betaFeedback

WorksPhysics and Astronomy

Unifying Time-Independent and Time-Dependent Franck–Condon Pictures in Quantum Mechanochemistry via the Linear Coupling Approximation

Abstract

Quantum chemistry has achieved impressive results in modeling and interpreting mechanochemical transformations at the molecular level. An essential question, the time profile of force application, has however attracted less attention. Here, the Franck–Condon theory of quantum mechanochemistry, previously formulated for the sudden-force limit, is generalized to arbitrary time-dependent force profiles using the linear-coupling approximation. Solving the driven harmonic oscillator yields a closed-form, time-dependent effective Huang–Rhys factor: a natural measure of vibrational excitations upon force application. It is governed by the Fourier transform of the force derivative and recovers the sudden- and fixed-force limits. At the conceptual level, this linear coupling picture links mechanochemistry to vibronic spectroscopy and Marcus electron theory. For a shock-front force ramp, vibrational excitation is shown to be suppressed in frequency, only low-frequency modes being excited under realistic laser-shock-peening conditions.

Introduction

Mechanical force has emerged as an important tool for activating chemical reactions[1, 2]. It can be applied using different techniques, including ultrasound[3], ball mills[4], single-molecule manipulation[5] via atomic-force microscopy[6] or optical tweezers[7], shock waves[8] (including laser shock peening, LSP[9, 10]), and other methods. Understanding these phenomena at the level of molecular quantum chemistry has progressed significantly in recent years[11, 12, 13].

The seminal work of Bell[14] established the linear dependence of the reaction barrier on the applied force and thus explained mechanochemical activation in a simple and compelling way. This idea was elaborated by including nonlinear terms in the force dependence of the barrier[15]. An alternative approach models force application by constrained geometry optimization (COGEF), leaving some intramolecular forces unrelaxed to mimic the external force[16].

Another generalization of Bell’s approach applied a linear force term to the entire potential energy surface (PES), yielding the force-transformed, or force-modified, PES (FM-PES)[17]. This idea has proven particularly fruitful because it allows the smooth application of well-established PES-exploration machinery to the FM-PES, including geometry optimization, transition-state theory (TST), and molecular dynamics (MD), including steered MD. In addition, several ingenious analysis techniques and concepts have arisen, including judgement of energy distribution (JEDI) analysis[18], Newton-trajectory analysis[19], and pulling corridors[20]. Importantly, the equivalence of explicit force inclusion (EFEI) and COGEF was established[16]; these approaches are related by a Legendre transformation, with force and internal coordinate being conjugate variables[17].

The FM-PES and EFEI approaches are agnostic to the nature of the force and usually treat it as constant in time. Several contributions have addressed the first issue by explicitly including the force in the electronic Hamiltonian[21] and by computing hydrostatic stress through the electronic Hamiltonian[22, 23]. These approaches bring mechanochemical models closer, at least in principle, to the physical conditions of real experiments. Regarding time dependence, a notable application considered a stepwise linear force ramp corresponding to the load rate estimated for ultrasound[24].

The basis for the systematic study of varying force was laid down by Smalø et al., who identified two limiting cases: fixed force and sudden force application[25]. The former is adiabatically slow, whereas the latter assumes the immediate application of the force at its full strength, in a manner analogous to a vertical electronic transition. This idea was developed further by the author of this contribution by recognizing that sudden (“vertical”) external force application corresponds to the Condon principle, in which the transition takes place not between different electronic states, but rather between the PES and the FM-PES[26]. Consequently, in addition to barrier lowering, further activation occurs through vibrational excitation on the FM-PES, which can be described using Franck–Condon (FC) factors. This framework explains the different fragmentation patterns of butane in the fixed- and sudden-force regimes, as well as the faster rates in the latter regime.

Realistic modeling of different experiments requires different force time profiles. Single-molecule force spectroscopy is almost certainly adiabatic, or close to the fixed-force limit, because the force ramps are many orders of magnitude slower than vibrational relaxation times. However, it is unclear whether the fixed-force regime corresponds to any experimental reality under current technological conditions. Importantly, the issue of unrealistically fast loading rates in classical biomolecular simulations of single-molecule force spectroscopy has been identified and discussed[27]. In quantum chemistry, however, the force-application protocol is often implicit: sudden force application can occur in molecular MD applications[28], whereas in ab initio MD simulations of ma terials and liquids under periodic boundary conditions, fixed-force models based on COGEF-like constraints, such as fixing terminal atoms or changing the simulation cell, are commonly used[29, 30]. In studies using static approaches such as TST, fixed-force application is typically assumed[31, 32].

In this work, I attempt to emphasize the role of the force-loading rate by generalizing the FC theory of quantum mechanochemistry to time-dependent external-force application. In doing so, the FC theory is simplified using the linear coupling approximation (LC). This gives several notable results linking mechanochemistry to spectroscopy and electron-transfer theory, even in the fixed-force regime. The final result is an expression for the time-dependent Huang–Rhys factors, which define the FC factors in the LC approximation, for a general force-ramp profile. This expression is used to estimate the amount of vibrational excitation under typical experimental conditions.

Time-Independent Theory

Harmonic Force-Modified PES as a Linear-Coupling Model

For a one-dimensional case, such as a diatomic molecule, the force-modified potential energy surface (FM-PES) in the harmonic approximation can be written as

V(q)=12k(q−q0)2−f(q−q0),(1)V(q) = \frac{1}{2}k(q-q_{0})^{2}-f(q-q_{0}), \tag*{(1)}

where qq is the coordinate, q0q_{0} is its equilibrium value in the absence of an external force, kk is the harmonic force constant, and ff is the force magnitude. Equation (1) differs from a standard harmonic PES, V(q)=12k(q−q0)2V(q)=\frac{1}{2}k(q-q_{0})^{2}, by the linear force-dependent term. This term allows the application of a linear-coupling model known from vibronic spectroscopy without further approximation[33], as shown in Figure 1.

Force-modified potential energy surfaces as a linear-coupling model

Figure 1. Force-modified PES as a linear-coupling model: gray, V(q)V(q), harmonic PES; orange, V(qf)V(q^{f}), force-modified PES (FM-PES); δq\delta q, coordinate displacement; q0q_{0} and q0fq_{0}^{f}, minima on the PES and FM-PES, respectively; ERE_{R}, reorganization energy; kk, harmonic force constant.

