A Theoretical Study of the Reactive Mechanisms of Alkali Metal Doped Ni-Based Oxygen Carrier During Chemical Looping Combustion

Abstract

Chemical looping combustion (CLC) is a promising technology for CO₂ capture, with the performance of the system largely dependent on the oxygen carrier. Although Ni-based carriers have been extensively investigated, their practical application is still constrained by inadequate reactivity. This study investigated the doping of alkali metals (Li, Na, K) into NiO to improve its performance in CLC. Through density functional theory calculations, the structural, electronic, and reactivity of doped NiO surfaces were systematically analyzed. Results reveal that doping induces lattice expansion and enhances CO adsorption, with adsorption energies strengthening to −0.53 eV for Li, −0.46 eV for Na, and −0.36 eV for K. Furthermore, alkali metal doping significantly reduces the energy barrier for CO₂ formation from 2.12 eV on pure NiO to 0.73 eV, 0.80 eV, and 0.99 eV on Li-, Na-, and K-doped surfaces, respectively. Oxygen vacancy formation energy also decreases from 3.60 eV to as low as 2.90 eV for K-doping, indicating markedly improved oxygen activity. Electronic structure analysis confirms that doping facilitates electron transfer and stabilizes key reaction intermediates. In conclusion, alkali metal doping substantially enhances the redox activity of NiO, providing an effective strategy for developing high-performance oxygen carriers in CLC.

1. Introduction

Carbon emissions from fossil fuel consumption represent a major source of greenhouse gases (GHGs), and the resulting greenhouse effect has become a critical global environmental concern [1]. The development of efficient carbon capture technologies is essential to mitigate global CO₂ emissions [2]. Chemical looping combustion (CLC) has emerged as a promising approach for achieving complete carbon capture from fuels with minimal energy penalty and at low cost [3,4]. As illustrated in Figure 1, the CLC process comprises two main reaction stages. In the fuel reactor (FR), the fuel (denoted as C_nH_2mO_p) is oxidized by a metal oxide oxygen carrier, producing a highly concentrated stream of CO₂ and H₂O. In the air reactor (AR), the reduced oxygen carrier is re-oxidized by air. For Ni-based carriers, this two-reactor design is essential to address their specific material challenges. The FR must be operated under well-controlled conditions to fully exploit the high oxygen capacity of NiO while minimizing carbon deposition caused by excessive reduction. Conversely, the AR must efficiently re-oxidize the reduced Ni while managing exothermic heat to mitigate sintering, a primary deactivation mechanism for Ni-based materials. The regenerated carrier is then separated and recycled to the fuel reactor to initiate the next cycle. A key advantage of CLC lies in its inherent separation of CO₂ during the process, thereby significantly reducing the energy penalty and cost associated with carbon capture.

$${\mathrm{C}}_{\mathrm{n}}{\mathrm{H}}_{2\mathrm{m}}{\mathrm{O}}_{\mathrm{p}}+(2\mathrm{n}+\mathrm{m}-\mathrm{p}){\mathrm{MeO}}_{\mathrm{x}}\to (2\mathrm{n}+\mathrm{m}-\mathrm{p}){\mathrm{MeO}}_{\mathrm{x}-1}+{\mathrm{nCO}}_2+{\mathrm{mH}}_2\mathrm{O}$$

(1)

$${\mathrm{O}}_2+2{\mathrm{MeO}}_{\mathrm{x}-1}\to 2{\mathrm{MeO}}_{\mathrm{x}}$$

(2)

