Volume 33 Issue 4
Sep.  2020
Turn off MathJax
Article Contents

Xiu-neng Song, Huan-yu Ji, Juan Lin, Ruo-yu Wang, Yong Ma, Chuan-kui Wang. Geometric and Electronic Structures of Pyrazine Molecule Chemisorbed on Si(100) Surface by XPS and NEXAFS Spectroscopy[J]. Chinese Journal of Chemical Physics , 2020, 33(4): 417-426. doi: 10.1063/1674-0068/cjcp1910180
Citation: Xiu-neng Song, Huan-yu Ji, Juan Lin, Ruo-yu Wang, Yong Ma, Chuan-kui Wang. Geometric and Electronic Structures of Pyrazine Molecule Chemisorbed on Si(100) Surface by XPS and NEXAFS Spectroscopy[J]. Chinese Journal of Chemical Physics , 2020, 33(4): 417-426. doi: 10.1063/1674-0068/cjcp1910180

Geometric and Electronic Structures of Pyrazine Molecule Chemisorbed on Si(100) Surface by XPS and NEXAFS Spectroscopy

doi: 10.1063/1674-0068/cjcp1910180
More Information
  • Corresponding author: Yong Ma, E-mail: mayong@sdnu.edu.cn
  • Received Date: 2019-10-19
  • Accepted Date: 2020-01-19
  • Publish Date: 2020-08-27
  • The geometric and electronic structures of several possible adsorption configurations of the pyrazine ({C$ _{4} $}{H$ _{4} $}{N$ _{2} $}) molecule covalently attached to Si(100) surface, which is of vital importance in fabricating functional nano-devices, have been investigated using X-ray spectroscopies. The Carbon K-shell (1s) X-ray photoelectron spectroscopy (XPS) and near-edge X-ray absorption fine structure (NEXAFS) spectroscopy of predicted adsorbed structures have been simulated by density functional theory with cluster model calculations. Both XPS and NEXAFS spectra demonstrate the structural dependence on different adsorption configurations. In contrast to the XPS spectra, it is found that the NEXAFS spectra exhibiting conspicuous dependence on the structures of all the studied pyrazine/Si(100) systems can be well utilized for structural identification. In addition, according to the classification of carbon atoms, the spectral components of carbon atoms in different chemical environments have been investigated in the NEXAFS spectra as well.
  • 加载中
  • [1] M. Schöning and H. Lüth, Phys. Status Solidi A 185, 65 (2001). doi:  10.1002/1521-396X(200105)185:1<65::AID-PSSA65>3.0.CO;2-Y
    [2] Y. Wen, W. Yi, L. Meng, M. Feng, G. Jiang, W. Yuan, Y. Zhang, H. Gao, L. Jiang, and Y. Song, J. Phys. Chem. B 109, 14465 (2005). doi:  10.1021/jp044256t
    [3] Z. Liu, A. Yasseri, J. Lindsey, and D. Bocian, Science 302, 1543 (2003). doi:  10.1126/science.1090677
    [4] T. He, J. He, M. Lu, B. Chen, H. Pang, W. Reus, W. Nolte, D. Nackashi, P. Franzon, and J. Tour, J. Am. Chem. Soc. 128, 14537 (2006). doi:  10.1021/ja063571l
    [5] T. Strother, W. Cai, X. Zhao, R. Hamers, and L. Smith, J. Am. Chem. Soc. 122, 1205 (2000). doi:  10.1021/ja9936161
    [6] Z. Lin, T. Strother, W. Cai, X. Cao, L. Smith, and R. Hamers, Langmuir 18, 788 (2002). doi:  10.1021/la010892w
    [7] W. Ng, J. Liu, and Z. Liu, Phys. Chem. Chem. Phys. 17, 16876 (2015). doi:  10.1039/C5CP01742D
    [8] F. Tao, S. Bernasek, and G. Xu, Chem. Rev. 109, 3991 (2009). doi:  10.1021/cr8003532
    [9] F. Billes, H. Mikosch, and S. Holly, J. Mol. Struct. 423, 225 (1998). doi:  10.1016/S0166-1280(97)00143-7
    [10] T. Omiya, H. Yokohara, and M. Shimomura, J. Phys. Chem. C 116, 9980 (2012). doi:  10.1021/jp300101t
    [11] W. Ng, J. Liu, and Z. Liu, J. Phys. Chem. C 117, 26644 (2013). doi:  10.1021/jp409041s
    [12] X. Lu, X. Xu, J. Wu, N. Wang, and Q. Zhang, New J. Chem. 26, 160 (2002). doi:  10.1039/b105774j
    [13] H. Huang, J. Huang, Y. Ning, and G. Xu, J. Chem. Phys. 121, 4820 (2004). doi:  10.1063/1.1781117
    [14] M. Shimomura, D. Ichikawa, Y. Fukuda, T. Abukawa, T. Aoyama, and S. Kono, Phys. Rev. B 72, 033303 (2005). doi:  10.1103/PhysRevB.72.033303
