Skip to main content
NCBI home page
ACS Omega logoLink to ACS Omega
. 2024 Mar 29;9(14):16592–16600. doi: 10.1021/acsomega.4c00495

Synergistic Effects of Carbon Vacancies in Conjunction with Phosphorus Dopant across Bilayer Graphene for the Enhanced Hydrogen Evolution Reaction

Huimin Hu †,‡,§,∥, Jin-Ho Choi †,‡,§,∥,*
PMCID: PMC11007721  PMID: 38617609

Abstract

graphic file with name ao4c00495_0009.jpg

Bilayer graphene (BLG) exhibits distinct physical properties under external influences, such as torsion and structural defects, setting it apart from monolayer graphene. In this study, we explore the synergistic effects of carbon vacancies, in conjunction with phosphorus dopants, across BLG, focusing on their impact on structural, magnetic, electrical, and hydrogen adsorption properties. Our findings reveal that the substitutional doping of a phosphorus atom into a single carbon vacancy in a graphene layer induces substantial structural distortion in BLG. In contrast, doping phosphorus into a double vacancy maintains the flat structure of graphene layers. These distinct layer structures affect the electron distribution and spin arrangement, leading to varied electronic configurations and intriguing magnetic behaviors. Furthermore, the presence of abundant unsaturated electrons around the vacancy promotes the capture and bonding of hydrogen atoms. Hydrogen adsorption on BLG results in substantial orbital hybridization, accompanied by significant charge transfer. The calculated Gibbs free energies for hydrogen adsorption on BLG range from −0.08 to 0.09 eV, indicating exceptional catalytic activity for the hydrogen evolution reaction. These findings carry implications for optimizing the properties of graphene layers, making them highly desirable for applications such as catalysis.

1. Introduction

Graphene, as the pioneering monolayer two-dimensional (2D) material successfully synthesized, stands out prominently within the realm of 2D materials.1,2 Comprising single-atom-thick hexagonal carbon structures, graphene’s carbon atoms engage in sp2 orbital hybridization, forming a π-bond on the surface. This unique electronic configuration results in a distinctive energy band structure characterized by a zero band gap and Dirac cone, contributing significantly to graphene’s exceptionally high electron mobility and electrical conductivity.3−5 Moreover, graphene boasts a high specific surface area and exceptional light absorption properties, among other advantageous traits. These versatile features render graphene widely applicable in the fields of electronics, energy storage, and catalysis.6,7

It is crucial to recognize that the intrinsic monolayer graphene lacks inherent catalytic activity for producing hydrogen (H2), a promising clean energy source. Currently, graphene is commonly employed as a substrate for active materials, forming a heterostructure catalyst. For instance, Chen et al. developed a nickel-doped molybdenum disulfide with graphene as a substrate, proving to be an efficient material for the hydrogen evolution reaction (HER) catalyst.8 Wu et al. theoretically confirmed the positive role of graphene substrates in the catalytic behavior of GeP3 and SnP3 monolayers.9

Alternatively, graphene can undergo external modulation to finely adjust its electronic properties to amplify active sites and ultimately to enhance its intrinsic catalytic activity. Manipulating graphene by doping or introducing carbon vacancies stands as a widely employed strategy to improve its catalytic activity in H2 production. For instance, Denis et al. investigated the dual-doped graphene systems and found that Al–O, S–N, P–O, and Si–B (para)-codoped graphene presented extraordinary chemical reactivity.10 Zhang et al. proposed nitrogen, phosphorus, and fluorine tridoped graphene as a multifunctional catalyst, demonstrating outstanding catalytic properties for the oxygen reduction reaction (ORR), oxygen evolution reaction (OER), and HER.11 Kumatani et al. revealed that the synergy of chemical doping and boundary engineering in graphene can significantly enhance its catalytic activity for HER.12

With extensive research on graphene, attention has gradually expanded to the study of bilayer graphene (BLG). In comparison to monolayer graphene, bilayer graphene possesses additional degrees of freedom such as stacking manners, interlayer spacing, and rotation angle between layers. These degrees of freedom contribute to the emergence of distinctive physical properties, enhancing the intrigue and fascination surrounding research on BLG. For example, the magic-angle graphene, achieved by precisely twisting BLG to a specific angle, triggers a strong correlation effect among electrons, leading to the manifestation of distinctive characteristics such as superconductivity and ferromagnetism.13,14

Bilayer graphene also holds promising applications in the field of energy storage and is anticipated to emerge as a favorable electrode material for secondary batteries.15−17 Several studies on the fundamental properties of doped bilayer graphene have been reported previously.18−21 Besides, bilayer graphene codoped with FeN4–HfN4 and FeN4–IrN4 has been reported for its remarkable catalytic prowess in the realms of HER, OER, and ORR.22 The bilayer heterojunction composed of CoN4-embedded graphene and B-doped graphene has been documented to exhibit outstanding oxygen reduction electrocatalytic activity.23 Recently, it was reported that phosphorus-doped monovacancy at the hollow site of bilayer graphene is a promising alternative to Pt/C catalysts under acidic conditions for the oxygen reduction reaction (ORR).24

Applying high-energy atomic/ion bombardment to graphene layers can generate C (mono- and bi-) vacancies, enabling the relatively high-energy process of single-atom doping.25−28 Consequently, the synthesis of P-doped graphene layers is feasible, and the creation of a composite material consisting of both P-doped and defective graphene layers is within reach.29−31

In contrast to single metal atom catalysts, nonmetallic single-atom catalysts demonstrate heightened stability and durability during catalytic reactions. Moreover, nonmetallic atoms offer environmental friendliness and cost-effectiveness in both preparation and utilization processes, further emphasizing their advantageous characteristics.32 However, thus far, research on nonmetallic atom-doped BLG in the field of HER catalysis remains scarce. Furthermore, defective configurations in BLG, along with the resulting structural characteristics and significant changes in physical properties, have garnered substantial interest due to their potential impact on HER catalysis.