Application of an external force shifts the PES minimum to a new position:

q0f=q0+fk=q0+δq.(2)q_{0}^{f}=q_{0}+\frac{f}{k}=q_{0}+\delta q. \tag*{(2)}

If the force application is instantaneous, or sudden[26], the Condon approximation[33, 34] applies and leads to a nonthermal population of vibrational energy levels on the FM-PES. The probabilities of vibrational excitation following sudden force application are given by Franck–Condon factors (FCFs). These are overlaps between the vibrational wave functions on the original PES, χn\chi_{n}, and those on the FM-PES, χmf\chi_{m}^{f}.

Evaluating FCFs in the linear-coupling model is straightforward because the force constant remains the same for the PES and the FM-PES. Therefore, the vibrational wave functions associated with the PES and FM-PES are harmonic oscillator eigenfunctions related by a shift of origin. For simplicity, we consider the zero-temperature case, in which only the vibrational ground state is populated before force application. We

introduce the dimensionless displacement parameter

Δ=μωℏδq,(3)\Delta=\sqrt{\frac{\mu\omega}{\hbar}}\delta q, \tag*{(3)}

where μ\mu is the reduced mass, ω\omega is the angular vibrational frequency, and ℏ\hbar is the reduced Planck constant. The probability of a transition from χ0\chi_{0} to χmf\chi_{m}^{f}, the Franck–Condon factor, is then

Fm0=∣⟨χmf|χ0⟩∣2=(Δ2/2)mm!e−Δ2/2=Smm!e−S,(4)\begin{aligned} F_{m0}=\left|\left\langle\chi_{m}^{f}\middle|\chi_{0}\right\rangle\right|^{2} \\ &=\frac{(\Delta^{2}/2)^{m}}{m!}e^{-\Delta^{2}/2}=\frac{S^{m}}{m!}e^{-S}, \tag*{(4)} \end{aligned}
S=Δ2/2,S=\Delta^{2}/2,

where SS is the Huang–Rhys factor[35]. One can show that the Huang–Rhys factor equals the molecular reorganization energy following sudden force application, expressed in units of the vibrational quantum:

ER=Sℏω.(5)E_{R}=S\hbar\omega. \tag*{(5)}

Since the vibrational levels are equidistant in the harmonic approximation, SS equals the average number of vibrational quanta in the state under the external force, or the level of vibrational excitation. Substituting Eq. (3) into Eq. (5), we obtain

ER=12μω2δq2=12f2k.(6)\begin{aligned} E_{R}=\frac{1}{2}\mu\omega^{2}\delta q^{2} \\ &=\frac{1}{2}\frac{f^{2}}{k}. \tag*{(6)} \end{aligned}

The reorganization energy, ER=12μω2δq2=12kδq2E_{R}=\frac{1}{2}\mu\omega^{2}\delta q^{2}=\frac{1}{2}k\delta q^{2}, is the same as the strain energy, which is the harmonic increase in energy at finite displacement, analogous to the energy required to stretch a spring. (5) establishes the dependence of the reorganization, or strain, energy on the applied force, which is quadratic.

The force dependence of the Huang–Rhys factor is

S=12μωℏδq2=12μωℏf2k2=f22ℏ2μω3.(7)S = \frac{1}{2}\frac{\mu\omega}{\hbar}\delta q^2 = \frac{1}{2}\frac{\mu\omega}{\hbar}\frac{f^2}{k^2} = \frac{f^2}{2\hbar^2\mu\omega^3}. \tag*{(7)}

It is worth noting that ERE_{\mathrm{R}} in the two-parabola picture in Figure 1 is conceptually analogous to the Marcus reorganization energy in electron-transfer processes[36].

Kinetics on the Force-Modified PES

Although a harmonic potential does not describe bond breaking, we assume that it provides a good approximation for vibrational motion, while the dissociation limit on the PES is still present, as is standard in this type of model. According to the simple Bell theory[14], the dissociation barrier, or activation energy EAfE_{\mathrm{A}}^{f}, depends linearly on the applied force:

EAf=EA−f(q−q0).(8)E_{\mathrm{A}}^{f} = E_{\mathrm{A}} - f(q-q_0). \tag*{(8)}

Thus, the larger the pulling force, the lower the dissociation barrier. This observation provides the simplest explanation of mechanical activation. In the sudden-force regime, additional activation originates from vibrational excitation, as quantified by the Huang–Rhys factor [Eq. (5)]. Within a simple Fridman–Macheret model, this excess energy, or strain/reorganization energy, is subtracted from the activation energy[37, 38]:

EAf(sudden)=EAf−ER=EAf−Sℏω.(9)E_{\mathrm{A}}^{f}(\mathrm{sudden}) = E_{\mathrm{A}}^{f} - E_{\mathrm{R}} = E_{\mathrm{A}}^{f} - S\hbar\omega. \tag*{(9)}

Combining Eqs. (9) and (6), we arrive at the activation-energy dependence on the external force in the sudden regime within the linear-coupling approximation:

EAf(sudden)=EA−f(q−q0)−12f2k.(10)E_{\mathrm{A}}^{f}(\mathrm{sudden}) = E_{\mathrm{A}} - f(q-q_0) - \frac{1}{2}\frac{f^2}{k}. \tag*{(10)}

The second term is present only in the sudden regime. It is more sensitive to the force magnitude than the barrier lowering in the linear Bell picture, although the more realistic model of Konda et al. assumes a quadratic dependence of the activation energy on the external force[15].

Generalization to Polyatomic Molecules

For a polyatomic molecule, the FM-PES in the basis of normal modes q\mathbf{q} reads

V(q)=12(q−q0)Tk(q−q0)−fT(q−q0),(11)V(\mathbf{q}) = \frac{1}{2}(\mathbf{q}-\mathbf{q}_0)^{\mathsf{T}}\mathbf{k}(\mathbf{q}-\mathbf{q}_0) - \mathbf{f}^{\mathsf{T}}(\mathbf{q}-\mathbf{q}_0), \tag*{(11)}

where k\mathbf{k} is the diagonal force-constant matrix and f\mathbf{f} is the force vector in the normal-coordinate frame.

The force-shifted minimum of the PES becomes

q0f=q0+k−1f,(12)\mathbf{q}_0^{f} = \mathbf{q}_0 + \mathbf{k}^{-1}\mathbf{f}, \tag*{(12)}

where k−1\mathbf{k}^{-1} is also known as the compliance matrix.