    [15] H. Lee, J. Park, I. Kim, H. Kim, B. Park, H. Shin, I. Lee, A. Singh, A. Thakur, and J. Kim, J. Phys. Chem. C 116, 722 (2011).
    [16] W. Ng, S. Sun, J. Liu, and Z. Liu, J. Phys. Chem. C 117, 15749 (2013). doi:  10.1021/jp4048665
    [17] C. Wang, H. Ge, Y. Zhao, S. Liu, Y. Zou, and W. Zhang, J. Magn. Magn. Mater. 423, 421 (2017). doi:  10.1016/j.jmmm.2016.09.128
    [18] N. Shirota, S. Yagi, M. Taniguchi, and E. Hashimoto, J. Vac. Sci. Technol. A 18, 2578 (2000). doi:  10.1116/1.1286392
    [19] X. Song, G. Wang, Y. Ma, S. Jiang, W. Yue, S. Xu, and C. Wang, Chem. Phys. Lett. 645, 164 (2016). doi:  10.1016/j.cplett.2015.12.005
    [20] X. Song, J. Hu, J. Lin, S. Wang, J. Zhang, S. Yang, Y. Ma, Y. Zhou, and C. Wang, Mol. Phys. 117, 794 (2019). doi:  10.1080/00268976.2018.1542167
    [21] S. Wang, J. Hu, J. Zhang, J. Lin, X. Song, and Y. Ma, Mol. Phys. 117, 507 (2019). doi:  10.1080/00268976.2018.1524937
    [22] Y. Ma, J. Zhang, S. Wang, J. Hu, J. Lin, and X. Song, Mol. Phys. 117, 635 (2019). doi:  10.1080/00268976.2018.1532540
    [23] X. Song, W. Hua, Y. Ma, C. Wang, and Y. Luo, J. Phys. Chem. C 116, 23938 (2012). doi:  10.1021/jp307834x
    [24] K. Kim, Y. Han, J. Zhu, J. Baik, H. Shin, H. Lee, and B. Kim, Curr. Appl. Phys. 16, 1120 (2016). doi:  10.1016/j.cap.2016.06.014
    [25] K. O'Donnell, O. Warschkow, A. Suleman, A. Fahy, L. Thomsen, and S. Schofield, J. Phys. Condens. Matter 27, 054002 (2014).
    [26] M. Frisch, G. Trucks, H. Schlegel, G. Scuseria, M. Robb, J. Cheeseman, G. Scalmani, V. Barone, B. Mennucci, G. Petersson, H. Nakatsuji, M. Caricato, X. Li, H. Hratchian, A. Izmaylov, J. Bloino, G. Zheng, J. Sonnenberg, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai, T. Vreven, J. Montgomery, Jr., J. Peralta, F. Ogliaro, M. Bearpark, J. Heyd, E. Brothers, K. Kudin, V. Staroverov, R. Kobayashi, J. Normand, K. Raghavachari, A. Rendell, J. Burant, S. Iyengar, J. Tomasi, M. Cossi, N. Rega, J. Millam, M. Klene, J. Knox, J. Cross, V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts, R. Stratmann, O. Yazyev, A. Austin, R. Cammi, C. Pomelli, J. Ochterski, R. Martin, K. Morokuma, V. Zakrzewski, G. Voth, P. Salvador, J. Dannenberg, S. Dapprich, A. Daniels, O. Farkas, J. B. Foresman, J. Ortiz, J. Cioslowski, and D. Fox, Gaussian 09 Revision A.02, Wallingford CT: Gaussian Inc., (2009).
    [27] K. Hermann, L. Pettersson, M. Casida, C. Daul, A. Goursot, A. Koester, E. Proynov, A. St-Amant, D. Salahub, V. Carravetta, H. Duarte, C. Friedrich, N. Godbout, J. Guan, C. Jamorski, M. Leboeuf, M. Leetmaa, M. Nyberg, S. Patchkovskii, L. Pedocchi, F. Sim, L. Triguero, and A. Vela, Stobe-demon Version 3.0, Stockholm, Sweden: StoBe Software, (2007).
    [28] A. Becke, Phys. Rev. A 38, 3098 (1988).
    [29] J. P. Perdew, Phys. Rev. B 33, 8822 (1986).
    [30] X. Song, Y. Ma, C. Wang, M. Paul, E. Wolfgang, Unger, and Y. Luo, J. Phys. Chem. C 116, 12649 (2012). doi:  10.1021/jp302716w
    [31] M. Nyberg, Y. Luo, L. Triguero, L. Pettersson, and H. Ǻgren, Phys. Rev. B 60, 7956 (1999).
    [32] W. Kutzelnigg, U. Fleischer, and M. Schindler, NMR Basic Principles and Progress, 3rd Edn., Heidelberg, Berlin, New York: Springer Verlag, (1990).
    [33] L. Triguero, L. Pettersson, and H. Ǻgren, Phys. Rev. B 58, 8097 (1998).
    [34] P. Bagus, Phys. Rev. 139, A619 (1965).
    [35] L. Triguero, O. Plashkevych, L. Pettersson, and H. Ǻgren, J. Electron. Spectrosc. 104, 195 (1999).
    [36] U. von Barth and G. Grossmann, Phys. Rev. B 25, 5150 (1982).
    [37] T. Privalov, F. Gel'mukhanov, and H. Ǻgren, Phys. Rev. B 64, 165115 (2001).
    [38] X. Song, J. Hu, S. Wang, Y. Ma, Y. Zhou, and C. Wang, Phys. Chem. Chem. Phys. 19, 32647 (2017). doi:  10.1039/C7CP06543D
  • 加载中
通讯作者: 陈斌, bchen63@163.com
  • 1. 

