Photolysis of Fluorinated Graphites with Embedded Acetonitrile Using a White-Beam Synchrotron Radiation

Fluorinated graphitic layers with good mechanical and chemical stability, polar C–F bonds, and tunable bandgap are attractive for a variety of applications. In this work, we investigated the photolysis of fluorinated graphites with interlayer embedded acetonitrile, which is the simplest representative of the acetonitrile-containing photosensitizing family. The samples were continuously illuminated in situ with high-brightness non-monochromatized synchrotron radiation. Changes in the compositions of the samples were monitored using X-ray photoelectron spectroscopy and near-edge X-ray absorption fine structure (NEXAFS) spectroscopy. The NEXAFS N K-edge spectra showed that acetonitrile dissociates to form HCN and N2 molecules after exposure to the white beam for 2 s, and the latter molecules completely disappear after exposure for 200 s. The original composition of fluorinated matrices CF0.3 and CF0.5 is changed to CF0.10 and GF0.17, respectively. The highly fluorinated layers lose fluorine atoms together with carbon neighbors, creating atomic vacancies. The edges of vacancies are terminated with the nitrogen atoms and form pyridinic and pyrrolic units. Our in situ studies show that the photolysis products of acetonitrile depend on the photon irradiation duration and composition of the initial CFx matrix. The obtained results evaluate the radiation damage of the acetonitrile-intercalated fluorinated graphites and the opportunities to synthesize nitrogen-doped graphene materials.


Introduction
Fluorinated graphites are layered carbon materials possessing a good chemical, mechanical, and thermal stability [1,2]. Fluorination of graphite using inorganic fluorides at room temperature produces compounds with a composition CF x , where x is usually below 0.5 [3,4]. The molecules from the reaction media fill the space between the fluorinated layers and they can be replaced by other inorganic or organic guests [5]. Since such compounds are stable at ambient conditions for a long time, the fluorinated graphites are considered as containers for the storage and transport of volatile and hazardous substances [6].
The distance between fluorinated layers varies from~0.6 to~1.5 nm depending on the size and packing of the guest molecules [7]. A very weak (if any) interaction between the layers causes a two-dimensional (2D) magnetic behavior of these materials [8]. Their exfoliation in appropriate solvents allows producing thin films of the fluorinated graphene layers [9], which are promising materials for gas sensors and energy applications [9,10]. Guest molecules affect the thermal stability of the fluorinated graphite compounds [5] and The content of carbon, fluorine, bromine, and nitrogen in the samples was determined from the analysis of products of high-temperature destruction of samples in an oxygen atmosphere [19]. According to the obtained data, the composition of the yellow sample (49 days with 13.89 wt% BrF 3 in Br 2 ) can be represented as CF 0.5 Br 0.005 0.070CH 3 CN and the green-brown sample (87 days with 5.04 wt% BrF 3 in Br 2 ) as CF 0.3 Br 0.005 0.054CH 3 CN. Bromine found in these compounds forms covalent bonds with carbon edges of graphite domains. Below, the studied samples will be denoted CH 3 CN@CF 0.5 and CH 3 CN@CF 0.3 .

Measurements
XPS spectra of initial samples were measured on a Specs PHOIBOS 150 spectrometer (Specs GmbH, Berlin, Germany) using an Al K α excitation radiation (1486.7 eV). The spot size of the photon beam was about 3 mm. The Casa XPS 2.3.15 software (Casa Software Ltd., Teignmouth, UK) was used for data processing. The C 1s, F 1s, and N 1s spectra were fitted by a product of Gaussian-Lorentzian (7:3) peaks after subtraction of a Shirley background. The binding energies were calibrated to the C(sp 2 ) component energy at 284.5 eV.