In this study, we explore the collaborative effects of phosphorus doping and carbon vacancies on the structural, magnetic, electronic, and catalytic properties of BLG by using first-principles density functional theory (DFT) calculations. The DFT calculations reveal that substituting a phosphorus (P) atom into the single carbon (C) vacancy of a graphene layer causes significant out-of-plane distortion in BLG, leading to nonflat layer structures. However, when doping occurs within a double C vacancy, the flat layer structure of graphene is well-preserved. These structural distortions also result in variations in the interlayer spacings of BLG. The interlayer interactions across BLG involving P dopants and C vacancies in different layers not only result in different geometrical configurations but also bestow distinct electronic, magnetic, and catalytic properties. Several combinations containing a P-doped layer exhibit outstanding catalytic activities for HER, as reflected by the H adsorption Gibbs free energies (−0.08–0.09 eV), approaching the optimal zero value. Hydrogen adsorption on BLG causes substantial orbital hybridization accompanied by significant charge transfer. These findings highlight the significant impact of interlayer interactions between the P dopant and C vacancies on the catalytic properties of graphene layers.

2. Computational Methods

Spin-polarized density functional theory (DFT) calculations were conducted using the Vienna Ab initio Simulation Package.33 The calculations employed the projector-augmented wave method and the Perdew–Burke–Ernzerhof exchange–correlation functional. To include the effects of long-range van der Waals interactions, a semiclassical dispersion correction scheme (DFT-D3) was used.34 The plane-wave cutoff energy was set to 550 eV. In structural optimization, all atoms underwent full relaxation until the forces acting on each atom reached values less than 0.01 eV/Å, adhering to the energy convergence criterion of 1 × 10–7 eV. We constructed a 4 × 4 supercell for the defective BLG, utilizing a 5 × 5 × 1 mesh to sample the 2D Brillouin zone.

The configurations of BLG with C vacancies and P dopants fall into four groups: (1) configurations with C vacancies only, (2) a single graphene layer doped with P, (3) two graphene layers doped with P, and (4) BLG with interstitial P doping. These BLG configurations are denoted by the number of C vacancies and the presence of a P dopant, 0V-1V represents a BLG configuration with one layer without a C vacancy and the other layer with one C vacancy, while P2V-1V signifies a BLG configuration with one layer having P doped substitutionally into a double C vacancy and the other layer with a single C vacancy. Additionally, BLG with interstitial P doping is denoted by BLG-P.

The interlayer binding energy per atom (EB) was calculated using the equation: EB = (ED-BLG – E1L – E2L)/N, where ED-BLG, E1L, and E2L is the total energy of defective BLG (D-BLG) and first and second graphene layer of D-BLG, respectively, and N represents the total atom number in D-BLG.

The formation energy (EF) of defects in BLG is defined as

2.

Here, EBLG represent the total energy of BLG, Np and Nc represent the number of P dopants and carbon vacancies, respectively, μp is the energy of single phosphorus atom in its bulk structure, and μc is per carbon atom in pristine graphene.

The H adsorption energy, denoted as ΔEH, is defined as the difference between the total energy of the investigated BLG system with adsorbed H (Esys+H), the total energy of the corresponding BLG system without H (Esys), and half of the total energy of H2 (EH2). This is expressed by the formula

2.

The overall HER reaction process begin with the Volmer reaction, expressed as H+ + e– + * → H*. In this context, H+, e–, * and H* represent the proton, electron, the active site, and the adsorbed H, respectively. Subsequently, the Tafel (H* + H+ + e– → H2) or Heyrovsky reactions (H* + H* → H2) take place.35 The Volmer reaction, recognized as the most critical component in the HER, can be evaluated through the assessment of the H adsorption Gibbs free energy (ΔGH), calculated by ΔGH = ΔEH + ΔEZPE – TΔS. ΔEZPE denotes the zero-point energy correction for H2, while ΔS signifies the entropy difference between the adsorbed H atom and molecule in the gas phase under the standard conditions. This relationship is expressed by the formula

2.

where S0H2 is the entropy of a H2 gas molecule.36

3. Results and Discussion

The investigation commenced with an analysis of pristine BLG. Figure 1a displays the optimized structure of the AB-stacked BLG. The pristine BLG exhibits nonmagnetic properties, with an interlayer spacing of 3.38 Å and a calculated interlayer binding energy of −24 meV/per atom, consistent with previous studies.37−41 Subsequently, our investigation explores the impact of C vacancies and P doping on the structural and energetic properties. Figure 1b–f highlights the lowest energy configurations when introducing only C vacancy (XV-YV, see the Computational Methods section for notation), revealing a slight increase in layer spacing by only 0.02–0.06 Å compared to pristine BLG. It is noteworthy that the flat layered structure is maintained. The interlayer binding energy changes slightly compared to that of pristine BLG, by −2.97 to 2.3 meV. For P doping, the BLG structure depends highly on the number of C vacancies of the P-doped layer: BLG containing a layer with P doped substitutionally into a single C vacancy (denoted as P1V) undergoes substantial structural distortion, while BLG with P2V layers maintains the flat layered structure. This trend is because double C vacancy can offer ample space for P doping.

Figure 1.

