---
title: von Zeipel-Lidov-Kozai Migration
url: https://www.emergentmind.com/topics/von-zeipel-lidov-kozai-migration
type: topic
---

# von Zeipel-Lidov-Kozai Migration

Searching arXiv for recent and foundational work on von Zeipel–Lidov–Kozai migration.
von Zeipel–Lidov–Kozai migration is a secular high-eccentricity migration process in hierarchical or mildly hierarchical multiple systems in which gravitational perturbations from a distant companion drive large-amplitude, long-term exchanges between eccentricity and inclination in an inner orbit, reducing the pericenter distance until dissipative processes such as tides or gravitational-wave emission can shrink and circularize the orbit. In the conservative limit, the dynamics are migration only in eccentricity–inclination space, with semi-major axis fixed; true radial migration arises when dissipation is coupled to the secular dynamics [1208.0237][1807.05901][2505.13780]. Across planetary, stellar, compact-object, satellite, and small-body contexts, the mechanism is commonly described through libration or circulation of the argument of periapsis, conservation or approximate conservation of a Kozai-like angular-momentum component, and the existence of resonant phase-space structures that regulate access to extreme eccentricities [1208.0237][2205.11238][2207.12954].

## 1. Historical definition and dynamical scope

The modern term combines contributions associated with von Zeipel, Lidov, and Kozai and refers to the secular dynamics of a hierarchical triple in which a distant perturber excites coupled oscillations of eccentricity and inclination in an inner orbit [2507.21251][1208.0237]. In the simplest quadrupole, test-particle formulation, the secular Hamiltonian is integrable, the argument of periapsis \(\omega\) may librate around \(90^\circ\) or \(270^\circ\), and the quantity
\[
\sqrt{1-e^2}\cos i
\]
is conserved, enforcing an anticorrelated exchange between \(e\) and \(i\) [1807.05901][1208.0237][1601.05814]. For initially circular inner orbits, the classical inclination window is
\[
i \in \big[\arccos(\sqrt{3/5}),\,\arccos(-\sqrt{3/5})\big] \approx [39.2^\circ,140.8^\circ]
\]
in the test-particle quadrupole limit [1807.05901].

In migration problems, this conservative secular exchange is only the first stage. The second stage occurs when the maximum eccentricity becomes large enough that the pericenter
\[
q = a(1-e)
\]
or \(q_1=a_1(1-e_1)\) becomes sufficiently small for tides, gravitational-wave emission, or related dissipative mechanisms to remove orbital energy during repeated pericenter passages [1208.0237][1807.05901][1601.05814]. The semi-major axis then shrinks while the orbit circularizes, producing hot Jupiters, close stellar binaries, or compact-object mergers depending on the physical setting [1208.0237][1807.05901][2606.20803].

The same secular architecture also appears outside the standard hierarchical binary-planet problem. Recent work treats low-hierarchy triples where the ZLK timescale approaches orbital periods [2505.13780], near-Earth asteroids affected by Jupiter [2507.21251], resonant trans-Neptunian objects inside mean-motion resonances with Neptune [2207.12954], and white-dwarf planetary systems in which migration is delayed until after stellar evolution [2010.04724]. This suggests that “von Zeipel–Lidov–Kozai migration” is best understood as a family of secular transport processes rather than a single specialized channel.

## 2. Conservative Hamiltonian structure

In the standard secular treatment of a hierarchical triple, the fast mean anomalies are averaged out and the dynamics reduce to slow evolution of orbital elements. For the inner orbit, a common canonical description uses Delaunay variables
\[
L = \sqrt{\mu a},\quad G = L\sqrt{1-e^2},\quad H = G\cos i,
\]
with conjugate angles \(l=M\), \(g=\omega\), and \(h=\Omega\) [2505.13780]. In the quadrupole test-particle problem, \(L\) is constant, so \(a\) is fixed, and axial symmetry makes \(H\) an integral, leaving a 1-degree-of-freedom Hamiltonian in \((G,g)\) [2505.13780][1208.0237].

