Quasi‐Freestanding Graphene via Sulfur Intercalation: Evidence for a Transition State

Sulfur intercalation of a carbon rich (63×63)R30∘$(6\sqrt {3}\times 6\sqrt {3})R30\,^\circ$ reconstruction on silicon carbide, also known as buffer layer, is reported. In a two‐zone furnace a sulfur rich precursor is heated and the gaseous species is transported for intercalation by an argon flow to the sample. Successful intercalation can be confirmed by X‐ray photoelectron spectroscopy and low‐energy electron diffraction. Angle‐resolved photoelectron spectroscopy reveals a p‐type doping of the intercalated samples. In some cases only partial intercalation appears with non‐intercalated sulfur on top of the remaining buffer layer areas. Further annealing of such samples leads to a migration of the non‐intercalated sulfur under the buffer layer areas, indicating that the sulfur bonded to the buffer layer constitutes a transition state.


Introduction
Geim and Novoselov's groundbreaking work on graphene [1] paved the way for the research area of two-dimensional (2D) materials.An important prerequisite for the implementation of graphene in applications is the large-scale and homogeneous growth of graphene, which can be achieved by high-temperature graphitization of silicon carbide (SiC). [2]During this process, SiC bilayers decompose and the carbon atoms arrange in a graphene lattice while volatile silicon atoms evaporate. [3]The growth on the Si terminated side of the semiconductor SiC starts with the formation of a (6 DOI: 10.1002/admi.202300725 it is partially covalently bound to the SiC substrate and therefore lacks the  band of graphene. [4]An increase of the growth temperature leads to the formation of a new 6 [6] The growth process of epitaxial graphene on SiC(0001) was first achieved in UHV, [2] later performed in argon (Ar) atmosphere at atmospheric pressure [7] and most recently optimized by the method of polymerassisted sublimation growth (PASG). [8]his has improved the homogeneity and quality of the graphene in recent years.
The isolation of graphene sparked the investigation of other 2D materials.
There is now a whole range of 2D materials with emerging properties contrasting their bulk species.For example, the transition metal dichalcogenide (TMDC) molybdenum disulfide (MoS 2 ) changes from an indirect (multi layer) to a direct semiconductor (single layer). [9,10]Furthermore, the combination of different 2D materials influences the electronic properties of the system.[17] Experiments show possible photodetecting and storage applications, where the samples were prepared by layer-by-layer dry transfer. [18,19]However, a high-quality and homogeneous growth process is needed for commercial utilization.Recently, the intercalation of epitaxial graphene on SiC has been utilized to grow 2D materials at the interface, e.g. the 2D metals gallium (Ga), indium (In) and tin (Sn). [20][23] In a similar fashion, the formation of 2D TMDCs by means of sequential intercalation is of interest to study proximity effects in epitaxial graphene.However, to the best of our knowledge, the successful intercalation of chalcogen atoms (S, Se, Te) has not been established so far.
Various methods for intercalation are known so far.However, the intercalant not only decouples the 6 √ 3 from the SiC substrate but also influences its electronic properties.For example, the decoupling of the 6

√
3 from the SiC substrate can be achieved by annealing in an atmosphere of hydrogen. [24,25]The intercalation of oxygen as another gaseous species can be achieved by annealing the sample either in molecular O 2 atmosphere [26] or in air. [27]oth, hydrogen [24] and oxygen [28] intercalation leads to p-type doped graphene.A different approach is necessary for solid intercalants.Elements like gold, [29] germanium, [30] and bismuth [31] can decouple the 6 3 by depositing a thin layer of the solid on top of the buffer layer and subsequent annealing in UHV.Furthermore, the intercalation of bismuth is possible by ion implantation of the atoms through the 6 3 in the SiC substrate followed by annealing in UHV. [32]Unlike the gaseous intercalants, all three solids result in p-or n-type doped graphene, depending on the amount of intercalated material.The intercalation of volatile elements like antimony (Sb) can be achieved by the deposition of a thick layer on top of the 6 3 and subsequent annealing in argon at atmospheric pressure, which leads to an almost undoped QFMLG. [33]Lanthanides such as ytterbium [34] or gadolinium [35] can be intercalated by evaporation of the atomic species and simultaneous annealing of the 6 √ 3 sample, whereby the decoupled graphene exhibits strong n-type doping up to the Van Hove singularity.Here, we present the successful intercalation of sulfur to the 6 √ 3/SiC interface by using a solid sulfur precursor, which is converted into the gaseous phase.Sulfur vapour is transported to a 6 √ 3 surface by an Ar flow at low pressure.