CLC has garnered significant interest due to its inherent advantages in efficient CO₂ separation. By employing oxygen carriers to directly transport oxygen to the fuel, CLC avoids nitrogen dilution and enables the production of a highly concentrated CO₂ stream [5]. Moreover, since fuel combustion occurs in the absence of air and oxygen carrier re-oxidation takes place at relatively low temperatures, the process significantly suppresses the formation of nitrogen oxides [6,7]. The performance of the oxygen carrier is critical to the successful implementation of CLC [8,9]. To date, over 500 different oxygen carrier materials have been investigated, including perovskites, natural ores, and transition metal oxides [10,11,12,13,14,15,16]. Among them, transition metal oxides such as Ni, Co, Fe, Mn, and Cu have been extensively studied as promising candidates due to their favorable redox properties and structural stability [14,15,16,17,18,19,20,21,22]. In particular, Ni-based oxygen carriers have attracted considerable attention owing to their excellent performance at high temperatures (900–1100 °C), satisfactory reaction kinetics, and remarkable catalytic efficiency in cleaving C–C and C–H bonds [23,24]. However, the practical application of pure NiO in CLC is hampered by several limitations, including carbon deposition and relatively low reactivity [25]. For instance, Cheng et al. reported a gradual decline in methane combustion efficiency over successive redox cycles [26], highlighting the need for further improvement in the performance of Ni-based oxygen carriers.

Doping modification is a widely employed strategy for enhancing the reactivity of functional materials. Studies have shown that the incorporation of different metal elements can distinctly alter material performance. For instance, Zhang et al. [27] reported that iron doping improves the reactivity of nickel-based oxides, whereas cobalt doping exerts a detrimental effect. Yuan et al. [28] systematically investigated the mechanism of hydrogen oxidation on copper-doped iron oxides, revealing that Cu doping facilitates the release of lattice oxygen in Fe₂O₃. Similarly, Wang et al. [29] demonstrated that Zr-doped CuO exhibits superior oxygen release capability compared to pure CuO. Furthermore, Qin et al. [30] found that introducing a low concentration of La into Fe-based oxides not only significantly enhances the reactivity of Fe₂O₃ with CO. This improvement is attributed to the reduced activation energy barriers for C–H and C–O bond cleavage during reactions with carbon-containing gases such as CH₄ and CO.

In the context of oxygen carriers, elemental doping has been extensively investigated as a means to improve their redox performance, with particular focus on alkali and alkaline earth metals. For instance, Imtiaz et al. [31] reported that potassium ions increase the concentration of active copper oxide species. Bao et al. [32] demonstrated that potassium doping significantly improves the porosity and specific surface area of carrier materials. Yin et al. [33] further showed that doping pyrite-slag-derived oxygen carriers with alkali/alkaline earth metals (Na, Mg, Ca, K) markedly enhances their reactivity in chemical looping processes. Among these dopants, potassium was identified as the most effective promoter, substantially reducing the oxygen vacancy formation energy from 4.31 eV to 0.49 eV, thereby facilitating deep reduction in the oxygen carriers. Liu et al. [34] synthesized Fe-based oxygen carriers doped with alkali metals (Na, K, Cs) and evaluated their reactivity and stability over 50 isothermal redox cycles in a fixed-bed reactor. The results indicated that alkali metal doping significantly improved long-term stability and mitigated carbon deposition, with K-doped samples exhibiting the best performance. Huang et al. [35] investigated the effect of Na addition (1–10 wt%) on the redox properties of Fe₂O₃/Al₂O₃ oxygen carriers. It was observed that 1 wt% Na loading significantly enhanced thermal stability, attrition resistance, and reduction performance. Ma et al. [15] confirmed that potassium doping effectively restores the redox activity of deactivated Fe₂O₃/Al₂O₃ by promoting the formation of a porous structure and increasing the surface oxygen vacancy concentration.

The survey of the existing literature indicates that alkali metal doping represents a promising strategy for enhancing the performance of Ni-based oxygen carriers. Nevertheless, a systematic understanding of its role in CLC—including its influence on redox behavior, structural evolution—remains incomplete. Thus, there is a clear need for dedicated investigations into the effects of alkali metal doping on Ni-based oxygen carriers under CLC conditions, in order to elucidate the underlying promotion mechanisms and assess their potential for practical application. Current investigations into chemical reaction mechanisms primarily rely on two approaches: experimental observation and theoretical simulation. While experimental techniques are effective for tracking macroscopic structural changes in oxygen carriers during reactions, they face considerable limitations in probing the microscopic mechanisms of CLC. Specifically, experimental characterization struggles to capture transient reaction intermediates and monitor dynamic reaction pathways over extremely short timescales [36,37]. In contrast, computational simulations have demonstrated distinct advantages in elucidating material surface structures, properties, and underlying reaction mechanisms. In particular, first-principles calculations can provide atomic-scale insights and energetic information that are often inaccessible through experiments, thereby enabling a deeper understanding of material structures, property predictions, and reaction pathways [38,39].

