graphene-based material, as observed in samples processed with a high number of accumulated pulses below the melting laser fluence threshold. This mechanism, which is common in the laser processing of many types of materials, is the so-called “incubation effect”.^24,27,32 Conversely, as mentioned, GO membranes irradiated in a liquid environment do not exhibit any apparent morphological modification. The causes could be directly linked to the nature and amount of photochemically-induced structural defects, as will be presented hereafter, and by the lower temperature development in the material, as revealed by our numerical simulations.

The observed laser-induced morphology evolution is closely related to the compositional modification of the GO material. X-ray photoelectron spectroscopy is an excellent tool for analyzing the chemical structure evolution at the GO surface. The atomic concentration of C, O, and N elements, calculated from the wide scan XPS spectra, are shown for the raw GO membrane (Ref), samples irradiated in gaseous (G50, G100) and liquid (L100) environments, as well as for a GO membrane immersed in liquid solution for about 1 hour (LRef) without laser treatment (Fig. 4a). The atomic concentrations of C, O, and N elements in the as-synthesized GO membrane (Ref) are 70.7, 28.4, and 0.9%, respectively. Laser irradiation of the GO membrane in a gaseous ammonia-rich atmosphere provokes a severe reduction process, since the C concentration increases to 91.8 and 91.7% for G50 and G100, respectively; whereas the oxygen concentration abruptly decreases to 7.0 and 6.8%, respectively. Nitrogen is incorporated to the rGO structure to some extent, since its concentration slightly increases to 1.2% (G50) and 1.5% (G100). We recall that G50 and G100 are obtained by accumulation of the same number of laser pulses. Therefore, doubling the laser fluence, despite the large difference in thermal behavior, provokes a rather small difference in GO reduction and nitrogen doping. The evolution of the GO membranes’ chemical composition under irradiation in a liquid environment is different. First of all, the non-irradiated GO membrane immersed in the solution (LRef) exhibits a slight decrease in C and O concentration to 69.4% and 27.8%, respectively. But more interestingly, the N content noticeably increases up to 2.8%. Laser irradiation at a fluence of 100 mJ cm⁻² (L100) leads to a rather limited reduction of the GO material, given that the C concentration slightly increases to 73.8% and the oxygen content moderately decreases to 19.8%. Conversely, the N concentration markedly increases up to 6.4%, which is more than four-fold the concentration reached after a similar irradiation process in gaseous conditions. High resolution C1s XPS spectra were recorded and deconvoluted in four peak contributions centered at 284.6 eV (Ci), 285.6 eV (Cii), 286.5 eV (Ciii) and 287.8 eV (Civ) (Fig. 4b). (Ci) is attributed to graphitic carbon sp² bonds (C=C), whereas the rest of the contributions are respectively assigned to carbon–oxygen (Cii) epoxide-hydroxyl (C–O–C, C–OH), (Ciii) carbonyl (C=O), and (Civ) carboxyl (O–C–OH) bonds (Fig. 4c).²² As observed, the Ci and Ciii signals are the main contributions in the raw GO membrane (Ref) spectrum. The spectrum of the G100 sample reveals an intense Ci signal, whereas those assigned to carbon–oxygen bonds show a large shrinkage in intensity, and the Civ signal is even not present. This result, which corroborates the elemental concentration results (Fig. 4a), certainly points to a high degree of deoxygenation of the GO material. Conversely, and as expected from the element concentration data, the spectrum of the L100 sample reveals a more intense carbon–oxygen signal than G100, with the Cii contribution being larger than Ciii and Civ. The deconvoluted N1s XPS spectrum of the Ref sample is characterized by the presence of two contributions: (Ni) centered at 400.0 eV and (Nii) centered at 402.0 eV. These signals are respectively attributed to amine and pyridinic N⁺–O⁻ chemical groups (Fig. 4c).²² In this sample, the Nii signal is larger than the Ni one, indicating that the small quantity of nitrogen present in the raw GO membrane is predominantly due to pyridinic N⁺–O⁻ groups created during the synthesis of GO sheets. Conversely, the spectrum of G100 reveals that the Ni signal increases, almost reaching the level of the Nii one. This indicates that additional nitrogen is indeed included in the rGO structure during laser irradiation, mainly as amine functionalities. This tendency is completely validated in the laser irradiation of the GO membranes in liquid medium, since the N1s spectrum of the L100 sample reveals a single and more intense Ni (amine) signal. It has to be noted that the presence of C–N bonds would contribute to some extent to the Civ signal.¹⁴