Figure 1

Open in a new tab

Optimized atomic geometries for (a) pristine BLG and defective BLG with a P dopant or C vacancies: (b) 1V–0V, (c) 2V–0V, (d) 1V–1V, (e) 1V–2V, (f) 2V–2V; (g) P1V-0V, (h) P1V-1V, (i) P1V-2V, (j) P2V-0V, (k) P2V-1V, (l) P2V-2V, (m) P1V-P1V, (n) P2V-P1V, and (o) P2V-P2V configurations. Gray and blue circles represent C and P atoms, respectively. The locations of the C vacancies are marked by red dashed circles.

The concurrent presence of a P dopant and C vacancies induces variation in the interlayer spacing of BLG, ranging from 3.35 to 3.89 Å. Simultaneously, the interlayer binding energy changes from −23.74 to 21.62 meV, as detailed in Table 1. To provide a comprehensive understanding, we also explored interstitial P doping between layers of BLG (BLG-P), as illustrated in Figure S1. The BLG-P configuration demonstrates a substantial increase in the interlayer spacing, measuring 4.89 Å. The P dopant in the P1V layer tends to be located toward to the C vacancies of the other layer, as shown in Figure 1h,i. This is in sharp contrast to the case of P1V-0V, where P protrudes outward of BLG (Figure 1g). The trend is understandable because C vacancies possess dangling bonds that are known to often attract other atoms or molecules with unsaturated electron configurations, such as P.42,43

Table 1. C Interlayer Spacing (d), Magnetic Moment (M), Relative Interlayer Binding Energy Per Atom (ΔEB), Defective Formation Energy (EF), H Adsorption Energy (ΔEH) and Corresponding Free Energy (ΔGH) on Extralayer (Interlayer) in BLG Configurations with and without Defectsa.

  d (Å) M (μB) ΔEB (meV) EF (eV) ΔEH (eV) ΔGH (eV)
BLG 3.38 0 0   1.37 (1.5) 1.6 (1.74)
1V-0V 3.41 1.54 0.042 7.94 –1.17 –0.93
2V-0V 3.42 0 0.58 8.56 –0.8 –0.56
1V-1V 3.4 3.36 –2.97 15.73 –1.11 –0.87
1V-2V 3.44 1.52 2.3 16.46 –1.94 –1.7
2V-2V 3.41 0 1.23 17.12 –0.43 –0.19
P1V-0V 3.35 1 –0.37 2.83 –0.32 (1.2) –0.08 (1.44)
P1V-1V 3.56 0.71 –23.74 9.27 0.22 (−1.32) 0.46 (−1.08)
P1V-2V 3.54 0 21.62 11.22 –0.03 (−2.09) 0.21 (−1.85)
P2V-0V 3.4 0 0.82 4.1 –0.19 (−0.01) 0.05 (0.23)
P2V-1V 3.38 1.12 –1.87 11.92 –0.15 (−1.88) 0.09 (−1.64)
P2V-2V 3.42 0 3.12 12.7 –0.21 (−0.82) 0.03 (−0.58)
P1V-P1V 3.89 0 1.23 5.68 0.04 (1.4) 0.28 (1.64)
P1V-P2V 3.46 0 –13.66 6.17 0.21 (0.76) 0.45 (1.0)
P2V-P2V 3.38 0 10.01 8.16 –0.187 (−0.08) 0.053 (0.16)
BLG-P 4.89 1   4.76 0.04 (−1.26) 0.28 (−1.02)
Open in a new tab
a

For ΔEB, the value of the pristine BLG is set to zero.

The formation energy (EF) of defects in BLG was investigated to determine the stability of the defective BLG configurations. The calculated EF values, as listed in Table 1, for all of the considered defective BLG configurations are positive, indicating that the formation of defects in BLG is endothermic. The difference in the EF between 1V-0V and 2V-0V (7.94 and 8.56 eV, respectively) is only 0.62 eV, suggesting that inducing a second C vacancy is facilitated when a single vacancy exists. The formation energy of P1V-0V is 2.83 eV, considerably lower than those of P2V-0V (4.1 eV) and BLG-P (4.76 eV), indicating that substitutional P doping into a single C vacancy is energetically more favorable.

The distinct structural features arising from different doping and vacancy configurations led to varying magnetic properties. It is widely recognized that a C vacancy within a graphene layer leads to the emergence of magnetic moments, stemming from the presence of unpaired electrons surrounding the defect.44 The calculated magnetic moment in this work is 1.53 μB, consistent with previous studies.45 In contrast, within a graphene layer containing a double C vacancy, these localized magnetic moments vanish. This occurs due to an even number of dangling bonds, allowing for the pairing of all unpaired electrons. Consequently, the total magnetic moment becomes zero. For the BLG, our DFT calculations revealed similar magnetic trends. The magnetic moments for the 1V-0V (Figure 2a) and 1V-2V (Figure 2c) configurations are nearly identical, measuring 1.54 μB and 1.52 μB, respectively. Notably, the magnetic moment of 1V-1V is slightly higher than twice the value of a single 1V layer (3.36 μB), implying the effects of interlayer interactions.

Figure 2.

Figure 2

Open in a new tab

Spin density results for magnetic BLG configurations. The yellow isosurface is drawn at a density of 0.004 e Å–3.

