Mechanism of Temperature-Programmed Photoelectron Emission (TPPE) from Cu₂O/Cu Surfaces: The Role of Oxygen Vacancies in Photoredox Activation

Abstract

The performance of coatings, corrosion barriers, photocatalysts, and tribological materials is greatly influenced by in situ surface properties, requiring highly sensitive and reproducible operando surface characterization methods. We previously developed a temperature-programmed photoelectron emission (TPPE) method to clarify electron transfer behavior on light-irradiated metal surfaces. TPPE is sensitive to surface temperature and prior chemical exposure, which affect the total photoemitted electron count (N_T), the photothreshold, and the activation energy derived from Arrhenius plots of N_T obtained during heating–cooling cycles. This study examines the reproducibility of TPPE data and the TPPE mechanisms for Cu₂O/Cu surfaces subjected to mechanical abrasion, cleaning, plasma treatment, and subsequent immersion in organic liquids. The resulting Arrhenius plots reveal both positive and negative activation energies, depending on the treatment conditions. Negative activation energies during cooling are associated with photoredox-mediated emission. TPPE is attributed to oxygen vacancies within the Cu₂O surface layer, which is interfaced with metallic Cu, serving as a direct probe of these vacancy-related states. The TPPE characteristics (N_T intensity and activation energy) following exposure to various polar and nonpolar organic molecules (e.g., acetone, toluene, hexane, and ethanol) correlate with the electronic properties of these vacancies, consistent with previous observations for ambient air, alcohol, and water vapor exposure. Under illumination, Cu₂O vacancy states enhance photocarrier extraction (electrons and holes) and accelerate surface redox reactions within adsorbed thin films, thereby improving photocatalytic performance. Notably, the solvent’s reciprocal dielectric constant significantly influences TPPE, indicative of electrostatic surface–solvent interactions. Finally, the TPPE mechanism is discussed in the context of antiviral inactivation at the metallic copper–environment interface.

1. Introduction

In the field of surface chemical technology, significant attention has been directed toward the electron emission from processed solid surfaces, commonly referred to as exoemission [1,2]. This type of emission originates from surface-specific defects and contamination layers rather than pristine (ideal) surfaces. Electron transfer on metallic surfaces with thin oxide layers plays a vital role in photocatalysis, photoelectrochemical processes, sensors, and the inactivation of viruses and bacteria. However, the temperature dependence of electron transfer behavior in various chemical environments remains poorly understood. When engineered materials are handled at industrial production sites after surface processing, various compounds—oxygen, water vapor, and other environmental components—adsorb onto the surface. Consequently, further efforts are required to develop sensitive monitoring technologies capable of tracking these dynamic surface and interface changes under practical operating conditions [1].

Copper surfaces in particular are critical in various fields, including antibacterial/antiviral applications [3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18] and sensor technology [19,20,21,22]. Furthermore, the mechanisms underlying electron transfer on copper surfaces are central to studies on oxygen vacancies [23,24,25,26,27,28], photocatalysis [29,30,31], corrosion and adsorption [32,33,34,35,36,37,38,39,40,41], and photoelectrochemical processes [42,43,44,45,46]. In the present study, we focus on electronic defects formed within the Cu₂O/Cu matrix as the primary origin of the photoelectron emission (PE) observed in our experiments. Crucially, the inclusion of oxygen vacancies is justified in the title of this study because it is well-established that oxygen vacancies introduce discrete in-gap defect states in Cu₂O and modify the density of states near the Fermi level [24]. Therefore, it is highly probable that the observed temperature-programmed photoelectron emission (TPPE) characteristics are modulated by these vacancy-derived configurations, acting as a governing electronic factor. In a subsequent study, we plan to examine the relationship between TPPE characteristics (apparent PE intensity and multi-step activation energy) and models of oxygen vacancies.

In a previous study [2], we examined the TPPE data from two types of copper surfaces (Sample A, which was mechanically abraded in alcohols, water, or air, and Sample C, which was non-abraded, covered by a stable oxide layer, and only ultrasonically cleaned in the same liquids) by determining the temperature dependence (while heating and cooling in the range of 25–350 °C) of the total photoemitted electron count (N_T) for Samples A and C exposed to different environments (after pretreatment and immersion in alcohols or water). For both samples, the N_T values gradually increased with temperature, reached a maximum (N_Tmax) at 250 °C (the temperature of maximum emission, T_max), and decreased thereafter. Notably, the two samples exhibited opposing trends regarding the influence of the liquid environment. For Sample A, N_Tmax decreased in the order of H₂O > CH₃OH > C₂H₅OH > (CH₃)₂CHOH > C₃H₇OH, although the last alcohol yielded T_max = 100 °C, whereas Sample C followed the exact opposite order. This two-way TPPE behavior was attributed to acid–base interactions between the oxygen-rich surface overlayer and the adsorbed molecules.

Furthermore, our previous report [2] confirmed that the emission intensity strongly correlates with the surface hydroxyl (−OH) groups before the TPPE measurement and with the orientation of the electric dipoles formed by organic molecules adsorbed on the −OH groups. After the measurement, only the O²⁻ component appears, and the N_T signal disappears. Although the detailed mechanism of the −OH group adsorbed at the surface was not clarified, it is hypothesized to behave as a radical with high electron affinity. This may allow it to attract electrons from the base metal via tunneling, subsequently triggering Auger emission. Hence, the PE stimulation spectra are thought to originate from this Auger emission process.

TPPE has proven to be a powerful method for examining how thin adsorbed films affect the electronic properties of metallic surfaces. Supported by X-ray photoelectron spectroscopy (XPS) data, we previously reported that the TPPE mechanism involves the partial transformation of –OH into oxide oxygen (Cu₂O), accompanied by H₂O desorption. We hypothesize that these –OH groups possess radical-like properties, attracting electrons from the base metal via tunneling, followed by Auger emission, which leads to PE under irradiation. Although the mechanism of TPPE remains unclear, this process is considered a key factor in increasing TPPE intensity.

The present understanding of TPPE focuses on several factors: (a) the polarity of –OH groups adsorbed on the metal surface; (b) the electron density of the oxygen in those groups, which is influenced by electric potential and light intensity; (c) the orientation of adsorbate molecules due to acid–base interactions; (d) the formation of electric dipoles; and (e) changes in the O 1s spectra before and after TPPE measurements. This study aims to investigate the transition from hydroxyl components or adsorbed oxygen to Cu₂O in greater detail and to extend the analysis to metallic Cu surfaces with various organic adsorption layers, thereby further clarifying the electronic performance of Cu₂O/Cu in practical applications.

To achieve these objectives and present a significant conceptual improvement over the previous publication [2], this study expands the analysis to address the following key boundaries: (a) the reproducibility of emission profiles after distinct surface preparation pathways (mechanical abrasion, ultrasonic cleaning, and RF plasma pretreatments), followed by immersion in polar and nonpolar organic solvents; (b) the implementation of dynamic thermal scanning (heating and subsequent cooling cycles) coupled with optical profiling (wavelength scans) to measure and analyze the PE intensity from the surfaces, thus deriving the apparent activation energy and photoemission probability; (c) quantitative XPS O 1s deconvolution to map the real-time transformation between Cu₂O and Cu(OH)₂; and (d) the correlation between TPPE characteristics (N_T intensity and activation energy) and the chemical properties and the dielectric constant of adsorbed organic molecules, especially regarding acid–base interactions and surface redox reactions.

2. Materials and Methods

2.1. Materials

Figure 1 illustrates the experimental procedure, resulting in nine sample groups, designated as environments (1)–(9). Rolled copper sheets (0.1 mm thickness; 99.9% purity) (Nilaco Corporation, Tokyo, Japan) were used as substrates, with sample dimensions of 20 mm × 30 mm for TPPE measurements and 3 mm × 3 mm for XPS.

