First-Principles and Device-Level Investigation of β-AgGaO₂ Ferroelectric Semiconductors for Photovoltaic Applications

Abstract

Ferroelectric semiconductors, with their inherent spontaneous polarization, present a promising approach for efficient charge separation, making them attractive for photovoltaic applications. The potential of β-AgGaO₂, a polar ternary oxide with an orthorhombic Pna2₁ structure, as a light-absorbing material is evaluated. First-principles computational analysis reveals that β-AgGaO₂ possesses an indirect bandgap of 2.1 eV and exhibits pronounced absorption within the visible spectral range. Optical simulations suggest that a 300 nm thick absorber layer could theoretically achieve a power conversion efficiency (PCE) of 20%. Device-level simulations using SCAPS-1D evaluate the influence of hole and electron transport layers on solar cell performance. Among the tested hole transport materials, Cu₂FeSnS₄ (CFTS) achieves the highest PCE of 14%, attributed to its optimized valence band alignment and reduced recombination losses. In contrast, no significant improvements were observed with the electron transport layers tested. These findings indicate the potential of β-AgGaO₂ as a ferroelectric photovoltaic absorber and emphasize the importance of band alignment and interface engineering for optimizing device performance.

1. Introduction

Ferroelectric semiconductors, due to their inherent spontaneous polarization properties, can facilitate the separation of photogenerated charge carriers without the need for an external electric field, making them significant in optoelectronic research [1]. In recent years, with the increasing demand for high-performance materials in optoelectronic devices, ferroelectric semiconductors have gradually become a prominent research direction [2]. For example, β-phase ABO₂ ternary oxides derived from the zinc-blende ZnO structure, with their Pna2₁ polar space group, show promising ferroelectric semiconductor properties [3]. First-principles calculations indicate that β-CuGaO₂ possesses both a high spontaneous polarization of 83.80 μC/cm² and a direct, narrow bandgap, making it a promising candidate for ferroelectric photovoltaic applications [4,5]. By employing the spectroscopic limited maximum efficiency (SLME) approach, our calculation indicates that the photovoltaic efficiency of β-CuGaO₂ approaches 30%, which is comparable to that of traditional direct bandgap photovoltaic materials like CdTe [6]. Furthermore, as a material possessing both ferroelectric and semiconducting properties, β-CuGaO₂ shows extensive application potential in fields such as energy harvesting and wastewater environmental remediation [7].

Building on the excellent properties of β-CuGaO₂, Wang et al. have employed unit co-doping strategies to identify four ferroelectric semiconductor materials, including β-CuInO₂, β-AgGaO₂, β-CuGaO₂, and β-AuInO₂, which exhibit outstanding photocatalytic performance and broad spectral absorption capabilities [8]. Furthermore, Guo et al. introduced cations such as P³⁺, As³⁺, Sb³⁺, and Bi³⁺ into β-phase ABO₂ ternary oxide ferroelectric semiconductors, thereby synthesizing 44 different types of β-AIBIIIO₂ photocatalysts with dual polarization characteristics, which significantly improve electron–hole pair separation efficiency [9]. Computational results indicate that the solar-to-hydrogen efficiency of β-TlBiO₂ can reach 17.2% [9]. In terms of experimental synthesis, β-AgGaO₂ has been successfully prepared, showing excellent visible light absorption and potential ferroelectric properties, indicating its important application value in photocatalytic hydrogen production and photovoltaic fields [10]. First-principles calculations and experimental measurements show that the bandgap of β-AgGaO₂ is approximately 2.1 eV, and its band edge positions can span the redox potential of water, thereby facilitating the photocatalytic water splitting process [8,11,12,13]. In contrast, BiFeO₃, a commonly used ferroelectric material, can be effectively tuned through La doping to adjust the bandgap, which typically ranges from 2.2 eV to 2.7 eV, and is widely used in photovoltaic and photocatalytic research [14,15]. Although both β-AgGaO₂ and BiFeO₃ exhibit similar ferroelectric properties and bandgaps, BiFeO₃ often requires an additional electron transport layer to enhance photoelectric conversion efficiency and improve photovoltaic performance [16,17]. In contrast, β-AgGaO₂, with its excellent visible light absorption, has emerged as a potential material for solar cell structures. To further optimize its device performance, β-AgGaO₂ can benefit from the incorporation of electron or hole transport layers, which facilitate efficient charge carrier separation and transport, thereby significantly enhancing overall photovoltaic efficiency and expanding its potential applications in the field of photovoltaics. In this study, density functional theory (DFT) calculations combined with numerical simulations were utilized to investigate the optical and electronic properties of β-AgGaO₂, as well as to assess its theoretical power conversion efficiency and photovoltaic device performance. The computational results demonstrate that the integration of a hole transport layer (HTL) into the β-AgGaO₂-based architecture can enhance photovoltaic performance by promoting more effective charge carrier separation and transport.

