Soft X-ray Heterogeneous Radiolysis of Pyridine in the Presence of Hydrated Strontium-Hydroxyhectorite and its Monitoring by Near-Ambient Pressure Photoelectron Spectroscopy

The heterogeneous radiolysis of organic molecules in clays is a matter of considerable interest in astrochemistry and environmental sciences. However, little is known about the effects of highly ionizing soft X-rays. By combining monochromatized synchrotron source irradiation with in situ Near Ambient Pressure X-ray Photoelectron Spectroscopy (in the mbar range), and using the synoptic view encompassing both the gas and condensed phases, we found the water and pyridine pressure conditions under which pyridine is decomposed in the presence of synthetic Sr2+-hydroxyhectorite. The formation of a pyridine/water/Sr2+ complex, detected from the Sr 3d and N 1s core-level binding energies, likely presents a favorable situation for the radiolytic breaking of the O-H bond of water molecules adsorbed in the clay and the subsequent decomposition of the molecule. However, decomposition stops when the pyridine pressure exceeds a critical value. This observation can be related to a change in the nature of the active radical species with the pyridine loading. This highlights the fact that the destruction of the molecule is not entirely determined by the properties of the host material, but also by the inserted organic species. The physical and chemical causes of the present observations are discussed.


S1. Core-levels fitting parameters
. C 1s fitting parameter at hv=750 eV in biased condition

S2. Differential charging and its elimination
Clays are insulating materials. Under irradiation by the X-ray beam, the holes created by the photoemission process cannot be replenished by electrons provided by the ground.
Therefore, the material becomes positively charged, and any reference with the analyzer Fermi level is lost. 1 Moreover differential charging can be observed as some parts of are better connected to the ground than others, and therefore charging is not uniform, laterally and in-depth. To eliminate charging we use an original procedure. We bias the sample positively (+30 V) with respect to the analyzer and the chamber walls that are both grounded. Positive biasing leads to an increase in the flow of stray negative species produced by the X-ray beam in the analysis chamber (electrons in UHV, electrons and anions in NAP conditions). This is a self-compensating flood gun effect, superior to usual flood guns as overcompensation (i.e. a negative charging of the surface) never occurs. The elimination of differential charging in the XPS probed layer can be monitored by minimizing the full-width at half maximum (FWHM) of the core-levels, especially Mg 2p, as this species nested at the center of phyllosilicate lath is not affected by chemical changes occurring at the surfaces.
We find that a bias of +30 V is optimal. At hν=450 eV and with a pass energy of 50 V, biasing is impossible for the N 1s spectrum, due to a reduction of the photoelectron kinetic energy to a too low value of ~20 eV. In all cases, the binding energies given here are corrected from the bias.
The Sr 3d and Mg 2p spectra of the grounded sample and biased sample are given in Figure S1 (a) and (b), respectively. They exemplify the beneficial effects of pressure increase (NAP conditions) and positive biasing. The corresponding full-width at half maximum (FWHM) of Mg 2p are given in Figure S1. Differential charging is particularly acute in UHV conditions, see Figure   S1 (a), bottom curve, for the grounded sample. The Sr 3d spectrum is fitted with two doublets, separated by 1.28 eV suggesting that two different chemical environments exist for the Sr 2+ ion (for Sr 3d fitting parameters see section S4). In fact, this is only apparent, as once the +30 V bias is applied, Figure S1(b), the spectrum can be fitted with a single doublet. The Mg 2p FWHM of the UHV spectrum also changes from 2.21 eV to 1.99 eV when the bias is applied.
The mere increase of pressure improves the spectral resolution when the sample is grounded, as shown in Figure S1 (a) and S1 (  Not only does the attenuation of differential charging decrease the FWHM of Mg 2p, but also the binding energy measured with respect to the gold substrate Fermi level decreases. For the biased spectra, the effect of increasing the pressure is shown in Figure S2 (b). When the sample is grounded, Figure S2 (a), the apparent binding energy of the Mg 2p spectrum stabilizes at 50.40 eV