Samples were prepared for TPPE and XPS measurements using a sequential treatment protocol. For each sample test group (hereafter identified based on their respective treatment and solvent exposure, as numbered below), metallic copper substrates were subjected to a consecutive series of three pretreatments: (1) mechanical abrasion with emery paper (Struers K.K. (Marumoto Struers), #3000, Tokyo, Japan) in ambient air; (2) ultrasonic cleaning (Branson 1210J, Yamato Scientific, Tokyo, Japan) in acetone (30 mL) for 30 min, followed by drying under vacuum (<1 × 10⁻³ Torr) for 10 min; and (3) RF plasma treatment (13.56 MHz, 30 W) under flowing argon (99.9995% purity) in a Pyrex glass chamber for 5 min at room temperature, which represents a typical cleaning process and ensures that surface contaminants (e.g., organic molecules) are removed. After pretreatment, samples were individually immersed in one of six reagent-grade solvents (Fujifilm Wako Pure Chemical Corp., Osaka, Japan) for 1 h at room temperature: (4) aniline (C₆H₅NH₂), (5) cyclohexane (C₆H₁₂), (6) benzene (C₆H₆), (7) toluene (C₆H₅CH₃), (8) hexane (C₆H₁₄), and (9) ethanol (C₂H₅OH). Following immersion, the samples were dried under vacuum for 10 min.

Ultimately, environments (4)–(9) refer to samples that underwent all three pretreatments, followed by immersion in one solvent (see Figure 1). Environment (2) is a notable exception to the standard protocol; these samples were subjected to ultrasonic cleaning without prior emery paper abrasion or subsequent treatment.

It should be noted that the durations of mechanical abrasion, plasma pretreatment, and solvent immersion cycles can significantly influence the resulting TPPE profiles. In the previous paper [2], we explored the effect of mechanical friction—such as surface abrasion driven by a rotating iron screw in selected environments—on the PE stimulation spectra. As abrasion time increased, the PE intensity decreased, reducing the surface’s reactivity with the environment. However, baseline control profiles from these historical benchmarks demonstrate that while mechanical abrasion modifies macro-roughness parameters, subsequent plasma exposure and low-temperature solvent immersion primarily alter the surface chemistry and surface energy. These treatments do not fundamentally modify the established micro-roughness or structural topography of the underlying copper matrix. This stability suggests that the order-of-magnitude variations observed in our TPPE data—particularly the sharp contrast between polar and nonpolar solvent behaviors—originate predominantly from electronic dipole interactions and chemical modifications rather than morphological alterations.

2.2. TPPE and XPS Measurements

Figure 2 illustrates the TPPE measurement apparatus (adapted from our previous work [2]). The system comprises a UV light source (Hamamatsu Photonics, D₂ lamp, L613, Hamamatsu, Japan), a monochromator, an electron counter with an anode set at 1400 V, and a sample holder equipped with a Nichrome wire heater. To accelerate and collect emitted electrons, a −94 V potential relative to the grounded grid (marked by a dotted line in Figure 2) was applied to the holder using a battery. The negative potential acts similarly to cathodic reduction in corrosion processes. Q gas (99% He and ~1% isobutane (iso-C₄H₁₀)) was used as the counter gas, with a flow rate of ~100 bubbles min⁻¹. The TPPE measurement was performed as follows. First, the pretreated sample was secured in the chamber using the sample holder. Then, the counter gas (99% He and 1% isobutane) was continuously flown through the cell at atmospheric pressure to maintain a stable, uniform environment. Once environmental equilibrium was established, the PE intensity was measured at the designated temperatures. Note that the sample was constantly in contact with the counter gas during the TPPE measurement under heating and subsequent cooling.

The PE intensity (counts min⁻¹) was measured by scanning the incident light wavelength from 300 to 170 nm at 20 nm min⁻¹. A 0.5 mm × 5 mm spot was illuminated on the sample using a D₂ lamp, with a mean incident light intensity of 102 nW for wavelengths between 240 and 200 nm. The resulting plot of PE intensity versus incident light wavelength is defined as the PE stimulation spectrum. For each sample, PE stimulation spectra were collected at 50 or 100 °C intervals during the heating (up scan) and cooling (down scan) cycles.

The XPS spectra—including Cu 2p, O 1s, C 1s, Cu_LMM, Cu 3p, and N 1s (in aniline)—were recorded before and after TPPE measurements using a Shimadzu ESCA 750 spectrometer (Kyoto, Japan) equipped with a Mg Kα X-ray source (8 kV and 30 mA) and an ESCAPAC 760 analyzer (Kyoto, Japan). To determine the elemental compositions from the measured spectra, peak areas were integrated over specific energy ranges and quantified using the relative sensitivity factors provided in the spectrometer’s operating manual: Cu 2p (965–925 eV, 24.1), O 1s (540–528 eV, 2.9), C 1s (300–280 eV, 1.0), Cu_LMM (350–325 eV), Cu 3p (82–70 eV), and N 1s (410–392 eV). Binding energies were calibrated and assigned by setting the main C 1s hydrocarbon component to 285 eV. XPS measurements were performed before and after the TPPE measurements. Samples were briefly exposed to ambient air (for less than 5 min) during transfer between the TPPE measurement apparatus and the XPS apparatus, in addition to the time used for drying (after immersion in solvents). The effects of ambient air and water adsorption on TPPE can be removed as background data [2], as described later in Section 3.2.

Overall, we aim to further clarify the role of Cu(OH)₂ (including surface hydroxyl groups) and Cu₂O. To achieve this, O 1s spectra are deconvoluted into two oxygen species, Cu(OH)₂ and Cu₂O, both before and after TPPE measurements. The analysis focused on quantifying these components for each environmental condition, determining the dependence of peak heights on the specific solvents used, and examining the relationship between Cu(OH)₂ and Cu₂O intensities before and after TPPE measurement.

2.3. Data Analysis and Theoretical Framework

The analysis of the TPPE characteristics is based on two modified formulas. The first one is an Arrhenius-type equation used to determine the activation energy [2,47].

N_T = A_0,N,T exp (−ΔE_NT/k_BT)

(1)

where N_T is the total count of photoemitted electrons obtained from the 6.5 min wavelength scan (300→170 nm) performed at specific temperatures during the heating (up scan) and cooling (down scan) processes. A_0,N,T represents the pre-exponential factor (expressed as the count of photoelectrons emitted in the given measurement period), ΔE_NT is the activation energy, k_B is the Boltzmann constant, and T is the absolute temperature.

The second formula characterizes the PE stimulation spectrum as a function of the incident light wavelength. Based on the Fowler–DuBridge theory [48,49,50,51] for clean surfaces, the photoelectric current (I) is expressed as follows:

I = αAT²f(x)

(2)

where α is a dimensionless proportionality factor related to the probability of an electron absorbing a quantum of the incident light, A is the Richardson constant (1.20 × 10⁶ Am⁻²K⁻²), and T is the absolute temperature in Kelvin. Note that αA is referred to as the electron photoemission probability herein. Furthermore, f(x) represents the universal function of the parameter x ≡ (hν − ϕ)/k_BT, which is given by the following equation:

f(x) = π²/6 + x²/2 − (e^−x − e^−2x/2² + e^−3x/3² − …) for x ≥ 0

(3)

where h is Planck’s constant, ν is the frequency of the incident light, ϕ is the photothreshold of the metal, k_B is Boltzmann’s constant, and hν is the photon energy of the incident light. Using the notation Y_FD and E_p instead of I and hν in Equation (2), respectively, and replacing Equation (3) with x²/2 as an approximation, we derive an approximate equation for the photoemission:

Y_FD = αA(E_p − ϕ)²/2k_B²

(4)

where Y_FD is the quantum yield. In the present experiments, we applied the Fowler–DuBridge formalism to oxidized Cu₂O/Cu surfaces. It must be noted that the values of αA and ϕ obtained through this fitting are derived from composite surfaces containing intrinsic defect states and structural oxide overlayers, rather than pristine, ultra-high-vacuum-cleaned Cu metal. Consequently, our calculated β parameter does not represent an absolute or intrinsic constant of the substrate. Instead, it serves as an effective, apparent PE descriptor that captures the qualitative behavioral modifications induced by the overlayer.

Physically, treating the thin Cu₂O oxide and its corresponding molecular adsorbate film as a single electronic filter or tunneling barrier means that while the overlayer severely attenuates, scatters, and electrostatically screens the escaping electron flux, it functions mathematically as a linear scaling operator on the total emission intensity. Because the underlying quadratic dependency of the quantum yield near the photothreshold originates fundamentally from the excitation density of states within the base metal matrix, the structural deviations of the film do not invalidate the core scaling relationship of Equations (2)–(4). Rather than yielding an absolute physical determination of the true quantum probability, the Fowler–DuBridge square-root relationship is employed as a standardized phenomenological basis to compare relative shifts in the apparent photoemission capacity and kinetic barriers across all 9 environmental conditions.

