Stochastic Thermodynamics

Σχετικά έγγραφα
Stochastic thermodynamics

Jesse Maassen and Mark Lundstrom Purdue University November 25, 2013

Connec&ng thermodynamics to sta&s&cal mechanics for strong system-environment coupling

Chapter 6: Systems of Linear Differential. be continuous functions on the interval

Chapter 6: Systems of Linear Differential. be continuous functions on the interval

the total number of electrons passing through the lamp.

Phys460.nb Solution for the t-dependent Schrodinger s equation How did we find the solution? (not required)

P AND P. P : actual probability. P : risk neutral probability. Realtionship: mutual absolute continuity P P. For example:

2 Composition. Invertible Mappings

A Bonus-Malus System as a Markov Set-Chain. Małgorzata Niemiec Warsaw School of Economics Institute of Econometrics

[1] P Q. Fig. 3.1

DETERMINATION OF DYNAMIC CHARACTERISTICS OF A 2DOF SYSTEM. by Zoran VARGA, Ms.C.E.

Lecture 2: Dirac notation and a review of linear algebra Read Sakurai chapter 1, Baym chatper 3

General 2 2 PT -Symmetric Matrices and Jordan Blocks 1

DERIVATION OF MILES EQUATION FOR AN APPLIED FORCE Revision C

Second Order RLC Filters

NonEquilibrium Thermodynamics of Flowing Systems: 2

Thermodynamics of the motility-induced phase separation

Bayesian statistics. DS GA 1002 Probability and Statistics for Data Science.

Fractional Colorings and Zykov Products of graphs

Probability and Random Processes (Part II)

Elements of Information Theory

Lifting Entry (continued)

HOMEWORK 4 = G. In order to plot the stress versus the stretch we define a normalized stretch:

The Spiral of Theodorus, Numerical Analysis, and Special Functions

Matrices and Determinants

DuPont Suva. DuPont. Thermodynamic Properties of. Refrigerant (R-410A) Technical Information. refrigerants T-410A ENG

Hadronic Tau Decays at BaBar

Απόκριση σε Μοναδιαία Ωστική Δύναμη (Unit Impulse) Απόκριση σε Δυνάμεις Αυθαίρετα Μεταβαλλόμενες με το Χρόνο. Απόστολος Σ.

DuPont Suva 95 Refrigerant

Solution Series 9. i=1 x i and i=1 x i.

Every set of first-order formulas is equivalent to an independent set

Coupling of a Jet-Slot Oscillator With the Flow-Supply Duct: Flow-Acoustic Interaction Modeling

Problem Set 3: Solutions

Biostatistics for Health Sciences Review Sheet

Optical Feedback Cooling in Optomechanical Systems

Study on Re-adhesion control by monitoring excessive angular momentum in electric railway traction

DuPont Suva 95 Refrigerant

Statistics 104: Quantitative Methods for Economics Formula and Theorem Review

Graded Refractive-Index

Space-Time Symmetries

Statistical Inference I Locally most powerful tests

Overview. Transition Semantics. Configurations and the transition relation. Executions and computation

Three coupled amplitudes for the πη, K K and πη channels without data

ST5224: Advanced Statistical Theory II

Laplace Expansion. Peter McCullagh. WHOA-PSI, St Louis August, Department of Statistics University of Chicago

Technical Information T-9100 SI. Suva. refrigerants. Thermodynamic Properties of. Suva Refrigerant [R-410A (50/50)]

Homework 8 Model Solution Section

Lecture 21: Scattering and FGR

1 String with massive end-points

Quantum Statistical Mechanics (equilibrium) solid state, magnetism black body radiation neutron stars molecules lasers, superuids, superconductors

SOLUTIONS TO MATH38181 EXTREME VALUES AND FINANCIAL RISK EXAM

Parametrized Surfaces

The kinetic and potential energies as T = 1 2. (m i η2 i k(η i+1 η i ) 2 ). (3) The Hooke s law F = Y ξ, (6) with a discrete analog

