Structural parameters for the optimized q1 and q2 nanoclusters are shown in Fig. 1. As
can be seen, upon optimization, q1 loses the structural symmetry and forms a twisted shape. In q2 on the other hand, three SiO4 tetrahedra form a closed triangle, which is the basic structural unit of quartz. The cohesive energy for these structures is evaluated from
E coh = E[(SiO2)n] − n E[Si] − 2n E[O] (1)
where E[(SiO2)n], E[Si] and E[O] are the total energies of the nanocluster and isolated silicon and oxygen atoms in a same supercell, respectively. Table 1 contains the cohesive energies per atom for q1 and q2. The cohesive energy is lower for q2, which means that q2 is more stable than q1.
The density of states (DOS) plots is shown for q1 and q2 structures. The computed DOS data are in good agreement with previous studies [35]. As can be seen, DOS for q1 shows different spin-up and spin-down densities, while q2 has almost similar characteristics for up and down states. This is more obvious near the Fermi level. For q2, there is a strong spin-up peak below the Fermi level and a similar spin-down peak above it. Figure 2 illustrates the highest occupied and lowest unoccupied molecular orbitals (HOMO and LUMO) for q2. The LUMO state is mainly spin-down, while the HOMO is mainly spin-up. The spin splitting near the Fermi level means different spin-up and spin-down transmissions along the structure in low bias voltages. Such splitting and different transmissions are favorable as fundamental characteristics in spintronics. Despite different values of band gaps, both nanoclusters are semiconductor. However, bulk silicon dioxide is an insulator. This feature has been already observed in metal oxide clusters [36]. In the calculated DOS for bulk silica by DFT [36] the peak in the top of the valence band is not present.
Figure 3 shows the spin-up and spin-down projected DOS (PDOS) plots of q2 along the z direction in different energies. In both plots, three main areas exist between 0.15 and 0.3 Å (place of outer atoms). These areas show that the main part of spin characteristics is due to the outer silicon atoms and are near the Fermi level, and about the energies of −0.25 and −0.75 eV. Despite edge Si atoms, inner Si atoms and oxygen atoms have not a major role in the DOS.
Figure 4 illustrates the effect of solvolysis on DOS of bulk surface silica and q1 and q2 nanoclusters. In solvent, because of high activity of the dendrite oxygen atoms, these atoms have a high tendency to bond to hydrogen. In fact, oxygen atoms are replaced by hydroxyl groups. Here, for ease of use, we denote fully hydroxyl terminated nanoclusters as qH. The hydroxyl terminated surface may change the electronic and magnetic properties of nanoclusters. Here, a sharp drop in the value of magnetization is observed and the DOS spectrum shifts to lower energies. This shift is larger for q2 than for q1. Moreover, the addition of hydrogen atoms to pristine silicon dioxide nanoclusters can convert them from p-type to n-type semiconductors.
Table 1
Magnetic moment (µ), total force (Ftot), binding energy (Eb), Fermi energy (Ef) from DFT (from DFTB+)and charge transfer (ΔQ) for bare and functionalized nanoclusters..
system | µ(bohr/cell) | Ftot (Ry/au) | Eb (eV/atom) | Ef (eV) | ΔQSi (e) | ΔQO (e) | ΔQH (e) | ΔQN (e) |
---|
q1 | 14.82 | 2.615 | −5.573* | −7.53 | 0.147 | −0.214 | | |
q1H | 0.77 | 0.740 | −5.990 | −5.59 | −0.238 | 0.345 | 0.238 | 0.238 |
q2 | 20.57 | 0.066 | −6.158* | −6.83 | 0.213 | −0.176 | | |
q2−NH2 | 19.96 | 2.775 | + 5.321 | −6.67 | −0.116 | 0.153 | | 0.116 |
q2H | −0.29 | 0.024 | −7.670 | −4.35 | −0.015 | 0.338 | 0.015 | 0.015 |
q2H−NH2 | −0.20(-0.32) | 0.957(0.95) | -7.87 | −4.33(-5.54) | −0.105 | 0.023 | 0.154 | 0.105 |
q2H−CH2NH2 | 0.24(-0.42) | 1.159(0.95) | −6.734 | −4.58(-5.54) | 0.107 | −0.008 | 0.065 | −0.107 |
q2H−C2H4NH2 | 0.24(-0.35) | 1.223(0.95) | −6.806 | −4.44(-5.54) | 0.102 | −0.003 | 0.069 | −0.102 |
q2H−C3H6NH2 | 0.01(-0.55) | 1.264(0.95) | −6.810 | −2.82(-5.54) | 0.122 | −0.109 | 0.069 | −0.122 |
* Cohesive energy per atom form (1).
