---
title: Hard Mode in Three-Mode Optomechanical Systems
url: https://www.emergentmind.com/topics/hard-mode
type: topic
---

# Hard Mode in Three-Mode Optomechanical Systems

A three-mode optomechanical system can exhibit a **hard excitation mode** in which phonon generation and excitation of a lower-frequency optical mode begin through a discontinuous, hysteretic jump rather than a continuous threshold crossing. In the configuration studied in "A control of threshold of hard excitaion mode of optomechanical system by a low-intensity seed wave" [2506.16269], two optical cavity modes and one mechanical mode satisfy a Brillouin-type frequency relation, one optical mode is driven by an off-resonant pump, and the lower-frequency optical mode is resonantly driven by a weak seed. The central result is that a low-intensity seed wave can significantly reduce the threshold of the hard excitation mode by changing the stability of solutions in the bistability region, so that switching from a non-generating state to a generating one occurs at lower pump intensity [2506.16269].

## 1. Three-mode optomechanical configuration

The system consists of a high-frequency optical mode of frequency $\omega_1$ with annihilation operator $\hat a_1$, a low-frequency optical mode of frequency $\omega_2$ with annihilation operator $\hat a_2$, and a mechanical phonon mode of frequency $\omega_b$ with annihilation operator $\hat b$. These are related by the Brillouin-type condition
$$
\omega_1 - \omega_2 = \omega_b .
$$
Two coherent external fields act on the cavity: an off-resonant pump of frequency $\omega$ drives the higher optical mode, while a seed wave at frequency $\omega_2$ resonantly drives the lower optical mode [2506.16269].

Under the rotating-wave approximation, the Hamiltonian is
$$
\begin{aligned}
\hat H &= \hbar \omega_1 \hat a_1^\dagger \hat a_1
      + \hbar \omega_2 \hat a_2^\dagger \hat a_2
      + \hbar \omega_b \hat b^\dagger \hat b \\
&\quad + \hbar g \left( \hat a_1^\dagger \hat a_2 \hat b
                     + \hat a_1 \hat a_2^\dagger \hat b^\dagger \right) \\
&\quad + \hbar \Omega_1 \left( \hat a_1^\dagger e^{-i\omega t}
                            + \hat a_1 e^{+i\omega t} \right) \\
&\quad + \hbar \Omega_2 \left( \hat a_2^\dagger e^{-i\omega_2 t}
                            + \hat a_2 e^{+i\omega_2 t} \right).
\end{aligned}
$$
The three-wave mixing term describes conversion between the optical modes via the phonon mode, $\hat a_1 \leftrightarrow \hat a_2 + \hat b$, with coupling strength $g$ [2506.16269].

Including dissipation with rates $\gamma_1$, $\gamma_2$, and $\gamma_b$, and neglecting noise, the semiclassical equations of motion for the mode amplitudes are
$$
\frac{d a_1}{dt} = (-\gamma_1 - i\omega_1)\,a_1 - i g\, a_2 b - i \Omega_1 e^{-i\omega t},
$$
$$
\frac{d a_2}{dt} = (-\gamma_2 - i\omega_2)\,a_2 - i g\, a_1 b^* - i \Omega_2 e^{-i\omega_2 t},
$$
$$
\frac{d b}{dt} = (-\gamma_b - i\omega_b)\,b - i g\, a_1 a_2^*.
$$
The analysis assumes $a_1(0)=a_2(0)=b(0)=0$ and both coherent drives switched on at $t=0$ [2506.16269].

## 2. Bistability and the hard excitation mode

In this system, **soft excitation** denotes a continuous onset of generation above threshold, with a unique stable steady state for each pump value. By contrast, **hard excitation mode** denotes a bistable regime in which a low-intensity non-generating state and a high-intensity generating state coexist, producing a jump-like switch and hysteresis as the pump is varied [2506.16269].

After transforming to rotating frames,
$$
a_1(t)=a_{1\mathrm{st}}e^{-i\omega t}, \qquad
a_2(t)=a_{2\mathrm{st}}e^{-i\omega_2 t}, \qquad
b(t)=b_{\mathrm{st}}e^{-i(\omega-\omega_2)t},
$$
the stationary equations become
$$
0 = (-\gamma_1 - i \delta\omega_1)\,a_{1\mathrm{st}} - i g\, a_{2\mathrm{st}} b_{\mathrm{st}} - i \Omega_1,
$$
$$
0 = (-\gamma_2 - i \Delta_2)\,a_{2\mathrm{st}} - i g\, a_{1\mathrm{st}} b_{\mathrm{st}}^* - i \Omega_2,
$$
$$
0 = (-\gamma_b - i \delta\omega_b)\,b_{\mathrm{st}} - i g\, a_{1\mathrm{st}} a_{2\mathrm{st}}^*.
$$
Here the pump detuning is $\delta\omega_1=\omega_1-\omega$, the seed is resonant so $\Delta_2=0$, and $\delta\omega_b$ is the effective mechanical detuning [2506.16269].