We now investigate the effects of P doping on the magnetic properties of BLG. In the P1V-0V configuration (Figure 2d), three of the valence electrons in P form three σ bonds with the neighboring C atoms, while the fourth replaces the missing C and forms one π bond. The remaining one unpaired valence electron exists as a lone pair and reduces the magnetic moment to 1 μB. Interestingly, the interlayer interactions of the P-doped P1V and defective 1V layers result in an interlayer ferrimagnetic spin ordering with opposite spin orientations, as illustrated in Figure 2e. The total magnetic moment of P1V-1V is 0.71 μB, even lower than that of P1V-0V. When combined with the nonmagnetic 2V layer, the P1V layer also becomes nonmagnetic following the 2V layer. The unsaturated electrons of P are uniformly transferred to the four C atoms around the vacancies in the 2V layer. As a result, the total magnetic moment of P1V-2V is zero. Similarly, the P-doped 2V layer (P2V) is nonmagnetic with zero magnetic moment. The defect geometry of the 2V layer significantly affects the bonding configuration of the P dopant in the layer. The five valence electrons of P are saturated by forming four C–P σ bonds and one π bond, resulting in no unpaired electron. Therefore, in BLG with P2V-0V and P2V-2V configurations, the magnetic moments are zero. In contrast, P2V-1V exhibits a nonzero magnetic moment of 1.12 μB, mainly distributed around a single vacancy in the 1V layer (see Figure 2f).

The interlayer interactions between two P dopants across the BLG can affect the magnetic ordering. For example, although P1V alone has a magnetic moment of 1 μB, BLG with both P1V-P2V and P1V-P1V configurations exhibit zero magnetic moments. When a P1V layer with unsaturated electrons encounters a structurally symmetrical P2V layer, the resulting nonmagnetic behavior aligns with that of P1V-2V. Nevertheless, in the scenario where two P1V layers with unsaturated electrons converge, the symmetrical configuration between the layers results in the charge transfer between P atoms to form a P–P bond. This interlayer interaction renders the P1V-P1V system showcasing a lack of magnetism.

In contrast, the flat P2V-P2V configuration, distinguished by the absence of redundant electrons, inherently exhibits a nonmagnetic nature. However, in the case of BLG-P, the magnetic moment with 1 μB becomes localized around the P atom, as illustrated in Figure S2. This phenomenon is attributed to the presence of one excess unsaturated electron associated with the P atom.

In our investigation of H adsorption on various defective BLG configurations, we further observe a significant enhancement in adsorption strength compared to pristine BLG. Specifically, in the case of interstitial P doping (BLG-P), H atoms exhibit the formation of a strong C–P bond with the interstitially doped P, as indicated by the H adsorption energy (ΔEH) of −1.26 eV. Furthermore, interstitial P doping enhances H adsorption on BLG surfaces, with a calculated ΔEH of 0.04 eV—significantly lower than the value (1.5 eV) observed in pristine BLG. In BLG with C vacancies, H atoms tend to adsorb onto vacancies within the graphene layer, as illustrated in Figure S3. All configurations display robust C–H bonds, with ΔEH ranging from −1.94 to −0.43 eV.

When P doping is introduced into the C vacancies of BLG, the H adsorption properties are examined in both the outer surface and inner surface. In cases where the P-doped layer coexists with the vacancy layer, the adsorption of H atoms on the inner surfaces tends to occur at the vacancy sites within the vacancy layer, as illustrated in Figure S4. The ΔEH values for P1V-1V, P1V-2V, P2V-1V, and P2V-2V are −1.32, −2.09, −1.88, and −0.82 eV, respectively. In contrast, for the configurations consisting solely of P-doped or pristine graphene layers (P1V-0V, P2V-0V, P1V-P1V, P2V-P2V, P1V-P2V), the inner-surface H adsorption is relatively weak, as reflected by ΔEH ranging from 1.4 to −0.08 eV. Interestingly, we observe that in the BLG configurations involving a P-doped layer with P1V, where the P dopant is located toward the other layer, the H atoms adsorbed on the outer surface tend to form tilted C–H bonds with the C atoms around P (see Figure 3b,c,g). On the contrary, in the other configurations, the adsorbed H atoms adhere to the P top site of the P2V layer (Figure 3d–f,h,i). Additionally, H atoms adsorbed on the outer surface exhibit relatively weakened adsorption compared to adsorption on the inner surface in configurations containing both vacancy and P-doped layers. More importantly, for the P1V-0V, P2V-0V, P2V-1V, P2V-2V, and P2V-P2V configurations featuring H adsorption at the P top site, the corresponding adsorption Gibbs free energy values (ΔGH) range from −0.08 to 0.05 eV, which closely approaches the optimal value (ΔGH = 0 eV) for HER, as shown in Table 1 and Figure 4. These results suggest their potential as promising hydrogen catalysts.

Figure 3.

Figure 3

Open in a new tab

Optimized hydrogen adsorption configurations on the surface of (a) P1V-0V, (b) P1V-1V, (c) P1V-2V, (d) P2V-0V, (e) P2V-1V, and (f) P2V-2V, (g) P1V-P1V, (h) P2V-P1V, and (i) P2V-P2V. Gray, blue, and pink circles represent C, P, and H atoms, respectively.

Figure 4.

Figure 4

Open in a new tab

Calculated hydrogen adsorption Gibbs free energy values (ΔGH) for the configurations with concurrent presence of the P dopant and C vacancies. The ΔGH in P1V-0V, P2V-0V, P2V-1V, P2V-2V, and P2V-P2V are close to 0 eV, where H adsorbed at the top site of P.