S3. Can we distinguish surface Sr 2+ from interlayer Sr 2+ in UHV conditions?
The depth profiling ( Figure S3) shows that while the Mg 2p FWHM is constant at 2 eV, the FWHM of Sr 3d increases from 1.72 eV (surface sensitive conditions, imfp of ~ 1.2nm) at hv=450 eV to 2.04 eV (more bulk condition, imfp of ~2.5 nm) at hv=1050 when fitted with a single doublet.
Vertical differential charging is excluded as the FWHM of Mg 2p remains constant. However the situation is different for the counterion. Its binding energy may be affected by an initial state effect, related to different electrostatic potential energies at the open surface and within the interlayer.
Final state effects may be also expected, as the relaxation energy 5 of a surface Sr 2+ may be also different from that of an interlayer Sr 2+ as it lies at the interface between an oxide and the vacuum, and is not sandwiched between two phyllosilicate sheets. However, Figure S3 shows that The broadening of the Sr 3d doublet due to the different chemical environments is so small that it cannot be resolved experimentally.

S4. Electrostatic potential energy variation between the center of the clay layer and the counter ion plane
To estimate the electric field value in the clay structure, we consider three infinite successive parallel planes, with charge densities of +σ/2, -σ, and +σ/2, plunged into a dielectric medium of permittivity . d is the distance between the between the -σ plane and the +σ/2 ones. By application of Gauss' law and of the superposition principle, the field is zero outside the three planes, equal to σ/2 between the + σ/2 plane and the -σ one and equal to -σ/2 between the -σ plane and +σ/2 one. Between the central -σ plane and the +σ/2 ones, the electrostatic potential energy varies by ∆qV =-eσd/2, where e is the elementary charge This 3-plane unit cell can be used to build larger tactoids comprising N such unit cells. The chosen cell, and the boundary conditions, it involves are such that the divergence of the electrostatic energy with increasing N is avoided, as it occurs when the stacking of N +σ/-σ alternately charged planes is considered. 6 Each unit cell (0.5240.909 nm 2 ) contains a charge of 0.8 e, therefore | | = 0.34 C.m -2 . The permittivity  is 0r, where 0 is the vacuum dielectric constant and r is the relative dielectric constant. For the 0W and 1W hydration states, r is estimated to be approximately 5 and 7, respectively. 7 In hydroxyhectorite the negatively charged plane is centered in the octahedral layer.
Therefore, taking d between 0.38 nm (half a sheet width) and 1.02 nm (the distance between the middle of the sheet and the middle of the interlayer in the 1W state) and r=7 one finds that |∆ |

S5. Insertion of pyridine molecules in 1W Sr 2+ -hydroxyhectorite
Let us consider the adsorption at the surface of the sheets or in the interlayer. Given the charge of the unit cell (0.8e) and the basal plane unit cell dimension of hectorite (0.524 nm × 0.909 nm 8 ), there is an average free area around each strontium ion of 1.21 nm 2 (corresponding to an average distance between Sr 2+ ions of 1.1 nm). This corresponds to ~5 ditrigonal cavities per intercalated ion. Assuming that at RH=7% the water molecules are all around the Sr 2+ (about 4 for the 1W state), 9 there is room between the cations for the pyridine molecule, of diameter ~0.5 nm, to physisorb on siloxanes or to lock in the unoccupied ditrigonal holes (DFT calculations of pyridine interaction with the central HOin the Sr 2+ free cavities lack).
We can get an estimate of the maximal pyridine insertion, based on the volume occupied by the hydrated Sr 2+ /pyridine complex (assuming a basal spacing increase of 1.2 nm) and on that occupied by the pyridine molecule itself. The average free area around one strontium ion is 1.21 nm². In the model given in Refs. 10 and 11 the basal spacing increases by 1.2 nm (swelling due to water and pyridine), thus the free space around the strontium ion is 1.45 nm 3 . Now the dimensions of the complex between the hydrated Sr 2+ and pyridine can be estimated to be 1.2 nm (height) ×0.9 nm (width) ×0.26 nm (depth, i.e. the size of the Sr 2+ solvation sphere 12,13 ). The complex occupies a volume of 0.28 nm 3 . Consequently, of the 1.45 nm 3 available space, 0.28 nm 3 are occupied by the complex, leaving 1.17 nm 3 free for "non H-bonded" pyridine molecules. Pyridine is a planar molecule with a diameter of 0.5 nm, whose minimal occupation volume can be estimated to be 0.063 nm 3 taking a face-to-face - stacking of 0.32 nm. This means that a maximum of 19 "non H-bonded" pyridine molecules per strontium ion can be inserted in the available space. Therefore, the maximum non "H-bonded" to "H-bonded" ratio is 19/4, that is 4.75.
A noticeable effect of pyridine adsorption onto the clay (at the surface and within the interlayer) is the strong damping of the Sr 3d and Mg 2p signals when pyridine partial pressures are greater than or equal to 0.3 mbar, as shown in Figure S4. See also the damping of the condensed-phase O 1s spectra in Figure 2 of the main paper.

