Time as Density of States: Unified Time Identity \kappa = \rho and Microscopic Scattering Mechanism of Gravitational Redshift
Abstract
In conventional physics, time plays incompatible roles in quantum theory and general relativity: in quantum mechanics it is an external evolution parameter generated by a Hamiltonian, whereas in general relativity proper time is a path-dependent functional of the spacetime metric. Within a quantum cellular automaton (QCA) and optical-path conservation framework, this work proposes a microscopic definition of time in terms of the density of quantum states. Building on Eisenbud--Wigner--Smith (EWS
Full text
Time as Density of States: Unied Time Identity κ=ρ and Microscopic Scattering Mechanism of Gravitational Redshift Haobo Ma 1 Wenlin Zhang 2 1 Independent Researcher 2 National University of Singapore Abstract In conventional physics, time plays incompatible roles in quantum theory and general relativity: in quantum mechanics it is an external evolution parameter generated by a Hamiltonian, whereas in general relativity proper time is a path-dependent functional of the spacetime metric. Within a quantum cellular automaton (QCA) and optical-path conservation framework, this work proposes a microscopic denition of time in terms of the density of quantum states. Building on EisenbudWignerSmith (EWS) time-delay theory and Krein FriedelLloyd (KFL) trace formulas, we show that the trace of the WignerSmith time-delay operator Q(E) = −iℏS(E)†∂ES(E) is universally related to the relative density of scattering states via the spectral shift function. This leads to a unied time identity κ(E) = 1 2πℏtr Q(E)=∆ρ(E), where ∆ρ(E) is the change of density of states (DOS) induced by an interaction with respect to a reference background. Physically, κ(E) quanties the rate at which a system traverses quantum states in Hilbert space per unit energy. We interpret κ(E) as an intrinsic time density, and argue that macroscopic clock rates are governed by the local DOS of the underlying microscopic degrees of freedom. On this basis we develop a microscopic picture of gravitational redshift: deep in a gravitational potential well, matter and radiation experience an enhanced local DOS due to modied phase space volume and eective refractive index of the vacuum. The increased DOS yields a larger WignerSmith delay per unit energy, thereby slowing local proper time relative to distant observers. We show how, in the weak-eld limit, the DOS-based time density reproduces the standard gravitational redshift relation ∆ν/ν ≃∆Φ/c2 when normalized to asymptotically at regions, and argue that the metric component g00 can be viewed as an emergent functional of the microscopic DOS eld. The timeDOS identity is embedded into Dirac-type QCA models where relativistic dynamics emerge from discrete, strictly causal unitary updates, demonstrating that relativistic time dilation and gravitational redshift can be reinterpreted as collective properties of state-traversal rates on a discrete quantum network. Finally, we connect the proposed time density with thermodynamic time in KMS equilibrium states and the thermal time hypothesis, showing that the DOS-weighted EWS time density denes a natural arrow of time compatible with modular ow in algebraic quantum eld theory. We outline experimental and engineering proposals in microwave cavities, wavechaotic scattering systems and optical clock networks to test the quantitative link between WignerSmith delays, DOS measurements and gravitational redshift. Keywords: time delay; density of states; EisenbudWignerSmith operator; spectral shift function; KreinFriedelLloyd formula; gravitational redshift; quantum cellular automaton; optical metric; KMS state; thermal time 1