Results and Discussion
Figure 1 shows XP survey spectra of a 6 √ 3 sample before (top) and after (bottom) intercalation.The spectrum of the pristine 6

√
3 sample shows its typical core level signals.We find the silicon core levels (Si2p at E b ≈ 101.5 eV, Si2s at E b ≈ 153.0 eV) of the SiC substrate.Furthermore, the 1s core level of the carbon atoms at E b ≈ 284.0 eV can be identified, originating from the 6 √ 3 surface and the substrate.There is also a small amount of oxygen at E b ≈ 531.5 eV (O1s) which, according to our experience, is bonded to tiny regions of the SiC substrate where no 6 √ 3 is grown.
After characterization of the 6 √ 3 sample, the intercalation process was performed.For this purpose a two-zone tube furnace was used where each zone is individually heatable.The sulfur source (precursor) for this experiment was pyrite (FeS 2 ), which was positioned in the first zone while the sample was positioned in the second zone (see Figure 10 in Section 4).Thermal decomposition of the precursor releases sulfur into the gas phase and an argon gas flow transports the sulfur to the sample.More information on the process can be found in the experimental section.In the survey spectrum of the sample after the process with pyrite (Figure 1, bottom) the emergence of sulfur core levels at E b ≈ 163.0 eV (S2p) and E b ≈ 227.0 eV (S2s) is observed.Otherwise no further signals can be detected, proving the advantage of the FeS 2 precursor compared to elemental sulfur (cf.Supporting Information).In order to determine whether or not the sulfur intercalation was successful, a closer look at the C1s, Si2p, and S2p core levels is required.
Figure 2 shows the C1s (a) and Si2p (b) signals of a pristine 6 3 sample (top) and a sample after the process with m FeS 2 = 266 mg pyrite (bottom), hereafter referred to as sample I1.Also shown is the S2p core level (c) of sample I1 after the process.We will briefly discuss the C1s and Si2p spectra of the pristine 6 3 sample to understand the spectra after the process with FeS 2 .In the C1s spectrum characteristic components of a 6 3 on SiC(0001) are visible. [4]The component labeled SiC at 283.9 eV corresponds to the carbon atoms of the SiC substrate.The other two components labeled S1 and S2 at 285.0 eV and 285.8 eV, respectively, originate from the buffer layer.Here, S1 can be attributed to carbon atoms which are bound in both the 6 3 and to the substrate while S2 belongs to carbon atoms, which are bound only in the buffer layer. [4]We now examine the corresponding Si2p core level of the 6 √ 3 sample.Two doublets with a branching ratio of 1/2 and a spin-orbit splitting of 0.6 eV are used to fit the spectrum.The component labelled SiC at 101.5 eV belongs to silicon atoms bound only to carbon atoms of the SiC.Furthermore, the second component Si C at 102.0 eV can be attributed (similar to S1 in the C1s) to silicon atoms of the topmost SiC bilayer which are covalently bonded to the 6 √ 3. Now that we understand the XPS results of a 6 3 on SiC, we will examine the core levels after the process with FeS 2 precursor.
The C1s core level of sample I1 (Figure 2a bottom) has changed significantly compared to the pristine 6 √ 3 sample.Only two components are needed to fit the data.The asymmetric peak labelled G at 284.1 eV can be attributed to graphene and indicates the decoupling of the 6 3 and its transformation into a quasifreestanding graphene layer.A comparison with the binding energy of neutral highly oriented pyrolytic graphite (HOPG) indicates a p-type doping of the sample. [36]The decoupling of the 6 √ 3 also leads to a component labeled SiC S at 283.3 eV, which belongs to the carbon atoms of the SiC substrate.The shift in binding energy compared to pristine 6 ) is a consequence of the modified interface and therefore altered surface band bending.Furthermore, no signals associated with the 6 3 are observed after the process, which can be interpreted as a complete intercalation of the sample.Similar behavior has already been reported for intercalation with hydrogen, [24,37] germanium, [30] or antimony. [33]he Si2p