---
title: 'Screw Dislocation: Theory & Applications'
url: https://www.emergentmind.com/topics/screw-dislocation
type: topic
---

# Screw Dislocation: Theory & Applications

A screw dislocation is a one-dimensional crystalline defect for which the Burgers vector is parallel to the line direction. In anti-plane elasticity, for a straight defect along \(e_3\), the only nonzero displacement component is the out-of-plane field \(u_z=(b/2\pi)\theta\), with \(u_r=u_\theta=0\), so that one circuit around the core produces a net shift \(\Delta z=b\). This Volterra description remains the canonical far-field model, but contemporary treatments extend it to lattice variational theories, gauge and gradient regularizations, nonlinear Riemann–Cartan geometry, and symmetry-based descriptions in electronic, phononic, and optical media [1707.06176, 2401.04392, 2405.16757].

## 1. Topological and kinematic characterization

In antiplane shear, the elastic field of a system of screw dislocations is represented by a planar strain field \(h=\nabla u\in\mathbb R^2\) satisfying
\[
\mathrm{div}\,h=0,\qquad \mathrm{curl}\,h=\sum_{i=1}^n b_i\,\delta_{z_i},
\]
where each scalar \(b_i\in\{\pm1\}\) is the Burgers modulus and \(b_i e_3\) is the Burgers vector of the \(i\)th line defect. Equivalently, for any loop \(\gamma_i\) enclosing exactly one core,
\[
b_i=\int_{\gamma_i} h\cdot t\,ds.
\]
This circulation law is the two-dimensional expression of the screw character: the defect is encoded not by a smooth displacement potential but by nontrivial winding around the core [1707.06176].

The continuum Volterra field can be written in cylindrical coordinates as
\[
u_r=0,\qquad u_\theta=0,\qquad u_z(r,\theta)=\frac{b}{2\pi}\theta.
\]
The same defect may be represented geometrically by the metric
\[
ds^2=d\rho^2+\rho^2\,d\varphi^2+\bigl(dz+\beta\,d\varphi\bigr)^2,\qquad \beta=\frac{b}{2\pi},
\]
which encodes the cut-and-glue operation of a screw dislocation. In this formulation, the single nonzero elastic strain component is \(S_{\varphi z}=S_{z\varphi}=\beta/(2\rho)\) [1002.0646].

Algebraic formulations replace the continuum branch cut by a covering-space construction. The exact sequence
\[
0\longrightarrow \mathbb Z \longrightarrow \mathbb R \xrightarrow{\exp(2\pi\,\cdot)} U(1)\longrightarrow 1
\]
is used to realize the defect as a pullback over the punctured plane; the resulting lifted section is an embedded helicoid whose vertical increment after one loop is exactly the Burgers vector. For simple cubic and body-centered cubic lattices, this viewpoint is implemented with free abelian groups and group rings, and the screw displacement is the logarithm of a lattice phase winding [1605.09550].

These equivalent descriptions emphasize that a screw dislocation is simultaneously a mechanical singularity, a topological winding defect, and a symmetry defect. That shared structure underlies the otherwise disparate continuum, atomistic, and wave-transport formulations discussed below.

## 2. Continuum field theories and core regularization

Classical elasticity gives the correct long-range \(1/r\) behavior but leaves the core singular. One major direction therefore replaces the Volterra defect by field theories with intrinsic lengths. In the translation gauge theory of dislocations, the elastic distortion and torsion are
\[
\beta_{ij}=u_{i,j}+\phi_{ij},\qquad
T_{ijk}=\phi_{ik,j}-\phi_{ij,k}=\beta_{ik,j}-\beta_{ij,k},
\]
with Bianchi identity \(\epsilon_{jkl}T_{ijk,l}=0\). For a screw dislocation in a functionally graded material whose moduli vary exponentially in \(x\) and \(y\), consistency requires
\[
\frac{c_1}{\mu+\gamma}=-\frac{c_2}{\mu-\gamma}\equiv \ell^2,
\]
which defines the intrinsic gauge length \(\ell\). The anti-plane distortion then satisfies the perturbed Helmholtz equation
\[
\bigl[1-\ell^2(\Delta+2\,a\cdot\nabla)\bigr]\beta_{zj}=\beta^0_{zj},\qquad j=x,y,
\]
and the torsion obeys the same operator with source \(b\,\delta(x)\delta(y)\). The explicit dislocation density becomes
\[
T_{zxy}(x,y)=\frac{b}{2\pi \ell^2}e^{-(a_1x+a_2y)}K_0(\kappa r),
\qquad
\kappa=\sqrt{\frac{1+a^2\ell^2}{\ell^2}},
\]
so the core is nonsingular, the full cylindrical symmetry is broken by gradation, and the fields decay faster in the direction of increasing moduli [1102.3815].

