Laboratory-based high pressure X-ray photoelectron spectroscopy: A novel and flexible reaction cell approach.

The last 10-15 years have witnessed a resurgence in the application of high pressure X-ray photoelectron spectroscopy, mainly through the development of new electron energy analyser designs and the utilization of high-brilliance synchrotron radiation sources. To continue this expansion of the technique, it is crucial that instruments are developed for the home-laboratory, considering that this is where the vast majority of traditional ultra-high vacuum (UHV) X-ray photoelectron spectroscopy is performed. The research presented here introduces a new addition to the field, an instrument capable of performing spectroscopy measurements from UHV to high pressure (25 mbar), achieved using a retractable and modular reaction cell design. The ease of use, stability (of analyser, X-ray source, and gas delivery, etc.), and overall capability of the instrument will be demonstrated.


I. INTRODUCTION
Chemical processes that occur at solid-gas, solid-liquid, and liquid-gas interfaces are of enormous practical significance, and a detailed mechanistic understanding of such processes is essential for progress in fields such as catalysis, 1 electrochemistry, 2 environmental science, 3 the development of bio-compatible materials, and many others. Photoelectron spectroscopy (PES) has long been established as a powerful technique for probing the electronic structure and chemical composition of surfaces; 4 however, performing PES experiments under conditions of practical relevance can often be very challenging. In particular, the use of pressures in excess of ∼10 5 mbar is not possible in a conventional spectrometer. There are two major difficulties that arise when trying to measure a photoelectron spectrum at higher pressures: first, the mean free path of the emitted photoelectrons becomes so short that very few photoelectrons will reach the detector without undergoing inelastic collisions; and second, electrical arcing will occur in the electron analyser due to the high voltages that are applied to achieve energy-filtering of the photoelectrons. A solution to these problems was achieved by introducing several differential-pumping stages between the sample and the entrance to the hemisphere of the analyser, in conjunction with electrostatic lenses to refocus the electrons to apertures at the exit of each differential pumping stage and increase transmission of photoelectrons from the point of emission from the sample surface. a) Author to whom correspondence should be addressed. Electronic mail: d.payne@imperial.ac.uk Historically, "unconventional" (i.e., non-UHV) X-ray photoelectron spectroscopy (XPS) measurements were first performed in the group of Kai Siegbahn 5 for the development of an instrument for PES measurements of liquids 6 and later by Wyn Roberts 7 who developed the first spectrometer for studies of solid surfaces under gas pressures of up to 1 mbar, although these electron analysers lacked the electrostatic lenses that characterise modern-day instruments. These differentially pumped electrostatic lenses were first developed by the group in the Advanced Light Source (ALS, Berkeley) 8 and then jointly between the ALS and the Fritz Haber Institute (Berlin). 9 Whilst the original instruments of Siegbahn and Roberts used an Al or Mg X-ray anode as the excitation source, most of the modern instruments are located at synchrotrons, due to the obvious advantages of higher photon flux and tuneable excitation energy. Nonetheless, there has recently been resurgence in the development of laboratorybased high-pressure photoelectron spectrometers. [10][11][12][13][14][15][16] Current state-of-the-art laboratory-based instruments have maximum operating pressures that are comparable to the saturated vapour pressure of water at room temperature (∼32 mbar) and are therefore suitable for the study of a wide range of surfaces of practical relevance. A major advancement in the controllability of the reaction environment within the spectrometer was the introduction of in situ reaction cells, where instead of filling the entire analysis chamber with the reactive gases only a small volume is exposed, whilst the surrounding chamber remains in vacuum. 17 A design of this type of in situ measurement cell was presented by Schnadt et al. in 2012. 18 The design, and also the layout of subsequent retractable high-pressure cells, is based on cells with a comparatively small volume of between 20 and 500 ml, where samples can be transferred into the cell through a door mechanism which is manually manipulated, e.g., by using a wobble stick. [18][19][20][21] In this article, we describe the functionality and operation of a new laboratory based high pressure X-ray photoelectron spectroscopy (HPXPS) system that includes a unique small volume, high-pressure reaction cell which is compression sealed against the high-pressure electron energy analyser and utilise a highly precise gas delivery system. The system has been specifically designed for maximum flexibility in the way the sample is introduced and handled, with the capability of switching between ultra-high vacuum and gas pressures (up to 25 mbar) using a retractable high pressure cell concept. The in situ cell is the first reported design without a door mechanism necessitating mechanical manipulation and has a highly improved internal design leading to a significant reduction of the gas-exposed surface area on the inside of the cell itself. This design and other aspects of the instrument result in smooth operation with sample transfer times from preparation chamber to measurement position in the order of ∼5 min, unparalleled stability of performance and measurement, from the X-ray source, electron energy analyser to gas delivery, evidence of which will be presented in this work. Figure 1 shows the outline of the current HPXPS system. It consists of three stainless steel (S304) chambers: the fast entry load lock (1), the preparation chamber (2), and the reaction cell entry lock and a µ-metal analysis chamber (3). The analyser is composed of two sections, the pre-lens (4) and the hemispherical analyser itself (5). The instrument has two monochromated radiation sources: an X-ray source (6) and a vacuum ultra-violet (VUV) source (7). Finally, the instrument has a mass flow controlled gas delivery system (8). These components will be individually described in detail in Secs. II A-II D and Sec. III C.

