Impact Factor 3.560 | CiteScore 3.1
More on impact ›

ORIGINAL RESEARCH article

Front. Phys., 11 January 2021 | https://doi.org/10.3389/fphy.2020.582992

Irreversibility in Active Matter: General Framework for Active Ornstein-Uhlenbeck Particles

  • 1Fakultät für Physik, Universität Bielefeld, Bielefeld, Germany
  • 2Nordita, Royal Institute of Technology and Stockholm University, Stockholm, Sweden

Active matter systems are driven out of equilibrium by conversion of energy into directed motion locally on the level of the individual constituents. In the spirit of a minimal description, active matter is often modeled by so-called active Ornstein-Uhlenbeck particles an extension of passive Brownian motion where activity is represented by an additional fluctuating non-equilibrium “force” with simple statistical properties (Ornstein-Uhlenbeck process). While in passive Brownian motion, entropy production along trajectories is well-known to relate to irreversibility in terms of the log-ratio of probabilities to observe a certain particle trajectory forward in time in comparison to observing its time-reversed twin trajectory, the connection between these concepts for active matter is less clear. It is therefore of central importance to provide explicit expressions for the irreversibility of active particle trajectories based on measurable quantities alone, such as the particle positions. In this technical note we derive a general expression for the irreversibility of AOUPs in terms of path probability ratios (forward vs. backward path), extending recent results from [PRX 9, 021009 (2019)] by allowing for arbitrary initial particle distributions and states of the active driving.

1. Introduction

Irreversible thermodynamic processes are characterized by a positive entropy change in their “Universe”, i.e., in the combined system of interest and its environment [1]. In macroscopic (equilibrium) thermodynamics, where entropy is a state variable, this change usually refers to the difference between the entropy in the final state of the “Universe” reached at the end of the process and in the initial state from where it started. In small mesoscopic systems on the micro- and nanometer scale, such as a colloidal Brownian particle diffusing in an aqueous solution, it has been established within the framework of stochastic thermodynamics [26] that the total entropy change should be evaluated from the entropy produced in the system and in its thermal environment along the specific trajectory the system follows during the process. This procedure remains valid even when the system is far from equilibrium, for example due to persistent currents or because it is driven by an external protocol realizing the thermodynamic process. The omnipresence of thermal fluctuations on the mesoscopic scale leads to a distribution of possible paths the system can take to go from the initial to the final state, and, accordingly to a distribution of entropy changes. A central result in stochastic thermodynamics is that the total entropy change ΔS along a specific realization of the system path (divided by Boltzmann’s constant kB) equals the log-ratio of probabilities for observing that specific path vs. observing the same path in a time-reversed manner, i.e., traversing the same trajectory, but from the final state to the original initial state [25]. As a direct consequence, the total entropy change ΔS fulfills a so-called fluctuation theorem, exp(ΔS/kB)=1 (the angular brackets denote an average over all trajectories connecting the initial and final states), which can be viewed as a generalization of the second law of thermodynamics to the non-equilibrium realm when deviations from equilibrium are induced by externally applied forces or gradients.

A fundamentally different class of non-equilibrium systems are so-called “active particles”, like Janus colloids with catalytic surfaces or bacteria [711], which have the ability to locally convert energy into self-propulsion, i.e., they move independently of external forces or thermal fluctuations. The source of non-equilibrium is the energy-to-motion conversion process on the level of the individual particle. This out-of-equilibrium process produces entropy, but the various degrees of freedom maintaining the self-propulsion are usually not observable in typical experiments with active particles, such that this entropy production can in general not be quantified. Moreover, for the (collective) behavior of active particles emerging from self-propulsion, as described, e.g., in Refs. 1215, the details of the propulsion mechanism and the amount of dissipation connected with it are largely irrelevant. In analogy to the stochastic thermodynamics of passive Brownian particles, a central question in active matter is therefore how the path probabilities for translational degrees of freedom of the active particles and the associated log-ratio of forward vs. backward path probabilities is connected to irreversibility and entropy production [1618]. We remark that this is an ongoing debate [16, 1924] which we will not resolve here. Rather, we will provide a central step toward an understanding of the role of the path probability ratio in active matter by providing exact analytical expressions for a simple but highly successful and well-established [15, 2531] model of active matter, namely the active Ornstein-Uhlenbeck particle (AOUP) [16, 1821, 24, 3244]. In this model, self-propulsion is realized via a fluctuating “driving force” in the equations of motion [7, 10] with Gaussian distribution and exponential time-correlation (see Section 2.1). By integrating out these active fluctuations, we derive an explicit analytical expression for the path weight of an AOUP, valid for arbitrary values of the model parameters, arbitrary finite duration of the particle trajectory and arbitrary initial distributions of particle positions and active fluctuations (see Section 4). Using this path weight, we then derive the irreversibility measure in form of the log-ratio of forward vs. backward path probabilities and comment on its physical implications (Section 5). Before establishing these general results, we briefly recall earlier findings from Ref. 18 for independent initial conditions of particle positions and active fluctuations, see Section 3. We conclude with a short discussion in Section 6, including potential applications of our results.

2. Setup

2.1. Model

The model for an active Ornstein-Uhlenback particle (AOUP) consists in a standard overdamped Langevin equation for a passive Brownian particle at position x in d dimensions with an additional fluctuating force, which represents the active self-propulsion and which we denote by 2Daη(t),

x˙(t)=1γf(x(t),t)+2Daη(t)+2Dξ(t).(1)

Here, the dot denotes the time-derivative, γ is the viscous friction coefficient, f(x,t) represents externally applied forces (conservative or non-conservative, and possibly time-dependent). Furthermore, ξ(t) are mutually independent Gaussian white noise sources modeling thermally fluctuating forces with δ-correlation in time, i.e., ξi(t)=0, ξi(t)ξj(t)=δijδ(tt), and D is the particle diffusion coefficient, related to the temperature T of the thermal bath via Einstein’s relation D=kBT/γ. All bold-face letters represent d-dimensional vectors with components usually labeled by subscripts i, j, etc. In analogy to the thermal fluctuations, we denote the strength of the active fluctuations η(t) by 2Da with an active “diffusion coefficient” Da. For an AOUP, the active fluctuations follow a Gaussian process with exponential time-correlations, which can be generated by a so-called Ornstein-Uhlenbeck process,

η˙(t)=1τaη(t)+1τaζ(t),(2)

where τa is the correlation time of the active noise fluctuations, i.e.,

ηi(t)ηj(t)=δij2τae|tt|/τa.(3)

2.2. Central Quantity of Interest

Our central goal is to evaluate the path weight P[x¯|xi] for particle positions alone, conditioned on the initial position xi for an arbitrary initial distribution pi(ηi|xi) of the active fluctuations given the specific value xi. By definition, we can write this path weight as

P[x¯|xi]=Dη¯P[x¯,η¯|xi,ηi]pi(ηi|xi),(4)

where the path integral over η¯:={η(t)}t=τiτf includes the initial configuration ηi, whereas the notation η¯:={η(t)}t>τiτf denotes the same history of active fluctuations without the initial configuration ηi, and similarly for x¯:={x(t)}t>τiτf. Moreover,

P[x¯,η¯|xi,ηi]exp{τiτfdt[(x˙tvt2Daηt)24D+(τaη˙t+ηt)22+vt2]}(5)

is the standard Onsager-Machlup path weight [4547] for the joint process (x¯,η¯), where we use the shorthand notation vt=ft/γ=f(x(t),t)/γ and xtx(t), ηtη(t), etc. The technical challenge consists in performing the integral over the active fluctuations η¯ without explicitly specifying the initial distribution pi(ηi|xi).