In recent years, first-principles calculations have been extensively employed to investigate a variety of oxygen carriers, including NiO, Fe₂O₃, and perovskite-type materials [22,40,41,42,43,44,45,46]. In addition, density functional theory (DFT) has emerged as a powerful computational approach for exploring the structure and reactivity of Ni-based oxygen carriers, offering valuable atomistic insights into their properties and reaction mechanisms. For example, Feng et al. [46] employed DFT calculations to elucidate the role of support materials in enhancing the reactivity of Ni-based oxygen carriers for chemical looping reforming. Their study revealed that the ZrO₂ support significantly reduces the energy barrier of the rate-limiting CO formation step and weakens the Ni–O bond, thereby making NiO/ZrO₂ a more reactive oxygen carrier compared to NiO/MgAl₂O₄. Xing et al. [47] utilized DFT to demonstrate that the NiAl₂O₄ support weakens the Ni-C bond via strong metal-support interactions, thereby reducing the activation energy barrier for CH₄ dehydrogenation and enhancing the reaction efficiency of Ni-based oxygen carrier in CLC. Yuan et al. [48] applied DFT calculations to elucidate the elementary reaction mechanisms of syngas combustion on Ni-based oxygen carriers, revealing that H₂ decomposition is the rate-determining step.

In this work, the DFT calculations was employed to systematically investigate the effects of alkali metal (Li, Na, K) doping on the reactivity and underlying mechanisms of Ni-based oxygen carriers in CLC. The crystal structure models of NiO with different alkali metal dopants were first constructed; the influence of doping on CO adsorption behavior was then analyzed; the CO oxidation reaction pathway was subsequently simulated; and finally, the oxygen vacancy formation energy was calculated to evaluate the doping-induced enhancement of reducibility. The findings provide fundamental insights for designing effective oxygen carriers with superior performance.

2. Results and Discussion

2.1. Surface Structure of Alkali Metal Doped NiO

All calculations in this study are based on the conventional cubic rock-salt phase of NiO. To investigate the structural influence of different alkali metal dopants (Li, Na, K) on NiO, surface models were constructed by substituting a central Ni atom in the first layer of the NiO(001) surface with Li, Na, or K. To preserve the net zero magnetic moment of the original antiferromagnetic structure, the doped atoms were assigned magnetic moments consistent with the host Ni atoms. All models—pristine NiO and each doped system—were optimized using the DFT + U approach, with the resulting structures shown in Figure 2. Since the reactivity of oxygen carriers in CLC is governed by surface bonding characteristics, the effect of doping on the bond parameters of surface oxygen atoms was examined. The binding energies of Li, Na, and K dopants were calculated to assess the relative stability of incorporation at Ni sites.

Table 1. The average bond length between the M atom and its neighboring O atom, and the corresponding binding energy of alkali metal doped NiO surfaces.

Model	Ebind/eV	LO–M/Å
NiO	—	2.085
Li-NiO	−1.67	2.146
Na-NiO	0.17	2.249
K-NiO	0.15	2.321

summarizes the average bond lengths (L_(O–M)) between the dopant site and adjacent oxygen atoms, together with the corresponding binding energies for each dopant.

As summarized in Table 1, the calculated binding energies for Li-, Na-, and K-doped NiO are −1.67 eV, 0.17 eV, and 0.15 eV, respectively. While the values for Na and K are similar and both positive—indicating endothermic incorporation—Na doping requires the highest energy input. In contrast, Li doping is exothermic and exhibits the lowest binding energy, suggesting that the Li–NiO configuration is the most stable among the three. Geometry optimization reveals that the Ni–O bond length in pristine NiO is 2.085 Å. Upon doping with Li, Na, or K, the average bond lengths between the dopant site and adjacent oxygen atoms increase to 2.146 Å, 2.249 Å, and 2.321 Å, respectively. These pronounced structural distortions facilitate the formation of oxygen vacancies near the dopant sites, indicating an effective activation of surface oxygen species and an enhancement in the redox capability of NiO.

