Infrared spectroscopy (IR) is one of the most commonly used spectroscopic techniques to identify specific functional groups of complex materials. Absorbent groups in the infrared region absorb within a specific wavelength range. In this way, IR spectroscopy can be very sensitive for determining the functional groups of the sample, since different functional groups absorb specific IR radiation frequencies. The absorption peaks of molecules are obtained by comparing their measurement with spectral databases1,2. This can also be used for qualitative and quantitative analysis of complex mixtures of similar compounds. Molecular bonds are subject to different vibrations and rotations; consequently, atoms within molecules are unstable under external forces. However, the presence of functional groups in 2D materials at the edges and center of the 2D surface is quite difficult to identify since the IR spectra usually assume the materials are 3D. Therefore, it is crucial to introduce analytical methods that allow the investigation of 2D carbon materials such as graphene and its derivatives using IR spectra.
Graphene quantum dots (GQD), a novel type of graphene nanomaterial, have attracted tremendous attention. GQD is nano-sized material that is of particular importance as functional material with a variety of applications across all disciplines. In order to examine the fluorescence of the synthesized GQD, optical spectroscopic studies are performed, which reveal a red shift in the photoluminescence (PL) emission when GQD is functionalized with amine groups at different pH values3. The pH modification causes the protonation or deprotonation of the functional groups. The functional group also causes the structural deformation of the GQD in the aromatic ring. This deformation causes an energy level in the mid-gap that allows electron transitions to occur in this state4.
Generally, when the oxygen functional group of particles grows three-dimensionally it increases particle size. However, it is reported that the changes in the oxygen functional group of GQD as reaction times elapse reveal a blue shift in PL spectra means particle size decrease5. Thus, there is no guarantee that the particles are 3D, they may form 2D materials with unique functional groups either in the center or at the edge that causes different PL responses. In this condition, GQD may possess complex structures with unpredictable vibrational motions and interactions6–13. Many researchers employed FTIR to investigate functional groups in GQD; however, atomic or molecular dislocation is rarely observed since considered 3D material. For instance, the FTIR applies to identify sulfonic-GQD as rich in sulfonic and hydroxyl groups on their surfaces7,14. These functional groups improved the hydrophilicity and stability of the sulfonic-GQD in an aqueous system. Due to the dislocation of various active groups, it is difficult to correlate the informational FTIR spectra with their hydrophobicity-enhancing abilities. In contrast, with nitrogen-containing GQDs such as pyridine and nitrogen with low oxygen content, remarkable blue shifts15 in PL spectra were observed. The composition and structure of GQD are characterized by various spectroscopic techniques such as X-ray photoelectron spectroscopy (XPS), and FTIR. The infrared spectra of the prepared GQD showed absorption bands at 1289 cm− 1 and 1223 cm− 1, which are attributed to the C-O-C stretching vibrations16,17. However, the results did not indicate the position of the functional group attached to the GQD. Some researchers also introduced a simulation method to investigate the functional group of GQD on their corresponding FTIR. The infrared simulation of the GQD structures with edge oxygen and amino groups, was studied using the ab initio method18. As a result, the functional groups at the edge of GQD change the relative order of their energy levels, particularly the energy levels of the highest occupied molecular orbital (HOMO) and the lowest unoccupied molecular orbital (LUMO). FTIR spectral data from ab initio calculations can be used to determine the chemical composition of the functional groups in the GQD. Therefore, the position and location of these functional groups still need to be investigated to give corresponding different FTIR characteristics. Instead of using ab initio calculation, methods with simple and faster approaches are needed. The semi-empirical method, Austin Model1 (AM1) gives comparable results with the best linearity between the calculated and experimental frequencies19–21. Therefore, AM1 consider the most helpful auxiliary tool for the FTIR spectroscopy identification of GQD.
In this study, we used a semi-empirical method to calculate the infrared spectra of pyrene-like molecules to provide insight for a simple analysis of functional GQD. The investigations explain the complex molecular level vibration of GQD correlated with their structure. These vibrations could also inform the sensitivity and selectivity of the molecule under investigation22. Therefore, this method may give an insight into the applications of FTIR spectra for 2D carbon materials analysis, especially pyrene-based molecules and other complex 2D materials. Geometry optimization and vibrational frequencies of the molecules (Benzene, Pyrene, and GQD) are determined and analyzed using the semi-empirical methods AM1. We analyzed the IR spectra of these molecules at the aromatics vibrations (1400–2000 cm− 1), which is identified as C = C stretching mode. At the same time, the AM1 method successfully predicted infrared intensities and a spectral intensity pattern of GQD. We also reported the excitation spectra to review the electronic transition of the molecules in the visible region (200 to 900 nm) by Zerner's Intermediate Neglect of Differential Overlap (ZINDO) method23. The excited spectra showed that the molecule under study was a GQD material.