Irradiations of samples by non-monochromatized SR light (white beam) and subsequent XPS and NEXAFS experiments were carried out at the Russian-German dipole beamline (RGBL Dipole, BESSY II, Berlin, Germany) operated by Helmholtz-Zentrum Berlin für Materialien und Energie. The total light intensity can be estimated as 50 mJ/cm 2 [27]. The samples are non-conducting; therefore, they were deposited on copper substrates with a scratched surface (roughness~100 µm) in the thinnest possible layers. The substrates were fixed on a holder and placed in a vacuum chamber providing a residual pressure of 10 −10 mbar. After acquiring the NEXAFS spectra in total electron yield (TEY) mode, the samples were exposed to a non-monochromatized photon beam for a certain period, and XPS and NEXAFS spectra were recorded directly for the irradiated spots (~1 × 1 mm). The measurements of the spectra accompanied each step of the sample irradiations. The irradiation experiments were repeated at different spots of the samples, and they showed the same trend in the spectral modifications depending on the exposure time. XPS spectra were excited by a photon energy of 830 eV. The binding energies were aligned to the position of the Au 4f 7/2 line at 84 eV recorded from a clean Au foil. Mass-spectra were registered upon an irradiation of CH 3 CN@CF 0.3 in a scan mode for m/z (~30 scans during 4 s). The residual gas analyzer Extorr XT100M (Extorr Inc., New Kensington, PA, USA) was operated with an electron impact ionizer with an energy of 70 eV.
A graphene fragment of a C 96 composition and D 6h symmetry was taken to construct the fluorinated models. Saturation of dangling bond of an edge carbon atom by one fluorine atom yielded the C 96 F 24 model. An attachment of fluorine atoms to both sides of the basal graphene plane, like in fluorographene [31], and bonding of an edge carbon atom with two fluorine atoms produced the C 96 F 120 model. The geometries of the models were optimized by an analytic gradient method to default convergence criteria. Then, we removed 34 central fluorine atoms from the C 96 F 120 fragment to form aromatic and polyene carbon areas in partially fluorinated graphene according to that observed experimentally [32]. The obtained partially fluorinated C 96 F 86 model was optimized at fixed positions of the boundary atoms.
Theoretical NEXAFS C K-and F K-edge spectra were plotted using the results of DFT calculations of the fluorinated models, where a carbon atom or a fluorine atom was replaced by an atom of nitrogen or neon, respectively. This so-called (Z + 1) approximation accounts for the effect of the core level hole on spectral profile [33]. To compensate for an increase in the number of valence electrons, the system charge was +1. The (Z + 1) approximation was used for the selected carbon or fluorine atoms located in structurally non-equivalent positions. The geometries of the structures with a neon were not optimized to avoid detachment of the neon atom. Spectral intensities were calculated as the sum of the squares of the coefficients, with which the atomic orbitals of nitrogen or neon participate in the formation of unoccupied molecular orbitals (MOs). The calculated intensities were broadened by Lorentz functions with a variable width of 1.4-4.0 eV, increasing with the photon energy, the spectral background was described by an arctan function [34]. X-ray transition energies were determined as the difference between the Kohn-Sham energies of the virtual MOs of the models calculated within the (Z + 1) approximation and energy of the core levels of the selected carbon or fluorine atoms, taken from the calculation of the ground state of the fluorinated model. The spectrum for a central nitrogen atom in the C 96 F 24 model calculated within the (Z + 1)-approximation ( Figure S1, Supporting Information) was aligned to the experimental C K-edge spectrum of graphite by the position of π* and σ* peaks. The obtained scaling formula was used to calibrate the energy of other theoretical C K-edge spectra. The calibration of energy for theoretical F K-edge spectra was done from the comparison of the calculated spectrum for the C 96 F 120 model and the experimental spectrum of fully fluorinated graphite (CF) n ( Figure S1, Supporting Information).