The electron emission intensity I_W (counts min⁻¹) in the PE stimulation spectra corresponds to the quantum yield for the given photon energy (E_P) during the incident light wavelength scan (300→170 nm). Plotting the square root of I_W against E_p results in a straight line, providing a slope, which represents β = (αA/2k_B²)^1/2, and the characteristic values of ϕ for the PE stimulation spectra. The values are obtained by fitting the PE data of the electron emission intensity as a function of E_p:

(I_W)^1/2 = (αA/2k_B²)^1/2(E_p − ϕ) = β(E_p − ϕ).

(5)

Note that ΔE_NT depends on the temperature, whereas β depends on the light wavelength. Overall, ΔE_NT refers to the activation energy needed for photoelectrons excited in the base Cu metal to transfer to the surface via the adsorbate overlayer, and β may be related to the ability of the electrons in the metal to absorb the photon energy irradiated through the adsorbate layer.

Using the data obtained by Equations (1) and (5), the relationship between the TPPE characteristics and the adsorbate properties is analyzed based on (a) the I_W values, which likely relate to the transport distance of electrons released from the base metal; (b) the threshold and peak photon energies needed to initiate and maximize I_W; (c) the activation energies (ΔE_N,T) derived from N_T during heating and cooling, using mean values (N_T,M); (d) the relation between heating and cooling activation energies; (e) the relationship between the cumulative N_T,M during heating and cooling and the peak intensity (N_T,Mmax); (f) the dependence of ΔE_N,T,M and β on the dielectric constant (ε/ε₀) of the solvent [52]; (g) the influence of chemical properties, such as the ionization energy (IE), Pearson’s absolute hardness (η) [53], and Gutmann’s acceptor number (AN), of the solvent molecules, and (h) the PE mechanisms for surface redox reactions of adsorbed water and organic molecules on Cu₂O/Cu. These parameters are critical for understanding the reducibility of the metal, a primary factor in multiple applications such as the contact killing of viruses by metallic copper surfaces [9].

3. Results

3.1. PE Stimulation Spectra and Analysis

Figure S1a,b in the Supplementary Materials present the PE stimulation spectra (wavelength range of 300–170 nm) obtained at selected temperatures during heating (up scan, 25→350 °C) and subsequent cooling (down scan, 350→25 °C) for environments (3)–(9). Note that the recorded PE intensity includes the background count level. Figure 3 shows the data for environment (9).

Using (9) ethanol during the heating cycle as an example (Figure 3a), the PE intensity progressively increases and then decreases with decreasing wavelength. This decline is due to the attenuation of UV light by air within the optical path. The emission peak consistently appears at approximately 205 nm. Furthermore, the emission intensity of the peak increases with temperature, reaching a maximum of 5.5 × 10³ count min⁻¹ (I_Wmax,Up) at the temperature of 250 °C (T_Wmax), before decreasing at higher temperatures. Additionally, the peak exhibits a slight blueshift at 350 °C, compared with the peak wavelength at 50 °C. Conversely, during the cooling cycle (Figure 3b), the PE intensity increases with decreasing temperature, reaching a maximum intensity of 5.0 × 10³ count min⁻¹ (I_Wmax,Down) at 50 °C. The peak initially appears between 200 and 210 nm and displays a slight redshift at 50 °C, compared with the peak wavelength at 350 °C.

To obtain the TPPE characteristics, the square root of the PE intensity, I_W, was plotted against the incident photon energy, E_p. Figure 4 shows representative results for (4) aniline at 150 °C during heating and (9) ethanol at 50 °C during cooling.

Table 1. Characteristics of the PE stimulation spectra peaks obtained during heating (up scan, 25→350 °C) and subsequent cooling (down scan, 350→25 °C).

Environment	Sample No. *	Scan Direction	IWmax/103 Count min−1	TWmax/°C	λWmax/nm	ϕWmax/eV	ϕ/eV	b/ (Count min−1)1/2 eV−1
(3) Ar plasma	69	Up	11.0	50	208	5.97	5.10	108.0
		Down	5.5	50	205	6.05	5.30	84.8
(4) Aniline	85	Up	12.0	150	210	5.91	5.00	103.4
		Down	5.5	50	208	5.97	5.20	67.9
(5) Cyclohexane	72	Up	11.0	150	208	5.97	5.10	112.6
		Down	5.5	50	208	5.97	5.30	90.8
(6) Benzene	70	Up	8.0	150	208	5.97	5.10	87.5
		Down	4.5	50	205	6.05	5.25	67.5
(7) Toluene	89	Up	7.0	200	205	6.05	5.15	77.3
		Down	3.5	50	205	6.05	5.30	69.8
(8) Hexane	76	Up	8.8	200	208	5.97	5.10	86.0
		Down	7.0	50	206	6.02	5.25	97.1
(9) Ethanol	92	Up	5.5	250	205	6.05	5.10	65.3
		Down	5.0	50	205	6.05	5.35	89.6

* The sample no. refers to the samples specified in Table S1. The plots of N_T value versus measurement temperature for each sample were very close to the mean N_T,M value in Table S2.

summarizes the resulting data—including the peak intensity (I_Wmax), peak temperature (T_Wmax), peak wavelength (λ_Wmax), peak photon energy (ϕ_Wmax), f, and b—for environments (3)–(9) during heating and cooling. An analysis of the PE stimulation spectra and their relationships with TPPE characteristics is described in Section 3.7.

3.2. TPPE Plots and Reproducibility

Figure S2 in the Supplementary Materials shows the plots of N_T versus temperature during heating and cooling for all copper samples, referred to as the TPPE plots. To ensure reproducibility and data reliability, we carried out TPPE measurements for two or three samples in each environment: n = 3 for (1) emery paper abrasion, n = 2 for (2) acetone, n = 3 for (3) Ar plasma treatment, n = 3 for (4) aniline, n = 2 for (5) cyclohexane, n = 3 for (6) benzene, n = 3 for (7) toluene, n = 3 for (8) hexane, and n = 3 for (9) ethanol. The corrected N_T values were obtained by subtracting the natural background count (1000 counts).

Regarding the potential influence of brief ambient air exposure during sample transfer before XPS and TPPE measurements, we consider the benchmarks established in our prior work. Specifically, as documented in Figure 3(1) (H₂O after abrasion) and 3(6) (ambient air after abrasion) of the previous paper [2], the N_T values recorded at 25 °C were 0.59 × 10⁴ and 0.48 × 10⁴ counts, respectively. Because these initial values are remarkably close, the electronic effect of short-term atmospheric adsorption during the sample transfer step is considered to be nearly identical to that of an intentionally introduced water film, even though the N_T profiles diverge as the temperature is subsequently increased. Furthermore, this ambient transfer window was kept identical and brief across all nine experimental conditions to ensure that any minor atmospheric contamination acts as a constant, uniform baseline rather than a variable error.

Distinct behaviors are observed in each environment during heating. Generally, N_T remains low in the initial stage, increases with increasing temperature, reaches a maximum (N_Tmax) at a specific temperature (T_max), and then decreases. Notably, samples subjected only to (1) emery paper abrasion show relatively low, temperature-independent N_T values. In contrast, (3) Ar plasma treatment samples exhibit high N_T values at low temperature, peaking at 50 °C and then gradually decreasing. Samples exposed to environments (2) and (4)–(9) produced broad, high-intensity peaks, with significantly varied N_Tmax values at T_max. During the subsequent cooling cycle (down scan); however, the N_T values increase gradually with decreasing temperature for all environments.

Table S1 in the Supplementary Materials summarizes the TPPE characteristics, including the sum of NT, NTmax, and Tmax during the heating cycle. Activation energies (ΔENT) and pre-exponential factors (A0,N,T) were derived from Arrhenius plots (lnNT vs. 1/T) based on Equation (1). In the heating cycle, these plots typically revealed two distinct linear regions (the first and second steps) in most environments, except for (2) and (9), which exhibited a single linear relationship. Cooling cycle plots consistently showed a single linear region. The reproducibility of the TPPE data was confirmed by determining the mean and standard deviation (SD) for each environment; SD values for the sum of NT ranged from 0.25 to 2.64, while SD for the first-step ΔENT (up scan) ranged from 0.026 to 0.091. The averaged characteristics for the mean NT values (NT,M) are further detailed in Table S2.

3.3. Averaged TPPE and Arrhenius Plots

Figure 5 presents the averaged TPPE plots, showing the N_T,M values versus temperature during heating and cooling for all copper sample pretreatments and environments. N_T,M represents the mean N_T value recorded at each measurement temperature during heating (up scan: 25→350 °C) and subsequent cooling (down scan: 350→25 °C).

