Cl-Terminated MXenes: Synthesis & Properties
- Cl-terminated MXenes are 2D transition metal carbides/nitrides featuring chloride groups that replace the A-layer in MAX phases.
- Synthesis methods include Lewis acidic molten salts, electrochemical etching, and vapor-phase chlorosilane processes, confirmed by STEM, XPS, and DFT studies.
- These MXenes exhibit robust thermal stability, metallic conductivity, and serve as precursors for advanced energy storage, superconductivity, and heterostructure applications.
Searching arXiv for the cited Cl-terminated MXene papers to ground the article in current indexed records. arxiv_search({"query":"Cl-terminated MXene Ti3C2Cl2 (Lu et al., 2019, Shen et al., 2021, Li et al., 2020, Cheng et al., 2023, Wang et al., 2022)", "max_results": 10, "sort_by": "relevance"}) Search results reviewed; fetching a broader query to ensure coverage of synthesis, structure, oxidation, superconductivity, and exfoliation work on Cl-terminated MXenes. arxiv_search({"query":"\"Cl-terminated MXene\" OR Ti3C2Cl2 OR Ti2CCl2 OR Nb2CClx", "max_results": 10, "sort_by": "lastUpdatedDate"}) Cl-terminated MXene denotes a class of two-dimensional transition-metal carbides and nitrides whose exposed metal surfaces are predominantly capped by chloride ligands after removal of the A layer from a parent MAX phase. A convenient formula is , where is an early transition metal, is C or N, and “” denotes two chloride terminations per unit. Experimentally established members include , , and , with the first demonstrations of fully saturated Ti-based examples emerging from Lewis acidic molten salts and later extensions using molten-salt-assisted electrochemical etching, chlorosilane vapor etching, and CdCl-based routes (Lu et al., 2019, Li et al., 2019, Shen et al., 2021, Wang et al., 2022, Wang et al., 14 Sep 2025). The topic is defined by unusually ordered halide termination layers, fluorine-free processing routes, and a surface chemistry that modifies interlayer spacing, electronic structure, thermal stability, and downstream reactivity.
1. Definition, compositional scope, and identification
Cl-terminated MXenes are most explicitly represented by and 0, for which fully saturated, single-component Cl termination was established by atomic-resolution HAADF-STEM imaging, image simulations, lattice-resolved EDX mapping, EELS, XRD, and XPS (Lu et al., 2019, Li et al., 2019). In Ti-based systems synthesized from Lewis acidic melts, the reported Ti:Cl ratios are very close to 1 for Ti2C and 3 for Ti4C5, consistent with 6 and 7 (Lu et al., 2019). In molten-salt-assisted electrochemical etching of Ti8AlC9, the dominant product is likewise described as Ti0C1Cl2, with XPS giving Ti:Cl 3 and EDS giving Ti/Cl atomic ratio 4 (Shen et al., 2021). Vapor-phase chlorosilane etching explicitly produced 5 and 6, while CdCl7 etching produced superconducting 8, modeled in DFT as 9 with 0 (Wang et al., 14 Sep 2025, Wang et al., 2022).
Identification of Cl termination relies on convergent structural and spectroscopic signatures. In Ti1C2Cl3, XRD shows the Ti4AlC5 (002) peak at 6 shifting to 7 after MS-E-etching, increasing the 8-lattice parameter from 9 Å to 0 Å (Shen et al., 2021). In ZnCl1-derived Ti2C3Cl4, the (0002) peak appears at 5, with 6 Å (Li et al., 2019). XPS resolves Ti–Cl bonding directly: Ti–Cl 7 at 8 eV in MS-E-etched Ti9C0Tx, Ti–Cl 1 at 2 eV and Cl 3 at 4 eV in Si-coated Ti5CCl6, and Ti–Cl 7 at 8 eV with Cl 2p peaks at 9 eV and 0 eV in ZnCl1-derived Ti2C3Cl4 (Shen et al., 2021, Wang et al., 14 Sep 2025, Li et al., 2019).
