Recently, well-crystallized and nearly phase-pure Lonsdaleite (hexagonal diamond) has been successfully synthesized by heating graphite under extreme pressure conditions. Despite ongoing investigations into its physical properties, there remains a lack of high-precision interatomic potentials capable of accurately capturing its thermal transport behavior. In this work, we constructed a high performance Neuroevolution potential (NEP) model for hexagonal diamond, trained on a data-set generated via ab initio molecular dynamics (AIMD) simulations. The accuracy of the NEP model was assessed by comparing phonon dispersion curves derived from the NEP and density functional theory (DFT) calculations, revealing excellent agreement. The vibration modes and phonon irreducible representation of hexagonal diamond at the Γ point were further analyzed. The Raman active modes include ${\mathrm{A}}_{1\mathrm{g}}, {\mathrm{E}}_{1\mathrm{g}}$, and ${\mathrm{E}}_{2\mathrm{g}}$, and the infrared active modes include ${\mathrm{A}}_{2\mathrm{u}}$ and ${\mathrm{E}}_{1\mathrm{u}}$. Subsequently, the lattice thermal conductivity of hexagonal diamond was calculated, with a particular focus on the modal contributions. A comprehensive analysis of temperature-dependent thermal conductivity in the range of 300–3000 K shows that the thermal conductivity decreases with increasing temperature. This trend is primarily attributed to increased phonon-phonon anharmonic scattering at elevated temperatures. Moreover, comparative analysis indicated that the thermal conductivity of hexagonal diamond is lower than that of diamond, which can be ascribed to enhanced acoustic-optical phonon scattering in the hexagonal lattice. These findings provide in-depth theoretical insights into the phonon transport characteristics of hexagonal diamond.

Understanding interfacial thermal transport is essential for improving thermal management in high-speed power electronic devices, where the efficient removal of excess heat is a critical challenge. In this study, a machine-learning interatomic potential with near-first-principles accuracy was employed to investigate the interfacial thermal conductance (ITC) between [111]-oriented cubic boron arsenide (cBAs) and [0001]-oriented 4H silicon carbide (4H-SiC), as well as its dependence on uniaxial stress. Among all possible bonding configurations at the cBAs(111)/4H-SiC(0001) interface, the B–C bonded interface was identified as the most energetically favorable. Nonequilibrium molecular-dynamics simulations revealed that, under ambient conditions (300 K and 0 GPa), the ITC of the B–C interface reaches $353\pm 6\mathrm{MW}{\mathrm{m}}^{-2}{\mathrm{K}}^{-1}$, and increases monotonically to $460\pm 3\mathrm{MW}{\mathrm{m}}^{-2}{\mathrm{K}}^{-1}$ under a uniaxial stress of 25 GPa perpendicular to the interface. For comparison, the As–C-bonded interface exhibits a lower ITC, increasing from $233\pm 7$ to $318\pm 6\mathrm{MW}{\mathrm{m}}^{-2}{\mathrm{K}}^{-1}$ over the same stress range. These results demonstrate that proper interfacial bonding and moderate uniaxial stress can significantly enhance thermal transport across the cBAs(111)/4H-SiC(0001) heterointerface, offering valuable insight for thermal design in next-generation power electronics.