A related regularization arises in second-gradient elasticity for exponentially graded media. There, the anti-plane displacement is written as \(u(x,y)=w(x,y)e^{-ay}\), and \(w\) solves
\[
\bigl[1-c_1^2(\nabla^2-a^2)\bigr]\bigl[1-c_2^2(\nabla^2-a^2)\bigr]w(x,y)=w^0(x,y),
\]
with two intrinsic lengths \(c_1,c_2\) derived from \(\ell_1,\ell_2\). The resulting closed-form fields contain modified Bessel functions \(K_0\) and \(K_1\); their short-range contributions cancel the classical \(1/r\) divergence, while the limit \(\ell_1,\ell_2\to0\) recovers the Volterra solution [1012.5108].

A nonlinear geometric formulation uses a Riemann–Cartan manifold \((\mathcal M,g,\nabla)\), a plastic frame \(\theta\), and the torsion relation
\[
\tau:=\star\alpha,\qquad T^i=d\theta^i.
\]
For an axisymmetric screw density, \(\tau^3=f(r)b\,dx^1\wedge dx^2\otimes E_3\). A Helmholtz decomposition of \(\theta\) and a variational solution of \(d\Theta=\tau\), followed by elastic embedding into \(\mathbb R^3\), produce stress fields that are finite at \(r=0\) and converge to the Volterra \(\mu b/(2\pi r)\) field outside the core. In this setting, the Ricci curvature determines the symmetry of the stress field, and the work identifies a stress–curvature duality for screw dislocations [2401.04392].

Across these formulations, the recurrent result is that the core singularity is not mandatory. It is a feature of the Volterra idealization, not of screw dislocations per se.

## 3. Renormalized energetics, Peach–Koehler forces, and boundary effects

In bounded domains, screw-dislocation energetics are commonly defined by removing small cores of radius \(\varepsilon\) and passing to a renormalized limit. For dislocations at \(z_1,\dots,z_n\), the finite part of the energy can be expressed through the Green function \(G_\Omega(x,y)=-(1/2\pi)\log|x-y|+k_\Omega(x,y)\) as
\[
E_n(z_1,\dots,z_n)
=\frac12\sum_{i=1}^n b_i^2\,h_\Omega(z_i)
+\sum_{1\le i<j\le n} b_i b_j
\Bigl(k_\Omega(z_i,z_j)-\frac1{2\pi}\log|z_i-z_j|\Bigr),
\]
with \(h_\Omega(x)=k_\Omega(x,x)\). The Peach–Koehler force is then
\[
f_i=-\nabla_{z_i}E_n.
\]
This separates one-body self-energy, pair interactions, and boundary-image effects in a mathematically explicit way [1707.06176].

For free boundaries, the leading asymptotic force is attractive. If one dislocation \(z_1\) lies at distance \(d_1(z_1)\) from \(\partial\Omega\), with nearest boundary point \(s\), then
\[
f_1(z)=\frac{\nu(s)}{4\pi\,d_1(z_1)}+\frac{C_{n,\sigma}(\gamma)}{2\pi},
\]
so the dominant term points along the outward normal and has magnitude \(1/(4\pi d)\). Under overdamped dynamics \(\dot z_i=-\nabla_{z_i}E_n\), this yields a collision-time estimate
\[
T_{\mathrm{coll}^\partial}\le 2\pi\,\delta_0^2+O(\delta_0^3)
\]
for a dislocation initially at boundary distance \(\delta_0\). For two nearby dislocations of opposite Burgers vector, the pair-collision time satisfies
\[
T_{\rm coll}^\pm \le
\frac{\pi\,\zeta_0^2\,\eta_0^2}
{2\bigl(\eta_0^2-\zeta_0^2-2(n-2)\zeta_0\eta_0\bigr)},
\]
and, in the leading-order regime discussed in the companion note, scales as \((\pi/2)\zeta^2\) [1707.06176, 1705.08121].