A representative quadrupole secular Hamiltonian is
\[
F=(2+3e^2)(3\cos^2 i-1)+15e^2\sin^2 i\cos 2\omega,
\]
with the companion invariant
\[
H=(1-e^2)\cos^2 i,
\]
or equivalently the “Kozai constant” \(h=\sqrt{1-e_1^2}\cos i\) in stellar-triple notation [1208.0237][1807.05901]. These constants organize the familiar phase portraits in the \((e,\omega)\) plane, separating circulation from libration islands [1208.0237][1807.05901].

For low-hierarchy systems, the standard double-averaged quadrupole Hamiltonian can fail because nonlinear short-period effects contaminate the long-term evolution. A recent formulation extends the Brown Hamiltonian by carrying the von Zeipel procedure to second order in the disturbing function, producing a normalized long-term Hamiltonian
\[
{\cal H}^{**}=-{\cal C}_0{\cal F},\qquad {\cal F}={\cal F}_{20}+\varepsilon_{21}{\cal F}_{21}+\varepsilon_{22}{\cal F}_{22},
\]
where \({\cal F}_{20}\) is the standard double-averaged quadrupole term, \({\cal F}_{21}\) is the classical Brown Hamiltonian, and \({\cal F}_{22}\) is a new extended Brown term induced by inner-orbit fast oscillations [2505.13780]. In element form,
\[
{\cal F}_{20} = \frac{1}{6}\left(2+3e^2\right)\left(3\cos^2 i - 1\right) + \frac{5}{2}e^2\sin^2 i\cos(2\omega),
\]
\[
{\cal F}_{21} = \frac{3}{16}\eta\cos i\left[ 2\sin^2 i + e^2\left(33+17\cos^2 i + 15\sin^2 i\cos 2\omega\right) \right],
\]
with \(\eta=\sqrt{1-e^2}\), and \({\cal F}_{22}\) adds higher harmonics such as \(\cos 4\omega\) [2505.13780]. The result remains integrable at quadrupole order in the restricted problem, but it modifies fixed points, resonant islands, precession rates, and the shape and period of ZLK cycles [2505.13780].

A complementary conceptual development identifies the eccentric ZLK effect at octupole order with apsidal secular resonance. In the restricted hierarchical planetary problem, after averaging over rotating ZLK cycles, one obtains an effective resonant Hamiltonian with slow angle
\[
\sigma = \varpi^*-\varpi_p,
\]
where \(\sigma\) is the apsidal angle difference. The resulting phase portraits contain multiple libration zones, separatrices, and flipping regions, and the paper concludes that the eccentric ZLK effect is dynamically equivalent to apsidal resonance in the test-particle limit [2205.11238]. This rephrasing is especially relevant for migration because the extreme-eccentricity channels that enable dissipation are localized in specific resonant zones rather than being generic across phase space [2205.11238].

## 3. Secular timescales, hierarchy, and the onset of extreme eccentricity

A standard quadrupole ZLK timescale used in low-hierarchy analyses is
\[
t_{\rm ZLK}  \simeq  \frac{16}{15} \left(\frac{1}{n}\right) \left(\frac{m_0}{m_p}\right) \left(\frac{a_{\rm p,eff}}{a}\right)^3,\qquad a_{\rm p,eff}=a_p\sqrt{1-e_p^2},
\]
while in hierarchical stellar and planetary migration problems one often uses
\[
t_{\rm KL} \simeq \frac{2T'^2}{3\pi T}\,\frac{M_*+m_p+m'}{m'}(1-e'^2)^2
\]
or
\[
t_{\rm KL} \sim \frac{M_{\star}+m_1}{m_2}\,\frac{P_2^2}{P_1}(1-e_2^2)^{3/2}
\]
depending on notation and context [2505.13780][1208.0237][2606.20803]. The common point is that ZLK is a secular process: much slower than orbital motion in strongly hierarchical triples, but not necessarily so in mildly hierarchical ones [2505.13780][2507.19709].