We present a first-principles study of the contrasting easy magnetization axes (EMAs) in the layered chromium trihalides ${\mathrm{CrCl}}_3$ and ${\mathrm{CrBr}}_3$, which exhibit in-plane and out-of-plane EMAs, respectively. Using density-functional theory calculations, we show that the EMA is determined by the interplay between spin–orbit coupling-induced magnetocrystalline anisotropy energy (SOC-MAE) and shape magnetic anisotropy energy (shape-MAE) arising from dipole–dipole interactions. While the Cr $d$ orbitals contribute similarly to the SOC-MAE in both compounds, the key difference stems from the halogen $p$ orbitals. In ${\mathrm{CrCl}}_3$, the localized Cl $3p$ orbitals favor spin-flip SOC interactions, particularly between the $(p_x,p_y)$ and $(p_y,p_z)$ channels. These channels contribute with opposite signs–negative and positive, respectively–leading to partial cancellation and a small net SOC-MAE. As a result, the shape-MAE exceeds the SOC-MAE in magnitude, favoring an in-plane EMA. In contrast, ${\mathrm{CrBr}}_3$ features more delocalized Br $4p$ orbitals, enhanced $p\text{−}d$ hybridization, and stronger SOC. This leads to stronger spin-conserving SOC interactions, with dominant contributions from both the $(p_x,p_y)$ and $(p_y,p_z)$ channels. In this case, the positive contribution from the $(p_x,p_y)$ channel outweighs the smaller negative contribution from the $(p_y,p_z)$ channel, resulting in a sizable net SOC-MAE. The SOC-MAE thus surpasses the shape-MAE and stabilizes an out-of-plane EMA. These findings demonstrate that the contrasting magnetic anisotropies in ${\mathrm{CrCl}}_3$ and ${\mathrm{CrBr}}_3$ originate from differences in the spatial distribution, SOC strength, and hybridization of the halogen $p$ orbitals, highlighting the critical role of orbital anisotropy and spin selection rules in governing magnetic behavior in layered semiconductors.

Van der Waals (vdW) magnetic materials have gained significant interest for both their fundamental physics and potential spintronic applications. One vdW magnet of great interest is layered chromium ditelluride ($\mathrm{CrT}{\mathrm{e}}_2$), which hosts intrinsic ferromagnetism above room temperature. The intercalation of Cr atoms in the vdW gaps leads to a versatile family of stable $\mathrm{C}{\mathrm{r}}_{1+\delta }\mathrm{T}{\mathrm{e}}_2$ compounds with highly tunable magnetic properties. However, experimental studies show complicated and sometimes inconsistent magnetic properties when the intercalation concentration (δ) varies. In this work, we employed first-principles calculations of three representative compounds, $\mathrm{CrT}{\mathrm{e}}_2$ ($\delta =0$), $\mathrm{C}{\mathrm{r}}_{1.25}\mathrm{T}{\mathrm{e}}_2$ ($\delta =0.25$), and $\mathrm{C}{\mathrm{r}}_{1.5}\mathrm{T}{\mathrm{e}}_2$, ($\delta =0.5$) to understand the precise role of the Cr intercalation on their structural, magnetic, and electronic properties. Our calculations suggest that $\mathrm{CrT}{\mathrm{e}}_2$ favors an in-plane magnetization, whereas intercalated Cr atoms transition it to easy-axis anisotropy in $\mathrm{C}{\mathrm{r}}_{1.25}\mathrm{T}{\mathrm{e}}_2$, and then to a canted spin configuration for $\mathrm{C}{\mathrm{r}}_{1.5}\mathrm{T}{\mathrm{e}}_2$. The variation of magnetic anisotropy can be understood based on spin-orbit coupling and orbital hybridization, with $p_x-p_y$ and $p_y-p_z$ interactions dominating in $\mathrm{CrT}{\mathrm{e}}_2$ and $\mathrm{C}{\mathrm{r}}_{1.25}\mathrm{T}{\mathrm{e}}_2$, respectively, and both contributions becoming comparably weak in $\mathrm{C}{\mathrm{r}}_{1.5}\mathrm{T}{\mathrm{e}}_2$. Charge doping switches the magnetic anisotropy by modifying the Te orbital hybridization. Using a Wannier function-based tight-binding Hamiltonian and Monte Carlo simulations, we find that intercalation introduces antiferromagnetic interlayer exchange in $\mathrm{C}{\mathrm{r}}_{1.25}\mathrm{T}{\mathrm{e}}_2$ and decreases the Curie temperature. In contrast, more ferromagnetic exchange channels in $\mathrm{C}{\mathrm{r}}_{1.5}\mathrm{T}{\mathrm{e}}_2$ enhance ferromagnetism. Our results highlight the critical role of self-intercalated Cr atoms in determining magnetic order and offer magnetic engineering paths for chromium tellurides for next-generation spintronic applications.