Given that the H atoms adsorbed on the inner surface in P2V-1V and P2V-2V form robust C–H bonds with C in the vacancy layer, our subsequent investigation delves into the adsorption of the second H atom. In P2V-1V, the ΔEH of the second H atom on the P top site (outer surface) is −0.11 eV, slightly higher than that (−0.15 eV) of the first H atom. On the other hand, the adsorption of the second H atom on the inner surface causes substantial structural distortion in the 1V layer due to the repulsive interactions between the two adsorbed H, as shown in Figure 5a. The calculated ΔEH is −0.98 eV, indicating that the inner-surface adsorption of the second H atom is stronger than that of the outer surface. For P2V-2V, the corresponding ΔEH of the outer (inner) surface is 0.17 (−0.21) eV. As a result, ΔGH closest to the optimal zero value is 0.03 eV for P2V-2V (inter surface) and 0.13 eV for P2V-1V (outer surface), suggesting that these configurations can still provide activity sites for other H atoms in the presence of already adsorbed H in the vacancy graphene layer.

Figure 5.

Figure 5

Open in a new tab

Adsorption configuration of the two H atoms in P2V-1V and P2V-2V, where the first H atom adsorbs on the vacancy sites and the second H atom adsorbs on the (a) inner and (b) outer surfaces of P2V-1V and the (c) inner and (d) outer surfaces of P2V-2V.

Our subsequent investigation reveals that distinct H adsorption behaviors of the different defect configurations are intricately linked to the electron distribution in the occupied states near the Fermi level. The partial electron density plot in Figure 6 indicates that, in comparison, the 1V configuration possesses the densest electron density around the vacancy site, followed by 2V, while the charge around the 0V configuration is nearly negligible. The abundant electron distribution serves as an active site, attracting hydrogen atoms for reaction and forming strong C–H bonds. Under the influence of P, the P1V layer in various defective BLG configurations manifests distinct charge distributions. In the P1V-0V configuration, electrons are primarily distributed around the P dopant. In the P1V-1V configuration, the P dopant located toward the inner side of BLG induces charge localization at the C atoms around P and the C vacancy in the 1V layer. The electrons are more localized compared with those of P1V-2V. The localized electron distribution results in the formation of robust C–H bonds. The P2V layer shows 4-fold symmetric charge distribution that is primarily concentrated around the P dopant, which accounts for the fact that P is the most favorable site for the H adsorption. In contrast, P2V-0V, P2V-1V, and P2V-2V configurations display distinguishing electron distribution mainly within the vacancy layer. As the 1V layer possesses a higher electron density, it demonstrates the most robust H adsorption in P2V-1V. Overall, the notable electron distribution on C atoms surrounding the vacancy results in the formation of strong C–H bonds, aligning with the calculated ΔEH.

Figure 6.

Figure 6

Open in a new tab

Partial electron densities obtained from the states within the energy range of EF −1 eV < E < EF. The iso-surfaces were taken at charge densities of 0.008 e Å–3.

The P1V-P1V and P1V-P2V configurations have less electronic states near the Fermi level, which can account for their higher ΔEH. The charge distribution in the P2V layer in P2V-P2V is nearly identical to that in P2V-0V. Consequently, these two configurations demonstrate nearly identical hydrogen adsorption behavior on the P2V surface, including the preferred H adsorption site and adsorption strength.

The outstanding catalytic activities of P2V-0V and P2V-2V can be interpreted by significant charge transfer. The differential charge density plots in Figure 7a,b highlight considerable charge transfer in P2V-0V and P2V-2V across BLG. Upon H adsorption, substantial charge transfer occurs between the P dopant and adsorbed H atom, as shown in Figure 7c,d. Our Bader charge analysis reveals that in P2V-0V (P2V-2V), the H atom gains 0.4e (0.335e), while the P dopant loses 0.23e (0.16e). The appropriate amount of charge transfer between the H and P atoms results in a moderate H adsorption strength, which is desirable for the HER.

Figure 7.

Figure 7

Open in a new tab

Differential charge density plots for (a, c) P2V-0V and (b, d) P2V-2V before and after H adsorption. The yellow and blue iso-surfaces are drawn at the density of 0.0003 and −0.0003 e Å–3 for (a, b) and 0.0004 and −0.0004 e Å–3 for (c, d), respectively.

The density of states (DOS) reveals that the states near the Fermi level originate primarily from C vacancies and P dopants in P2V-0V and P2V-2V. These electronic structures facilitate orbital hybridization with H atoms, leading to the adsorption preference on the vacancy and dopant. In contrast to the narrow-gap semiconducting characteristics of pristine BLG (Figure 8a), the P2V-0V configuration exhibits an upward shift of the valence band maximum, increasing the DOS at EF (Figure 8c). The calculated band gaps for pristine BLG and P2V-0V are 0.019 and 0.23 eV, respectively. In contrast, the P2V-2V configuration is completely metallic (Figure 8e). The H adsorption gives rise to a prominent peak at the EF in BLG (Figure 8b), signifying the presence of highly localized electrons. This suggests a subtle orbital hybridization between the adsorbed H atoms and the C atom of the adsorption site. In contrast, P2V-0V and P2V-2V show no prominent peak at the Fermi level due to substantial orbital hybridization of P and H. As depicted in Figure 8d,f, the projected densities of states of P and H align well with each other, indicating s–p orbital hybridization. This interaction weakens the electronic hybridization between P and C, resulting in a downward shift of the C states near the Fermi level.

Figure 8.

Figure 8

Open in a new tab

Calculated density of states for (a) BLG, (b) H adsorption on BLG, (c) P2V-0V, (d) H adsorption on P2V-0V, (e) P2V-2V, and (f) H adsorption on P2V-2V. The Fermi energy level is set to be 0 eV.