    沈阳化工大学材料科学与工程学院 沈阳 110142

  1. 本站搜索
  2. 百度学术搜索
  3. 万方数据库搜索
  4. CNKI搜索

Figures(11)  / Tables(5)

Article Metrics

Article views(6) PDF downloads(2) Cited by()

Proportional views
Related

Geometric and Electronic Structures of Pyrazine Molecule Chemisorbed on Si(100) Surface by XPS and NEXAFS Spectroscopy

doi: 10.1063/1674-0068/cjcp1910180

Abstract: The geometric and electronic structures of several possible adsorption configurations of the pyrazine ({C$ _{4} $}{H$ _{4} $}{N$ _{2} $}) molecule covalently attached to Si(100) surface, which is of vital importance in fabricating functional nano-devices, have been investigated using X-ray spectroscopies. The Carbon K-shell (1s) X-ray photoelectron spectroscopy (XPS) and near-edge X-ray absorption fine structure (NEXAFS) spectroscopy of predicted adsorbed structures have been simulated by density functional theory with cluster model calculations. Both XPS and NEXAFS spectra demonstrate the structural dependence on different adsorption configurations. In contrast to the XPS spectra, it is found that the NEXAFS spectra exhibiting conspicuous dependence on the structures of all the studied pyrazine/Si(100) systems can be well utilized for structural identification. In addition, according to the classification of carbon atoms, the spectral components of carbon atoms in different chemical environments have been investigated in the NEXAFS spectra as well.

Xiu-neng Song, Huan-yu Ji, Juan Lin, Ruo-yu Wang, Yong Ma, Chuan-kui Wang. Geometric and Electronic Structures of Pyrazine Molecule Chemisorbed on Si(100) Surface by XPS and NEXAFS Spectroscopy[J]. Chinese Journal of Chemical Physics , 2020, 33(4): 417-426. doi: 10.1063/1674-0068/cjcp1910180
Citation: Xiu-neng Song, Huan-yu Ji, Juan Lin, Ruo-yu Wang, Yong Ma, Chuan-kui Wang. Geometric and Electronic Structures of Pyrazine Molecule Chemisorbed on Si(100) Surface by XPS and NEXAFS Spectroscopy[J]. Chinese Journal of Chemical Physics , 2020, 33(4): 417-426. doi: 10.1063/1674-0068/cjcp1910180
  • The covalent attachment of organic materials to silicon surfaces, as a promising approach linking biomolecules with microelectronics [1], has been one of the most attractive subjects in surface science. In the past few decades, silicon has played an important role in the field of microelectronics because of its excellent performance, such as high strength, small thermal expansion coefficient and good electrical conductivity. However, with the rapid development of energy and information technology, high requirements for the optical and electrical properties of traditional semiconductor materials have been put forward. Silicon surface modified by organic molecule with abundant adjustable properties (include flexibility, chemical affinity and conductivity) will provide more new functionalities to the traditional silicon materials and fabricate intelligentized and functionalized surfaces. Adsorption of organic molecules to the silicon surface has advantages that are unmatched by other surfaces, for example, a single crystal plane having an atomic level flatness is more easily obtained than the metal and the conduction state can be adjusted by the degree of doping. In previous studies, they have shown that the covalent binding of the organic molecules on the silicon surface, combining biological, physical and chemical properties, plays a key role in designing novel molecular electronic devices, such as molecular switches [2], molecular memory [3], and field effect transistor [4]. In addition, the technique for the preparation of self-assembled monolayers (SAMs) using semiconductor silicon material as a substrate is a promising organic functional film preparation technology. Strother et al. firstly assembled olefinic acid onto a silicon surface by covalent bonding to form a monolayer film, and then attached DNA molecules to its surface by electrostatic interaction [5, 6]. This research is conducive to the development of new biochips and biosensors. In the light of the peculiar photoelectric properties and broad application prospects of the functional interfaces between organic materials and silicon, many efforts have been made to the theoretical research of chemical adsorption on the silicon surface.