3. Path Weight for Independent Initial Conditions

3.1. The Results From Ref. 18

We start by summarizing the main results from Ref. 18. In Ref. 18 we gave the path weight for trajectories x¯={x(t)}t=0τ, running from initial time τi=0 to final time τf=τ, assuming that the active noise is initially independent of the particle positions and in its steady state, i.e., pi(η0|x0)=pss(η0)=τa/πeτaη02. We found

P(0,τ]ind[x¯|x0]exp{14D0τdt0τdt[x˙tvt]T[δ(tt)DaDΓ[0,τ]ind(t,t)][x˙tvt]120τdtvt},(6)

with the memory kernel

Γ[0,τ]ind(t,t):=(12τa2λ)×κ+2eλ|tt|+κ2eλ(2τ|tt|)κ+κ[eλ(t+t)+eλ(2τtt)]κ+2κ2e2λτ,(7)

where λ:=1+Da/D/τa and κ±:=1±1+Da/D.

3.2. Stationary-State Scenario

If we have a trajectory x¯={x(t)}t=τiτf running from arbitrary times τi to τf instead, we can shift time as ttτi and identify τ=τfτi as the duration of the trajectory to convert (0,τ] path weights to those running from τi to τf. Performing these replacements, the memory kernel Eq. 7 turns into

Γ[τi,τf]ind(t,t):=(12τa2λ)×κ+2eλ|tt|+κ2eλ[2(τfτi)|tt|]κ+κ[eλ(t+t2τi)+eλ(2τftt)]κ+2κ2e2λ(τfτi).(8)

Consequently, the corresponding path weight for a trajectory starting at xi at time τi reads

P(τi,τf]ind[x¯|xi]exp{14Dτiτfdtτiτfdt[x˙tvt]T[δ(tt)DaDΓ[τi,τf]ind(t,t)][x˙tvt]12τiτfdtvt}.(9)

Letting τi (stationary-state scenario), the memory kernel becomes

Γ(,τf]ind(t,t)=12τa2λ[eλ|tt|κκ+eλ(2τftt)].(10)

For “infinitely long” stationary-state trajectories, for which also τf, this expression further reduces to

Γ(,)ind(t,t)=12τa2λeλ|tt|.(11)

The latter special case has been derived independently in Ref. 24 via Fourier transformation, see eq. 25 in Ref. 24, in order to analyze “entropy production” based on path-probability ratios. Similar Fourier-transform techniques for Langevin systems have been used in Ref. 48 for deriving a fluctuation relation at large times, with findings for the non-local “inverse temperature” as integration kernel in the “entropy production” corresponding to those in Ref. 24, and to our Eq. 11.

4. Path Weight for Arbitrary Initial Conditions

In this section, we generalize the path weight Eq. 9 to allow for arbitrary joint initial distributions pi(xi,ηi) of particle positions and active fluctuations. Keeping in mind that we can time-shift final results between trajectories running during a time interval (0,τ] and during arbitrary intervals (τi,τf] as in Section 3.2, we here consider without loss of generality trajectories with τi=0 and τf=τ. For notational simplicity we drop the subscripts (0,τ] or [0,τ] on P and Γ.

We start in Section 4.1 by first calculating Γ for a general Gaussian initial distribution of η0 independent of x0, which has variance σ2 and is centered at η^0,

pη^0,σ(η0)=12πσ2e(η0η^0)2/2σ2.(12)

Then, in Section 4.2, we show how this result can be used to cover any arbitrary initial distribution pi(η0|x0).

4.1 Gaussian Initial Distribution

With Eqs 4 and 5, and the initial distribution pi(ηi|xi)=pi(η0|x0)=pi(η0)=pη^0,σ(η0) from Eq. 12, the path weight we want to evaluate reads

Pη^0,σind[x¯|x0]12πσ2Dη¯×exp{0τdt[(x˙tvt2Daηt)24D+(τaη˙t+ηt)22+vt2](η0η^0)22σ2}.(13)

The superscript “ind” emphasizes again that we use statistically independent initial conditions for x0 and η0. After partial integration of the η˙t terms, similarly as in Ref. 18, we can express the path integral as

Pη^0,σind[x¯|x0]12πσ2exp{0τdt[(x˙tvt)24D+vt2]η^022σ2}×Dη¯exp{120τdt0τdtηtTVσ(t,t)ηt+0τdtηtT[2Da2D(x˙tvt)+δ(t)η^0σ2]},(14)

with the differential operator

Vσ(t,t):=δ(tt)[τa2t2+1+DaD+δ(t)(τa2tτa+1σ2)+δ(τt)(τa2t+τa)].(15)

Performing the Gaussian integral over η¯ in Eq. 14, we obtain

Pη^0,σind[x¯|x0](DetVσ)1/22πσ2exp{14D0τdt0τdt(x˙tvt)T×[δ(tt)DaDΓσ(t,t)](x˙tvt)+0τdt[2Da2D(x˙tvt)TΓσ(t,0)σ2η^0vt2]+[Γσ(0,0)σ21]η^022σ2},(16)

where Γσ(t,t) denotes the operator inverse of Vσ(t,t) in the sense that 0τdtVσ(t,t)Γσ(t,t)=δ(tt). It can be constructed similarly to the procedure in Ref. 18. In particular, we can also write Γσ(t,t)=G(t,t)+Hσ(t,t). Here G(t,t) is the Green’s function defined by [τa2t2+(1+Da/D)]G(t,t)=δ(tt) and G(0,t)=G(τ,t)=0. The second ingredient, H(t,t), is a solution of the associated homogeneous problem, [τa2t2+(1+Da/D)]H(t,t)=0, fixing the boundary terms as prescribed by Eq. 15. More details are given the Appendix. We find

Γσ(t,t)=(12τa2λ)[κ+(1σ2τaκ)κ(1σ2τaκ+)e2λτ]1[κ+(1σ2τaκ)eλ|tt|+κ(1σ2τaκ+)eλ(2τ|tt|)κ+(1σ2τaκ+)eλ(t+t)κ(1σ2τaκ)eλ(2τtt)].(17)

We note that Eq. 12 includes the steady-state distribution, pss(η0)=τa/πeτaη02, which arises for the active noise when evolving independently of the Brownian particle, as a special case for η^0=0 and σ2=1/(2τa). Accordingly, we recover Eqs 6 and 7 when plugging η^0=0 and σ2=1/(2τa) into Eqs 16 and 17, using 1κ±/2=κ/2.

