A Review of Strategies for the Synthesis of N-Doped Graphene-Like Materials

Abstract

Methods for synthesizing nitrogen-doped graphene-like materials have attracted significant attention among the scientific community because of the possible applications of such materials in electrochemical devices such as fuel cells, supercapacitors and batteries, as well as nanoelectronics and sensors. The aim of this paper is to review recent advances in this scientific niche. The most common synthesis technique is nitridization of as-deposited graphene or graphene-containing carbon mesh using a non-equilibrium gaseous plasma containing nitrogen or ammonia. A variety of chemical bonds have been observed, however, it is still a challenge how to ensure preferential formation of graphitic nitrogen, which is supposed to be the most favorable. The nitrogen concentration depends on the processing conditions and is typically few at.%; however, values below 1 and up to 20 at.% have been reported. Often, huge amounts of oxygen are found as well, however, its synergistic influence on N-doped graphene is not reported. The typical plasma treatment time is several minutes. The results reported by different authors are discussed, and future needs in this scientific field are summarized. Some aspects of the characterization of graphene samples with X-ray photoelectron spectroscopy and Raman spectroscopy are presented as well.

1. Introduction

Graphene is a promising material for future applications in electrochemical devices. It can be in the form of a single or multilayer film or as vertically oriented graphene sheets, which are often called carbon nanowalls (CNWs). The high capacitance enabled by the large specific surface area of CNWs makes them useful for supercapacitors [1]. The capacitance may be further altered by the introduction of heteroatoms such as O and N [1]. Graphene can also be used for fuel cells, field emitters and batteries [2,3]. In all cases, the electrical properties of the graphene are very important. Both the electrical and chemical properties of graphene can be changed by nitrogen doping. Therefore, the synthesis of N-doped graphene has become a significant scientific challenge in the last few years. Furthermore, N-doped graphene materials are also crucial for the development of metal-free carbon-based catalysts for application in fuel cells [4]. The synthesis of vertically-oriented graphene sheets (CNWs) or similar structures is often performed by plasma-enhanced chemical vapor deposition (PECVD) in CH₄ (or, rarely, in other gaseous hydrocarbons) atmosphere with the presence of other reactive gases that may enhance the quality of CNWs. A review of PECVD methods for CNW deposition was published recently in [5]. However, the incorporation of nitrogen into CNWs and the other graphene-like structures is still an important task, especially during the growth of the sample (direct synthesis), which would enable the homogenous distribution of nitrogen in the deposit. Nowadays, nitrogen is often doped by post-treatment of graphene-like materials in plasmas containing nitrogen; however, this may lead to nitrogen concentrated preferentially at the surface of the graphene-containing film. The conductivity of N-doped graphene depends on the concentration of nitrogen and its chemical binding [6]. Therefore, the characterization of nitrogen peak in X-ray photoelectron spectroscopy (XPS) spectra is crucial. The most common nitrogen configurations that are usually found in N-doped carbon nanoparticles are pyridinic, pyrrolic and quaternary (graphitic) nitrogen [6,7,8]. Schematically they are shown in Figure 1. For obtaining the enhanced electrical properties, graphitic nitrogen is essential [9,10]; however, other nitrogen configurations are also always present, and their concentration may significantly alter the physical and chemical properties of N-doped graphene. Therefore, the optimal procedure for obtaining N-doped graphene should enable the formation of more graphitic and less pyridinic and pyrrolic nitrogen, which may even reduce the conductivity [11,12]. For graphitic nitrogen configuration, n-type doping effect was reported, whereas pyridinic and pyrrolic nitrogen lead to p-type doping effect [13].

In this paper we give a review on one- or two-step synthesis procedures of N-doped graphene-like materials by plasma techniques. For comparison, also some classical chemical methods are summarized. Furthermore, in Appendix A we report about the peculiarities of application of Raman spectroscopy for characterization of graphene-like materials, whereas in Appendix B problems and challenges of XPS characterization of graphene-like materials are given.

2. Plasma-Assisted Synthesis of N-Doped Graphene

2.1. One-Step Plasma Deposition Procedures

Only a few authors have reported direct plasma syntheses of N-doped graphene layers and CNWs. Bundaleska et al. [14] performed a direct single-step synthesis of N-doped graphene using an atmospheric pressure microwave plasma created in ethanol and ammonia with Ar as a carrier gas at 2 kW. The nitrogen configuration in the material was additionally altered by applying infrared (IR) and ultraviolet (UV) radiation. The relative amount of nitrogen N/C was 0.004. The ratio of various nitrogen sub-peaks corresponding to different configurations depended on whether the sample was exposed to radiation or not. The relative amount of pyridinic and pyrrolic nitrogen was higher for the irradiated samples.

Tatarova et al. [15] used a powerful microwave discharge at atmospheric pressure for graphitization and formation of nitrogen-doped graphene sheets. Argon, ethanol, and nitrogen were used as processing gases. The discharge power was varied between 500 and 2000 W. The gas temperature was estimated to as high as 4000 K. The precursors partially decomposed upon high temperature, and the radicals condensate in the flowing afterglow to form clusters of graphene flakes. The authors also applied soft UV radiation to kick off the epoxy oxygen groups and sp³ carbons. If nitrogen in the processing gas was absent, the sp² bonds in graphene at 284.4 eV and the sp³ at 285.2 eV were found using asymmetric fit parameters for C1s obtained on HOPG. The ether and hydroxyl groups were found as well at 286.3 eV and a broad shake-up feature at 290.6 eV. When nitrogen was added, only symmetric fitting was used, and the C-N groups were found at 285.7 eV, the epoxy and carbonyl groups at 287.6 eV, and carboxylate groups at 289 eV. The concentration of nitrogen in the graphene sheaths was 0.2 at.% and oxygen 8 at.%. The authors concluded that the XPS undoubtedly confirmed the doping of graphene by nitrogen atoms, mainly in pyridinic form.

Boas et al. [8] also used a microwave discharge and managed to synthesize thin films of N-doped graphene at low pressure. They were able to sustain plasma at voltage as low as 85 V and the pressure of 43 Torr using a 500-W discharge power. Different concentrations of nitrogen and methane were adopted. The treatment time was 150 s, and the substrate temperature 760 °C. The XPS C1s peaks were fitted with three symmetric sub-peaks, i.e., the main peak at 284.6 eV assigned to sp² hybridized carbon, as well as α-sp² at 286 eV and α-sp³ at 288.5 eV, reflecting the binding of O and N atoms to sp² C and sp³ C, respectively. The XPS N1s peaks were fitted with two sub-peaks, i.e., 398.4 eV corresponding to pyridinic nitrogen and at 397.5 eV corresponding to another type of nitrogen defects. The experimental results were compared to simulations to conclude the chemical state changes in the graphene structure. The concentration of nitrogen was between 0.2 and 4.2 at.%.

Low-pressure gaseous plasma sustained in H₂/CH₄/N₂ gases was also used by Terasawa and Saiki [16] for deposition of graphene islets on copper substrates, but the experimental details were not provided. Despite the weak XPS N1s peak, the authors managed to conclude that the majority of nitrogen was incorporated in the graphitic sites. Also Kumar et al. [17] used H₂/CH₄/N₂ plasma excited by microwaves at low pressure. The pressure was 10 Torr and the discharge power up to 400 W. The treatment time of 4–5 min enabled the deposition of nitrogen-doped few-layer graphene film. The nitrogen content, as deduced from XPS survey spectra, was about 2 at.%. The deconvolution of the N1s spectra revealed that the majority of nitrogen was in graphitic form.