Recent low-hierarchy work introduces two correction parameters,
\[
\varepsilon_{21} = \left(\frac{n_p}{n}\right)\left(\frac{m_p}{m_0+m_p}\right) \frac{1}{(1-e_p^2)^{3/2}\left(1+\frac{2}{3}e_p^2\right)},
\]
\[
\varepsilon_{22} = \left(\frac{n_p}{n}\right)^2\left(\frac{m_p}{m_0+m_p}\right) \frac{1}{(1-e_p^2)^3}\left(1+3e_p^2+\frac{3}{8}e_p^4\right),
\]
with \(\varepsilon_{21}\) essentially the single-averaging parameter and \(\varepsilon_{22}\) measuring how strongly inner-orbit fast variations contaminate the secular dynamics [2505.13780]. For nearly circular outer orbits and \(m_0\ll m_p\),
\[
\varepsilon_{21} \simeq \frac{1}{\sqrt 3}\left(\frac{a}{r_H}\right)^{3/2} \sim \frac{8}{15\pi}\frac{P_{\rm out}}{t_{\rm ZLK}},
\qquad
\varepsilon_{22} \simeq \frac{1}{3}\left(\frac{a}{r_H}\right)^3 \sim \frac{8}{15\pi}\frac{P_{\rm in}}{t_{\rm ZLK}},
\]
so the hierarchy can be read directly from ratios of orbital and secular timescales [2505.13780]. In the irregular-satellite examples treated there, \(t_{\rm ZLK}\) is comparable to \(P_{\rm out}\) and only \(5\)–\(6\) times \(P_{\rm in}\), making the classical double average unreliable [2505.13780].

At higher order in the hierarchy expansion, the octupole parameter
\[
\varepsilon_{\rm oct} =(a/a_p)\,e_p/(1-e_p^2)
\]
measures the strength of non-axisymmetric secular forcing [2505.13780]. In hierarchical migration literature the corresponding quantity is often written
\[
\epsilon_{\rm Oct} \equiv \frac{a_0}{a_{\rm out}}\frac{e_{\rm out}}{1-e_{\rm out}^2},
\]
or
\[
\epsilon \equiv \frac{a_1}{a_2}\frac{e_2}{1-e_2^2},
\]
and controls orbit flips, very large eccentricity spikes, and chaotic modulation of the quadrupole ZLK cycles [1601.05814][2606.20803]. This suggests a practical separation of regimes: quadrupole-dominated ZLK for nearly circular outer companions and strong hierarchy, octupole-dominated or eccentric ZLK when \(e_{\rm out}\) is significant, and Brown-corrected quasi-secular ZLK when \(t_{\rm ZLK}\) approaches orbital times [2505.13780][2507.19709].

## 4. Dissipative coupling and the mechanism of migration

The conservative Hamiltonian alone does not change the semi-major axis in the restricted secular problem; instead,
\[
L=\sqrt{\mu a}=\text{const}\Rightarrow a=\text{const}
\]
in the absence of non-conservative forces [2505.13780]. Migration therefore requires coupling the secular eccentricity pumping to a dissipative channel. In planetary systems this is usually equilibrium tides or, in some regimes, chaotic dynamical tides; in stellar triples it is tidal friction inside the inner stars; in compact-object problems it may be gravitational-wave emission [1208.0237][1807.05901][2010.04724].

A standard high-eccentricity migration picture has two stages. First, ZLK cycles at large \(a\) drive the eccentricity toward \(e_{\max}\), periodically reducing \(q=a(1-e)\) while leaving \(a\) nearly fixed [1208.0237][1807.05901]. Second, dissipation concentrated near pericenter removes orbital energy. Since the orbital angular momentum is approximately conserved during this stage, the post-circularization orbit satisfies relations such as
\[
a_f \approx a_{\rm init}(1-e_{\max})
\]
or, in the white-dwarf planetary case,
\[
a_{\rm p,f} \approx 2\,a_{\rm p,0} (1-e_{\max})
\]
depending on the adopted limit and notation [2606.20803][2010.04724].