A. Fast entry load lock and preparation chamber
The system is equipped with a fast entry load lock, where samples mounted on Omicron-style sample plates are FIG. 1. 3D schematic overview of the HPXPS setup. The system comprises a load lock chamber, a preparation chamber, an analysis chamber, a reaction cell load lock chamber located below the analysis chamber, and a gas delivery chamber.
introduced. The load lock is fitted with a linear drive for transferring samples between the fast entry load lock and the preparation chamber and is pumped by a turbo pump (Oerlikon Leybold SL80) which itself is backed by a 10 m 3 /h scroll pump (Edwards Vacuum XDS10). The preparation chamber is equipped with a 4-axis manipulator (Vacgen Miniax) for positioning and heating samples to a maximum of 1200 K, an argon ion sputter gun (Omicron ISE5), a low energy electron diffraction (LEED) (Scienta Omicron SpectaLEED), and a rotatable linear drive for transferring samples between the preparation chamber and the analysis chamber. The preparation chamber is pumped by 300 l/s turbo pump (Oerlikon Leybold MAG300) and backed by a 10 m 3 /h scroll pump (Edwards Vacuum XDS10).

B. Analysis chamber
The analysis chamber houses the two excitation sources: an Al Kα monochromated X-ray source and a retractable monochromated VUV source (both are described in more detail in Section II C). A hemispherical analyser (Scienta Omicron HiPP2) with a differentially pumped pre-lens, which has been demonstrated to operate at pressures of up to 30 mbar in the main analysis chamber, 15 is mounted vertically on top of the analysis chamber. In addition, a flood gun (Scienta, FG300) for charge compensation during UHV measurements and a solar simulator (150 W Xe-arc lamp, LOT-Oriel LSH102) are built into the system. The analysis chamber houses three manipulators: a standard 4-axes manipulator for UHV measurements, a specially designed manipulator for the reaction cell (described in more detail in Section III B), and a manipulator to handle the top part of the reaction cell (described in Section III A).
To maintain excellent vacuum in the analysis chamber, there are a total of twelve turbo molecular pumps supported by seven scroll pumps (this number includes the pumps on the analyser). The first stage of the analyser pre-lens is pumped by two 300 l/s turbo pumps (Oerlikon Leybold MAG300) and separately backed by two 30 m 3 /h scroll pumps (Edwards Vacuum XDS35i). The second stage of the pre-lens is pumped by two 300 l/s turbo pumps (Oerlikon Leybold MAG300) and backed by a 15 m 3 /h scroll pump (Edwards Vacuum nXDS15). The third stage of the pre-lens and the hemispherical analyser are pumped each by one 300 l/s turbo pump (Oerlikon Leybold MAG300) and a 10 m 3 /h scroll pump (Edwards Vacuum XDS10). The analysis chamber is pumped by a 400 l/s turbo pump (Oerlikon Leybold MAG400) backed by a combination of a 80 l/s turbo molecular pump (Oerlikon Leybold SL80) and a 10 m 3 /h scroll pump (Edwards Vacuum XDS10). This combination improves the ultimate vacuum pressure in the analysis chamber considerably, where the base pressure (after baking) is 10 10 mbar. The X-ray monochromator chamber is pumped by an 80 l/s turbo molecular pump (Oerlikon Leybold SL80) and backed by using a backing pump of the analysis chamber. The VUV source has three differential stages; the first and second stages are pumped by two 80 l/s turbo molecular pumps (Oerlikon Leybold SL80) and backed by one 10 m 3 /h scroll pump (Edwards Vacuum XDS10). The monochromator of the UV source is pumped by a 300 l/s turbo pump (Oerlikon Leybold MAG300) and backed by a 10 m 3 /h scroll pump (Edwards Vacuum XDS10).