4. Conclusions

In summary, we have investigated the collaborative effects of P doping and C vacancies on the structural, magnetic, electronic, and catalytic properties of BLG. The findings reveal that the P dopants within the P1V layer tend to be situated toward the inner layer of BLG, while the P2V layer maintains a flat configuration. The distinct defective configurations result in varied electron distribution, giving rise to intriguing magnetic behaviors. Notably, the electron distribution near the Fermi level has a crucial impact on H adsorption behavior. More importantly, P1V-0V, P2V-0V, P2V-1V, P2V-2V, and P2V-P2V exhibited ΔGH close to 0, with H adsorb at the top site of P, signifying excellent catalytic activity for HER. Further investigations demonstrate that the charge transfer between H atoms and adsorbed P atoms plays a facilitating role in the catalytic reaction. It is also noteworthy that, in the scenario where a H atom initially occupies the vacancy site in the 2V layer of P2V-1V and P2V-2V, the catalytic activity remains excellent for the second H adsorption. These findings suggest that structural changes by the interactions of the P dopant and C vacancies also modify the electronic distribution and spin ordering significantly. This work offers implications for the application of graphene layers in HER catalysis.

Acknowledgments

This work was supported by the National Natural Science Foundation of China (Grants No. 11874044 and No. 52350610260). The calculations were performed on TianHe-1(A) at the NSCC in Tianjin.

Supporting Information Available

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acsomega.4c00495.

  • The optimized configuration and spin density results of BLG-P and H adsorption configurations in BLG with the vacancy layer (PDF)

The authors declare no competing financial interest.

Supplementary Material

ao4c00495_si_001.pdf (226.4KB, pdf)