The migration threshold is commonly expressed as a required minimum pericenter. For hot-Jupiter production by ZLK in stellar binaries, a critical migration pericenter is
\[
r_{p,\rm mig} \equiv 2.4\times 10^{-2}\,{\rm AU} \left(\frac{R_p}{R_{\rm J}}\right)^{5/7} \left(\frac{\chi}{10}\right)^{1/7} \left(\frac{M_p}{M_{\rm J}}\right)^{-1/7}
\left(\frac{a_0}{1\,{\rm AU}}\right)^{-1/7}
\left(\frac{k_{2p}}{0.37}\right)^{1/7}
\left(\frac{M_*}{M_\odot}\right)^{2/7},
\]
with the corresponding migration eccentricity
\[
e_{\rm mig}\equiv 1-\frac{r_{p,\rm mig}}{a_0}
\]
[1601.05814]. The disruption threshold is similarly written
\[
r_{p,\rm dis} = \eta R_p\left(\frac{M_*}{M_p}\right)^{1/3},
\qquad
e_{\rm dis}\equiv 1-\frac{r_{p,\rm dis}}{a_0},
\]
so successful migration requires reaching \(e_{\rm mig}\) without exceeding \(e_{\rm dis}\) [1601.05814][2010.04724].

Short-range forces determine whether the required eccentricities are dynamically accessible. In quadrupole ZLK with short-range forces, the limiting eccentricity \(e_{\rm lim}\) is set by balancing the secular torque against apsidal precession from GR, tidal bulges, and rotation [1601.05814][2010.04724]. In low-hierarchy or high-eccentricity planetary migration, a similar role is played by GR precession
\[
\dot\omega_{\rm GR} \sim \frac{3 G(M_*+m_p)}{a c^2 (1-e^2)} n
\]
and tidal/rotational precession; once these dominate over the ZLK torque, the resonance is quenched and the orbit circularizes along a nearly constant-angular-momentum path [1208.0237][2405.19511].

The HAT-P-11 analysis shows that structural evolution of the migrating planet can be dynamically essential. There, the orbit-averaged tidal luminosity
\[
\langle \mathcal{L} \rangle
\]
is used to update the planetary radius through a fitted luminosity–radius relation
\[
\frac{R}{R_\oplus}=A x^4+B x^3+C x^2+D x+E,\qquad x=\log_{10}\left(\frac{\mathcal{L}}{L_\odot}\right),
\]
with the consequence that a larger \(R_b\) greatly enhances tidal precession because \(\dot\omega_{\rm Tide}\propto R_b^5\) [2405.19511]. In that system, significant tidally driven radius inflation is required for ZLK cycles to be quenched at the semi-major axis that later circularizes to the observed orbit; fixed-radius integrations over-migrate the planet inward [2405.19511]. This suggests that in some sub-Jovian migration problems, internal thermal response is part of the secular dynamics rather than a secondary correction.

## 5. Regime-dependent formulations and representative astrophysical settings

A major recent theme is that the secular backbone depends strongly on hierarchy and on whether the perturber is stellar, planetary, or embedded in a more complex multi-body architecture. The table summarizes regimes that are explicitly developed in recent work.

| Regime | Representative formulation | Main consequence for migration |
|---|---|---|
| Strongly hierarchical triples | Double-averaged quadrupole or octupole secular Hamiltonian [1208.0237][1807.05901][1601.05814] | Classical ZLK cycles; migration set by \(e_{\max}\), GR, and tides |
| Mildly hierarchical triples | Brown-corrected secular Hamiltonians [2505.13780][2507.19709] | Periods, amplitudes, and flip boundaries can be misestimated by DA models |
| Planet–planet high-e migration | Octupole-order double-averaged Hamiltonian with tides and spin evolution [2606.20803] | Short-period hot Jupiters preferentially arise from ZLK with inclined companions |
| Secular resonance plus ZLK in small-body dynamics | Direct \(N\)-body or semi-secular resonant models [2507.21251][2207.12954] | Migration is mainly in perihelion distance or resonant action rather than circularization |
| White-dwarf systems | ZLK plus tides after stellar-evolution-driven orbital expansion [2010.04724] | Delayed migration becomes possible only in the WD phase |

In stellar triples, one concrete application is the production of short-period inner binaries through ZLK cycles plus tidal friction. In a secular octupole-plus-GR-plus-tides framework, migrating systems preferentially have small \(|h|\), where
\[
h=\sqrt{1-e_1^2}\cos i,
\]
and are not determined by inclination alone: the initial \(e_1\) and \(\omega_1\) also control whether the system reaches the large \(e_1\) required for migration [1807.05901]. A key result is that systems starting near the center of a KL libration island can remain non-migrating because the eccentricity amplitude is too small, even though they are formally in the ZLK regime [1807.05901].