core level of sample I1 (Figure 2b bottom) also shows differences compared to the 6 √ 3 sample.Again, two components are needed to fit the data.The component labelled Si S at E b, 3/2 = 101.4eV belongs to the topmost silicon atoms of the SiC bilayer which are bound to the intercalated sulfur atoms.The second component with the highest intensity at E b, 3/2 = 101.0eV is attributed to the silicon atoms of the substrate below the intercalated regions in the SiC bulk.Here, the difference between the SiC substrate component of the pristine 6 3 sample and that of the intercalated areas is ΔE SiC = −0.5 eV, which is in good 3 sample before (top) and after the process with FeS 2 (bottom), respectively.Spectra are offset from each other for clarity.c) XP spectrum of the S2p core level of a buffer layer after the process with FeS 2 .
agreement with the results obtained from the C1s.As already seen in the C1s spectrum of sample I1, the Si2p spectrum shows no contribution from the 6 3 and supports the interpretation of a complete decoupling of the 6 √ 3. Figure 2c shows the corresponding S2p signal of sample I1.The spectrum can be described by a single doublet at E b, 3/2 = 162.2eV with a branching ratio of 1/2 and a spin-orbit splitting of 1.2 eV.We attribute this component to intercalated sulfur atoms bonded to the silicon atoms of the topmost SiC bilayer.Concluding, the XPS results of sample I1 demonstrate a complete intercalation of the sample with sulfur.
Another method of assessing a successful intercalation is LEED.Figure 3 shows the LEED patterns of a 6 3 sample shows the indicated diffraction spots of graphene (G) and the substrate (SiC), as well as the superstructure spots of the (6 [4,38] After the process with pyrite, we can see sharp spots associated with SiC and graphene. [38]However, compared to the pattern of the pristine 6 3 sample, the SiC spots are weaker in relation to the graphene spots.Moreover the superstructure spots are strongly suppressed, which indicates a decreased interaction between the carbon layer and the SiC substrate.This supports the interpretation of a successful intercalation in good agreement with XPS.Since there are no spots other than those already mentioned, and since the spots are sharp one could infer a 1 × 1 periodicity with respect to SiC(0001) of the intercalated sulfur atoms.On the other hand, it is possible that the sulfur atoms do not have a  long-range order.In this case, also no other spots are expected to appear.
The third evidence of successful intercalation is provided by ARPES. Figure 4 shows ARPES measurements in the vicinity of the K point of the graphene Brillouin zone of a sample after intercalation using FeS 2 precursor.The appearance of a delocalized -band around K is a further proof of the decoupling of a buffer layer and its transformation into a quasi-freestanding graphene layer. [24,28][41] The Dirac point is determined by a tight-binding fit at E D − E F = −0.19eV, indicating a p-type doping in good agreement with XPS, and the carrier concentration can be calculated to be p = 2.2 • 10 12 cm −2 .
The results of sample I1 presented up to here suggest a reliable and reproducible method for the complete decoupling of a 6 √ 3 on SiC by sulfur intercalation.However, as discussed below, not every intercalation experiment leads to the exact same results.As an example, Figure 5 depicts the C1s (a), Si2p (b) and S2p (c) core level spectra of sample I2 after the process with m FeS 2 = 53 mg pyrite, measured under 0°(top) and 60°(bottom) emission angles, respectively.More components are needed to fit the data of sample I2 when comparing the spectra (0°) with those of sample I1 (Figure 2).
In the C1s spectrum (Figure 5a 3.This has already been reported for other intercalation experiments. [33,42,43]The energy difference between the two substrate components ΔE SiC = −0.5 eV is the same as for sample I1, confirming the same origin of the component SiC S for both samples.Furthermore, the peak area ratio of the substrate components is I(SiC)/I(SiC S ) = 0.22.Since the substrate component of the pristine 6