The reorganization energy is a sum of contributions from the individual vibrational modes:

ER=12∑iωi2δqi2=∑iSiℏωi=∑iERi.(13)E_{\mathrm{R}} = \frac{1}{2}\sum_i \omega_i^2 \delta q_i^2 = \sum_i S_i\hbar\omega_i = \sum_i E_{\mathrm{R}}^i. \tag*{(13)}

This equation is equivalent to the central equation of the JEDI analysis written in normal coordinates. As established previously, the strain energy in JEDI equals the linear-coupling reorganization energy ERE_{\mathrm{R}}. This highlights the quantum nature of the molecule, in which ERE_{\mathrm{R}} has a discrete spectrum for each particular external force f\mathbf{f}, with the values of EiE_i forming a Franck–Condon profile.

The equivalence of strain energy and reorganization energy allows the apparatus of linear-coupling Franck–Condon theory to be applied not only to sudden-force cases, but also to fixed-force cases and static strain analysis, because the strain energy can be expressed in terms of Huang–Rhys factors.

The magnitude of the Huang–Rhys factor SiS_i for each vibration depends on the force magnitude and its direction, through the projection of the force onto the normal mode. It may happen that the external force f\mathbf{f} has no component along the normal mode corresponding to the reaction coordinate, or that it compresses the bond, thereby increasing the barrier EAE_{\mathrm{A}} rather than decreasing it. Even in this case, sudden force application can still accelerate the reaction by creating vibrational excitation, that is, by heating the molecule. Although the harmonic model considered here cannot per se explain intramolecular vibrational energy redistribution, such redistribution will still occur from excited modes with large force projections into reaction-relevant modes that are not directly excited by the initial force application.

Time-Dependent Theory

Time-Dependent Hamiltonian and Basis

If the external force is time-dependent, the Hamiltonian takes the form

H^(t)=p^22μ+12k(q−q0)2⏟H^0−f(t)(q−q0),(14)\hat{H}(t) = \frac{\hat{p}^{2}}{2\mu} + \underbrace{\frac{1}{2}k(q-q_0)^2}_{\hat{H}_0} - f(t)(q-q_0), \tag*{(14)}

where H^0\hat{H}_0 is the unperturbed harmonic-oscillator Hamiltonian. (14) defines the well-known problem of the driven harmonic oscillator.

Although its solution is generally known[39], we present it here using notation tailored to mechanochemical applications. The force ramps from f(−∞)=0f(-\infty)=0 to f(+∞)=fmax⁡f(+\infty)=f_{\max}. In the sudden-force regime, the force is switched on immediately as a step function, whereas in the fixed-force regime it increases adiabatically slowly. These are limiting cases; here we solve the problem for a general force profile f(t)f(t).

Following the previous derivation [Eq. (2)], the instantaneous FM-PES minimum is

q0f(t)=q0+f(t)k.(15)q_0^{f(t)} = q_0 + \frac{f(t)}{k}. \tag*{(15)}

The final FM-PES at f=fmax⁡f=f_{\max} has eigenstates χmf\chi_m^f with total equilibrium displacement

δq=fmax⁡k,Δ=μωℏ δq.(16)\delta q = \frac{f_{\max}}{k}, \qquad\Delta= \sqrt{\frac{\mu\omega}{\hbar}}\,\delta q. \tag*{(16)}

We are interested in the vibrational distribution over the states χmf\chi_m^f. For a driven harmonic system, this distribution is Poissonian[40]:

Pm=e−SeffSeffmm!,⟨m⟩=Seff,(17)P_m=e^{-S_{\mathrm{eff}}}\frac{S_{\mathrm{eff}}^m}{m!}, \qquad\langle m\rangle=S_{\mathrm{eff}}, \tag*{(17)}

fully parameterized by SeffS_{\mathrm{eff}}. The static result gives the sudden-limit value for fmax⁡f_{\max}.

It is convenient to write the Hamiltonian in second quantization. We introduce the bosonic ladder operators

a^=μω2ℏ[(q−q0)+ip^μω],q−q0=ℏ2μω(a^+a^†),(18)\hat{a}=\sqrt{\frac{\mu\omega}{2\hbar}}\left[(q-q_0)+\frac{i\hat{p}}{\mu\omega}\right], \qquad q-q_0=\sqrt{\frac{\hbar}{2\mu\omega}}(\hat{a}+\hat{a}^{\dagger}), \tag*{(18)}

which give the Hamiltonian

H^(t)=ℏω(a^†a^+12)−f(t)ℏ2μω(a^+a^†).(19)\hat{H}(t)=\hbar\omega\left(\hat{a}^{\dagger}\hat{a}+\frac{1}{2}\right)-f(t)\sqrt{\frac{\hbar}{2\mu\omega}}(\hat{a}+\hat{a}^{\dagger}). \tag*{(19)}

Effective Huang–Rhys Factor

The FM-PES Hamiltonian has the same force constant kk and frequency ω\omega as the original, but is shifted by δq\delta q. Its ladder operator is b^=a^−Δ\hat{b}=\hat{a}-\Delta.

The FM-PES ground state ∣χ0f⟩|\chi_0^f\rangle is the vacuum state of b^\hat{b}:

b^∣χ0f⟩=0⟹a^∣χ0f⟩=Δ∣χ0f⟩,(20)\hat{b}|\chi_0^f\rangle=0 \Longrightarrow\hat{a}|\chi_0^f\rangle=\Delta|\chi_0^f\rangle, \tag*{(20)}

so ∣χ0f⟩|\chi_0^f\rangle is a coherent state of a^\hat{a} with amplitude Δ\Delta:

∣χ0f⟩=∣Δ⟩=D^(Δ)∣χ0⟩,D^(Δ)=exp⁡(Δa^†−Δ∗a^),(21)|\chi_0^f\rangle=|\Delta\rangle=\hat{D}(\Delta)|\chi_0\rangle,\qquad\hat{D}(\Delta)=\exp(\Delta\hat{a}^{\dagger}-\Delta^*\hat{a}), \tag*{(21)}

where ∣χ0⟩|\chi_0\rangle is the ground vibrational state of the original, force-free PES and D^(Δ)\hat{D}(\Delta) is the displacement operator. All FM-PES Fock states ∣χmf⟩|\chi_m^f\rangle are created from this coherent ground state by repeated application of b^†=a^†−Δ\hat{b}^{\dagger}=\hat{a}^{\dagger}-\Delta:

∣χmf⟩=(b^†)mm!∣χ0f⟩=(a^†−Δ)mm!D^(Δ)∣χ0⟩.(22)|\chi_m^f\rangle=\frac{(\hat{b}^{\dagger})^m}{\sqrt{m!}}|\chi_0^f\rangle=\frac{(\hat{a}^{\dagger}-\Delta)^m}{\sqrt{m!}}\hat{D}(\Delta)|\chi_0\rangle. \tag*{(22)}

Since H^(t)\hat{H}(t) is linear in a^\hat{a} and a^†\hat{a}^{\dagger}, the driven state remains a coherent state ∣Δ(t)⟩|\Delta(t)\rangle of a^\hat{a} at all times:

a^∣Δ(t)⟩=Δ(t)∣Δ(t)⟩,(23)\hat{a}|\Delta(t)\rangle=\Delta(t)|\Delta(t)\rangle, \tag*{(23)}

where Δ(t)\Delta(t) is the Schrödinger-picture coherent amplitude. Moving to the interaction picture removes the free precession:

Δ~(t)=Δ(t)eiωt.(24)\widetilde{\Delta}(t)=\Delta(t)e^{i\omega t}. \tag*{(24)}

For the FM-PES, the ground-state amplitude in the interaction picture is

Δ~FM(t)=Δeiωt.(25)\widetilde{\Delta}_{\mathrm{FM}}(t)=\Delta e^{i\omega t}. \tag*{(25)}

The Huang–Rhys factor is related to the displacement parameter in the corresponding displacement operator [Eq. (4)][41] and measures vibrational excitation on the final FM-PES relative to its ground state ∣χ0f⟩|\chi_0^f\rangle. Here it is the squared distance between the driven amplitude and the FM-PES ground-state amplitude, both in the interaction picture:

Seff=lim⁡t→∞∣Δ~(t)−Δ~FM(t)∣2=lim⁡t→∞∣Δ~(t)−Δeiωt∣2.(26)S_{\mathrm{eff}}=\lim_{t\to\infty}\left|\widetilde{\Delta}(t)-\widetilde{\Delta}_{\mathrm{FM}}(t)\right|^2=\lim_{t\to\infty}\left|\widetilde{\Delta}(t)-\Delta e^{i\omega t}\right|^2. \tag*{(26)}

Solving the Heisenberg equation of motion, as detailed in Appendix A, gives the central result:

Seff(ω)=∣f˙~(ω)∣22μℏω3,f˙~(ω)=∫−∞∞f˙(t)eiωt dt.(27)S_{\mathrm{eff}}(\omega)=\frac{\left|\widetilde{\dot{f}}(\omega)\right|^2}{2\mu\hbar\omega^3},\qquad\widetilde{\dot{f}}(\omega)=\int_{-\infty}^{\infty}\dot{f}(t)e^{i\omega t}\,dt. \tag*{(27)}

The excitation is governed by the power spectrum of the force derivative, filtered by ω−3\omega^{-3}. Only the time variation of the force produces vibrational excitation on the FM-PES.

Limiting and Intermediate Cases

In the sudden-force case, f(t)f(t) is a Heaviside step function:

f(t)=fmax⁡Θ(t),Θ(t)={0,t<0,1,t>0.(28)f(t)=f_{\max}\Theta(t),\qquad\Theta(t)=\begin{cases}0,&t<0,\\1,&t>0.\end{cases} \tag*{(28)}

Its time derivative is

f˙(t)=fmax⁡δ(t)⟹f˙~(ω)=fmax⁡.(29)\dot{f}(t)=f_{\max}\delta(t)\Longrightarrow\widetilde{\dot{f}}(\omega)=f_{\max}. \tag*{(29)}

Therefore,

Seffsudden(ω)=fmax⁡22μℏω3=S,(30)S_{\mathrm{eff}}^{\mathrm{sudden}}(\omega)=\frac{f_{\max}^2}{2\mu\hbar\omega^3}=S, \tag*{(30)}

recovering the static Franck–Condon result of Eq. (7).

In the adiabatic limit, the ramp is infinitely slow, so f˙(t)→0\dot{f}(t)\to0 at all finite frequencies:

f˙~(ω)→0⟹Seffadiabatic(ω)=0.(31)\widetilde{\dot{f}}(\omega)\to0\Longrightarrow S_{\mathrm{eff}}^{\mathrm{adiabatic}}(\omega)=0. \tag*{(31)}

The molecule follows the FM-PES adiabatically, with no vibrational excitation.

For the physically motivated shock-front profile with rise time τ\tau (see Figure 2 and Table I):

Force-ramp profiles

Figure 2. Force-ramp profiles (upper panel), shown as error functions, and their time derivatives (lower panel) for different regimes: gray, sudden; blue, adiabatic or fixed; orange, intermediate.

f(t)=fmax⁡2[1+erf⁡(t2τ)].(32)f(t)=\frac{f_{\max}}{2}\left[1+\operatorname{erf}\left(\frac{t}{\sqrt{2}\tau}\right)\right]. \tag*{(32)}

Its time derivative is a Gaussian,

f˙(t)=fmax⁡2πτexp⁡(−t22τ2),(33)\dot{f}(t) = \frac{f_{\max}}{\sqrt{2\pi}\tau} \exp\left(-\frac{t^2}{2\tau^2}\right), \tag*{(33)}

with Fourier transform

f~(ω)=fmax⁡exp⁡(−ω2τ22).(34)\widetilde{f}(\omega) = f_{\max} \exp\left(-\frac{\omega^2\tau^2}{2}\right). \tag*{(34)}

Substitution into Eq. (27) gives

Seff(ω)=Sexp⁡(−ω2τ2).(35)S_{\mathrm{eff}}(\omega) = S\exp\left(-\omega^2\tau^2\right). \tag*{(35)}
LimitConditionSeff(ω)S_{\mathrm{eff}}(\omega)
Suddenωτ→0\omega\tau\to0SS
Generalωτ\omega\tau finiteSe−ω2τ2S e^{-\omega^2\tau^2}
Adiabaticωτ→∞\omega\tau\to\infty00

Table 1. Limiting behavior of the effective Huang–Rhys factor for the error-function force ramp.