2.2. Plasma Post-Treatment Procedures

As it will be shown in this section, N-doped graphene samples are mostly prepared by using a two-step procedure, including: (1) synthesis of graphene-like material by using mostly plasma-enhanced chemical vapour deposition (PECVD) as a first step, and (2) post-treatment of as-deposited materials with nitrogen-containing plasma, such as N₂ or NH₃ to obtain N-doped graphene-like structures. The second step is often performed in-situ just after accomplishing the first step.

Evlashin et al. [18] performed post-plasma treatment of CNWs in DC glow discharge in N₂ plasma with the pressure below 3 Pa. The modification was performed on a time scale of 2 h. Such modification led to the incorporation of 3 at.% of nitrogen (and also almost 30 at.% of oxygen). N-doped CNWs were found useful for supercapacitor fabrication. In another paper, Evlashin et al. [1] used DC plasma created in O₂, N₂, or their mixtures to investigate the mechanisms that lead to the increased specific capacitance. Four at.% of nitrogen and ten at.% of oxygen was found in CNWs. Oxygen-plasma treated CNWs exhibited higher specific capacitance (8.9 F/cm³) than those treated in nitrogen (4.4 F/cm³).

Cho et al. [12] investigated the electrical conductivity of CNWs after post-plasma treatment in N₂ plasma sustained with a capacitively coupled radio-frequency discharge (CCP). Treatment time was varied between 30 and 300 s. They found increased electrical conductivity after 30 s of treatment. However, with further increasing of treatment time, the electrical conductivity decreased. They found that that carrier density was decreasing with treatment time, whereas the carrier mobility was increasing. This was correlated with an increase of the nitrogen content and density of defects in CNWs, followed by their decrease with increasing treatment time. Nitrogen to carbon ratio increased from 0.095 at 30 s to 0.22 at 300 s. No remarkable changes in surface morphology of CNWs after nitrogen plasma treatment were observed.

Singh et al. [11] used a post-plasma treatment procedure using NH₃ gas. They treated graphene oxide structures in NH₃ plasma for various treatment times and found the best properties at short treatment times up to 5 min, where an increase of sp² carbon occurred, and graphitic nitrogen prevailed, leading to the n-type conductivity. Prolonged plasma treatment caused a decrease of sp² carbon and a decrease of electrical conductivity, which was explained by the high concentration of pyridinic nitrogen; however, graphitic nitrogen still persisted. Nitrogen content was increasing with treatment time up to N/C = 0.25. For 5 min treatment, N/C was 0.15. In general, the time evolution of functional groups was as follows: at the initial stages, pyrrolic nitrogen was formed with a small contribution of pyridinic and graphitic. After 5 min of treatment, a significant increase of graphitic content occurred, whereas pyrrolic nitrogen saturated. For long treatments, pyridinic nitrogen significantly increased together with a continuous increase of graphitic nitrogen. Oxygen content did not depend much on treatment time and the O/C ration was up to 0.27; however, the smallest ratio of 0.15 was found for a treatment time of 5 min.

McClure et al. [19] performed plasma-post treatment of CNWs in N₂/Ar low-pressure plasma created at various RF powers. Nitrogen concentration after doping was 4–20 at.%, depending on the power used. Significant amounts of oxygen (39–52 at.%) were found as well. With the increasing RF power, a shift of XPS low-binding energy N1s peaks (398.7–400.3 eV) towards high-binding energy N1s peaks (~401–404 eV) occurred. The results of Raman spectroscopy showed that I_D/I_G ratio of non-doped CNWs was higher if they were synthesized at a higher temperature, indicating the formation of smaller crystallite sizes and grains. After N-doping, the change in the I_D/I_G ratio was less pronounced for CNWs with the smaller crystallite size.

Yen et al. [20] performed in-situ doping of CNWs in NH₃ plasma immediately after accomplishing their synthesis. Interestingly, and opposite to what has been observed by other authors, only sp² C-N (pyridinic) nitrogen was found, whereas other nitrogen configurations such as pyrrolic and graphitic were not observed. Nitrogen concentration was about 7.8 at.%.

Zhao et al. [21] used plasma treatment of vertically-oriented graphene nanowalls in the PECVD reactor with NH₃ plasma. The substrate was heated to 350 °C, and the sample was treated for 30 min. They found a small amount of nitrogen at about 1.2 at.%. Nitrogen was in the form of amino groups because only one peak at 399.6 eV was found in the XPS spectrum. Oxygen, which was present already on pristine graphene, was reduced for 7.4% after the NH₃-plasma treatment. They found enhanced field emission properties of N-doped samples.

Achour et al. [22] treated CNWs in the PECVD reactor using Ar/N₂ or Ar/O₂ plasma. The same reactor was first used to synthesize pristine CNWs; however, because of a rather high pressure of residual atmosphere (about 100 Pa), already virgin CNWs contained significant concentration of nitrogen (5.1 at.%) as well as oxygen (8.7 at.%). After the plasma treatment in Ar/N₂ or Ar/O₂ mixture, nitrogen concentration was increased to 12.4 and 13.5 at.%, respectively, and oxygen to 23.5 and 33.9 at.%, respectively. Both plasma treatments caused an increased concentration of pyrrolic nitrogen, especially Ar/N₂. SEM images of N-doped CNWs showed more branched CNWs, especially in the case of oxygen plasma treatment, which was explained by etching of the CNWs.

Jeong et al. [23] synthesized N-doped graphene for ultracapacitors with 4-times larger capacitance than for pristine graphene. They used the PECVD reactor first to reduce graphene oxide in H₂ plasma to graphene, followed by subsequent N₂ plasma treatment that induced 1.7 to 2.5 at.% of nitrogen in the form of pyridinic, pyrrolic, and graphitic nitrogen. Improved capacity was correlated with a certain N-configuration at basal planes. Oxygen was found as well in concentrations of about 16–25 at.%.

McManus et al. [24] employed two-step remote plasma treatments in Ar, followed by NH₃/H₂ treatment to form n-type N-doped graphene. Argon plasma was used to make defects first (vacancy defects rather than edge defect), which were then healed with NH₃/H₂-plasma treatment to incorporate nitrogen. Nitrogen content was 2.5 at.%, and 7% of this nitrogen was in graphitic form, which contributed to the n-doping of the sample. Raman results showed a significant increase of D component and decrease and broadening of 2D after Ar treatment. When NH₃/H₂ was applied, the D peak was reduced, and 2D slightly recovered. The shifts of Raman 2D and G peaks, which are an indicator of doping, were found to be shifted to higher values after Ar treatment (increased p-doping), whereas after further NH₃/H₂ treatment, they were shifted back to lower values (increased n-doping). A similar synthesis was also performed by McEvoy et al. [19]; however, they used first remote O₂ plasma treatment followed by further NH₃/H₂ treatment. The authors were mostly focused on defects induced by O₂ plasma treatment for various treatment times, and results for NH₃/H₂ treatment are only briefly mentioned. Nitrogen content was about 3%.

Lin et al. [25] applied ECR enhanced MW nitrogen plasma source in two different modes of operation, i.e., ion- or atom-mode. In the latter case, a metal ion trap was used to remove ions and allow only neutral atoms to react with graphene. Ion and neutral-atom fluxes were 4 × 10¹² s⁻¹ cm⁻² and 2.5 × 10¹⁵ s⁻¹ cm⁻², respectively. The working pressure was 5 × 10⁻⁵ mbar. When exposed to ions, n-type was formed attributed to graphitic nitrogen. When exposed to atoms, pyridinic nitrogen prevailed, causing only minor n-doping. Annealing the sample at 850 °C helped in removing weak nitrogen adsorbates that may have a negative effect on the electronic structure.