Ternary transition metal (TM) nitrides have gained significant attention in thin film research due to their promising properties for a broad range of applications. Particularly, some of the ternary TM nitrides have been predicted to adopt layered structures that make them interesting for thermoelectric conversion and quantum materials applications. Unfortunately, synthesis of TM ternary nitride films by physical vapor deposition often favors disordered 3D structures rather than the predicted 2D-like layered structure. In this study, we investigate the structural interplay in the Sc-Ta-N ternary system using a combinatorial approach. Combinatorial libraries $\mathrm{S}{\mathrm{c}}_x\mathrm{T}{\mathrm{a}}_{1-x}\mathrm{N}$ are synthesized following a two-step method: First, deposit film precursors by cosputtering and then process the resulting 3D-structured samples with rapid thermal annealing. Synchrotron grazing-incidence wide-angle x-ray scattering on films annealed at $1200{}^{\circ }\mathrm{C}$ for 20 min leads to the nucleation of ${\mathrm{ScTaN}}_2$ layered structure ($P{6}_3/mmc$) near stoichiometry. We find that the layered structure can accommodate large off-stoichiometry in the Ta-rich region ($x<0.5$), facilitated by the alloying with quasi-isostructural ${\mathrm{Ta}}_5{\mathrm{N}}_6$ compound that exists on a composition tie line at $x=0$. While focusing on ${\mathrm{ScTaN}}_2$, we estimate the long-range order parameter in near-stoichiometric films to be 0.86, corresponding to a fraction of Sc/Ta antisites of 7%. Transport measurements on ${\mathrm{ScTaN}}_2$ reveal a nearly temperature-independent high carrier density $\left({10}^{21}\mathrm{c}{\mathrm{m}}^{-3}\right)$, suggesting a heavily doped semiconductor or semimetallic character, consistent with a small positive Seebeck coefficient of +19 µV/K. The carrier mobility at 2 K is relatively small ($9.5\mathrm{c}{\mathrm{m}}^2{\mathrm{V}}^{-1}{\mathrm{s}}^{-1}$) and the residual-resistivity ratio is minor, suggesting that electrical conduction is dominated by defects or disorder. Measured magnetoresistance suggests possible weak antilocalization at 2 K. This paper highlights the interplay between ${\mathrm{ScTaN}}_2$ and ${\mathrm{Ta}}_5{\mathrm{N}}_6$ crystal structures in stabilizing layered materials, emphasizes the importance of cation order/disorder for potential tunable alloys, and suggests that ${\mathrm{ScTaN}}_2$ is a promising platform for exploring electronic properties.

The study of Eu intermetallic compounds has allowed the exploration of valence fluctuations and transitions in $4f$ electron systems. Recently, a ${\mathrm{Eu}}_2{\mathrm{Pt}}_6{\mathrm{Al}}_{15}$ phase synthesized by arc-melting followed by a thermal treatment was reported M. Radzieowski et al. [J. Am. Chem. Soc. 140, 8950 (2018)], which undergoes a transition upon cooling below 45 K that was interpreted as a valence transition from ${\mathrm{Eu}}^{2+}$ to ${\mathrm{Eu}}^{3+}$. In this paper, we present the discovery of another polymorph of ${\mathrm{Eu}}_2{\mathrm{Pt}}_6{\mathrm{Al}}_{15}$ obtained by high temperature solution growth, which presents different physical properties than the arc-melted polycrystalline sample. Despite the similarities in crystal structure and chemical composition, the Eu valence transition is almost fully suppressed in the solution-grown crystals, allowing the moments associated with the ${\mathrm{Eu}}^{2+}$ state to order antiferromagnetically at around 14 K. A detailed analysis of the crystal structure using single crystal x-ray diffraction reveals that, although the solution grown crystals are built from the same constituent layers as the arc-melted samples, these layers present a different stacking. The effect of different thermal treatments is also studied. Different anneal procedures did not result in significant changes in the intrinsic properties, and only by arc-melting and quenching the crystals we were able to convert them into the previously reported polymorph.

