Bistable Transition Point (BP)

• BP is a threshold in bistable systems that marks the point where the transition mechanism between two stable states changes qualitatively.
• It encompasses diverse roles such as barrier tops in energy landscapes, spinodal boundaries in driven systems, and interface thresholds in spatial media.
• Understanding BP helps in designing experiments and devices by pinpointing critical transition states and optimizing system response.

Searching arXiv for the cited papers and related terminology to ground the article in current indexing.

“Bistable Transition Point” (BP) does not denote a single universally standardized object across the arXiv literature. Instead, it names a family of threshold notions attached to bistable dynamics: an unstable equilibrium or saddle on an energy landscape, a bifurcation or spinodal boundary in control-parameter space, a transition-state location projected onto a reaction coordinate, or a moving interface separating phases in spatially extended media. What these uses share is a qualitative change in how a system passes between two stable states, rather than bistability of the states alone (Zijlstra et al., 2019, Shirai et al., 2019, Hamel, 2013, Ehrmann et al., 30 Jun 2026).

1. Terminological scope and principal meanings

In many papers, the exact phrase “bistable transition point” is absent, and the relevant object must be identified from the model’s threshold structure. The same acronym therefore covers distinct, though closely related, mathematical roles.

Setting	BP interpretation	Representative papers
Energy landscape	Barrier top, saddle, or transition state	(Zijlstra et al., 2019, Kusumaatmaja et al., 2015, Ehrmann et al., 30 Jun 2026)
Coupled stochastic dynamics	Pathway-crossover or path-bifurcation threshold	(Chen et al., 2014, Tian et al., 2016, Kato et al., 2024)
Driven/open systems	Dynamical spinodal or bifurcation point	(Shirai et al., 2019, Wang et al., 13 Jul 2025)
Spatially extended media	Interface location, unstable threshold, or transition set	(Hamel, 2013, Shen et al., 2015, Herrmann et al., 2012)
Devices with hysteresis	Critical load or control-parameter switching threshold	(Sarkar et al., 2013, Reina et al., 2019, Wolde et al., 2023)

A useful taxonomy follows from this diversity. In state-space formulations, BP is the unstable object separating basins of attraction. In parameter-space formulations, BP is the locus where a monostable regime becomes bistable, or where one bistable mechanism changes into another. In front-propagation problems, BP is often not a point at all but an interface $\Gamma_t$, a level-set location $X_\lambda(t)$, or even a selected propagation speed rather than a spatial threshold (Guo et al., 2018). This suggests that BP is best regarded as a model-dependent transition organizer.

2. BP as transition state on an energy landscape

A canonical meaning of BP is the unstable configuration at the top of the barrier between two metastable wells. In a bistable optical trap, the intrinsic transition point is the maximum of the one-dimensional potential of mean force

$G_{1D}(x)=-k_B T\ln P_{1D}(x),$

or, in the full three-dimensional description, the saddle of

$G_{3D}(\mathbf r)=-k_B T\ln P_{3D}(\mathbf r).$

The paper distinguishes this intrinsic barrier-top definition from the operational “transition region” used to extract reactive trajectories: in practice, trajectories are segmented between boundaries placed halfway in energy between each well minimum and the barrier top, so the measured transition object is a zone rather than a single point (Zijlstra et al., 2019). A related dynamical marker is that the transition-path velocity profile has a minimum at the barrier top, making the unstable equilibrium the slowest local point along the crossing.

In liquid-crystal free-energy landscapes, BP is the transition state on a minimum-energy pathway. For the planar bistable device studied within two-dimensional Landau–de Gennes theory, the relevant object is an index-1 saddle connecting stable equilibria. Its structure depends strongly on anchoring strength $W$: for $W\ge 6.5\times 10^{-3}$, the transition state is defect-mediated; for $1.4\times 10^{-3}\le W<6.5\times 10^{-3}$, it is defect-free; and for $W<1.4\times 10^{-3}$, a cusp-catastrophe regime is recovered in which the rotated branch acts as the transition state connecting two diagonal states (Kusumaatmaja et al., 2015). Here BP is simultaneously a state-space saddle and a parameter-space threshold where the nature of the saddle changes.