S6. Ionization cross sections at h=450 and h=750 eV
In order to discuss the eh pair generation by the soft X-ray beam, we first calculate the total cross sections corresponding to the Sr0.4Mg5.2Li0.8Si8.0O20(OH)4 formula at h=450 eV and h=750 eV, see Grounded high cross-section, but the metal atom contribution is less than at h=450 eV. We give also the total cross-sections for the pyridine molecule. As at h=750 eV the energy distances from the C and N K edges are greater than at 450 eV, the total cross-section is smaller.

Orbital
Cross-sections in Mbarn at

S7. Hole-electron pair formation and irradiation doses
According to Cazaux, 15 the number of eh pairs n(eh) produced by one (absorbed) photon of energy h and having excited a given core-level is: where E(eh) the energy for creating a eh pair, the kinetic energy of the photoelectron, and the kinetic energy of the Auger electron, and the probability of filling the core-hole via an Auger decay. is close to one for the light elements of the 2 nd period, C, N, O (KLL transitions) and of the 3 rd period, Mg, Si (LVV transitions). With = ℎ − , ≈ , and ≈ 1 (where BE is core-level binding energy) one gets: Concerning the heavier element Sr (5 th period) the higher binding energy level that is excited is the M1 level (3s). Its decay probability is also close to one.

E(eh)
is twice to three times the material gap. It is 17 eV for SiO2, 15 for a band gap of 9.3 eV. 16 Given that the band gap of hectorite is 4.0-4.5 eV, 17 we will take E(eh) equal to 9 eV in our calculations. The pair generation factor per unit volume and per unit time ( ℎ) may also be expressed as a function of Φ 0 the incident photon flux, and the linear absorption coefficient of the specimen for the X-ray photons of interest as: The mass absorption coefficient μ* in dry Sr 2+ -hydroxyhectorite of composition corresponding to a dose rate of 3×10 9 Gy/s, two orders of magnitude greater than in the present case. However, typical doses received by the samples are 1.4×10 5 Gy for the whole experiment, three orders of magnitude smaller than during a typical XPS scan of 25 s.

S8. C 1s spectrum of the grounded sample
As shown in Figure S5, the C 1s peak of gas phase pyridine is found at a binding energy of 286.6 eV (maximum). The ionization energy (with respect to vacuum level) of the o-, m-and p-carbons are at 291.1, 290.5 and 290.8 eV respectively. 22 Gas phase C 1s ionization energies of typical small CxHy molecules CH4, C2H2, C2H4 and C2H6 are measured at 290.9, 291.2, 290.7 and 290.76 eV, respectively. 23 Therefore, any small non nitrogenous CxHy molecule going into the gas phase after pyridine dissociation cannot be distinguished from gaseous pyridine. Figure S5. C 1s spectrum measured at h=750 eV (grounded sample). Pyridine Gas phase