The migration and release of lattice oxygen in oxygen carriers are closely related to their electronic structure, and the density of states (DOS) provides a precise description of electronic configurations and distributions. Comparing the DOS of pristine and doped systems allows a deeper understanding of how dopants modify the electronic structure. To investigate the influence of Li, Na, and K doping on the electronic properties of NiO, we computed the total DOS for both undoped and doped NiO. As shown in Figure 3, the overall electronic structure of NiO remains largely unaffected by doping with Li, Na, or K. These results demonstrate that despite inducing local lattice distortions, the doping process preserves the overall stability and the cubic framework of NiO.

2.2. CO Adsorption on the Pure and Doped NiO Surfaces

The adsorption behavior of CO molecules on undoped, Li-, Na-, and K-doped NiO surfaces was systematically investigated, with particular focus on the influence of different adsorption sites. Multiple adsorption configurations were considered, including vertical adsorption on Ni-top, O-top, and vacancy sites, as well as C-end and O-end vertical orientations. In addition, parallel adsorption configurations on M–O, M…Ni, and O…O bridge sites (denoted as P1, P2, and P3, respectively, where M = Ni, Li, Na, or K) were evaluated. A schematic of the considered adsorption sites is provided in Figure 4.

Geometry optimization of each configuration, combined with adsorption energy and structural parameter analysis, revealed that CO can stably adsorb on all Ni-based oxygen carriers. Through geometry optimization of a wide range of potential CO adsorption configurations, the most stable configuration was identified as C-end vertical adsorption on the Ni-top site for the pure, Li-, and Na-doped NiO surfaces. In contrast, for the K-doped surface, it involves C-end vertical adsorption on a K…Ni bridge site. This clearly indicates that the Ni-top site serves as the preferred adsorption site for pure, Li-, and Na-doped NiO, while the K…Ni bridge site is the preferred site for K-doped NiO. The optimized structures for the most stable adsorption configurations in each system are shown in Figure 5, with corresponding adsorption energies and structural parameters summarized in

Table 2. Adsorption energy, structural parameters and preferred adsorption site of CO following adsorption on pure and doped surfaces.

Oxygen Carrier	Eads/eV	Mulliken Charge/e	LC–O/Å	LC–Ni/Å	Preferred Adsorption Site
CO	—	0	1.154	—	—
NiO	−0.41	0.04	1.155	2.085	Ni-top site
Li-NiO	−0.53	0.02	1.158	2.052	Ni-top site
Na-NiO	−0.46	−0.02	1.159	2.054	Ni-top site
K-NiO	−0.36	−0.07	1.156	2.063	K…Ni bridge site

.

For the most stable configuration on the undoped NiO surface, the adsorption energy was −0.41 eV, with a C–Ni distance of 2.085 Å and a C–O bond length of 1.155 Å—close to that of the free CO molecule (1.154 Å)—indicating no significant activation of CO upon adsorption. These results are consistent with previous computational data [49], validating the present methodology. On the Li-, Na-, and K-doped NiO surfaces, the adsorption energies were −0.53 eV, −0.46 eV, and −0.36 eV, respectively, with the strongest exothermic adsorption observed for Li-NiO. Compared to the undoped surface, Li and Na doping enhanced CO adsorption strength. Moreover, the C–Ni bond lengths in the doped systems (2.052–2.063 Å) were shorter than that in undoped NiO, indicating stronger binding between CO and the surface. The C–O bond lengths of adsorbed CO slightly increased (1.156–1.159 Å) in all doped systems, suggesting that Li, Na, and K doping promote activation of the CO molecule.