1 Introduction & Historical Context The problem of time pervades the development of physical theory. Special and general relativity take the four-dimensional spacetime manifold (M, gµν) as the fundamental object, describing proper time τ as a line element integral along worldlines. Quantum theory, by contrast, is built on Hilbert space and the Hamiltonian H , where time t is an external continuous parameter and evolution is generated by the unitary operator U(t) = exp(−iHt/ℏ) . While the two frameworks are compatible in the semiclassical limit, a unied microscopic ontology of time is lacking. Pauli's theorem indicates that in systems with an energy spectrum bounded from below, there exists no self-adjoint time operator satisfying canonical commutation relations with H , seemingly ruling out attempts to operatorize time within traditional quantum mechanics. On the other hand, time delay has long appeared as an observable in scattering theory. Eisenbud, Wigner, and Smith dened the WignerSmith group delay matrix from the energy derivative of the scattering matrix S(E) as Q(E) = −iℏS(E)†∂ES(E), whose diagonal elements Qαα(E) give the time delay in each channel, while the trace tr Q(E) characterizes the total temporal response of the system to a family of scattering states. In parallel, Krein, Friedel, Lloyd and others established the celebrated KreinFriedelLloyd formula in scattering spectral theory, relating the energy derivative of scattering phase shifts to the change in DOS ∆ρ(E) . In nite-box normalization, this formula can be written simply as ∆ρ(E) = 1 2πitrS(E)†∂ES(E)=1 π∂Eδtot(E), where δtot(E) is the total scattering phase shift. Modern spectral shift function theory further shows that the derivative of the spectral shift function ξ(E) is precisely the DOS dierence induced by the interaction, and ξ(E) can itself be characterized by the determinant phase of S(E) . These results hint at a profound fact: the time delay and the density of states in scattering systems are not two independent quantities, but rather dierent projections of the same spectral structure. Building on this foundation, the present work proposes a unied time identity that strictly equates the time ow rate κ(E) with the DOS ρ(E) . On the other hand, general relativity interprets gravitational redshift as the spatial variation of the g00 component in static spacetime metrics. In the weak-eld limit, the static gravitational potential Φ(x) and the metric satisfy g00(x)≃ −1 + 2Φ(x)/c2, and the proper time of stationary observers satises dτ=√−g00dt , so that for two positions x1,x2 at dierent potential values we have ∆ν ν≃Φ(x2)−Φ(x1) c2, a relation that has been precisely veried in the PoundRebka γ -ray experiment, the Hafele Keating around-the-world atomic clock ight, and recent optical lattice clock height comparison experiments. However, the geometric description of gravitational redshift still leaves the origin of time ow in the continuous eld of the metric tensor, lacking a direct connection to the structure of quantum states. Quantum cellular automaton and discrete quantum walk studies have shown that under appropriate symmetry conditions, the Dirac and Weyl equations can emerge as the continuum limit of local unitary discrete dynamics, thus providing rigorous models for the discrete picture of the universe as quantum computation. 2
In such a discrete ontology, the time step is a discrete parameter of the local update rule, rather than an external continuous variable. The natural question is: how can we reconstruct continuous proper time within this discrete framework, and relate it to gravitational redshift? The basic viewpoint of this paper is: time is an alias for density of states . More precisely, through the EWS time-delay operator and the KFL trace formula, we prove that under quite general scattering settings, the following unied time identity holds: κ(E) = 1 2πℏtr Q(E)=∆ρ(E) where κ(E) is interpreted as the time ow density per unit energy interval, i.e., the eective time delay experienced by the system per unit energy in the neighborhood of energy E . We further show that, under appropriate coarse-graining and normalization choices, the spatial variation of this quantity can restate gravitational redshift in general relativity. 2 Model & Assumptions 2.1 Scattering System and Density of States Let H0 be the free Hamiltonian and H=H0+V the interacting Hamiltonian containing a local potential V . We make the following assumptions: 1. H0, H are self-adjoint operators on the same Hilbert space H , and V is a suciently rapidly decaying bounded or relatively H0 -bounded perturbation, so that the wave operators Ω±= s-lim t→±∞ eiHt/ℏe−iH0t/ℏ exist and are complete. 2. The scattering operator S= (Ω+)†Ω− decomposes in the energy representation into a family of unitary matrices S(E) , acting on the channel space HE on each energy shell. 3. The spectral measures of both free and interacting systems are absolutely continuous plus a nite number of bound states, so that the well-dened state-counting functions exist: N0(E)=#{λn(H0)≤E}, N(E)=#{λn(H)≤E}, whose derivatives give the densities of states ρ0(E), ρ(E) . The change in DOS induced by the interaction is ∆ρ(E) = ρ(E)−ρ0(E). The spectral shift function ξ(E) is dened by the Krein trace formula, via trf(H)−f(H0)=Z+∞ −∞ f′(E)ξ(E) dE for suciently smooth functions f . An approximation to the indicator function yields ξ(E) = N(E)−N0(E) , whose derivative satises ξ′(E) = ∆ρ(E). On the other hand, the determinant of the scattering matrix and the spectral shift function are related by det S(E) = exp−2πiξ(E), so that 1 2πi∂Eln det S(E) = ξ′(E) = ∆ρ(E). 3