C. Radiation sources
The X-ray source (Scienta Omicron MX 650) is mounted at an angle of 62.5 • with respect to the symmetry of the analyser pre-lens and produces a 1 × 3 mm spot of monochromated Al Kα radiation (hν = 1486.6 eV). It consists of a water-cooled aluminium anode which is bombarded by a fine-focus electron beam from a tungsten cathode and a seven quartz crystal monochromator. Under standard operating conditions, a current of 16.7 mA in the filament and 12 V accelerating voltage (and a gun power of 200 W) are used. The seven toroidal crystals are mounted on a close packed array with one central crystal surrounded by six crystals in a Rowland circle configuration with a diameter of 650 mm. 22 The crystals are heated to 50 • C to ensure a good uniformity of the lattice spacing of the crystals. Each of the crystal positions can be adjusted manually using a set of three screws on the outside of the monochromator housing. This X-ray source has been described in previous articles when applied to HPXPS. For such applications, it is necessary to introduce a pressure separator, i.e., an X-ray transparent window, between the high vacuum of the monochromator and the elevated pressure around the sample in the analysis chamber. Eriksson et al. report the properties of the source using a reinforced aluminium-coated window 15,23 and Newberg et al. use a silicon nitride window. 13 In this paper, we present a third approach using a polyimide window built into the reaction cell itself. This greatly minimises the size of the window required, when compared to the separator window between the monochromator and analysis chamber, found in other systems.
The VUV source (Scienta Omicron VUV5000) is mounted at a 60 • angle to the analyser pre-lens and can produce He i (21 eV and 23 eV) and He ii (41 eV) radiations with exceptionally high intensity and narrow bandwidths. The helium plasma is generated using a microwave generator (200 W) coupled with a discharge cavity in a magnetic field tuned to the required microwave frequency exploiting the electron cyclotron resonance (ECR) technique. The system is equipped with a monochromator containing a toroidal focusing mirror separating He iα and He iβ, and the resulting photons have an energy resolution of 1 meV. The resulting light passes through a small capillary giving a spot size of 2 mm on the sample.

D. Analyzer
The high-pressure analyzer (Scienta Omicron HiPP-2) is a combination of a differentially pumped pre-lens attached before the receiving lens of a modified (additional pumping ports and sectioned pumping compartments) UHV hemispherical electron energy analyser (Scienta Omicron R4000). 24 The R4000 has a mean radius of 200 mm and is equipped with 9 interchangeable slits of variable dimensions (0.2 × 25 mm 2 to 4 × 30 mm 2 ) and shapes (straight and curved). A set of electrostatic lens elements enables operation at pass energies from 5 eV to 200 eV. A 2-dimensional detector is used, which consists of a FireWire CCD camera and a multi-channel plate (MCP). The energy range detected simultaneously is 8% of the pass energy. The pre-lens comprises a number of differentially pumped stages and apertures. The first aperture is located in the cone of the gas reaction cell in opposite to other instruments developed where the cone is a part of the analyser. 13,15 A second aperture is located at a short distance (10 mm) downstream to ensure a rapid decrease in pressure. A third aperture is located at the end of the pre-lens stage. The design and performance of the HiPP-2 analyser are described in more detail in Ref. 15