√
sample is visible, its components S1 and S2 should also present, but since their intensity is very low, they are only minor contributions to the total spectrum.However, measurement under 60°emission angle (Figure 5a, bottom) clearly shows the signals S1 and S2 of the non-intercalated 6 3. At this emission angle, the relative intensities of atoms close to the surface increases compared to atoms in deeper layers.Therefore, the intensity of the graphene component is also increased compared to the substrate components.
Furthermore, the measurement under normal emission can be used to estimate the degree of intercalation D by peak area ratios with For sample I2 this means D = 84 %.It can be assumed that the reduced amount of precursor is the reason for the lower degree of decoupling.Interestingly, even a comparable amount of pyrite precursor as for sample I1 did not always lead to a complete decoupling of the 6 √ 3 surface (see Figure S3, Supporting Information).A possible explanation could be a different quality of the 6 3 surfaces and indicates that a high quality of the pristine 6

√
3 sample accompanied by a low defect density could be the limiting factor for the decoupling process.This points toward a defect-mediated intercalation process as it has already been observed for other intercalation experiments. [21,24,44]For successful intercalation with lead (Pb), even the defect density had to be increased by ion sputtering. [43]he Si2p spectrum (Figure 5b, top) is composed of three peaks.Two doublets have almost the same intensity and binding energy.The signal at E b = 101.5 eV labeled SiC was already present in the spectrum of the pristine 6 3 sample and corresponds to the substrate underneath non-intercalated regions of the sample.Furthermore, the second component Si S at E b = 101.3eV describes silicon atoms in the topmost SiC bilayer bound to intercalated sulfur atoms.Finally, the component with the highest intensity SiC S at E b = 101.0eV belongs to silicon atoms of the substrate deeper layers below intercalated areas.The energy difference between the bulk components ΔE SiC = −0.5 as well as their peak area I(SiC)/I(SiC S ) = 0.20 is comparable to the fit of the C1s spectrum.The measurement under 60°emission angle (Figure 5b, bottom) highlights the surface related components Si S and Si C (E b = 101.9eV) compared to the bulk related components.Here, the component Si C corresponding to the silicon atoms of the topmost SiC bilayer bonded to the non-intercalated 6