A frequent misconception is that any detected chlorine in MXene automatically proves an exclusive basal-plane termination. Atom probe tomography of wet-etched Ti5C6 detected Cl at 7 at% within the MXene region of interest, but the study states that APT “does not resolve whether Cl is bound as a surface termination (T = Cl) on basal planes or edges, intercalated between layers, or present as residual species associated with byproducts” (Krämer et al., 2023). That caveat applies to wet-chemical Ti8C9, not to the ordered Ti0CCl1 and Ti2C3Cl4 structures established by STEM and image simulation in molten-salt-derived systems.
2. Synthetic routes and reaction frameworks
Several distinct synthetic strategies generate Cl-terminated MXenes, but they share a common logic: selective A-layer extraction coupled to chloride delivery at exposed metal sites. In the original ZnCl5-based replacement route, Al-MAX powders were mixed with ZnCl6 and heated at 7 for 8 h under Ar; an Al-MAX:ZnCl9 ratio of 0 yielded Zn-MAX phases, whereas excess ZnCl1 at 2 enabled exfoliation to Ti3C4Cl5 and Ti6CCl7 (Li et al., 2019). In the related Ti8AlC/Ti9AlC0 Lewis acidic melt route, powders were mixed with ZnCl1 and heated at 2 for 3 h, then cleansed with aqueous HCl and washed with deionized water, producing fully saturated Ti4CCl5 and Ti6C7Cl8 (Lu et al., 2019). Molten-salt-assisted electrochemical etching used a LiCl–KCl eutectic (1:1 wt%) at 9, a typical applied cell voltage of 00 V, and a total etching time of 01 h, with the process constrained so that the anode potential remained lower than 02 V vs Ag/AgCl to avoid Ti dissolution; the reported overall etching yield was 03 (Shen et al., 2021). Chlorosilane vapor etching used SiCl04 as a Lewis acid, with a 05 Al-MAX:SiCl06 ratio favoring formation of Cl-terminated MXene plus amorphous Si, while CdCl07 etching of Nb08AlC used a 09 CdCl10:Nb11AlC ratio, ball milling for 12 h, then heat treatment at 13 for 14 h followed by 15 for 16 h (Wang et al., 14 Sep 2025, Wang et al., 2022).
| Route | Representative products | Key conditions/features |
|---|---|---|
| Lewis acidic ZnCl17 melt | Ti18C19Cl20, Ti21CCl22 | Al-MAX:ZnCl23 = 24, 25, 26 h, Ar (Li et al., 2019) |
| Lewis acidic melt from Ti27AlC/Ti28AlC29 | Ti30CCl31, Ti32C33Cl34 | ZnCl35, 36, 37 h; post-HCl wash (Lu et al., 2019) |
| MS-E-etching in LiCl–KCl | Ti38C39Cl40 | LiCl–KCl (1:1 wt%), 41, 42 V, 43 h, yield 44 (Shen et al., 2021) |
| SiCl45 vapor etching | Ti46CCl47, Ti48C49Cl50 | Al-MAX:SiCl51 = 52 favors MXene + Si(s) (Wang et al., 14 Sep 2025) |
| CdCl53 molten salts | Nb54CCl55 | 56/8 h, then 57/36 h, Ar with 5% H58 (Wang et al., 2022) |
The reaction chemistry is explicit in several cases. For chlorosilane etching, the generic relation is
59
with Ti-containing examples
60
and
61
A redox-potential model was proposed to rationalize why Ti-containing MAX phases can yield Cl-terminated MXenes under SiCl62, whereas Nb/Ta/Cr/V systems favor Si-substituted MAX phases under comparable conditions (Wang et al., 14 Sep 2025).
For MS-E-etching, the reported overall scheme is
63
with AlCl64 volatility providing the thermochemical driving force for Al removal and Cl65 supplying the termination chemistry (Shen et al., 2021). In ZnCl66 melts, the exfoliation step was written as
67
which the authors described as conceptually analogous to HF etching but mediated by Zn68 extraction and Cl69 termination rather than H70 extraction and F71 termination (Li et al., 2019).