The TPPE characteristics are summarized as follows, according to the environment:

Pretreatments (1)–(3). For environment (1), N_T,M remains nearly constant during both heating and cooling. For environment (2), N_T,M initially remains constant during heating, but increases and reaches a broad maximum (N_T,Mmax) at 250 °C (T_max,M) before declining. For environment (3), N_T,M remains high during heating at low temperatures, reaches a maximum (N_T,Mmax) at 50 °C (T_max,M), as the temperature rises, and then gradually decreases. For both (2) and (3), N_T,M progressively increases with decreasing temperature during cooling.

Organic Solvents (4)–(7). During heating, N_T,M is initially low, then increases to a maximum (N_T,Mmax) at T_max,M, and then declines. Notably, (4) C₆H₅NH₂ exhibited a broad, high-intensity peak between 100 and 300 °C. This sustained emission is attributed to a strong acid–base reaction between the amino group (−NH₂) and surface hydroxyls: C₆H₅NH₂ + CuOH → C₆H₅NH₃⁺ + Cu–O⁻, producing a strong surface dipole with the positive pole oriented outward.

Organic Solvents (8) and (9). These samples followed a similar heating trend. However, (9) C₂H₅OH produces the highest T_max,M (approximately 250 °C), suggesting stronger bonding to the Cu surface than the other solvents. The hydroxyl group of the C₂H₅OH molecule is likely hydrogen-bonded to the oxygen atom of CuOH on the Cu surface, generating a surface dipole with the negative pole oriented outward.

All samples follow a similar trend during cooling. N_T,M gradually increases with decreasing temperature and reaches a maximum at 25 °C. The N_T,M intensities and DE_NT values shown in Table S2 serve as the basis for subsequent discussion.

Figure 6 shows the Arrhenius plots (lnN_T,M vs. 1/T) during heating (squares, shown in red) and cooling (circles, shown in blue) for environments (4) and (9). Figure S3 shows the plots for the environments (1)–(3) and (5)–(8). For environments (1) and (3), and (5)–(8), the heating plots can clearly be divided into two stages (first and second steps) as the temperature rises. In contrast, environments (2) and (9) exhibit only a single stage during heating. During cooling, however, environments (1)–(9) all display consistent linear relationships over the temperature range. The bifurcation of the heating plots into two steps corresponds to the appearance of broad N_T,Mmax peaks in the TPPE plots (Figure 5). These thermal characteristics are detailed in Table S2.

3.4. XPS Spectra and O 1s Curve Resolution

XPS spectra recorded before and after TPPE measurements are shown in Figure S4 for environments (2), (3), (5), (6), (7), and (8) and in Figure 7 for environments (4) and (9). Assignments for the copper and oxygen species are based on previous work [54,55,56,57,58,59,60]. The binding energies for the oxidation states of copper are 932.5 eV (Cu⁰ and Cu⁺) and 933.5 eV (Cu²⁺) for Cu 2p_3/2, and 335 eV (Cu⁰) and 337 eV (Cu⁺ and Cu²⁺) for Cu_LMM. A characteristic shake-up satellite structure is observed 6–10 eV above the main Cu 2p peak in Cu²⁺ compounds.

Regarding the O 1s region, the peak at 529.9 eV is assigned to lattice oxygen bonded to copper, and the peak at 532.5 eV is ascribed to adsorbed species, including OH, Cu(OH)₂, and H₂O [58,59]. Considering that the copper samples were exposed to air during transfer, these higher-binding-energy components are expected across all pretreated surfaces. According to a recent study [60] on tarnished copper plates (C1220, Cu purity > 99.9%), the binding energies are 932.6 eV (Cu), 932.7 eV (Cu₂O), and 933.1 eV (CuO) for the Cu 2p peak; 943–948 eV (Cu₂O) and 940–945 eV (CuO) for the satellite peak; and 531.7 eV (Cu (OH)₂), 530.7 eV (Cu₂O), and 529.8 eV (CuO) for the O 1s peak.

Detailed views of the O 1s spectra before and after TPPE are provided in Figure S5 for environments (2), (3), (5), (6), (7), and (8) and in Figure 8 for environments (4) and (9).

Table 2. Characteristics of curve-resolved components of O1s spectra: Cu(OH)₂ and Cu₂O before and after TPPE measurement for copper samples subjected to environments (2)–(9).

Environment	Before TPPE Measurement						After TPPE Measurement
	Cu2O			Cu2O			Cu(OH)2			Cu2O
	Peak Binding Energy/eV	SD *	Peak Height	Peak Binding Energy/eV	SD *	Peak Height	Peak Binding Energy/eV	SD *	Peak Height	Peak Binding Energy/eV	SD *	Peak Height
(2) Ultrasonicate	532.2	1.07	42.1	530.0	0.57	1.8	531.7	1.10	21.8	530.9	0.45	11.8
(3) Ar plasma	532.0	1.12	58.3	531.2	0.44	14.5	531.9	1.02	44.7	531.1	0.44	26.5
(4) Aniline	531.8	1.13	52.5	531.0	0.43	8.3	532.0	1.12	41.0	531.1	0.44	23.6
(5) Cyclohexane	532.1	1.15	35.8	531.0	0.42	10.1	531.8	1.05	25.6	530.9	0.43	15.8
(6) Benzene	532.2	1.24	40.2	531.1	0.42	11.3	531.9	1.08	29.6	531.0	0.44	16.6
(7) Toluene	532.1	1.07	53.0	531.0	0.43	5.8	532.2	1.05	51.5	531.2	0.47	27.5
(8) Hexane	532.1	1.07	64.4	531.0	0.43	7.3	531.9	1.02	41.1	531.1	0.44	24.4
(9) Ethanol	532.2	1.14	58.3	531.1	0.45	6.7	531.9	1.02	43.5	531.1	0.44	25.7
Average	532.1	1.12	50.6	530.9	0.45	8.2	531.9	1.06	37.4	531.0	0.44	21.5

* SD, standard deviation.

quantifies the peak-height changes for both oxygen species in each environment.

The XPS spectra for each environment are summarized as follows:

(2) Acetone Cleaning. Before TPPE, the Cu_LMM spectrum shows peaks corresponding to both oxide Cu₂O (~337 eV) and metallic Cu (~335 eV) with weak intensities. The O 1s spectrum shows one distinct component, adsorbed oxygen (Cu (OH)₂) at ~532.5 eV. After TPPE, the metallic Cu peak decreases, whereas the Cu₂O peak increases. Correspondingly, the adsorbed oxygen signal diminishes, and the oxide oxygen (Cu₂O: 530.7 eV) becomes prominent. This change can be attributed to thermal oxidation, which correlates with the observed decrease in N_T,M beyond T_max,M (Figure 5). After TPPE, the Cu 2p spectrum also shows the growth of peaks attributable to Cu₂O and CuO (~940–945 and ~934 eV, respectively). Moreover, high initial C 1s levels suggest that residual organic contamination is present after ultrasonic cleaning in acetone but decreases after TPPE.

(3) Ar Plasma Treatment. Before TPPE, the Cu_LMM spectrum exhibits both Cu₂O (~337 eV) and metallic Cu (~335 eV) features. The O 1s spectrum displays a mix of adsorbed oxygen (Cu(OH)₂: ~532 eV) and oxide oxygen (Cu₂O: ~531 eV). After TPPE, the metallic Cu peak in the Cu_LMM spectrum decreases, while the Cu₂O peak increases. Simultaneously, the adsorbed oxygen component in the O 1s spectrum diminishes, and the oxide oxygen component becomes more prominent. Although this change is due to an increase in temperature during TPPE measurement, the decrease in N_T,M beyond 50 °C (Figure 5) is a distinctive feature of samples subjected to Ar plasma after ultrasonic cleaning. The Cu 2p spectrum also shows peaks attributable to Cu₂O (~940–945 eV) and CuO (~934 eV) before and after TPPE. The C 1s spectrum remains largely unchanged before and after TPPE, indicating minimal organic contamination.

(4) Aniline Adsorption. Metallic Cu remains observable even after TPPE, suggesting that the surface oxide overlayer is relatively thin. Before TPPE, the O 1s spectrum shows both adsorbed and oxide oxygen species, giving a slightly lower proportion of adsorbed oxygen than (8) hexane, (9) ethanol, and (7) toluene. Before TPPE, a distinct N 1s peak (~400 eV) from adsorbed aniline can be observed; after TPPE, the signal is reduced. The C 1s peak also decreases after TPPE, indicating aniline desorption.