XPS C 1s and F 1s Spectra
XPS measurements were used to evaluate the changes in the composition of CH 3 CN@CF 0.3 and CH 3 CN@CF 0.5 samples before and after illumination with high-brilliance non-monochromatized SR light. The content of the elements was determined from the survey spectra (not shown) taking into account atomic subshell photoionization crosssections of elements at a given excitation energy. Atomic concentrations of main elements are collected in Table 1. The content of fluorine and nitrogen in the two studied samples differs by a factor of two in line with the data of elemental analysis. The XPS-derived F/C ratio is 0.16 for CH 3 CN@CF 0.3 and 0.37 for CH 3 CN@CF 0.5 . Since XPS is a surfacesensitive method, it detects the low content of fluorine in the upper surface layers of the samples as a result of their partial de-fluorination due to the contact with H 2 O present in laboratory atmosphere [35]. Higher oxygen content on the surface of CH 3 CN@CF 0.3 than for CH 3 CN@CF 0.5 may indicate an easier replacement of fluorine by oxygen in this sample. The weakness of C-F bonds in CH 3 CN@CF 0.3 results in almost complete removal of surface fluorine under the photon irradiation. While the CH 3 CN@CF 0.5 sample keeps about 3 at% of fluorine even after 200-s exposure to polychromatic synchrotron light. XPS also detects nitrogen from CH 3 CN molecules in both initial samples and after each step of the irradiation. Table 1. XPS determined content (at%) of main elements in CH 3 CN@CF 0.3 and CH 3 CN@CF 0.5 samples before and after exposure to polychromatic synchrotron light for 80 and 200 s. The ratio of the areas of the C-CF to C-F components in the XPS C 1s spectra (last column).  Figure 1 compares XPS C 1s and F 1s spectra of the samples. C 1s spectra of initial CH 3 CN@CF 0.5 and CH 3 CN@CF 0.3 are fitted by four components (Figure 1a,b). The binding energies and the relative areas of the components are listed in Table S1. A weak component at~287 eV corresponds to carbon in guest acetonitrile molecules [36]. The low-energy component at 284.5 eV originates from sp 2 -carbon areas remaining in the fluorinated layers and its intensity is higher for the CF 0.3 matrix than for CF 0.5 . The peaks at 288.7 and 286.1 eV in the CH 3 CN@CF 0.5 spectrum characterize carbon atoms covalently bonded to fluorine (C-F) and located at CF groups (C-CF), respectively [37]. These peaks are downshifted by 0.7 eV for CH 3 CN@CF 0.3 due to the weakening of C-F bonds [38,39]. The intensity of the C-F peak correlates with fluorine content in the samples ( Table 1). The ratio of the C-F component to the total area of the C 1s spectrum gives matrix stoichiometry CF 0.24 for the CH 3 CN@CF 0.3 sample and CF 0.43 for the CH 3 CN@CF 0.5 sample. The C-CF/C-F ratio gives the average number of bare carbon atoms near CF groups. The ratio 1 for CH 3 CN@CF 0.5 ( Table 1) indicates an average of one bare carbon neighbor for a CF group. Such a ratio can be realized when CF chains alternate with bare carbon chains [40,41]. An increase in the ratio value for CH 3 CN@CF 0.3 is associated with the shortening of CF chains and increase in numbers of two bare carbon neighbors for CF groups located at the edges of short CF chains. The relative intensity of the C-F component decreases in the C 1s spectra of irradiated samples (Figure 1a,b) due to the removal of fluorine. The shift of the C-F and C-CF components to lower binding energies indicates the weakening of C-F bonds as compared to those in the initial samples. New components located at~289.5 and~292 eV are especially noticeable in the spectrum of CH 3 CN@CF 0.5 after 200 s of the irradiation. They are assigned to carbon bonded with two (CF 2 ) and three fluorine atoms (CF 3 ) [42]. Thus, white beam partially destroys graphitic lattice. The detached carbon and fluorine atoms are combined with the CF 2 and CF 3 groups that bind to the edges of vacancies [43]. Analysis of the XPS C 1s spectra reveals that the composition of the CH 3 CN@CF 0.5 and CH 3 CN@CF 0.3 samples irradiated for 200 s is CF 0.18 and CF 0.10 , respectively.
XPS F 1s spectra of initial CH 3 CN@CF 0.5 and CH 3 CN@CF 0.3 exhibit a single peak at 687.2 and 686.6 eV, respectively (Figure 1c,d). These binding energies correspond to fluorine covalently bonded with carbon [15,44]. The exposure of the samples to the nonmonochromatized light leads to the emergence of fluorine states possessing higher binding energies. The F 1s components at 689.0 and 691.0 eV can be attributed to fluorine in CF 2 and CF 3 groups [45], or the atoms located in densely fluorinated regions [44] like in (CF) n . However, the C-CF/C-F ratio in the XPS C 1s spectra of the samples (Table 1) indicates that most CF groups have one or two bare carbon atoms as their neighbors, and this differs from the fluorine arrangement in (CF) n . Amounts of CF 2 and CF 3 groups are larger in the irradiated CH 3 CN@CF 0.5 than in the irradiated CH 3 CN@CF 0.3.