4.2. Arbitrary Initial Distribution

To cover arbitrary initial distributions pi(ηi|xi)=pi(η0|x0) in η0, we introduce a δ-distribution of the form δ(η0η^0)=limσ0e(η0η^0)2/2σ2/2πσ2 and rewrite Eq. 4 (with τi=0, τf=τ) as

P[x¯|x0]=Dη¯P[x¯,η¯|x0,η0]pi(η0|x0)=Dη¯P[x¯,η¯|x0,η0]dη^0δ(η0η^0)pi(η^0|x0)=limσ0dη^0pi(η^0|x0)[12πσ2Dη¯P[x¯,η¯|x0,η0]e(η0η^0)2/2σ2].(18)

In view of Eq. 5 we see that the term in brackets is exactly Pη^0,σind[x¯|x0] as defined in Eq. 13. Since we can also write P[x¯|x0]=dη^0P[x¯,η^0|x0]=dη^0pi(η^0|x0)P[x¯|x0,η^0] we conclude that

P[x¯|x0,η^0]=limσ0Pη^0,σind[x¯|x0](19)

is the path weight conditioned on an initial position x0 and initial state of the active noise η^0 with arbitrary distributions.

With the explicit result Eq. 16 for Pη^0,σind[x¯|x0] we thus see that we have to calculate the σ0 limit of the expressions (σ2DetVσ)1/2, [Γσ(t,0)/σ21]/σ2, Γσ(0,0)/σ2 and Γσ(t,t). From Eq. 15 we observe that σ2Vσ has a constant term (independent of σ) and contributions quadratic in σ such that (σ2DetVσ)1/2 reduces to an (irrelevant) constant as σ0. Next, setting t=0 in Eq. 17 and using τaκ+τaκ=2τa2λ we get

Γσ(t,0)σ2=κ+eλtκeλ(2τt)κ+(1σ2τaκ)κ(1σ2τaκ+)e2λτσ0κ+eλtκeλ(2τt)κ+κe2λτ.(20)

If t=0, too, we obtain

Γσ(0,0)σ2=κ+κe2λτκ+(1σ2τaκ)κ(1σ2τaκ+)e2λτ,(21)

such that

[Γσ(0,0)σ21]1σ2=κ+κτa(1e2λτ)κ+(1σ2τaκ)κ(1σ2τaκ+)e2λτσ0κ+κτa(1e2λτ)κ+κe2λτ.(22)

Furthermore, we define

Γ(t,t):=limσ0Γσ(t,t)=(12τa2λ)×κ+eλ|tt|+κeλ(2τ|tt|)κ+eλ(t+t)κeλ(2τtt)κ+κe2λτ(23)

as the memory kernel for the path weight conditioned on an arbitrary initial configuration (x0,η^0) of particle positions and active fluctuations. Altogether, Eq. 19 for this path weight then becomes

P[x¯|x0,η^0]exp{14D0τdt0τdt(x˙tvt)T[δ(tt)DaDΓ(t,t)](x˙tvt)+0τdt[2Da2D(x˙tvt)Tκ+eλtκeλ(2τt)κ+κe2λτη^0vt2]Da2D[τa(1e2λτ)κ+κe2λτ]η^02}(24)

where we have used κ+κ=Da/D in the fourth line.

Finally, we can shift trajectories similarly as in Section 3.2 to obtain the path weight for arbitrary trajectories x¯={x(t)}t=τiτf conditioned on the joint initial state (xi,ηi) of position and active noise,

P(τi,τf][x¯|xi,ηi]exp{14Dτiτfdtτiτfdt[x˙tvt]T[δ(tt)DaDΓ[τi,τf](t,t)][x˙tvt]+τiτfdt[2Da2D[x˙tvt]Tκ+eλ(tτi)κeλ(2τftτi)κ+κe2λ(τfτi)ηivt2]Da2D[τa(1e2λ(τfτi))κ+κe2λ(τfτi)]ηi2}(25)

with

Γ[τi,τf](t,t):=(12τa2λ)×κ+eλ|tt|+κeλ(2(τfτi)|tt|)κ+eλ(t+t2τi)κeλ(2τftt)κ+κe2λ(τfτi).(26)

Given an initial distribution pi(ηi|xi) of the active fluctuations conditioned on the initial particle position, we can then compute the position-only path weight of an arbitrary trajectory by averaging over pi(ηi|xi),

P(τi,τf][x¯|xi]=dηiP(τi,τf][x¯|xi,ηi]pi(ηi|xi).(27)

Equations 2527 represent the first central result of the present contribution, a general expression for the path weight of active Ornstein-Uhlenbeck particles in position space only, for arbitrary trajectories with arbitrary initial and final times and arbitrary initial distributions. There is no approximation involved, so that our results are valid for any values of thermal and active noise parameters D and Da, τa.

We expect that the specific initial configuration becomes irrelevant for steady-state trajectories, i.e., in the limit τi. As τi, the second line vanishes in Eq. 25, because κ+eλ(tτi)κeλ(2τftτi)/κ+κe2λ(τfτi)0. The third line enters into the integral over the initial configuration ηi (see Eq. 27) and thus decouples from the trajectory x¯ resulting in an irrelevant prefactor. The only relevant contribution as τi is therefore the first line in Eq. 25 with the integral kernel Γ[τi,τf](t,t) reducing to

Γ(,τf](t,t)=12τa2λ[eλ|tt|κκ+eλ(2τftt)],(28)

the same expression as Γ(,τf]ind(t,t) from Eq. 10. This illustrates that the system loses its memory about the initial state as τi.

Another comparison to our previous results from Section 3.1 [18] is obtained by plugging the stationary state distribution pi(ηi|xi)=pss(ηi)=τa/πeτaηi2 into Eq. 27 and performing the Gaussian integral over ηi. In that case, we should get back the result Eq. 8, Eq. 9 for independent initial conditions. Indeed, including only the terms from Eq. 25 which involve ηi, we evaluate the Gaussian integral over ηi, yielding

dηiexpdηiexp{τiτfdt[2Da2D[x˙tvt]Tκ+eλ(tτi)κeλ(2τftτi)κ+κe2λ(τfτi)]ηiτa2[2+DaD(1e2λ(τfτi))κ+κe2λ(τfτi)]ηi2}=exp{14Dτiτfdtτiτfdt[x˙tvt]T[DaDΓ[τi,τf]ini(t,t)][x˙tvt]}(29)

with

Γ[τi,τf]ini(t,t)=(1τa)×κ+2eλ(t+t2τi)+κ2eλ(4τftt2τi)2κ+κe2λ(τfτi)[eλ(tt)+eλ(tt)](κ+κe2λ(τfτi))(κ+2κ2e2λ(τfτi)).(30)

A somewhat tedious but straightforward calculation then confirms Γ[τi,τf](t,t)+Γ[τi,τf]ini(t,t)=Γ[τi,τf]ind(t,t), as expected.

5. Irreversibility

In stochastic thermodynamics [26], irreversibility is quantified by comparing the probability P[x¯]=P[x¯|xi]pi(xi) of observing a specific trajectory x¯={x(t)}t=τiτf in a given experimental setup with the probability P˜[x˜¯] of observing the exact same trajectory traced out backwards when providing identical experimental conditions. In other words, P˜[x˜¯] is the probability of observing the “time-reversed” trajectory

x˜¯={x˜(t)}t=τiτf={x(τf+τit)}t=τiτf,(31)

with x˜(τi)=x(τf) and x˜(τf)=x(τi), under the time-reversed experimental protocol f˜(x,t):=f(x,τf+τit) (note that we disregard for convenience the possibility that parts of the forces could be odd under time reversal; it is straightforward to adapt the expressions below accordingly if necessary). For passive Brownian motion, it has been shown that the log-ratio of these path probabilities is related to the dissipation occurring along the trajectory x¯, quantified as the total change of entropy in the thermal bath and the system. This fundamental connection makes the “irreversibility measure”

ΔΣ[x¯]=kBlnP˜[x˜¯]P[x¯](32)

a central quantity of interest also for active particles. Indeed, its connection with dissipation and entropy is under lively debate [16, 1824].