(5) Cyclohexane and (6) Benzene Adsorption. Before TPPE, both environments show a significantly lower proportion of adsorbed oxygen than that observed for (8) hexane, (9) ethanol, and (7) toluene. After TPPE, surface oxidation is observed, and the C 1s peak slightly decreases, suggesting the desorption of organic molecules.

(7) Toluene Adsorption. Before TPPE, the adsorbed oxygen peak is higher than the oxide peak, and after TPPE, the oxide oxygen grows, similar to that for (4) aniline, (8) hexane, and (9) ethanol. Cu_LMM spectra confirm surface oxidation after TPPE, although metallic Cu remains slightly, and the C 1s peak decreases, indicating toluene desorption.

(8) Hexane and (9) Ethanol Adsorption. These environments exhibit higher intensities for adsorbed oxygen relative to oxide oxygen, similar to that of (7) toluene. This contrasts with the O 1s spectra observed for (5) cyclohexane, (6) benzene, and (4) aniline. In the case of ethanol, the Cu_LMM spectrum exhibits a relatively high metallic Cu peak before TPPE. After TPPE, both environments show characteristic surface oxidation and a decrease in C 1s levels, consistent with previous observations.

The O 1s spectra were deconvoluted into two primary components, adsorbed oxygen (Cu(OH)₂ and H₂O) and oxide oxygen (O²⁻ from Cu₂O), by fitting to Gaussian functions using the non-linear least-squares method in Microsoft Excel Solver, as previously described [61]. These two primary components were selected for deconvolution because in the XPS O 1s measurement, a distinct, molecular H₂O component at the highest binding energy could not be unambiguously distinguished across all samples due to peak broadening and spectral overlap. Therefore, to maintain analytical consistency across all datasets, this molecularly adsorbed water component was systematically incorporated into the broader higher-binding-energy envelope assigned to the adsorbed oxygen/hydroxyl species. Note that the formation of a Cu(I)-OH_ad−H₂O_ad complex is considered essential for lowering the energy barrier of water dissociation [37]. Figure 8 and Figure S5 show the observed O 1s spectra alongside the curve-resolved components for environments (2)–(9) before and after TPPE. The averaged binding energies and peak heights for these components are listed in Table 2.

On average, the O 1s peak of adsorbed oxygen (Cu(OH)₂) was located at 532.1 eV with a height of 50.6 before TPPE, shifting to 531.9 eV and 37.4 after TPPE, respectively. Conversely, the oxide oxygen (Cu₂O) peak shifted from 530.9 eV (height: 8.2) before TPPE to 531.0 eV (height: 21.5) after TPPE. These shifts confirm that TPPE, specifically the heating cycle, promotes the transformation of surface hydroxyls into lattice oxides. This chemical transition is likely responsible for the broad N_T,M peaks observed in the up scans and the subsequent increase in N_T,M during the down scans (Figure 5). The binding energies of Cu(OH)₂ and Cu₂O remained relatively constant around the averages before and after TPPE.

In our previous report [2], it was unclear why this transition leads to an increase in PE intensity. Herein, to explore environment-specific effects, we examined the relative peak heights of Cu(OH)₂ and Cu₂O across environments (2)–(9), as shown in Figure 9a–c.

Figure 9a shows that while the Cu(OH)₂ intensity generally decreases owing to the thermal desorption of H₂O and the growth of Cu₂O, a significant amount of adsorbed oxygen is still present or regenerated after TPPE. Figure 9b reveals an inverse relationship; environments with lower initial Cu₂O levels exhibit a greater capacity for oxide growth during the measurement. Specifically, Figure 9c shows that (7) toluene and (9) ethanol have a higher ability to enhance both Cu₂O and Cu(OH)₂ than (2) acetone, (5) cyclohexane, and (6) benzene.

Finally, the surface compositions were analyzed via the O 1s/Cu 2p and C 1s/Cu 2p ratios before and after TPPE measurement for environments (3)–(9) (Figure 10). Before TPPE, the C 1s/Cu 2p ratios are significantly higher in the organic environments than in (3) Ar plasma treatment, likely due to the high surface energy and activation after plasma exposure [1]. After TPPE, the C 1s/Cu 2p ratios sharply decrease across all environments. This is attributed to the thermal desorption of adsorbed organic molecules, hydroxyl groups, and water, particularly in the cases of (7) toluene and (8) hexane.

3.5. Chemical Properties of the Different Environments

To elucidate the dependence of the TPPE characteristics (N_T,M intensity and PE energy barrier) for the Cu(OH)_x/CuO/Cu₂O/Cu surface on the chemical properties of the environment, we analyzed the molecules of the solvents used in this study. Following Huheey’s classification [62], the species and compounds are categorized according to the hard and soft acid–base theory. This study involves borderline (Cu²⁺) and soft (Cu⁺, metallic Cu) acids, as well as hard (H₂O, OH⁻, O²⁻, ROH), borderline (C₆H₅NH₂), and soft (C₆H₆) bases. According to Pearson [63], hard acids preferentially bind to hard bases, and soft acids to soft bases. Furthermore, hard acid–base interactions are predominantly electrostatic, whereas soft acid–base interactions are predominantly covalent.

Table 3. Chemical properties of the liquid environments.

Environment	IE a/eV (Gas)	AN b	ε/ε0 (20–25 °C) c	(ε/ε0)−1
Water H2O d	12.62	54.8	78.39	0.0128
(2) Acetone (CH3)2CO	9.71	12.5	20.70	0.0483
(4) Aniline C6H5NH2	7.72	-	6.89	0.1451
(5) Cyclohexane C6H12	9.90	-	2.023	0.4943
(6) Benzene C6H6	9.24	8.2	2.275	0.4396
(7) Toluene C6H5CH3	8.82	-	2.379	0.4203
(8) Hexane C6H14	10.15	-	1.8799	0.5319
(9) Ethanol C2H5OH	10.54	37.1	24.55	0.0407

^a IE: Experimental gas-phase ionization energies [64]. ^b AN: Gutmann’s acceptor number [65]. ^c Values obtained from a previous study [52]. ε/ε₀: dielectric constant. ^d Water forms as a thin film on the metal surface exposed to ambient air [32].

lists the chemical properties—IE [64], AN [65], and ε/ε₀ [52].

Pearson’s absolute hardness (defined as η = (IE − EA)/2) measures a species’ resistance to charge transfer or changes in electron density. In this work, η is not computed because experimental EA values are unavailable; instead, the trend in hardness is discussed based on the experimental IE. The absolute hardness trend based on IE (largest to smallest) is water—hardest > ethanol > hexane > cyclohexane > acetone > benzene > toluene > aniline—softest. This trend matches HSAB expectations: water is very hard, aliphatic alcohols and alkanes are hard, and aromatic and nitrogen-containing molecules are softer. Hard solvents (water, ethanol, and hexane) interact differently than softer ones (cyclohexane, acetone, benzene, toluene, and hexane). In HSAB terms, hard solvents generally have larger η, whereas borderline solvents (aniline) and soft solvents exhibit intermediate to smaller η values. AN quantifies a solvent’s propensity to accept electrons (Lewis acidity). Solvents with high AN (water and ethanol) act as strong electron acceptors from the oxygen lone pair of surface hydroxyl groups, owing to hydrogen bonding and polarity. Acetone has moderate acceptor ability because of its carbonyl group, whereas cyclohexane, benzene, toluene, and hexane are nonpolar or aromatic, resulting in negligible electron acceptor behavior.

The reciprocal dielectric constant (1/(ε/ε₀)) serves as an inverse index of solvent polarity and electron screening capacity. Figure S6a–c illustrates the complex relationships among IE, AN, and 1/(ε/ε₀). For the polar compounds, 1/(ε/ε₀) decreases with IE and AN, and for nonpolar groups, 1/(ε/ε₀) increases with IE. As shown in Figure 11, the solvents can be clearly separated into two groups: lower 1/(ε/ε₀) with high polarity (water, ethanol, acetone, and aniline) and higher 1/(ε/ε₀) with low polarity (hexane, cyclohexane, benzene, and toluene).

Generally, more polar solvents (high ε/ε₀, low 1/(ε/ε₀)) are expected to enhance surface polarization and screen charges more efficiently, thereby lowering the energy barrier for electron emission and increasing intensities or reducing photothresholds. Conversely, nonpolar solvents (low ε/ε₀, high 1/(ε/ε₀)) typically provide less charge screening, sometimes resulting in higher photothresholds or kinetic barriers. However, our TPPE data do not align with these standard predictions.