6.1. Dirac Equation. Hamiltonian. Dirac Eq.

Thi=Τ1. Thο=Τ2. Tci=Τ3. Tco=Τ4. Thm=Τ5. Tcm=Τ6

Second Order Partial Differential Equations

TMA4115 Matematikk 3

Homomorphism in Intuitionistic Fuzzy Automata

Dr. D. Dinev, Department of Structural Mechanics, UACEG

EE512: Error Control Coding

High Performance Voltage Controlled Amplifiers Typical and Guaranteed Specifications 50 Ω System

Markov chains model reduction

Non-Gaussianity from Lifshitz Scalar

STABILITY FOR RAYLEIGH-BENARD CONVECTIVE SOLUTIONS OF THE BOLTZMANN EQUATION

Tridiagonal matrices. Gérard MEURANT. October, 2008

65W PWM Output LED Driver. IDLV-65 series. File Name:IDLV-65-SPEC

J. of Math. (PRC) 6 n (nt ) + n V = 0, (1.1) n t + div. div(n T ) = n τ (T L(x) T ), (1.2) n)xx (nt ) x + nv x = J 0, (1.4) n. 6 n

Figure A.2: MPC and MPCP Age Profiles (estimating ρ, ρ = 2, φ = 0.03)..

Lecture 2. Soundness and completeness of propositional logic

ΗΜΥ 220: ΣΗΜΑΤΑ ΚΑΙ ΣΥΣΤΗΜΑΤΑ Ι Ακαδημαϊκό έτος Εαρινό Εξάμηνο Κατ οίκον εργασία αρ. 2

D Alembert s Solution to the Wave Equation

SMBJ SERIES. SMBG Plastic-Encapsulate Diodes. Transient Voltage Suppressor Diodes. Peak pulse current I PPM A with a 10/1000us waveform See Next Table

Chapter 2. Stress, Principal Stresses, Strain Energy

Abstract Storage Devices

Teor imov r. ta matem. statist. Vip. 94, 2016, stor

Aluminum Electrolytic Capacitors

상대론적고에너지중이온충돌에서 제트입자와관련된제동복사 박가영 인하대학교 윤진희교수님, 권민정교수님

Free Energy and Phase equilibria

Lifting Entry 2. Basic planar dynamics of motion, again Yet another equilibrium glide Hypersonic phugoid motion MARYLAND U N I V E R S I T Y O F

ΠΑΝΕΠΙΣΤΗΜΙΟ ΠΕΙΡΑΙΑ ΤΜΗΜΑ ΝΑΥΤΙΛΙΑΚΩΝ ΣΠΟΥΔΩΝ ΠΡΟΓΡΑΜΜΑ ΜΕΤΑΠΤΥΧΙΑΚΩΝ ΣΠΟΥΔΩΝ ΣΤΗΝ ΝΑΥΤΙΛΙΑ

Data sheet Thick Film Chip Resistor 5% - RS Series 0201/0402/0603/0805/1206

A Determination Method of Diffusion-Parameter Values in the Ion-Exchange Optical Waveguides in Soda-Lime glass Made by Diluted AgNO 3 with NaNO 3

Thermoelectrics: A theoretical approach to the search for better materials

The Simply Typed Lambda Calculus

Rating to Unit ma ma mw W C C. Unit Forward voltage Zener voltage. Condition

Ordinal Arithmetic: Addition, Multiplication, Exponentiation and Limit

Metal Oxide Leaded Film Resistor

Physique des réacteurs à eau lourde ou légère en cycle thorium : étude par simulation des performances de conversion et de sûreté

Geodesic paths for quantum many-body systems

Nowhere-zero flows Let be a digraph, Abelian group. A Γ-circulation in is a mapping : such that, where, and : tail in X, head in

Mechanical Behaviour of Materials Chapter 5 Plasticity Theory

Systems with unlimited supply of work: MCQN with infinite virtual buffers A Push Pull multiclass system

