Maryam Sadeghi and
Bahram Khoshnevisan*
Faculty of Physics, University of Kashan, Kashan, Iran. E-mail: b.khosh@kashanu.ac.ir
First published on 25th June 2024
In this study, we modeled a drug delivery system consisting of Ti3C2 MXene nanosheets as a carrier and 5-fluorouracil (FU) as a selected drug molecule using density functional theory (DFT) computations. During the adsorption procedure, electronic, magnetic and structural properties were calculated. Our results showed that the adsorption of FU drugs on the Ti3C2 surface is thermodynamically favorable. Our spin-polarized calculations also determined that the magnetization of Ti3C2 after FU adsorption does not change significantly, which is an important factor for magnetic hyperthermia and drug delivery. In addition, our calculations indicate that in the slightly acidic environment of tumor tissue, FU could start to be released (by increasing distance from the MXene surface and then instability of the complex) from the Ti3C2 surface without any substantial change in the structural properties. This study could provide a deep understanding of the interaction mechanism of 2-dimensional (2D) MXene materials with drugs at the atomistic scale and have an important contribution to the discovery and application of novel 2D materials as drug delivery systems.
Another group of 2D layered materials discovered by two groups of researchers from Drexel University are MXenes.11 The new family with the general formula of Mn+1AXn is called the MAX phase and includes carbide, nitride and carbonitride of transition metals. In this formula, M, A and X are: transition metal atom, a main group element (mostly group 13 or 14 element from the periodic table), and carbon or nitrogen atoms, respectively, and n can be 1 to 3. MXenes are obtained by the selective etching of the A-layers of the MAX phase either in pure form (Mn+1Xn), or with surface functional groups, Mn+1XnTx, where Tx can be fluorine (F), hydroxyl (OH) and oxygen(O).12
MXenes have excellent properties in terms of electrical, optical and thermal stability, so most conducted studies on them have focused on energy storage,13,14 catalysis,15 and sensors.16,17 MXenes also have (i) high specific surface area, which is an efficient factor for high drug loading, (ii) tunable layered structure, and (iii) hydrophilic nature, which make them to be considered as new inorganic nanostructures for biological and biomedical applications.18 MXenes also have high photothermal conversion efficiency, which makes these materials suitable for photothermal therapy and hyperthermia.19,20
Ti3C2 is the first MXene that was discovered in 2011 by Michael Naguib et al.21 So far, many theoretical and experimental studies have been performed on it in various areas. In 2017 and 2020, A. M. Jastrzębska et al. investigated the cytotoxic effects of Ti3C2. The results revealed that these nanosheets exhibited the highest cytotoxic effect on the cancerous cell line of A549 while the normal cell line HaCaT showed no changes across all concentrations and remained unharmed.22 Furthermore, they deposited a layer of Ti2O3 on the Ti3C2Tx MXene using ultrasound and mild thermal oxidation after the synthesis process and indicated that it could probably regulate the cytotoxicity of samples to cancer cells. These samples were toxic to all cancer cell lines up to 375 mg l−1.23 In 2018, C. Xing et al. developed a MXene/DOX@cellulose hydrogel nano-platform, where MXene comprised Ti3C2 nanosheets, to study the release of the DOX model drug and its photothermal performance. This platform showed desirable biocompatibility, good