3. Atomic structure, termination registry, and interlayer geometry
Atomic-resolution microscopy and first-principles calculations show that Cl occupies ordered surface sites rather than a random adlayer. In Ti72CCl73 and Ti74C75Cl76, plan-view STEM images display honeycomb Ti arrangements, and contrast analysis indicates that Cl preferentially occupies fcc hollow sites, not the hcp hollow or atop sites (Lu et al., 2019). HAADF-STEM image simulations with Cl on fcc sites reproduce the experimental intensity distributions for both Ti77CCl78 and Ti79C80Cl81, and cross-sectional imaging from both 82 and 83 orientations confirms fcc-site occupancy (Lu et al., 2019). DFT independently found that fcc hollow sites are favored over hcp or mixed configurations for both Ti84C and Ti85C86 (Lu et al., 2019).
A defining structural feature of Cl termination is its large vertical offset from the outer Ti layer. Experimentally, the vertical separation between the Cl termination layer and the nearest Ti layer is 87 Å for Ti88C and 89 Å for Ti90C91, corresponding to Ti–Cl bond lengths of 92 Å and 93 Å when 94 Å is used (Lu et al., 2019). DFT gives Ti–Cl vertical separation 95 Å for Ti96C and 97 Å for Ti98C99, with Ti–Cl bond length 00 Å for both (Lu et al., 2019). Prior work on pristine O/F-terminated Ti01C02 gives Ti–O/F 03 Å and termination–Ti layer separation 04 Å, so Cl sits significantly farther from the Ti surface (Lu et al., 2019). This suggests increased local surface corrugation and, in multilayer stacks, a larger average interlayer spacing that can benefit ion accessibility.
The crystallographic response is consistent across synthesis routes. In MS-E-etched Ti05C06Cl07, the (002) peak shift from 08 to 09 corresponds to 10 Å (Shen et al., 2021). In ZnCl11-derived Ti12C13Cl14, the (0002) peak at 15 corresponds to 16 Å, close to the DFT value of 17 Å (Li et al., 2019). In the SiCl18-derived Ti19CCl20 composite, the (002) reflection appears at 21, giving 22 Å when Cu K23 radiation is used (Wang et al., 14 Sep 2025). Cl-terminated sheets are also reported to be laterally translated relative to each other such that fcc sites of adjacent layers do not stack directly atop one another, unlike common O-terminated multilayers (Lu et al., 2019).
Spectroscopy supports the structural model. In situ EELS from Ti24C25Cl26 shows the Cl-27 edge at 28 eV and Ti-29 at 30 eV (Lu et al., 2019). In MS-E-etched Ti31C32Cl33, 34C NMR shifts from 35 ppm in Ti36AlC37 to 38 ppm in Ti39C40Tx (Shen et al., 2021). XPS shows no Al in the SiCl41-derived Ti42CCl43 composite and no F 1s signal above the detection limit in Nb44CCl45 (Wang et al., 14 Sep 2025, Wang et al., 2022).
4. Electronic structure, bonding strength, and thermal behavior
Cl-terminated Ti-based MXenes are metallic. DFT calculations on Ti46CCl47 and Ti48C49Cl50 show finite density of states at the Fermi level, with the Fermi level dominated by Ti 51-orbitals and strong hybridization between Ti 52 and C/Cl 53 states between 54 and 55 eV (Lu et al., 2019). Ti56C57Cl58 has more Ti and C bands, and a small gap present around 59 eV in Ti60CCl61 is closed in Ti62C63Cl64 (Lu et al., 2019). Formation energies quantify strong binding:
65
with 66 eV per formula unit for Ti67CCl68 and 69 eV per formula unit for Ti70C71Cl72 (Lu et al., 2019). For comparison, Ti73C74O75 was reported as 76 eV/f.u. and Ti77C78N79 as 80 eV/f.u., so Cl binds more strongly than N but less strongly than O (Lu et al., 2019).
Thermal robustness is one of the most distinctive features of Cl termination. In situ STEM/EELS heating experiments on Ti81C82Cl83 showed that Cl terminations remain largely intact up to 84 (Lu et al., 2019). Above 85, Cl gradually desorbs, most markedly at thin multilayer edges, consistent with diffusion-limited desorption; at 86 only residual Cl remains, the lamination persists, and nanoscale voids appear, interpreted as microstructural damage due to local gas evolution as Cl leaves (Lu et al., 2019). The reported stability hierarchy is O 87 Cl 88 F (Lu et al., 2019). A separate molten-salt reduction study further showed that Ti89C90Cl91 annealed in eutectic LiCl–KCl at 92 under inert atmosphere without Na exhibited no structural or compositional change, supporting the conclusion that CdCl93-derived Cl terminations are stable at 94 in LiCl–KCl without a reductant (Ding et al., 8 Jul 2025).