The electron distribution of hydrogenated and bare q2 structures are illustrated in Fig. 5. The electrons are localized on oxygen atoms. This confirms that these atoms are favorable sites for electron-accepting functional groups. In fact, the electron-donating functional groups such as amino groups attract to silicon atoms and cause stress in nanoclusters.
After the evaluation of the active bonding sites for amine group, the next step was to find the activation energy along that path, so that kinetics of the recombination reaction could be forcasted. An initial path was constructed and represented by a discrete set of images of the system connecting the initial and final states. To calculate the activation energy barrier the nudged elastic band (NEB) method implemented in the Quantum ESPRESSO package is used. Results of activation energy for reaction between q2H and water molecule indicates that are suitable process. Functionalization of silica nano-cluster in an exothermic reaction occurs where a water molecule is released.
Based on a previous work, the most favorable situation is to replace one of the edge oxygen atoms by amino groups [38]. Figure 7 shows the optimized structures for amine group functionalization of q2.
As can be seen, these functional groups can deform the cluster structure. Recently, the amino-functionalized silica nanoparticles have been successfully fabricated [39]. In this work, the amino functional groups are added to the nanocluster along the (0, 0, 1) surface.
All terminated O-H groups have a potential to replacing by amino groups and making silica core-shell. Core–shell particles with different ranges of pore sizes in the shell are used in many applications such as chromatography and drug delivery. The productivity of manufacturing core–shell silica particles is low because high tendency of silica fragments to aggregation. Silica spheres are employed to prepare various inorganic/composite core–shell structures although silica can also be removed by acid etching or alkaline washing to produce hollow structures. Core–shell silica particles further coated with a layer of carbon because carbon is chemically inert and highly stable. Therefore, coating silica nano particle by carbon shell is desirable. According to results of this investigation, silica nano structure functionalized with amide chain due to homogenous charge density around amide chain can be stable. In next step, the core-shell particle of silica nano structure coated with amide shell is investigated. In next we investigate the stable structure of core silica nanostructure coated by amine rings. Table 1 shows the binding energies for the functionalized q2 clusters. Functionalization of q2 with amino group is an endothermic process. In addition, as data shows, the amino group has no significant effect on magnetic moments of pristine and hydrogenated nanoclusters.
On the other hand, as Table 1 shows, the functionalized nanoclusters with amine groups are energetically more stable than the pristine ones. In addition, these processes are exothermic and the stability of the system increases by increasing the number of carbon atoms in the hydrocarbon chain of amine. Amine-group functionalization causes charge transfer from Si to amine group. In fact, the resonance electron affinity of nitrogen atom overcomes the electronegativity. Because of that, electron density on nitrogen atom becomes higher and the reactivity of nanocluster for binding to other groups increases. However, in biosystems, the free electron pair on nitrogen can covalently bind to other electron acceptors and form a strong chemical bond. These strong bonds can affect or even damage the structure of the group which is bonded to the N atom. Therefore, in these applications, such as in drug delivery, a modest bond is more favorable [7, 8] and the amine groups connect to silica nanocluster via a hydrocarbon link. According to the value of charge transfer between N and Si atoms for q2H−CH2NH2 (Table 1) the presence of a hydrocarbon interface between amino group and silica nanostructure causes a balanced electron distribution on N−C bond at the interface.
The projected density of states plots for q2H-amine systems are shown in Fig. 8–11.
By increasing the length of the amine hydrocarbon chain, the occupation of energy states near the Fermi level gradually becomes less. PDOS results show that this effect is due to amine group. It seems that in amine-functionalized silica nanoparticles the distance of functional group from silicone atoms affects the occupation of energy states of nanocluster.
Density functional theory and self-consistence charge density functional tight bonding are employed for calculating density of sates. According to Fig. 8, results of these two methods are dramatically different. As density of states for carbon atoms shows, the main source of such difference is in calculating the energy occupation for carbon atoms. This nonconformity determines that generalized gradient approximation in interpretation of orbitals combination is weak. In self-consistence charge density functional tight bonding method, the hybrid Slater-Koster files are employed. Results of SCC-DFTB reflects that density functional theory and self-consistence charge density functional tight bonding calculate the occupation of each state of this nano-cluster totally different.