Zeng et al. [26] used RF N₂ plasma treatment at various powers from 30 to 70 W to tune the properties of N-doped CNWs. The n-type doping was obtained, and only graphitic nitrogen was found, whose concentration was increasing with increasing discharge power. Electron concentration was increasing as well. Because of increased disorder and degraded crystallinity, I_D/I_G ratio was also increasing with increasing RF power.

Baraket et al. [27] investigated the functionalization of graphene with NH₃ plasma created at various pressures for DNA detection. NH₃ pressure was varied between 25 to 90 mTorr. Nitrogen concentration was increasing with increasing pressure from 5 (at 25 mTorr) to 20 at.% (at 90 mTorr). Raman D-peak intensity was also increasing with the pressure. The presence of amino groups was proven by XPS chemical derivatization, and 45% of nitrogen was assigned to amino groups. Amine functionalized graphene showed enhanced sensitivity for DNA detection.

Lee et al. [28] used an atmospheric-pressure plasma jet (APPJ) to treat mono or multilayer graphene. Treatment time and a distance of the APPJ jet to the sample surface was varied. N₂ was used as a feed gas. To prevent the mixing of the surrounding atmosphere with the jet, APPJ was closed in a box filled with N₂. After the treatment, the surface became hydrophilic, as the water contact angle dropped from 84° to 44°. Nitrogen was mostly in pyridinic configuration, although other configurations were present as well (amine, pyrrolic, graphitic). Thinner graphene layers were more prone to modification because a higher increase of I_D/I_G was observed. I_D/I_G was increasing with increasing treatment time and with decreasing distance of APPJ to the sample.

Santhosh et al. [29] compared the doping of CNWs treated in N₂ or NH₃ plasma. A source of nitrogen radicals was inductively coupled RF plasma, and the samples were placed in the afterglow region. Samples were treated for different times up to 40 s or 25 s for N₂ or NH₃ plasma, respectively. To prevent heating, they were treated in pulses. As revealed from SEM, N₂-plasma afterglow caused etching of CNWs, that was not observed for NH₃. The maximum nitrogen content was 8.0 and 2.8 at.% for N₂ and NH₃, respectively. All three nitrogen configurations were found; however, for NH₃ plasma, also amino groups were found, whereas for N₂, some oxygenated nitrogen species were observed. I_D/I_G ratio was almost linearly decreasing with increasing treatment time in N₂ plasma. In the case of NH₃ plasma, they found that after the initial decrease of I_D/I_G with treatment times up to 15 s, the ratio I_D/I_G increased at longer treatment times. A decrease in electrical conductivity was observed as a result of afterglow treatments, especially when plasma was sustained in N₂. High nitrogen concentrations (above 5 at.%) were not found beneficial for conductivity.

Manojkumar et al. [30] used nitrogen ion implantation at 2 kV for 10, 20, and 30 min to induce defects in CNWs. The samples were biased (negative pulsed DC voltage) during implantation. The source of nitrogen ions was RF plasma. Nitrogen was implanted at the CNW edges and also a few nanometers within the samples. This caused the formation of defects as observed by Raman spectroscopy and revealed from a decrease of I_D/I_G and I_2D/I_G ratio. Depending on the treatment time, the nitrogen concentration was between 7.6 to 8.8 at.%. Oxygen (~13 at.%) was also present. Nitrogen was deconvoluted to three peaks, which were attributed to lone-pair localized nitrogen (N1) at 309.6 eV (pyridinic/pyrrolic/nitrile), lone-pair delocalized at 400.9 eV (N2), and quaternary nitrogen at 402.6 eV (N3). The concentration of nitrogen in the N1 configuration was increasing with treatment time, whereas the concentration of N2 and N3 was decreasing. Nevertheless, N1 was always dominating. A significant transformation of sp² carbon to sp³ was observed after nitrogen implantation, what was explained by defect formation. No noticeable changes in the morphology of CNWs were observed after 20 min of implantation; however, extensive sputtering was observed for the sample treated for 30 min.

The literature about post-deposition plasma treatment indicates that different authors have used plasmas sustained in nitrogen or ammonia. Sometimes also an addition of another gas was used. Most authors used low-pressure high-frequency discharges to sustain the gaseous plasma. The discharge parameters varied significantly, and a few authors also stated the fluxes of reactive species. The surface finish should depend on the fluxes. The variation of the fluxes with discharge parameters is complex because numerous details may be important. For example, the density of reactive gaseous species depends on the type of discharge, the discharge coupling and power, the pressure, the gas purity, the geometry of the plasma reactor etc. Taking into account all these effects, it is difficult to draw general conclusions based on comparison or surface finish reported by different authors. One observation that is common among all authors is that ammonia plasma will favourite functionalization with amino groups. In contrast, nitrogen plasma treatment will result in the formation of various other nitrogen-containing functional groups. A few authors also mention etching of nano-carbon materials during exposure to either ammonia or nitrogen plasma. The etching should be caused by sputtering when the samples are negatively biased, but chemical etching should prevail in cases where samples were left at a floating potential. Obviously, chemical etching only occurs upon exposure to afterglows.

3. Non-Plasma Synthesis of N-Doped Graphene

Generally, non-plasma synthesis procedures are often used by authors, comprising mostly chemical vapor deposition (CVD) using various chemical precursors acting as a source of carbon and nitrogen. Opposite to post-plasma treatment techniques described above, non-plasma CVD techniques enable a direct synthesis of N-doped graphene. Some of them are mentioned below.

Usachov et al. [9] proposed a CVD method for the growth of N-doped graphene from s-triazine precursor. The concentration of nitrogen in graphene was 1–2 at.%. Nitrogen was mostly in the pyridinic form. Therefore, they developed an additional procedure to convert nitrogen to graphitic form. This was done with the help of Au intercalation, followed by annealing up to 635° for 30 min.

Deng et al. [31] synthesized N-doped graphene via the reaction of tetrachloromethane with lithium nitride. Nitrogen content N/C was in the range of 4.5–16.4%. For the sample containing low nitrogen content, graphitic nitrogen prevailed, whereas for the sample with a high nitrogen concentration, pyridinic and pyrrolic nitrogen dominated. The authors also investigated the thermal stability of N-doped graphene at 600 °C. The nitrogen concentration was slightly reduced; however, a still significant amount of nitrogen was retained, indicating that nitrogen species were rather thermally stable. The most thermally unstable was pyrrolic nitrogen. Results of Raman spectroscopy showed a shift of G band of N-doped graphene with a low N content to the lower frequency (typical for n-type), whereas for the sample with a high N content to the higher frequency (typical for p-type), implying different doping effect. They concluded that graphitic nitrogen, as found in the sample with a low N content, is the n-type dopant, whereas pyridinic and pyrrolic found at higher N concentration are p-type dopants. Contrarily, Lu et al. [32] observed transformation from p-type N-doped graphene to n-type, when nitrogen content was increased from 2.1 to 5.6%. They used CVD of 1,3,5-triazine to prepare N-doped samples. They found that pyridinic and pyrrolic N plays an important role in the transport behavior of carriers.

Qu et al. [4] used CVD to prepare N-doped graphene from CH₄ and NH₃ precursors for the application as electrocatalysts for oxygen reduction reaction (ORR) in fuel cells. Nitrogen (4 at.%) was in the form of two configurations (pyridinic and pyrrolic), both important for the ORR process. A very low I_D/I_G ratio was obtained by Raman spectroscopy for N-doped samples (0.06–0.25), indicating high crystallinity. The samples showed better characteristics than Pt/C electrode. Also, Amano et al. [33] synthesized N-doped nanographene for ORR in polymer electrolyte fuel cells. However, they used in-liquid plasma created in a mixture of ethanol and iron phthalocyanine. Nitrogen content was 6 or 11 at.%, depending on the solvent used to disperse phthalocyanine. Four different nitrogen configurations were found: pyridinic, Fe-N, pyrrolic, and graphitic. The D to G ratio I_D/I_G was 1.66 or 1.25, and the estimated grain size L_a = 15.3 or 11.6 nm, depending on the solvent used. Samples showed high catalytic activity, which can also be attributed to the presence of Fe-N bonds.