The same framework also yields confinement results under prescribed boundary data. When one imposes tangential boundary conditions \(h\cdot\tau=f\) with total circulation \(2\pi n\), the limiting finite energy
\[
\mathcal F(a_1,\dots,a_n)
=
\sum_{i=1}^n \pi\log d_i
+\frac12\int_\Omega\Bigl|\nabla v+\sum_{i=1}^n K_{a_i}\Bigr|^2dx
+\sum_{i<j}\int_\Omega K_{a_i}\cdot K_{a_j}\,dx
\]
diverges as any \(d_i\to0\), where \(d_i\) is the minimum distance to the boundary or to another core. Consequently, minimizers stay strictly in the interior and remain well separated [1707.06176].

Zero-temperature lattice models sharpen this picture at the atomistic level. A variational anti-plane model on the triangular lattice establishes existence of a global minimizer with net Burgers vector \(1\), despite non-coercivity of the energy in the displacement variable. A related analysis proves locally stable multi-dislocation states provided dislocations are sufficiently far from one another and from the boundary, and shows that linear elasticity remains accurate beyond a few lattice spacings from the cores [1304.2500, 1403.0518].

## 4. Core structure, cross-slip, and mobility in crystalline metals

In HCP Mg, the perfect screw \(\langle a\rangle\) dislocation dissociates on the basal plane into two Shockley partials bounding an \(I_1\)-type stacking fault of width \(d\simeq 7.7\) Å in DFT or \(\simeq 12\) Å in EAM. Density-functional calculations give two nearly degenerate basal-dissociated cores and a finite-size-scaled cross-slip barrier
\[
\Delta E_{\rm Peierls}=61.4\pm2.0\ {\rm meV}/b.
\]
A line-tension model based on the DFT Peierls potential gives a jog-pair formation enthalpy \(\Delta H^*\simeq1.4\)–\(1.7\) eV at zero applied \(\tau_{zx}\), or \(\simeq1.1\)–\(1.3\) eV at \(100\) MPa. Molecular dynamics further shows jerky prismatic glide at low temperature, the appearance of plateau segments above \(\sim250\) K, and a \(\sim20\)–\(30\%\) reduction of cross-slip CRSS when a \(6\)-Å hard sphere is placed on the basal plane at \(50\) K and \(150\) K [1506.03538].

In fcc Al, real-space orbital-free DFT indicates that the energetic core size of a perfect screw is \(\approx 7|{\bf b}|\), substantially larger than the \(1\)–\(3|{\bf b}|\) estimate obtained from displacement fields alone. Upon relaxation, the perfect screw dissociates into two Shockley partials with separations \(d_{\rm screw}=8.24\) Å and \(d_{\rm edge}=6.59\) Å. The relaxed core energy per unit length is nearly linear in applied strain,
\[
E_c(\epsilon_{ij})\approx E_{c0}+\sum_{i\le j} S_{ij}\epsilon_{ij},
\]
with \(E_{c0}=0.284\) eV/Å and \(S_v\approx -1.0\) eV/Å for equi-triaxial volumetric strain. In discrete dislocation networks, the resulting core forces include \(1/d^2\), \(1/d\), and line-tension-type terms, and the type-I contribution can remain at least \(10\%\) of the Peach–Koehler force out to \(10\)–\(15\) nm [1701.08912].

For the \(1/2[111]\) screw dislocation in \(\alpha\)-Fe, ab initio calculations resolve an additional core field beyond the Volterra screw field. Its symmetry is that of a biaxial dilatation with
\[
M_{xx}=M_{yy}=650\pm50\ {\rm GPa\ \AA^2},\qquad M_{zz}=0.
\]
Including this field yields converged core energies \(219\pm1\) meV/Å for the easy core and \(227\pm1\) meV/Å for the hard core when \(r_c=3\) Å. The same core field strengthens short-range interactions, increasing the dipole passing stress by up to \(20\%\) for \(h<15\) Å and increasing carbon binding by \(\sim0.05\) eV for sites within \(10\)–\(15\) Å of the dislocation [1112.4936].