We here provide a general expression for ΔΣ based on our result Eqs. 2527 for the path weight P[x¯]. Since the time-reversed trajectory x˜¯ is supposed to occur under identical conditions as the forward trajectory x¯, we can express its probability density via Eqs. 2527 as well, if we replace v(x,t) by the time-reversed protocol v˜(x,t)=f˜(x,t)/γ (see below Eq. 31). Using Eq. 31 we then rewrite the path weight for the reversed path in terms of the forward path (and the original protocol vt=v(x(t),t)). The resulting expression for P˜[x˜¯] is formally similar to Eq. 25, just with the sign inverted for all x˙(t) terms and all initial coordinates replaced by final ones. Plugging the path weights P[x¯] and P˜[x˜¯] into Eq. 32, and denoting the conditional average over the initial configuration ηi of the active fluctuations dηi()pi(ηi|xi) in Eq. 27 by ηi|xi and the corresponding one over final configurations dηf()pf(ηf|xf) by ηf|xf, we find

ΔΣ[x¯]=1Tτiτfdtτiτfdtx˙tTft[δ(tt)DaDΓ[τi,τf](t,t)]kBlnp(xf)p(xi)kBlnexp{τiτfdt[2Da2D[x.t+ vt]Tκ+eλ(t τi)κ eλ(2τft τi)κ+κ e2λ(τfτi)    ηf]}ηf|xfexp{+τiτfdt[2Da2D[x.t vt]Tκ+eλ(t τi)κeλ(2τft τi)κ+κe2λ(τfτi)   ηi]}ηi|xikBlnexp{Da2D[τa(1e2λ(τfτi))κ+κe2λ(τfτi)]ηf2}ηf|xfexp{Da2D[τa(1e2λ(τfτi))κ+κe2λ(τfτi)]ηi2}ηi|xi.(33)

This expression constitutes the second central result of this work. Given any spatial trajectory x¯={x(t)}t=τiτf, the measure ΔΣ[x¯] quantifies how irreversible this single trajectory is in the sense of the definition Eq. 32. A trajectory with ΔΣ[x¯]=0 is reversible, i.e., movement of the AOUP forward or backward along the trajectory occurs with equal probability, but the larger ΔΣ[x¯] the (exponentially) less likely it is to observe the backward movement.

Central properties of the active fluctuations which drive the particle motion are represented by the parameters Da (the strength of the active fluctuations) and τa (their correlation time, hidden in λ=1+Da/D/τa). Moreover, our general result Eq. 33 contains averages over the distributions of the active fluctuations ηi and ηf at the beginning of the particle trajectory and at the beginning of the reversed trajectory (see also Eq. 27). We therefore presuppose that we have some knowledge or control over these distributions when setting up the experiment, even though the (microscopic) degrees of freedom related to the active fluctuations typically are inaccessible, and so are specific realizations of η(t) or the specific values of ηi and ηf. For artificial active colloids [10], or in computer experiments, we may imagine, e.g., to let the particles orient randomly before “switching on” the activity, possibly with a specific strength (distribution).

In the spirit of quantifying irreversibility by asking how likely it is to observe a reversed trajectory compared to its forward twin when starting from identically prepared experimental setups (except for the initial particle position, which is xi for the forward path and xf for the backward path), we may take the distributions for ηi and ηf to be the same, or to be “mirror images” of each other under sign-inversion, depending on the physical situation modeled by the active fluctuations η(t) (see the discussion in Ref. 18)1. Moreover, we may imagine the experiment to be prepared in a way that the initial distributions of the active fluctuations for forward and backward motion are independent of particle positions (a notable exception arising, if the experiment starts from a joint steady state). For such independent initial conditions with identical (or “mirrored”) distributions, the third line in Eq. 33 vanishes. The second line, however, is still non-zero, and can be interpreted to quantify the contribution to irreversibility from the initial configuration of the active fluctuations.

The first line in Eq. 33 is independent of ηi and ηf, and thus measures the irreversibility associated with the time-evolution of the spatial particle position alone. It contains three terms (two in the double-integral and a boundary term), which all represent different contributions to irreversibility. The boundary term kBln[p(xf)/p(xi)] does not involve any parameters characterizing the thermal bath or the active fluctuations, and is usually interpreted as the change in system entropy of the AOUP between the beginning and end of the trajectory x¯ [18]. The integral involving δ(tt) is independent of the active parameters Da and τa, and is formally identical to the entropy produced in the thermal bath along the trajectory x¯ as known for passive Brownian motion [4]. However, in the present case of an AOUP it does not capture the full heat dissipation, because in addition to the force f(x(t),t) also active self-propulsion “forces” 2Daη(t) drive the particle and contribute to dissipation, i.e., even though the δ(tt)-integral can be interpreted as the “thermal contribution” to irreversibility due to the AOUP being in contact with a heat bath, it cannot be identified with the entropy produced in this thermal environment (see also the detailed discussion in Ref. 18). The second, proper double-integral encodes the (statistical) characteristics of the active fluctuations via its kernel Γ[τi,τf](t,t) and, furthermore, vanishes if active propulsion is “switched off”, i.e., for Da=0. Hence, it can be interpreted to measure the irreversibility “produced” by the active fluctuations along the trajectory x¯, and we will refer to it as the “active contribution” to irreversibility.

These two contributions to irreversibility from the particle trajectory x¯, the thermal one and the active one, are non-zero only if external forces f(x,t)=γv(x,t) are present in addition to the active self-propulsion (likewise for the integral in the second line, i.e., for the contribution associated with the initial preparation of the system), implying that the trajectories of “free active motion” appear reversible. For non-conservative forces, both contributions typically lead to a time-extensive increase of irreversibility with the trajectory length τ. For conservative forces f(x)=U(x) derived from a stationary confining potential U(x), the thermal contribution reduces to the boundary term [U(xi)U(xf)]/T and thus is non-extensive in τ. Due to the double-integral nature of the active part with the non-local kernel Γ[τi,τf](t,t) a similarly obvious argument does not apply. Indeed, the question whether or not, or in how far, the trajectory of an AOUP in a confining potential appears (ir)reversible is still not fully answered [16, 18, 49]. We can, however, draw some conclusions from considering the limiting cases of small and large correlations times τa0 and τa. In the first case, τa0, the active fluctuations become white and thus behave like a thermal bath, such that the AOUP can be imagined to be a passive Brownian particle in contact with a heat bath at effective temperature γ(D+Da)/kB, trapped in a confining potential. Accordingly, irreversibility production is non-extensive. In the second case, τa, the active fluctuations become constant and thus behave like a bias force which slightly tilts the confining potential. Again, the situation is similar to a trapped passive Brownian particle with non-extensive irreversibility production. We can therefore expect that the active contribution to irreversibility in a confining potential may become maximal at some intermediate value of τa.