photothermal efficiency and high capacity for Dox drug loading, and it was useful for immediate tumor destruction and preventing its recurrence.24 For the first time, the performance of Ti3C2 MXene for chemotherapy drug delivery was investigated by X. Han in 2018. The drug was released by pH-responsive method and NIR laser technique, and a high synergistic therapeutic outcome was obtained in the treatment of cancer.25 G. Liu et al. used a multifunctional nanoplatform (Ti3C2–DOX), in which the nanosheets have a small lateral size (∼100 nm) and included the stable surface functional group Al(OH)4. The work offered a new effective strategy for cancer therapy based on surface-modified Ti3C2 nanosheets.26 B. Zhu et al. combined gold nanorods (GNRs) with Ti3C2 nanosheets and prepared intelligent sandwich-like Ti3C2@GNRs/PDA/Ti3C2 nanohybrids, which were used for drug delivery with synergistically enhanced NIR drug release behavior. The results showed that this platform had pH/NIR-responsive drug release. Overall, it was promising for use in photothermal therapy and in remote controllable drug delivery.27 In 2020, Y. Liu et al. developed a heterostructured titanium carbide–cobalt nanowires (Ti3C2–CoNWs) nanocarrier. When DOX was used as a model drug, this nanocarrier had a high drug loading ability and showed drug release behavior induced by pH/NIR stimulations.28 In 2023, Z. Bai et al. showed that under the irradiation of a NIR-I laser, the Ti3C2 core in the DOX/Ti3C2/Apt-M therapeutic platform enhanced the effect of PTT and promoted the release of the DOX to enhance the chemotherapy effect.29
To the best of our knowledge, theoretical calculations have not been reported so far on using Ti3C2 as a drug delivery system for fluorouracil. Fluorouracil (abbreviated: 5-FU or FU) is one of the common cancer treatment drugs for some cancers, such as breast, stomach and colon.30
In this work, based on Density Functional Theory (DFT), we have modeled Ti3C2 MXene nanosheets as a carrier for FU. First, we investigated the adsorption behavior of 5-FU on Ti3C2 nanosheets. Then, we calculated the electronic structure and magnetic properties of Ti3C2@FU systems. Finally, drug release from the carrier surface, which is an important part of the drug delivery process, has been investigated. We hope this study provides a deep understanding of the interaction mechanism at the atomistic scale, and provides some insight for biomedical applications of Ti3C2 MXene nanosheets.
Fig. 1 (a) Ti3C2 unit cell, (b) side and (c) top views of the 3 × 3 × 1 supercell of Ti3C2 MXene. Three special sites for the absorption process are marked on the (c). |
In the relaxed structure of the Ti3C2 monolayer, the a-lattice parameter of its hexagonal structure was determined to be 3.07 Å, and the bond lengths of Ti1–C and Ti2–C were found to be 2.05 and 2.20 Å, respectively. The thickness of the sheet (distance between Ti1 atoms at the upper and lower level of the surface) was 4.72 Å. For comparison, other theoretical and experimental results are summarized in Table 1, and it can be seen that our results are in good agreement with them.36–48
a (Å) | Ti1–C (Å) | Ti2–C (Å) | Thickness (Å) | Magnetization (μB) | |
---|---|---|---|---|---|
a * means experimental reports. | |||||
This work | 3.07 | 2.05 | 2.20 | 4.72 | 1.85 |
Other reports | 3.07 (ref. 36 and 48) | 2.064 (ref. 47) | 2.21 (ref. 36) | 4.66 (ref. 36) | 1.9 (ref. 36), 1.93 (ref. 47) |
3.10 (ref. 46) | 2.062 (ref. 37) | 2.22 (ref. 37) | 4.64 (ref. 37) | 1.87 (ref. 38) | |
3.08 (ref. 39) | 2.083 (ref. 39) | 2.21 (ref. 40) | 4.60 (ref. 41) | 2.2 (ref. 42) | |