2. Density Functional Theory (DFT) Computational Approach

All first-principles calculations within the framework of density functional theory (DFT) were performed using the plane-wave pseudopotential approach as implemented in the VASP package [18,19]. The exchange-correlation interactions were treated with the generalized gradient approximation (GGA) in the form proposed by Perdew, Burke, and Ernzerhof (PBE) [20]. Equilibrium structural parameters, including lattice constants and atomic positions, were obtained by minimizing the total energy with the conjugate gradient algorithm. A kinetic energy cutoff of 550 eV was employed for the plane-wave basis set expansion, and k-point sampling grids were generated with the aid of the pre-processing tool VASPKIT [21]. The Gamma scheme was chosen for k-point mesh of 6 × 5 × 6. The relaxation of electronic degrees of freedom was terminated when the total energy and band structure energy changes between two electron steps were less than 10⁻⁶ eV. The convergence criterion for ionic steps was set to 0.01 eV Å⁻¹. The calculated lattice parameters of β-AgGaO₂ are a = 5.64 Å, b = 7.18 Å, and c = 5.59 Å as shown in Figure 1. The CIF file of the crystal structure is provided in the Supporting Information.

These values are in excellent agreement with the experimentally reported lattice parameters of a = 5.56 Å, b = 7.14 Å, and c = 5.46 Å [10]. The Modified Becke–Johnson (MBJ) exchange potential, which corrects the underestimation of the bandgap in conventional DFT-PBE calculations, was used to compute the electronic structure and optical absorption spectra [22]. The parameter C in the MBJ functional was set to 1.5, and the computed bandgap showed good agreement with previous computational results and experimental data. Ab initio molecular dynamics (AIMD) simulations based on first-principles calculations were also conducted to evaluate the thermodynamic stability of AgGaO₂, as shown in Figure S1 of the Supporting Information.

3. Computational Analysis of Electronic and Optical Properties

3.1. Electronic Properties

The optical and electronic characteristics of materials are fundamentally determined by their band structures. The band gap and density of states (DOS) for β-AgGaO₂ are depicted in Figure 2.

The bandgap calculation employs the Modified Becke–Johnson (MBJ) exchange potential, which generally yields more accurate results. The calculated bandgap is in good agreement with experimental data and HSE06 computational results (approximately 2.18 eV) [10,11,12]. As illustrated in Figure 2, the valence band maximum is mainly constituted by O-2p orbitals, whereas the conduction band minimum is chiefly attributed to Ag-5s states. The effective masses of charge carriers (holes and electrons) were calculated using a second-order derivative method, as outlined in Equation (1).

$$m^{*}=\hslash {\left({\displaystyle \frac{d^{2}E}{d{k}^2}}\right)}^{-1}$$

(1)

Here, E denotes the band edge energy as a function of the wave vector k. The computed effective masses for electrons and holes are 0.17 m₀ and −3.45 m₀, respectively, which are consistent with previously reported values of 0.1 m₀ and −2.9 m₀ in the literature [8]. The effective density of states in the conduction band (Nc) and at the valence band maximum for holes (Nv) were determined using the following expressions, with the temperature T set to 300 K.

$$N_c=2{\left({\displaystyle \frac{m_e^{*}{k}_{B}T}{2\pi {\hslash }^2}}\right)}^{{\displaystyle \frac{3}{2}}}$$

(2)

$$N_v=2{\left({\displaystyle \frac{{2\pi m}_h^{*}{k}_{B}T}{h^2}}\right)}^{{\displaystyle \frac{3}{2}}}$$

(3)