References

  1. Tiwari S. K.; Sahoo S.; Wang N.; Huczko A. Graphene research and their outputs: Status and prospect. J. Sci.: Adv. Mater. Devices 2020, 5 (1), 10–29. 10.1016/j.jsamd.2020.01.006. [DOI] [Google Scholar]
  2. Edwards R. S.; Coleman K. S. Graphene synthesis: relationship to applications. Nanoscale 2013, 5 (1), 38–51. 10.1039/C2NR32629A. [DOI] [PubMed] [Google Scholar]
  3. Novoselov K. S.; Geim A. K.; Morozov S. V.; Jiang D.; Katsnelson M. I.; Grigorieva I. V.; Dubonos S. V.; Firsov A. A. Two-dimensional gas of massless Dirac fermions in graphene. Nature 2005, 438 (7065), 197–200. 10.1038/nature04233. [DOI] [PubMed] [Google Scholar]
  4. Wang G. X. A series of two-dimensional carbon allotropes with Dirac cone structure. Phys. Chem. Chem. Phys. 2023, 25 (23), 15815–15821. 10.1039/D3CP01293J. [DOI] [PubMed] [Google Scholar]
  5. Teng C.; Xie D.; Wang J. F.; Yang Z.; Ren G. Y.; Zhu Y. Ultrahigh Conductive Graphene Paper Based on Ball-Milling Exfoliated Graphene. Adv. Funct. Mater. 2017, 27 (20), 1700240 10.1002/adfm.201700240. [DOI] [Google Scholar]
  6. Miller D. L.; Kubista K. D.; Rutter G. M.; Ruan M.; de Heer W. A.; First P. N.; Stroscio J. A. Observing the Quantization of Zero Mass Carriers in Graphene. Science 2009, 324 (5929), 924–927. 10.1126/science.1171810. [DOI] [PubMed] [Google Scholar]
  7. Hirsch A. The era of carbon allotropes. Nat. Mater. 2010, 9 (11), 868–871. 10.1038/nmat2885. [DOI] [PubMed] [Google Scholar]
  8. Chen L. X.; Chen Z. W.; Wang Y.; Yang C. C.; Jiang Q. Design of Dual-Modified MoS2 with Nanoporous Ni and Graphene as Efficient Catalysts for the Hydrogen Evolution Reaction. ACS Catal. 2018, 8 (9), 8107–8114. 10.1021/acscatal.8b01164. [DOI] [Google Scholar]
  9. Wu H. H.; Huang H.; Zhong J.; Yu S.; Zhang Q.; Zeng X. C. Monolayer triphosphates MP(3) (M = Sn, Ge) with excellent basal catalytic activity for hydrogen evolution reaction. Nanoscale 2019, 11 (25), 12210–12219. 10.1039/C9NR03255J. [DOI] [PubMed] [Google Scholar]
  10. Denis P. A.; Pereyra Huelmo C. Structural characterization and chemical reactivity of dual doped graphene. Carbon 2015, 87, 106–115. 10.1016/j.carbon.2015.01.049. [DOI] [Google Scholar]
  11. Zhang J. T.; Dai L. M. Nitrogen, Phosphorus, and Fluorine Tri-doped Graphene as a Multifunctional Catalyst for Self-Powered Electrochemical Water Splitting. Angew. Chem., Int. Ed. 2016, 55 (42), 13296–13300. 10.1002/anie.201607405. [DOI] [PubMed] [Google Scholar]
  12. Kumatani A.; Miura C.; Kuramochi H.; Ohto T.; Wakisaka M.; Nagata Y.; Ida H.; Takahashi Y.; Hu K.; Jeong S.; Fujita J. I.; Matsue T.; Ito Y. Chemical Dopants on Edge of Holey Graphene Accelerate Electrochemical Hydrogen Evolution Reaction. Adv. Sci. 2019, 6 (10), 1900119 10.1002/advs.201900119. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Wu F. C.; Das Sarma S. Ferromagnetism and superconductivity in twisted double bilayer graphene. Phys. Rev. B 2020, 101 (15), 155149 10.1103/PhysRevB.101.155149. [DOI] [Google Scholar]
  14. Hsu Y. T.; Wu F. C.; Das Sarma S. Topological superconductivity, ferromagnetism, and valley-polarized phases in moire systems: Renormalization group analysis for twisted double bilayer graphene. Phys. Rev. B 2020, 102 (8), 085103 10.1103/PhysRevB.102.085103. [DOI] [Google Scholar]
  15. Ji K. M.; Han J. H.; Hirata A.; Fujita T.; Shen Y.; Ning S. C.; Liu P.; Kashani H.; Tian Y.; Ito Y.; Fujita J.; Oyama Y. Lithium intercalation into bilayer graphene. Nat. Commun. 2019, 10, 1603421 10.1038/s41467-018-07942-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Yang S.; Fecher S.; Wang Q. X.; Kühne M.; Smet J. H. Device level reversible potassium intercalation into bilayer graphene. 2D Mater. 2022, 9 (2), 025020 10.1088/2053-1583/ac58a1. [DOI] [Google Scholar]
  17. Li J. R.; Yang S. B.; Zhang X.; Zhang C. X. Y. Exploring the effect of layer spacing of graphene for lithium single-atom diffusion using first principles calculations. Comput. Theor. Chem. 2022, 1217, 113937 10.1016/j.comptc.2022.113937. [DOI] [Google Scholar]
  18. Denis P. A. Band gap opening of monolayer and bilayer graphene doped with aluminium, silicon, phosphorus, and sulfur. Chem. Phys. Lett. 2010, 492 (4–6), 251–257. 10.1016/j.cplett.2010.04.038. [DOI] [Google Scholar]
  19. Denis P. A.; Iribarne F. The effect of the dopant nature on the reactivity, interlayer bonding and electronic properties of dual doped bilayer graphene. Phys. Chem. Chem. Phys. 2016, 18 (35), 24693–703. 10.1039/C6CP02481E. [DOI] [PubMed] [Google Scholar]
  20. Denis P. A. Theoretical characterization of codoped bilayer graphene. Comput. Theor. Chem. 2023, 1221, 114035 10.1016/j.comptc.2023.114035. [DOI] [Google Scholar]
  21. Denis P. A. When noncovalent interactions are stronger than covalent bonds: Bilayer graphene doped with second row atoms, aluminum, silicon, phosphorus and sulfur. Chem. Phys. Lett. 2011, 508 (1–3), 95–101. 10.1016/j.cplett.2011.04.018. [DOI] [Google Scholar]
  22. Hu R.; Li Y.; Wang F.; Shang J. Rational prediction of multifunctional bilayer single atom catalysts for the hydrogen evolution, oxygen evolution and oxygen reduction reactions. Nanoscale 2020, 12 (39), 20413–20424. 10.1039/D0NR05202G. [DOI] [PubMed] [Google Scholar]
  23. Xiao B. B.; Zhang Z.; Yu L. B.; Huang Q. Y.; Wu J.; Song E. H.; Wang L. L. Boosting oxygen reduction electrocatalysis of graphene-based bilayer heterojunction. Surfaces and Interfaces 2022, 33, 102232 10.1016/j.surfin.2022.102232. [DOI] [Google Scholar]
  24. Pham N. N. T.; Nguyen V. K. T.; Guo H.; Lee S. G. Influence of phosphorus-doped bilayer graphene configuration on the oxygen reduction reaction in acidic solution. Carbon 2023, 210, 118012 10.1016/j.carbon.2023.118012. [DOI] [Google Scholar]