As previously reported [2], the effects of AN and 1/(ε/ε₀) on the TPPE significantly vary depending on the initial state (treatment) of the metal surface (e.g., mechanically abraded vs. ultrasonically cleaned Cu surfaces). The Cu surfaces in the present experiment behave similarly to those in Sample C [2], in which hard, high-AN solvents (water, ethanol, and acetone) interact strongly with the thin oxide layer via hydrogen bonds and dipole interactions, thereby affecting the surface potential and PE. In contrast, soft, nonpolar solvents (cyclohexane, benzene, toluene, and hexane) primarily interact through van der Waals forces, inducing minimal changes in the surface properties.

To fully understand these variations and address the limitations of macroscopic descriptors, microscopic molecular properties—such as gas-phase molecular dipole moments, steric hindrance, and adsorption orientations—must be considered. While a bulk metric like the dielectric constant suggests a uniform electrostatic behavior within each group, the local surface dipole is ultimately dictated by the net orientation of the chemisorbed or physisorbed molecules on the Cu(OH)₂/Cu₂O matrix. For instance, aromatic bases such as benzene and toluene can adopt a flat or tilted orientation relative to the surface plane, depending on the presence of co-adsorbed species or surface defects. A flat alignment minimizes the normal component of the molecular dipole, offering localized charge screening without significantly decreasing the work function. Conversely, sterically unhindered aliphatic chains, such as hexane, interact strictly via weak, isotropic dispersion forces, creating a loose overlayer that fails to establish an ordered interfacial dipole array. This microscale disparity explains why macroscale dielectric correlations exhibit specific outliers: the geometry of the first adsorption monolayer alters the electron tunneling probability independently of the bulk solvent polarity. The detailed correlation between the TPPE characteristics and these chemical properties is discussed in the following section.

3.6. Correlation Analysis: TPPE Characteristics, Dielectric Constants, and Oxygen Species

Figure 12a–d and Figure S7e–l show the correlation between N_T,M and the energy barrier and pre-exponential factor obtained during heating and cooling, and further relates them to the dielectric constants of the solvents and the concentrations of surface oxygen species obtained from the Cu(OH)₂ and Cu₂O peak heights (Table S2, Table 2 and Table 3). The nonpolar solvents (C₆H₁₄, C₆H₁₂, C₆H₆, and C₆H₅CH₃), which are distinguished by their values of AN, η, and 1/(ε/ε₀), as shown in Figure 11, are highlighted in yellow.

The findings in Figure 12a–d and continued in Figure S7e–l are summarized as follows:

Total N_T,M vs. N_T,Mmax (Heating) (a). These values are greatly influenced by the solvent environment. The total N_T,M intensity increases linearly with N_T,Mmax during heating, suggesting that the overall PE yield can be estimated from the peak intensity. The decrease in the C 1s/Cu 2p ratios after TPPE across environments (4)–(9) shown in Figure 10 follows the same order as the sum of N_T,M during heating in Figure 12a.

Total N_T,M vs. 1/(ε/ε₀) (Heating) (b). The total N_T,M generally increases as 1/(ε/ε₀) decreases during heating. This indicates that higher solvent polarity (lower 1/(ε/ε₀)) weakens the Coulomb field at the surface overlayer, thereby facilitating electron emission.

ΔE_N,T,M (First step, Heating) vs. 1/(ε/ε₀) (c). Similar to the total intensity, the activation energy for the first step increases as 1/(ε/ε₀) decreases. This inverse correlation shows that both the electron emission yield and the energy barrier tend to rise as the solvent’s dielectric constant increases, consistent with our previous findings for Sample A [2].

Cooling Cycle Characteristics (d), (e). The total N_T,M increases as the negative activation energy becomes more negative during cooling. Furthermore, the absolute value of this negative activation energy decreases as 1/(ε/ε₀) drops, exhibiting a similar trend to that described for the heating cycle in Figure 12c.

Activation Energy Interdependence (f), (g). The negative ΔE_N,T,M of the cooling cycle becomes smaller as ΔE_N,T,M of the first-step heating cycle increases. Conversely, it becomes less negative as the second-step heating barrier decreases. However, the reason for this opposing dependence remains unclear.

Pre-exponential Factors (h), (i). During heating (first step), A_0N,T,M increases gradually with increasing activation energy. We hypothesize that this is related to the adsorption strength of organic molecules on the Cu surface: more tightly bound molecules (higher ΔE_N,T,M and A_0N,T,M) are attached directly to the overlayer, whereas less strongly bound molecules occupy the outer layers. During cooling, A_0N,T,M progressively increases as the negative ΔE_N,T,M becomes less negative. Nonpolar hydrocarbon molecules appear to be less tightly bound, characterized by highly negative ΔE_N,T,M and low A_0N,T,M values.

Surfaces Species and N_T,Mmax (j), (k). During heating, N_T,Mmax increases as the peak heights of both Cu(OH)₂ and Cu₂O (measured after TPPE) decrease. Notably, the gradient for the Cu₂O peak height is approximately twice that of the Cu(OH)₂ peak after TPPE, suggesting that a reduction in Cu₂O has a stronger effect on enhancing N_T,M intensity during heating.

Cooling Cycle vs. Cu₂O (l). Similarly, the total N_T,M during cooling increases as the Cu₂O peak height decreases after TPPE (excluding outliers: Ar plasma and benzene), confirming that Cu₂O depletion significantly facilitates PE in both scan directions.

Figure S8a–i presents the comparison and correlation of the TPPE characteristics across environments (1)–(9).

The TPPE characteristics are summarized as follows:

N_T,M Intensity and Peak Maxima. Across all environments, the N_T,M intensity during heating is considerably higher than that during cooling. For (1) emery paper only, both the heating and cooling N_T,M intensities are the lowest, but for (2) acetone cleaning, where rolled samples were used; both values are markedly higher. This finding can be attributed to carbon contamination, including acetone, on the surface. For (3) Ar plasma treatment, the heating-cycle N_T,M intensity is much higher than that for (1) emery paper and comparable to those for the solvent-treated samples, despite the activated adsorption of oxygen species (H₂O and O₂) from the air by plasma exposure. However, the cooling-cycle intensity for (3) Ar plasma is significantly lower, which may be related to the desorption of these oxygen species. N_T,M gradually decreases as the solvents change from (4) to (7), then increases for (8) and (9).

ΔE_N,T,M (Heating, First step). The energy barrier for (3) Ar plasma is slightly negative, whereas the other environments have positive values. Moreover, a paradoxical result is observed for (4) aniline, which exhibits both the highest energy barrier and the highest N_T,Mmax. This finding suggests that the PE involves electrons escaping from energetically demanding yet highly emissive surface states, which may be rationalized using a model involving surface–dipole interactions that modify the work function [66]. For the nonpolar solvents (C₆H₁₂, C₆H₆, C₆H₅CH₃, and C₆H₁₄), the energy barrier becomes highest for C₆H₅CH₃ while N_T,Mmax is minimized, demonstrating a consistent inverse relationship between the barrier height and the PE intensity.

ΔE_N,T,M (Heating, Second step). No second step is observed for environments (2) and (9). All other environments exhibit negative energy barrier values, and those for (3), (5), (6), (7), and (8) are much more negative than those for (1) and (4). Among the nonpolar solvents, as the magnitude of the first-step energy barrier increases, that of the second-step energy barrier tends to become less negative, reaching a minimum for (7) toluene.

ΔE_N,T,M (Cooling). All environments exhibit large negative activation energies during cooling. This trend is similar to that of the second step during heating (except for environments (2) and (9)), although the magnitudes are smaller. ΔE_N,T,M Down becomes extremely small for (4) aniline, despite having the highest first-step heating barrier. For the nonpolar solvents, (7) toluene shows the lowest negative barrier and the lowest N_T,M during cooling, emphasizing the role of the energy barrier in controlling emission.