    Aromatic molecule is an important organic group in organic materials, and the adsorption of aromatic compounds on the silicon surface through covalent bonds is a typical reaction between $ \pi $-conjugated molecules and the Si dimer. To date, a variety of these types of adsorbed structures have been produced in both theoretical and experimental studies according to the sides, polarity, substituent groups and the types of heteroatoms in the aromatic ring [7, 8]. In the aromatic family, pyrazine ({C$ _{4} $}{H$ _{4} $}{N$ _{2} $}), lightly modified by inserting two para-nitrogen atoms into the six-membered aromatic ring, is isoelectronic to benzene and pyridine, and may be an ideal building block for various synthetic compounds and new functional materials [8, 9]. The well-ordered organic one-dimensional (1D) lines achieved in the chemisorption of pyrazine even on the reactive Si(100) [10] holds the potential for fabricating functional nanodevices. This is an interesting observation in recent study and has successfully attracted our attention to the chemisorbed structures of pyrazine on the silicon surface. The presence of three double bonds and two nitrogen atoms in the pyrazine ring lead to an increase in the number of molecular adsorption configurations. According to the report in Ref.[11], we have presented several feasible binding modes, which are most likely to exist in the chemisorption experiments of pyrazine on the silicon surface, are shown in FIG. 1 (Mode Ⅰ-Ⅹ). Many experimental and theoretical investigations so far have been performed mainly on the three relatively stable cluster modes, Mode Ⅰ, Ⅱ and Ⅵ. The theoretical prediction of binding energy on a Si$ _{9} $H$ _{12} $ cluster mode suggested that the N-end-on adsorbed pyrazine would be the primary species at low temperature, whereas the side-on adsorbed pyrazine (Mode Ⅱ in FIG. 1), which is bonded to the surface dimer through two carbon atoms at opposite site, can be regarded as the main adsorbed configuration at the elevated temperatures [12]. Subsequently, by the analysis of the high-resolution electron energy loss spectroscopy (HREELS) and X-ray photoelectron spectroscopy (XPS), the researchers presented another different adsorption model (Mode Ⅰ in FIG. 1) where two $ para $-nitrogen atoms directly bonded to the silicon dangling bonds to form a kind of 1, 4-N, N-dihydropyrazine-like adduct at 300 K [13]. Recently, by using scanning tunneling microscopy (STM), photoelectron diffraction (PED) and density functional theory (DFT), one definite 1D molecular chain arranged along the direction of the dimer bond was proposed (Mode Ⅵ from FIG. 1), in which pyrazine molecule is located between two dimer rows [14]. To the best of our knowledge, the selectivity, configuration and mechanism of the cross-row configuration have been well discussed in previous studies [15, 16]. The structural discrepancies of several proposed configurations will inevitably result in differences in the nature of the adsorption modes, so it is of important significance to characterize different configurations. The main object of the present work is to characterize ten possible adsorption configurations of pyrazine on the Si(100) (shown in FIG. 1).

    Figure 1.  Optimized configurations for a pyrazine molecule chemically adsorbed on Si(100).

    A series of spectroscopic techniques, such as Raman spectroscopy, infrared(IR) spectroscopy and X-ray spectroscopy [17, 18, 19] are applied to research the surface of organic functional semiconductors. X-ray spectroscopy corresponding to the excitation or de-excitation process of nuclear orbital electrons of the molecular system is efficient in determining the elemental composition of the material and the chemical state of the elements contained [20, 21, 22], so it is a useful means to probe the interfacial structure between pyrazine molecule and the silicon surface. Previous studies have demonstrated that the NEXAFS spectra are commonly used to probe electronic structures, which can provide information about the geometric orientation and chemical bonding of adsorbate [23, 24, 25], and the XPS spectra related to the ionization of nuclear electrons are advisable to detect the composition and chemical state of the involved element. In this context, the C K-edge NEXAFS and XPS spectra, both of which are sensitive to the local chemical environment of adsorbed structure, were performed in first-principles simulations with cluster representations. The relationship between the model structure and the theoretical simulated spectra can serve to pinpoint or quantify the actual surface species in the experiment and is of vital importance for better understanding of the experimental spectra.

  • In calculation strategies, we selected a fragment of the Si(100) surface containing single or double dimers as a substrate, and considered ten possible bonding modes in terms of pyrazine molecules attached to Si(100) surface. For different adsorption structures, the appropriate cluster modes adopted include 17, 29 and 31 silicon atoms respectively, while silicon atoms at the boundary of heterocyclic molecules were saturated by hydrogen atoms. All geometries of predicted cluster modes were optimized using B3LYP functional and standard 6-31G(d, p) basis set in the Gaussian09 [26] package. Then in the light of the optimized structures, we calculated the XPS and NEXAFS spectra for all the carbon atoms implemented by StoBe program [27] with DFT method. For accurate calculations, the gradient corrected Becke (BE88) exchange functional [28] and the Perdew (PD86) correlation functional [29] were taken into account to furnish precise oscillator strength consistent with the experimental results [30, 31]. In terms of the basis set, IGLO-Ⅲ (localized orbital) [32] was utilized for the description of core-excited carbon atoms, whereas DZVP (double-$ \zeta $ plus valence polarization) was set for silicon atoms and TZVP (triple-$ \zeta $ plus valence polarization) for the others. To facilitate self-consistent field (SCF) convergence, miscellaneous auxiliary basis sets and model core potentials were also applied to corresponding atoms. Furthermore, there was a augmented diffuse basis set (19s, 19p, 19d) [33] added to the transition dipole moment and the transition energy calculation of the nuclear excited state.