Section 8.3 Trigonometric Equations

SCITECH Volume 13, Issue 2 RESEARCH ORGANISATION Published online: March 29, 2018

Mean bond enthalpy Standard enthalpy of formation Bond N H N N N N H O O O

Metal Oxide Leaded Film Resistor


Matrices and vectors. Matrix and vector. a 11 a 12 a 1n a 21 a 22 a 2n A = b 1 b 2. b m. R m n, b = = ( a ij. a m1 a m2 a mn. def

An Inventory of Continuous Distributions

Transcript:

Leiden, Hot nanostructures, October 213 Stochastic Thermodynamics Udo Seifert II. Institut für Theoretische Physik, Universität Stuttgart Review: U.S., Rep. Prog. Phys. 75 1261, 212. 1

Thermodynamics of macroscopic systems [19 th cent] λ 111111 111111 111111 111111 111111 W T λ t 111 111 111 111 111 First law energy balance: W = E +Q = E +T S M Second law: S tot S + S M > W > E T S F W diss W F > 2

Stochastic thermodynamics for small systems W λ T, p λ t driving: mechanical (electro)chemical (bio)chemical First law: how to define work, internal energy and exchanged heat? fluctuations imply distributions: p(w; λ(τ))... entropy: distribution as well? 3

Stochastic thermodynamics applies to such systems where non-equilibrium is caused by mechanical or chemical forces ambient solution provides a thermal bath of well-defined T and µ i fluctuations are relevant due to small numbers of involved molecules Main idea: Energy conservation (1 st law) and entropy production (2 nd law) along an individual stochastic trajectory Precursors: notion stoch th dyn by Nicolis, van den Broeck mid 8s ( ensemble level) early fluct diss response relations: Hänggi & Thomas early 8 s stochastic energetics (1 st law) by Sekimoto late 9s work theorem(s): Jarzynski, Crooks late 9s, Imparato & Peliti early s fluct theorem: Evans, Cohen, Galavotti, Kurchan, Lebowitz & Spohn 9s quantities like stochastic entropy by Crooks, Qian, Gaspard in early s... 4

Paradigm for mechanical driving: x x 4 x 1 x 6 x 3 x 5 x 2 λ(τ) V (x, λ) λ(τ) x f(λ) 1111 1 1 111111 111 1111111 1111 1 1 1 1 1111111111 11111111 V (x,λ) Langevin dyn s ẋ = µ[ V (x,λ)+f(λ)] }{{} F(x,λ) external protocol λ(τ) +ζ ζζ = 2µk B T δ(...) }{{} ( 1) First law [(Sekimoto, 1997)]: dw = du+dq applied work: dw = λ V(x,λ)dλ+f(λ)dx internal energy: du = dv dissipated heat: dq = dw du = F(x,λ)dx = Tds m 5

Experimental illustration: Colloidal particle in V(x, λ(τ)) [V. Blickle, T. Speck, L. Helden, U.S., C. Bechinger, PRL 96, 763, 26] work distribution non-gaussian distribution Langevin valid beyond lin response [T. Speck and U.S., PRE 7, 66112, 24] 6

Stochastic entropy [U.S., PRL 95, 462, 25] Fokker-Planck equation τ p(x,τ) = x j(x,τ) = x (µf(x,λ) D x )p(x,τ) [D = µk B T] Common non-eq ensemble entropy [k B 1] S(τ) dx p(x,τ)lnp(x,τ) Stochastic entropy for an individual trajectory x(τ) s(τ) lnp(x(τ),τ) with s(τ) = S(τ) s tot s m + s exp[ s tot ] = 1 s tot integral fluctuation theorem for total entropy production arbitrary initial state, driving, length of trajectory 7

Heat engines at maximal power Carnot (1824) Curzon-Ahlborn (1975) Q h T h Q h = α(t h T m h ) T m h T h W W T m c Qc η c 1 T c /T h but zero power Tc Qc = β(t m c Tc) efficiency at maximum power η ca 1 T c /T h Tc 8