To elucidate the adsorption characteristics of CO on NiO and its doped systems, a detailed electronic structure analysis was conducted. This investigation examined the Mulliken charge populations before and after CO adsorption and compared the partial density of states (PDOS) of free CO with that of CO adsorbed at active sites. As summarized in Table 2, the net charge on CO changes from 0 e in the free state to 0.04 e, 0.02 e, −0.02 e, and −0.07 e upon adsorption on the NiO, Li-NiO, Na-NiO, and K-NiO surfaces, respectively. The Mulliken charge analysis indicates electron transfer between CO and the oxide surfaces: CO gains electrons on the undoped and Li-doped surfaces but donates electrons to the Na- and K-doped surfaces. Notably, the K-doped system, which exhibits the most significant electron transfer to CO, also shows the lowest adsorption energy. The distinct adsorption configuration on the K-doped surface, with CO adsorbed on the K…Ni bridge site, can be attributed to the larger ionic radius and more diffuse electron density of K⁺ compared to Li⁺ and Na⁺. This results in a more protruded local geometry and a modified electronic environment around the dopant, making the neighboring Ni atom a favorable site for CO adsorption. Consequently, this configuration leads to an overall weaker CO adsorption energy and the more pronounced electron transfer observed.

The PDOS spectra for both free and adsorbed CO are presented in Figure 6. The electronic configuration of the free CO molecule is [(1σ)²(2σ)²(3σ)²(4σ)²(1π)⁴(5σ)²(2π)⁰]. Its PDOS spectrum features four main peaks at −4.95 eV, −2.60 eV, 0.00 eV, and 6.80 eV, corresponding to the 4σ, 1π, 5σ, and 2π* orbitals, respectively. The 5σ orbital, located near the Fermi level, is identified as the highest occupied molecular orbital (HOMO), while the 2π* orbital represents the lowest unoccupied molecular orbital (LUMO). The predominant contribution of the C atom to both the 5σ and 2π* orbitals facilitate covalent bonding with surface metal atoms. Analysis of the PDOS for adsorbed CO reveals a hybridization between the Ni 3d and C 2π* orbitals, evidenced by an energy overlap of the 5σ and 1π orbitals and the emergence of new peaks. This orbital hybridization stabilizes the chemical bond between CO and the NiO surface. Furthermore, a shift in the CO orbital levels (4σ, 1π, 5σ, and 2π*) to lower energies upon adsorption confirms the stabilized interaction. Slight differences in the PDOS profiles across the doped surfaces correlate with variations in the extent of electron transfer, consistent with the Mulliken charge analysis.

2.3. Mechanism of CO Oxidation on Pure and Doped NiO Surfaces

The CO oxidation reaction on both pristine and doped NiO surfaces was systematically investigated through simulation. The oxidation of CO on the NiO surface proceeds through three elementary steps: adsorption of CO, formation of CO₂, and desorption of CO₂. In modeling the oxidation process, the adsorbed CO structure on the oxygen carrier surface was taken as the initial state (IS), the formed CO₂ as the intermediate state (IM), and the desorbed CO₂ as the final state (FS). The transition state (TS) between IS and IM was located for each system. The reaction pathways and energy profiles for CO oxidation on undoped and alkali-metal-doped NiO surfaces were computed, considering only the most stable adsorption configurations. The resulting reaction pathways and energy changes are illustrated in Figure 7 and Figure 8.

As shown in Figure 8, the energy barrier for CO₂ formation on the undoped NiO surface is 2.12 eV. In contrast, the corresponding barriers on Li-, Na-, and K-doped NiO surfaces are substantially lower, at 0.73 eV, 0.80 eV, and 0.99 eV, respectively. Compared with the undoped surface, alkali metal doping markedly reduces the activation barrier for CO₂ formation, with Li doping exhibiting the most pronounced effect—lowering the barrier from 2.12 eV to 0.73 eV. These results indicate that doping with different alkali metals effectively facilitates CO₂ formation by reducing the reaction barrier, thereby significantly enhancing the reactivity of the oxygen carrier. The overall reaction energy from CO adsorption to CO₂ desorption is 1.19 eV on the undoped NiO surface. On Li-, Na-, and K-doped surfaces, the reaction energies are 0.03 eV, −0.10 eV, and −0.84 eV, respectively. The negative values for Na- and K-doped surfaces indicate exothermic reactions, suggesting that CO oxidation proceeds more readily on these surfaces. Although the reaction remains endothermic on the Li-doped surface, both its activation barrier and reaction energy are lower than those on the undoped surface, implying a reduced energy input requirement. The aforementioned research findings indicate that doping with four alkali metals effectively enhances the surface oxidation of CO on NiO.