The Gaussian filter e−ω2τ2e^{-\omega^2\tau^2} is the central result. High-frequency modes, for which ωτ≫1\omega\tau\gg1, are exponentially suppressed, whereas low-frequency modes, for which ωτ≪1\omega\tau\ll1, retain the sudden-limit excitation to a large degree. For a laser-shock front with τ∼50\tau\sim50–100fs100\mathrm{fs}, this suppression becomes significant already around ν~∼300\tilde{\nu} \sim300–500cm−1500\mathrm{cm}^{-1}, confirming that only low-frequency normal modes are efficiently excited by the mechanical shock.

Applications

Numerical Estimates with Molecular Parameters

Here, we compute effective Huang–Rhys factors for molecular vibrations in intermediate force-ramp regimes using parameters derived from LSP experiments. The parameters for the error-function force-ramp profile, fmax⁡f_{\max} and τ\tau, are estimated in Appendix B. Because the kinetics are exponential, the logarithmic scale used in the right-hand panels gives an indication of the vibrational contributions to the activation energy in Eq. (9).

Results for several parameter sets are shown in Figure 3. At moderate force time ramp values lower-frequency vibrations acquire significant vibrational excitation, namely bending, torsional and acoustic modes in the solid state. This excess vibrational energy can be transferred to bond-stretching vibrations through mechanisms of intermolecular vibrational redistribution; estimating such effects requires quantum-dynamics simulations and is system dependent.

Effective Huang–Rhys factors as a function of molecular vibrational frequency for different force values and ramp times

Figure 3. Effective Huang–Rhys factors as a function of molecular vibrational frequency for different force values and ramp times.

Stretching nn-Butane

Next, we present an application of the theory to the mechanical stretching of nn-butane, which was considered in previous works[25, 26]. For consistency, we used the same protocol and quantum-chemical approach: DFT at the B3LYP[42, 43]/6-311G**[44] level. The geometry was first optimized without constraints and in the absence of applied forces, followed by a standard Wilson harmonic analysis[45] using PySCF[46]. We then applied an external stretching force along the line connecting the two terminal carbon atoms (see Figure 4(a)). The force magnitude was 2.95 nN, corresponding to 0.75 of the critical value of 3.9 nN determined elsewhere[25]. This was the maximum force value used for the static analysis in the first study of Franck–Condon theory[26].

Application to *n*-butane

Figure 4. Application to n-butane: (a) stretching of the n-butane molecule by external force of 2.95 nN; (b) effective Huang–Rhys factors of the selected normal modes; (c) relative half-life for different force-ramp times.

Time-dependent Huang–Rhys factors were subsequently calculated for ramp times of 10 and 20 fs using Equation (35). The sudden-limit values were evaluated using Equation (7), with the force components determined by projecting the external force onto the normal modes.

The results are shown in Figure 4. As expected, significant Huang–Rhys factors are obtained for only two modes: the symmetric chain-bending (deformation) mode, with ν5=424 cm−1\nu_5 = 424\ \mathrm{cm}^{-1}, and the symmetric stretching mode, with ν8=840 cm−1\nu_8 = 840\ \mathrm{cm}^{-1}. For nonzero τ\tau, however, the contribution of the latter becomes negligible. We observe a drastic decrease in SS with increasing ramp time.

To illustrate the effect of these vibrational excitations on the reaction kinetics, relative first-order reaction half-lives were calculated using Equation (9) for the activation energy and the Arrhenius equation, assuming that both the pre-exponential factor and the static energy barrier are independent of the ramp time. Owing to the exponential dependence in the Arrhenius equation, even a modest additional activation can lead to considerable acceleration, as shown in panel (c) of Figure 4.

Conclusions and Outlook

Introducing the linear-coupling approximation into the Franck–Condon theory of quantum mechanochemistry, previously formulated for sudden force application, allows this vibronic-like picture to be generalized to adiabatic (fixed-force) and intermediate force-application regimes. The main conceptual conclusion of this work is the dependence of mechanochemical activation efficiency not only on the magnitude and direction of the external force, but also on the rate and time profile of its application. These may result in additional vibrational excitations accelerating mechanochemical reactions: such an effect is in principle measurable provided force-ramp profile f(t)f(t), direction and magnitude are controllable. Figure 4 illustrates these effects for nn-butane.

The amount of additional vibrational excitation, or heating, arising from external mechanical force is characterized by the effective time-dependent Huang–Rhys factor, SeffS_{\mathrm{eff}}, which depends on the time-ramp profile of the force f(t)f(t) and its time derivative f˙(t)\dot{f}(t), sometimes called the yank. In the limiting cases of fixed and sudden force action, SeffS_{\mathrm{eff}} reaches its extreme values: zero and its maximum value, respectively.

For fixed force, the linear-coupling approximation for the harmonic PES is exactly equivalent to the force-modified PES model with a linear force term, which is the most commonly used approach in mechanochemical theory and applications. This makes it possible to establish terminological connections between spectroscopy and mechanochemistry. In particular, the strain energy is equivalent to the reorganization energy of a vibronic transition, and also to that associated with charge transfer. The contribution of each vibrational mode is given by the corresponding Huang–Rhys factor.

Estimation of SeffS_{\mathrm{eff}} for typical LSP experiments and computing those for stretching of nn-butane show that direct excitation of valence-bond-stretching vibrations is unlikely under currently accessible ramp times and force magnitudes. Such a regime may remain futuristic; therefore, care must be taken when relating theoretical calculations that implicitly adopt sudden force application to experiments. On the other hand, bending, torsional and acoustic vibrations, as well as slow modes in biomolecules[27], are likely to undergo non-negligible excitation and can affect the kinetics through intermolecular vibrational redistribution[47]. Other types of experiments, such as femtosecond shocks[48], different ramp functions, and advances in LSP techniques may reveal a different picture. More realistic theoretical models incorporating anharmonicity and quantum dynamics will likely show a more nuanced behavior.

At this point a special note must be made regarding mechanochemical reactions in solution[49]. Due to high collision rates vibrational energy dissipates rapidly in this environment, which may mitigate vibrational heating discussed here. On the other hand, energy transfer from excited solvent molecules to the substrate is also possible opening interesting opportunities for theorizing and practice.

I believe that this simple model will be useful for qualitative estimation of finite-time force-application effects in mechanochemistry, whereas the terminological and conceptual conclusions may stimulate further understanding of the field.

Acknowledgments

This work has received funding from the German Federal Ministry of Research, Technology and Space (BMFTR) within the PhoQuant project (Grant No. 13N16103).

