--- title: Periodic Lattice Distortion (PLD) url: https://www.emergentmind.com/topics/periodic-lattice-distortion-pld type: topic --- # Periodic Lattice Distortion (PLD) Searching arXiv for recent and foundational papers on periodic lattice distortion to ground the article in published work. Periodic lattice distortion (PLD) denotes a static or quasi-static modulation of atomic positions in which each nucleus is displaced from its high-symmetry site by a periodic field. In CDW materials, PLD is the structural counterpart of the charge modulation: conduction electrons condense at wavevector(s) $\mathbf{q}$, and the lattice follows by displacing its nuclei so as to lower the combined electronic and elastic free energy. A general representation is $$ \mathbf{u}(\mathbf{r})=\sum_j A_j\cos(\mathbf{q}_j\cdot \mathbf{r}+\phi_j), $$ or, in a complex three-$q$ form, $$ \mathbf{u}(\mathbf{r})=\sum_{j=1}^3\left[A_j\hat e_j e^{i\mathbf{q}_j\cdot\mathbf{r}}+\text{c.c.}\right]. $$ Commensurate PLDs use $\mathbf{q}_j$ that are simple rational fractions of reciprocal-lattice vectors, whereas nearly commensurate and incommensurate states deviate slightly in magnitude or direction. Although the concept emerged most prominently in CDW studies, the same language also describes spin-Peierls distortions, charge-ordered oxides, and torsional relaxation fields in twisted two-dimensional moiré systems [1609.09486], [1710.09107], [2203.06510]. ## 1. Mathematical form and symmetry content A PLD is a displacement field rather than a density field. In the simplest one-mode description, $$ \mathbf{u}(\mathbf{r})=A\cos(\mathbf{q}\cdot \mathbf{r}+\phi), $$ while for two coexisting modulations one may write $$ \Delta_i(\mathbf{r})=A_i\sin(\mathbf{q}_i\cdot \mathbf{r}+\phi_i). $$ Here $\mathbf{q}_i$ fixes the periodicity, $A_i$ is the amplitude vector, and $\phi_i(\mathbf{r})$ carries local information about defects, dislocations, and domain walls. In BSCMO, the amplitude vectors are transverse to $\mathbf{q}_i$; in twisted bilayers, the analogous field is written as $\mathbf{r}'=\mathbf{r}_0+\mathbf{u}(\mathbf{r}_0)$ with $\mathbf{u}_N(\mathbf{r})=\sum_{n=1}^N A_n\sum_{i=1}^3 \hat a_i\sin(n\mathbf{q}_i\cdot \mathbf{r})$ [1707.00221], [2203.06510]. The symmetry content of a PLD is often more informative than its amplitude alone. In the commensurate $2\times2\times2$ CDW of $1T$-TiSe$_2$, the three principal wavevectors are $\mathbf{q}_1=(\tfrac12,0,0)$, $\mathbf{q}_2=(0,\tfrac12,0)$, and $\mathbf{q}_3=(0,0,\tfrac12)$ in reciprocal-lattice units, with dominant in-plane polarizations and an antiphase relation between adjacent layers, $$ \mathbf{u}_{\text{layer 2}}(\mathbf{r})=-\mathbf{u}_{\text{layer 1}}(\mathbf{r}), $$ which restores inversion symmetry and forbids net handedness along $c$ [1710.09107]. In phenomenological form, the coupled electronic and lattice sectors may be expressed through $$ F[\Delta,u]=\alpha(T)|\Delta|^2+\beta|\Delta|^4+\gamma|u|^2+\lambda(\Delta u^*+\Delta^*u)+\cdots, $$ or, alternatively, through a charge-lattice functional $$ F[\rho,u]=\tfrac12\alpha|\rho|^2+\tfrac14\beta|\rho|^4+\tfrac12 C|\nabla u|^2+g\,\mathrm{Re}[\rho\,\nabla\cdot u]+\cdots, $$ both of which encode the lock-in between electronic order and ionic displacement [2010.09826], [1710.09107]. ## 2. Canonical realizations in layered chalcogenides Layered chalcogenides provide the standard exemplars of PLD physics because their electronic phase transitions are frequently inseparable from real-space superstructures. In $1T$-TaS$_2$, the low-temperature commensurate phase contains three principal CDW/PLD wavevectors, $$ \mathbf{q}_1=\frac{3}{13}\mathbf{b}_1+\frac{1}{13}\mathbf{b}_2,\qquad \mathbf{q}_2=-\frac{1}{13}\mathbf{b}_1+\frac{4}{13}\mathbf{b}_2,\qquad \mathbf{q}_3=-\mathbf{q}_1-\mathbf{q}_2, $$ forming the “Star-of-David” $13\times13$ supercell. Each vector lies roughly $13.9^\circ$ from the nearest $\Gamma$–M direction and has magnitude $|\mathbf{q}_C|\simeq0.28\,\text{\AA}^{-1}$, corresponding to a $1.2\,\text{nm}$ periodicity. In the nearly commensurate phase, reached for $280\,\text{K}17\,\text{GPa}$ [1807.04993]. | System | Modulation vector / supercell | Quantitative signature | |---|---|---| | $1T$-TaS$_2$ | $13\times13$ “Star-of-David” | $|\mathbf{q}_C|\simeq0.28\,\text{\AA}^{-1}$, $1.2\,\text{nm}$ period, $A_j=11$–$15\,\text{pm}$ | | $1T$-TiSe$_2$ | $2\times2\times2$ | Se displacement $\approx0.028\,\text{\AA}$, $z<5\,\text{pm}$, antiphase stacking | | SnSe$_2$ under pressure | $\sqrt{3}\times\sqrt{3}\times1$ at $(1/3,1/3,1)$ | onset above $\approx17\,\text{GPa}$, Raman modes at $\approx25$ and $\approx40\,\text{cm}^{-1}$ | These examples show that PLD may be nearly commensurate, strictly commensurate, or pressure induced, and that the relevant length scale may range from picometer displacements to nanometer supercells [1609.09486], [1710.09107], [1807.04993]. ## 3. Microscopic driving mechanisms The standard starting point is a coupled electron-phonon Hamiltonian, $$ H=\sum_k \epsilon_k c_k^\dagger c_k+\sum_q \Omega_q b_q^\dagger b_q+\sum_{k,q} g_q c_{k+q}^\dagger c_k(b_q+b_{-q}^\dagger), $$ with a static lattice distortion $u(\mathbf{Q})=\langle b_{\mathbf{Q}}+b_{-\mathbf{Q}}^\dagger\rangle$ below the transition and an electronic order parameter $\Delta_k\simeq \sum_q g_q u(q)\delta_{k,k-q}$. This framework permits both the canonical Peierls interpretation and a broader class of strong-coupling scenarios in which lattice, charge, and sometimes excitonic degrees of freedom are co-primary [2010.09826]. In SnSe$_2$, density functional theory attributes the pressure-induced PLD to the combined effect of strong Fermi surface nesting and electron-phonon coupling at $\mathbf{q}=(1/3,1/3,1)$. The static nesting function $\chi(\mathbf{q})$ develops a pronounced peak at precisely that wavevector under compression, while the corresponding DFPT phonon branch softens to $\omega_{\mathbf{q}_{\rm PLD},\nu}\to0\,\text{meV}$ and the mode-resolved EPC constant becomes anomalously large, of order unity. This is explicitly contrasted with canonical CDW-bearing TMDs, where significant Fermi surface nesting is not usually involved [1807.04993]. In $1T$-TiSe$_2$, a different microscopic route has been formulated. An exciton-condensate mean-field model with order parameter $$ \Delta(T)=\Delta_0\sqrt{1-(T/T_c)^2}, $$ using $\Delta_0=115\,\text{meV}$ and $T_c=200\,\text{K}$, generates a condensate-induced force on the lattice through the electron-phonon term. Minimization of the phonon and exciton-phonon energies yields a static normal-coordinate shift and, for the Ti transverse mode at the $L$ point, a calculated displacement amplitude $u_{\rm Ti}^{\rm theo}\approx0.025\,\text{\AA}$ at $T\to0$, compared with the neutron-diffraction value $u_{\rm Ti}^{\rm exp}\approx0.04\,\text{\AA}$ at $77\,\text{K}$. The calculation reproduces the order of magnitude of the experimental PLD amplitude without invoking an independent Jahn-Teller instability [1012.1119]. The monolayer limit of $1T$-TiSe$_2$ reinforces the mixed electronic-lattice character. DFT finds a Kohn-type soft mode of $A_g$ symmetry at the $M$ point in the undistorted $1\times1$ structure, a relaxed $2\times2$ PLD with in-plane displacements $\Delta r_{\rm Ti}\simeq0.090\,\text{\AA}$ and $\Delta r_{\rm Se}\simeq0.029\,\text{\AA}$, and