Another important implication of the result Eq. 33 is that the rate at which irreversibility is produced in the stationary state (i.e., upon letting τi, cf., Section 3.2) is independent of the specific initial distribution pi(xi,ηi). Indeed, the terms in the second and third lines of Eq. 33 vanish as τi, and the memory kernel Γ[τi,τf](t,t) in the first line assumes the form Eq. 28, independent of the initial distribution (see the discussion in Section 4.2). Hence, if we are only interested in long-time properties, the initial configuration in particular of the self-propulsion drive is irrelevant. While we might have intuitively expected that the long-time irreversibility production rate is independent of the details of the initial setup, it is not completely obvious in the presence of memory effects [50, 51]. The fact that we can verify it for AOUPs is reassuring though, in so far as control over the initial state is limited in typical active particle systems (as already mentioned above). For infinitely long trajectories τf in the stationary state τi, the expression for ΔΣ[x¯] reduces further to

ΔΣ(,)[x¯]=1Tτiτfdtτiτfdtx˙tTft[δ(tt)DaDΓ(,)(t,t)](34)

with Γ(,)(t,t) from Eq. 11 (see the discussion around Eq. 28), in agreement with the findings in Refs. 24 and 48.

We conclude this discussion with a remark concerning the relation between the expression for ΔΣ[x¯] found earlier in Ref. 18 and the result Eq. 33 derived here. In Ref. 18, we have calculated ΔΣ[x¯] under the assumption that the active fluctuations are in their stationary state, independent of initial particle positions, at the beginning of the forward path and at the beginning of the backward path, i.e., pi(ηi|xi)=pss(ηi)=τa/πeτaηi2 and pf(ηf|xf)=pss(ηf)=τa/πeτaηf2. The resulting expression for Eq. 33 looks formally identical to the first line in Eq. 33, but with Γ[τi,τf](t,t) substituted by Γ[τi,τf]ind(t,t) from Eq. 8 (compare with eqs. 42b and 40a in Ref. 18). As we can see from the calculation at the end of Section 4.2 above, the “amount of irreversibility” stemming from the particular stationary-state initial configuration of the active fluctuations has been absorbed into Γ[τi,τf]ind(t,t), and is therefore not explicitly visible in Ref. 18 as an additional term analogous to the second line in Eq. 33.

6. Conclusion

What can we learn about the non-equilibrium nature of an active system by observing particle trajectories, i.e., the evolution of particle positions over time? Within the framework of a minimal model for particulate active matter on the micro- and nanoscale, the active Ornstein-Uhlenbeck particle [15, 2531] (see Eqs. 1 and 2), we here contribute an essential step toward exploring this question by deriving an exact analytical expression for the path weight (Eqs. 2527), which is valid for any values of the model parameters, any external driving forces, arbitrary initial particle positions and configurations of the active fluctuations, and arbitrary trajectory durations. We use this general expression to calculate the log-ratio of path weights for forward vs. backward trajectories (see Eq. 33). In analogy to the stochastic thermodynamics of passive Brownian particles [26], such an irreversibility measure may provide an approach toward a thermodynamic description of active matter [16, 1824].

In future works we may build on these results to further explore the non-equilibrium properties of AOUPs. A highly interesting problem is a possible thermodynamic interpretation of the path probability ratio ΔΣ[x¯] [18], e.g., via exploring its connection to active pressure [14, 52], to the different phases observed in active matter [53], or to the arrow of time [54, 55] in these systems. Such a thermodynamic interpretation, in particular concerning the role of dissipation, may finally allow to quantify efficiency fluctuations in stochastic heat engines operating between active baths [56], in analogy to passive stochastic heat engines [5759]. Other important questions which can be approached directly by using our general result for the path weight P[x¯] include the analysis of the response behavior under external perturbations [43] or of violations of the fluctuation-response relation [60, 61] due to the inherent non-equilibrium character of active matter, and their potential for probing properties of the active fluctuations [61]. Finally, it would be interesting to explore if our analytical methods used here to integrate out the simple Ornstein-Uhlenbeck fluctuations Eq. 2 can be extended to treat more general active fluctuations, like the ones considered in Ref. 62.

Data Availability Statement

The original contributions presented in the study are included in the article/Supplementary Material, further inquiries can be directed to the corresponding authors.

Author Contributions

All authors contributed equally to this work.

Funding

This research has been funded by the Swedish Research Council (Vetenskapsrådet) under the Grants No. 2016-05412 (RE) and by the Deutsche Forschungsgemeinschaft (DFG) within the Research Unit FOR 2692 under Grant No. 397303734 (LD).

Conflict of Interest

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Acknowledgments

We thank Stefano Bo for stimulating discussion. LD gratefully acknowledges support by the Nordita visiting PhD program.

Appendix

Evaluation of Γσ(t,t)

We here outline the calculation of Γσ(t,t) as the inverse of the differential operator

Vσ(t,t):=δ(tt)[τa2t2+1+DaD+δ(t)(τa2tτa+1σ2)+δ(τt)(τa2t+τa)],(35)

i.e., Γσ(t,t) is a solution of the equation 0τdtVσ(t,t)Γσ(t,t′′)=δ(tt′′). In fact, the operator Vσ(t,t) is “diagonal” in the time arguments such that Γσ(t,t) solves the differential equation

[τa2t2+1+DaD+δ(t)(τa2tτa+1σ2)+δ(τt)(τa2t+τa)]Γσ(t,t)=δ(tt).(36)

Note that t is essentially a fixed parameter here, just like D, Da, τa and σ. To find the solution, we follow the procedure from Ref. 18, i.e., we compose Γσ(t,t) from two parts as Γσ(t,t)=G(t,t)+Hσ(t,t). First, we construct the function G(t,t) as the Green’s function solving the equation [τa2t2+(1+Da/D)]G(t,t)=δ(tt) with homogeneous boundary conditions G(0,t)=G(τ,t)=0. Second, we determine Hσ(t,t) as a solution of the homogeneous problem [τa2t2+(1+Da/D)]Hσ(t,t)=0 such that the boundary terms are fixed as prescribed by Eq. 36.

We can construct both parts, G(t,t) and Hσ(t,t), from the general solution

Γ(t)=a+eλt+aeλt,λ=1τa1+DaD,a±=const(37)

of the homogeneous ordinary differential equation

[τa2t2+1+DaD]Γ(t)=0(38)

associated with Eq. 36. The Green’s function G(t,t) is exactly the same as in Ref. 18. Accordingly, its construction is completely analogous to the procedure outlined in Appendix B of Ref. 18, and we only recall the result here,

G(t,t)=12τa2λeλ(τ|tt|)eλ(τtt)+eλ(τ|tt|)eλ(τtt)eλτeλτ.(39)

The difference between the present calculation and the one in Ref. 18 is the boundary term at t=0, which contains a contribution from the arbitrary (Gaussian) initial distribution of the active fluctuations. To take both boundary terms in Eq. 36 into account, we make an ansatz of the form Eq. 37 for the function Hσ(t,t), i.e., Hσ(t,t)=a+eλt+aeλt. The coefficients a± are fixed by ensuring that the full solution Γσ(t,t)=G(t,t)+Hσ(t,t) fulfills Eq. 36. Plugging G(t,t)+Hσ(t,t) into Eq. 36, and using [τa2t2+(1+Da/D)]G(t,t)=δ(tt) and [τa2t2+1+DaD]Hσ(t,t)=0, we are left with