NEXAFS C K-Edge and F K-Edge Spectra
NEXAFS spectra measured before and after sequential irradiation of CH 3 CN@CF 0.5 and CH 3 CN@CF 0.3 for 20, 80, and 200 s are presented in Figure 2. The difference in the binding energies of the XPS F 1s peak and the C-F component of the XPS C 1s spectrum (Figure 1) is used for the energy alignment of NEXAFS C K-and F K-edge spectra of the particular sample.  (Figure 2a,b) assigned to the electron transitions from C 1s levels onto unoccupied π-type and σ-type states for sp 2 -hybridized carbon, respectively [46][47][48]. The peaks, which appeared between these resonances at 287.8 and 288.8 eV and labeled C 1 and C 2 , correspond to carbon bonded with fluorine [49,50]. In the spectrum of starting CH 3 CN@CF 0.5 , these peaks are more prominent, while the π* resonance has the lowest intensity (Figure 2a). The letter C denotes the position of σ*-edge for the fluorinated areas because it coincides with the last intense peak (labeled F) of the F K-edge spectra (Figure 2c,d). The shoulder F 1 at 686.5 eV and the peak F 2 at 687.4 eV align with peaks C 1 and C 2 of the C K-edge spectra and therefore they refer to the C-F bonds. The illumination of CH 3 CN@CF 0.5 and CH 3 CN@CF 0.3 samples with polychromatic synchrotron beam results in the suppression of C 1 and C 2 peaks in C K-edge spectra and F 1 and F 2 peaks in F K-edge spectra and the growth of relative intensity of π* resonance from sp 2 -carbon. The changes are stronger with increasing exposure time and correlate with the behavior observed in the XPS C 1s spectra of the samples (Figure 1).
To interpret NEXAFS C K-edge and F K-edge spectra in detail, NEXAFS spectra for structurally nonequivalent carbon and fluorine atoms present in the partially fluorinated graphitic monolayer are constructed (Figure 3). The spectra of the starting and irradiated for 200 s CH 3 CN@CF 0.5 sample are chosen for the modeling (Figure 3a,b). The calculated fluorinated graphene fragment is shown in Figure 3c. Theoretical spectra are constructed for the carbon and fluorine atoms from CF groups surrounded by three (CF-3), two (CF-2), one (CF-1), and none (CF-0) fluorinated carbon atoms. We also calculate the C K-edge spectra for bare carbon atoms from polyene-like chain (C-ch) and aromatic naphthalene-like area (C-ar). Comparison of the C K-edge spectrum of initial CH 3 CN@CF 0.5 with the calculated spectra shows that energy of π* resonance corresponds to the position of the low-energy intense peak in the spectrum of C-ch (Figure 3a). This result indicates that most of the sp 2 -hybridized carbon atoms in the fluorinated CF 0.5 layers form polyene-like chains, that is in agreement with the previous data [40,51]. Peak C 1 originates from carbon in isolated CF groups (CF-0), while carbon atoms from CF-1 and CF-2 groups, where CF groups have one and two CF neighbors, respectively, contribute to the peak C 2 in the experimental spectrum. Fluorine atoms from CF-1 and CF-2 groups are responsible for intense peaks F 2 and F 3 in the F K-edge spectrum of CH 3 CN@CF 0.5 (Figure 3b). Shoulder F 1 is assigned to fluorine from isolated CF-0 groups. Analysis of MOs calculated in the (Z + 1)-approximation reveals that spectral features F 1 , F 2 , and F 3 correspond to C-F bonds of σ*-type ( Figure S2, Supporting Information). The difference in energy is due to the different local distribution of electron density between this bond and the neighbors. The high-energy peak F is formed by an overlapping of F 2p x,y orbitals with neighboring C-C σ-bonds.