Next, the solvents are divided into two groups based on the reciprocal dielectric constant (Figure 11). The first group ((4) aniline, (2) acetone, and (9) ethanol) has a lower reciprocal dielectric constant (higher polarity), and the second group ((8) hexane, (5) cyclohexane, (6) benzene, and (7) toluene) has a higher reciprocal dielectric constant (lower polarity). In the first group (Figure S8f,g), 1/(e/e₀) decreases as N_T,Mmax decreases during heating. During cooling, the magnitude of the negative barrier decreases in the order of (2) acetone > (4) aniline > (9) ethanol, and the total N_T,M decreases accordingly. Acetone appears to form the most surface electronic traps. In the second group (Figure S8h,i), 1/(e/e₀) decreases in the order of (8) hexane > (5) cyclohexane > (6) benzene > (7) toluene, while the first-step heating barrier increases. Meanwhile, the negative magnitudes of the second-step heating and cooling barriers decrease in the same order. For (5) cyclohexane, (6) benzene, and (7) toluene, N_T,Mmax and the total N_T,M correspond well with 1/(e/e₀) and ΔE_N,T,M during heating. Hexane is an outlier because the intensity remains lower than expected. These findings collectively suggest that electron transfer to the surface can be controlled by the energy barriers imposed by specific adsorbate–surface interactions.

3.7. PE Stimulation Spectra and TPPE Characteristics

The relationships between the PE stimulation spectra and the TPPE characteristics for the different environments are displayed in Figures S9a(a−h) and S9b.

Intensity vs. β (Heating) (a), (b). The total N_T,M and I_Wmax all show a positive correlation with β. Because β is proportional to the PE probability (αA), these results confirm that β is a primary driver of PE intensity. Specifically, values for (4) aniline and (5) cyclohexane are significantly higher than those for (9) ethanol and (7) toluene.

Intensity vs. Photon Energy and Wavelength (Heating) (c), (d). A paradoxical phenomenon is observed: I_Wmax decreases as the photon energy (ϕ_Wmax) increases, inverse to typical photoemission behavior. This decline in I_Wmax may be related to an increase in electron density on surface oxygen species (Cu(OH)₂ and Cu₂O) at higher photon energies, which alters the interaction modes of adsorbed organic molecules [67]. Notably, (7) toluene and (9) ethanol exhibit the lowest intensities at the highest photon energies (shortest wavelengths).

Cooling-Cycle Intensity vs. β (Cooling) (e), (f). Similar to the heating cycle, the cooling-cycle total N_T,M and I_Wmax progressively increase with β. Hexane exhibits the highest intensity, whereas toluene and benzene exhibit the lowest.

Influence of Solvent Polarity (1/(ε/ε₀)) (g). For the nonpolar solvents having higher 1/(ε/ε₀), I_Wmax increases with 1/(ε/ε₀) during cooling. This correlates with the increasingly negative ΔE_N,T,M values (Figure S8e). That is, lowering the energy barrier facilitates the transfer of photoelectrons through these solvents.

β Dependence and IE (h), (i). While β tends to decrease from environment (3) to (9) (although those for environments (5) and (8) are slightly higher) during heating, the dependence of β during cooling is closely related to the chemical properties of the solvents, namely the molecular IE. The solvents fall into two groups: a low IE group (aniline, toluene, and benzene) with lower β (cooling) values, and a high IE group (hexane, cyclohexane, and ethanol) with higher β (cooling) values.

The TPPE mechanism is linked to the electronic properties of Cu₂O. As a p-type semiconductor, the band structure comprises holes in the valence band (VB), acceptor levels in the bandgap, and excited electrons in the conduction band (CB). Holes in the VB promote oxidation reactions by accepting electrons from adsorbed molecules. This helps cleave C–H and C–C bonds and is central to catalytic transformations like alcohol dehydrogenation (C₂H₅OH → CH₃CHO + 2H⁺ +2e⁻) and oxygen formation from water (2H₂O → O₂ + 4H⁺ + 4e⁻). Electrons in the CB can participate in the reduction in water (2H₂O + 2e⁻ → H₂ + 2OH⁻). In these processes, adsorbed molecules donate electrons to the holes, promoting further electron excitation in the CB. Thus, the more holes that are produced in the VB, the more electrons that can be excited. We hypothesize that this PE is accelerated during cooling. For the high IE group (hexane, cyclohexane, and ethanol), hole donation is slower than for the low IE group (aniline, toluene, and benzene).

Crucially, the trap states and charge-transfer pathways within this Cu₂O matrix are strongly governed by the specific charge states of its oxygen vacancies. In binary oxides, oxygen vacancies do not exist merely as static defects. However, they can transition between neutral, singly ionized, and doubly ionized electronic configurations, operating similarly to F-type color centers [68]. Under the simultaneous influence of low-energy UV irradiation and thermal scanning used in our TPPE experiments, these vacancy states undergo dynamic shifts in their electronic structures. A neutral vacancy, or standard F-center, traps two electrons and introduces localized states deep within the bandgap. As the heating cycle (up scan) progresses, thermal activation and optical excitation drive the ionization of these neutral centers into singly ionized, paramagnetic states (F⁺-centers) or fully ionized states, releasing delocalized electrons into the conduction band matrix.

This transient thermal ionization model provides an elegant electronic explanation for the observed variations in the photoemission probability factor (β). The generation of these intermediate, highly polarizable ionized vacancy centers effectively introduces a high density of shallow donor traps just below the conduction band edge, facilitating electron transport and temporarily lowering the kinetic barrier for electron extraction toward surface hydroxyl radicals. Conversely, during high-temperature oxidation steps or subsequent cooling cycles, the redistribution of these localized defect charges directly modifies the availability of reactive surface traps, influencing both the local work function and the baseline photoemission yields across different chemical environments.

3.8. Proposed Mechanism and Its Relation to Activation Energy

The mechanistic roles of Cu(OH)₂ and Cu₂O in TPPE are characterized in several key stages, including the formation of surface hydroxyl groups from adsorbed water films; the production of different –OH trap levels with high electron affinity depending on the environment; the attraction of electrons from the Cu base metal to these hydroxyl groups, leading to Auger emission; and the formation of surface dipoles via acid–base interactions, where the reciprocal dielectric constant significantly modulates the emission barrier. Furthermore, the process involves the growth of Cu₂O during thermal cycling, as well as surface redox reactions facilitated by photogenerated holes in the VB and excited electrons in the CB of Cu₂O. Finally, oxygen vacancies formed between the VB and CB appear to serve as essential mediators for electron transport.

To maintain scientific rigor, these interpretive pathways must be evaluated alongside alternative surface physical phenomena. Because TPPE measures a comprehensive, macroscopic electronic output from a complex composite film, the experimental signal represents a convolution of multiple competing processes. In addition to our proposed radical-tunneling model, local variations in the work function significantly influence the effective emission threshold. The adsorption of organic molecules establishes an electrostatic dipole layer that either opposes or assists the surface barrier, depending on molecular orientation. Concurrently, dynamic oxide growth at elevated temperatures changes the physical thickness of the tunneling barrier, while localized defect-assisted transport through the expanding Cu₂O semiconductor matrix creates localized, high-conductivity pathways. Finally, low-energy photoelectron collection can be influenced by transient surface charging potentials across the less-conductive regions of the organic overlayer. While isolating these individual sub-processes experimentally within a single macroscopic measurement is challenging, our proposed framework aims to integrate these factors by analyzing their combined effects on the apparent activation energy and photoemission probability.

This multifactorial interpretation strongly complements recent advances in operando surface and interface characterization. Specifically, Tian et al. [69] highlighted that in crystalline, thin-film architectures, interfacial charge transport is profoundly sensitive to defect-mediated pathways and specific structural surface reconstructions. Our observation of multi-stage activation energies provides a direct kinetic manifestation of these interface phenomena, demonstrating how the transition from surface hydroxyl states to highly polarizable p-type semiconductor domains creates transient donor traps that accelerate electron transport. Furthermore, this dynamic tuning of surface electronic states explains broader macroscopic applications, such as path-dependent virus inactivation. Glover et al. [70] demonstrated that the rate of pathogen inactivation on high-touch copper surfaces is highly dependent on the surrounding chemical assay media, which alters oxidative surface release. Our TPPE results complement this materials science approach by uncovering the underlying electronic mechanism: the surrounding environment establishes distinct chemical dipoles and changes the surface potential array, directly modulating the availability of localized reactive electrons and surface species necessary for driving contact-mediated redox reactions.

Figure 13 outlines the temperature dependence of N_T,M (TPPE plot) for all environments. The adsorption states of water, oxygen species, and organic molecules change significantly with temperature and light irradiation. These changes strongly influence the observed PE and the associated activation energies. The TPPE plot can generally be divided into three parts. However, no second stage is observed for environments of (2) and (9): the first stage refers to the first step of heating (N_T,M increases to a maximum as temperature rises; positive activation energy), the second stage refers to the second step of heating (N_T,M declines after the peak; negative activation energy), and the third stage refers to cooling (N_T,M increases as temperature decreases; negative activation energy).