    In practice, taking full account of relaxation, XPS spectra were obtained by broadening the C1s ionization potentials (IP), which was the energy difference between the core-ionized state, i.e. full core hole (FCH) state and the ground state (GS) in the $ \Delta $Kohn-Sham ($ \Delta $KS) scheme [34, 35]. With regards to the NEXAFS, including the initial state and the final state, the exact adsorption spectrum of finite-sized molecule systems could be obtained just from the final-state wave function, that is, the final-state rule [36, 37]. In the calculative process of the X-ray spectrum, we chose FCH approximation regarding the core-ionized state as the reference state that could show the characteristics of all transitions. The averaged absorption oscillator strength was estimated over the $ x $, $ y $, and $ z $ directions for simulating the magic incident angle during the experiment:

    Here, $ \psi_{\rm{i}} $ and $ \psi_ \rm{f}$ involve two molecular orbits, i.e., the initial state of i, the final state of f in the X-ray adsorption and $ \varepsilon_{ \rm{if}} $ represents the energy-level interval. Next, in order to obtain the absolute energy positions of each peak, the exact transition energy from 1s to the lowest unoccupied molecular orbital (LUMO) employed for the calibration of the whole adsorption spectra was calculated by the $ \Delta $KS scheme:

    Where the value of transition energy from 1s to the LUMO getting in this way is assumed as the energy difference between the GS and the fully optimized core-excited state. Then, the relativistic effect associated with the introduced core-hole was treated with a small shift value of +0.2 eV to the computed IP's and transition energy. Ultimately, the XPS and NEXAFS spectra were generated from line broadening for some thermodynamic and optical reasons. In terms of XPS, the obtained IP's was convoluted by a Gaussian function with full width at half maximum (FWHM) of 0.2 eV. For the NEXAFS, taking full account of the vibronic interaction, a convolution with FWHM of 0.3 eV was used by a Gaussian function in the region below the IP, while it was carried out by a Stieltjes imaging approach in the continuous part over the IP. Although the actual observed peak width is usually higher than 1 eV, we have chosen a finer FWHM value for a better distinction of the adsorption configurations. Our computational theories and methods have already been confirmed to be credible and accurate [38].

  • The chemisorbed products of pyrazine molecules covalently bonded to Si(100) through the Si-C or Si-N bonds are depicted in FIG. 1 (Mode Ⅰ-Ⅹ). Considering diverse conjunctions of pyrazine molecule with the excellent electron acceptor silicon atoms, we made a rough division of the predicted structures corresponding to the number of chemical bonds formed between the pyrazine and the silicon surface. The first type of adsorption structure is to combine the pyrazine molecule with the Si(100) through two bonds. There are three ways to form this type in these possible modes: in the so-called "butterfly" structure, the organic molecules and the silicon surface are held together by two equivalent Si = C bonds or Si = N bonds on the same dimer. Another configuration named "tilted", is constructed by the connecting two adjacent atoms of the organic ring to the silicon surface. However, different from the above two structures, the "Cross-Row Bridge" geometries with two binding bonds are formed through two dimer rows. The second type of adsorption structure is established by adsorption of pyrazine molecules on two silicon dimers along the same row, which is labeled as "Tight-Bridge" structure. For convenient and careful comparison, the former three species (Mode Ⅰ-Ⅵ) possessing two bonding linkages are discussed respectively from the four-bond saturated "Tight-Bridge" structure (Mode Ⅶ-Ⅹ). In spectral studies, the bond lengths, buckling and adsorption heights of the ten optimized geometries were also investigated. The bond lengths of C-C, C-N and N-Si, C-Si for the adsorption of one pyrazine on silicon surface are shown in Table Ⅰ. Moreover, the dihedral angles and adsorption heights of the pyrazine molecules are listed in Table Ⅱ. As can be seen from Table Ⅰ, Mode Ⅰ is bonded to the surface by two N-Si bonds at a distance of 1.82 Ǻ. However, the distance of the N-Si bond in Mode Ⅳ where only one nitrogen atom is bonded to the silicon surface is shortened to 1.78 Ǻ. Mode Ⅸ and Ⅹ with the same orientation of pyrazine molecules have similar Si-Si bond lengths compared to Mode Ⅶ and Ⅷ. The dihedral angles of the pyrazine molecules reflect the degree of distortion of the pyrazine molecular backbone caused by the chemisorption behavior. The near planar geometries of pyrazine in Mode Ⅲ and Ⅴ conclude that the chemisorption behavior of these two forms induces little disturbance to the aromatic ring. The pyrazine molecules in Mode Ⅲ and Ⅴ are near to the planar structures but their adsorption heights are different.

    Table Ⅰ.  C-C, C-N and N-Si, C-Si bond lengths of the ten adsorption configurations. Redundant results for equivalent bond are omitted. The indices of C, N and Si atoms are shown in FIG. 1. C and N atoms in C-Si, N-Si are atoms bonded to silicon atoms on the surface of Si(100).

    Table Ⅱ.  Dihedral angle of the non-planar structure of pyrazine and adsorption heights of the ten adsorption configurations.