Superconducting thin films and layered crystals display a Nernst signal generated by short-lived Cooper pairs above their critical temperature. Several experimental studies have broadly verified the standard theory invoking Gaussian fluctuations of a two-dimensional (2D) superconducting order parameter. Here, we present a study of the Nernst effect in granular NbN thin films with a thickness varying from 4 to 30 nm, exceeding the short superconducting coherence length and putting the system in the three-dimensional (3D) limit. We find that the Nernst conductivity decreases linearly with reduced temperature (${\alpha }_{xy}\propto \frac{T-T_c}{T_c}$), but the amplitude of ${\alpha }_{xy}$ scales with thickness. While the temperature dependence corresponds to what is expected in a 2D picture, scaling with thickness corresponds to a 3D picture. We argue that this behavior indicates a $2+1\mathrm{D}$ situation, in which the relevant coherence length along the thickness of the film has no temperature dependence. We find no visible discontinuity in the temperature dependence of the Nernst conductivity across ${\mathrm{T}}_c$. Explaining how the response of the superconducting vortices evolves to the one above the critical temperature of short-lived Cooper pairs emerges as a challenge to the theory.

Transition-metal (TM) atom-functionalized nanomaterials are promising candidates for hydrogen storage due to their ability to adsorb multiple hydrogen molecules through Kubas interactions. However, achieving efficient hydrogen desorption at ambient conditions remains a critical challenge for practical use. Here, we present a novel approach to modulate the desorption temperature of hydrogen in TM-intercalated bilayer graphene (BLG) using external mechanical forces. By employing first-principles density functional theory and thermodynamic occupancy probability calculations, we demonstrate that adjusting the interlayer distance allows for precise control over the interaction energy of ${\mathrm{H}}_2$, thereby facilitating its desorption at ambient conditions. Complete hydrogen desorption occurs when the interlayer distance is reduced below $4.7$, $5.3$, and $5.1\AA$ for Sc-, Ti-, and V-intercalated BLG, respectively. Our findings suggest that external mechanical forces can effectively bring hydrogen occupancy to zero by minimizing charge transfer from the TM $d$-orbitals to ${\mathrm{H}}_2$ antibonding orbitals. Notably, while the total charge transferred from the TM atoms remains nearly constant at varying interlayer distances, its redistribution between the graphene layers and ${\mathrm{H}}_2$ fine-tunes the interaction strength. This approach can be extended to large interlayer distances, as supported by recent experiments on graphene oxide membranes [ACS Nano 12, 9309 (2018)]. Furthermore, recent experimental advances in noble gas and alkali-metal intercalation in BLG highlight the potential of this approach to overcome the long-standing challenge of high desorption temperatures in TM-functionalized layered nanomaterials.

Quantum coherence in superconducting circuits has increased steadily over the last decades because of a growing understanding of the various loss mechanisms. Recently, tantalum (Ta) emerged as a promising material to address microscopic sources of loss found on niobium (Nb) or aluminum (Al) surfaces. However, the effects of film and interface microstructure on low-temperature microwave loss are still not well understood. Here, we present a systematic study of the structural and electrical properties of Ta and Nb films sputtered on c-plane sapphire at varying growth temperatures. As growth temperature was increased, our results show that the onset of epitaxial growth of $\alpha$-phase Ta correlates with lower Ta surface roughness, higher critical temperature, and higher residual resistivity ratio, but surprisingly also correlates with a significant increase in loss at microwave frequency. Notably, this high level of loss is not observed in Nb films prepared in the same way and having very similar structure. By experimentally controlling the surface on which the Ta film is nucleated, we determine that the source of loss was only present in samples having an epitaxial Ta/sapphire interface and show that it was apparently mitigated by either growing a thin, epitaxial Nb interlayer between the Ta film and the substrate or by intentionally treating, and effectively damaging, the sapphire surface with an in situ argon plasma before Ta growth. In addition to elucidating this interfacial microwave loss, this work provides adequate process details to aid reproducible growth of low-loss Ta films across fabrication facilities.