Opposite to other authors, who found nitrogen in various concentrations, Luo et al. [34] synthesized single-layer graphene doped with pure pyridinic nitrogen only. The deposition was performed on Cu substrate at 900 °C by CVD in H₂ and C₂H₄ with the presence of NH₃. The authors found that pyridinic nitrogen, which is often regarded to be important for ORR activity, is not an efficient stimulant for ORR.

Another approach to control the nitrogen configuration was made by Sui et al. [35]. The authors performed CVD synthesis of N-doped graphene from NH₃ and CH₄ gases at various temperatures from 880 to 1050 °C. They found that nitrogen concentration was decreasing with increasing temperature from 4.5 at.%, obtained at 880 °C, to only 0.7 at.%, at 1050 °C. The authors also observed linear relation between the XPS N1s peak position and the temperature. At the highest temperature, the N1s peak was positioned at 397.7 eV, corresponding to pyridinic N, whereas at the lowest temperature, the N1s peak was found at 400.2 eV, belonging to pyrrolic N. Effect of the growth temperature during the synthesis of N-doped graphene by free-radical reaction using pentachloropyridine on the nitrogen configuration was also investigated by Zhang et al. [13]; however, the authors used much lower temperatures, i.e., 230–600 °C. The synthesized films were dominated either with graphitic nitrogen (230–300 °C) or pyrrolic nitrogen (400–600 °C). The sample with graphitic-N configuration exhibited strong n-type doping and much higher electron mobility than for the sample with pyrrolic N. Also, Wei et al. [36] managed to synthesize graphene with dominant graphitic nitrogen by CVD in NH₃ and CH₄ at a temperature of 800 °C.

Although we can find more such publications about the chemical synthesis of N-doped graphene, we can notice from the papers mentioned above, that similar as for plasma procedures, also in the case of chemical synthesis, there is a problem of controlling nitrogen concentration, its configuration, and thus properties of N-doped graphene. However, there are few reports where authors obtained mostly one nitrogen configuration, or they were using temperature to manipulate the nitrogen configuration. The influence of temperature was not performed yet for plasma techniques, where samples can be heated because of exothermic reactions of plasma radicals on the surface of graphene, and the temperature is thus changing when plasma treatment proceeds. Therefore, it is difficult to control and keep the temperature constant during plasma synthesis.

4. Summary of the Literature Review

All the above-reported literature review is summarized in

Table 1. Overview of published literature on post-synthesis of N-doped graphene.