    In order to compare the stability of the several chemically adsorbed configurations, we also calculated the adsorption energy (binding energy) of pyrazine adsorption on the Si(100) surface, which is defined as:

    Where $ E_{ \rm{C}_{4} \rm{H}_{4} \rm{N}_{2}} $, $ E_{ \rm{Si}} $, and $ E_{ \rm{C}_{4} \rm{H}_{4} \rm{N}_{2}/ \rm{Si}} $ represented the total energies of an isolated pyrazine molecule, corresponding silicon surface, and the chemisorbed system, respectively. The evaluation of adsorption energy is the difference between the total energy of the corresponding system prior to the adsorption of organism onto the silicon surface ($ E_{ \rm{C}_{4} \rm{H}_{4} \rm{N}_{2}} $+$ E_{ \rm{Si}} $) and the energy of the system after adsorption ($ E_{ \rm{C}_{4} \rm{H}_{4} \rm{N}_{2}/ \rm{Si}} $). The calculated results of adsorption energies for all adsorption configurations are summarized in Table Ⅲ. We found in our binding energy computations that the Mode Ⅳ occurring exclusively through a C-Si and a N-Si bonding is the most stable configuration in both "Butterfly" and "Tilted" types. It is also worth noting that the Mode Ⅴ of the Cross-Row structure with the adsorption energy -37.65 kcal/mol is not much stable compared to Mode Ⅱ, while the binding energy is about 12.5 kcal/mol lower than that of another "Cross-Row Bridge" structure, i.e. Mode Ⅵ. For the last four "Tight-Bridge" structures, pyrazine molecule attached on surface accompany with a 90$ ^\circ $ rotation in Mode Ⅶ and Mode Ⅸ, their binding energies are surprisingly equal. However, the adsorption energies in Mode Ⅷ and Mode Ⅹ are quite different.

    Table Ⅲ.  The number of non-equivalent carbon atoms $ C_{ \rm{non}} $, as well as binding energy of pyrazine.

  • The XPS for the detection of occupied states in the core region has become one of the most widely used spectroscopic techniques in surface analysis. To identify all the predicted structures above, the C1s excited XPS spectra for any angle incidence of pyrazine molecule chemisorbed on Si(100) are displayed in FIG. 2 and FIG. 3. In order to obtain a detailed and careful comparison of the ten calculated XPS spectra, the energy values of the main spectral features marked in the spectra are listed in Table Ⅳ. For the ten studied binding modes, the presences of four spectral features a, b, c, and d only in the spectrum of Mode Ⅳ play a key role in the spectral identification, hence Mode Ⅳ can be firstly identified from the other adsorption structures by X-ray photoelectron spectra. From these two diagrams, we can observe that the XPS spectra of Mode Ⅰ, Ⅵ, Ⅶ, and Ⅸ exhibit a single obvious spectral peak a, while the spectral intensities of such feature existing in Mode Ⅰ and Mode Ⅵ are visually strengthened compared with the other two model systems. Moreover, Mode Ⅰ and Mode Ⅵ have the similar peak intensities of the peak a on account of two equivalent nitrogen-silicon bonds to the silicon surface, however, the energy position of this feature in the spectrum of Mode Ⅵ at around 289.7 eV is 0.3 eV lower than the former. Compared to Mode Ⅰ and Mode Ⅵ, Mode Ⅶ and Mode Ⅸ with one relatively weaker peak a, both the widths of spectral character a and the peak intensities of these two cluster modes exhibit discrepancies. Firstly, it is noticed that the spectral feature a at 289.5-289.7 eV of Mode Ⅸ, where heterocyclic molecules have a 90$ ^\circ $ rotation to Mode Ⅶ, shows a wider energy range from 289.19 eV to 289.84 eV compared with Mode Ⅶ. In addition, it is not hard to find that a significant difference appears between Mode Ⅶ and Mode Ⅸ in the spectral intensity. In terms of Mode Ⅱ, Ⅲ, Ⅴ, Ⅷ, and X, two conspicuous features a and b with the similar spectral intensities presented in the spectra of the former three configurations are somewhat different from the Mode Ⅷ and Mode Ⅹ. However, the strong character b and peculiar weak shoulder a arising at a lower energy position in the spectra of Mode Ⅷ and Mode Ⅹ could not be employed to identify these two modes on account of the similarity in their photoelectron spectral characteristics. The roughly similar spectral profiles were observed in patterns Ⅱ, Ⅲ, and Ⅴ, all of which exhibit two nearly same spectral peaks at different energy positions. However, a small energy interval of about 0.2 eV from the peak a to peak b found in Mode Ⅱ is about 0.1 eV lower than that presented in both Mode Ⅲ and Mode Ⅴ within the corresponding energy ranges. For two Mode Ⅲ and Ⅴ, spectral peaks a and b possessed by the spectrum of Mode Ⅴ show a slight shift towards higher energy position in contrast with Mode Ⅲ, but it is deficient for discrimination between two adsorption structures. The N1s excited XPS spectra for any angle incidence of pyrazine molecule chemisorbed on Si(100) are also displayed in FIG. 4. It can be seen from the figure that the spectrum of Mode Ⅳ can be easily identified from the other modes, but the peaks a in the spectra of Mode Ⅱ, Ⅲ and Ⅴ have the similar energy positions that can not be identified from each other. And, the XPS spectra can not identify Mode Ⅶ, Ⅷ and Ⅹ, which have the rough similar profiles. Our study results indicate that the XPS spectra could not provide sufficient evidence to distinguish all the adsorbed structures, so the investigation of the NEXAFS spectra is expected to provide the possible identification of the adsorption configurations.