To gain deeper insight into the electron transfer and bonding interactions between CO and the oxygen carrier surface, the charge distribution during CO₂ formation was analyzed. Figure 9a–d display the charge density distribution for CO stably adsorbed at the active site. The C atom is surrounded by a region of low charge density, whereas the adjacent O atom appears in red, indicating significant electron accumulation around oxygen. This charge distribution suggests electron transfer from the C atom to the O atom. Moreover, no pronounced charge transfer is observed between the adsorbed CO molecule and the surface, indicating that CO is physiosorbed on the surface. Figure 9e–h illustrate the charge distribution between the generated CO₂ and the oxygen carrier surface after the oxidation reaction. Significant electron cloud overlap is observed between the O atom of the NiO surface and the C atom, accompanied by the breaking of the bond between that surface O atom and the bulk. These findings indicate the formation of a new covalent bond between the surface O atom and the CO molecule, leading to the generation of CO₂.

2.4. Oxygen Activity for Alkali Metal Doped NiO Surfaces

An essential function of oxygen carriers in the CLC process is to supply active lattice oxygen for surface reactions with fuel molecules. When CO molecules are oxidized, oxygen vacancies are generated on the surface of the oxygen carrier. Consequently, the oxygen vacancy formation energy is widely regarded as an indicator of lattice oxygen reactivity at the oxygen carrier surface. The oxygen vacancy formation energies for NiO surfaces doped with different alkali metals were calculated to evaluate the influence of dopants on surface oxygen activity. As shown in Figure 10, the oxygen vacancy formation energy for the undoped NiO surface is 3.60 eV, while those for the Li-, Na-, and K-doped systems (Li-NiO, Na-NiO, K-NiO) are reduced to 3.05 eV, 2.98 eV, and 2.90 eV, respectively. Compared with the undoped system, alkali-metal doping notably reduces the energy barrier for oxygen vacancy formation. Among them, K-doping shows the most significant effect, with a reduction of approximately 0.7 eV. This finding is consistent with the conclusion in existing studies, which indicates that K doping can enhance the reactivity of the oxygen carrier [33,34]. This trend can be attributed to the substitution of Ni²⁺ ions by aliovalent M⁺ ions (M = Li, Na, K), which creates an electron-deficient environment in the doped region and weakens the bonding between oxygen and neighboring metal ions. These results uncover an important subtlety in dopant functionality: whereas Li doping promotes the adsorption of the initial reactant (CO), K doping proves to be the most efficient in activating and liberating lattice oxygen, as evidenced by its lowest oxygen vacancy formation energy. This indicates that different dopants predominantly enhance different elementary steps in the catalytic redox cycle.

The preceding analyses on structure, CO adsorption, oxidation kinetics, and oxygen activity collectively indicate that doping with Li, Na, and K differentially enhances the reactivity of NiO. To present a clear and comprehensive overview, the key quantitative metrics for all systems are compared in

Table 3. Comparative overview of key performance metrics for alkali metal-doped NiO.

Performance Metric	Pure NiO	Li-Doped	Na-Doped	K-Doped	Ranking (Best → Good)
CO adsorption energy, (eV)	−0.41	−0.53	−0.46	−0.36	Li > Na > NiO > K
Barrier for CO2 formation, (eV)	2.12	0.73	0.80	0.99	Li < Na < K < NiO
Oxygen vacancy formation energy, (eV)	3.60	3.05	2.98	2.90	K < Na < Li < NiO
Overall reaction energy, (eV)	1.19	0.03	−0.10	−0.84	K < Na < Li < NiO
Overall promotional effectiveness	–	Effective	Moderately Effective	Effective	Li = K > Na

. This table ranks the dopants according to the critical descriptors for CLC performance: the adsorption strength of the reactant (CO), the kinetic barrier for the crucial CO₂ formation step, and the ease of lattice oxygen release (reflected by the oxygen vacancy formation energy). An overall evaluation of the promotional effectiveness is also provided. The data conclusively demonstrates that doping with Li and K provides a more balanced and superior enhancement across all major performance criteria, with Na following.