Fig. 4 (a) XPS atomic concentration of C, O, and N species; (b) C1s and N1s XPS spectra of different samples. (c) Scheme, according to the Lerf–Klinowski model, of the observed chemical groups in rGO structure.

In summary, the as-synthesized GO membrane structure is mainly composed of graphene sheets highly decorated with carbonyl, epoxide, hydroxyl and carboxyl chemical groups, in addition to a tiny amount of pyridinic N⁺–O⁻ and amine moieties. Upon laser irradiation in gaseous NH₃–H₂O–N₂ conditions, a prominent removal of oxide-based functional groups is obtained, especially double bonded carbon–oxygen ones, alongside a slight incorporation of nitrogen-based functionalities in the form of amine groups. It is known that sp² π-conjugated domains present in GO confers semiconducting properties to the material, with a band gap in the 2.5–4.0 eV range, which is determined by the extent of oxygenation of the graphene backbone.^5,33–37 In our experiments, 4.7 eV photons are used for irradiation, exceeding this band gap energy range. Then, photoexcitation of sp² domains would take place, resulting in the generation of electron–hole pairs which would react with GO functional groups and ambient molecules. The presence of water is crucial in the reduction process of GO by means of the following reactions:

GO + hν → GO(h⁺ + e⁻) (1)

4h⁺ + 2H₂O → O₂ + 4H⁺ (2)

GO +4e⁻ + 4H⁺ → rGO + 2H₂O (3)

Matsumoto et al. propose that H₂O molecules firstly react with epoxy and hydroxyl groups, followed by the remaining oxide functional groups, during UV photoreduction of GO in the presence of H₂O, leading to the formation of structural defects in the graphene’s basal plane.³⁵ These reactions would combine carbon atoms from the structure of graphene with oxygen, producing CO₂ as a by-product, leaving a high density of holes and edges in rGO sheets. These statements would support the relation between the large deoxygenation process, the formation of large quantities of structural defects, and the observed morphology variation in the G100 sample. However, the G100 and L100 XPS spectra clearly show that, in our case, doubled bonded carbon–oxygen groups are mainly removed, which is probably due to the cross-chemical activity of NH₃ molecules.¹⁷ Additionally, it must be pointed out that given the much higher density of NH₃ and H₂O molecules in the liquid than in the gas, higher photochemical reactivity would be expected in the former case. Nevertheless, the obtained degree of deoxygenation and laser-induced variation of the morphology in GO processed in gas is much larger than in liquid solution (Fig. 1, 2 and 4). Taking into account that numerical simulations point to the development of much larger temperatures in samples irradiated in a gaseous environment than in liquid (Fig. 3), a greater impact of photothermal rather than photochemical mechanisms in the reduction of GO, and the associated formation of structural defects and morphology changes could be suggested. Indeed, the thermal reduction of GO is reported to take place in a rather low temperature range, i.e. 400–1300 K.^38,39 Laser treatment of the GO membrane in liquid solution actually reaches ca. 1400 K, leading to thermal reduction of GO, but no remarkable morphological change is observed. The noticeable distortion of the GO surface, as well as the formation of micrometric mountain-like features observed in samples obtained in gaseous conditions, could be mainly ascribed to high reduction, dense defect creation and large mechanical stress developed by highly extensive thermal processes.³⁶ It is worth noting that a GO material with a high density of structural defects is preferred for catalytic applications, since the structural defects on the graphene sheets could lead to stronger catalyst–support interactions, thus contributing to an enhanced catalytic performance.^35,40 As for the nitrogen functionalities, the initial pyridinic N⁺–O⁻ (Nii) and amine (Ni) groups present in the raw GO membrane evolve to a higher amount of amine groups, especially in samples irradiated in a liquid solution, where significant formation of amine groups takes place. Recent work points to the formation of amine groups in the GO structure when submitted to low temperature treatments in an ammonia environment,⁴¹ supporting our results. It must also be noted that partially reduced GO containing amine species would exhibit photocatalytic activity with improved efficiency in overall water decomposition into H₂ and O₂ at visible wavelengths.¹⁷ Therefore, the presented laser processing method could be a simple and versatile way to obtain photoactive materials for catalytic applications.