    Figure 2.  Calculated C1s XPS spectra of six binding modes (Mode Ⅰ-Ⅵ) for pyrazine molecule adsorbed on Si(100), and the major feature peak a, b, c, d are labeled.

    Figure 3.  Calculated C1s XPS spectra of four Tight-Bridge configurations (Mode Ⅶ-Ⅹ) for pyrazine molecule adsorbed on Si(100), and the major features a, b are labeled.

    Figure 4.  Calculated N 1s XPS spectra of the ten Modes (Ⅰ-Ⅹ) and the major features of each spectrum are marked.

    Table Ⅳ.  The energies of the adsorption configurations for spectral features shown in the calculated XPS spectra.

  • The theoretical C1s NEXAFS spectra of the ten adsorbed structures in core excited state that could give accurate information about relative energies and spectral profiles have been simulated. In our view, compared to four "Tight-Bridge" configurations (Mode Ⅶ-Ⅹ), other three types of adsorption models possessing roughly similar and more concise spectral profiles can be discussed together in terms of their spectral identification. The NEXAFS spectra of six predicted configurations belonging to "Butterfly", "Tilted", as well as "Cross-Row Bridge" structures are presented in FIG. 5, and absorbtion spectra of four Tight-Bridge configurations are shown in FIG. 6. Several distinctive spectral characteristics described in each spectrum, which show obvious dependence of the NEXAFS spectra on the chemical environment of carbon atoms, could be employed for the discrimination of all the studied systems. In addition, in order to obtain a detailed and careful comparison of the ten calculated absorption spectra, the energy values of the main spectral features marked in the spectra are listed in Table Ⅴ. In FIG. 5, a clearly weak shoulder b arises at about 285.1 eV on the right of peak a in the spectrum of Mode Ⅳ, but it does not show up in the other five total spectra, hence peak b in the spectrum of Mode Ⅳ can be taken as an evident difference from other binding modes. We can observe that the spectral intensities of the common peaks a corresponding to the lowest photon energy in the spectra of Mode Ⅰ and Ⅵ are stronger than this feature in Mode Ⅱ, Ⅲ, Ⅳ, and Ⅴ. Moreover, the spectral peak c exists in the spectra of Mode Ⅱ, Ⅲ, Ⅳ, and Ⅴ, which is not visible in Mode Ⅰ and Ⅵ, thus it is reasonable to divide two cluster models bonded by two nitrogen atoms from other four adsorption patterns. Obviously, for these two modes (Mode Ⅰ and Ⅵ), the relative spectral intensities between peak d and peak e are reversed, which suggests that Mode Ⅰ and Mode Ⅵ can be readily distinguished by their NEXAFS spectra. In the total spectra of Mode Ⅱ, Ⅲ, Ⅴ, the energy separation between spectral peak a and peak c of the spectrum of Mode Ⅲ is approximately 0.5 eV higher than that possessed by both Mode Ⅱ and Mode Ⅴ within the corresponding ranges of two features. Furthermore, the energy intensity of the other conspicuous spectral feature e exhibited in the spectrum of Mode Ⅲ at about 287.7 eV, however, is stronger than this feature presented in Mode Ⅱ and Mode Ⅴ, therefore Mode Ⅲ can be readily discriminated from the three modes on their NEXAFS spectra. In terms of the spectra of Mode Ⅱ and Mode Ⅴ, four peaks a, c, d and e appear at near the same energy position, while the appearance of peak c in the spectrum of Mode Ⅴ of "Cross-Row Bridge" structure is much weaker than the common characteristic of Mode Ⅱ, so feature peak c in the spectrum of Mode Ⅴ can be taken as a fingerprint to tell apart two modes.

    Figure 5.  NEXAFS spectra at the C-edge of six binding modes (Mode Ⅰ-Ⅵ) for pyrazine molecule adsorbed on Si(100), and the major feature peaks a, b, c, d, e are labeled.

    Figure 6.  NEXAFS spectra at the C-edge of four Tight-Bridge configurations (Mode Ⅶ-Ⅹ) for pyrazine molecule adsorbed on Si(100), and the major feature peaks a, b, c, d, e, f, g are labeled.

    Table Ⅴ.  The energies for main spectral features shown in the calculated NEXAFS spectra of the adsorption configurations.