Electronic-structure effects also control alloying trends in bcc Mo. For a/2\(\langle111\rangle\) screw cores, Ta and W raise both the complex formation energy and the zero-stress Peierls barrier, whereas Os, Ir, and Pt lower them. At the first-nearest-neighbor site, the reported values are \(E_{\rm form}=+0.70\) eV and \(\Delta E_P(0)=0.42\) eV for Ta, \(+0.62\) eV and \(0.40\) eV for W, \(-0.08\) eV and \(0.36\) eV for Os, \(-0.22\) eV and \(0.33\) eV for Ir, and \(-0.30\) eV and \(0.30\) eV for Pt. The fitted stress-dependent barrier for pure Mo is
\[
\Delta E_{\rm Mo}(\sigma)=0.38\,[1-(\sigma/3.20\,{\rm GPa})^{0.50}]^{1.50}\ {\rm eV},
\]
with analogous parameter changes under solute addition [2404.04897].

At larger scales, recent atomistic simulations in Nb, Mo, dilute Nb–Mo alloys, and equiatomic NbMo at \(300\) K over strain rates from \(10^3\) to \(10^7\ {\rm s}^{-1}\) and line lengths from \(15\) to \(50\) nm show that cross-kinks form not only in concentrated alloys but also in pure BCC metals. High-rate depinning proceeds predominantly via vacancy–interstitial cluster formation, whereas low-rate, long-line configurations exhibit lateral cross-kink migration, three-dimensional forward–backward cross-slip, and prismatic loop formation [2602.16883].

## 5. Screw symmetry as a generator of electronic, phononic, and optical states

In a three-dimensional strong \(\mathbb Z_2\) topological insulator, a screw dislocation can bind one-dimensional helical electronic modes. For weak indices \((\nu_x,\nu_y,\nu_z)=(1,1,1)\), the Ran–Zhang–Vishwanath criterion is
\[
\mathbf b\cdot \mathbf M=1\quad (\bmod\,2),\qquad \mathbf M=(1,1,1).
\]
Thus, for Burgers vector \(\mathbf b=(0,0,Na)\), odd \(N\) gives gapless dislocation modes and even \(N\) does not. In the continuum Dirac description, the defect is encoded by the twisted boundary condition
\[
\psi(r,\theta+2\pi,z)=
\exp\!\Bigl[i\frac{b\,k_z}{2\pi}\tau_z\Bigr]\psi(r,\theta,z),
\]
and the bound states have dispersion \(E(k_z)=\pm v_z k_z\). Transport calculations separate \(T=T_{\rm dis}+T_{\rm side}\); under periodic boundary conditions in \(x\) and \(y\), odd \(N\) gives \(T\approx4\) at low \(E_F\), while even \(N\) gives \(T\ll1\). A Zeeman gap on the side surfaces suppresses \(T_{\rm side}\) without affecting \(T_{\rm dis}\) [2405.16757].

A closely related symmetry mechanism appears in phononics, but without requiring a nontrivial bulk topological invariant. In an HCP phononic crystal with a \(6_3\) screw dislocation, the defect preserves a local screw symmetry
\[
S_n:(\theta\to\theta+2\pi/n,\ z\to z+b/n),
\qquad S_n^n=T_b.
\]
Modes can be labeled by Bloch momentum and screw eigenvalue, with the selection condition \(e^{ik_z b}e^{i2\pi m/n}=1\). Within the bulk band gap \(3.75\)–\(5.20\) kHz, exactly two defect modes appear for each \(k_z\), with angular indices \(m=\pm1\). Their transverse displacement fields have \(u_x\) and \(u_y\) exactly \(90^\circ\) out of phase, and near \(\Gamma\) they are described by
\[
H_{\rm eff}(k_z)\simeq \omega_0\sigma_0+v k_z \sigma_z,
\qquad
\omega_\pm(k_z)=\omega_0\pm v k_z.
\]
Propagation direction and helicity are therefore locked by the screw symmetry itself [2404.18347].