The laser-induced modification of the morphology and composition of GO membranes is expected to significantly influence its functional properties, such as electrical conductivity. Fig. 5a shows the macroscopic resistance measured by the van der Pauw method in 5 × 5 mm² samples. Accordingly, the raw GO membrane (Ref) displays a high sheet resistance (ca. 1 MΩ sq⁻¹) due to its high degree of oxygenation. Furthermore, the laser irradiation of GO in a gaseous environment (G100) leads to the reduction of the sheet resistance to around 1.1 kΩ sq⁻¹, whereas the analogous treatment in liquid (L100) provokes a decrease of resistance to about 54 kΩ sq⁻¹. Many reported works demonstrate that the charge propagation in GO-based materials mainly occurs over sp² conducting pathways. In particular, the charge transport would take place via variable range hopping between graphitic regions separated by sp³ functional clusters.^9,42 Accordingly, the observed reduction of resistance in the GO material after laser irradiation could be ascribed to the propagation and merging of the sp² domains, which create percolating pathways for charge flow across the rGO layers.⁴² XPS results indeed support the direct correlation between the amount of oxygenated functionalities and the macroscopic resistance. The G100 sample, obtained by laser irradiation in gaseous conditions, is the most deoxygenated one and simultaneously exhibits the lowest sheet resistance, while the L100 sample, obtained in liquid conditions, presents an intermediate amount of oxygenated functional groups and sheet resistance as compared to Ref and G100. The influence of nitrogen-based functional groups in the overall resistance of rGO is not clearly discernible, though it seems to not be substantial. Besides the chemical composition, the influence of the structural defects in the electrical conductivity of rGO is an important matter. It is expected that structural defects contained in sp²-conjugated domains, which provoke the observed morphological changes, induce electron scattering mechanisms, leading to the increase of electrical resistance.⁴³ In order to go further in the study of the correlation between structural defects and conduction paths, SPM-based electric characterization was carried out. Local current–voltage spectroscopic measurements (IVs) were acquired in [−1,1] volts range at several sites of the raw GO membrane (Ref), G100, and L100 samples’ surface. Typical IVs of the samples are shown in Fig. 5b. The IVs taken in the Ref sample exhibit a very resistive nature, with the surface-tip flowing current at 1 V bias in the range of 0.1–10 pA. Higher conductivity at negative voltages is acquired, pointing to a Schottky rectifying behavior in the SPM tip-surface junction, given the p-type nature of partially reduced GO.⁸ To the contrary, IVs obtained in the G100 and L100 samples reveal symmetric behavior between forward and reverse biasing, as well as a linear relationship between current and applied voltage in the studied range. These characteristics disclose the ohmic nature in the material, alongside the absence of rectification behavior in the contact region between the sample’s surface and the Pt tip. The current ranges at 1 V biasing are about 100 μA and 1 nA for the G100 and L100 samples, respectively. Fig. 5c presents resistance maps obtained by applying 0.5 V between the samples’ surface and an SPM tip. As observed, the resistance maps taken in the Ref sample are very uniform and do not show any characteristic features. The measured average resistance value is higher than 1 TΩ, which is the maximum detectable limit of the system. Interestingly, the resistance maps obtained in the G100 sample show, in addition to micrometric irregular areas with a high resistance (hundreds of kΩ), regions with parallel filament-like features of about tens of kΩ (Fig. 5c and 6a). The resistive filaments, about one hundred nanometers in width and up to a few microns in length, have similar dimensions to topographic filament-like structures observed by AFM (Fig. 1c and 2b). In order to correlate the resistance features with the surface morphology, three-dimensional high resolution topographic maps are coded with their corresponding resistance color (Fig. 6b). As observed, filamentary structures mostly show higher resistance at their topmost sites (crests) than in the trenches between filaments. The measured difference in resistance is about 7%. It should be noted that this difference in resistance is not provoked by inaccuracy in the SPM topographic measurement, since our calculations from the SPM deflection data point to around 1% resistance deviation. Similarly, highly-resistive irregular regions present greater roughness than less resistive ones (Fig. 6c). Therefore, it can be pointed out that the higher resistance regions are those which exhibit a large density of structural defects. Besides, filament-like structures would be formed by somewhat self-organization of structural line defects, leading to the creation of resistive-conductive paths which promote percolation and charge transport along large distances, contributing to the decrease of macroscopic resistance. Conversely, the L100 sample, which does not exhibit the formation of significant structural defects (Fig. 1d and 2c), reveals no formation of parallel conductive-resistive filament-like paths, but randomly distributed irregular regions with different resistance values (Fig. 5c). Moreover, the 3D topography-resistance maps show that the highly resistive areas are indeed not correlated with the morphology at all, since they do not exhibit different morphologies compared to their surroundings (Fig. 6d). It can be pointed out that these high resistance regions correspond to GO areas with a high density of sp³-bonded functional groups. In consequence, the higher the oxidation degree, the larger resistance measured. Due to the larger content of oxygen functionalities and their random distribution, there is poor percolation and few conductive paths are created, leading to higher macroscopic resistance than in the G100 sample. These results clearly indicate that laser-induced structural defects and chemical alterations play a multifaceted role in the macroscopic conductivity of processed GO materials.