Understanding Short-Range Order (SRO) in alloys is essential in atomistic materials science, as it reveals the underlying mechanisms that drive their mechanical behavior. A precise description of local atomic environments plays a central role in characterizing and simulating complex alloys at the atomic scale. This study proposes a complementary method to the traditional SRO characterization methods by leveraging high-dimensional representations of atomic structures. These representations, commonly used as internal feature vectors, known as descriptors, in machine learning potentials, have matured significantly in recent years. We investigate how local chemical, geometric, and combined atomic effects influence the mechanical properties of medium- and high-entropy alloys (HEAs), and propose a method to identify regions with similar structural and chemical features in the multidimensional space defined by these invariant descriptors. Using the widely adopted bispectrum projection, we capture the short-range order factors—geometry and chemistry—and apply various techniques to explore the structure of the resulting high-dimensional space: (i) statistical distance analysis, which reveals the non-Gaussian nature of descriptor distribution associated with studied HEA systems, (ii) traditional Bayesian Gaussian mixture model, and (iii) a novel iterative scheme combining both approaches and a descriptor embedding. We demonstrate the efficiency of both methods on two alloys, NiCoCr (a medium-entropy alloy) and CoCuFeNiPd (a high-entropy alloy), making a priori predictions of zones of interest, called Gaussian ordered states. Finally, we show the influence of short-range order under tensile stress, opening new possibilities for predicting mechanical behavior based on these Gaussian ordered states.

Using high-pressure single-crystal x-ray diffraction combined with thermodynamic measurements and density-functional calculations, we uncover the microscopic magnetic model of the mineral brochantite, ${\mathrm{Cu}}_4{\mathrm{SO}}_4{\left(\mathrm{OH}\right)}_6$, and its evolution upon compression. The formation of antiferromagnetic spin chains with the effective intrachain coupling of $J\simeq 100\mathrm{K}$ is attributed to the occurrence of longer Cu–Cu distances and larger Cu–O–Cu bond angles between the structural chains within the layers of the brochantite structure. These zigzag spin chains are additionally stabilized by ferromagnetic couplings $J_2$ between second neighbors and moderately frustrated by several antiferromagnetic couplings that manifest themselves in the reduced Néel temperature of the material. Pressure tuning of the brochantite structure keeps its monoclinic symmetry unchanged and leads to the growth of antiferromagnetic $J$ with the rate of 3.2 K/GPa, although this trend is primarily caused by the enhanced ferromagnetic couplings $J_2$. Our results show that the nature of magnetic couplings in brochantite and in other layered ${\mathrm{Cu}}^{2+}$ minerals is controlled by the size of the lattice translation along their structural chains and by the extent of the layer buckling.

Superconducting radio-frequency cavities, critical components of modern particle accelerators and quantum computing hardware, fundamentally rely on the surface properties of niobium. However, native oxide formation and impurity uptake in the near-surface region can degrade superconducting performance and increase radio-frequency losses during operation. In this study, we present a systematic in situ x-ray reflectivity investigation of (110) niobium single-crystal surfaces with and without an atomic layer deposited ${\text{Al}}_2{\text{O}}_3$ capping layer under ultrahigh-vacuum conditions up to $650{}^{\circ }\mathrm{C}$. Our results reveal a temperature-dependent reduction of the native niobium oxide layers in both capped and uncapped samples, with similarities in the overall behavior but clear differences in the reduction pathways. The ${\text{Al}}_2{\text{O}}_3$ capping layer modifies the reduction process, prevents oxide regrowth upon air exposure, and protects the niobium surface against impurity uptake during thermal treatment. These findings demonstrate that ${\text{Al}}_2{\text{O}}_3$ capping is an effective strategy to suppress native oxide formation during thermal cycling, offering clear benefits for the performance and operational lifetime of superconducting radio-frequency cavities and related quantum computing technologies.