Brownian heat engine at maximal power with universal bounds [T. Schmiedl and U.S., EPL 81, 23, (28)] V 1 V p a p b 1 V 4 T h T c 3 V 2 η.8.6.4.2 α = 1/2 α = α CA α = 1 p a p b 1 2 3 4 5 η C (T h T c )/T c η = η c 2 αη c with α 1 implies η c 2 η η c 2 η c Curzon-Ahlborn neither universal nor a bound cf. Esposito, Kawai, Lindenberg, vd Broeck, PRL 1,... 9

Experimental realization in the Stirling version H transition isochoric [V. Blickle and C. Bechinger, Nature phys. 8, 143, 212] 4 2 a W [k B T c ] -2-4 -6 W [k B T c ] 2 1 (1) (2) (1) (2) (1) (2) w n T =86 C h (3) (4) -8 (3) (4) (3) (4) (3) (4) -1 time[s] 3-1 5 1 15 2 w n [k B T c ] 1 b time[s] 1 c -.2..2 position [µm] p isothermal expansion macroscopic Stirling cycle -.2..2 position [µm].5-1 1 cycle number n -1 1 p(w n ).2 (3).. isochoric transition (2) (2) (4) (1) V (1) P [k B T c /s] -.5 -.1 -.6 1 2 3 4 5.2 a -.2 -.4 w [k B T c ] T =22 C c -.2..2 position [µm] isothermal compression -.2..2 position [µm] Efficiency. b -.2 1 2 3 4 5 [s] 1

NESSs: Examples and common characteristics f(λ) 1111 1 1 111111 111 1111111 1111 1 1 1 1 1111111111 11111111 V (x,λ) Time-independent driving beyond linear response regime Broken detailed-balance Persistent currents with permanent dissipation Stationary (non-boltzmann) distribution 11

Fluctuation theorem p( s tot )/p( s tot ) = exp( s tot ) long-time limit: Evans et al (1993), Gallavotti & Cohen (1995), Kurchan (1998), Lebowitz & Spohn (1999)... finite times: U.S., PRL 5 experimental data [Speck, Blickle, Bechinger, U.S., EPL 79 32 (27)] ~ a ) t = 2 s t = 2 s -5-1 -15-2 -25 f(λ) V (x,λ) 2 4 6 8 1 b ) -5-1 -15-2 t = 2 s -25 t = 2 s t = 2 1. 5 ẋ = µ[ x V(x)+f]+ζ, 2 4 6 8 1 to ta l e n tr o p y p r o d u c tio n s t o t [k B ] 12

F theorem and slow hidden degrees of freedom [J. Mehl, B. Lander, C. Bechinger, V. Blickle and U.S., PRL 18, 2261, 212] total entropy production in the NESS s tot t dτ[ x 1 ν 1 (x 1,x 2 )+ x 2 ν 2 (x 1,x 2 )] with ν 1 (x 1,x 2 ) x 1 x 1,x 2 x 1 x 2 (a) + - R f(λ) 1111 1 1 111111 111 1111111 1111 1 1 1 11 1 1 111111111 R f 1 V (x,λ) f(λ) 1111 1 1 111111 111 1111111 1111 1 1 1 11 1 1 111111111 V (x,λ) min.,15,165,18,195,21,225,24,255,27,285,3 max. +1 x ( R) obeys FT p( s tot )/p( s tot ) = exp s tot B x 1 (b) -1 +1 suppose x 2 is hidden: ν 1 (x 1 ) ν(x 1,x 2 )p(x 2 x 1 )dx 2 apparent entropy production s tot t dτx 1 ν 1 (x 1 ) obeys FT?? + - R potential (units of k B T) x 2 6 3-3 U 1 R f 2 U 2-6 -1 +1 x 1, x 2 ( R) (c) (d) -1 +1-1 -1 +1 x 1 ( R) x ( R) x ( R) 13