In inverse-designed bistable nanostructures, BP is the transition-state location $r_t$ on a prescribed energy profile $E(r)$ along a geometric reaction coordinate $X_\lambda(t)$0, with metastable states at $X_\lambda(t)$1 and $X_\lambda(t)$2. The design hierarchy is explicit: barrier heights and minima positions are readily programmable, whereas controlling $X_\lambda(t)$3 or the full shape of $X_\lambda(t)$4 requires additional design freedom. The optimization loss

$X_\lambda(t)$5

makes BP itself a design target rather than merely a by-product of the landscape (Ehrmann et al., 30 Jun 2026).

3. Pathway-restructuring BPs in coupled bistable systems

In coupled stochastic systems, BP often denotes a threshold where the dominant escape mechanism changes topology. A paradigmatic example is the gradient system of two overdamped coupled bistable oscillators,

$X_\lambda(t)$6

with effective potential

$X_\lambda(t)$7

Here the relevant thresholds are the coupling-dependent boundaries $X_\lambda(t)$8, found numerically from collisions and annihilations of stationary points. For identical oscillators $X_\lambda(t)$9, the primary BP is

$G_{1D}(x)=-k_B T\ln P_{1D}(x),$0

where serial two-step activation through $G_{1D}(x)=-k_B T\ln P_{1D}(x),$1 or $G_{1D}(x)=-k_B T\ln P_{1D}(x),$2 disappears and the mechanism becomes one-step direct activation. A secondary BP at $G_{1D}(x)=-k_B T\ln P_{1D}(x),$3 reduces two equivalent one-step paths to a single one-step path. For nonidentical oscillators, $G_{1D}(x)=-k_B T\ln P_{1D}(x),$4, creating an intermediate mixed-pathway regime in which one candidate route remains two-step while the other is already one-step (Chen et al., 2014). An important distinction made there is that nonmonotonic transition-rate maxima are related to the same competition of mechanisms but are not themselves the formal BP.

A closely related notion appears in coupled bistable gene circuits. There the relevant threshold is not loss of bistability of the deterministic steady states but loss of local action-minimizing stability of the symmetric most probable escape path. The decisive quantity is the transverse Hessian eigenvalue $G_{1D}(x)=-k_B T\ln P_{1D}(x),$5: if $G_{1D}(x)=-k_B T\ln P_{1D}(x),$6 along the entire path, the non-bifurcating path is stable; if $G_{1D}(x)=-k_B T\ln P_{1D}(x),$7 somewhere, the transition path bifurcates into symmetry-broken branches. For the parameter set studied, the forward transition becomes bifurcating at approximately $G_{1D}(x)=-k_B T\ln P_{1D}(x),$8 (Tian et al., 2016). In this usage, BP is a symmetry-breaking point in path space.

Periodic forcing adds another variant. In the extended Stuart–Landau oscillator, the autonomous bistable boundary is the unstable limit cycle separating attraction to the stable fixed point from attraction to the stable oscillation. Under periodic forcing with quadratic feedback, the decisive transition is governed by homoclinic, SNIC, and Hopf bifurcations in the averaged rotating-frame dynamics, with oscillation quenching occurring by homoclinic bifurcation (Kato et al., 2024). BP is therefore the separatrix in the unforced system and a critical forcing-frequency value in the forced system.