Fig. 5 (a) Macroscopic van der Pauw resistance measurements, (b) local current–voltage spectra, and (c) resistance maps of Ref, G100 and L100 samples. Resistance maps were obtained at 0.5 V.

Fig. 6 (a) High resolution resistance map of G100. Inset: resistance profile along the indicated line. 3D topographic merged with resistance-coded color maps acquired in (b and c) G100 and (d) L100 samples. The maps were recorded whilst applying 0.5 V.

Conclusions

Ultraviolet laser irradiation of a flexible, free-standing graphene oxide membrane in ammonia-rich gas and liquid environments was carried out. The samples submitted to analogous laser treatments exhibit remarkably different morphological and compositional properties depending on the nature of the surrounding medium. Thus, samples irradiated in gaseous conditions undergo a significant deoxygenation process, a slight incorporation of nitrogen species into the reduced graphene oxide structure and a large morphological modification. Conversely, the samples processed in a liquid environment reveal a rather low reduction process, but the formation of significant amount of nitrogen functionalities occurs, mainly in form of amine moieties, accompanied by a slight morphological alteration. Though photochemical interactions play an important role in the deoxygenation development of GO, numerical simulations point to a major influence of photothermal processes in the compositional change of the irradiated material, as well as the formation of structural defects in the graphene backbone. Resistance measurements evidence the direct influence of the composition and surface morphology configuration in the macroscopic conductivity of the processed material. The developed laser direct write method is facile, versatile, eco-friendly and easy to be implemented in industrial processes for the development of flexible devices, based on N-doped rGO materials.

Acknowledgements

The authors acknowledge the financial support of the Spanish Ministry of Economy and Competitiveness under the project ENE2014-56109-C3-3-R, in addition to the Romanian National Authority for Scientific Research and Innovation, CNCS – UEFISCDI, under the Grants PN-II-ID-PCE-2012-4-0292 and PN-II-RU-TE-2014-4-1194. ICMAB acknowledges financial support from the Spanish Ministry of Economy and Competitiveness, through the “Severo Ochoa” Programme for Centres of Excellence in R&D (SEV-2015-0496).