Ref	Method	Gas	Material	Treatment Parameters for N-Doping	Methods for Characterization	Most Important Conclusions	Possible Application
[1]	Post-plasma treatment	N2, O2 or mixture O2:N2 = 25:75	CNWs	DC discharge, 2kV, 80 mA, pressure: 0.2 Torr, treatment time: 1–120 min (90 min)	SEM, XPS, Raman (/), CV	- 4 at.% N + 10 at.% O, - increased specific capacitance - ID/IG increased from 0.81 to 0.86,1.63, and 1.38 for O2, N2, and O2/N2, respectively; - I2D/IG decreased from 0.95 to 0.72, 0.32, and 0.64 for O2, N2, and O2/N2, respectively, - La decreased from 23.8 nm to 22.4, 11.8, and 14 nm for O2, N2, and O2/N2, respectively	supercapacitors
[18]	Post-plasma treatment	N2	CNWs	DC discharge, pressure: 3 Pa, treatment time: 2 h	SEM, XPS, Raman (/), electrochemical measurements	3 at.% N + 30 at.% O ID/IG increased from 0.78 to 0.90 after doping	supercapacitors
[12]	Post-plasma treatment	N2	CNWs	CCP plasma, power: 400 W, N2 flow 10 sccm, treatment time: 30–300 s	OES, SEM, XPS, Raman (/), van der Pauw-Hall measurements	Electrical properties depended on treatment time. N/C = 9.5% (30 s)–22.2% (300 s)	electronic application
[19]	Post-plasma treatment	N2/Ar(1:2)	CNWs	Pressure: 2 Pa, RF power: 200, 300, 600 W, treatment time: 15 min	XPS, Raman (632.8 nm), TEM, RDE	4–20 at.% of N 39–52 at.% of O, A change of ID/IG after doping depended on the crystallite size La of non-doped CNWs.	Electrocatalyst for polymer electrolyte membrane fuel cells
[20]	Post-plasma treatment	NH3	Vertical graphene nanowalls	MW PECVD, In-situ doping after CNW synthesis using NH3	SEM, TEM, XPS, Raman (532 nm), EIS, CV, galvanostatic measurements	7.8 at.% of N—only pyridinic N was found. Capacitance: 991.6 F/g, Energy density: 275.4 Wh/kg, Power density: 14.8 kW/kg	supercapacitors
[30]	Ion implantation	N2	Vertical graphene nanowalls	RF source of ions, Sample biased with pulsed DC voltage of 2 kV; Treatment time: 10, 20, 30 min	SEM, Raman (514.5 nm), XPS, AFM/AFAM	7.6–8.8 at.% of N and ~13 at.% of O. Reduction of ID/IG from 2.5 to 1.3. No modification in CNWs morphology up to 20 min.	/
[21]	Post-plasma treatment	NH3	Vertically aligned few-layer graphene (FLG)	PECVD, RF power: 20 W, pressure: 1.4 Torr, NH3 flow rate 50 sccm, substrate temperature: 350 °C, treatment time: 30 min	SEM, TEM, XPS, Raman (514.5 nm), field emission properties	1.2 at.% of N in the form of amino groups. IG/I2D decreased from 1.53 to 1.03, whereas ID/IG increased from 1.94 to 2.20. Lower work function and enhanced electron emission properties.	Field emitters
[22]	Post*-plasma treatment *already pristine CNW contaminated with N	Ar/N2 or Ar/O2	CNWs	PECVD, RF power 50 W Ar flow 100 sccm, N2 or O2 flow 10 sccm, pressure: 0.2 Pa, treatment time: 5 min	SEM, XPS, Raman (514 nm), CV, EIS	12.5–13.5 at.% of N. Pyrrolic N was found to be important for improvement of electrochemical transaction. I2D/IG decreased from 0.5 to 0.4 and 0.2, whereas ID/IG decreased from 1.47 to 1.38 and 1.27 for Ar/N2 and Ar/O2, respectively	Electrochemical transductors
[23]	Post-plasma treatment	N2	Graphene layer	PECVD, power: 500 W, pressure: 14 Torr, flow rate 91 sccm, treatment time: up to 3 min followed by annealing at 300 °C for 3 h	TEM, SAED, XRD, SPEM, Raman (/), CV, galvanostatic measurements	1.7–2.5 at.% of N, 16–25 at.% of O, capacitance 4× larger than for the undoped graphene (280 F/gelectrode), excellent cycle life	ultracapacitors
[11]	Post-plasma treatment	NH3	Graphene oxide monolayer	DC plasma, power: 10 W, pressure: 1 Pa, treatment time 1–20 min	SEM, AFM, XPS, UPS, Raman (514 nm), electrical conductivity	N/C = 6–25%, O/C = 15–27%. Pyridinic, pyrrolic, and graphitic N content depended on treatment time. The best results obtained at low treatment time (n-type). ID/IG increased from 1.5 to 1.9 only for long times.	/
[24]	Post-plasma treatment	Ar followed by NH3/H2	Graphene	MW, remote two-step procedure: Ar plasma (60 s), followed by NH3/H2 (300 s), Ar flow = 200 sccm (2 Torr), NH3 flow = H2 flow = 50 sccm (1 Torr), sample position 30 cm downstream	XPS, Raman (532 nm), electrical measurements	2.5 at.% of N, n-type I2D/IG = 1.2, ID/IG = 0.02 for pristine graphene. After Ar treatment ID/IG ~ 2.5, after Ar/NH3/H2 treatment ID/IG ~ 1	transistors
[37]	Post-plasma treatment	Ar/N2	Self-standing graphene sheets	MW power 600 W, Pressure: 100 Pa, total flow 50 sccm, N2:Ar = 10:90, treatment time: 5, 10, and 15 min	Raman (633 nm), XPS, TEM, OES	Pyridinic, pyrrolic and quaternary N, high doping level 5.6%, increase of sp2/sp3 ratio	/
[38]	Post-plasma treatment	NH3	Bilayer graphene	Dose: 3 × 1014 cm−2, other details not specified	XPS, Raman (633 nm)	Doping level: 1.5 × 1013 cm−2. I2D/IG changed from 1.7 to 0.7	/
[39]	Post-plasma treatment	NH3	Graphene sheet	RF 13.56 MHz, with/without an additional Cu grid in the discharge tube after the coil. Power 20 (45 W) with (without) a grid. Remote treatment at a distance 75 (45) cm and treatment time 60 (10) min with (without) a grid.	AFM, Raman (632.8 nm), electrical measurements	Graphene preferably doped near the edge. Doping density: 1.7 × 1012 cm−2 for mild treatment (with a grid).	/
[25]	Post-plasma treatment	N2	Graphene monolayer	Tunable hybrid ECR-MW plasma source, two modes of operation: (1) an ion-mode with a flux: 4 × 1012 ions s−1 cm−2, energy 35 eV, and (2) an atom-mode (by using an ion trap) with a flux of atoms 2.5 × 1015 s−1 cm−2, sample at 850 °C, pressure 5 × 10−5 mbar, treatment time: 10 min	ARIPES, XPS, LEED	Ion-mode treatment: n-type dopping attributed to 8.7 at.% of graphitic N. Atom-mode treatment:mainly pyridinic N is formed, minor n-doping	/
[40]	Post-plasma treatment	N2	Few-layer graphene	Ion irradiation, DC power supply, negative bias 300–350 V, pressure: 460 Pa, treatment time: 20 and 40 s	XPS, Raman (532 nm), TEM, EIS	4.4 and 2.8 at.% of N for 40 and 20 s. Mostly pyridinic and pyrrolic N, graphitic only in a minor concentration. 3-times higher energy conversion efficiency.	Solar cells
[28]	Post-plasma treatment	N2	Mono-, few-, and multi-layer graphene	APPJ (15 kV, 25 kHz, AC), flow rate 15 slm, APPJ positioned in N2 surrounding atmosphere, treatment times: 1–30 s, jet distances from the sample: 1, 2, 3 cm	XPS, TEM, WCA, Raman (532 nm)	Pyridinic nitrogen prevailed. ID/IG increased with plasma treatment time from 0.22 to 0.6. Surface change to hydrophilic (contact angle 44°) because of the OH and COOH groups.	/
[41]	Post-plasma treatment	N2	Monolayer graphene	RF plasma 13.56 MHz, power: 10 W, pressure: 0.12 Torr, treatment time: 0–16 s	XPS, Raman (514.5 nm), CV, RDE	Pyridinic N prevailed, followed by pyrrolic and graphitic N. Enhanced electrocatalytic activity and charge transfer.	Hydrogen production
[42]	Post-plasma treatment	N2	Graphene	Harrick model PDC-32G plasma cleaning unit, power: 100 W, pressure: 0.75 Torr, treatment time: 20, 40, 60, 100 min	XPS, TEM, CV,	1.35 at.% of N and 28 at.% of O. High electrochemical activity for reduction of H2O2. Fast direct electron transfer kinetics for glucose oxidase	Biosensors
[43]	Post-plasma treatment	N2	Graphene sheet	PC2000—Plasma Cleaner, RF 13.56 MHz, power 140 W, pressure: 0.2 Torr, treatment time: 20 min, DC bias 990 V	XPS, Raman (514.5 nm), ORR, CV	8.5 at.% of N and 8.6 at.% of O. Nitrogen was in all typical configurations with the highest pyrrolic content. Higher electrochemical activity toward oxygen reduction	ORR (fuel cells, biosensors)
[26]	Post-plasma treatment	N2	Graphene films	RF, powers: 30, 50, 70 W, flow rate: 50 sccm, pressure: 0.7 Pa, treatment time: 5 min	XPS, Raman (532 nm), Scanning Kelvin Probe, Van der Pauw-Hall measurements	n-type, mostly graphitic N. ID/IG increased from 0.42 to 0.45, 0.60, 0.81 for 30, 50, and 70 W, respectively. Increased power caused increased graphitic content, increased electron concentration, and a shift of Fermi level to higher energy. Work function decreased.	optoelectronics
[27]	Post-plasma treatment	Ar/NH3	Graphene films	Electron beam plasma, 2 kV, 5% NH3, pressure: 25–90 mTorr, total treatment time: 60 s (equivalent plasma exposure time 6 s)	XPS, Raman (/)	N content increased with increasing pressure from 5 to 20 at.%. Raman D peak also increased with pressure.	biosensors
[29]	Post-plasma treatment	N2 or NH3	Graphene nanowalls	IC RF, power: 300 W, flow rate: 100 sccm, pressure: 30 Pa, post-glow region i.e., 10 cm away from the glow, treatment time: 4, 8, 12, 25 s (for NH3) and 10, 20, 30, 40 s (for N2), pulsed treatment to keep the sample < 50 °C	XPS, SEM, NEXAFS, Raman (633 nm), van der Pauw measurements	8.0 and 2.8 at.% of N for N2 and NH3, respectively. All three N types were found as well as amine for NH3 treatment. N2 caused etching, which was not observed for NH3. ID/IG was in general decreasing with increasing treatment time: from 2.8 to 2.28 (40 s, N2) or to 2.68 (12 s, NH3).	/

,

Table 2. Overview of published literature on direct-synthesis of N-doped graphene.