4. BP as parameter-space bifurcation, hysteresis threshold, or finite-size spinodal

In driven-dissipative optical bistability, BP can denote a genuine dynamical transition in the thermodynamic limit and a finite-size crossover in the quantum system. For the periodically driven open Dicke-type model analyzed via the Floquet dissipative map

$G_{1D}(x)=-k_B T\ln P_{1D}(x),$9

the mean-field system has two separated limit cycles for $G_{3D}(\mathbf r)=-k_B T\ln P_{3D}(\mathbf r).$0 and one limit cycle for $G_{3D}(\mathbf r)=-k_B T\ln P_{3D}(\mathbf r).$1. The critical point is

$G_{3D}(\mathbf r)=-k_B T\ln P_{3D}(\mathbf r).$2

where a linearized relaxation rate vanishes as

$G_{3D}(\mathbf r)=-k_B T\ln P_{3D}(\mathbf r).$3

At finite $G_{3D}(\mathbf r)=-k_B T\ln P_{3D}(\mathbf r).$4, however, tunneling leaves a unique asymptotic Floquet state, so BP appears as a spinodal-like crossover diagnosed by the closing of the Floquet dissipative gap,

$G_{3D}(\mathbf r)=-k_B T\ln P_{3D}(\mathbf r).$5

with scaling

$G_{3D}(\mathbf r)=-k_B T\ln P_{3D}(\mathbf r).$6

(Shirai et al., 2019). This is a precise example of the distinction between a true bifurcation and its finite-size precursor.

Several device papers use BP in a more operational hysteresis sense. In the Bloch oscillating transistor, the formal threshold is $G_{3D}(\mathbf r)=-k_B T\ln P_{3D}(\mathbf r).$7, where the gain diverges as $G_{3D}(\mathbf r)=-k_B T\ln P_{3D}(\mathbf r).$8, and the experimentally extracted bifurcation current $G_{3D}(\mathbf r)=-k_B T\ln P_{3D}(\mathbf r).$9 is obtained by extrapolating $W$0 on the $W$1–$W$2 plane (Sarkar et al., 2013). In a VO$W$3-based thermomechanical oscillator driven by near-field heat exchange, BP is not named, but the bistable interval in substrate temperature is explicitly

$W$4

with the authors interpreting the switching phenomenology as consistent with hysteretic multistability bounded by saddle-node-type turning points (Reina et al., 2019). In the dual phosphorylation cycle described by a two-dimensional Chemical Master Equation, the deterministic/stochastic transition from unimodal to bimodal stationary behavior occurs near

$W$5

in the detailed-balance symmetric example, while non-equilibrium currents can delay or anticipate the onset of bimodality (Bazzani et al., 2011). Across these examples, BP is the threshold for non-uniqueness of stationary response.

5. Interface-based BPs in spatially extended and lattice systems

For bistable reaction–diffusion fronts, BP is often distributed along an interface rather than concentrated at a single scalar threshold. In $W$6, the closest analogues are the unstable threshold value $W$7 in the bistable nonlinearity, the interface set $W$8, and level sets $W$9, especially $W\ge 6.5\times 10^{-3}$0. The large-scale propagation quantity is not a point but the global mean speed

$W\ge 6.5\times 10^{-3}$1

independent of front geometry (Hamel, 2013). In exterior domains and multiple-branch geometries, the corresponding selected quantity is again the unique planar speed $W\ge 6.5\times 10^{-3}$2, provided complete propagation occurs (Guo et al., 2018).

In inhomogeneous one-dimensional media, the local bistable threshold $W\ge 6.5\times 10^{-3}$3 or $W\ge 6.5\times 10^{-3}$4 is the natural state-space BP, but it need not organize the global dynamics into a coherent front. The paper on mixed bistable-ignition reactions proves existence and uniqueness of transition fronts under a quantitative small-tail condition, yet also constructs spatially periodic and temporally periodic pure bistable examples with no transition front in the Berestycki–Hamel sense (Zlatos, 2015). This directly refutes the common identification of a local unstable threshold with a global moving BP.

For nonlocal bistable equations in time-heterogeneous media, the sharpest interface-level BP is the level-set location $W\ge 6.5\times 10^{-3}$5, defined by

$W\ge 6.5\times 10^{-3}$6

When a space-nonincreasing transition front exists, $W\ge 6.5\times 10^{-3}$7 is $W\ge 6.5\times 10^{-3}$8, has bounded speed, and satisfies

$W\ge 6.5\times 10^{-3}$9

Uniform steepness, exponential tail estimates, uniqueness up to spatial shift, and asymptotic speeds in periodic or uniquely ergodic media all make $1.4\times 10^{-3}\le W<6.5\times 10^{-3}$0 a precise dynamical BP (Shen et al., 2015).