The calculated results for Nc and Nv are 2.4 × 10¹⁸ and 1.5 × 10²⁰ cm⁻³, respectively. The band edge energies of β-AgGaO₂ referenced to the vacuum level were estimated using the following empirical relationships [23]:

$$E_{VBM}=\chi +0.5E_g$$

(4)

$$E_{CBM}=\chi -0.5E_g$$

(5)

$$\chi ={\left({\chi }_{Ag}^1+{\chi }_{Ga}^1+{\chi }_O^2\right)}^{{\displaystyle \frac{1}{1+1+2}}}$$

(6)

χ represents the average electronegativity of β-AgGaO₂, and E_g denotes the band gap. The absolute electronegativity values for ${\chi }_{Ag}$, ${\chi }_{Ga}$, and ${\chi }_O$ are 4.44 eV, 3.2 eV, and 7.54 eV, respectively. The calculated values of the valence band maximum (E_VBM) and conduction band minimum (E_CBM) are 4.3 eV and 6.4 eV, respectively.

The carrier mobility of β-AgGaO₂ was computed using the AMSET code, which employs the momentum relaxation time approximation (MRTA) method [24]. All required input parameters, including the elastic constants (

Table 1. The calculated elastic constants of β-AgGaO₂ (GPa).

C11	C12	C13	C22	C23	C33	C44	C55	C66
103.2	80.2	87.9	157.3	90.6	131.1	16.9	15.2	5.1

), dielectric constants (

Table 2. The calculated static (ɛ₀), high frequency (ɛ_∞) dielectric constant, and effective phonon frequency (ω).

Material	ɛ0			ɛ∞			ω (THz)
	x	y	z	x	y	z
AgGaO2	14.06	10.81	13.10	6.04	6.11	6.30	10.41

), phonon frequencies (Table 2), piezoelectric tensor (

Table 3. The calculated piezoelectric tensor (C m⁻²).

		XX	YY	ZZ	XY	YZ	ZX
AgGaO2	x	0.00	0.00	0.00	0.00	0.00	−0.37
	y	0.00	0.00	0.00	0.00	−0.75	0.00
	z	−0.52	−0.86	1.29	0.00	0.00	0.00

), and deformation potentials (

Table 4. The calculated deformation potentials for the lower conduction bands and upper valence bands.

Material			DXX	DYY	DZZ
AgGaO2	VBM(eV)	DXX	1.23	0.39	0.76
		DYY	0.39	2.59	0.17
		DZZ	0.76	0.17	0.1
	CBM(eV)	DXX	0.05	0.31	0.00
		DYY	0.31	0.71	0.03
		DZZ	0.00	0.03	0.54

), were derived from our DFT calculations. The resulting carrier mobilities are summarized in

Table 5. The calculated electrons (μ_e) and holes (μ_h) mobility for AgGaO₂ $\left({\mathrm{c}\mathrm{m}}^2{\mathrm{V}}^{-1}{\mathrm{S}}^{-1}\right)$.

Material		ND = 1015 /cm−3	ND = 1016 /cm−3	ND = 1017 /cm−3	ND = 1018 /cm−3	ND = 1019 /cm−3	ND = 1020 /cm−3
AgGaO2	μe(n)	107	105.2	100.3	90.7	82.0	44.4
	μh(p)	0.4	0.3	0.3	0.3	0.3	0.3

.

The findings demonstrate that, under moderate impurity concentrations, the electron mobility can achieve 100 cm²/V·s, while the hole mobility is 0.3 cm²/V·s. These values substantially exceed those of the intrinsic ferroelectric material BiFeO₃, whose carrier mobility typically ranges from 1.6 to 1.8 cm²/V·s, with the electron mobility of β-AgGaO₂ being particularly pronounced [25]. In BiFeO₃, significant enhancement in electron mobility generally necessitates doping with ions such as La, which can increase electron mobility up to 80 cm²/V·s [16,17]. The mobility of charge carriers in semiconductors is influenced by various scattering mechanisms, primarily including ionized impurity (IPM), polar optical phonon (POP), piezoelectric (PIE), and acoustic deformation potential scattering (ADP), as shown in Figure 3. Consistent with other polar semiconductors, the carrier mobility in β-AgGaO₂ is predominantly governed by polar optical phonon scattering [26]. The relatively low hole mobility in β-AgGaO₂ is primarily attributed to strong scattering effects from polar optical phonons and piezoelectric interactions. This limited mobility significantly influences device performance and suggests the need for further optimization through material engineering or interface modification.