Cl also acts as an electronically consequential surface group. In Ti95C96Cl97, the authors of the Na-mediated reduction study state that “−Cl terminations, with high electronegativity (98 vs. 99), extract electron density from outer Ti sites, lowering near-Fermi electron density and increasing work function” (Ding et al., 8 Jul 2025). In that work, the work function decreased from 00 eV for Ti01C02Cl03 to 04 eV after Na-mediated dechlorination, with carrier concentration increasing by 05, mobility by 06, and conductivity by 07 under the optimal reduction condition (Ding et al., 8 Jul 2025). These values do not characterize Cl-terminated MXene itself as a poor conductor in every context; rather, they isolate the electronic role of Cl as an electron-withdrawing terminal in that specific Ti08C09Cl10 system.
5. Functional properties and application-specific behavior
The large Cl–Ti separation and preserved metallicity make Cl-terminated MXenes relevant to electrochemical transport, but the performance outcome depends strongly on the device mechanism. For Ti-based systems, the unusually large Ti–Cl vertical separation “suggests enhanced interlayer spacing in multilayers and improved ion accessibility,” and theory predicts “a very large voltage window (3.5–4 V) for exclusively Cl-terminated Ti11C” (Lu et al., 2019). In MS-E-etched Ti12C13Cl14, Cl is the direct intermediate in one-pot conversion to O- and S-containing terminations: after etching, a pellet of Li15O or Li16S is added directly into the same molten LiCl–KCl bath, and XPS shows disappearance of Ti–Cl peaks with emergence of Ti–O or Ti–S/Ti–O components (Shen et al., 2021). The O-terminated product exhibited capacitance of 17 and 18 F/g at 19 and 20 A/g, with retention above 21 after 22 cycles at 23 A/g (Shen et al., 2021). Here, Cl termination functions as a controllable precursor state rather than the final electrochemically optimized state.
By contrast, in aqueous Zn-ion batteries, Ti24C25Cl26 showed no distinct redox peaks and a quasi-linear charge/discharge profile, with discharge capacity 27 mAh g28 at 29 A g30 (Li et al., 2020). Ti31C32(OF) behaved similarly, while Br- and I-containing MXenes exhibited distinct discharge platforms and higher capacities, including 33 mAh g34 for Ti35C36Br37 and 38 mAh g39 for Ti40C41I42 (Li et al., 2020). The paper attributes the lower conversion activity of Cl to stronger Ti–Cl bonding (43 eV) and smaller interlayer spacing (44 Å) relative to Br and I (Li et al., 2020). This corrects another common overgeneralization: Cl termination expands the MXene property space, but it is not the universally best terminal for every electrochemical duty cycle.
Cl termination also enables functionalities outside conventional energy storage. Nb45CCl46 is superconducting, with Meissner onset and zero-resistivity-derived 47 K, 48 T, 49–50 T, coherence length 51 Å, penetration depth 52 Å, and Ginzburg–Landau parameter 53, establishing Nb54CCl55 as a type-II superconductor (Wang et al., 2022). DFT gave electron–phonon coupling constant 56 and 57 K for 58, while the F-terminated analogue was dynamically unstable with imaginary phonons at the M point (Wang et al., 2022).
Cl-terminated Ti59C60 can also serve as a chemically addressable precursor for topological reconstruction. A “simple topological reaction between chlorine-terminated MXenes and selected metals” produced metal-bonded atomic layers; in the Al case, interlayer spacing shrank from 61 Å in pristine Ti62C63Cl64 to 65 Å in Al–Ti66C67Cl68, sheet resistance dropped from 69 70 to 71 72, and the total EMI shielding effectiveness reached 73 dB at a thickness of 74 75m (Cheng et al., 2023). Here Cl acts as a removable handle: Al–Cl bond energy (76 kJ/mol) exceeds Ti–Cl (77 kJ/mol), AlCl78 sublimates at 79, and residual Al bridges adjacent Ti80C81 slabs (Cheng et al., 2023).