Discrete lattice models admit a related interface notion. In subsonic phase-transition waves for atomic chains with double-well interaction, the transition layer is the narrow region where the strain crosses the spinodal interval $1.4\times 10^{-3}\le W<6.5\times 10^{-3}$1. That interface persists under localized perturbations of the bi-quadratic force law, remains sharply localized near the crossing point, and is selected physically by causality and the associated radiation condition (Herrmann et al., 2012). More recent constructions of changing-shape bistable transition fronts by mixing finite planar fronts show that even when the front geometry evolves, the interface can still be described in terms of finite facets and entire solutions on a finite-dimensional manifold of planar-front data (Guo et al., 2024).

6. Measurement, design, and conceptual distinctions

The most technologically explicit recent BP appears in hybrid quantum sensing. In the NV-center–microwave Van der Pol self-oscillator, the control parameter

$1.4\times 10^{-3}\le W<6.5\times 10^{-3}$2

locates the endpoint of the bistable oscillation phase, with the BP at

$1.4\times 10^{-3}\le W<6.5\times 10^{-3}$3

At zero spin detuning, three steady-state frequency solutions—two stable and one unstable—meet there. Near the BP, the frequency response obeys

$1.4\times 10^{-3}\le W<6.5\times 10^{-3}$4

while nonlinear saturation prevents the divergent noise penalty characteristic of linear exceptional-point sensing. The measured outcome is a $1.4\times 10^{-3}\le W<6.5\times 10^{-3}$5 enhancement in SNR and a sensitivity of $1.4\times 10^{-3}\le W<6.5\times 10^{-3}$6 (Wang et al., 13 Jul 2025). In this setting, BP is explicitly a nonlinear state-coalescence point rather than a defective linear eigenspace.

Mechanical switching devices often realize BP as a snap-through threshold. In the single-input bistable mechanism built from a near-bifurcation column, a nonlinear spring, and a bistable state element, the natural BP analogues are the critical loads $1.4\times 10^{-3}\le W<6.5\times 10^{-3}$7 and $1.4\times 10^{-3}\le W<6.5\times 10^{-3}$8 for forward and reverse switching. Experimentally, the two snap-through thresholds occur at input displacements of approximately $1.4\times 10^{-3}\le W<6.5\times 10^{-3}$9 mm and $W<1.4\times 10^{-3}$0 mm, and a nonlinear spring with quadratic stiffness is reported as crucial for regulating the state-switching behavior (Wolde et al., 2023). Here BP is operationally the loss of the current equilibrium path under cyclic loading.

Several recurrent ambiguities follow from these examples. BP should not be conflated with mere bistability of the wells; in the coupled-oscillator problem, the system remains bistable while the activated mechanism changes from serial to direct switching (Chen et al., 2014). It should not be conflated with the experimentally defined transition region; in the optical-trap problem, the barrier top and the operational trajectory-segmentation zone are related but not identical (Zijlstra et al., 2019). It should also not be conflated with a true asymptotic phase transition in finite systems; in Floquet optical bistability, finite $W<1.4\times 10^{-3}$1 replaces the mean-field bifurcation by metastability and gap closing (Shirai et al., 2019).

Taken together, these results support a compact general definition: BP is the model-specific threshold—state-space, parameter-space, or interface-level—at which the mechanism mediating passage between two stable states changes qualitatively. The threshold may be a saddle, a fold boundary, a path-bifurcation condition, a level-set location, or a critical coupling/load/temperature. What makes it “bistable” is not only the coexistence of two stable states, but the presence of a distinguished transition structure whose reorganization can be identified mathematically and, increasingly, exploited experimentally.