3.2. Optical Properties

There exists a close interrelation between the band gap, dielectric constant, and optical properties. Within the framework of the random phase approximation (RPA), the optical properties are derived from the real (ε₁) and imaginary (ε₂) parts of the dielectric function. From ε₁ and ε₂, the refractive index (n) and extinction coefficient (k) can be determined as follows [27]:

$$n\left(E\right)={\displaystyle \frac{1}{\sqrt{2}}}\sqrt{{\epsilon }_1+\sqrt{{\epsilon }_1^2\left(E\right)+{\epsilon }_2^2\left(E\right)}}$$

(7)

$$k\left(E\right)={\displaystyle \frac{1}{\sqrt{2}}}\sqrt{{-\epsilon }_1+\sqrt{{\epsilon }_1^2\left(E\right)+{\epsilon }_2^2\left(E\right)}}$$

(8)

Consequently, the reflectance (R) and absorption coefficient (α) are calculated from n and k using the following expressions [27]:

$$R\left(E\right)={\displaystyle \frac{{\left(n-1\right)}^2+K^2}{{\left(n+1\right)}^2+K^2}}$$

(9)

$$\alpha \left(E\right)={\displaystyle \frac{4\pi k}{hc/E}}$$

(10)

As illustrated in Figure 4a, β-AgGaO₂ demonstrates strong optical absorption in the visible region, with absorption coefficients on the order of 10⁵ cm⁻¹, which highlights its significant potential for applications in solar energy harvesting.

The photoelectric conversion efficiency is evaluated using the following physical model [28]:

$$PCE={\displaystyle \frac{{\int }_0^{{\lambda }_{max}}W\left(\lambda \right)\left[1-R\left(\lambda \right)\right]\left(1-e^{-\alpha \left(\lambda \right)d}\right)C\left(\lambda \right)d\lambda }{{\int }_0^{\infty }W\left(\lambda \right)d\lambda }}$$

(11)

In this model, λ denotes the wavelength, and λ_max is the maximum absorption wavelength corresponding to the band gap (Eg), given by:

$${\lambda }_{max}={\displaystyle \frac{hc}{E_g}}$$

(12)

The conversion coefficient C(λ) signifies the fraction of excitation energy that results in the creation of an electron–hole pair via the minimum band gap, described as:

$$C\left(\lambda \right)=\lambda {\displaystyle \frac{E_g}{hc}}$$

(13)

The calculated absorption coefficient α and reflectance R are depicted in Figure 4a and Figure 4b, respectively. The photoelectric conversion efficiencies, as shown in Figure 4c, reach 13% and 20% for film thicknesses of 100 nm and 300 nm, respectively, with the system approaching absorption saturation (21%) at 300 nm. In our device simulation, the thickness of β-AgGaO₂ is chosen as 300 nm, a value that is commonly used for BiFeO₃ thin films in experiments [16,29,30]. The physical model described in Equation (11) calculates the theoretical maximum photovoltaic conversion efficiency, which serves as an initial reference for evaluating the photovoltaic potential of the material. To obtain a more accurate representation of the actual device performance, we construct a detailed solar cell device model in the following sections for further calculation of the solar cell characteristics of β-AgGaO₂.

4. Photovoltaic Performance Analysis of β-AgGaO₂ Devices with and Without Transport Layers

The β-AgGaO₂ solar cell structures were designed and analyzed utilizing SCAPS-1D, a one-dimensional simulation software for solar cells developed by the Department of Electronics and Information Systems at Ghent University, Belgium [31]. SCAPS-1D enables the simulation of homojunction, heterojunction, multi-junction, as well as Schottky junction solar cells through numerical solutions of the Poisson equation [32,33].

$${\displaystyle \frac{d^2\phi \left(x\right)}{d{x}^2}}={\displaystyle \frac{q}{\epsilon }}\left[n\left(x\right)-p\left(x\right)-N_D^{+}\left(x\right)+N_A^{-}\left(x\right)-p_t\left(x\right)+n_t\left(x\right)\right]$$