Notes and references

1. A. C. Ferrari, F. Bonaccorso, V. Fal’ko, K. S. Novoselov, S. Roche, P. Boggild, S. Borini, F. H. L. Koppens, V. Palermo and N. Pugno, et al., Nanoscale, 2015, 7, 4598–4810.
2. S. Borini, R. White, D. Wei, M. Astley, S. Haque, E. Spigone, N. Harris, J. Kivioja and T. Ryhanen, ACS Nano, 2013, 7, 11166–11173.
3. H.-C. Hsu, I. Shown, H.-Y. Wei, Y.-C. Chang, H.-Y. Du, Y.-G. Lin, C.-A. Tseng, C.-H. Wang, L.-C. Chen, Y.-C. Lin and K.-H. Chen, Nanoscale, 2013, 5, 262–268.
4. K. Krishnamoorthy, R. Mohan and S. J. Kim, Appl. Phys. Lett., 2011, 98, 244101.
5. D. R. Dreyer, S. Park, C. W. Bielawski and R. S. Ruoff, Chem. Soc. Rev., 2010, 39, 228–240.
6. T. Musso, P. V. Kumar, A. S. Foster and J. C. Grossman, ACS Nano, 2014, 8, 11432–11439.
7. S. E. Yalcin, C. Galande, R. Kappera, H. Yamaguchi, U. Martinez, K. A. Velizhanin, S. K. Doorn, A. M. Dattelbaum, M. Chhowalla, P. M. Ajayan, G. Gupta and A. D. Mohite, ACS Nano, 2015, 9, 2981–2988.
8. X. Wang, X. Li, L. Zhang, Y. Yoon, P. K. Weber, H. Wang, J. Guo and H. Dai, Science, 2009, 324, 768–771.
9. G. Venugopal, K. Krishnamoorthy, R. Mohan and S.-J. Kim, Mater. Chem. Phys., 2012, 132, 29–33.
10. A. Iwase, Y. H. Ng, Y. Ishiguro, A. Kudo and R. Amal, J. Am. Chem. Soc., 2011, 133, 11054–11057.
11. T.-F. Yeh, J.-M. Syu, C. Cheng, T.-H. Chang and H. Teng, Adv. Funct. Mater., 2010, 20, 2255–2262.
12. T. Schiros, D. Nordlund, L. Palova, D. Prezzi, L. Zhao, K. S. Kim, U. Wurstbauer, C. Gutierrez, D. Delongchamp, C. Jaye, D. Fischer, H. Ogasawara, L. G. M. Pettersson, D. R. Reichman, P. Kim, M. S. Hybertsen and A. N. Pasupathy, Nano Lett., 2012, 12, 4025–4031.
13. T.-F. Yeh, C.-Y. Teng, S.-J. Chen and H. Teng, Adv. Mater., 2014, 26, 3297–3303.
14. S. Seo, Y. Yoon, J. Lee, Y. Park and H. Lee, ACS Nano, 2013, 7, 3607–3615.
15. J. O. Hwang, J. S. Park, D. S. Choi, J. Y. Kim, S. H. Lee, K. E. Lee, Y.-H. Kim, M. H. Song, S. Yoo and S. O. Kim, ACS Nano, 2012, 6, 159–167.
16. H. Nolan, B. Mendoza-Sanchez, N. A. Kumar, N. McEvoy, S. O’Brien, V. Nicolosi and G. S. Duesberg, Phys. Chem. Chem. Phys., 2014, 16, 2280–2284.
17. T.-F. Yeh, S.-J. Chen, C.-S. Yeh and H. Teng, J. Phys. Chem. C, 2013, 117, 6516–6524.
18. S. Pei and H.-M. Cheng, Carbon, 2012, 50, 3210–3228.
19. R. Trusovas, G. Raciukaitis, G. Niaura, J. Barkauskas, G. Valušis and R. Pauliukaite, Adv. Opt. Mater., 2016, 4, 37–65.
20. W. Gao, N. Singh, L. Song, Z. Liu, A. L. M. Reddy, L. Ci, R. Vajtai, Q. Zhang, B. Wei and P. M. Ajayan, Nat. Nanotechnol., 2011, 6, 496–500.
21. L. Guo, Y.-L. Zhang, D.-D. Han, H.-B. Jiang, D. Wang, X.-B. Li, H. Xia, J. Feng, Q.-D. Chen and H.-B. Sun, Adv. Opt. Mater., 2014, 2, 120–125.