In amorphous optical media, the same defect geometry acts as an effective gauge field for paraxial propagation. With metric \(ds^2=d\rho^2+\rho^2d\varphi^2+(dz+\beta\,d\varphi)^2\), the paraxial envelope obeys
\[
-2ik\,\partial_z\Psi=
\Bigl[\partial_\rho^2+\frac1\rho\partial_\rho
+\frac1{\rho^2}(\partial_\varphi-\beta k)^2\Bigr]\Psi,
\]
equivalently with gauge potential \(A_\varphi=\beta k\). The orbital angular momentum shifts as \(m\to m+\beta k\), allowing vortex annihilation or generation, while the elasto-optic birefringence induces polarization precession and an optical Hall effect [1002.0646].

In GaN, exact restoration of the screw-dislocation group algebra yields band-connectivity and optical-selection rules. For the \(6_2\) threading screw, the eigenvalues of \(S=\{C_n|mc/n\}\) satisfy
\[
\lambda_\mu(k_z)=\exp[i(k_zmc/n+2\pi\mu/n)],
\]
and the band-flow rule under \(k_z\to k_z+2\pi/c\) is \(\Delta\mu=+2\ (\mathrm{mod}\ 6)\). Electric-dipole transitions obey
\[
\mu_f-\mu_i\equiv m \quad (\mathrm{mod}\ 6).
\]
The computed recombination coefficients are \(B_{\rm rad}\simeq2.8\times10^{-14}\ {\rm cm^3/s}\) and \(C_{\rm nonrad}\simeq9.0\times10^{-10}\ {\rm cm^3/s}\) at \(\Delta E=0.7\) eV and \(T=300\) K, with \(C_{\rm nonrad}\gg B_{\rm rad}\) by four orders of magnitude. The reported mechanism is a spiraling piezoelectric potential that spatially separates electrons and holes and suppresses radiative recombination [2601.19240].

These results directly contradict the common reduction of screw dislocations to purely mechanical defects. In the cited systems, the defect line is a symmetry-bearing channel that can bind, filter, or suppress excitations.

## 6. Unifying themes across scales and models

Several themes recur across otherwise different theories. First, the core is environment-dependent rather than universal. Gauge theory introduces \(\ell\) and an effective inverse length \(\kappa\) in graded media; second-gradient elasticity introduces \(c_1\) and \(c_2\); nonlinear geometry replaces a Dirac core by a smooth density \(f(r)\); atomistic calculations in Al assign an energetic core radius of \(\approx7|{\bf b}|\); and \(\alpha\)-Fe requires an additional biaxial core field beyond the Volterra solution [1102.3815, 1012.5108, 1701.08912, 1112.4936].

Second, the far field is robust while the near field is model-sensitive. The variational lattice analyses show that linear elasticity is essentially correct outside a few lattice spacings, and the nonlinear geometric construction recovers the Volterra \(\mu b/(2\pi r)\) field away from the core. By contrast, cross-slip barriers, strain-dependent core energies, short-range solute binding, and symmetry-filtered waveguiding are all controlled by the detailed core structure [1403.0518, 2401.04392, 1506.03538].

Third, screw symmetry itself is operational. In phononics it protects \(m=\pm1\) helical modes; in topological insulators it determines whether a dislocation binds conducting channels through \(\mathbf b\cdot\mathbf M\); in GaN it organizes band connectivity and dipole selection rules; and in optics it enters as an Aharonov–Bohm-type phase shift. A plausible implication is that screw dislocations should be treated not only as defects to be regularized, but also as one-dimensional symmetry objects that can be exploited or mitigated depending on context [2404.18347, 2405.16757, 2601.19240].

Finally, boundaries and external fields are not secondary perturbations. Free surfaces attract screw dislocations; prescribed boundary data can confine them; applied strain modifies core energy in Al; hard-sphere obstacles catalyze cross-slip in Mg at low temperature; and ferromagnetic coating can isolate dislocation transport in a topological insulator by gapping side-surface states [1705.08121, 1701.08912, 1506.03538, 2405.16757].

Within this modern view, the screw dislocation is best understood as a topological line defect whose mechanics, energetics, and transport signatures are inseparable from intrinsic length scales, local symmetry, and environmental coupling.

Source: https://www.emergentmind.com/topics/screw-dislocation