Author Declarations

Conflict of Interest

The author has no conflicts to disclose.

Author Contributions

Vladimir V. Rybkin. Conceptualization, Formal analysis, Investigation, Methodology, Visualization, Writing – original draft, Writing – review & editing.

Data Availability

Data sharing is not applicable to this article as no new data were created or analyzed in this study.

Derivation of the Effective Huang–Rhys Factor Using the Heisenberg Equation of Motion

The Heisenberg equation for the annihilation operator, da^/dt=(i/ℏ)[H^(t),a^]d\hat{a}/dt=(i/\hbar)[\hat{H}(t),\hat{a}], takes the form

da^dt=−iωa^+if(t)2μℏω(36)\frac{d\hat{a}}{dt}=-i\omega\hat{a}+\frac{if(t)}{\sqrt{2\mu\hbar\omega}} \tag*{(36)}

after using the standard commutator relations [a^†,a^]=−a^[\hat{a}^{\dagger},\hat{a}]=-\hat{a} and [a^†,a^]=−1[\hat{a}^{\dagger},\hat{a}]=-1.

The coherent amplitude is the expectation value of the annihilation operator, so its time derivative is

Δ˙(t)=−iωΔ(t)+if(t)2μℏω.(37)\dot{\Delta}(t)=-i\omega\Delta(t)+\frac{if(t)}{\sqrt{2\mu\hbar\omega}}. \tag*{(37)}

Differentiating the interaction-picture amplitude and substituting Eq. (37), we obtain

Δ~˙(t)=Δ˙(t)eiωt+iωΔ(t)eiωt=(−iωΔ(t)+if(t)2μℏω)eiωt+iωΔ(t)eiωt.(38)\begin{aligned} \dot{\widetilde{\Delta}}(t)=\dot{\Delta}(t)e^{i\omega t}+i\omega\Delta(t)e^{i\omega t} \\ &=\left(-i\omega\Delta(t)+\frac{if(t)}{\sqrt{2\mu\hbar\omega}}\right)e^{i\omega t}+i\omega\Delta(t)e^{i\omega t}. \tag*{(38)} \end{aligned}

The ±iωΔ(t)\pm i\omega\Delta(t) terms cancel exactly:

Δ~˙(t)=if(t)2μℏωeiωt.(39)\dot{\widetilde{\Delta}}(t)=\frac{if(t)}{\sqrt{2\mu\hbar\omega}}e^{i\omega t}. \tag*{(39)}

Integrating from −∞-\infty to tt with the initial condition Δ~(−∞)=0\widetilde{\Delta}(-\infty)=0 gives

Δ~(t)=i2μℏω∫−∞tf(t′)eiωt′ dt′.(40)\widetilde{\Delta}(t)=\frac{i}{\sqrt{2\mu\hbar\omega}}\int_{-\infty}^{t}f(t')e^{i\omega t'}\,dt'. \tag*{(40)}

Integration by parts yields

∫−∞tf˙(t′)eiωt′ dt′=f(t′)eiωt′iω∣−∞t−1iω∫−∞tf˙(t′)eiωt′ dt′.(41)\int_{-\infty}^{t} \dot{f}(t')e^{i\omega t'}\,dt'=\left.\frac{f(t')e^{i\omega t'}}{i\omega}\right|_{-\infty}^{t}-\frac{1}{i\omega}\int_{-\infty}^{t}\dot{f}(t')e^{i\omega t'}\,dt'. \tag*{(41)}

The lower limit vanishes because f(−∞)=0f(-\infty)=0. Using i/2μℏω⋅1/(iω)=1/2μℏω3i/\sqrt{2\mu\hbar\omega}\cdot1/(i\omega)=1/\sqrt{2\mu\hbar\omega^3}, we arrive at

Δ~(t)=f(t)eiωtω2μℏω+12μℏω3∫−∞tf˙(t′)eiωt′ dt′.(42)\widetilde{\Delta}(t)=\frac{f(t)e^{i\omega t}}{\omega\sqrt{2\mu\hbar\omega}}+\frac{1}{\sqrt{2\mu\hbar\omega^3}}\int_{-\infty}^{t}\dot{f}(t')e^{i\omega t'}\,dt'. \tag*{(42)}

Substituting Eq. (42) into Eq. (26):

Δ~(t)−Δeiωt=(f(t)−fmax⁡)eiωtω2μℏω+12μℏω3∫−∞tf˙(t′)eiωt′ dt′.(43)\widetilde{\Delta}(t)-\Delta e^{i\omega t}=\frac{(f(t)-f_{\max})e^{i\omega t}}{\omega\sqrt{2\mu\hbar\omega}}+\frac{1}{\sqrt{2\mu\hbar\omega^3}}\int_{-\infty}^{t}\dot{f}(t')e^{i\omega t'}\,dt'. \tag*{(43)}

As t→∞t\to\infty, f(t)→fmax⁡f(t)\to f_{\max}, so the first term vanishes. The remaining integral converges to the full Fourier transform of f˙\dot{f}, giving our central result:

Seff(ω)=∣f˙~(ω)∣22μℏω3,f˙~(ω)=∫−∞∞f˙(t)eiωt dt.(44)S_{\mathrm{eff}}(\omega)=\frac{\left|\widetilde{\dot{f}}(\omega)\right|^2}{2\mu\hbar\omega^3},\qquad\widetilde{\dot{f}}(\omega)=\int_{-\infty}^{\infty}\dot{f}(t)e^{i\omega t}\,dt. \tag*{(44)}

Estimation of Shock Force and Rise Time

The mechanical force acting on a single molecule during laser shock peening (LSP) can be estimated from the shock pressure PP and an effective molecular cross-sectional area A∼d2A\sim d^2, where dd is a typical intermolecular distance:

fmax⁡=Pd2.(45)f_{\max}=Pd^2. \tag*{(45)}

For a small molecule with d∼3 A˚d\sim3\,\text{\AA} and pressures characteristic of nanosecond LSP, P∼1P\sim1–10 GPa10\,\text{GPa}[50], and femtosecond LSP, P∼100P\sim100–$300,\text{GPa}, {xref}Eq. (B1)` gives the approximate range

fmax⁡∼0.09–25 nN.(46)f_{\max}\sim0.09\text{--}25\,\text{nN}. \tag*{(46)}

The shock-front rise time can be estimated from the shock-front thickness dfrontd_{\mathrm{front}} and the shock propagation velocity vsv_s. Experimental measurements suggest that the shock front spans only a few molecular diameters[51], $d_{\mathrm{front}}\sim1,\text{nm}, whereas typical shock velocities in condensed matter are $v_s\sim3$–$6\,\text{km}\,\text{s}^{-1}${citep}51,52`. The time for the shock front to sweep across a molecule is therefore