δ(t)[τa2tG(t,t)|t=0+a+(1/σ2τaκ+)+a(1/σ2τaκ)]+δ(τt)[τa2tG(t,t)|t=τ+a+τaκ+eλτ+aτaκeλτ]=0,(40)

where κ±=1±λτa=1±1+Da/D. Requiring that the terms in the two square brackets each vanish, we can solve for the coefficients a±, yielding

a+=(1τa)(1σ2τaκ)[eλ(2τt)eλ(2τ+t)]κ[eλ(2τ+t)eλ(4τt)]κ+(1σ2τaκ)κ(1σ2τaκ+)e2λτ,(41)
a=(1τa)(1σ2τaκ+)[eλ(2τt)eλ(2τ+t)]+κ+[eλteλ(2τt)]κ+(1σ2τaκ)κ(1σ2τaκ+)e2λτ.(42)

Substituting these coefficients into the above ansatz for Hσ(t,t) [see below Eq. 39] and combining it with G(t,t) from Eq. 39 according to Γσ(t,t)=G(t,t)+Hσ(t,t), we obtain the result stated in Eq. 17 of the main text.

Evaluation of Γσ(t,t)

We here outline the calculation of Γσ(t,t) as the inverse of the differential operator

Vσ(t,t):=δ(tt)[τa2t2+1+DaD+δ(t)(τa2tτa+1σ2)+δ(τt)(τa2t+τa)],(35)

i.e., Γσ(t,t) is a solution of the equation 0τdtVσ(t,t)Γσ(t,t′′)=δ(tt′′). In fact, the operator Vσ(t,t) is “diagonal” in the time arguments such that Γσ(t,t) solves the differential equation

[τa2t2+1+DaD+δ(t)(τa2tτa+1σ2)+δ(τt)(τa2t+τa)]Γσ(t,t)=δ(tt).(36)

Note that t is essentially a fixed parameter here, just like D, Da, τa and σ. To find the solution, we follow the procedure from Ref. 18, i.e., we compose Γσ(t,t) from two parts as Γσ(t,t)=G(t,t)+Hσ(t,t). First, we construct the function G(t,t) as the Green’s function solving the equation [τa2t2+(1+Da/D)]G(t,t)=δ(tt) with homogeneous boundary conditions G(0,t)=G(τ,t)=0. Second, we determine Hσ(t,t) as a solution of the homogeneous problem [τa2t2+(1+Da/D)]Hσ(t,t)=0 such that the boundary terms are fixed as prescribed by Eq. 36.

We can construct both parts, G(t,t) and Hσ(t,t), from the general solution

Γ(t)=a+eλt+aeλt,λ=1τa1+DaD,a±=const(37)

of the homogeneous ordinary differential equation

[τa2t2+1+DaD]Γ(t)=0(38)

associated with Eq. 36. The Green’s function G(t,t) is exactly the same as in Ref. 18. Accordingly, its construction is completely analogous to the procedure outlined in Appendix B of Ref. 18, and we only recall the result here,

G(t,t)=12τa2λeλ(τ|tt|)eλ(τtt)+eλ(τ|tt|)eλ(τtt)eλτeλτ.(39)

The difference between the present calculation and the one in Ref. 18 is the boundary term at t=0, which contains a contribution from the arbitrary (Gaussian) initial distribution of the active fluctuations. To take both boundary terms in Eq. 36 into account, we make an ansatz of the form Eq. 37 for the function Hσ(t,t), i.e., Hσ(t,t)=a+eλt+aeλt. The coefficients a± are fixed by ensuring that the full solution Γσ(t,t)=G(t,t)+Hσ(t,t) fulfills Eq. 36. Plugging G(t,t)+Hσ(t,t) into Eq. 36, and using [τa2t2+(1+Da/D)]G(t,t)=δ(tt) and [τa2t2+1+DaD]Hσ(t,t)=0, we are left with

δ(t)[τa2tG(t,t)|t=0+a+(1/σ2τaκ+)+a(1/σ2τaκ)]+δ(τt)[τa2tG(t,t)|t=τ+a+τaκ+eλτ+aτaκeλτ]=0,(40)

where κ±=1±λτa=1±1+Da/D. Requiring that the terms in the two square brackets each vanish, we can solve for the coefficients a±, yielding

a+=(1τa)(1σ2τaκ)[eλ(2τt)eλ(2τ+t)]κ[eλ(2τ+t)eλ(4τt)]κ+(1σ2τaκ)κ(1σ2τaκ+)e2λτ,(41)
a=(1τa)(1σ2τaκ+)[eλ(2τt)eλ(2τ+t)]+κ+[eλteλ(2τt)]κ+(1σ2τaκ)κ(1σ2τaκ+)e2λτ.(42)

Substituting these coefficients into the above ansatz for Hσ(t,t) [see below Eq. 39] and combining it with G(t,t) from Eq. 39 according to Γσ(t,t)=G(t,t)+Hσ(t,t), we obtain the result stated in Eq. 17 of the main text.

Footnotes

1In fact, to us these seem to be the only proper choices, if we want ΔΣ to quantify irreversibility. For arbitrary, unrelated distributions of ηi and ηf, we would compare forward and backward paths generated under different experimental conditions.

References

1. Callen HB. Thermodynamics & an introduction to thermostatistics. Hoboken, NJ: John Wiley & Sons (2006). 512 p.

Google Scholar

2. Seifert U. Stochastic thermodynamics: principles and perspectives. Eur Phys J B. (2008). 64:423–31. doi:10.1140/epjb/e2008-00001-9

CrossRef Full Text | Google Scholar

3. Jarzynski C. Equalities and inequalities: irreversibility and the second law of thermodynamics at the nanoscale. Annu Rev Condens Matter Phys. (2011). 2:329–51. doi:10.1146/annurev-conmatphys-062910-140506

CrossRef Full Text | Google Scholar

4. Seifert U. Stochastic thermodynamics, fluctuation theorems and molecular machines. Rep Prog Phys. (2012). 75:126001. doi:10.1088/0034-4885/75/12/126001

CrossRef Full Text | Google Scholar

5. Van den Broeck C, Esposito M. Ensemble and trajectory thermodynamics: a brief introduction. Phys Stat Mech Appl. (2015). 418:6–16. doi:10.1016/j.physa.2014.04.035

CrossRef Full Text | Google Scholar

6. Seifert U. Stochastic thermodynamics: from principles to the cost of precision. Phys Stat Mech Appl. (2018). 504:176–91. doi:10.1016/j.physa.2017.10.024

CrossRef Full Text | Google Scholar

7. Romanczuk P, Bär M, Ebeling W, Lindner B, Schimansky-Geier L. Active brownian particles. Eur Phys J Spec Top. (2012). 202:1–162. doi:10.1140/epjst/e2012-01529-y

CrossRef Full Text | Google Scholar

8. Cates ME. Diffusive transport without detailed balance in motile bacteria: does microbiology need statistical physics? Rep Prog Phys. (2012). 75:042601. doi:10.1088/0034-4885/75/4/042601