√
3 is required to fit the data.This component has already been used to fit the measurement under 0°, but is almost negligible due to its low intensity.The positions of the peaks corresponding to the intercalated regions are in good agreement with those of sample I1.
Finally, we want to discuss the XPS results of the S2p core level depicted in Figure 5c.In contrast to sample I1, two doublets are used to fit the measurement under 0°emission angle (top) for sample I2.The component labeled S Si at E b, 3/2 = 162.0eV belongs to the intercalated sulfur atoms and has already been seen in the spectrum of sample I1.The second component labeled S at E b, 3/2 = 164.0eV can be found at comparable binding energies as pure sulfur. [45,46]The question is whether the atoms corresponding to the component S are on the surface of the sample or under the QFMLG but not bound to the substrate.Therefore, we compare the measurement with the corresponding one under 60°emission angle (Figure 5c, bottom).It is obvious that the peak area ratio I(S)/I(S Si ) increases, indicating that the sulfur atoms corresponding to the component S are located above those corresponding to the components S Si .The following equation is used to determine the position of the sulfur atoms corresponding to the component S in the S2p spectrum relative to the carbon atoms of graphene in the C1s spectrum: The ratio for the component S is R S = 1.10, which means that the sulfur atoms of this component are located above the graphene.By analogy with Equation (2), the relative position of the sulfur atoms belonging to the component S Si is determined.Here, the ratio R S Si = 0.84 means that the corresponding sulfur atoms are located below the graphene.In conclusion, we can say that besides the intercalated sulfur atoms, there is also non-intercalated sulfur on the surface of the sample.This supports the statement that the partial decoupling of sample I2 is not due to the reduced amount of FeS 2 precursor, as sulfur is still present on the sample surface.Experiments with other samples have shown similar results to sample I2, indicating that partial intercalation occurs more frequently.Figure 6 shows the relative amount of non-intercalated sulfur on the surface as a function of the degree of intercalation D for different sulfur intercalation experiments.Obviously, the less the sample is intercalated, the more non-intercalated sulfur can be found on the sample surface.In addition, each partially intercalated sample has non-intercalated sulfur on the surface.This suggests, that the non-intercalated sulfur atoms are preferentially located on top of the 6 √ 3 surface.The question arises whether subsequent annealing can remove the remaining surface sulfur and further improve the degree of intercalation.For this purpose, sample I2 was annealed in UHV in ΔT = 50 K steps in the temperature range of T = 300 °C … 900 °C for 30 min, respectively.This will be discussed in the following.
Figure 7 shows the C1s (a) and S2p (b) core levels of sample I2 after each UHV annealing step.Even a rough overview of the spectra shows a shift of the peaks in the C1s and S2p data.A more precise interpretation of the measurement is possible after fitting each core level.
The binding energies of the components S Si in the corresponding S2p spectrum and SiC S in the corresponding C1s spectrum are plotted as a function of the annealing temperature in Figure 8a.Here, the shape of the curves is similar for both signals.The binding energies remain constant to 400 °C and shift to higher E b between 400 and 800 °C.This striking agreement in the temperature dependence indicates a structural rearrangement of the intercalated sulfur layer accompanied by a change in the surface band bending.Similar behavior was observed for a bismuth intercalated QFMLG [31] and a ytterbium intercalated QFMLG [34] annealed in UHV.Further increase of the annealing temperature reduces the binding energies back to the initial values.This suggests a similar arrangement of the intercalated atoms immediately after the intercalation, as well as after an annealing at 900 °C.Structural analysis such as scanning tunneling microscopy (STM) could help to provide more information.
Interesting insights into the intercalation process can be gained from Figure 8b, which shows the temperature dependence of the degree of intercalation D and the relative amount of surface sulfur A S .Up to 700 °C no change is observed for A S .However, increasing of the annealing temperature further reduces the amount of non-intercalated sulfur on the surface, which is completely absent after an annealing at 900 °C.According to literature, elemental sulfur vaporizes at 444.6 °C at atmospheric pressure [47] and this temperature should be even lower for UHV conditions.A stronger bonding of the sulfur atoms to the 6 √ 3 surface must be assumed, as higher temperatures are needed to destroy it.This is also supported by the binding energy of the component S, which is comparable to that of pure sulfur. [45,46]However, the electronegativity of sulfur (2.5 [47] ) and carbon (2.5 [47] ) is the same and a binding of these atomic species does not have much influence on the peak position in XPS.A closer look at the degree of intercalation D is helpful to find out, where the surface sulfur goes to.The amount of decoupled graphene does not change up to 700 °C, which was already observed for A S .In contrast, at higher annealing temperatures an increase of D is observed, which is the exact opposite of the behavior of A S .After heating to 900 °C, the sample is completely intercalated, with no non-intercalated sulfur left on the sample surface.This indicates that the non-intercalated surface sulfur is a transition state prior to the intercalation and that the sulfur intercalation investigated here can be considered as a two-step process, where sulfur atoms are bound to the 6 √ 3 surface in a first step and move to the 6 √ 3/SiC interface to decouple it in a second step.The occurrence of non-intercalated sulfur only on partially intercalated samples (see Figure 6) also supports this interpretation.