(14)

where $\phi \left(x\right)$ is the electrostatic potential, q is the elementary charge, ε is the permittivity, n(x) and p(x) are the electron and hole densities, $N_D^{+}\left(x\right)$ and $N_A^{-}\left(x\right)$ are the ionized donor and acceptor concentrations, and $n_t\left(x\right)$ and ${ p}_t\left(x\right)$ represent trapped charge carrier densities. SCAPS also solves the continuity equations for electrons and holes:

$${\displaystyle \frac{\partial n}{\partial t}}={\displaystyle \frac{1}{q}}{\displaystyle \frac{\partial J_n}{\partial x}}+G-R$$

(15)

$${\displaystyle \frac{\partial p}{\partial t}}=-{\displaystyle \frac{1}{q}}{\displaystyle \frac{\partial J_p}{\partial x}}+G-R$$

(16)

where Jn and Jp are the current densities of electrons and holes, G is the generation rate, and R is the recombination rate. This allows for a detailed understanding of charge transport and recombination processes across multilayer device architectures. The input parameters used in the simulation are presented in

Table 6. Input Parameters of ITO, ETLs, and Absorption layer.

Material Property	ITO	ZnO	PCBM	C60	WS2	AgGaO2
Thickness “t” (nm)	500	50	50	50	100	300
Bandgap “Eg” (eV)	3.5	3.3	2	1.7	1.8	2.1
Electron affinity (eV)	4	4	3.9	3.9	3.95	4.3
Relative permittivity “Er”	9	9	3.9	4.2	13.6	13.1
C.B”Nc” (cm−3)	2.2 × 1018	3.7 × 1018	2.5 × 1021	8.0 × 1019	1 × 1018	2.4 × 1018
V.B”Nd” (cm−3)	1.8 × 1019	1.8 × 1019	2.5 × 1021	8.0 × 1019	2.4 × 1019	1.5 × 1020
Electron thermal Velocity “Ve” (cm/s)	1 × 107	1 × 107	1 × 107	1 × 107	1 × 107	1 × 107
Hole thermal velocity “Vh” (cm/s)	1 × 107	1 × 107	1 × 107	1 × 107	1 × 107	1 × 107
Electron mobility “μn” (cm2/V·s)	20	100	0.2	8.0 × 10−2	100	100
Hole mobility “μh” (cm2/V·s)	10	25	0.2	3.5 × 10−3	100	0.3
Donor density “ND” (cm−3)	1 × 1021	1 × 1018	2.93 × 1017	1 × 1017	1 × 1018	1 × 1019
Acceptor density “NA” (cm−3)	0	0	0	0	0	0
Defect density “Nt” (cm−3)	1 × 1015	1 × 1015	1 × 1015	1 × 1015	1 × 1015	1 × 1015
References	[35,36]	[37,38]	[36,39]	[40]	[35,41]

,

Table 7. Input Optimization Parameters for the Studied HTLs.

Material Property	CuSCN	P3HT	PEDOT:PSS	CuI	CFTS
Thickness “t” (nm)	50	50	50	100	100
Bandgap “Eg” (eV)	3.6	1.7	1.6	3.1	1.3
Electron affinity (eV)	1.7	3.5	3.4	2.1	3.3
Relative permittivity “Er”	10	3	3.9	6.5	9
C.B”Nc” (cm−3)	2.2 × 1019	2 × 1021	2.2 × 1018	2.8 × 1019	2.2 × 1018
V.B”Nd” (cm−3)	1.8 × 1018	2 × 1021	1.8 × 1019	1 × 1019	1.8 × 1019
Electron thermal Velocity “Ve” (cm/s)	1 × 107	1 × 107	1 × 107	1 × 107	1 × 107
Hole thermal velocity “Vh” (cm/s)	1 × 107	1 × 107	1 × 107	1 × 107	1 × 107
Electron mobility “μn” (cm2/V·s)	100	1.8 × 10−3	4.5 × 10−2	100	21.98
Hole mobility “μh”(cm2/V·s)	25	1.86 × 10−2	4.5 × 10−2	43.9	21.98
Donor density “ND” (cm−3)	0	0	0	0	0
Acceptor density “NA” (cm−3)	1 × 1018	1 × 1018	1 × 1018	1 × 1018	1 × 1018
Defect density “Nt” (cm−3)	1 × 1015	1 × 1015	1 × 1015	1 × 1015	1 × 1015
References	[37]	[37]	[42,43]	[44]	[45]

and

Table 8. Specifications of Defect Layers’ Interface.