*3.08 (ref. 43), *3.09 (ref. 44) | *1.87 (ref. 39 and 45) |
FU is one of the common drugs used to treat cancer and has five different structures: the Diketo (contains two carbonyl groups), Dienol (contains two hydroxyl groups), and three Ketoenol tautomers (with one carbonyl and one hydroxyl groups). We have investigated all five structures in our simulations. It also must be noted that Ketoenol has three different configurations, and we took the most stable of them in our study (Fig. 2).49,50
At first, each structure was simulated and optimized, and the bond lengths after relaxation calculations for each of the three drug forms are given in the first row of Tables 2–4. In comparison with the literature, there is a good agreement between our results and the previous theoretical and experimental reports.51–53
N5–H9 | C3–O12 | C1–F8 | C2–H11 | N4–H10 | C6–O7 | C6–N5 | N5–C3 | C3–C1 | C1–C2 | C2–N4 | N4–C6 | |
---|---|---|---|---|---|---|---|---|---|---|---|---|
Diketo | 1.01 | 1.23 | 1.35 | 1.09 | 1.01 | 1.22 | 1.39 | 1.40 | 1.46 | 1.35 | 1.37 | 1.39 |
Ref. 51 | — | 1.22 | 1.34 | — | — | 1.22 | 1.40 | 1.40 | 1.47 | 1.35 | 1.38 | 1.39 |
⊥Ti1 | 1.02 | 1.23 | 1.37 | 1.09 | 1.01 | 1.22 | 1.39 | 1.40 | 1.45 | 1.35 | 1.37 | 1.39 |
⊥Ti2 | Not converge | — | — | — | — | — | — | — | — | — | — | — |
⊥C | 1.02 | 1.31 | 1.37 | 1.08 | 1.01 | 1.22 | 1.39 | 1.39 | 1.40 | 1.36 | 1.39 | 1.39 |
After release ⊥C | 1.02 | 1.31 | 1.36 | 1.08 | 1.01 | 1.22 | 1.39 | 1.38 | 1.41 | 1.36 | 1.38 | 1.39 |
H9–O7 | O7–C6 | C6–N5 | N5–C3 | C3–C1 | C1–C2 | C2–N4 | N4–C6 | C2–H10 | C1–F8 | C3–O12 | N5–H11 | |
---|---|---|---|---|---|---|---|---|---|---|---|---|
Ketoenol | 0.98 | 1.35 | 1.36 | 1.42 | 1.45 | 1.36 | 1.37 | 1.30 | 1.09 | 1.35 | 1.23 | 1.02 |
Ref. 52 and 53 | 0.98 | 1.34 | 1.36 | 1.42 | — | 1.37 | — | 1.30 | — | 1.35 | 1.22 | 1.02 |
⊥Ti1 | 0.98 | 1.34 | 1.36 | 1.42 | 1.44 | 1.36 | 1.37 | 1.30 | 1.09 | 1.38 | 1.23 | 1.02 |
⊥Ti2 | 0.98 | 1.35 | 1.36 | 1.42 | 1.44 | 1.36 | 1.37 | 1.30 | 1.09 | 1.37 | 1.23 | 1.02 |
⊥C | 0.98 | 1.34 | 1.36 | 1.42 | 1.45 | 1.36 | 1.37 | 1.30 | 1.09 | 1.36 | 1.23 | 1.02 |
After release ⊥C | 0.98 | 1.34 | 1.36 | 1.4 | 1.45 | 1.36 | 1.37 | 1.3 | 1.1 | 1.35 | 1.23 | 1.02 |
H9–O7 | O7–C6 | C6–N4 | N4–C2 | C2–O12 | O12–H11 | C2–C1 | C1–C3 | C1–F8 | C3–H9 | C3–N5 | N5–C6 | |
---|---|---|---|---|---|---|---|---|---|---|---|---|
Dienol | 0.98 | 1.35 | 1.33 | 1.33 | 1.34 | 0.98 | 1.40 | 1.38 | 1.35 | 1.09 | 1.34 | 1.34 |
Ref. 52 and 53 | 0.98 | 1.35 | 1.34 | 1.33 | — | — | — | 1.38 | 1.35 | — | — | 1.34 |
⊥Ti1 | 0.98 | 1.34 | 1.34 | 1.32 | 1.34 | 0.98 | 1.40 | 1.38 | 1.38 | 1.09 | 1.34 | 1.34 |
⊥Ti2 | 0.98 | 1.34 | 1.33 | 1.32 | 1.36 | 0.10 | 1.39 | 1.38 | 1.38 | 1.09 | 1.34 | 1.34 |
⊥C | 0.98 | 1.34 | 1.34 | 1.32 | 1.34 | 0.98 | 1.40 | 1.38 | 1.38 | 1.09 | 1.34 | 1.34 |
After release ⊥Ti2 | 0.98 | 1.35 | 1.34 | 1.33 | 1.35 | 0.98 | 1.40 | 1.38 | 1.36 | 1.09 | 1.34 | 1.34 |
The selected drug molecules were approached to the Ti3C2 surface at the ⊥Ti1, ⊥Ti2 and ⊥C sites. For example, Fig. 3 displays the most stable Ti3C2@ FU systems before and after the relaxation calculations, and other configurations before and after the relaxation can be seen in Fig. 1S–3S in the ESI† Section (Appendix). In the cases of Ketoenol and Diketo, ⊥C is the most stable configuration. Meanwhile, in the Dienol case, ⊥Ti2 is the most stable site for the drug. The bond lengths of the drug molecules after adsorption were calculated and are summarized in Tables 2–4. It can be seen that the bond lengths of the drugs have not changed after adsorption, which confirmed that the structures of the drugs were preserved.