Summary and Conclusion
We have demonstrated the intercalation of a 6 √ 3 on SiC by annealing the sample in a two-zone tube furnace with a pyrite precursor as sulfur source.Using spectroscopic and diffraction techniques, we have clearly presented the decoupling of the buffer layer and its transformation into quasi-freestanding graphene.A fully decoupled graphene layer is p-type doped with a Dirac point at E D − E F = −0.19eV and a carrier concentration of p = 2.2 • 10 12 cm −2 .
In some cases only a partial intercalation is achieved, with nonintercalated sulfur remaining on top of the 6 √ 3 surfaces.Here, the amount of non-intercalated sulfur on the sample surface is greater the lower the degree of intercalation.However, we have shown that the non-intercalated sulfur is bound to the remaining 6 √ 3 surface and constitutes a transition state.Further annealing in UHV reduces the sulfur on the surface and at the same time increases the degree of intercalation.A schematic representation of this is shown in Figure 9.We have also demonstrated the thermal stability of the intercalated species up to 900 °C, while structural rearrangements may still occur in the temperature range between 400 and 800 °C.Further structural analysis is needed to understand this behavior.For future experiments on possible TMDC intercalation, sulfur-intercalated graphene samples can be used as a starting point.

Experimental Section
Sample Preparation: The SiC samples used in this study were cut from n-type 6H-SiC(0001) wafers purchased from Pam-Xiamen.Prior to the growth of the buffer layer, the SiC samples underwent a wet-chemical cleaning procedure in order to remove organic as well as metallic contaminants from the surface.A polymer-assisted sublimation growth process was employed to obtain the 6 √ 3 samples. [8]For this purpose, an ultrathin polymer layer (Merck AZ 5214E) was spin-coated on the substrate surface prior to the growth of the buffer layer.The latter was carried out in argon atmosphere in two annealing steps at 1200 and 1425 °C for 10 and 5 min, respectively.
The sulfur intercalation process was performed in a two-zone furnace, where each zone could be heated separately.The setup is shown schematically in Figure 10.Precursor and sample were positioned in the same  quartz tube but within two separately heatable zones.An argon gas flow of 0.05 standard liter per minute (slm) transports the gaseous sulfur from the precursor to the sample while the pressure was around 6 mbar.As sulfur source (precursor) pyrite (FeS 2 ) powder purchased from Sigma-Aldrich or Alfa Aesar was used.For different experiments and sample preparations a varying amount of m FeS 2 = 50 mg … 270 mg FeS 2 was placed in a quartz boat in the first heating zone.Annealing of pyrite at a temperature of T FeS 2 = 400 °C results in a thermal decomposition [48] according to the equation FeS 2 ←→ FeS + S ( 4 ) While FeS remains as solid material, S enters into gas phase.The sample was positioned in a second quartz boat in heating zone 2. During the process, first the temperature of zone 2 was increased to T 6 √ 3 = 850 • C. When the desired temperature of zone 2 was reached, the temperature of zone 1 was raised.The precursor was positioned in a temperature region of T FeS 2 =400 °C± 35 °C.The process took t = 1 h once the pyrite was at the desired temperature.
Prior to all measurements, the samples were annealed in ultrahigh vacuum (UHV) at a temperature of T = 310 °C.
Initially, an unsuccessful intercalation attempt was made using elemental sulfur powder as a precursor.The results of these experiments are discussed in the Supporting Information.
XPS: The samples were characterized by X-ray photoelectron spectroscopy (XPS) with Al K  radiation from a Specs XR 50 M X-ray source monchromatized by a Specs Focus 500 monochromator and analyzed with a Specs Phoibos 150 MCD-9.The angle of incidence of the X-rays on the sample, in normal emission geometry, was about 50°with this setup.The chamber pressure was better than 3 × 10 −10 mbar during the measurement.
ARPES: Angle-resolved photoelectron spectroscopy (ARPES) was conducted using a Specs UVS 300 He-lamp with a Specs TMM304 monochromator and a Specs Phoibos 150 analyzer with a 2D-CCDdetector.HeI radiation (h = 21.22 eV) was used for the ARPES measurements.During the measurement, the pressure in the chamber was better than 9 × 10 −10 mbar, with residual gas being He from the UV lamp.
LEED: The surface crystal structure had been investigated by means of low energy electron diffraction (LEED) using an ErLEED 150 system (SPECS Surface Nano Analysis GmbH, Berlin, Germany).The chamber pressure was better than 5 × 10 −10 mbar during the measurement.