A. Design
The reaction cell has been designed to be completely nonmagnetic and able to withstand temperatures between 100 K and 1200 K on the sample even when the cell is pressurised to 25 mbar. The two main parts of the cell are machined from solid blocks of titanium, providing good tensile strength and high heat conductivity. Figure 2 shows a 3-dimensional view of the cell, including its two parts: the bottom cell (light grey in Figure 2) and the top cell (dark grey in Figure 2). The top cell can be removed using a linear drive to allow for samples to be transferred easily into the cell from the preparation chamber. This cell design eliminates the necessity to have a door mechanism for sample transfer as in previous cell designs, including the use of wobble sticks to mechanically manipulate the door under vacuum. [18][19][20][21] The major improvement of the overall in situ cell design compared to previous approaches is a dramatic reduction in the surface area that is exposed to the reactive gas. In combination with a very small overall inner cell volume of 14 ml, exchange times between UHV and high pressure conditions are greatly minimised.
The bottom cell is water cooled (Figure 2, item 9) to keep the O-ring gaskets (Figure 2, item 5) below 200 • C under all operating conditions. The sample plate is mounted on a molybdenum sample holder equipped with a pyrolytic boron nitride (PBN) heater ( Figure 2, item 8). The sample holder is fixed onto a copper tank (Figure 2, item 7) which is liquid nitrogen cooled during low temperature measurements. The sample holder and the copper tank are located on a second z-axis which enables the sample height to be precisely adjusted with respect to the cone aperture. The pressure in the cell is measured using a Baratron gauge located on the pumping port 500 mm away from the sample. Further down the line and protected by a capillary of 0.007 in. of inner diameter and 1 m length there is a residual gas analyzer (RGA) for mass spectrometry analysis (SRS RGA100 Residual Gas Analyser) of the gases in the main chamber. The excess of gas can be pumped away with an 80 l/s turbo molecular pump (Oerlikon Leybold SL80) backed by a 10 m 3 /h scroll pump (Edwards Vacuum XDS10) which are located on the same pumping pipework.
The top cell consists of a titanium housing, the cone ( Figure 2, item 1), an X-ray window (Figure 2, item 2), and three other windows (two for cameras and one for the secondary light source (e.g., solar simulator)). The top cell housing is indirectly water cooled by the cooling ring at the entrance of the analyser pre-lens and the bottom cell housing so that the X-ray window and the O-ring gaskets are kept to a temperature below 200 • C. A set of molybdenum shields (Figure 2, item 6) also helps to protect and reduce the temperature of the inner cell walls and windows. There are three different top cells available, adopted to different pressure ranges: a 0.3 mm cone aperture, a 0.5 mm cone aperture, and one with a 0.8 mm cone aperture. 15 Crucially, the top cell can be removed and remains in the analysis chamber, whilst the bottom half of the cell can be retracted behind a gate valve. The entire manipulator, including the bottom half of the cell, can then be removed from the analysis chamber.

B. Reaction cell load lock chamber
The reaction cell load lock chamber houses a purpose built wide bore z-axis translator (Vacgen Wide Bore Omniax) that carries the gas reaction cell and can be used to move it between the reaction cell load lock chamber and the analysis chamber. The manipulator is mounted on a moveable frame stage to enable the quick removal of the whole gas reaction cell from the system. The chamber is pumped by one 80 l/s turbo molecular pump (Oerlikon Leybold SL80) and backed by one 10 m 3 /h scroll pump (Edwards Vacuum XDS10). During the initial pump down, the turbo molecular pump of the preparation chamber is also used to speed up the pump down time. Before an experiment, samples can be transferred into the gas reaction cell via the fast entry lock chamber or via the reaction cell load lock chamber.

C. Gas delivery system
The gas delivery system is designed to control the pressure and mixtures of the gas in the reaction cell. Currently, there is a maximum of 5 different types of gases (including water vapour) that are used with the delivery system. The gases are dosed using mass flow controllers (MFCs) and water vapour is dosed using a bubbler which provides a pressure (in the gas reaction cell) of up to 30 mbar at room temperature. When using flammable gases, the scroll pumps of the analyser and the cell are purged with nitrogen gas to a concentration in the exhaust lower than the corresponding lowest explosive limit (LEL). The gas manifold is enclosed in a ventilated cabinet to ensure that any potential gas leak is removed safely from the laboratory.

D. Use of the gas cell and gas pod
Before a sample can be introduced into the reaction cell, the top cell has to be removed and is parked on a transfer fork. The sample plate is inserted into the sample holder and the top cell is put back in place. The reaction cell is compressed against the lens of the analyzer until the pressure rise detected in the analysis chamber is less than 3 × 10 10 mbar with a pressure of 40 mbar in the cell. During high-pressure experiments, the reaction cell is pumped through the cone aperture into the pumping system of the analyzer lens. The chosen gas pressure and ratio are achieved by using a proportionalintegral-derivative (PID) which controls the inlet flow of the gases. For water vapour measurements, the control of the pressure is achieved by varying the opening pulse of the bubbler valve and a PID controlling the opening of the throttling valve connected to the scroll pump.