Approximately 25-nm-thick $\mathrm{M}{\mathrm{n}}_{4-x}\mathrm{P}{\mathrm{d}}_x\mathrm{N}$ epitaxial thin films were deposited on ${\mathrm{SrTiO}}_3$(001) by molecular beam epitaxy, and the existence of a magnetization compensation composition $\left(x_{\mathrm{MC}}\right)$ of approximately 0.15 at room temperature was demonstrated. The substitution of Mn by Pd at the corner sites was confirmed by x-ray absorption near-edge structure measurements. The magnetic field-dependent x-ray magnetic circular dichroism (XMCD) spectra revealed that the magnetic structure of $\mathrm{M}{\mathrm{n}}_{4-x}\mathrm{P}{\mathrm{d}}_x\mathrm{N}$ was significantly different below and above $x_{\mathrm{MC}}$. This result is different from previously reported Ni- and Co-doped ${\mathrm{Mn}}_4\mathrm{N}$ epitaxial thin films. The magnetic structure of $\mathrm{M}{\mathrm{n}}_{4-x}\mathrm{P}{\mathrm{d}}_x\mathrm{N}$ was assessed based on various magnetic properties, such as magnetization curves, anomalous Hall effect loops, magneto-optical Kerr effect loops, and XMCD spectra.

Highly stable $n$-type carbon nanotubes (CNTs) are essential for the development of high-performance CNT-based field-effect transistors. However, current $n$-type CNTs predominantly rely on the contact of a low-work-function metal electrode, but they are prone to oxidation and consequently exhibit increased contact resistance. Therefore, in this work we proposed a feasible strategy for high-throughput screening of cubic-phase nanowire infillings to achieve $n$-type single-walled CNTs (SWCNTs) by first-principles calculations. Firstly, two key nanowire descriptors (low work function and large internal band gap) were introduced which not only maintain the original transport properties but also realize highly stable electron injection of SWCNTs. Moreover, taking the cubic space group $Pm\overline{3}m$ as an example, ten promising cubic-phase nanowire candidates were identified through descriptor-based high-throughput screening. Meanwhile, we further verified the $n$-type SWCNTs with nanowires infilling by calculating the charge transfer, the difference charge density, and the electronic structures. Crucially, the transferred charge is positively correlated with the work-function difference between the nanowires and the SWCNT host. Our findings establish a practical strategy for selecting cubic-phase nanowire fillers to enable low-power CNT-based electronic devices.

Iron garnets (IGs) are materials with unique radio-frequency (rf) properties. They are already being used as GHz oscillators, magneto-optical filters, and circulators. Recently used growth methods, including the rf sputtering technique, now allow the fabrication of nm-thick films and multilayers, compatible with lithography techniques for device fabrication. In some cases (two-step growth), IGs fabricated using physical vapor deposition techniques are crystallized following deposition at room temperature. This work focuses on an in situ structural characterization of IGs during the annealing process to determine the crystallization mechanism and dynamics. It is found that IG crystallization leads to heteroepitaxy but the crystallized thickness does not grow linearly with time and competes with diffusion from the substrate. Modeling the crystallization dynamics demonstrates a two-dimensional solid phase epitaxy (SPE) mechanism using a Johnson-Mehl-Avrami-Kolmogorov approach. This SPE is controlled by interface nucleation, and differs from the expected one-dimensional SPE from the interface to the surface considered before. The determined crystallization and diffusion activation energies are in agreement with the ones found in the literature. The obtained results allow us to find the best compromise between annealing time and temperature needed for IG crystallization in order to minimize the effect of substrate diffusion, especially Gd magnetic atoms, which are detrimental to the IG rf quality factor.