τ=dfrontvs∼1×10−9 m5×103 m s−1=200 fs.(47)\tau=\frac{d_{\mathrm{front}}}{v_s}\sim\frac{1\times10^{-9}\,\text{m}}{5\times10^{3}\,\text{m}\,\text{s}^{-1}}=200\,\text{fs}. \tag*{(47)}

References

  1. [1]1 M. K. Beyer and H. Clausen-Schaumann, “Mechanochemistry: The mechanical activation of covalent bonds,” Chemical Reviews 105, 2921–2948 (2005).pubmed.ncbi.nlm.nih.gov/16092823
  2. [2]2 S. Pagola, “Outstanding advantages, current drawbacks, and significant recent developments in mechanochemistry: A perspective view,” Crystals 13, 124 (2023).DOI
  3. [3]3 G. Cravotto, E. C. Gaudino, and P. Cintas, “On the mechanochemical activation by ultrasound,” Chemical Society Reviews 42, 7521–7534 (2013).DOI
  4. [4]Speight, I. R., Ardila-Fierro, K. J., Hernández, J. G., Emmerling, F., Michalchuk, A. A. L., García, F., Colacino, E., & Mack, J. (2025). Ball milling for mechanochemical reactions. Nature Reviews Methods Primers, 5, 29.
  5. [5]5 K. C. Neuman and A. Nagy, “Single-molecule force spectroscopy: optical tweezers, magnetic tweezers and atomic force microscopy,” Nature Methods 5, 491–505 (2008).DOI
  6. [6]6 M. Rief, M. Gautel, F. Oesterhelt, J. M. Fernandez, and H. E. Gaub, “Reversible unfolding of individual titin immunoglobulin domains by afm,” Science 276, 1109–1112 (1997).pubmed.ncbi.nlm.nih.gov/9148804
  7. [7]7 M. S. Z. Kellermayer, S. B. Smith, H. L. Granzier, and C. Bustamante, “Folding-unfolding transitions in single titin molecules characterized with laser tweezers,” Science 276, 1112–1116 (1997).pubmed.ncbi.nlm.nih.gov/9148805
  8. [8]Dlott, D. D. (2008). New developments in the physical chemistry of shock compression. Annual Review of Physical Chemistry, 57, 575–597.
  9. [9]Pairand, B. P., & Clauer, A. H. (1979). Laser generation of high-amplitude stress waves in materials. Journal of Applied Physics, 50, 1497–1502.
  10. [10]10 P. Peyre and R. Fabbro, “Laser shock processing: a review of the physics and applications,” Optical and Quantum Electronics 27, 1213–1229 (1995).DOI
  11. [11]Stauch, T., & Drew, A. (2016). Advances in quantum mechanochemistry: Electronic structure methods and force analysis. Chemical Reviews, 116, 14137–14180.
  12. [12]12 M. Alonso, T. Bettens, J. Eeckhoudt, P. Geerlings, and F. De Proft, “Wandering through quantum-mechanochemistry: from concepts to reactivity and switches,” Physical Chemistry Chemical Physics 26, 21–35 (2024).DOI
  13. [13]13 J. Ribas-Arino and D. Marx, “Covalent mechanochemistry: Theoretical concepts and computational tools with applications to molecular nanomechanics,” Chemical Reviews 112, 5412–5487 (2012).DOI
  14. [14]14 G. I. Bell, “Models for the specific adhesion of cells to cells,” Science 200, 618–627 (1978).pubmed.ncbi.nlm.nih.gov/347575
  15. [15]15 S. S. Konda, J. N. Brantley, C. W. Bielawski, and D. E. Makarov, “Chemical reactions modulated by mechanical stress: extended Bell theory,” The Journal of Chemical Physics 135, 164103 (2011).DOI
  16. [16]16 M. K. Beyer, “The mechanical strength of a covalent bond calculated by density functional theory,” The Journal of Chemical Physics 112, 7307–7312 (2000).DOI
  17. [17]17 J. Ribas-Arino, M. Shiga, and D. Marx, “Understanding covalent mechanochemistry,” Angewandte Chemie International Edition 48, 4190–4193 (2009).DOI
  18. [18]Stauch, T., & Drew, A. (2014). A quantitative quantum-chemical analysis tool for the distribution of mechanical force in molecules. The Journal of Chemical Physics, 140, 134107.
  19. [19]19 W. Quapp and J. M. Bofill, “A contribution to a theory of mechanochemical pathways by means of newton trajectories,” Theoretical Chemistry Accounts 135, 113 (2016).DOI
  20. [20]20 W. Quapp, J. M. Bofill, and J. Ribas-Arino, “Analysis of the acting forces in a theory of catalysis and mechanochemistry,” The Journal of Physical Chemistry A 121, 2820–2838 (2017).DOI
  21. [21]21 K. Wolinski and J. Baker, “Theoretical predictions of enforced structural changes in molecules,” Molecular Physics 107, 2403–2417 (2009).DOI
  22. [22]22 T. Stauch, “A mechanochemical model for the simulation of molecules and molecular crystals under hydrostatic pressure,” The Journal of Chemical Physics 153, 134503 (2020).DOI
  23. [23]23 R. Weiß, F. Zeller, and T. Neudecker, “Calculating high-pressure vibrational frequencies analytically with the extended hydrostatic compression force field approach,” The Journal of Chemical Physics 160, 084101 (2024).DOI
  24. [24]Stauch, T., & Drew, A. (2017). Force-induced retro-click reaction of triazoles competes with adjacent single-bond rupture. Chemical Science, 8, 5567–5575.
  25. [25]Smal, H. S., Rybkin, V. V., Klopper, W., Helgaker, T., & Uggerud, E. (2014). Mechanochemistry: The effect of dynamics. The Journal of Chemical Chemistry A, 118, 7683–7694.
  26. [26]26 V. V. Rybkin, “Franck–Condon theory of quantum mechanochemistry,” The Journal of Physical Chemistry A 121, 5758–5762 (2017).DOI
  27. [27]27 J. Sheridan, F. Gräter, and C. Daday, “How fast is too fast in force-probe molecular dynamics simulations?” The Journal of Physical Chemistry B 123, 3658–3664 (2019).DOI