Ref	Method	Gas/Precursor	Material	Treatment Parameters for N-Doping	Methods for Characterization	Most Important Conclusions	Possible Application
[9]	CVD	s-triazine	Graphene monolayer	vapor pressure: 1 × 10−6 mbar, deposition time: 30 min, temperature: 540–635 °C.	XPS, ARPES, NEXAFS	1–2 at.% of N (0.4 at.% of graphitic N) Bandgap 0.2 eV	semiconductors
[14]	Direct plasma synthesis	Ethanol + NH3	Free-standing graphene	MW at atmospheric pressure, additional IR and UV treatment for sp2 C and N-type manipulation. Deposition yield: 1.3 mg/min	XPS, SEM, TEM, Raman (633 nm), CV, van der Pauw method, OES and FTIR	N/C = 0.4%,O/C = 1.5%ID/IG ~ 0.9 after irradiation; Higher relative amount of pyridinic and pyrrolic N for the irradiated CNWs.	supercapacitors
[44]	Thermal segregation		Few-Layer graphene	Annealing of a substrate consisting of N-containing boron and C-containing Ni films	XPS, Raman (514.5 nm), AFM, electrical characteristic of fabricated field-effect transistors	Higher N doping caused lower La. La reduced from 65 nm to 21 and 8 nm for N/C = 0.6 and 2.9%, respectively Doping level 4 × 1013 cm−2, bandgap 0.16 eV, n-type	nanoelectronics
[31]	Chemical synthesis	CCl4 + Li3N	Few-layer graphene	Reaction of CCl4 with Li3N	STM, TEM, XPS, Raman (633 nm), thermal stability tests	N/C = 4.5–16.4%. In the sample with a high N content, pyridinic and pyrrolic N dominated (p-type). For the sample with a low N content, graphitic N dominated (n-type).	nanoelectronics
[32]	CVD	1,3,5-triazine	Graphene sheets	Chemical vapor deposition of 1,3,5-triazine to Cu substrate at different temperatures 700, 800 and 900 °C	XPS, Raman (473 nm), AFM, SEM, TEM, electrical measurements	N/C = 2.1–5.6%. A lwer temperature was favorable to obtain higher N doping. Increasing of N-doping content caused the transformation of p-type to n-type.	nanoelectronics
[4]	CVD	CH4 + NH3	Few-layer graphene	NH3/CH4/H2/Ar = 10/50/65/200 sccm for 5 min, followed by NH3/Ar for another 5 min, temperature 1000 °C	AFM, TEM, Raman (514.5 nm), XPS, SEM, XRD, RDE	4 at.% of N, pyridinic, and pyrrolic N-configuration. ID/IG = 0.06–0.25 Improved electrocatalytic activity and stability.	Fuel cells
[33]	In-liquid plasma	ethanol + Fe-phthalocy-anine	Nano-graphenes	In-liquid plasma synthesis from ethanol and Fe-phthalocyanine	SEM, XPS, Raman (/), ORR, CV	6–11 at.% of N, N-configurations: pyridinic, Fe-N, pyrrolic and graphitic. ID/IG = 1.25–1.66, La = 11.6–15. 3 nm	Polymer electrolyte fuel cells
[16]	PECVD	H2/CH4/N2	Mono- to multilayer graphene	Flow rates of H2/CH4/N2 = 20/5/1 sccm. Power 300 W, pressure 1.08 Pa, growth time: 5 min, annealing to 500 and 950 °C.	XPS, Raman (532 nm)	N content: 0.5, and 1.1% at 950 and 500 °C, respectively. N mostly in the graphitic form. I2D/IG = 2.1 (decreasing with N content). ID/IG = 1–1.5. An island like growth.	/
[17]	PECVD	H2/CH4/N2	Few-layer graphene	MW, first H2/CH4 treatment at 500 W, followed by N2/CH4 treatment at 150 W. Pressure: 10 Torr, flow: H2 = CH4 = 10 sccm, N2 = 50 sccm. Total growth time: 5 min. Temperature: 800 °C.	Raman (532 nm), XPS, SEM, TEM	2 at.% of N in the form of pyridinic, graphitic, and oxygenated form. ID/IG increased from 1.34 to 2.3, and I2D/IG decreased from 1.0 to 0.28.	/
[15]	PECVD	Ar/ethanol/N2 +UV	Free- standing graphene	MW, power: 2 kW, Ar flow 1200 sccm, ethanol flow 15 sccm, N2 flow 5 or 10 sccm	XPS, SEM, FTIR, NEXAFS, Raman (532 nm)	0.2 at.% of N and 8 at.% of O, mostly pyridinic nitrogen and some graphitic, growth yield 2 mg/min.	/
[8]	PECVD	H2/CH4/N2	Graphene bilayers	MW, power: 500 W, N2:CH4 = 2:1, 3:1, or 5:1, H2 flow: 10 sccm, pressure: 43 Torr, deposition time: 2.5 min, temperature: 760 °C	XPS, Raman (532 nm), simulations	2.0–4.2 at.% of N, pyridinic, and another peak related to other type of N defects. Formation of interlayer bonds mediated by nitrogen defects. ID/IG increased from 0.6 to 2, I2D/IG decreased from 1.5 to 0.7.	/
[34]	CVD	H2/C2H4/ NH3	Single-layer graphene	Thermal deposition in H2/C2H4/NH3 at various NH3 flow rates	XPS, Raman (532 nm), SIMS, UPS, RDE voltametry	N/C = 1.6–16%. Depending on NH3 flow, pure pyridinic N formation.	ORR
[35]	CVD	CH4/NH3	Graphene domain film	Thermal deposition in NH3 and CH4 at various temperatures 880–1050 °C	XPS, Raman (532 nm)	Control of N configuration by growth temperature. At high temperatures, mostly pyridinic N was formed, and pyrrolic N at low temperatures. The N concentration was decreasing with increasing temperature. N = 4.5 and 0.7 at.% at 880 and 1050 °C, respectively.	/
[13]	Free-radical reaction	Penta-chloro-pyridine	Graphene films	Free-radical reaction from pentachloropyridine at various growth temperatures 230–600 °C	XPS, Raman, STM, electronic properties	Control of N configuration by growth temperature. Graphitic N dominated at 230–300 °C and pyrrolic N at (400–600 °C). ID/IG = 0.48–1.91 (minimum at 400 °C) La = 7.4–19.6 nm	/

and

Table 3. Summary of discharge parameters and XPS results for the direct and post-plasma synthesis of N-doped graphene.