For the unseeded case, $\Omega_2=0$, one steady-state branch is the trivial non-generating solution
$$
a_{2\mathrm{st}}=0,\qquad b_{\mathrm{st}}=0,\qquad
a_{1\mathrm{st}}^{(0)}=\frac{\Omega_1}{\gamma_1+i\delta\omega_1}.
$$
A nontrivial generating solution with finite $a_{2\mathrm{st}}$ and $b_{\mathrm{st}}$ can coexist with it [2506.16269].

The possibility of hard excitation is governed by the inequality
$$
\delta\omega_1(\delta\omega_2+\omega_b) > \gamma_1(\gamma_2+\gamma_b).
$$
When this holds, there is a bistable interval
$$
\Omega_{\mathrm{ex}} \le |\Omega_1| \le \Omega_{\mathrm{th}},
$$
with analytic boundaries
$$
\Omega_{\mathrm{ex}}=
\frac{\sqrt{\gamma_b\gamma_2}}{|g|}
\left|
\delta\omega_1+
\frac{(\delta\omega_2+\omega_b)\gamma_1}{\gamma_2+\gamma_b}
\right|,
$$
and
$$
\Omega_{\mathrm{th}}=
\frac{1}{|g|}
\frac{\sqrt{\gamma_b\gamma_2}}{\gamma_2+\gamma_b}
\sqrt{(\gamma_1^2+\delta\omega_1^2)
\left[
1+\left(\frac{\delta\omega_2+\omega_b}{\gamma_2+\gamma_b}\right)^2
\right]}.
$$
Within this interval, the low- and high-intensity branches are both stable; above $\Omega_{\mathrm{th}}$, the low-intensity branch loses stability and the system must jump to the generating state [2506.16269].

## 3. Seed-wave control of the threshold

The lower-frequency optical mode is directly seeded by the resonant drive $\Omega_2$. Once $\Omega_2\neq 0$, the strictly zero solution
$$
a_{2\mathrm{st}}=0,\qquad b_{\mathrm{st}}=0
$$
no longer exists. Instead, two qualitatively distinct steady-state branches remain: a low-intensity branch continuously connected to the unseeded trivial state as $\Omega_2\to 0$, and a high-intensity branch corresponding to the generating state [2506.16269].

Numerical solutions of the dynamical equations show that the hard excitation persists in the seeded system: at low $\Omega_1$ the system follows the low-intensity branch, and with increasing pump it still undergoes an abrupt jump to the high-intensity branch. The decisive modification is that the jump occurs at a smaller pump amplitude, and the effective switching point moves downward as $\Omega_2$ increases [2506.16269].

The paper states that, with seed present, no closed analytic threshold formula is given; the threshold is obtained numerically. The principal trend is that increasing $\Omega_2$ shifts the effective switching point from $\Omega_{\mathrm{th}}$ toward $\Omega_{\mathrm{ex}}$. The total optical power required for switching, counted as pump plus seed, can drop by **a factor of several** relative to the single-pump threshold [2506.16269].

This threshold reduction is specific to the hard excitation regime. In the soft excitation regime, where the bistability condition is not satisfied, the seed only slightly shifts the generation threshold and does not produce a dramatic reduction or a discontinuous jump [2506.16269].

## 4. Stability mechanism and threshold reduction

The stability analysis is based on linearization around a steady solution,
$$
a_1(t)=a_{1\mathrm{st}}+\delta a_1(t),\qquad
a_2(t)=a_{2\mathrm{st}}+\delta a_2(t),\qquad
b(t)=b_{\mathrm{st}}+\delta b(t),
$$
leading to a linear system
$$
\frac{d}{dt}\mathbf{X}=\mathbf{J}\mathbf{X},
$$
where $\mathbf{J}$ is the Jacobian. Stability is determined by the eigenvalues $\lambda_j$: the steady state is stable when all $\Re(\lambda_j)<0$ and unstable when at least one eigenvalue has positive real part [2506.16269].

Without seed, at $\Omega_1=\Omega_{\mathrm{th}}$ one real part crosses zero, so the non-generating solution loses stability and the system switches to the generating state. With seed, the eigenvalue spectrum of the low-intensity branch changes in two linked ways. First, at fixed seed amplitude, increasing $\Omega_1$ makes the least damped eigenvalue less negative. Second, at fixed pump amplitude, increasing $\Omega_2$ likewise reduces the magnitude of the most weakly damped eigenvalue [2506.16269].

