Environmentally Triggered Dark Energy via Dark-Sector Coupling and Explicit Vacuum Bias

We present a coupled dark sector model wherein late-time cosmic acceleration is realized through an environmentally triggered thawing quintessence mechanism. A real scalar field $\phi$ with a regularized runaway potential $V(\phi) = \Lambda_0^4 / \sqrt{1 + (\phi/\mu)^2} + \epsilon \Lambda_0^4 [1 - \tanh(\phi/\mu)]$ couples conformally to cold dark matter (CDM) via a quadratic metric interaction $A(\phi) = \exp(\beta\phi^2 / 2M_{\text{Pl}}^2)$, while Standard Model fields and baryons remain strictly uncoupled. Throughout radiation and early matter domination, the ambient dark matter density establishes a positive effective curvature ($m_{\text{eff}}^2 > 0$) at $\phi \approx 0$; severe Hubble friction ($3H \gg m_{\text{eff}}$) overdamps the field, dynamically locking it at the stable minimum with $w_\phi = -1$ and vanishing scalar activity during early epochs. As cosmic expansion dilutes the dark matter density below a critical threshold $\rho_{\text{crit}} = \Lambda_0^4 M_{\text{Pl}}^2 / (\beta\mu^2)$, an environmental transition unfreezes the field into a rolling regime that drives cosmic acceleration today. In the uncoupled limit $\beta \to 0$, Hubble overdamping ($3H \gg |m_0|$) smoothly preserves the frozen state, recovering flat $\Lambda$CDM without physical or numerical singularities. Primordial inflationary vacuum selection, augmented by a technically natural spurion bias term $\epsilon \sim 10^{-5}$, prevents domain wall formation while guaranteeing asymptotic stability ($V \ge 0$) against future Big Crunch turnaround singularities. A perturbative stability analysis confirms that scalar fluctuations are ghost-free ($Q_s > 0$) with sound speed $c_s^2 = 1$, eliminating gradient instabilities. Because baryons are uncoupled, equivalence principle tests are identically satisfied at tree level. Integrating the scalar field profile across realistic dark matter halos demonstrates that the fifth force is effectively uniform and Planck-suppressed, evaluating to $F_{\text{fifth}}/F_{\text{grav}} \approx 4.9 \times 10^{-4}$ today ($\le 3.4 \times 10^{-3}$ for $\phi \le M_{\text{Pl}}$), satisfying astrophysical bounds without invoking non-linear screening. Confronting the model with joint compressed \textit{Planck} 2018 CMB priors, DESI 2024 baryon acoustic oscillations, Pantheon+ supernovae, and RSD growth data yields an observational upper bound $\beta < 0.068$ (95\% CL). While the model provides an acceptable fit ($\chi^2/\text{dof} \approx 0.961$) and is not excluded by current data, model selection metrics indicate that it is not statistically preferred over flat $\Lambda$CDM ($\Delta\text{BIC} = +2.8$).