Ref	Gas	Discharge	Material	Treatment Time	Power	Pressure/Flow	N and O Content as Obtained by XPS
[1]	N2, O2, or mixture	DC	CNWs	1–120 min (90 min)		0.2 Torr	N: 4 at.% O: 10 at.% N/C = 4.7%
[18]	N2	DC	CNWs	2 h		3 Pa	N: 3 at.% O: 30 at.% N/C = 4.5%
[12]	N2	CCP	CNWs	30, 180, 300 s	400 W	10 sccm	N/C = 9.5% at 30 s N/C = 16.4% at 180 s N/C = 22.2% at 300 s
[19]	N2/Ar	RF	CNWs	15 min	200, 300, and 600 W	2 Pa	N: 5, 7, or 18 at.%, O: 41, 52 or 39 at.%, N/C = 9, 17 or 42% for 200, 300 and 600 W, respectively (pristine CNW deposited at 860 °C). N: 4, 6, or 20 at.% O: 47, 45, 39 at.% N/C = 8, 12 or 49% for 200, 300 and 600 W, respectively (pristine CNW deposited at 730 °C)
[20]	NH3	MW PECVD	Vertical graphene nanowalls				N: 7.8 at.%, pyridinic N
[30]	N2	RF/DC biased	Vertical graphene nanowalls	10, 20, 30 min			N: 7.6–8.8 at.% O: ~13 at.% N/C ~ 10%
[21]	NH3	RF PECVD	Vertically aligned few-layer graphene (FLG)	30 min	20 W	1.4 Torr 50 sccm	N: 1.2 at.%, amino groups
[22]	Ar/N2 or Ar/O2	RF PECVD	CNWs	5 min	50 W	0.2 Pa 100/10 sccm	N: 12.5–13.5 at.% Pyridinic, pyrrolic, graphitic, and oxygenated N
[23]	N2	PECVD	Graphene layer	0.5–3 min + followed by annealing 3 h	500 W	14 Torr 91 sccm	N: 1.7 at.%, O: 25.5 at.%, N/C = 2.3% for 0.5 min N: 1.9 at.%, O: 15.9 at.%, N/C = 2.3% for 1 min N: 2.2 at.%, O: 21.8 at.%, N/C = 2.8% for 1.5 min N: 2.4 at.%, O: 16.9 at.%, N/C = 3.0% for 2 min N: 2.5 at.%, O: 19.6 at.%, N/C = 3.2% for 3 min Pyridinic, pyrrolic and graphitic. Graphitic content was decreasing with increasing treatment time.
[11]	NH3	DC	Graphene oxide monolayer	1–20 min	10 W	1 Pa	N/C = 6%, O/C = 27%, for 1 min N/C = 9%, O/C = 25% for 2 min N/C = 15%, O/C = 15% for 5 min N/C = 20%, O/C = 22% for 10 min N/C = 25%, O/C = 26% for 20 min Pyridinic, pyrrolic and graphitic
[24]	Ar + NH3/H2	MW	Graphene	60 s in Ar + 300 s in NH3/H2		1 Torr, 200/50/50 sccm	N: 2.5 at.%
[37]	Ar/N2	MW	Self-standing graphene sheets	5, 10, 15 min	600 W	100 Pa 45/5 sccm	N/C ~ 8%, O/C ~ 19% for 5 min N/C ~ 4%, O/C ~ 26% for 10 min N/C ~ 6%, O/C ~ 115% for 15 min (estimated from the graph) Pyridinic, pyrrolic, and graphitic N
[25]	N2	ECR-MW	Graphene monolayer	Two modes of operation: (1) ion-mode, flux = 4 × 1012 ions s−1 cm−2, (2) atom-mode, flux = 2.5 × 1015 s−1 cm−2,		0.005 Pa	Ion-mode treatment: N: 8.7 at.%, mostly graphitic Atom-mode treatment: Minor doping, mainly pyridinic
[40]	N2	DC biased	Few–layer graphene	20 s, 40 s		460 Pa	N: 4.4 at.% at 40 s N: 2.8 at.% at 20 s Mostly pyridinic and pyrrolic, graphitic N in a minor concentration
[41]	N2	RF	Monolayer graphene	14 s	10 W	0.12 Torr	2.2 at.%, pyridinic prevails, followed by pyrrolic and graphitic
[42]	N2	Harrick model PDC-32G plasma cleaning unit	Graphene	20, 40, 60, 100 min	100 W	0.75 Torr	N: 1.35 at.% O: 28 at.% N/C = 1.9%
[43]	N2	RF biased (PC2000—Plasma Cleaner)	Graphene sheets	20 min	140 W	0.2 Torr	N: 8.5 at.% O: 8.6 at.% N/C = 10% Pyridinic, pyrrolic (the highest content), and graphitic
[26]	N2	RF	Graphene films	5 min	30, 50, 70 W	0.7 Pa	N: 2.5, 2.8 and 3.2 at.%, for 30, 50 and 70 W, respectively Mostly graphitic
[27]	Ar/NH3	Electron beam plasma	Graphene films	total treatment time 60 s (equivalent plasma exposure time 6 s)		25–90 mTorr	N: 5 at.% at 3.3 Pa N: 10 at.% at 6.7 Pa N: 17 at.% at 10 Pa N: 20 at.% at 12 Pa
[29]	N2 or NH3	RF	Graphene nanowalls	4, 8, 12, 25 s (for NH3) 10, 20, 30, 40 s (for N2)	300 W	30 Pa	N: 8.0 and 2.8 at.% for N2 and NH3, respectively Pyridinic, pyrrolic and graphitic N, as well as amine in the case of NH3 treatment
[14]	EtOH/ NH3	MW	Free-standing graphene	/		Atmospheric	N/C = 0.4%, O/C = 1.5% Higher amounts of pyridinic and pyrrolic N, if irradiated.
[33]	EtOH and Fe-phthalo-cyanine	In-liquid plasma	Nano-graphenes				N: 6–11 at.% Pyridinic, Fe-N, pyrrolic, and graphitic N
[16]	H2/CH4/N2	PECVD	Mono- to multilayer graphene	5 min	300 W	1.08 Pa	N: 0.5%–1.1% for 950 and 500 °C, respectively Mostly graphitic N.
[17]	H2/CH4/N2	PECVD	Few-layer graphene	5 min	500 W	10 Torr 10/10/50 sccm	N: 2 at.% Pyridinic, graphitic and oxygenated N
[8]	N2/H2/CH4	MW PECVD	Graphene bilayers	2.5 min	500 W	43 Torr	N: 2.0 at.% for N2:CH4 = 2:1 N: 4.2 at.% for N2:CH4 = 3:1 and 5:1 Pyridinic N and another one related to other type of defects
[15]	Ar/EtOH/N2	MW PECVD + UV	Free- standing graphene	/	2000 W	1200/15/(5 or 10) sccm	N: 0.2 at.% O: 8 at.% Mostly pyridinic and some graphitic

. In Table 1 and Table 2 a summary of the methods for post-synthesis and direct- synthesis of N-doped graphene-like materials, is shown, respectively, whereas in Table 3 results of XPS characterization and treatment conditions of N-doped graphene synthesized by plasmas are summarized.

5. Discussion

Because of the various conditions used by authors performing plasma synthesis of N-doped graphene, such as type of the discharge, gas mixture, pressure, power, treatment time, substrate temperature, sample biased or not, etc. it is difficult to draw general correlations regarding the treatment parameters and the resulting materials’ properties. It was found that all these parameters may strongly influence the nitrogen content. Moreover, the fluxes of reactive nitrogen species from the discharge to the surface were rarely reported. Furthermore, also detailed surface chemical composition is often not reported, because the N/C ratio is often provided instead of the full composition. Many authors reported significant concentration of oxygen, for example. As shown in Table 3, the oxygen content may be very high, even exceeding the carbon content as in [37]. This points out that N-doped graphene is more susceptible to functionalization with oxygen as elaborated in [8]. Anyway, the following observations are common to the many reviewed papers:

(1)

No general correlation between the nitrogen content and the characteristics of N-doped graphene samples was found. Sometimes, even very low nitrogen content of about 1 at.% was found beneficial. The induced defects increase with increasing nitrogen concentration, but this observation should be taken with a precaution because the effect of oxygen was not always included.

(2)

Nitrogen is usually present in various configurations such as pyridinic, pyrrolic, and graphitic. Sometimes also oxidized nitrogen groups were reported. Usually, all three typical nitrogen configurations are found, but they differ in concentration among authors. There are only a few papers where mostly only one nitrogen configuration was reported i.e., pyridinic [34] or graphitic [36]. Therefore, it is still a challenge to control the type of nitrogen incorporated into the graphene-like structures.

(3)

Treatment times for N-doping were mostly of the order of 10 min (without taking into account the time needed for the preparation of the pristine graphene samples in the case of the two-step procedure). In rare cases, treatment times of the order of 10 s were reported [29,41]. RF plasma was used in both cases.

(4)

Systematic investigation of N-doping versus treatment parameters were provided in several papers. For example, the discharge power was varied systematically in [19,26], pressure in [27], and treatment time [1,11,24]. It should be stressed that such systematic experiments take time, especially spectra acquisition and the interpretation. Comparison of surface finishes obtained by different gases (N₂ and NH₃) using the same experimental system was reported in [29].

(5)

Nitrogen content in graphene-like materials is generally higher when using N₂-plasma than NH₃-plasma. Furthermore, in addition to pyridinic, pyrrolic, and graphitic nitrogen, the presence of amino groups was reported for NH₃-plasma treatments.

It is difficult to draw any clear correlation between the nitrogen concentration, its configuration, and final properties of N-doped graphene-like materials, for example its conductivity. e.g., Cho et al. found an increased conductivity of N-doped CNWs in comparison to pristine CNWs [12]. They also found that the conductivity was decreasing with increasing nitrogen content in the range of N/C from 9.5 to 22%). Different results were reported by Santhosh et al. [29], who found decreasing conductivity of N-doped CNWs for both N₂- and NH₃-plasma treatment, where the maximum nitrogen content was 8 and 2.8 at.% for N₂ and NH₃ plasma, respectively. No correlation between the measured conductivity and nitrogen content was found by Santhosh et al.