Processing-specific properties have also been quantified for few-layer Cl-terminated Ti82C83. Gaseous scissor-mediated electrochemical exfoliation yielded few-layer Ti84C85Cl86 nanoflakes with ultrahigh yield of 87, average thickness 88 nm, work function 89 eV, water contact angle 90, and stable colloids for 91 weeks (Fan et al., 2024). In the same study, Cl-terminated MXene lubricants improved tribovoltaic nanogenerator output from a dry short-circuit current of 92 nA to the 93A range, though Ti94C95Br96 outperformed Ti97C98Cl99 because of lower work function and greater hydrophobicity (Fan et al., 2024).
6. Stability in real environments, ambiguities, and open questions
The strongest evidence for Cl robustness concerns vacuum and inert-environment thermal stability, not comprehensive environmental stability. In situ STEM/EELS demonstrated stability up to 00 under the electron beam and microscope vacuum (Lu et al., 2019), and CdCl01-derived Ti02C03Cl04 remained unchanged at 05 in LiCl–KCl without Na (Ding et al., 8 Jul 2025). However, several studies explicitly note that ambient and aqueous stability remain unresolved. The Ti-based STEM/DFT study states that “stability of Cl terminations in air, water, and electrolytes, and their resistance to hydrolysis or halide exchange, remain to be characterized,” while the chlorosilane study states that systematic studies under humidity, thermal cycling, solvents, or electrochemical cycling were not reported (Lu et al., 2019, Wang et al., 14 Sep 2025).
Wet-chemical systems illustrate why this distinction matters. Atom probe tomography of HCl–LiF-etched Ti06C07 found Cl at 08 at% in as-synthesized MXene and 09 at% in oxidized TiO10 nanowires, with the authors emphasizing that halogens and alkalis are “inevitable” in wet synthesis and remain incorporated during oxidation (Krämer et al., 2023). Oxidation produced TiO11 nanowires enriched in Li and Na, while Al decreased from 12 at% to 13 appm (Krämer et al., 2023). This shows that chlorine can persist through degradation pathways, but it does not by itself define a pure Cl-terminated MXene state.
Scalability and compositional generality are likewise only partly resolved. ZnCl14 and LiCl–KCl routes clearly produce Ti15CCl16 and Ti17C18Cl19, and MS-E-etching was also extended to Ti20SiC21 in supporting information (Li et al., 2019, Shen et al., 2021). Chlorosilane etching established a redox-potential framework in which Ti-based MAX phases can be driven to Cl-terminated MXenes, whereas MAX phases with 22 favor Si substitution under comparable conditions (Wang et al., 14 Sep 2025). This suggests that Cl-termination is not controlled solely by chloride availability; the accessible MXene chemistry is also gated by the relative positions of 23, 24, and 25.
Several mechanistic gaps are explicitly identified across the literature. No NEB diffusion/desorption barriers or Bader charges were reported in the Ti26CCl27/Ti28C29Cl30 study (Lu et al., 2019). Work functions were not reported there either (Lu et al., 2019). The chlorosilane work did not quantify conductivity, long-term stability, colloid dispersion, or ease of delamination (Wang et al., 14 Sep 2025). The metal-bonded Ti31C32Cl33 study did not report DFT or XPS core-level analysis for the bonded interfaces (Cheng et al., 2023). A plausible implication is that Cl-terminated MXenes are already structurally well established, but the predictive structure–property map linking termination density, work function, intercalation energetics, ambient stability, and reaction selectivity remains incomplete.
In aggregate, Cl-terminated MXenes constitute a distinct termination-defined branch of MXene chemistry: fluorine-free in their canonical syntheses, frequently highly ordered, metallic in Ti-based cases, thermally robust up to about 34 in vacuum, and sufficiently versatile to support termination exchange, superconductivity, metal-bonded heterostructures, few-layer colloids, and composite formation with in-situ amorphous Si (Lu et al., 2019, Shen et al., 2021, Wang et al., 2022, Cheng et al., 2023, Fan et al., 2024, Wang et al., 14 Sep 2025). Their central scientific significance lies less in a single benchmark property than in the demonstration that chloride can function as a stable, ordered, and synthetically addressable terminal, thereby expanding the accessible MXene termination space beyond the mixed F/O/OH surfaces typical of aqueous HF-derived processing.