Molecular Model And Computational Methods
The initial molecular structure model.
Benzene (C6H6) and Pyrene (C16H10) consist of their original condition of one and four-membered rings (Fig. 1a & 1b). Figure 1c and 1d show a nine and sixteen-ring pyrene extension model as pyrene-like GQD (pGQD). Throughout the manuscript, these pyrene-like GQDs with 9 and 16-ring structures are referred to as p1GQD and p2GQD. The model of the functionalized GQD is developed by adding five hydroxyl groups (-OH) and one methyl group (-CH3) on the edge and four oxygen atoms in the center of the surface of p1GQD, which are then termed functional GQDs (fGQD) as in Fig. 1e.
Geometry optimization and IR simulation with the AM1 method.
All calculations were performed on an Intel Core i5 computer with a 3.2 GHz CPU and 16.0 GB of physical memory. Optimizing the structures, vibrational modes, and excitation spectra were done using a Winmostar V10.0.7 package containing an AM1 semi-empirical method24. The initial geometrical structure of the pyrene molecule assumes to consist purely of regular networks with a CC distance of 1.397 Å, a CH distance of 1.084 Å, and bond angles of 120°. This approximation of the poly aromatic hydrocarbon (PAHs) refers to reported works25. Initially, all compounds were optimized using the molecular mechanic method. Then, an additional re-optimization by employing the AM1 semi-empirical method26. The convergence limit is determined based on orientation observations, i.e., after reaching the gradient limit of the energy change of 0.05 kcal/mol/Angstrom. The geometry optimization determines based on a second-order Taylor energy expansion around the current point27. Once a stable structure is available, the system provides the result of calculation data containing the energy and the electronic structure as an output file. A force calculation must then be carried out to obtain the vibration modes. The result matched against experimental data for that vibrational mode. The following calculation determines the absolute and percentage error in the predicted frequency. If the match between the calculated and the experimental frequencies is not close enough (within 10%), then the semi-empirical technique is not considered good for predicting that type of vibrational mode28.
Excitation spectra determination according to the ZINDO method.
The excitation spectra simulations for the structure of the geometry optimization results with AM1 were performed by one-step or single-point calculations using the semi-empirical ZINDO method23. The calculation consists of two parts. The first is the ground state calculation, which gives molecular orbital coefficients and eigenvalues. The ground state calculation is then followed by a configuration interaction calculation. The energy corresponding to a singlet-singlet transition between pure configurations is given by:
$$\varDelta {E}_{ia}={\epsilon }_{a}-{\epsilon }_{i}-{J}_{ia}+2{K}_{ia}$$
1
where \({\epsilon }_{a}\) and \({\epsilon }_{i}\) are the orbital energies of orbitals a and i, respectively. \({J}_{ia}\) is the molecular Coulomb integral \(\left(ii\right|aa)\) and \({K}_{ia}\) is the molecular exchange integral \(\left(ia\right|ai).\) The calculated transition energies are then entered as diagonal elements of the Hamiltonian of the configurational interaction (CI). The off-diagonal elements of the CI Hamiltonian are:
\(⟨{}^{1}{\phi }_{0}\left|H\right|{}^{1}{\phi }_{i\to a}⟩\) = 0 (2)
\(⟨{}^{1}{\phi }_{i\to a}\left|H\right|{}^{1}{\phi }_{j\to b}⟩\) = 2\(\left(ai|jb\right)-\left(ab\right|ij)\) (3)
The calculation is done with CI restrictions and with a single excited CI, then the orbital calculation command is activated, and an electronic transition spectrum is generated in the form of a wavelength (λ) range and its oscillator strength. This determination of the excitation spectra was performed to examine the electronic transition level of pGQD in the visible range (200–900 nm). In addition, these data confirm that the studied compound is a quantum dot material.
Analysis methods of vibrational mode and excitation spectra.
The pyrene molecular mode vibrational pattern is used to refer to the original graphene quantum dot vibrational mode model (pyrene-like). By analyzing the intensity and pattern of the C = C (sp2) vibration at wavenumber 1400–1800 cm− 1 on the pyrene molecule, they can be used to determine the microscopic properties of the pyrene-like GQD structure. It can be practically used in experiments to quickly analyze specific structures in pure GQD or functionalized GQD. After analyzing the shape of the GQD molecular structure based on the vibrational mode pattern of the pyrene molecule, the next step is to investigate whether the excitation spectrum is in the UV (200 to 400 nm for pyrene) or in the visible (400 to 700 nm for pGQD) region. From the above two analyses, an electron distribution analysis using the molecular electrostatic potential (MESP) map was performed for the pristine and functionalized pyrene-like GQD models to know that they consequently affect the excitation. All of these analyzes were used to propose a mechanism, such as the electronic transition properties of the GQD pyrene-like model due to functionalization.