~ ~ Experimental data with and without coupling [rarely:] p(ds tot )(1-2 ) ~ 8 6 4 2-3 -2-1 1 2 3 ~ Ds tot ln[p(ds tot )/p(-ds tot )] 3 2 1-1 -2 (c) -3-3 -2-1 1 2 3 ~ Ds tot ~ ~ ln[p(ds tot )/p(-ds tot )] 3 2 1-1 -2-3 -3-2 -1 1 2 3 ~ Ds tot (a) slope α 1,1 1, slope a,9,8,7 (a),6 1 2 3 G,5 1, 1,5 2, t(s) 14

Fluctuation-dissipation (response) theorem (FDT) in equilibrium perturbation with a field f : E E fb observable A T δa(t 2) δf(t 1 ) = t 1 A(t 2 )B(t 1 ) any variable A formalizes Onsager s regression hypothesis in a NESS often (phenomenologically) modified by an effective temp: T eff δa(t 2 ) δf(t 1 ) = t 1 A(t 2 )B(t 1 ) Culgiandolo, Kurchan & Peliti, 97 Berthier and Barrat, 2... 15

Paradigm for an FDT in a NESS 1111 1 1 111111 111 1111111 1111 1 1 1 1 1111111111 [T. Speck and U.S., Europhys. Lett. 74, 391, 26] f(λ) 111111111 V (x,λ) FDT in eq: T δ ẋ(t 2) δf(t 1 ) f= = ẋ(t 2 )ẋ(t 1 ) eq extended FDT in non-eq: T δ ẋ(t 2) δf(t 1 ) f = ẋ(t 2 )ẋ(t 1 ) ness ẋ(t 2 )ν s (x(t 1 )) ness with ν s (x) ẋ x = j s /p s (x) additive modification (rather than multiplicative) 16

Integrated version: Generalized Einstein relation [Blickle, Speck, Lutz, U.S., Bechinger, PRL 98, 2161 (27)] f(λ) 1111 1 1 111111 111 1111111 1111 1 1 1 1 1111111111 111111111 V (x,λ) δ ẋ(t 2 ) = ẋ(t 2 )ẋ(t 1 ) neq ẋ(t 2 )ν s (x(t 1 )) neq δf(t 1 ) f µ eff (f) = D eff (f) dt I(t) with I(t) ẋ(t)ν s(x()) ν s (x) 2 1.5 1..5 v io la tio n in te g r a l m o b ility D cf giant diffusion [Reimann et al 21]..4.6.8.1.12.14.16 17

General FDT in a NESS [U.S and T. Speck, EPL 89: 17, 21] NESS : δa(t 2 ) δf(t 1 ) = A(t 2 ) f ṡ(t 1 ) = A(t 2 ) f ṡ m (t 1 ) }{{} ness + A(t 2 ) f ṡ tot (t 1 ) ness EQ : T δa(t 2) δf(t 1 ) = A(t 2 ) f Ė(t 1 ) eq stochastic entropy replaces energy add to eq form the term conj to total entropy production proof: (1) pert theory of FPE + (2) use of det bal in equilibrium 18

Sheared colloidal suspension with 3d Brownian dynamics [B. Lander, U.S. and T. Speck, EPL 92 581, 21; PRE 85, 2113, 212] driving: linear shear flow perturbating field: force on tagged particle 19

velocity-force FDT in a NESS 1.5 γ = 1., φ =.1 γ = 1., φ =.3 R ij (t) = δ v i(t) δf j () 1 C xx R xx 1.5 1 C xx R xx C ij (t) = v i (t)v j ().5.5.5 1 1.5 2 2.5 t.2.4.6.8 t γ = 1., φ =.1 γ = 1., φ =.3 effective temperature? θ i C ii(t) R ii (t) 1+a γ2 works well at larger densities 1.5 1.4 1.2 1 θ x θ y θ z.2.4.6.8 1 γ 1.5 1.8 1.5 1.2.9 θ x θ y θ z.2.4.6.8 1 γ.5 1 1.5 2 2.5 t.2.4.6.8 t 2