E. Gas and water delivery and analysis pressure control
Constant gas pressure in the reaction cell is achieved using a PID to control the incoming gas flow. In the case of a gas mixture, the mass flow that is needed to produce the partial gas pressure of the gas in the reaction cell is calculated for each gas and then fixed in the mass flow controller (MFC). The pressure can be controlled over a range of 0.05-30 mbar with an accuracy of ±0.001 mbar. A plot of pressure of N 2 in the gas cell and MFC flow rate vs time can be seen in Figure  3. To deliver a pressure of 0.1 mbar in the reaction cell, it can be seen that the pressure overshoots to ∼0.19 mbar before stabilising back to 0.1 mbar after about 150 s. The pressure is then held constant and the variation in measurement arises from noise in the digital readout. The pressure during water vapour experiment can be controlled with an accuracy of ±0.1 mbar for pressures between 3 and 30 mbar, ±0.15 mbar for pressure between 1 and 3 mbar, and ±0.2 mbar for pressures in the 0.1-1 mbar range. The minimum pressure that can be provided during a water vapour experiment is 0.17 mbar (with

A. Energy resolution and intensity in ultra-high vacuum
The optimum resolution of the Scienta Omicron HiPP-2 analyzer when combined to the monochromated X-ray source Scienta MX650 was calculated to be 413 meV using the empirical expression given by Olivero and Longbothum 22 as the product of the convolution of the Gaussian instrumental contributions from the analyzer (75 meV for a 200 mean radius hemispherical analyzer at a pass energy of 100 eV and a slit width of 0.3 mm), the X-ray source (168 meV), the Gaussian temperature broadening (90 meV for measurements at room temperature), and the Lorentzian lifetime broadening of the photoinduced core hole (310 meV for the Ag 3d 5/2 transition). 15 To test the energy resolution and intensity of the instrument, Ag 3d core levels were measured at room temperature. A polycrystalline silver sample was sputtered until no O 1s signal was detected and the C 1s contribution was less than 5% of the Ag 3d signal. The full width at half maximum (FWHM) and count rates were obtained after subtracting a Shirley background. The Ag 3d core level measured for selected slits at pass energies 100 eV and 200 eV is shown in Figures 4(a) and 4(b), respectively. A resolution of 450 meV on the Ag 3d 5/2 is achieved when using the same measurement conditions as those used in the calculation. As seen in the figure, decreasing the slit size improves the energy resolution but also lowers the count rate. For example, when using a 1.5 mm straight slit at a pass energy of 100 eV, the FWHM of the Ag 3d 5/2 transition peak is 490 meV higher (40 meV higher) but the count rate is 4.7 times greater than the one obtained using the 0.3 mm curved slit. On the other hand, using a pass energy of 200 eV with the 1.5 mm straight slit, the FWHM is 610 meV and the count rate is 6.1 times higher than that for the 0.3 mm curved slit with a pass energy 100 eV. In terms of testing the performance of the analyser with the VUV source,   Figure 6 shows the Ag 3d 5/2 core level measured at varying nitrogen pressures for the 0.3 mm cone aperture using the standard lens mode. Similarly to the work of Eriksson et al., 15 as the pressure increases the signal intensity is increasingly affected by the electron losses in the gas phase and the count rate decreases rapidly; therefore, higher acquisition times are required to obtain spectra with equivalent signal to noise ratio. The measured signal attenuation at different pressures was fitted using the intensity attenuation equation of the low energy electron in a gas, assuming the attenuation path to be equal to the distance between the sample and the analyser cone and a scattering cross section σ e (KE) of 2.8 × 10 16 cm 2 . 8 Similarly to the work of Edwards et al., 26 the scattering cross section σ e (KE) is relatively higher than the values found in the literature, and the use of the gas cell does not affect the measurements. 27,28 This is most likely caused by the additional signal intensity loss that occurs in the first pumping stage of the FIG. 6. Ag 3d 5/2 core level intensity measured at varying N 2 pressures. The inner spectra represent the measurements of the Ag 3d 5/2 for a given pressure using the standard lens table, a 4 mm straight slit, a pass energy of 200 eV, a step size of 33 meV, a dwell time of 14 ms, and a cone of 0.3 mm. The straight line represents the calculated intensity attenuation equation of the low energy electron using a cross section of σ e (KE) of 2.8 × 10 16 cm 2 . 8 analyser. Despite the high acquisition time at higher pressure, the position of the Ag 3d 5/2 core level remains stable at 368.3 eV with an equivalent FWHM at 800 ± 10 meV as seen in the inset plots.