Fig. 3 (a) Ti3C2@Dienol, (b) Ti3C2@Ketoenol, and (c) Ti3C2@Diketo before (left) and after (right) the adsorption process in the most stable configurations. |
In order to obtain the most stable structure, the adsorption energy was calculated for each of the three configurations. The formula for calculating adsorption energy is:
Ead = (ETi3C2@FU) − (ETi3C2 − EFU) | (1) |
The structure with the highest negative adsorption energy was known as the most stable Ti3C2@drug system. Table 5 shows the calculated adsorption energies of each structure in the considered adsorption sites. Ead of the Diketo on the ⊥C site was calculated to be −3.39 eV, which is the most stable system. Moreover, ⊥the Ti2 site was the most stable configuration for Dienol with Ead equal to −2.36 eV. For Ketoenol, the ⊥C site with Ead: −2.03 eV was the most stable structure.
Ti3C2@Diketo | Ti3C2@Ketoenol | Ti3C2@Dienol | |
---|---|---|---|
⊥C | −3.39 | −2.03 | −2.01 |
⊥Ti2 | Not converged | −1.94 | −2.36 |
⊥Ti1 | −2.01 | −1.99 | −2.01 |
We compared our results with literature (Table 6), and our calculated Ead were higher than the adsorption of the FU drug on some nanosheets, such as SiG, BN, InN, AlN, ZnO, SWCNTs and also BNNT.50,55–60 In these nanosheets and nanotubes, it is necessary to increase the adsorption energy using some process such as doping because the Ead is low. On the other hand, our results were lower than those from the Diketo adsorption on GaN.62 In this case, there are problems with the drug releasing procedure. Our Ead were comparable with FU adsorbed on GNS and in some sites on graphene oxide, respectively.61,63
Ead (eV) | Carrier surface | Form of FU drug | Ref. |
---|---|---|---|
−0.083, −0.172, −0.082 | (SiG) silicon graphene nanosheet | Diketo/Ketoenol/Dienol | 50 |
−0.56 | BNNS (boron nitride nanosheet) | Diketo | 55, and 56 |
−1.3, 0.81, −0.84 | InN, AlN, GaN (nanosheet) | Diketo | 55 |
−0.423 | ZnONS (zinc oxide nanosheet) | Diketo | 57 |
−0.358 | (4,0) SWCNT | Diketo | 58 |
−0.13 | (8,0) BNNT (boron nitride nanotube) | Diketo | 59, and 60 |
−1.74 | (GONS) graphene oxide nanosheets | Diketo | 61 |
−5.41, −4.29, −1.37 | (GaN) gallium nitride | Diketo/Ketoenol/Dienol | 62 |
−3.116 | GNS (graphene nanosheet) | Diketo | 63 |
Fig. 4 shows the results of the partial DOS and total DOS diagrams of the Ti3C2@FU complexes. A careful comparison between the DOS of Ti3C2 and Ti3C2@FU systems reveals that the insertion of a drug into Ti3C2 altered the amount of the DOS at the Fermi surface. Actually, the amount of DOS at the Fermi surface increased slightly with respect to the Ti3C2 surface; these new states belonged to the drugs. Further investigation showed that the partial DOS of the drugs did not change after adsorption. This means that the electronic properties of the drug were slightly affected during the interaction, and it can be said that the structure of the drug was preserved during the adsorption.