Interface	Defect Type	Capture Cross-Section: Electrons /Holes (cm 2)	Energetic Distribution	Reference for Defect Energy Level	Total Density (cm−2)
ETL/AgGaO2	neutral	1.0 × 10−17	single	Above the highest EV	1.0 × 1010
		1.0 × 10−18
AgGaO2/HTL	neutral	1.0 × 10−18	single	Above the highest EV	1.0 × 1010
		1.0 × 10−19

. Table 6 primarily lists the parameters of the electrodes, absorption layer, and electron transport layers (ETLs). Table 7 focuses on the parameters of the hole transport layers (HTLs), while Table 8 provides the parameters of the interfacial layers. In this study, the parameters for the hole transport layers (HTLs) and electron transport layers (ETLs) were primarily selected based on values reported in the literature (Table 6, Table 7 and Table 8), which is a widely adopted approach in solar cell device modeling. However, recent simulations of CsSnI₂Br-based perovskite solar cells have gone a step further by actively tuning the properties of charge transport layers to enhance photovoltaic performance [34]. Given that both charge transport layers’ interface and intrinsic defect concentrations in β-AgGaO₂ can also influence device efficiency, we have included corresponding simulation results in the Supporting Information (Tables S1–S4) to provide a more comprehensive assessment of their impact. Material properties for the electrodes, HTLs, and ETLs were obtained from the literature, whereas the parameters for β-AgGaO₂ were mainly derived from our DFT calculations.

The work function assigned to the gold (Au) electrode is 5.1 eV. The photovoltaic performance was evaluated across three device configurations: (i) a structure without transport layers (ITO/β-AgGaO₂/Au), (ii) a structure incorporating a hole transport layer (ITO/β-AgGaO₂/HTL/Au), and (iii) a structure incorporating an electron transport layer (ITO/ETL/β-AgGaO₂/Au).

4.1. Photovoltaic Performance Without Transport Layers (ITO/β-AgGaO₂/Au)

The performance of the solar cell without hole transport layer (HTL) and electron transport layer (ETL) was analyzed using current–voltage (J–V) characteristics and band diagrams (Figure 5).

The power conversion efficiency (PCE) of the device is 1.80%, with a short-circuit current density (Jsc) of 6.01 mA/cm² and an open-circuit voltage (Voc) of 0.39 V. This indicates low charge extraction efficiency and significant recombination losses, resulting in reduced overall device performance. This performance is similar to that of ferroelectric La-doped BiFeO₃ (FTO/BLFO/Au) photovoltaic devices, which have a PCE of 1.5%, making it difficult to achieve high-performance photovoltaic devices [17].

The band diagram further explains this phenomenon in Figure 5b: the conduction band (Ec) alignment between AgGaO₂ and ITO facilitates effective electron extraction and transport. However, the poor valence band (Ev) alignment suppresses hole extraction, leading to enhanced interface recombination and hindered charge transport. Since the electron affinity was estimated using a semi-empirical formula, some degree of uncertainty may be present. To assess its influence, we varied the electron affinity from 4.0 eV to 4.6 eV in increments of 0.1 eV and recalculated the corresponding power conversion efficiencies (PCEs). As shown in

Table 9. The calculated power conversion efficiency (PCE, %) of photovoltaic devices with different electron transport layers (ETLs), hole transport layers (HTLs), and no transport layer as a function of electron affinity. Values in parentheses represent PCE without considering interface recombination.