3. Calculation Method and Model

All DFT calculations were performed using the Cambridge Sequential Total Energy Package (CASTEP) within the Materials Studio 2019 Software [50]. The Perdew–Burke–Ernzerhof (PBE) [51] generalized gradient approximation (GGA) [52] was employed to treat electron exchange and correlation, and ultrasoft pseudopotentials were used to represent the ionic cores. The total energy was evaluated via self-consistent field (SCF) iterations. All DFT + U calculations incorporated spin polarization to properly describe the antiferromagnetic nature of NiO. A range of Hubbard U values (0–9 eV) was systematically evaluated, including 1, 2.5, 4, 6.3, 8, and 9 eV. A U value of 6.3 eV was selected, as it yielded results in good agreement with previously reported theoretical benchmarks. Furthermore, the optimized lattice parameter for bulk NiO was calculated to be 4.163 Å, consistent with the experimental value of 4.170 Å, supporting the reliability of the chosen computational approach [53,54].

The NiO(001) surface was chosen as the model system in this work due to its high reactivity in adsorption processes, as established in previous CLC studies. Experimental evidence also indicates that the (001) plane cleaves readily, yielding a flat and largely defect-free surface, further supporting its selection for investigating NiO-mediated reactions. The NiO(001) surface was modeled using a (2 × 2) periodic supercell with a six-layer slab, representing both the undoped and doped systems. Alkali metal doping was introduced via substitutional replacement of a Ni atom in the first surface layer by Li, Na, or K, denoted as M-NiO (where M = Li, Na, K). A 15 Å vacuum layer was included to minimize periodic interactions between adjacent slabs. During structural optimization, the top two layers along with any adsorbates were fully relaxed, while the bottom four layers remained fixed in their bulk positions, as illustrated in Figure 11.

All calculations employed a plane-wave cutoff energy of 330 eV. Geometry optimization was performed under the following convergence criteria: maximum force per atom ≤ 0.05 eV/Å, energy change ≤ 2.0 × 10⁻⁵ eV/atom, SCF tolerance ≤ 2.0 × 10⁻⁶ eV/atom, and maximum atomic displacement ≤ 0.002 Å. A 3 × 3 × 1 Monkhorst–Pack k-point mesh was used for sampling the Brillouin zone. The CO molecule was optimized in a 1 nm³ cubic cell, yielding a C–O bond length of 1.154 Å, in excellent agreement with previously reported theoretical and experimental values [55].

The present study systematically computes the energies of multiple configurations, including the primitive cell, surface models, doped systems, and adsorption structures. Among these, the binding energy serves as a key physical parameter for evaluating the feasibility of incorporating dopant atoms into the host material. The binding energy associated with alkali metal doping, E_bind, is defined by the following expression:

$${\mathrm{E}}_{\mathrm{bind}}{=\mathrm{E}}_{\mathrm{M}-\mathrm{NiO}}-\left({\mathrm{E}}_{\mathrm{NiO}}{-\mathrm{E}}_{\mathrm{Ni}}\right){-\mathrm{E}}_{\mathrm{M}}$$

(3)

where E_M–NiO denotes the total energy of the alkali metal doped NiO supercell, E_NiO corresponds to the total energy of the pure NiO supercell, and E_Ni and E_Li represent the energies of an isolated Ni atom and Li atom, respectively.

To quantify the strength of CO adsorption on both pure and doped surfaces, the adsorption energy (E_ads) was calculated according to the following expression:

$${\Delta \mathrm{E}}_{\mathrm{ads}}{=\mathrm{E}}_{\mathrm{OC}+\mathrm{CO}}-\left({\mathrm{E}}_{\mathrm{OC}}{+\mathrm{E}}_{\mathrm{CO}}\right)$$

(4)

where E_OC+CO denotes the total energy of the adsorbed system, while E_OC and E_CO represent the total energies of the clean surface model and the isolated CO molecule, respectively. A more negative value of E_ads generally corresponds to a stronger adsorbate–surface interaction and greater stability of the adsorption configuration.