We report the characteristic domain wall (DW) patterns observed in rare-metal-free plastic/ferroelectric films of $\left[\mathrm{MDABCO}\right]\left[{\mathrm{PF}}_6\right]$, where ${\left[\mathrm{MDABCO}\right]}^{+}$ is a cage-like molecular ion with a protrusion. In its paraelectric plastic phase, $\left[\mathrm{MDABCO}\right]\left[{\mathrm{PF}}_6\right]$ adopts a cubic lattice. However, in the ferroelectric phase, the lattice undergoes significant distortion into a right prism (pseudocube) with a kite-shaped quadrilateral base. This distortion results from the alignment of the ${\left[\mathrm{MDABCO}\right]}^{+}$ protrusions parallel to the base planes of the prism lattice, with spontaneous polarization along the $\langle 110\rangle$ direction of the pseudocubic $\left(pc\right)$ lattice. This leads to multiaxial ferroelectricity with six possible polarization directions. We demonstrate that two distinct types of thin films, each with unique crystal orientations, can be fabricated depending on the film growth method, resulting in different DW patterns. Films grown from a confined solution layer between top and bottom substrates exhibit the pseudocubic ${\left\{100\right\}}_{pc}$ plane nearly parallel to the substrate, producing a DW pattern corresponding to two possible in-plane polarization orientations. In contrast, blade-coated films show the diagonal ${\left\{110\right\}}_{pc}$ plane of the pseudocubic lattice aligned parallel to the substrate surface. These films predominantly feature polarized domains with in-plane polarization components, while thin, streak-like domains with out-of-plane polarization components intersect at a consistent angle of approximately 54°. Furthermore, blade-coated films allow for polarization switching across the entire film surface, generating out-of-plane polarization components. These findings advance our understanding of polarization domain control in plastic/ferroelectrics and highlight their potential for use in printable ferroelectric-based devices.

In this study, we constructed an efficient framework based on phonon self-consistent theory and unified transport theory combined with machine learning potential, and systematically reveal some interesting lattice dynamics in simple transition metal halide crystals ${\mathrm{CaXF}}_6$ (X = Ti, Zr, Hf) with cubic structures. Although ${\mathrm{CaTiF}}_6$ exhibits a lighter average atomic mass, its lattice thermal conductivity is lower than that of ${\mathrm{CaZrF}}_6$ and ${\mathrm{CaHfF}}_6$. Such behavior is in contrast to the conventional Keyes theory that typically associates light atomic systems with higher thermal conductivity. Further analysis indicates that this deviation from the expected mass-dependent behavior stems from the stronger anharmonic scattering induced by the weaker $s\text{−}d$ orbital bonding in the Ti-F bond. Notably, the rotating mode vibrations in the soft phonon dispersion curves of ${\mathrm{CaTiF}}_6$ exhibit a double-potential-well, and lead to more pronounced quartic residuals (quartic anharmonicity). As a result, the strong four-phonon scattering reduces the room-temperature thermal conductivity of ${\mathrm{CaTiF}}_6$ by 59%, a suppression markedly stronger than in the Zr and Hf systems. Moreover, the stronger anharmonic scattering (including three- and four-phonon processes) induces substantial coherent phonons in ${\mathrm{CaTiF}}_6$, resulting in the wave-like coherent thermal conductivity even exceeding the particle-like propagate thermal conductivity at high temperatures $(>700 \mathrm{K})$. Our results not only deepen the fundamental understanding of the lattice dynamics in transition metal halides but also provide new perspectives for designing thermal management devices with desirable thermal transport properties.