CrossRef Full Text | Google Scholar

9. Elgeti J, Winkler RG, Gompper G. Physics of microswimmers-single particle motion and collective behavior: a review. Rep Prog Phys. (2015). 78:056601. doi:10.1088/0034-4885/78/5/056601

CrossRef Full Text | Google Scholar

10. Bechinger C, Di Leonardo R, Löwen H, Reichhardt C, Volpe G, Volpe G. Active particles in complex and crowded environments. Rev Mod Phys. (2016). 88:045006. doi:10.1103/RevModPhys.88.045006

CrossRef Full Text | Google Scholar

11. Patteson AE, Gopinath A, Arratia PE. Active colloids in complex fluids. Curr Opin Colloid Interface Sci. (2016). 21:86–96. doi:10.1016/j.cocis.2016.01.001

CrossRef Full Text | Google Scholar

12. Tailleur J, Cates ME. Statistical mechanics of interacting run-and-tumble bacteria. Phys Rev Lett. (2008). 100:218103. doi:10.1103/PhysRevLett.100.218103

CrossRef Full Text | Google Scholar

13. Speck T, Bialké J, Menzel AM, Löwen H. Effective cahn-hilliard equation for the phase separation of active brownian particles. Phys Rev Lett. (2014). 112:218304. doi:10.1103/physrevlett.112.218304

CrossRef Full Text | Google Scholar

14. Takatori SC, Yan W, Brady JF. Swim pressure: stress generation in active matter. Phys Rev Lett. (2014). 113:028103. doi:10.1103/physrevlett.113.028103

CrossRef Full Text | Google Scholar

15. Farage TFF, Krinninger P, Brader JM. Effective interactions in active brownian suspensions. Phys Rev E. (2015). 91:042310. doi:10.1103/PhysRevE.91.042310

CrossRef Full Text | Google Scholar

16. Fodor É, Nardini C, Cates ME, Tailleur J, Visco P, van Wijland F. How far from equilibrium is active matter? Phys Rev Lett. (2016). 117:038103. doi:10.1103/PhysRevLett.117.038103

CrossRef Full Text | Google Scholar

17. Nardini C, Fodor É, Tjhung E, van Wijland F, Tailleur J, Cates ME. Entropy production in field theories without time-reversal symmetry: quantifying the non-equilibrium character of active matter. Phys Rev X. (2017). 7:021007. doi:10.1103/physrevx.7.021007

CrossRef Full Text | Google Scholar

18. Dabelow L, Bo S, Eichhorn R. Irreversibility in active matter systems: fluctuation theorem and mutual information. Phys Rev X. (2019). 9:021009. doi:10.1103/physrevx.9.021009

CrossRef Full Text | Google Scholar

19. Marconi UMB, Puglisi A, Maggi C. Heat, temperature and clausius inequality in a model for active brownian particles. Sci Rep. (2017). 7:46496. doi:10.1038/srep46496

CrossRef Full Text | Google Scholar

20. Mandal D, Klymko K, DeWeese MR. Entropy production and fluctuation theorems for active matter. Phys Rev Lett. (2017). 119:258001. doi:10.1103/physrevlett.119.258001

CrossRef Full Text | Google Scholar

21. Puglisi A, Marconi UMB. Clausius relation for active particles: what can we learn from fluctuations. Entropy. (2017). 19:356. doi:10.3390/e19070356

CrossRef Full Text | Google Scholar

22. Caprini L, Marconi UMB, Puglisi A, Vulpiani A. Comment on “entropy production and fluctuation theorems for active matter”. Phys Rev Lett. (2018). 121:139801. doi:10.1103/physrevlett.121.139801

CrossRef Full Text | Google Scholar

23. Mandal D, Klymko K, DeWeese MR. A reply to the comment by Mandal, Klymko, and DeWeese. Phys Rev Lett. (2018). 121:139802. doi:10.1103/physrevlett.121.139802

CrossRef Full Text | Google Scholar

24. Caprini L, Marconi UMB, Puglisi A, Vulpiani A. The entropy production of Ornstein-Uhlenbeck active particles: a path integral method for correlations. J Stat Mech. (2019). 2019:053203. doi:10.1088/1742-5468/ab14dd

CrossRef Full Text | Google Scholar

25. Fily Y, Marchetti MC. Athermal phase separation of self-propelled particles with no alignment. Phys Rev Lett. (2012). 108:235702. doi:10.1103/PhysRevLett.108.235702

CrossRef Full Text | Google Scholar

26. Maggi C, Paoluzzi M, Pellicciotta N, Lepore A, Angelani L, Di Leonardo R. Generalized energy equipartition in harmonic oscillators driven by active baths. Phys Rev Lett. (2014). 113:238303. doi:10.1103/PhysRevLett.113.238303

CrossRef Full Text | Google Scholar

27. Argun A, Moradi A-R, Pinçe E, Bagci GB, Imparato A, Volpe G. Non-Boltzmann stationary distributions and nonequilibrium relations in active baths. Phys Rev E. (2016). 94:062150. doi:10.1103/PhysRevE.94.062150

CrossRef Full Text | Google Scholar

28. Maggi C, Paoluzzi M, Angelani L, Di Leonardo R. Memory-less response and violation of the fluctuation-dissipation theorem in colloids suspended in an active bath. Sci Rep. (2017). 7:17588. doi:10.1038/s41598-017-17900-2

CrossRef Full Text | Google Scholar

29. Chaki S, Chakrabarti R. Entropy production and work fluctuation relations for a single particle in active bath. Phys Stat Mech Appl. (2018). 511:302–15. doi:10.1016/j.physa.2018.07.055

CrossRef Full Text | Google Scholar

30. Marconi UMB, Maggi C. Towards a statistical mechanical theory of active fluids. Soft Matter. (2015). 11:8768–81. doi:10.1039/C5SM01718A

CrossRef Full Text | Google Scholar

31. Shankar S, Marchetti MC. Hidden entropy production and work fluctuations in an ideal active gas. Phys Rev E. (2018). 98:020604. doi:10.1103/physreve.98.020604

CrossRef Full Text | Google Scholar

32. Koumakis N, Maggi C, Di Leonardo R. Directed transport of active particles over asymmetric energy barriers. Soft Matter. (2014). 10:5695–701. doi:10.1039/c4sm00665h

CrossRef Full Text | Google Scholar

33. Szamel G. Self-propelled particle in an external potential: existence of an effective temperature. Phys Rev E. (2014). 90:012111. doi:10.1103/physreve.90.012111

CrossRef Full Text | Google Scholar

34. Szamel G, Flenner E, Berthier L. Glassy dynamics of athermal self-propelled particles: computer simulations and a nonequilibrium microscopic theory. Phys Rev E. (2015). 91:062304. doi:10.1103/physreve.91.062304

CrossRef Full Text | Google Scholar

35. Maggi C, Marconi UMB, Gnan N, Di Leonardo R. Multidimensional stationary probability distribution for interacting active particles. Sci Rep. (2015). 5:10742. doi:10.1038/srep10742

CrossRef Full Text | Google Scholar

36. Flenner E, Szamel G, Berthier L. The nonequilibrium glassy dynamics of self-propelled particles. Soft Matter. (2016). 12:7136–49. doi:10.1039/C6SM01322H