Fig. 4 (a) The partial density of state (PDOS) of the optimized Ti3C2 MXene. (b–d) The density of state (DOS) of the Ti3C2@FU surface complex. |
The DOS diagram of the bare Ti3C2 shows the metallic behavior, and the Ti1 atoms have the main contribution on the Fermi level (Fig. 4a).
The molecular structure stability and reactivity could be predicted by chemical potential and chemical hardness. We have calculated the chemical potential and hardness using the following equations:
(2) |
(3) |
The results are shown in Table 7. It is found that after the adsorption, the values of the HOMO–LUMO energy gap (H/L gap) and chemical hardness for all of the systems were reduced. This indicated that the chemical stability of the drugs would be diminished; therefore, its chemical activity would be increased.
HOMO (eV) | LUMO (eV) | H/L gap | η (eV) | η (eV) | ΔN | |
---|---|---|---|---|---|---|
Ti3C2 surface | 1.31 | 1.32 | 0.01 | 0.005 | 1.32 | |
Diketo (⊥C) | 2.09 | 2.09 | 0.00 | 0.002 | 2.09 | 3.05 |
Ketoenol (⊥C) | 2.006 | 2.009 | 0.003 | 0.002 | 2.08 | 2.80 |
Dienol (⊥Ti2) | 2.10 | 2.11 | 0.01 | 0.005 | 2.11 | 2.68 |
To consider the charge transfer between the Ti3C2 surface and adsorbed drug molecule, we also calculated the ΔN parameter, which shows the fractional electron's charge that transfers from system A to system B:
ΔN = (μB − μA)/(ηB + ηA) | (4) |
Iso surfaces of the most stable structures are also shown in Fig. 5. We can see the charge transfer between the drug and the surface during the adsorption process, which implied the interaction between the adsorbent and the surface in each of the three states.
Fig. 5 Iso surfaces of the most stable structures: (a) Ti3C2@Diketo, (b) Ti3C2@Ketonel and (c) Ti3C2@Dienol. |
To explore the magnetic behavior of Ti3C2 during the adsorption process, we performed spin-polarized density functional theory calculations. The total magnetic moment of the 3 × 3 × 1 supercell of the Ti3C2 surface was calculated to be 20.28 μB per cell. Conversely, the average magnetic moment of each Ti1 atom on the Ti3C2 surface was about 0.49 μB. For Ti2 atoms in the middle of the surface, the calculated value was ∼0.024 μB, which confirms that the magnetic properties of Ti3C2 are primarily due to the Ti1 atoms.35,47,65,66
Fig. 6 shows the total magnetic moment of the most stable Ti3C2@FU systems. As it can be seen, the total magnetic moments varied in the range of 17.87 to 20.32 μB per cell. For the Ti3C2@Ketonel and Ti3C2@Dienol systems, the total magnetic moment did not change significantly before and after drug adsorption. In other words, drug adsorption did not destroy the magnetic properties of these systems. For the Ti3C2@Diketo case on the ⊥C site of the surface, the total magnetic moment value decreased to 17.82 μB per cell. Ti3C2@Diketo has the highest Ead. The oxygen atom of the Diketo bonded with one of the surface titanium atoms during the adsorption process, and caused the decrease of the magnetization. In contrast, for the two former cases, there was not any oxygen atom interactions. So, the bonding state plays a key role in the magnetic moment values. In all cases, after the drug adsorption on the carrier surface, in spite of the Ti1 atoms being around the drug adsorption site, the value of the magnetic moment of the other Ti1 atoms did not change significantly. However, Ti1 atoms that were located around the drug adsorption site on the surface during the adsorption process showed a change in the range of 0.33 to 0.55 μB.