effective confinement for moderate/large densities Cxx,Rxx 1..5. -.5 m=1,k=1.25,θ x =2.6 A mc xx /θ x mr xx. 1. 2. 3. 4. 5. 6. 7. 8. t/m effective temperature also for trapped particle? similar math. structure in confined suspensions C ii (t) θ i R ii (t)+ γ 2 β i I i (t) with max I i (t) = 1 βx Cxx,Rxx Cxx,Rxx 1. m=1,k=12.5,θ x =1.16 B.5. -.5. 1. 2. 3. 4. 5. 6. 7. 8. t/m 1. m=.5,k=25,θ x =1.4 C.5. -.5. 1..1 1. 2. 3. 4. 5. 6. 7. 8. t/m 3 D E 2 X x θ x k m=1 1 m=.1 θ x 1.1 1. 1. 1 2 k γ 21

An isothermal nano-rotor: F1-ATP-ase [K. Hayashi,... H. Noji, PRL 14, 21813 (21)] kinetics vs thermodynamics first law? efficiency(ies)? 22

Hybrid model [E. Zimmermann and U.S., New J. Phys. 14, 1323, 212] probe particle ẋ = µ( y V(y)+f ex )+ζ with y(τ) n(τ) x(τ) motor w + /w = exp[ µ V(n+d,x) V(n,x)] local detailed balance condition 23

Simulated trajectories 1 8 6 4 2.1.2.3.4.5.6.7 t [s] c ATP =2.962 1-6 M x [12 ] c ATP =5.9494 1-5 M c ATP =1.4747 1-7 M probe motor 24

F1-ATPase and the fluctuation theorem [K. Hayashi,... H. Noji, PRL 14, 21813 (21)] Γ θ = N +ζ ζ 1 ζ 2 ) = 2Γk B Tδ(τ 1 τ 2 ) ln[p( θ)/p( θ)] = N θ/k B T independent of friction coefficient cf f theorem for total entropy production time-dependence? ln[p( s tot )/p( s tot )] = s tot /k B 25

FT-slope from simulations vs experiment p( x) 6 5 4 3 2 1 -.5.5 1 1.5 x [d] t = 1 ms t = 2 ms t = 5 ms t = 1 ms t = 2 ms 16 14 12 1 c ATP = 1.4747 1-7 c ATP = 8 1-7 c ATP = 2.962 1-6 c ATP = 5.9494 1-5 α 8 6 4 2.1.2.3.4.5.6.7 t 26

Efficiencies Thermodynamic efficiency [Parmeggiani et al, PRE 1999] η f ex v/ µ for f ex = : pseudo efficiency [Toyabe et al, PRL 21] η Q Q p / µ 27

Experimental determination of η Q [S. Toyabe et al, PRL 14, 19813 (21)] Harada-Sasa relation [PRL 26] µ Q P = v 2 + dω[cẋ(ω) 2k B T ReRẋ(ω)] from violation of fluc-diss-theorem 28

Comparison to experiment quite reasonable agreement for small θ + future: substeps of 9 + 3 degrees 29

Efficiency of autonomous nano-machines at maximum power [U.S., PRL 16, 261, 211] output: w out = fd input: w in = µ mean currents: j ± local detailed balance: j /j + = exp( S) = exp(w out w in ) power: P out,in = (j + j ) w out,in = j + (w out,w in ) [1 exp(w in w out )] w out,in efficiency: η P out /P in (= w out /w in, for unicyclic) 3

Efficiency at maximum power (EMP) case I: fix w in ( µ), maximize P out with respect to w out ( f): EMP: η w out w in 1/2+(x out +1/2)w in. with x out dlnj + /dw out 1/2 universal in LR for strong coupling (Kedem and Caplan, 1965; vd Broeck and Esposito 25...) 1. x out 5 Η.8.6.4.2. x out 1 2 x out 2 1 2 3 4 5 6 Ω in 31