a total-energy lowering $$ \Delta E(PLD\to1\times1)\simeq -4.7\,\text{meV} $$ in GGA, or $-3.7\,\text{meV}$ in LDA. The relaxed phase opens a direct gap of $\approx82\,\text{meV}$ in GGA and shows a “Mexican-hat” valence-band profile. The paper identifies these results as evidence that lattice dynamics and electron-electron interactions conspire to stabilize the CDW/PLD state [1702.08329]. In $2H$-NbSe$_2$, time-resolved optical spectroscopy provides direct evidence against a purely electronic ordering picture. At equilibrium, CDW gap opening and PLD formation occur simultaneously at $T_{CDW}\simeq33\,\text{K}$, but under moderate photoexcitation the electronic order is destroyed while the PLD survives for tens of picoseconds. The same material is also discussed in terms of finite-$Q$ phonon softening and strong anharmonic phonon-phonon interactions rather than an adequate low-energy nesting instability [2010.09826]. Independent STM work on $2H$-TaS$_2$, TaSe$_2$, and NbSe$_2$ likewise reports dominant lattice contributions rather than the electronic modulation expected from a weak-coupling Peierls transition, with measured PLD amplitudes of $(7.2\pm1.0)\,\text{pm}$ in $2H$-TaS$_2$, $(8.1\pm1.2)\,\text{pm}$ in $2H$-NbSe$_2$, and $(5.4\pm0.8)\,\text{pm}$ in $2H$-TaSe$_2$ [1303.0956]. ## 4. Imaging and quantifying PLD The recent literature treats PLD as a directly measurable vector field rather than an inferred secondary order parameter. In exfoliated $1T$-TaS$_2$, atomic-resolution HAADF-STEM at room temperature and $95\,\text{K}$ was performed at $300\,\text{kV}$ with a semi-angle of $\sim27\,\text{mrad}$, with plan-view data on the $<001>$ zone axis and cross-section data on $<100>$. Ta atomic columns were fitted to six-parameter Gaussians and centroided with sub-$3\,\text{pm}$ precision; FFTs of both raw images and centroid-extracted positions were then used to isolate genuine displacement peaks. In the commensurate phase at $95\,\text{K}$, the resulting Ta displacements show clear $1.2\,\text{nm}$ periodic modulations with amplitudes $11$–$15\,\text{pm}$ [1609.09486]. A related STEM strategy was developed for the charge-ordered manganite Bi$_{0.35}$Sr$_{0.18}$Ca$_{0.47}$MnO$_3$. Aberration-corrected HAADF-STEM at $300\,\text{kV}$ with a $30\,\text{mrad}$ convergence beam and $68$–$340\,\text{mrad}$ collection range was combined with rigid and non-rigid registration of $20$–$30$ frames and two-dimensional Gaussian fits of atomic columns to achieve $\lesssim2\,\text{pm}$ precision. The PLD field was extracted by damping selected FT satellites, inverse transforming to construct a reference lattice, and subtracting the reference positions from the original coordinates. This yielded peak amplitudes $A_{A\text{-site}}\simeq6.2\,\text{pm}$ and $A_{B\text{-site}}\simeq8.2\,\text{pm}$ [1707.00221]. STM accesses PLD through several distinct observables. In $1T$-TiSe$_2$, intentionally introduced interstitial Ti atoms in the van der Waals gap serve as local markers of the PLD: their asymmetric triangular contrast at $V_{\rm bias}\approx+150\,\text{mV}$, together with Tersoff-Hamann simulations, identifies which Se sites are displaced and which correspond to CDW maxima. At a monoatomic step of height $\simeq6\,\text{\AA}=c_0$, defects on the two terraces reveal opposite handedness and directly image the $\pi$ phase shift between adjacent sandwiches [1710.09107]. In $2H$-TaS$_2$, TaSe$_2$, and NbSe$_2$, constant-current topographs at opposite biases are decomposed as $$ S(\mathbf{r})=\frac{T^+(\mathbf{r})+T^-(\mathbf{r})}{2},\qquad A(\mathbf{r})=\frac{T^+(\mathbf{r})-T^-(\mathbf{r})}{2}, $$ thereby separating