22. A. Pérez del Pino, E. György, C. Logofatu, J. Puigmartí-Luis and W. Gao, Carbon, 2015, 93, 373–383.
23. A. Pérez del Pino, A. Datcu and E. György, Ceram. Int., 2016, 42, 7278–7283.
24. D. Bäuerle, Laser Processing and Chemistry, Springer, Berlin, 2000.
25. D. C. Marcano, D. V. Kosynkin, J. M. Berlin, A. Sinitskii, Z. Sun, A. Slesarev, L. B. Alemany, W. Lu and J. M. Tour, ACS Nano, 2010, 4, 4806–4814.
26. D. Konios, M. M. Stylianakis, E. Stratakis and E. Kymakis, J. Colloid Interface Sci., 2014, 430, 108–112.
27. A. Perez del Pino, E. Gyoergy, L. Cabana, B. Ballesteros and G. Tobias, J. Appl. Phys., 2014, 115, 093501.
28. J. C. Meyer, C. Kisielowski, R. Erni, M. D. Rossell, M. F. Crommie and A. Zettl, Nano Lett., 2008, 8, 3582–3586.
29. L. D. Carr and M. T. Lusk, Nat. Nanotechnol., 2010, 5, 316–317.
30. Z. Qin, M. Taylor, M. Hwang, K. Bertoldi and M. J. Buehler, Nano Lett., 2014, 14, 6520–6525.
31. A. M. Alsayed, M. F. Islam, J. Zhang, P. J. Collings and A. G. Yodh, Science, 2005, 309, 1207–1210.
32. T. T. D. Huynh, M. Vayer, A. Sauldubois, A. Petit and N. Semmar, Appl. Phys. Lett., 2015, 107, 193105.
33. A. Bagri, C. Mattevi, M. Acik, Y. J. Chabal, M. Chhowalla and V. B. Shenoy, Nat. Chem., 2010, 2, 581–587.
34. M. Acik, C. Mattevi, C. Gong, G. Lee, K. Cho, M. Chhowalla and Y. J. Chabal, ACS Nano, 2010, 4, 5861–5868.
35. Y. Matsumoto, M. Koinuma, S. Ida, S. Hayami, T. Taniguchi, K. Hatakeyama, H. Tateishi, Y. Watanabe and S. Amano, J. Phys. Chem. C, 2011, 115, 19280–19286.
36. C. Gomez-Navarro, J. C. Meyer, R. S. Sundaram, A. Chuvilin, S. Kurasch, M. Burghard, K. Kern and U. Kaiser, Nano Lett., 2010, 10, 1144–1148.
37. G. Eda, G. Fanchini and M. Chhowalla, Nat. Nanotechnol., 2008, 3, 270–274.
38. Z. Wei, D. Wang, S. Kim, S.-Y. Kim, Y. Hu, M. K. Yakes, A. R. Laracuente, Z. Dai, S. R. Marder, C. Berger, W. P. King, W. A. de Heer, P. E. Sheehan and E. Riedo, Science, 2010, 328, 1373–1376.
39. R. Larciprete, S. Fabris, T. Sun, P. Lacovig, A. Baraldi and S. Lizzit, J. Am. Chem. Soc., 2011, 133, 17315–17321.
40. S. Moussa, A. R. Siamaki, B. F. Gupton and M. S. El-Shall, ACS Catal., 2012, 2, 145–154.
41. B. J. Schultz, R. V. Dennis, J. P. Aldinger, C. Jaye, X. Wang, D. A. Fischer, A. N. Cartwright and S. Banerjee, RSC Adv., 2014, 4, 634–644.
42. C. Gomez-Navarro, R. T. Weitz, A. M. Bittner, M. Scolari, A. Mews, M. Burghard and K. Kern, Nano Lett., 2007, 7, 3499–3503.
43. A. H. Castro Neto, F. Guinea, N. M. R. Peres, K. S. Novoselov and A. K. Geim, Rev. Mod. Phys., 2009, 81, 109–162.

---