The TS geometry and energy barrier along reaction pathways were located using the linear/quadratic synchronous transit (LST/QST) method [56]. The energy barrier, E_b, is defined as:

$${\Delta \mathrm{E}}_{\mathrm{b}}{=\mathrm{E}}_{\mathrm{TS}}{-\mathrm{E}}_{\mathrm{IS}}$$

(5)

where E_IS and E_TS represent the total energies of the initial state and transition state, respectively. A higher energy barrier corresponds to a more kinetically hindered process and a slower reaction rate.

The oxygen vacancy formation energy (E_vac) serves as a key metric for evaluating the oxygen activity on oxygen carrier. This quantity is calculated using the following expression:

$${\mathrm{E}}_{\mathrm{vac}}{=\mathrm{E}}_{\mathrm{def}}{-\mathrm{E}}_{\mathrm{per}}+0.5{\mathrm{E}}_{{\mathrm{O}}_2}$$

(6)

where E_per and E_def represent the total energies of the pure NiO supercell and the defective supercell containing one oxygen vacancy, respectively, while E_O2 denotes the energy of an isolated O₂ molecule. A lower value of E_vac generally indicates facilitated oxygen release and enhanced surface oxidation capability.

4. Conclusions

This study systematically investigated the enhancement of Ni-based oxygen carriers for CLC through alkali metal doping (Li, Na, K) using DFT calculations. The investigation focused on constructing doped NiO(001) surface models, analyzing structural and electronic properties, evaluating CO adsorption behavior, simulating the CO oxidation pathway, and calculating oxygen vacancy formation energies. The results demonstrate that incorporating alkali metals induces significant local lattice distortion, while preserving the overall electronic structure. Doping markedly improves CO adsorption strength, with adsorption energies increasing to −0.53 eV, −0.46 eV, and −0.36 eV for Li-, Na-, and K-doped surfaces, respectively, compared to −0.41 eV for pure NiO. More importantly, alkali metal doping drastically reduces the energy barrier for CO₂ formation from 2.12 eV on pristine NiO to 0.73 eV (Li), 0.80 eV (Na), and 0.99 eV (K), with Li exhibiting the most pronounced promotional effect. Furthermore, oxygen vacancy formation energies decrease significantly from 3.60 eV for undoped NiO to 3.05 eV, 2.98 eV, and 2.90 eV for the Li-, Na-, and K-doped systems, respectively, indicating a substantial enhancement in lattice oxygen reactivity.

In summary, alkali metal doping (Li, Na, K) confers multiple synergistic benefits to enhance the performance of NiO as a CLC oxygen carrier. (1) Structural modification: Doping induces local lattice expansion and distortion, which weakens the metal–oxygen bonds and facilitates oxygen release. (2) Enhanced surface reactivity: The adsorption strength of key reactant CO is improved, promoting its activation. (3) Lowered kinetic barriers: The energy barrier for the key CO₂ formation step is dramatically reduced, significantly accelerating the oxidation kinetics. (4) Improved redox activity: The oxygen vacancy formation energy is substantially decreased, indicating a marked increase in the mobility and reactivity of lattice oxygen. Li doping creates the most thermodynamically stable doped structure and enhances the initial CO adsorption strength. In contrast, K doping, by inducing the greatest lattice distortion, most effectively weakens the metal–oxygen bonds and facilitates oxygen vacancy generation, thereby optimizing the oxygen transfer capability. This distinction implies that the choice of the optimal dopant could be tailored depending on whether the rate-limiting step in a specific process. These findings provide a theoretical foundation for designing highly active Ni-based oxygen carriers, offering theoretical guidance for advancing efficient CLC systems and supporting the development of practical carbon capture technologies. However, it should be noted that these theoretical predictions, while providing fundamental atomistic insights, are derived from calculations under idealized conditions and await experimental validation.