The cumulative theoretical C K-edge spectrum obtained by summing the spectral intensity of carbon from C-ch, CF-0, CF-1, and CF-2 taken in a ratio of 1.8:1:1:1.8 perfectly repeats the shape of the experimental spectrum of initial CH 3 CN@CF 0.5 (two upper curves in Figure 3a). The cumulative F K-edge spectrum being a sum of the theoretical spectra of fluorine from CF-0, CF-1, and CF-2 taken in a proportion of 1:1:1.8 also agrees well with the experimental spectrum of CH 3 CN@CF 0.5 (two upper curves in Figure 3b). CF-3 groups with three CF neighbors are not necessary to define all spectral features; probably, they are hardly formed in the synthesis conditions used. Fluorine distribution in the layers of a CF 0.5 composition is mainly realized as CF chains separated by polyene-like carbon chains.
Exposure of CH 3 CN@CF 0.5 to white beam for 200 s causes an increase and broadening of π* resonance and a significant decrease in the intensity of C 1 and C 2 peaks of the C K-edge spectrum (Figure 3a). To describe this spectral profile, the spectra of CF-0, C-ch, C-ar, and central atom in the graphene model ( Figure S1) are taken in a ratio of 1:1:2:2. The F K-edge spectrum of the irradiated CH 3 CN@CF 0.5 shows mainly a decrease in the intensity of F 2 peak (Figure 3b) and only isolated fluorine atoms from CF-0 groups are needed to simulate the experimental profile. These results indicate that long-term irradiation of the fluorinated graphitic layers leads to their strong defluorination. The remaining fluorine atoms are separated from each other. A significant removal of fluorine occurs after the first 20-s irradiation and it is more pronounced for the CH 3 CN@CF 0.3 sample (Figure 2).

Electronic State of Nitrogen
Electronic state of nitrogen from acetonitrile molecules embedded between the fluorinated graphitic layers is revealed using XPS N 1s and NEXAFS N K-edge spectra. The XPS N 1s spectrum of initial CH 3 CN@CF 0.5 exhibits a single symmetrical peak at~399 eV (Figure 4a). The white beam illumination of the sample for 80 s causes the appearance of two new components located at~398.1 and~400.5 eV and assigned to pyridinic N and -NH-species in carbon rings (pyrrolic N), respectively [52]. The fraction of the pyrrolic N increases with the irradiation duration. This result indicates that CH 3 CN molecules are decomposed under photon-beam treatment and the released nitrogen and hydrogen atoms are incorporated into the surrounding CF x layers. Insertion of nitrogen into fluorinated graphitic layers was early observed for similar CH 3 CN@CF x samples heated at 250 • C in a vacuum [53].
NEXAFS measurements were performed to examine the initial stages of sample irradiation in more details, thus Figure 4b compares N K-edge spectra of starting CH 3 CN@CF 0.5 and that after exposure to white beam during 2, 5, 20, 80, and 200 s. The irradiation of CH 3 CN@CF 0.5 for 2 s already results in degradation of acetonitrile. The pre-edge peak C≡N located in the initial spectrum at~399.9 eV shifts by 0.3 eV to the low-energy region, its intensity decreases, and new peak around 401.0 eV appears. Our DFT calculations show the shift of the C≡N peak can be attributed to the formation of HCN ( Figure S3, Supporting Information). The peak at about 401.0 eV corresponds to pyrrolic N species [54] and N 2 molecules [55]. The intensity of this peak strongly reduces in the N K-edge spectrum of CH 3 CN@CF 0.5 (Figure 4b) and CH 3 CN@CF 0.3 ( Figure S3  NEXAFS N K-edge spectra of starting and irradiated CH 3 CN@CF 0.3 sample measured in a range of 397.5-403.0 eV are shown with a purpose to study the pre-edge peaks in detail (Figure 4c). The resonance emerging around 401.0 eV is resolved into five peaks characteristic of vibrations of N 2 molecules [56]. This proves the formation of N 2 molecules as a result of the photolysis of CH 3 CN and the retention of these molecules between the fluorinated graphitic layers. The content of the trapped N 2 molecules decreases with continuing irradiation and the molecules are not detected after 80-s irradiation. The peak at 400.6 eV (Figure 4b,c) corresponding to pyrrolic N at the boundaries of vacancies in CF x layers is identified according to the DFT calculations ( Figure S3, Supporting Information). Incorporation of pyridinic N occurs at the first stages of samples irradiation and raises the shoulder at 398.8 eV in the N K-edge spectra (Figure 4b,c).