For exoplanet migration, Gliese 436b provides a classic two-phase example. Starting near its current orbit, GR and tides suppress ZLK and the planet simply circularizes; starting \(5\)–\(10\) times farther out allows active Kozai cycles, followed by a second phase in which the system leaves the resonance and rapidly shrinks at high eccentricity [1208.0237]. The paper emphasizes that the total effective circularization time can exceed the standard constant-\(a\) estimate by a factor larger than \(50\), making it possible to observe the planet in an intermediate migration stage [1208.0237].

In planet–planet secular high-e migration, recent simulations identify two limiting channels: mutually inclined companions driving ZLK and nearly coplanar eccentric companions driving CHEM-like migration. The shortest-period hot Jupiters arise preferentially from ZLK with highly inclined companions and display a broad range of stellar obliquities, whereas longer-period hot Jupiters are produced over longer timescales by nearly coplanar high-e migration and remain aligned [2606.20803]. The paper derives an analytic envelope relating final mutual inclination and final semi-major axis,
\[
I_{\rm mut,f} = \cos^{-1} \left[ \sqrt{\frac{a_{\rm init} (1-e_{\rm init}^2)}{a_f}} \cos I_{\rm mut,init} \right],
\]
which organizes the simulated obliquity–period anti-correlation [2606.20803].

For near-Earth asteroids, the same label is used in a somewhat different sense. In the Miorita study, a von Zeipel–Lidov–Kozai secular resonance with Jupiter is identified numerically through long-period anticorrelated \(e\)–\(i\) oscillations of period \(\sim 30\) kyr, while a simultaneous apsidal resonance \(\varpi-\varpi_5\approx 180^\circ\) confines the semimajor axis and enables metastable low-perihelion states [2507.21251]. There, “migration” is chiefly migration of perihelion distance, not of semimajor axis, and some objects are ultimately driven into the Sun [2507.21251].

## 6. Efficiency, limitations, and current controversies

The modern literature treats von Zeipel–Lidov–Kozai migration as robust but conditional. A central result from analytic work on giant planets in stellar binaries is that the total migration fraction
\[
F_{\rm mig}=F_{\rm HJ}+F_{\rm dis}
\]
is much less sensitive to planet mass, host type, and tidal strength than the surviving hot-Jupiter fraction \(F_{\rm HJ}\), because the main geometric determinant is the fraction of initial inclinations that can reach the migration pericenter [1601.05814]. For fiducial giant-planet populations at \(a_0\in[1,5]\) AU, the analytic calculations give \(\mathcal{F}_{\rm mig}^{\rm LK}\sim 12\text{–}15\%\), while \(\mathcal{F}_{\rm HJ}^{\rm LK}\) is only a few percent or less depending on mass and dissipation [1601.05814].

Population synthesis of stellar-binary KL migration yields a stronger negative conclusion. One large Monte Carlo study finds that KL migration produces hot Jupiters with semi-major axes generally smaller than observed and can explain the data only if both of the following hold: tidal dissipation at high eccentricities is at least \(\sim 150\) times more efficient than the upper limit inferred from the Jupiter–Io interaction, and highly eccentric planets are tidally disrupted at distances \(\gtrsim 0.015\) AU [1405.0280]. On the basis of occurrence rate and semi-major-axis distribution, that work argues that KL migration in stellar binaries can produce at most \(\sim 20\%\) of observed hot Jupiters and almost no intermediate-period planets with \(a\sim 0.1\)–\(2\) AU [1405.0280]. This remains one of the clearest statements of the channel’s limitations.

A related limitation is that low-hierarchy corrections can qualitatively change flip criteria and phase-space geometry. In mildly hierarchical triples, Brown corrections break the symmetry of flipping regions across \(i=90^\circ\), and double-averaged quadrupole-plus-octupole models can predict flips where \(N\)-body integrations show none, or miss flips that do occur [2507.19709]. The extended pendulum approximation developed there shows that Brown terms shift the resonant center away from \(j_z=0\), so the flip region is no longer symmetric in prograde and retrograde initial conditions [2507.19709]. For migration predictions that depend sensitively on the extreme-eccentricity tail, this is not a small correction.