Figure 2 .
Figure 2. XP spectra of the C1s (a) and the Si2p (b) core levels.Depicted is the measurement of a 6 √3 sample before (top) and after the process with FeS 2 (bottom), respectively.Spectra are offset from each other for clarity.c) XP spectrum of the S2p core level of a buffer layer after the process with FeS 2 .

√ 3
sample before (a) and after (b) the process with FeS 2 , taken at an energy of E = 95 eV.The pristine 6 √

Figure 3 .
Figure 3. LEED images of (a) a 6 √ 3 sample and (b) a sample after the process with FeS 2 precursor at 95 eV.

Figure 4 .
Figure 4. ARPES measurement in the vicinity of the K-point of a sulfur intercalated sample.The delocalized electrons of the -orbital of the QFMLG are clearly visible.
, top) three components stand out.The components labelled G (284.3 eV) and SiC S (283.3 eV) have already been seen in the spectrum of sample I1 and belong to the quasi-freestanding graphene and the SiC substrate below intercalated regions.Finally, the third component labeled SiC at E b = 283.8eV corresponds to the substrate of the original 6 √ 3 surface, indicating residual non-intercalated regions of the sam-ple and only a partial decoupling of the 6 √

Figure 5 .
Figure 5. XP spectra of the C1s (a), the Si2p (b) and the S2p (c) core levels.Depicted is a sample after the process with FeS 2 under normal emission (top) and under 60°emission angle (bottom), respectively.Spectra are offset from each other for clarity.

Figure 6 .
Figure 6.Relative amount of non-intercalated surface sulfur A S (Equation (3)) as a function of the degree of intercalation D (Equation (1)) for different samples after the intercalation process.

Figure 7 .
Figure 7. Annealing of sample I2 in UHV after intercalation with sulfur.XPS data of the C1s (a) and S2p (b) core levels after every annealing step.The sample was annealed from 300 °C (bottom) to 900 °C (top) in 50 K steps for 30 min, respectively.

Figure 8 .
Figure 8. Fitting results of the C1s and Si2p core levels of sample I2 after annealing in UHV.a) Binding energy of the components SiC S of the C1s and S Si of the S2p after the corresponding annealing step.b) Degree of intercalation D (Equation (1)) and relative amount of non-intercalated surface sulfur A S (Equation (3)) after the corresponding annealing step.

Figure 9 .
Figure 9. Schematic representation of a sample after the sulfur intercalation process.Starting with a 6 √3 sample (left), after the process with pyrite (middle) the sample is partially intercalated and the component S Si belonging to the intercalated sulfur appears in the S2p core level spectrum.Furthermore, non-intercalated sulfur atoms S are bound to the 6 √ 3 surface.After subsequent annealing in UHV (right) the non-intercalated sulfur atoms have moved to the interface and a completely decoupled QFMLG has formed.

Figure 10 .
Figure 10.Schematic illustration of the furnace used for sulfur intercalation.