  25. Wang H.; Wang Q.; Cheng Y.; Li K.; Yao Y.; Zhang Q.; Dong C.; Wang P.; Schwingenschlogl U.; Yang W.; Zhang X. X. Doping monolayer graphene with single atom substitutions. Nano Lett. 2012, 12 (1), 141–144. 10.1021/nl2031629. [DOI] [PubMed] [Google Scholar]
  26. Ramasse Q. M.; Seabourne C. R.; Kepaptsoglou D. M.; Zan R.; Bangert U.; Scott A. J. Probing the Bonding and Electronic Structure of Single Atom Dopants in Graphene with Electron Energy Loss Spectroscopy. Nano Lett. 2013, 13 (10), 4989–4995. 10.1021/nl304187e. [DOI] [PubMed] [Google Scholar]
  27. Tripathi M.; Markevich A.; Böttger R.; Facsko S.; Besley E.; Kotakoski J.; Susi T. Implanting Germanium into Graphene. ACS Nano 2018, 12 (5), 4641–4647. 10.1021/acsnano.8b01191. [DOI] [PubMed] [Google Scholar]
  28. Ullah S.; Shi Q.; Zhou J.; Yang X.; Ta H. Q.; Hasan M.; Ahmad N. M.; Fu L.; Bachmatiuk A.; Rümmeli M. H. Advances and Trends in Chemically Doped Graphene. Adv. Mater. Interfaces 2020, 7 (24), 105021 10.1002/admi.202000999. [DOI] [Google Scholar]
  29. Rao Y.; Yuan M.; Luo F.; Wang Z.; Li H.; Yu J.; Chen X. One-step laser fabrication of phosphorus-doped porous graphene electrodes for high-performance flexible microsupercapacitor. Carbon 2021, 180, 56–66. 10.1016/j.carbon.2021.05.001. [DOI] [Google Scholar]
  30. Feng L.; Qin Z.; Huang Y.; Peng K.; Wang F.; Yan Y.; Chen Y. Boron-, sulfur-, and phosphorus-doped graphene for environmental applications. Sci. Total Environ. 2020, 698, 134239 10.1016/j.scitotenv.2019.134239. [DOI] [PubMed] [Google Scholar]
  31. Zhan X.; Tong X.; Gu M.; Tian J.; Gao Z.; Ma L.; Xie Y.; Chen Z.; Ranganathan H.; Zhang G.; Sun S. Phosphorus-Doped Graphene Electrocatalysts for Oxygen Reduction Reaction. Nanomaterials 2022, 12 (7), 1141 10.3390/nano12071141. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Zhang F.; Zhang Q. J.; Chen Y. C.; Xu L. Y.; Li Z. L.; Wang Q. T.; Zhang Q. F.; Li X. N. Advances and mechanistic insights in the catalytic semi-hydrogenation of acetylene over non-metallic catalysts. Appl. Catal., A 2023, 667, 119447 10.1016/j.apcata.2023.119447. [DOI] [Google Scholar]
  33. Kresse G.; Furthmuller J. Efficiency of ab-initio total energy calculations for metals and semiconductors using a plane-wave basis set. Comput. Mater. Sci. 1996, 6 (1), 15–50. 10.1016/0927-0256(96)00008-0. [DOI] [Google Scholar]
  34. Grimme S.; Antony J.; Ehrlich S.; Krieg H. A consistent and accurate ab initio parametrization of density functional dispersion correction (DFT-D) for the 94 elements H-Pu. J. Chem. Phys. 2010, 132 (15), 154104 10.1063/1.3382344. [DOI] [PubMed] [Google Scholar]
  35. Huang H. J.; Yan M. M.; Yang C. Z.; He H. Y.; Jiang Q. G.; Yang L.; Lu Z. Y.; Sun Z. Q.; Xu X. T.; Bando Y.; Yamauchi Y. Graphene Nanoarchitectonics: Recent Advances in Graphene-Based Electrocatalysts for Hydrogen Evolution Reaction. Adv. Mater. 2019, 31 (48), 1903415 10.1002/adma.201903415. [DOI] [PubMed] [Google Scholar]
  36. Nørskov J. K.; Bligaard T.; Logadottir A.; Kitchin J. R.; Chen J. G.; Pandelov S.; Norskov J. K. Trends in the exchange current for hydrogen evolution. J. Electrochem. Soc. 2005, 152 (3), J23–J26. 10.1149/1.1856988. [DOI] [Google Scholar]
  37. Chakarova-Käck S. D.; Schröder E.; Lundqvist B. I.; Langreth D. C. Application of van der Waals density functional to an ex-tended system: Adsorption of benzene and naphthalene on graphite. Phys. Rev. Lett. 2006, 96 (14), 146107 10.1103/PhysRevLett.96.146107. [DOI] [PubMed] [Google Scholar]
  38. Coello-Fiallos D.; Tene T.; Guayllas J. L.; Haro D.; Haro A.; Vacacela Gomez C. DFT comparison of structural and electronic properties of graphene and germanene: monolayer and bilayer systems. Mater. Today: Proc. 2017, 4 (7), 6835–6841. 10.1016/j.matpr.2017.07.011. [DOI] [Google Scholar]
  39. Hess P. Thickness of elemental and binary single atomic monolayers. Nanoscale Horiz. 2020, 5 (3), 385–399. 10.1039/C9NH00658C. [DOI] [PubMed] [Google Scholar]
  40. Chen X.; Tian F.; Persson C.; Duan W.; Chen N. X. Interlayer interactions in graphites. Sci. Rep. 2013, 3, 3046 10.1038/srep03046. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Herrero C. P.; Ramírez R. Elastic properties and mechanical stability of bilayer graphene: molecular dynamics simulations. Eur. Phys. J. B 2023, 96 (11), 147 10.1140/epjb/s10051-023-00616-w. [DOI] [Google Scholar]
  42. Rocha A. R.; Padilha J. E.; Fazzio A.; da Silva A. J. R. Transport properties of single vacancies in nanotubes. Phys. Rev. B 2008, 77 (15), 153406 10.1103/PhysRevB.77.153406. [DOI] [Google Scholar]
  43. Qu Y. T.; Wang L. G.; Li Z. J.; Li P.; Zhang Q. H.; Lin Y.; Zhou F. Y.; Wang H. J.; Yang Z. K.; Hu Y. D.; Zhu M. Z.; Zhao X. Y.; Han X.; Wang C. M.; Xu Q.; Gu L.; Luo J.; Zheng L. R.; Wu Y. E. Ambient Synthesis of Single-Atom Catalysts from Bulk Metal via Trap-ping of Atoms by Surface Dangling Bonds. Adv. Mater. 2019, 31 (44), 1904496 10.1002/adma.201904496. [DOI] [PubMed] [Google Scholar]
  44. Centrone A. Probing the paramagnetic interactions between the unpaired electronic spins of carbon atoms and the nuclear spins of hydrogen molecules with Raman spectroscopy. J. Raman Spectrosc. 2011, 42 (12), 2165–2167. 10.1002/jrs.2983. [DOI] [Google Scholar]
  45. Valencia A. M.; Caldas M. J. Single vacancy defect in graphene: Insights into its magnetic properties from theoretical modeling. Phys. Rev. B 2017, 96 (12), 125431 10.1103/PhysRevB.96.125431. [DOI] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

ao4c00495_si_001.pdf (226.4KB, pdf)

Articles from ACS Omega are provided here courtesy of American Chemical Society

RESOURCES