    As demonstrated in FIG. 6, the energy intervals between the first feature a and the second feature b of the total spectra of Mode Ⅶ and Mode Ⅸ are about 0.2 eV lower than that of Mode Ⅷ and Mode Ⅹ, therefore four modes bound to the silicon surface by four bonds can be divided into two categories giving a consideration of two spectral peak a and b in each spectrum. Mode Ⅶ and Ⅸ showing the roughly similar spectral profiles are the first type and Mode Ⅷ and Ⅹ are the second type. It is worth noting that pyrazine molecule on the surface has a 90$ ^\circ $ rotation in the structure for the above two categories, but the discrepancies in their spectral profiles are well-marked. For the spectra of Mode Ⅶ and Mode Ⅸ, the shape and energy position of peak d, as well as the presence of shoulder feature f in the spectrum of Mode Ⅶ can be used to discern two configurations. As for Mode Ⅷ and Ⅹ, although they exhibit substantially similar XPS spectra, obvious distinctions on the NEXAFS spectral profile for two modes demonstrate the capacity of absorption spectra to identify both structures. Firstly, the energy positions of two conspicuous weak features c and d occur at about 286.5 eV and 287.2 eV, respectively, in the spectrum of Mode Ⅹ, however, there is only one weak spectral characteristic e between two relatively strong feature b and feature g in the spectrum of Mode Ⅷ. Secondly, the peak intensity of spectral feature b, which appears within a broad energy range of Mode Ⅷ, is significantly weaker than the same peak in Mode Ⅹ. The N 1s NEXAFS spectra of ten configurations are also shown in FIG. 7. In this figure, the major features of each spectrum are marked. The spectra profiles and energy positions of main peaks in the ten calculated spectra are so different that they can be used to identify the ten adsorption structures. According to the identification method of the above C 1s NEXAFS spectra, the absorption spectra of ten adsorption configurations can be easily distinguished by their N excited absorbtion spectra. Here, we do not make a detailed analysis for the distinction of the ten absorption spectra.

    Figure 7.  Calculated N 1s NEXAFS spectra of the ten Modes (Ⅰ-Ⅹ) and the major features of each spectrum are marked.

    Different adsorption configurations have a large influence on the molecular orbitals, especially when the pyrazine molecules exhibit a severe distortion. So, we calculated the orbits of major spectral peaks to observe the changes of molecular orbital in different molecular orientations of the pyrazine molecules. Molecular orbital of each labeled peak in the front view is shown in FIG. 8. The fist peaks of all modes come from the C1s$ \rightarrow $LUMO resonances. However, we did not find a more obvious law between the configurations and the type of orbits. In order to drill down to see the origin of main features discussed above in the total spectra and a better comprehension of the dependence of the NEXAFS spectra on the local chemical environment, the NEXAFS spectra of a pyrazine molecule on silicon of known orientation, as well as the decompositions of the total spectra obtained on the local bonding environments of carbon atoms, are illustrated in FIG. 9 and FIG. 10. We can see from the investigation of the individual components of total spectra that the carbon atom at the C$ _{1} $ site generates the first absorption characteristics a of all the structures except for Mode Ⅳ, whose peak a is mainly contributed by carbon atom at the C$ _{4} $ position. Obviously, the decomposed spectra coming from the transition of carbon atoms at the same position in the pyrazine ring do not exhibit completely uniform spectral features, however, the total spectra of Mode Ⅰ and Mode Ⅵ derived from the carbon atom at the C$ _{1} $ site due to the same chemical environment of four carbon atoms in the structure of these two modes show similar spectral profiles. As for the components of total spectra of Mode Ⅱ, Ⅲ, Ⅴ, Ⅶ and Ⅸ, carbon atoms with non-binding to the silicon atoms generate more concise spectral profiles in the comparison to another carbon atoms bound to the silicon atoms. Moreover, four carbon atoms in the Mode Ⅳ, Ⅷ, and Ⅹ show more detailed spectral differences, which manifest sensitivities and specificities of the NEXAFS spectra to the local structure of adsorbed modes.

    Figure 8.  Molecular orbital of some labeled peaks a, c, e, g in the front view (contour isovalue = 0.02).

    Figure 9.  C1s NEXAFS spectra and the individual components according to the classification of carbon atoms for the six adsorption structures (Mode Ⅰ-Ⅵ).

    Figure 10.  C1s NEXAFS spectra and the individual components according to the classification of carbon atoms for four configurations belonging to Tight-Bridge types (Mode Ⅶ-Ⅹ).

  • In combination with DFT and cluster calculations, we have computed the geometric and electronic structures of ten possible adsorption configurations of a pyrazine molecule chemisorbed on the Si(100) surface, and investigated theoretical identification of all the predicted adsorbed structures by means of the XPS and NEXAFS spectroscopy. Our calculation results of binding energy show that "Cross-Row Bridge" structure is the most likely adsorption configuration presented in experiments. In the theoretical simulation of the XPS and NEXAFS spectra, while the XPS spectrum is deficient in the identification of all the adsorption structures, the NEXAFS showing a strong dependence on the local structure of carbon atoms is competent to discern ten different pyrazine/Si(100) molecules. Furthermore, the decompositions of the total NEXAFS spectra based on the local bonding environments of carbon atoms have illuminated the origin of main features, which manifested the first excited energy was generated by the carbon atoms near the nitrogen atom rather than the carbon atoms bonded to the silicon atoms.

  • This work is supported by the National Natural Science Foundation of China (No.11874242, No.11804196, No.11804197). The support provided by China Scholarship Council (CSC) for Yong Ma to Royal Institute of Technology (KTH) is acknowledged. Thanks to the support of the Taishan Scholar Project of Shandong Province.

Reference (38)

Catalog

    /

    DownLoad:  Full-Size Img  PowerPoint
    Return
    Return