The following summary integrates the proposed mechanistic steps and experimental observations into a coherent description of how the copper surface overlayer transforms and facilitates Auger-type emission and redox reactions.

1. Initial State: Thick Adsorption Overlayer Before TPPE

Before TPPE measurements begin, the copper surface is covered by a relatively thick, heterogeneous adsorption layer comprising water molecules, surface hydroxyl groups (OH), oxides (Cu₂O/CuO), and various organic molecules. These species accumulate naturally under ambient exposure and form a complex overlayer that significantly modifies the electronic properties of the Cu surface and creates a photothreshold distinct from the work function of clean metallic Cu.

2. Early Heating Stage: Desorption of Weakly Bound Species

Initial heating and irradiation trigger the desorption of weakly bound molecules, gradually thinning the overlayer and leaving a more stable, tightly bound film composed primarily of Cu(OH)₂, Cu₂O, and strongly adsorbed water or organic molecules. This thinning process alters the tunneling barrier for electrons, thereby modifying the PE intensity.

3. PE Through the Adsorption Layer

Even with the adsorption layer, PE from the underlying copper metal still occurs. This emission can be interpreted via two primary pathways: direct photoemission traversing the film and an Auger-type mechanism facilitated by surface hydroxyl radicals. These electrons must traverse the adsorption layer before escaping, meaning that the effective photothreshold is governed by the electronic properties and dipole characteristics of the adsorbed film.

4. Photogeneration of Charge Carriers in Cu₂O

Light irradiation also excites electrons and holes within the Cu₂O surface layer:

Cu₂O → h⁺ + e⁻.

The process creates mobile charge carriers that participate in subsequent surface reactions. Photogenerated holes (h⁺) are critical for oxidizing adsorbed water molecules, while electrons (e⁻) can contribute to PE or reduction reactions.

5. Role of Hydroxyl Radicals in PE

Photogenerated holes oxidize adsorbed water [71]:

H₂O + h⁺ → ∙OH + H⁺

The resulting hydroxyl radical (∙OH) has a high electron affinity (1.83 eV). In the early stage of TPPE, electrons near the Fermi level of Cu can tunnel to these ∙OH species, producing PE via an Auger-type mechanism. This tunneling-assisted emission represents one of the primary pathways for PE at lower temperatures.

6. Surface Dipoles and Thermal Activation

As temperature increases, water and organic molecules adsorb onto the surface hydroxyl groups of Cu(OH)₂, forming electric dipoles with their positive ends oriented outward. These dipoles modify the surface potential and influence electron tunneling. Meanwhile, oxygen vacancies in Cu₂O become thermally activated, further affecting charge transport.

The PE intensity (N_T,M signal) increases with temperature up to a maximum. Two mechanisms contribute:

• Auger-type emission from Cu VB electrons tunneling to ∙OH.
• Bandgap-mediated emission involving photogenerated electrons in Cu₂O that traverse the thin film.

Additionally, water adsorbed directly on Cu₂O undergoes reduction by photoexcited electrons [72]:

H₂O + e⁻ → ½H₂ + OH⁻

and the resulting OH⁻ becomes a new adsorption site for organic molecules.

During the first heating step, electrons must overcome the energy barrier of the adsorption film, thereby making the activation energy positive.

It should be noted that oxygen vacancies are thermally activated in Cu₂O (for a review of oxygen vacancies, see ref. [23]). Oxygen vacancies exhibit localized states just below the conduction band. These defect states possess high polarizability and strong affinity for polar molecules, making them preferential adsorption sites for H₂O, resulting in the formation of protonated surfaces. Crucially, this structural pathway contributes to the increase in N_T,M intensity during the first heating step.

7. High-Temperature Stage: Desorption and Oxidation Processes

The N_T,M intensity reaches a maximum at T_max,M. At higher temperatures (in the second heating step), water produced from Cu(OH)₂ decomposition, as well as adsorbed organic molecules, gradually desorb, resulting in continuous depletion of the protonated surfaces generated via oxygen vacancies–H₂O interactions, leading to a steady decline in PE intensity. Three main factors contribute to this behavior:

• The loss of surface dipoles as organic molecules desorb.
• The oxidation of water by photogenerated holes, which produces O₂ and promotes CuO formation.
• The thermal degradation of the active protonated surface layer that was generated from the interactions between oxygen vacancies and adsorbed water molecules.

This oxidation reaction reduces the Cu₂O content, altering the electronic pathways for PE and resulting in negative activation energies for many environments (emery paper, Ar plasma, aniline, cyclohexane, benzene, toluene, and hexane). Acetone and ethanol, however, do not support negative activation energies, suggesting differences in adsorption strength or reaction pathways.

Because the formation of oxygen vacancies in Cu₂O is thermally activated, vacancy generation alone cannot account for a decline in intensity at elevated temperatures. Instead, this high-temperature stage is governed by a kinetic competition where structural oxidation and dipole quenching dominate over vacancy activation, leading to the apparent negative activation energies observed during heating.

8. Post-Heating State at 350 °C

After heating to ~350 °C, organic molecules are essentially fully desorbed, but adsorbed water molecules remain. As the temperature decreases during cooling, the N_T,M intensity gradually increases, yielding negative activation energies across all evaluated environments. This indicates that cooling enhances PE through mechanisms involving the reconstruction of protonated surfaces, as residual moisture interacts with oxygen vacancies and forms stable interfacial dipole arrays at lower temperatures.

9. Cooling: PE Enhancement at Lower Temperatures

During cooling, several processes enhance PE:

• Stable water adsorption leads to the formation of hydroxyl groups that create local dipoles, facilitating electron emission, and protonation of the surface with oxygen vacancies promotes PE;
• The oxidation of water by photogenerated holes proceeds via [72]
  2H₂O → O₂ + 4H⁺ + 4e⁻
  donating electrons to Cu₂O;
• These electrons migrate through the thin film and contribute to PE at the surface. As a result, N_T,M intensity increases during cooling, and the activation energy becomes negative.

Notably, the negative activation energy arises from competing parallel phenomena: (1) the exothermic re-adsorption of residual moisture forming structured, dipole-enhancing OH networks, and (2) temperature-dependent changes in the carrier density of the p-type Cu₂O film, where lowered temperatures alter the mobility and lifetime of photoexcited holes. This multi-variable kinetic view deviates from the idea of a single photoredox step.

4. Conclusions

This study leveraged TPPE to characterize the electronic properties of pretreated Cu₂O/Cu surfaces in various organic environments. TPPE data revealed distinct electronic behaviors for copper surfaces subjected to different pretreatments during the heating cycle. Maximum photoelectron counts were correlated with activation energies, demonstrating the influence of the chemical environment on PE energy barriers. High-polarity solvents (e.g., ethanol, aniline) enhance emission by forming surface dipoles that lower the tunneling barrier. Furthermore, the observation of both positive and negative activation energies during heating and cooling cycles reveals a complex interplay between thermal activation, surface dipole formation, and the transition from hydroxyl groups to lattice oxides. XPS confirmed that the heating cycle induces a significant transformation of Cu(OH)₂ into Cu₂O. The decline in PE intensity at high temperatures correlates with the depletion of surface hydroxyls and the growth of the oxide layer.

Moreover, the TPPE process is governed by the p-type semiconductor properties of Cu₂O. Photogenerated holes and electrons drive surface redox reactions, where oxygen vacancies act as critical intermediates for electron transfer. A mechanism was proposed wherein electrons tunnel from the base metal to surface hydroxyl groups, followed by Auger-type emission. XPS analysis emphasized the involvement of surface ·OH species in the TPPE process, as evidenced by the O 1s spectra. The reactive electrons and surface species characterized by TPPE were hypothesized to play a critical role in the inactivation of viruses on copper surfaces. It must be emphasized that this specific biological correlation remains a mechanistic hypothesis, as direct virological assays were outside the scope of the present study. Finally, referring to the findings obtained during the first and second heating and cooling cycles for propanol (C₃H₇OH), as shown in Figure 4 of our previous report [2], the examination of subsequent cooling and heating cycles provided evidence that excess oxygen in the Cu₂O overlayer governs the low activation energies observed after thermal treatment. Ultimately, the TPPE method serves as a powerful operando tool for evaluating the surface chemical technology of copper, particularly in optimizing photocatalysts and antimicrobial surfaces, where electron transfer is the primary mechanism of action.