The results summarized in Table 3 hardly reveal any correlation between the treatment parameters and reported concentration of nitrogen in various binding sites. Despite the fact that no direct comparison is possible because of various discharge parameters, we nevertheless present the results from Table 3 graphically in Figure 2, Figure 3, Figure 4, Figure 5 and Figure 6. Many authors reported increased N-concentration with increased treatment time at their particular conditions. This is shown in Figure 2. From Figure 2 we can also see scattering of measured points obtained by between different authors, meaning that treatment time is not the only decisive parameter. When comparing applications of N₂ or NH₃ gas (Figure 3), it seems that more nitrogen is introduced into the graphene structure if authors used N₂ gas. Furthermore, in the case of ammonia, authors reported the presence of amino groups [29]. There is also no clear correlation versus the gas pressure (Figure 4) and discharge power (Figure 5), although one might speculate that the N concentration is increasing with increasing power and pressure as found by some authors at their particular conditions. Nevertheless, scattering of the results can be explained as a consequence of a different concentrations and doses of reactive plasma species in the discharge operating at different powers, pressures, or treatment times. In Figure 6 is shown the reported N/C ratio versus two parameters, i.e., plasma treatment time and discharge power, but also here there is no clear correlation.

Finally, we should also mention the problems associated with a uniform modification of vertically-oriented graphene (CNWs). As already mentioned, such structures are typically first deposited in an atmosphere free from nitrogen, followed by a post-treatment with plasma sustained in N₂ or NH₃. The interaction of reactive species from nitrogen plasma with CNWs can be predicted using the schematic illustration in Figure 7. Depending on discharge parameters, there are many reactive species in nitrogen plasma, such as positively charged ions, metastable molecules, and neutral atoms. The uppermost graphene sheets are subjected to a larger flux of ions than those deep in the film. Furthermore, if the sheets are not perfectly perpendicular (very likely), the surfaces not facing plasma are not affected by ions at all. Opposite to ions, the atoms do not feel the electric field and they can move randomly, as shown schematically in Figure 7. Because the distance between the neighboring sheets is smaller than the mean free path, the collisions with surfaces will prevail. Even if the probability for recombination of N atoms is low, the huge number of collisions will create a large gradient of N-atoms across the film. In one experiment, it has been shown that the CNWs exhibit an extremely high recombination coefficient for oxygen atoms because of the porous structure [45]. The coefficient for nitrogen atoms has never been reported, but in analogy with oxygen, it can be assumed that the loss of N atoms by surface recombination to parent molecules is also large enough to create strong gradients of N atoms across the CNW film. As a result, it can be expected that CNWs deep below the surface will be under-treated.

Gradients in both positive ions and atoms are therefore likely to occur upon treatment of CNWs or similar materials with nitrogen plasma. The result is the overtreatment of the surface layer (which may lead to etching) and undertreatment of the graphene sheets deep in the structure. It is expected that the N concentration in the graphene network will be high on the surface and negligible close to the interface with the substrates. Unfortunately, we have not found a single paper on the depth profiling of CNWs to reveal uniformity of modification versus depth. Any problems with gradients are avoided if a one-step procedure is accomplished. To the best of our knowledge, no report on the growth of vertically oriented graphene structures rich in nitrogen has appeared in the scientific literature. A possible explanation is a well-known fact that any attempt to use a mixture of hydrocarbon gas and nitrogen at low pressure leads preferentially to the formation of HCN molecules rather than the deposition of nitrogenated carbon nanostructures.

6. Conclusions and Perspectives

N-doped graphene-like materials can be synthesized by different techniques. The most straight-forward one is chemical vapor deposition: gaseous precursors are introduced into a furnace where they gradually decompose due to high temperature and a film containing carbon and other elements grows on a substrate. This procedure will normally lead to the formation of a rather compact film, so it is used predominantly for deposition of horizontally oriented films. The technique is scalable but not very interesting for application in electrochemical devices where an extremely high surface-to-mass ratio is required. Most authors used techniques based on the application of gaseous plasma.

Gaseous plasma can be used for one-step synthesis, which is typically performed by introducing a mixture of gaseous precursors into the plasma reactor. Methane (or any other light hydrocarbon) is used as a source of carbon, and nitrogen or ammonia gas is a source of nitrogen in the graphene-like film. This technique also seems to be useful only for deposition of horizontally-oriented structures.

Three-dimensional structures of N-doped graphene-like materials of high porosity are preferably synthesized using a two-step procedure. In the first step, carbon nanowalls free from nitrogen are deposited using a PECVD technique, and these materials are then functionalized with nitrogen in the second step, where plasma free from hydrocarbons is used. The three-dimensional carbon nanomaterials are doped with nitrogen at various experimental configurations, from nitrogen or ammonia afterglows (where pure chemical interaction occurs) to biased samples where ion implantation is the predominant mechanism.

The properties of N-doped graphene-like materials vary significantly depending on their structure, which is related to the method used for their synthesis. One limitation of the methods is their ability to produce N-doped samples with controllable concentration and configuration of nitrogen, and another limitation is the upscaling of the method to mass production on the industrial scale. Two-step procedures are too time-consuming to be interesting for the industrial production of N-doped CNWs. Direct one-step procedures are thus much better alternatives. However, enabling deposition on large surface areas and line-production still remains a challenge.

The science of N-doped vertically-oriented graphene structures is still in its infancy. Numerous authors have used different plasma techniques for synthesizing such materials, which are desired in several applications, in particular the fuel cells, batteries, and supercapacitors. The obtained structures differ significantly, so no correlation between the processing parameters and the resultant structures could be drawn on the basis of reviewed literature. The concentration of nitrogen in the materials probed by XPS was from less than 1 at.% and up to about 20 at.%. Typically, it was about a few at.%. No clear correlation between the amount of nitrogen and improved characteristics of N-doped graphene samples was found. Even when nitrogen concentrations were very low at about 1 at.%, the authors reported improved characteristics of the samples. In addition, too high nitrogen concentration may again be not good enough. Several types of chemical bonds between N-atoms in the graphene structure have been confirmed. At the current level of the state-of-the-art, it is not possible to deduce the processing parameters that would lead to preferential incorporation of N-atoms to a specific binding site; therefore, this still remains a scientific challenge. The review also reveals that most authors investigated the N-incorporation in the surface film as probed by XPS and Raman spectroscopy. The results may not be representative for thick films of vertically-oriented graphene because strong gradients of fluxes of reactive nitrogen particles are expected.

The synthesis of vertically oriented N-doped graphene such as nanowalls, nanomesh, or similar morphological forms on the surface of a smooth substrate is currently limited to a two-step procedure. In the first step, a film containing nitrogen-free carbon structures is deposited, preferably by PECVD, and in the next step, the structures are exposed to nitrogen plasma. A one-step procedure currently enables deposition of horizontally oriented graphene structures of limited applicability. There is a need for inventing a method for a one-step deposition because the gradients in nitrogen concentration could be suppressed if not avoided using such a technique. Any attempt to deposit vertically oriented N-doped carbon nanostructures using hydrogenated carbon precursors in plasma with nitrogen or ammonia admixtures failed, probably due to preferential etching and formation of volatile hydrogen cyanide.

The deposition of vertically oriented carbon nanostructures with an appropriate concentration of nitrogen in various chemical bonds remains a scientific challenge. In order to enlighten the kinetics of the formation of various nitrogen configurations in graphene structure, the correlations between the fluxes and/or fluences of reactive nitrogen plasma particles and surface finish are desired. Once the correlations are known, further studies will be necessary to form desired structures in a controllable and repeatable manner.