The paper interprets this as a reduction of the stability of the low-intensity branch inside the bistable window. Because the high-intensity branch already exists there, weakening the stability of the low-intensity solution makes switching to the generating branch occur at lower pump intensity. The stated mechanism is therefore not the appearance of a new branch, but a seed-induced change in the stability landscape of the bistability region [2506.16269].

A plausible implication is that the seed acts as a controlled perturbation of the gain-loss balance in the coupled $(a_2,b)$ subsystem. The paper explicitly notes that the seed provides an additional coherent input to $a_2$, enhances the effective couplings $a_1a_2^*$ and $a_1b^*$ in the linearized dynamics, and thereby pushes an eigenvalue toward zero at a lower pump value [2506.16269].

## 5. Numerical regime and representative parameters

For the hard excitation regime discussed in the main numerical examples, the paper uses
$$
\gamma_1/2\pi = \gamma_2/2\pi = 19~\text{MHz},\qquad
\gamma_b/2\pi = 121~\text{MHz},
$$
$$
\delta\omega_1/2\pi = \delta\omega_b/2\pi = -6~\text{GHz},\qquad
\Delta_2/2\pi = 0,
$$
and
$$
g/2\pi = 1.6~\text{kHz}.
$$
For these parameters, the one-pump hard excitation threshold is
$$
\Omega_{\mathrm{th}} = 5\times 10^{14}~\text{s}^{-1},
$$
and the hard excitation condition is satisfied [2506.16269].

The plotted examples compare the unseeded case with seeded cases such as
$$
\Omega_2 = 0.03\,\Omega_{\mathrm{th}},\qquad
\Omega_2 = 0.07\,\Omega_{\mathrm{th}}.
$$
In the unseeded case, the intensities remain very low until $\Omega_1$ reaches $\Omega_{\mathrm{th}}$, where they jump abruptly to high values. With seed, the intensities no longer vanish at low pump, and the jump to the high-intensity branch occurs at smaller $\Omega_1$, between $\Omega_{\mathrm{ex}}$ and $\Omega_{\mathrm{th}}$. Increasing $\Omega_2$ moves the jump point toward $\Omega_{\mathrm{ex}}$ and reduces the unstable region [2506.16269].

In the soft-excitation example, the paper states that parameters are chosen so that the bistability inequality is not satisfied, and quotes
$$
\Omega_{\mathrm{th}} = 3.6 \times 10^{-12}\,\text{s}^{-1}.
$$
In that regime all intensity curves are smooth, with only a slight threshold shift under seeding [2506.16269].

These numerical results support a sharp distinction between two operating regimes. In the bistable regime, a weak resonant seed can move the switching boundary substantially. Outside bistability, the same seed acts only as a modest perturbation.

## 6. Functional interpretation and broader context

The paper interprets the seeded hard excitation regime as a possible basis for **all-optical transistors and logical elements**. In this mapping, the output is the intensity of the generated low-frequency optical mode or of the phonon mode; the pump amplitude $\Omega_1$ acts as a main input; and the seed amplitude $\Omega_2$ acts as a control input. A low-intensity output corresponds to a logical “0,” while the high-intensity generating state corresponds to a logical “1” [2506.16269].

The proposed transistor-like mode of operation is to bias the pump near the original hard excitation threshold. With no seed, the system remains in the low-intensity state. Turning on a weak seed lowers the effective threshold below the fixed pump value and triggers the jump to the high-intensity state. Because the output intensity change is large, the paper identifies high on/off contrast and lower pump requirements as potential advantages of this mechanism [2506.16269].

The same bistability also suggests memory-like behavior. Since two stable states coexist over the same pump interval, seed pulses can in principle trigger transitions between them, and hysteresis can maintain the chosen state after the control input is removed. This suggests optical memory and logic functionality, although the paper presents this as an implication rather than a device-level implementation [2506.16269].

Within the broader nonlinear-optical context, the work is compared to optical bistability in Kerr cavities or saturable absorbers, injection seeding and injection locking in lasers, and stimulated Brillouin scattering with a Stokes seed. The specific feature emphasized here is that the nonlinearity is mediated by the optomechanical three-wave interaction, so phonons play an explicit role in the switching dynamics [2506.16269].

The main limitation identified is that hard excitation requires a specific detuning condition and a narrow bistable operating window. The paper also notes sensitivity to noise, the challenge of realizing two optical modes with frequency separation $\omega_b$ and sufficient coupling $g$ in an integrated platform, and the absence of a closed analytic threshold formula when the seed is present [2506.16269]. Within those constraints, the study establishes seed-controlled destabilization of the low-intensity branch as a concrete mechanism for reducing the hard excitation threshold in a three-mode optomechanical system.

Source: https://www.emergentmind.com/topics/hard-mode