Hydrogen, H₂(g), by near-ambient-pressure soft x-ray synchrotron-radiation photoelectron spectroscopy

The continuously increasing popularity of near-ambient-pressure x-ray photoelectron spectroscopy (NAP-XPS) at synchrotron facilities (Refs. 1–11) has renewed interest in gas-phase XPS. The introduction of commercial NAP-XPS systems based on traditional, lab-based x-ray sources has further increased the motivation to publish reference spectra for gas molecules (Refs. 12–18). To carefully analyze spectra and to obtain detailed insight into the surface in the presence of a gas phase, it is paramount to (1) correctly identify peaks directly stemming from the gas phase and to (2) understand how the interaction with the gas phase influences photoelectrons originating from the surface.

This contribution to the Focused Topic Collection: Near-ambient Pressure X-ray Photoelectron Spectroscopy presents the photoelectron spectra of H₂(g). Atomic hydrogen has only one (partially) filled orbital, which in H₂ forms a (filled) σ_1s molecular orbital. The H 1s photoionization probability (or cross section) rapidly decreases with increasing photon energy, e.g., at the photon energy of Mg K_α (1253.6 eV) and Al K_α (1486.7 eV), the cross section to excite the H 1s orbital is ∼4000 and 5000 times lower, respectively, than to excite the C 1s orbital (Ref. 19). Therefore, using lab-sources, it is impracticable to measure the H 1s orbital, which has, incorrectly, lead to the conclusion that it is impossible to measure H 1s photoelectrons in XPS (Refs. 20–22).

Soft x-ray beamlines using undulators provide a tunable, intense x-ray source that allows measuring the H 1s orbital (Refs. 23 and 24). The lower photon energies achievable at these beamlines help the measurement of the H 1s orbital. For example, at the photon energy used in this study, 400 eV, the H 1s orbital is “only” a factor 2000 weaker than the C 1s. The use of low photon energies with an intense x-ray beam makes measuring the H 1s orbital feasible.

The presence of H₂(g) in NAP-XPS can affect the photoelectron signal from the surface of the material; small features are discernable at the higher-binding-energy side of all peaks. These features are electron-energy loss features and are ascribed to electronic excitation of H₂(g) by photoelectrons originating from the surface. The excitation corresponds to a loss of 12.6 ± 0.1 eV in agreement with values in the literature of 12.6 (Ref. 25) and 12.7 eV (Ref. 26). Their relative intensity to the main, zero-loss peak is highly pressure dependent (cf. Fig. N01690-03 at 67 Pa, Ref. 25 at 300 Pa, and Ref. 26 at 400 Pa). After evacuating gas-phase H₂, the H₂(g) loss feature disappears immediately (Fig. N01690-07).

The given analyzer resolution is based on the Gaussian full-width at half maximum (FWHM) of the Ag 3d_5/2, fitted with a Voigt function, and collected with the same beamline and analyzer settings. As such, it reflects the total experimental resolution (beamline and analyzer).

Host Material: Hydrogen, H₂(g)

CAS Registry #: 1333-74-0

Host Material Characteristics: Homogeneous; gas; amorphous; dielectric; inorganic compound; other

Chemical Name: Hydrogen, H₂(g)

Source: Matheson research purity (99.999%)

Host Composition: Hydrogen, H₂(g)

Form: Gas

Structure: HH, H—H

History & Significance: H₂ is one of the most common reactants in heterogeneous catalysis and important as renewable fuel.

As Received Condition: Compressed gas cylinder

Analyzed Region: Hydrogen gas molecules encountered by the x-ray beam

Ex Situ Preparation/Mounting: N/A

In Situ Preparation: For the PdAg(111) sample, approximately 1 monolayer of Pd was evaporated on an Ag(111) single crystal surface, which was close to room temperature. To obtain the Ag 3d spectra in 01690-03, the surface was annealed to 475 K.

Charge Control: N/A

Temp. During Analysis: Room temperature

Pressure During Analysis: 146 Pa (survey/H 1s) and 67 Pa [H₂(g) loss and Ag 3d]

Pre-analysis Beam Exposure: 30 s

Manufacturer and Model: SPECS Phoibos 150R6 NAP

Analyzer Type: Spherical sector

Detector: Surface Concept 1D-DLD detector (model number: 1D-DLD64_2-150HV)

Number of Detector Elements: 100

Analyzer Mode: Constant pass energy

Throughput (T = E^N): N = −1

Excitation Source Window: 100-nm-thick SiN_x coated with 5 nm Cr and 10 nm Au (SPI supplies)

Excitation Source: Elliptically polarizing undulator at beamline 23-ID-2 of the National Synchrotron Light Source II (Ref. 27)

Source Energy: 400.02 eV (survey/H 1s) and 451.3 eV [H₂(g) loss and Ag 3d]

Source Strength: 10¹²–10¹³photons/s

Source Beam Size: 16 × 70 μm²

Signal Mode: Multichannel direct

Incident Angle: N/A

Source-to-Analyzer Angle: 70°

Emission Angle: N/A

Specimen Azimuthal Angle: N/A

Acceptance Angle from Analyzer Axis: 22°

Analyzer Angular Acceptance Width: 44°

Energy Scale Correction: The binding-energy scale in spectra 0169-01 and 0169-02 was corrected for the work function of the detector and photon energy. The work function of the detector was determined by setting the kinetic energy of the Ag M₄NN transition of a clean reference Ag(111) sample to 353.38 eV (Ref. 28). The photon energy was determined by setting the Ag 3d_5/2 binding energy of a pure reference Ag(111) sample to 368.20 eV. This correction procedure yields the Fermi level of a reference Ag(111) crystal to be 0.00 eV.

The binding-energy scale in 01690-03 [Ag 3d and H₂(g) loss] is relative to the Fermi level of an Ag(111) sample alloyed with a small amount of Pd. The Ag 3d_5/2 peak position was 367.97 eV.

The kinetic-energy scale of 01690-07 (H loss feature) is relative to Ag 3d_5/2 energy.

Fitting: The H₂(g) loss feature was fitted with a Gaussian profile (FWHM = 1.6–1.7 eV).

Recommended Energy Scale Shift: 0 eV.

Peak Shape and Background Method: Gaussian [H₂(g) loss] and Voigt peak shapes (H 1s in survey and Ag 3d). Linear (survey and H 1s) and Shirley [H₂(g) loss and Ag 3d] background subtraction.

Quantitation Method: N/A

The staff of the IOS beamline, especially Iradwikanari Waluyo and Adrian Hunt, are gratefully acknowledged for their support. This work was supported as part of the Integrated Mesoscale Architectures for Sustainable Catalysis, an Energy Frontier Research Center funded by the U.S. Department of Energy, Office of Science, Basic Energy Sciences, under Award No. DE-SC0012573. This research used resources of the IOS (23-ID-2) beamline of the National Synchrotron Light Source II and the Center for Functional Nanomaterials, which are U.S. Department of Energy (DOE) User Facilities at Brookhaven National Laboratory operated under Contract No. DE-SC0012704.