the symmetric PLD component from the antisymmetric electronic modulation [1303.0956]. In Bi$_2$Sr$_2$CaCu$_2$O$_{8+x}$, affine correction and the Lawler-Fujita drift-correction algorithm applied at coarse-graining lengths $L\approx30\,\text{\AA}$ and $L\approx6\,\text{\AA}$ yield a bond-length map with $\lesssim5\,\text{pm}$ precision over a $50\,\text{nm}\times50\,\text{nm}$ field of view [2305.10707]. Reciprocal-space diffraction remains central, but in several systems it now serves as a quantitative PLD metrology rather than only a superstructure detector. In twisted bilayer graphene and related moiré systems, a transverse torsional PLD produces superlattice peaks whose $m$th-order intensity scales as $$ I_m(\mathbf{k})\propto |J_m(\mathbf{k}\cdot \mathbf{A})|^2, $$ allowing the harmonic amplitudes to be extracted directly from selected-area electron diffraction with picometer sensitivity [2203.06510]. ## 5. Domains, defects, stacking, and non-equilibrium dynamics PLD fields are rarely uniform over macroscopic length scales. In $1T$-TaS$_2$, cryogenic STEM resolves one–Ta–Ta–bond-length stacking faults in which the trigonal $A..A..$ registry switches to hexagonal $A..B..$ over a narrow $\sim3\,\text{nm}$ boundary. The lateral shift is $1.94\,\text{\AA}$, and cross-sectional geometric phase analysis finds $>1.5\%$ local compressive strain at the boundary. Yet FFTs on either side continue to show nearly commensurate PLD peaks, so the electronic CDW persists across the stacking fault. The same study reports that NC PLDs exist inside both the stacking domains and their boundaries [1609.09486]. In BSCMO, the combined displacement field $\Delta(\mathbf{r})=\Delta_1+\Delta_2$ reveals locally unidirectional stripe domains as small as $\sim5\,\text{nm}$, despite an apparently bidirectional Fourier pattern on larger length scales. A $\Delta_1/\Delta_2$ domain boundary can host an edge-type PLD dislocation in which one wavefront terminates abruptly, the Burgers vector equals one PLD wavelength, the phase winds by $2\pi$ around the core, and the local amplitude collapses in an inlet extending from the boundary to the defect. Shear strain extracted from the phase field, $$ \epsilon_s(\mathbf{r})=\tfrac12\left(\hat q_1^\perp/|\mathbf{q}_1|\right)\cdot \nabla \phi_1(\mathbf{r}), $$ maximizes where $|\Delta_1|$ is minimal and nascent $|\Delta_2|$ appears [1707.00221]. Ultrafast experiments show that PLD can persist even when the associated electronic order is transiently suppressed. In $2H$-NbSe$_2$, a $1.03\,\text{eV}$ probe detects a sign change in the slow differential-reflectivity component $I_2$ at $T\approx33\,\text{K}$ and a divergent relaxation time $\tau_2$ on approaching $T_{CDW}$ from below. At $20\,\text{K}$, the fluence dependence shows a critical window between $F_{c1}\approx60\,\mu\text{J}/\text{cm}^2$ and $F_{c2}\approx250\,\mu\text{J}/\text{cm}^2$ in which the electronic CDW gap collapses while the PLD remains, yielding a non-equilibrium state with $\Delta\to0$ but $u\neq0$ for tens of picoseconds [2010.09826]. In K$_{0.3}$MoO$_3$, femtosecond x-ray diffraction tracks the PLD coordinate $Q(t)$ in an effective quartic potential, $$ V(Q,t)=\tfrac12 a(t)Q^2+\tfrac14 bQ^4, $$ with equation of motion $$ \frac{d^2Q}{dt^2}+2\gamma(t)\frac{dQ}{dt}+\omega_0^2[a(t)Q+Q^3]=0, $$ where $\omega_0=2\pi\times1.53\,\text{THz}$. A two-pulse optical scheme with $\Delta t_{12}\approx0.30\,\text{ps}$ produces a second revival of the PLD superlattice intensity at $t\approx0.60\,\text{ps}$ by re-exciting the electronic population and further suppressing the damping. This directly demonstrates optical control over PLD coherence in a