Mass spectrum of ion species measured upon the irradiation of CH 3 CN@CF 0.3 sample is presented in Figure 5. The background ion peaks from residual air and molecular ion peaks are highlighted in black and red, respectively. The signal of CO 2 + ions (m/z = 44) arising from the sample surface is taken as~100%. Note that the amplitude of background H + and H 2 O + ions is an order of magnitude larger than this signal. The spectrum detects the ions being the decomposition products of CH 3

Discussion
Fluorinated graphites of the composition CF 0.5 and CF 0.3 are insulators, and their XPS spectra are measured in a laboratory spectrometer where charging of sample under X-ray photon exposure is compensated. Analysis of XPS data indicates that C-F bonds are covalent and they are weaker in CF 0.3 layers. These layers also contain larger fractions of aromatic areas and bare carbon atoms located nearby CF groups as compared to CF 0.5 layers. The DFT modeling of NEXAFS C K-and F K-edge spectra of initial CF 0.5 reveals that fluorine atoms in the layers preferably form CF chains, alternating with polyene-like carbon chains. The CF chains are shorter in CF 0.3 layers.
NEXAFS C K-and F K-edge spectra of both studied samples exhibit a large decrease in the intensity of the peaks corresponding to the C-F bonds already after polychromatic photon beam exposure for 20 s (Figure 2). A probing depth of the spectra acquired in the TEY mode is a few nanometers [38], thus, we estimate that at least ten upper layers lose fluorine. The areas of C-F components in XPS C 1s spectra of both samples decrease by about 2.4 times after irradiation for 200 s (Table S1, Supporting Information). The shape of the F K-edge spectrum of the irradiated CF 0.5 well corresponds to the electronic state of fluorine in isolated CF groups, i.e., not adjacent to other CF groups (Figure 3b).
A part of fluorine atoms removed from basal graphitic planes is attached to their edges as CF 2 and CF 3 groups and amounts of these groups are markedly larger for CF 0.5 layers (Figure 1a,b). The formation of these groups means that a fluorine atom may leave a fluorinated graphitic layer together with the carbon neighbor. The CF + and CF 3 + ion signals detected in the mass spectrum upon the irradiation of CH 3 CN@CF 0.3 sample ( Figure 5) confirm this. According to our DFT calculations, fluorine atoms located in the center of CF chains have larger binding energies than fluorine atoms at the chain edges (Table S2, Figure S4, Supporting Information). Therefore, the C-F bonds in long CF chains that are present in CF 0.5 layers are stronger than bonds between bare carbon atoms and fluorinated ones (C−CF). The latter C−CF bonds break more easily, producing vacancies in the carbon network. The edge C-F bonds, which are predominant in short CF chains of CF 0.3 layers, dissociate under the radiation and this explains the less efficient destruction of C-C bonds observed for this sample.
NMR study of fluorinated graphites with embedded acetonitrile molecules revealed that interactions between the guests and matrices have van-der-Waals character [57]. Such weak interactions should not influence the radiation stability of the constituents. According to NEXAFS N K-edge spectra the CH 3 CN molecules located between fluorinated graphene layers dissociate under irradiation for 2 s (Figure 4b,c). The features detected in the spectra measured at various stages of the irradiation are assigned to HCN and N 2 molecules, and pyridinic and pyrrolic nitrogen atoms at vacancies edges of CF x layers. The fractions of these nitrogen forms are determined from the decomposition of N K-edge spectra ( Figure S5, Supporting Information).
