Unraveling the Covalency of the Ti Oxidation State in Ti3C2Tx via Electron Energy-loss Spectroscopy

Ti3C2Tx is a 2D MXene material that has been heavily studied due its potential for applications such as wearable electronics, photocatalysis, energy storage, etc.[1, 2] Ti3C2Tx is made by chemically removing the Al from the MAX phase material Ti3AlC2. The etching process results in dangling bonds on the surface which are passivated by -O, -OH, and -F. The effect of different surface terminations on the materials properties has been well studied through various methods including electron energy-loss spectroscopy (EELS), however these studies have primarily focused on either the low-loss region or the C K-edge while the effect on the Ti has been largely ignored. [3, 4] The structure of Ti3C2Tx consists of two C layers sandwiched between three Ti layers as shown in Fig. 1A. For a single Ti3C2Tx delaminated sheet, the organic groups mentioned above attach to the top and bottom Ti layers resulting in two different Ti species: i) surface Ti in octahedral coordination with 3 Tx groups and 3 C (Tis) and ii) middle Ti in octahedral coordination with 6 C (Tim). It has been well established that the energy-loss near edge fine structure is sensitive to the bonding environment and oxidation of the ionized atom, therefore we should be able to distinguish between Tim and Tis. Fig. 1B shows Ti L2,3 edge of three different Ti3C2Tx flakes all with t/λ value of ∼0.16. The EELS data matches XAS data by Yang et al. showing a small split on the L3 peak and slight asymmetry on the L2 peak. The L3/L2 branching ratio, which can be used to determine oxidation state, is also consistent with XAS. Through Bader charge analysis, Yang et al. propose that the Tis oxidation state is 2.35 and Tim is 2.5-2.6. The EELS and XAS data are consistent but there are two issues that need to be addressed regarding the interpretation of the spectral features. The first concerns the Bader charge analysis which indicates that the average Ti oxidation state is non-integer and is between 2 and 3. This implies that there is a mix of oxidation states, probably +2 and +3, that involve some d-orbital occupancy in the ground state. This means that atomic multiplet effects need to be considered in the interpretation. This exacerbates the second concern which is that both the XAS and plan-view EELS data in Fig. 1B are averaging the Tis and Tim signals. This makes identification of the signal from the different Ti sites rather challenging. The traditional method of preparing a TEM grid via drop casting often results in a range of sheet thicknesses oriented along the Z direction in which the three Ti layers are aligned. We therefore prepared a cross-section FIB lamella of Ti3C2Tx sheets that were drop cast on a silicon substrate (Fig. 2) allowing for distinction of the Tis and Tim layers. In this contribution, we will further discuss the Ti oxidation state in Ti3C2Tx flakes. Through atomic resolution EELS analysis of a cross-section lamella, we isolate the Tis and Tim signals to further understand the effect of surface groups on the Ti oxidation state and L2,3 edge. A) Schematic of the structure of MAX phase and Ti3C2Tx MXene. B) EELS of the Ti L2,3 edge of three different Ti3C2Tx sheets of similar thicknesses. Low magnification (left) and high magnification (right) HAADF images of a cross-section lamella of Ti3C2Tx flakes dispersed on a Si substrate.