photoinduced high-symmetry potential [1902.00388]. ## 6. Beyond canonical CDWs: cuprates, spin-Peierls systems, moiré materials, and controversies PLD is not restricted to transition-metal dichalcogenides. In Bi$_2$Sr$_2$CaCu$_2$O$_{8+x}$, the bond-length map $B(\mathbf{r})$ has mean $d_0\approx3.83\,\text{\AA}$ and a full width at half maximum $\Delta d\approx0.02d_0$, corresponding to $\sim2\%$ disorder. Fourier analysis of $B_x(\mathbf{r})$ and $B_y(\mathbf{r})$ yields peaks at $q_B\approx0.24\pm0.03\,(2\pi/d_0)$, indicating an approximately four-unit-cell modulation. Group-theoretic projection onto the $C_{4v}$ representations identifies PLD components in the $B_1$ and $E$ channels, and the same wavevector appears in the d-symmetry CDW measured via the spectroscopic ratio $R(\mathbf{r},E)=I(\mathbf{r},+E)/I(\mathbf{r},-E)$. The paper interprets the $E$-sector PLDs as a locally frozen version of the bond-stretching soft phonon at $Q\approx(0.25,0)2\pi/a$ [2305.10707]. A spin-Peierls realization is proposed for CeRu$_2$Al$_{10}$. There, the relevant distortion involves Al(1)–Al(4) atoms moving toward their unique nearest-neighbor Ce atoms, while Ce, Ru, and Al(5) remain immobile. The displacements are written as $\Delta\mathbf{r}_i=\delta_i\hat u_i$ with magnitudes of order $\sim10^{-3}\,\text{\AA}$, and the resulting three-dimensional order is characterized by $$ \mathbf{Q}=(0,0,1). $$ This pattern removes the base-centering translation, is compatible with the appearance of forbidden reflections such as $(1\,0\,1)$, and accounts for the two-site splitting of the $^{27}$Al-NQR lines for Al(1)–Al(4) below $T_0=27\,\text{K}$ while leaving Al(5) unsplit [1005.5017]. Moiré systems introduce an additional structural category: torsional PLD. In twisted bilayer graphene, the amplitude of the single-harmonic torsional order parameter obeys $$ A_1(\theta)=\frac{A_{\max}}{1+(\theta/\theta_c)^2}, $$ with $\theta_c\approx3.89^\circ$ and $A_{\max}\approx20.35\,\text{pm}$. The torsional PLD begins to form below $3.89^\circ$, reaches $8\,\text{pm}$ around the magic angle of $1.1^\circ$, and has a geometric upper bound of $22.6\,\text{pm}$. At very low twist angles, additional harmonics arise, the second, third, and fourth becoming $>1\,\text{pm}$ at approximately $0.9^\circ$, $0.45^\circ$, and $0.30^\circ$, respectively, thereby sharpening Bernal domains and narrowing solitonic boundaries to a minimal width of $\sim6\,\text{nm}$. Similar torsional distortions are reported in twisted WS$_2$, CrI$_3$, and WSe$_2$/MoSe$_2$ [2203.06510]. Several controversies in the PLD literature concern causality and symmetry. In the canonical 2H dichalcogenides, STM separation of symmetric and antisymmetric topographic channels finds that the PLD contribution dominates over the purely electronic modulation, challenging a weak-coupling Peierls interpretation [1303.0956]. In $2H$-NbSe$_2$, nonequilibrium measurements go further by exhibiting a state in which electronic order is melted but the lattice distortion remains [2010.09826]. In $1T$-TiSe$_2$, direct STM imaging with interstitial Ti markers shows inversion-symmetric achiral stacking and explicitly contradicts claims of helical CDW stacking and associated chiral order; FFT spot intensities on adjacent terraces are identical to within $5\%$ [1710.09107]. Taken together, these results place PLD not as a passive consequence of electronic order in every material, but as an experimentally resolvable degree of freedom whose symmetry, amplitude, defects, and dynamics can determine the character of the ordered phase itself. Source: https://www.emergentmind.com/topics/periodic-lattice-distortion-pld