CrossRef Full Text | Google Scholar

37. Paoluzzi M, Maggi C, Marconi UMB, Gnan N. Critical phenomena in active matter. Phys Rev E. (2016). 94:052602. doi:10.1103/physreve.94.052602

CrossRef Full Text | Google Scholar

38. Marconi UMB, Gnan N, Paoluzzi M, Maggi C, Di Leonardo R. Velocity distribution in active particles systems. Sci Rep. (2016). 6:23297. doi:10.1038/srep26215

CrossRef Full Text | Google Scholar

39. Szamel G. Evaluating linear response in active systems with no perturbing field. Europhys Lett. (2017). 117:50010. doi:10.1209/0295-5075/117/50010

CrossRef Full Text | Google Scholar

40. Sandford C, Grosberg AY, Joanny JF. Pressure and flow of exponentially self-correlated active particles. Phys Rev E. (2017). 96:052605. doi:10.1103/physreve.96.052605

CrossRef Full Text | Google Scholar

41. Caprini L, Marconi UMB, Vulpiani A. Linear response and correlation of a self-propelled particle in the presence of external fields. J Stat Mech. (2018). 2018:033203. doi:10.1088/1742-5468/aaa78c

CrossRef Full Text | Google Scholar

42. Fodor É, Cristina Marchetti M. The statistical physics of active matter: from self-catalytic colloids to living cells. Phys Stat Mech Appl. (2018). 504:106–20. doi:10.1016/j.physa.2017.12.137

CrossRef Full Text | Google Scholar

43. Dal Cengio S, Levis D, Pagonabarraga I. Linear response theory and green-kubo relations for active matter. Phys Rev Lett. (2019). 123:238003. doi:10.1103/physrevlett.123.238003

CrossRef Full Text | Google Scholar

44. Bonilla L. Active ornstein-uhlenbeck particles. Phys Rev. (2019). 100:022601. doi:10.1103/physreve.100.022601

CrossRef Full Text | Google Scholar

45. Onsager L, Machlup S. Fluctuations and irreversible processes. Phys Rev. (1953). 91:1505–12. doi:10.1103/PhysRev.91.1505

CrossRef Full Text | Google Scholar

46. Machlup S, Onsager L. Fluctuations and irreversible process. II. Systems with kinetic energy. Phys Rev. (1953). 91:1512–5. doi:10.1103/PhysRev.91.1512

CrossRef Full Text | Google Scholar

47. Cugliandolo LF, Lecomte V. Rules of calculus in the path integral representation of white noise Langevin equations: the Onsager-Machlup approach. J Phys Math Theor. (2017). 50:345001. doi:10.1088/1751-8121/aa7dd6

CrossRef Full Text | Google Scholar

48. Zamponi F, Bonetto F, Cugliandolo LF, Kurchan J. A fluctuation theorem for non-equilibrium relaxational systems driven by external forces. J. Stat. Mech. (2005). 2005:P09013. doi:10.1088/1742-5468/2005/09/p09013

CrossRef Full Text | Google Scholar

49. Dabelow L, Bo S, Eichhorn R. How irreversible are steady-state trajectories of a trapped active particle? (unpublished).

Google Scholar

50. Harris RJ, Touchette H. Current fluctuations in stochastic systems with long-range memory. J Phys Math Theor. (2009). 42:342001. doi:10.1088/1751-8113/42/34/342001

CrossRef Full Text | Google Scholar

51 . Puglisi A, Villamaina D. Irreversible effects of memory. Europhys Lett. (2009). 88:30004. doi:10.1209/0295-5075/88/30004

CrossRef Full Text | Google Scholar

52. Solon AP, Stenhammar J, Wittkowski R, Kardar M, Kafri Y, Cates ME, et al. Pressure and phase equilibria in interacting active brownian spheres. Phys Rev Lett. (2015). 114:198301. doi:10.1103/physrevlett.114.198301

CrossRef Full Text | Google Scholar

53. Cates ME, Tailleur J. Motility-induced phase separation. Annu Rev Condens Matter Phys. (2015). 6:219–44. doi:10.1146/annurev-conmatphys-031214-014710

CrossRef Full Text | Google Scholar

54. Roldán É, Neri I, Dörpinghaus M, Meyr H, Jülicher F. Decision making in the arrow of time. Phys Rev Lett. (2015). 115:250602. doi:10.1103/physrevlett.115.250602

CrossRef Full Text | Google Scholar

55. Roldán É, Barral J, Martin P, Parrondo JM, Jülicher F. Arrow of time in active fluctuations. Preprint repository name [Preprint] (2018). Available from: https://arxiv.org/abs/1803.04743.

Google Scholar

56. Krishnamurthy S, Ghosh S, Chatterji D, Ganapathy R, Sood AK. A micrometre-sized heat engine operating between bacterial reservoirs. Nat Phys. (2016). 12:1134–8. doi:10.1038/nphys3870

CrossRef Full Text | Google Scholar

57. Verley G, Esposito M, Willaert T, Van den Broeck C. The unlikely Carnot efficiency. Nat Commun. (2014). 5:4721. doi:10.1038/ncomms5721

CrossRef Full Text | Google Scholar

58. Verley G, Willaert T, Van den Broeck C, Esposito M. Universal theory of efficiency fluctuations. Phys Rev E. (2014). 90:052145. doi:10.1103/physreve.90.052145

CrossRef Full Text | Google Scholar

59. Manikandan SK, Dabelow L, Eichhorn R, Krishnamurthy S. Efficiency fluctuations in microscopic machines. Phys Rev Lett. (2019). 122:140601. doi:10.1103/physrevlett.122.140601

CrossRef Full Text | Google Scholar

60. Harada T, Si S. Equality connecting energy dissipation with a violation of the fluctuation-response relation. Phys Rev Lett. (2005). 95:130602. doi:10.1103/physrevlett.95.130602

CrossRef Full Text | Google Scholar

61. Gnesotto FS, Mura F, Gladrow J, Broedersz CP. Broken detailed balance and non-equilibrium dynamics in living systems: a review. Rep Prog Phys. (2018). 81:066601. doi:10.1088/1361-6633/aab3ed

CrossRef Full Text | Google Scholar

62. Sevilla FJ, Rodríguez RF, Gomez-Solano JR. Generalized Ornstein-Uhlenbeck model for active motion. Phys Rev E. (2019). 100:032123. doi:10.1103/physreve.100.032123

CrossRef Full Text | Google Scholar

Keywords: active matter, stochastic thermodynamics, non-equilibrium, active Brownian motion, active Ornstein-Uhlenbeck particle, irreversibility, path integrals

Citation: Dabelow L and Eichhorn R (2021) Irreversibility in Active Matter: General Framework for Active Ornstein-Uhlenbeck Particles. Front. Phys. 8:582992. doi: 10.3389/fphy.2020.582992

Received: 13 July 2020; Accepted: 07 October 2020;
Published: 11 January 2021.

Edited by:

Ayan Banerjee, Indian Institute of Science Education and Research Kolkata, India

Reviewed by:

Francisco J. Sevilla, Universidad Nacional Autónoma de México, Mexico
Ignazio Licata, Institute for Scientific Methodology (ISEM), Italy

Copyright © 2021 Eichhorn and Dabelow. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

*Correspondence: Ralf Eichhorn, eichhorn@nordita.org