Fig. 6 Total magnetic moment of the Ti1 atom and Ti3C2@FU systems at the most stable adsorption sites. |
The added protons can attack different places of the complex, including nucleophilic and the other sites as well.56,68 Therefore, we investigated the drug release behavior by considering the protons' attack on the drug, as well as the carrier surface around the adsorption site in the most stable structures.
In our modeling, the numbers of hydrogen atoms were added first to the F atom of the drug, but there was no sign of drug release. Therefore, we concluded that to weaken the surface Ti bindings with the drug, they should be attacked with protons gradually until the release features have been started.
The proton number was selected based on the sites expected to impact the release process, and the release energy was calculated per proton. According to the calculations in Section 3.3, it can be expected that the added protons and the carrier surface interact with each other. The charge is transferred from the surface to the protons, weakening (or breaking) the bond between the drug and the surface. After relaxation calculation of the Ti3C2@FU systems (Fig. 7), the calculated drug release energy was as follows: for the Ketoenol form in the ⊥C state, the release energy was obtained at +1.75 eV and the distance of the drug from the surface during the structural optimization was increased by 0.87 Å. The F bond with the Ti3C2 surface weakened and finally broke, which gave rise to the release of Ketoenol from the Ti3C2 surface.
Fig. 7 (a) Ti3C2@Ketonel, (b) Ti3C2@Dienol, and (c) Ti3C2@Diketo before and after drug release. The distance of the drug from the surface and its amount before and after release is shown. |
In Dienol in the ⊥Ti2 state, due to the fact that the O atom is closer to the surface than the F atom, a hydrogen atom was added to the oxygen of the drug and the Ti atoms around it so the following release energy was obtained: +1.47 eV. The distance of the O atom from the Ti3C2 MXene surface increased from 2.32 to 2.90 Å and the F atom from 2.39 to 2.98 Å. Therefore, Dienol starts to be released in the slightly acidic environment of the tumor tissue, which reduces the toxins and side effects of this drug. In the Diketo form in the ⊥C state, the distance of drug from Ti3C2 surface change about 0.08 Å with the change of the environment pH and release energy was obtained +1.57 eV, and it can be said that the release process has been started.
During the release of the drug from the carrier, the added protons approach the Ti3C2 nanosheets, which shows that the hydrogen atoms tend to interact with Ti atoms of the surface so the binding strength of these atoms with the drug could be weakened, respectively.
In order to ensure that the structure of the drug has not changed after release, the bond lengths of the drugs are compared with its the optimal model. The results are represented in row 6 of Tables 2–4. Our results showed that the structures of the drugs were not changed, so the properties of the drugs were preserved.
Our calculations confirm the stability of the Ti3C2@FU systems and exothermic interaction between FU and the Ti3C2 surface. Furthermore, charge flows from the Ti3C2 surface to drugs during the adsorption process. The amount of Ti3C2 surface DOS at the Fermi surface increases slightly after drug adsorption. The total magnetic moment of the 3 × 3 × 1 supercell of the Ti3C2 surface was calculated to be 20.28 μB per cell. For the Ti3C2@Ketonel and Ti3C2@Dienol systems, the total magnetic moment did not change significantly with respect to that of Ti3C2, so drug adsorption did not destroyed the magnetic properties of these systems. However, for the Ti3C2@Diketo case, the total magnetic moment value decreased to 17.82 μB per cell.
Drug release can be achieved in the low pH environment of cancerous cells; the release energy was obtained at +1.75, +1.47 and +1.57 eV for Ketoenol, Dienol and Diketo, respectively. Results of the calculations provide a deep understanding of the interaction mechanism of 2D MXene materials with drugs at the atomistic scale.
Footnote |
† Electronic supplementary information (ESI) available. See DOI: https://doi.org/10.1039/d4ra02399d |
This journal is © The Royal Society of Chemistry 2024 |