Efficiency at maximum power η case III: maximize P out with respect to w out ( f) and w in ( µ) three regimes (determined by two structural parameters) 32

Thermoelectrics [Snyder et al, Nat. Mat. 7 15, 28] 33

Paradigm for thermoelectrics: Single level quantum dot heat engine [Linke et al 22..., Esposito, vd Broeck et al 25...] particle current j = (w + 1 w+ 2 w 1 w 2 )/Σw th dynamic consistency: w + 1 /w 1 = exp(µ 1 E))/T 1, w + 2 /w 2 =... output power P out = (µ 2 µ 1 )j input power (heat from hot reservoir) P in = (E µ 1 )j efficiency η P out /P in = (µ 2 µ 1 )/(E µ 1 ) η C maximize P out wrt (µ 2 µ 1 ) at fixed T 2 T 1,E : η η C /2+O(η 2 C )... and with respect to E : η η C /2+η 2 C /8 34

Thermoelectrics with magnetic field μ L T L B C μ R T R speculation: Carnot efficiency at finite power? Benenti et al, PRL 211 linear response regime ( Jρ J q ) = ( Lρρ L ρq L qρ L qq )( Fρ F q ) fundamental constraints 2 nd law Onsager relations 4L ρρ L qq (L ρq +L qρ ) 2 L ρρ,l qq L ρq (B) = L qρ ( B) 35

Three-Terminal Model [K. Brandner, K. Saito, U.S., PRL 11 763, 213] effective Onsager matrix ( Lρρ L ρq L qρ L qq ) = L LL L LP L 1 PP L PL μ L T L μ P B T P C μ R T R multiterminal Landauer formula L AB = F(E) [ new bound de F(E) ( (E µ)e (E µ)e (E µ) 2 e 2 4k B hcosh 2 ( E µ k B T )] 1 ) (δab S AB(E,B) 2) 4L ρρ L qq (L ρq +L qρ ) 2 3(L ρq L qρ ) 2 36

Efficiency dimensionless parameters y L ρq L qρ /(L ρρ L qq L ρq L qρ ) x L ρq /L qρ η max/ηc 1.8.6.4.2 bound on η max η max η C = bound on η(p max ) η (P max ) η C = x 2 x+1 x 1 x 2 x+1 x 1 x 2 4x 2 6x+4 η (Pmax)/ηC 1.8.6.4.2-4 -2-1 1 2 4 x -4-2 -1 1 2 4 x 37

Multi-terminal model [K. Brandner, U.S., NJP 15 153, 213] 1 μ 1 T 1.8 μ3 T3 μ 2 T 2 B μn T n ηmax(x)/ηc.6.4.2... 1-4 -3-2 -1 1 2 3 4 x μ4 T4.8 generalized bound 4L ρρ L qq (L ρq +L qρ ) 2 tan ( ) π n (L ρq L qρ ) 2 η (x)/ηc.6.4.2-4 -3-2 -1 1 2 3 4 x 38

A classical Nernst engine [J. Stark, K. Brandner, K. Saito, U.S., arxiv:131.1195] 5 B -βµ 2 5 A.4.8.12.16 figure of merit ZT = L 2 ρq L ρρ L qq L 2 ρq = (NB)2 σt κ 2 nd law ZT 1 new bound ZT 1/2 η max η C = 1 1 ZT 1+ 1 ZT 3 2 2.172 η(p max ) η C = ZT 4 2ZT 1/6 39

Summary and Perspectives Stochastic thermodynamics along individual trajectories formulation of the 1 st law refinement of the 2 nd law NESS extended FDT ( Green-Kubo for a NESS) emergence of an effective temperature in sheared suspensions Efficiency thermolectric transport molecular motors information machines (in cell biology) 4