  28. [28]Ong, M. T., Leiding, J., Tao, H., Virshup, A. M., & Martínez, T. J. (2009). First. Journal of the American Chemical Society, 131, 6377–6379.
  29. [29]29 P. Dopieralski, J. Ribas-Arino, P. Anjukandi, M. Krupicka, J. Kiss, and D. Marx, “The janus-faced role of external forces in mechanochemical disulfide bond cleavage,” Nature Chemistry 5, 685–691 (2013).DOI
  30. [30]30 G. S. Kedziora, J. Moller, R. Berry, and D. Nepal, “Ab initio molecular dynamics modeling of single polyethylene chains: Scission kinetics and influence of radical under mechanical strain,” The Journal of Chemical Physics 155, 024102 (2021).DOI
  31. [31]31 H. S. Smalø and E. Uggerud, “Breaking covalent bonds using mechanical force, which bond breaks?” Molecular Physics 111, 1563–1573 (2013).DOI
  32. [32]32 M. F. Pill, A. L. L. East, D. Marx, M. K. Beyer, and H. Clausen-Schaumann, “Mechanical activation drastically accelerates amide bond hydrolysis, matching enzyme activity,” Angewandte Chemie International Edition 58, 9787–9790 (2019).DOI
  33. [33]33 M. Frank-Kamenetskii and A. V. Lukashin, “Electron-vibrational interactions in polyatomic molecules,” Soviet Physics Uspekhi 18, 391–409 (1975).DOI
  34. [34]34 E. Condon, “A theory of intensity distribution in band systems,” Physical Review 28, 1182 (1926).DOI
  35. [35]Huang, K., & Rhys, A. F. (1950). Theory of light absorption and non-radiative transitions in f-centres. Proceedings of the Royal Society of London. Series A. Mathematical and Physical Sciences, 204, 406–423.
  36. [36]36 R. A. Marcus, “On the theory of oxidation-reduction reactions involving electron transfer. i,” The Journal of Chemical Physics 24, 966–978 (1956).DOI
  37. [37]Rusanov, V. D., Fridman, A. A., & Sholin, G. V. (1981). The physics of a chemically active plasma with nonequilibrium vibrational excitation of molecules. Uspekhi Fizicheskikh Nauk, 134, 185–235.
  38. [38]38 S. O. Macheret and J. W. Rich, “Nonequilibrium dissociation rates behind strong shock waves: classical model,” Chemical Physics 174, 25–43 (1993).DOI
  39. [39]Cohen-Tannoudji, C., Diu, B., & Laloë, F. (1977). Quantum Mechanics, Volume 1: Basic Concepts, Tools, and Applications.
  40. [40]LaPierre, R. (2018). Getting started in quantum optics.
  41. [41]41 J.-L. Eickmann, K.-H. Luo, M. Roiz, J. Lammers, S. Atzeni, C. Pandey, F. Lütkewitte, R. G. Shirazi, F. Schlue, B. Brecht, V. V. Rybkin, M. Stefszky, and C. Silberhorn, “Bridging chemistry and gaussian boson sampling: a photonic hierarchy of approximations for molecular vibronic spectra,” npj Quantum Information 12, 89 (2026).DOI
  42. [42]42 A. D. Becke, “Density-functional thermochemistry. iii. the role of exact exchange,” The Journal of Chemical Physics 98, 5648–5652 (1993).DOI
  43. [43]43 C. Lee, W. Yang, and R. G. Parr, “Development of the colle–salvetti correlation-energy formula into a functional of the electron density,” Physical Review B 37, 785–789 (1988).pubmed.ncbi.nlm.nih.gov/9944570
  44. [44]44 R. Krishnan, J. S. Binkley, R. Seeger, and J. A. Pople, “Self-consistent molecular orbital methods. xx. a basis set for correlated wave functions,” The Journal of Chemical Physics 72, 650–654 (1980).DOI
  45. [45]45 E. B. Wilson, J. C. Decius, and P. C. Cross, Molecular Vibrations: The Theory of Infrared and Raman Vibrational Spectra (McGraw–Hill, New York, 1955) p. 388.DOI
  46. [46]Sun, Q., Berkelbach, T. C., Blunt, N. S., Booth, G. H., Guo, Z., Li, J., Lu, J. D., McClain, J. C., Sayfutyarova, S., Sharma, S., Wouters, S., & Chan, G. K.-L. (2018). PySCF: The python-based simulations of chemistry framework. WIREs Computational Molecular Science, 8, e1340.
  47. [47]47 A. Tokmakoff, M. D. Fayer, and D. D. Dlott, “Chemical reaction initiation and hot spot formation in shocked energetic molecular materials,” The Journal of Physical Chemistry 97, 1901–1913 (1993).DOI
  48. [48]Evans, R., Badger, A. D., Fallies, M., Mahdieh, M., Hall, T. A., Audibert, J., Geindre, J.-C., Gauthier, A., Mysmyrowicz, G., Grillot, F., & Antonetti, M. T. (1996). Time- and space-resolved optical probing of femtosecond-laser-driven shock waves in aluminum. Physical Review Letters, 77, 3359–3362.
  49. [49]49 S. L. James, C. J. Adams, C. Bolm, D. Braga, P. Collier, T. Friščić, F. Grepioni, K. D. M. Harris, G. Hyett, W. Jones, A. Krebs, J. Mack, L. Maini, A. G. Orpen, I. P. Parkin, W. C. Shearouse, J. W. Steed, and K. C. Waddell, “Mechanochemistry,” Chemical Society Reviews 41, 413–447 (2012).DOI
  50. [50]50 R. Fabbro, J. Fournier, P. Ballard, D. Devaux, and J. Virmont, “Physical study of laser-produced plasma in confined geometry,” Journal of Applied Physics 68, 775–784 (1990).DOI
  51. [51]51 D. D. Dlott, “The new wave in shock waves,” The Journal of Physical Chemistry B 102, 2121–2130 (1998).DOI

Paper details

vlnc:W_xux22n6w
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 0

downloads · 0

A view counts once per browser per day, a download each time the PDF is requested.

  • Physics and Astronomy

Not recorded for this paper

View statement
Looking for related papers…
Contents