Electron Affinity(eV)	ETL/HTL-Free	TiO2	WS2	ZnO	C60	PCBM	CuI	PSS	CuSCN	P3HT	CFTS
4.0	3.50	3.36 (3.38)	0.87 (0.87)	3.48 (3.48)	1.16 (1.16)	1.30 (1.30)	6.30 (8.05)	7.63 (9.08)	7.09 (8.49)	6.61 (6.98)	8.90 (14.47)
4.1	2.93	2.92 (2.92)	0.72 (0.72))	2.92 (2.93)	0.97 (0.97)	1.13 (1.13)	5.73 (7.50)	6.79 (8.31)	6.52 (7.93)	5.95 (6.68)	8.29 (13.07)
4.2	2.36	2.36 (2.36)	0.57 (0.57)	2.36 (2.36)	0.77 (0.77)	0.93 (0.93)	5.17 (6.95)	5.96 (7.54)	5.95 (7.38)	5.27 (6.31)	8.19 (11.51)
4.3	1.80	1.80 (1.80)	0.43 (0.43)	1.80 (1.80)	0.58 (0.58)	0.72 (0.72)	4.61 (6.41)	5.13 (6.78)	5.39 (7.91)	4.60 (5.96)	8.10 (9.97)
4.4	1.26	1.26 (1.26)	0.28 (0.00)	1.26 (1.26)	0.00 (0.37)	0.50 (0.50)	4.04 (5.86)	4.35 (6.03)	0.00 (0.00)	3.93 (5.44)	7.80 (8.45)
4.5	0.73	0.73 (0.73)	0.00 (0.00)	0.73 (0.73)	0.00 (0.00)	0.00 (0.28)	3.48 (5.32)	4.01 (0.00)	0.00 (0.00)	3.27 (4.89)	6.88 (6.97)
4.6	0.27	0.27 (0.27)	0.00 (0.00)	0.15 (0.15)	0.00 (0.00)	0.00 (0.00)	2.97 (0.00)	3.84 (0.00)	0.00 (0.00)	2.65 (0.00)	0.00 (0.00)

, a decrease in electron affinity tends to improve band alignment. The results suggest that even small variations in electron affinity can cause noticeable changes in PCE, indicating the model’s sensitivity to this parameter. Accordingly, the use of experimental techniques such as UPS or XPS in future studies may help obtain more accurate band alignment data and improve the reliability of simulation results. The multi-point computational analysis presented here may serve as a useful reference for subsequent investigations.

When the electron affinity is 4.0 eV, the PCE increases from 1.80% to 3.50%. Although the PCE improved, the interface losses remain significant due to the lack of HTL and ETL, particularly in hole extraction, ultimately leading to a reduction in efficiency. Therefore, the introduction of HTL and ETL plays a crucial role in optimizing band alignment, reducing recombination losses, and improving carrier transport efficiency, significantly enhancing the overall performance of the device.

4.2. The Impact of the Hole Transport Layer (ITO/β-AgGaO₂/HTL/Au)

To evaluate the effect of the HTL on photovoltaic performance, materials such as CFTS, CuI, P3HT, PSS, and CuSCN were tested. J–V characteristics in Figure 6 show that the choice of HTL significantly impacts the device performance.

CFTS showed the best performance, with a Jsc of 17.05 mA/cm² and a PCE of 8.10%, demonstrating excellent hole extraction and transport capabilities. According to previous reports, CFTS exhibits excellent thermal stability and could be compatible with low-temperature solution processing, making it a promising candidate for integration with oxide semiconductors in photovoltaic or photocatalytic applications [46]. However, due to the current lack of experimental studies on the direct interface formation between CFTS and β-AgGaO₂, further investigations are needed to verify the interface quality and assess the feasibility of practical device fabrication. Although CuSCN exhibited the highest Voc (1.07 V), its Jsc was low (5.88 mA/cm²), resulting in a PCE of 5.39%, indicating that the interface energy barrier limits charge transport. P3HT and CuI showed lower PCEs (around 4.6%) due to lower Jsc and Voc values. PSS exhibited slightly better performance with a Jsc of 8.85 mA/cm² and a PCE of 5.13%. The band diagram analysis in Figure 7 reveals that band alignment plays a critical role in performance.