The white-dwarf case adds a different limitation: the ZLK route is viable only after imposing survival through RGB and AGB evolution. For WD 1856b, the allowed planet masses and initial semimajor axes are constrained simultaneously by the requirement of survival during giant-star phases and the requirement of later WD-phase ZLK migration, leaving a narrow range \(M_p\simeq 0.7\)–\(3M_{\rm J}\) and main-sequence semimajor axis \(\simeq 2\)–\(2.5\) AU [2010.04724]. The inferred occurrence rate of Kozai-migrated planets around white dwarfs is only \({\cal O}(10^{-3}{-}10^{-4})\) [2010.04724].

The small-body literature raises a different interpretive issue. In the NEA and trans-Neptunian settings, ZLK migration often refers to inward or outward transport in perihelion distance or resonant action rather than tidal circularization [2507.21251][2207.12954]. This suggests that the phrase “ZLK migration” is used in at least two technically distinct ways: strict high-eccentricity migration producing compact circularized orbits, and secular transport in \(e\)–\(i\) space that changes encounter geometry or perihelion without large semimajor-axis decay. A plausible implication is that context-specific clarification is necessary whenever the term is used outside the exoplanet and stellar-binary literature.

## 7. Prospects and open directions

Several current developments indicate where the subject is moving. First, the conservative backbone is being improved in the quasi-secular regime. The extended Brown Hamiltonian expresses both the Hamiltonian and the mean–osculating transformation in closed form with respect to the inner and outer eccentricities, avoiding small-\(e\) expansions and the Laplace-limit restriction that affects Hansen-series approaches [2505.13780]. Because dissipation is applied to osculating orbits, this mean–osculating map is particularly relevant when the dissipation rate depends sensitively on pericenter distance [2505.13780].

Second, the octupole and Brown frameworks are beginning to converge. The mildly hierarchical eccentric-ZLK study extends Brown corrections to include octupole–octupole coupling, shows that quadrupole–octupole coupling vanishes in the chosen gauge, and demonstrates that Brown corrections are the key factor responsible for breaking the symmetry of flipping regions [2507.19709]. This suggests a route toward secular migration calculations that remain accurate when \(t_{\rm ZLK}\sim P_{\rm out}\), a regime relevant for planets in stellar binaries and for binaries near supermassive black holes [2507.19709].

Third, obliquity has become a diagnostic of migration channel. In planet–planet secular migration, the shortest-period hot Jupiters tend to arise from ZLK driven by highly inclined companions and thus preserve a broad range of stellar obliquities, while longer-period systems are preferentially aligned because they arise from nearly coplanar secular migration [2606.20803]. The paper predicts that the shortest-period hot Jupiters should have distant planetary companions with broadly distributed mutual inclinations, whereas the companions of longer-period systems should be nearly coplanar [2606.20803]. Gaia astrometry is explicitly identified there as a key observational test.

Finally, some systems now require coupled dynamical and thermodynamic modeling. The HAT-P-11 study shows that scattering, ZLK cycles, tidal dissipation, and radius inflation must be evolved together to reproduce the observed semimajor axis and eccentricity [2405.19511]. This suggests that a purely orbital secular model may be insufficient for sub-Saturn and Neptune-mass planets undergoing high-e migration, especially when tidal luminosity can strongly alter the planetary radius and therefore the precession rate [2405.19511].

Taken together, the literature supports a synthesis in which von Zeipel–Lidov–Kozai migration is a secular transport mechanism whose conservative core is now understood at several levels of refinement—quadrupole, octupole, apsidal-resonant, Brown-corrected, and resonant-in-MMR—while the actual astrophysical outcome depends on how that core is coupled to GR, tides, spins, internal structure, stellar evolution, or additional resonances [2505.13780][2205.11238][1807.05901][2606.20803]. This suggests that the central problem is no longer whether ZLK migration exists, but rather which specific formulation is appropriate for a given hierarchy, and which subset of systems can traverse the narrow corridor between insufficient eccentricity excitation and catastrophic disruption.

Source: https://www.emergentmind.com/topics/von-zeipel-lidov-kozai-migration