The evolution of nitrogen forms in CH 3 CN@CF 0.5 and CH 3 CN@CF 0.3 with the exposure time is illustrated in Figure 6a,b. The main product of the CH 3 CN photolysis is HCN molecules. Their preferable formation according to the path CH 3 CN→ CH + H + HCN was previously observed under the action of UV light [23], electron beam [20], and photon beam [22]. The HCN concentration after the first stage of the irradiation for 2 s is~64% for CH 3 CN@CF 0.5 and~57% for CH 3 CN@CF 0.3 , and it reduces to~49% for CH 3 CN@CF 0.5 and~47% for CH 3 CN@CF 0.3 after the 200-s exposure. Since the pyrrolic N is detected in the NEXAFS N K-edge spectra measured for the samples after the 80-s exposure, we suppose that decomposition of HCN with prolonged irradiation of samples contributes to the formation of this kind of nitrogen species. The initial photolysis of CH 3 CN molecules yields also N 2 molecules and pyridinic N atoms (Figure 6a,b). A weak signal of nitrogen ions was observed when gaseous CH 3 CN was photoionized by the monochromatic SR beam with a photon energy of 42 eV [20]. The CF x matrix probably facilitates the abstraction of the nitrogen atoms from acetonitrile molecules in our case. Mass spectrum of photo-induced ions detects the N 2 + signal ( Figure 5). The N 2 molecule fraction is twice the fraction of pyridinic N developed in CF x layers. However, with an increase in the exposure time, N 2 content decreases, which is accompanied by a growth of the content of pyridinic N. This behavior indicates photodissociation of N 2 and incorporation of the produced nitrogen atoms in the surrounding defective CF x layers. The process is faster for the CH 3 CN@CF 0.5 sample. The reason may be a larger amount of atomic vacancies in the layers as the XPS C 1s and F 1s spectra of the sample detect (Figure 1). Figure 6c schematically presents changes that occurred for fluorinated graphene layers with acetonitrile guests under the influence of non-monochromatized SR light. The matrix layers lose a part of fluorine and acquire nitrogen atoms. These atoms are located at the boundaries of vacancies, produced when fluorine atoms are removed from the layer together with carbon. The photolysis of acetonitrile produces N 2 and HCN molecules. DFT calculations show that these molecules are readily adsorbed on the nitrogen-doped CF x layer. Pyridinic N atoms and CF groups create the preferred positions for HCN and N 2 molecules, respectively.

Conclusions
Fluorinated graphites with the composition of the layers CF 0.3 and CF 0.5 were synthesized using a fluorinating agent BrF 3 at room temperature. DFT modeling of NEXAFS C K-edge and F K-edge spectra showed that fluorine atoms form the fluorinated carbon chains alternating with polyene-like carbon chains in CF 0.5 layers. These chains were shorter in the CF 0.3 layers, where the CF groups have two bare carbon neighbors on average. The interlayer space of the fluorinated graphites was filled by CH 3 CN. Photolysis of CH 3 CN@CF 0.3 and CH 3 CN@CF 0.5 was carried out using the zero-order light from the Russian-German dipole Beamline of the synchrotron source BESSY II. The photon irradiation led to a partial defluorination of the layers and the formation of vacancy defects. The XPS C 1s and F 1s spectra showed that the amount of vacancies is larger in the layers with an initial composition of CF 0.5 . C-F bonds in these layers were stronger than the bonds between the fluorinated carbon and bare carbon neighbor that caused the preferred breakage of the latter bonds. CH 3 CN molecules were completely decomposed during the first two seconds of the SR zero-order light exposure. The main products were HCN and N 2 molecules and pyridinic N atoms, introduced into the CF x layers at the vacancy boundaries. Upon further irradiation, N 2 molecules dissociated and the released nitrogen atoms gave mainly pyridinic N defects in the fluorinated graphene layers. This dissociation was faster in the CF 0.5 layers. The products of HCN photolysis contributed to the formation of pyrrolic N species. The study shows that the products of the photolysis of CH 3 CN depend on the time of irradiation and the fluorine loading of the fluorographitic matrix. Our results can be crucial when using CH 3 CN@CF x systems in environments with intense light from UV to soft X-rays.