The valence band of CFTS is well-matched with β-AgGaO₂, facilitating easy hole transport and low interface recombination losses, and forming an effective barrier in the conduction band to suppress electron leakage, ensuring high carrier selectivity. Although CuI has a higher Voc, the high valence band creates an energy barrier, leading to delayed charge extraction and increased recombination. The valence band misalignment in P3HT and PSS also limits hole extraction and collection efficiency. Although CuSCN’s valence band theoretically favors hole extraction, its low Jsc and PCE of 5.39% may be due to poor film quality, interface defects, or insufficient hole mobility. Both the interface defect density and the intrinsic defect density of β-AgGaO₂ were found to degrade the photovoltaic performance of the ITO/β-AgGaO₂/HTL/Au device structure. The corresponding simulation results are provided in Tables S1 and S2 of the Supporting Information. As shown in Figure 8 and Table 9, the results indicate that as the electron affinity decreases, the band alignment improves, which benefits hole transport, reduces interface recombination losses, and ultimately leads to higher photovoltaic efficiency.

In particular, if interface defects are ignored, the PCE could be further enhanced, with CFTS as the hole transport layer potentially reaching a PCE of 14.47%. Therefore, band alignment is crucial for improving performance, but material quality, defect control, and carrier mobility cannot be overlooked. CFTS shows the best overall performance, while CuI and CuSCN, despite ideal valence band positions, are limited by large energy barriers or material issues. P3HT and PSS are constrained by valence band misalignment and interface recombination.

4.3. The Impact of the Electron Transport Layer (ITO/ETL/β-AgGaO₂/Au)

Although the ETL is theoretically expected to improve performance, the devices using ETL materials such as C₆₀, WS₂, PCBM, ZnO, and TiO₂ did not demonstrate better performance than those without transport layers in Figure 9.

The Jsc for ZnO and TiO₂ was around 6.00 mA/cm², and the PCE was 1.80%, comparable to the performance of devices without transport layers, indicating that interface defects, poor carrier collection, or increased recombination paths still exist. C₆₀ and WS₂ performed worse, with Jsc values of 2.11 mA/cm² and 1.60 mA/cm², and PCEs of 0.58% and 0.43%, respectively. PCBM performed slightly better, with a Jsc of 2.58 mA/cm² and a PCE of 0.72%.

Band diagram analysis in Figure 10 shows that C₆₀ and WS₂ exhibit significant conduction band misalignment with β-AgGaO₂, creating large energy barriers that limit electron extraction.

Moreover, their valence band positions may cause hole accumulation and increased recombination, explaining their poor performance. PCBM shows better conduction band alignment but still presents an energy barrier that limits effective electron transport. ZnO and TiO₂ have good conduction band alignment but did not show significant performance improvements, possibly due to interface defects, carrier traps, or poor ETL film quality. Additionally, valence band misalignment could lead to hole accumulation and recombination, further limiting performance improvements. The interface defect density and intrinsic defect density of β-AgGaO₂ further reduce the photovoltaic performance of the ITO/ETL/β-AgGaO₂/Au device structure. The corresponding simulation results are provided in Tables S3 and S4 of the Supporting Information. Based on the calculations in Figure 11 and Table 9, it is evident that the introduction of ETL does not significantly improve the device’s photovoltaic efficiency under varying electron affinity conditions.

This suggests that future strategies to enhance ETL performance should focus on interface passivation, optimizing film quality, or using alternative materials with better alignment and superior transport properties.

5. Conclusions

In this work, we combine first-principles calculations with device-level simulations to systematically investigate the structural, electronic, optical, and photovoltaic properties of the ferroelectric polar ternary oxide β-AgGaO₂. Our density functional theory (DFT) calculations reveal that β-AgGaO₂ has an indirect band gap of 2.1 eV, strong absorption in the visible region, and favorable electron mobility. Polar optical phonon scattering is identified as the primary factor limiting carrier mobility. Nevertheless, optical modeling of thin-film photovoltaic devices predicts a theoretical power conversion efficiency (PCE) of up to 21%. Device simulations using SCAPS-1D indicate that incorporating a hole transport layer (HTL) substantially enhances device performance. Among the HTLs evaluated, CFTS was found to be the most effective, as it facilitates improved carrier extraction and suppresses interface recombination. Conversely, the addition of an electron transport layer did not improve performance, suggesting that further interface optimization or the exploration of alternative materials is necessary. Overall, our findings not only demonstrate the promising photovoltaic potential of β-AgGaO₂ but also underscore the critical role of energy band alignment and interface engineering in developing high-efficiency ferroelectric solar cells.