C. Performance of gas delivery system
Gas phase measurements are here used to demonstrate the capability of the gas delivery system to provide very small and precise gas mixtures which when combined to a stable monochromated X-ray source and a stable electron detection system enable the measurement of high resolution, highpressure XPS spectra with an extremely low noise to signal ratio. Figure 7 shows the gas phase XPS measurements of the C 1s and O 1s core levels for (a) pure CO 2 and (b) a CO 2 /O 2 mix (3% O 2 ) taken at 1.5 ± 0.001 mbar. The spectra are dominated by the C 1s and O 1s main lines with binding energies of 293.5 eV and 537.2 eV and FWHM of 600 meV and 700 meV corresponding to the C 1s and O 1s single hole states. It can be seen that at 10.7 eV higher binding energy above the main C 1s and O 1s core lines of the CO 2 gas, the instrument clearly resolves the fine structure of the shake-up satellites. 29,30 Additionally, two small features attributed to the paramagnetic nature of O 2 appear between 538 and 541 eV. 31 The peak fitting of these main lines of O 2 and CO 2 gas allows the quantification of the amount of O 2 in the CO 2 /O 2 mix to be 3.4% ± 0.5% which matches well with the values set for the MFCs of 2.9% ± 0.15% by the gas delivery system. There have been relatively few studies utilising high pressure XPS to study the gas phase, and we hope that the spectra displayed here and this instrument will stimulate further work, particularly in the area of environmental chemistry, which is critically an important research topic.

D. Purification of gases
A final, but important point to make is the importance of purifying gases before using them in HPXPS experiments in the reaction cell. Even minute contamination levels as, e.g., present even in the highest standard grade available for most gases, can completely change the interaction between a solid sample and a gas phase, and ultimately influence the observed surface chemistry. Therefore, there is a crucial need to add additional gas purifiers to remove any unwanted level of contamination present in the delivered gases. It is not always clear in the previous work that the consideration of gas purity levels is described in enough detail, if at all. To demonstrate the importance of gas purity, we exposed a polycrystalline copper foil to CO 2 with and without these purification systems in place. Cu foil lends itself to such experiments, as its surface is highly reactive due to a multitude of defects, and is therefore very sensitive to any impurity within the overall gas flow. The sputter-annealed Cu foil was exposed to a constant gas flow of 1 mbar CO 2 gas of N5.0 grade with and without the additional purifier in place. After exposure to standard N5.0 grade CO 2 , a multitude of adsorbate peaks was observed in the C 1s spectrum (see Figure 8(a)). As such a large number of different states for the interaction of CO 2 with Cu seemed unlikely, the experiment was repeated with the additional purifier (SAES, MC1-804F) in place. This purifier removes impurities including H 2 O, O 2 , CO, and H 2 to a level of <1 ppbV and acids and organics to a level of <5 pptV, after initial conditioning by flowing 100 sccm of CO 2 through it for 4 h. The C 1s spectrum recorded after exposing the Cu foil to a constant gas flow of 1 mbar CO 2 shows a much "cleaner" spectrum with the number of peak contributions drastically reduced (see FIG. 8. C 1s core level on the surface of a Cu foil while under a constant gas flow of 1 mbar CO 2 (a) without and (b) with gas purification using a step size of 50 meV and a dwell time of 1000 ms. Figure 8(b)). Even this rather simple example of a gas-solid interaction clearly demonstrates the importance of gas purity, as well as the cleanliness of the overall sample environment, to gain a realistic and well founded understanding of the number and nature of adsorbed species evolving on a surface.

V. CONCLUSIONS
This paper describes the design principles and construction of a newly developed laboratory based HPXPS system which is able to acquire spectra at pressures up to 25 mbar and temperatures ranging from 100 K to 1200 K. The excellent performance of the instrument is due to the combination of the extremely stable electrostatic lens in the high-pressure analyser and the stability and intensity of the monochromated X-ray source, the versatility of the small volume reaction cell, and the precision and control of the gas delivery system. The top cell (and therefore the cone aperture) can be changed without breaking vacuum in the main chamber allowing the optimisation of the signal intensity for different pressure regimes. The sensitivity and energy resolution of the spectrometer have been investigated using an Ag foil. The best energy resolution at high vacuum for the Ag 3d 5/2 peak was 450 meV which is close to the theoretical energy resolution of 413 meV and so the use of the cell does not affect the performance of the spectrometer. Raising the pressure in the reaction cell to 25 mbar did not affect the energy resolution but resulted in signal loss which varies exponentially with the pressure. The CO 2 and CO 2 /O 2 gas phase data demonstrate the flexibility and stability of the system, as well as the high sensitivity and excellent resolution of the analyser, and the need for careful consideration of gas purification should be taken into account when performing HPXPS measurements.