2.2 WignerSmith Time-Delay Operator In the energy representation, the WignerSmith group delay matrix is dened as Q(E) = −iℏS(E)†∂ES(E), whose diagonal element Qαα(E) gives the time delay of channel α , and the trace tr Q(E) = −iℏtrS(E)†∂ES(E) characterizes the total group delay of all channels. The self-adjointness and observability of Q(E) can be veried by dierentiating the unitarity of the scattering matrix S(E)†S(E) = I . Taking the trace and comparing with the spectral shift relation from the previous section, we immediately obtain ∆ρ(E) = 1 2πi∂Eln det S(E) = 1 2πi 1 det S(E)∂Edet S(E) = 1 2πitrS†∂ES=1 2πℏtr Q(E). This is the spectral-theoretic foundation of the unied time identity in the present work. 2.3 QCA Continuum Limit and Scattering Quantum cellular automata are a class of strictly local, discrete-time, unitary evolutions U de- ned on a lattice Λ , which in the single-particle subspace are equivalent to discrete-time quantum walks. A considerable body of work has shown that, in the limit of suciently small wavevector and mass, appropriately constructed QCA evolutions can approximate Dirac equations in various dimensions and their generalizations in curved spacetime. In the QCA framework, time is essentially the discrete step number n∈Z , and energy is given by the quasi-energy spectrum e−iε(k) of the single-step evolution operator U . Local defects or external potentials can be implemented by modifying the local update operator in a nite region, corresponding to Floquet scattering scenarios where the scattering matrix S(ε) is closely related to continuous-time scattering. Through the continuum-limit mapping ε7→ E , the FloquetEWS time delay in QCA can be unied with Q(E) in continuous scattering theory. We make the following model assumptions: 1. The universe can be described microscopically by a class of Dirac-type QCA models, whose single-particle subspace approximates the standard Dirac equation in the long-wavelength limit. 2. Macroscopic gravitational elds correspond to slow spatial variations of the QCA background cell update rules, i.e., gradual changes in local propagation velocity and phase response. This can be likened to the construction of quantum walks in curved spacetime. 3. In this framework, all macroscopic time observations can ultimately be reduced to phase dierences and group delay measurements in some kind of scattering or interference experiments, and therefore can be uniformly described by the EWSKFL structure. 3 Main Results (Theorems and Alignments) This section presents the unied time identity and its main consequences for gravitational redshift, the thermodynamic arrow of time, and the QCA continuum limit. Denition 3.1 (Time Density) . For a scattering system satisfying the above assumptions, dene the time density at energy E as κ(E)≡1 2πℏtr Q(E), where Q(E) = −iℏS(E)†∂ES(E) is the WignerSmith group delay operator. 4
Theorem 3.2 (Unied Time Identity) . Under the assumptions that the KreinFriedelLloyd formula and spectral shift function exist and are dierentiable, the time density and DOS change satisfy κ(E)=∆ρ(E) = ρ(E)−ρ0(E). In other words, the time density equals the relative DOS introduced by the interaction . This theorem unies the following three objects as dierent expressions of the same function: 1. The energy derivative of the total scattering phase φ(E) = arg det S(E) : κ(E) = 1 2π∂Eφ(E). 2. The DOS change ∆ρ(E) induced by the interaction. 3. The normalized trace of the WignerSmith delay matrix (2πℏ)−1tr Q(E) . See Appendix A for the detailed proof. Proposition 3.3 (Time Flow Rate as Function of DOS) . Suppose a macroscopic clock is driven by a family of scattering states concentrated near energy E0 , and its output period T is determined by the energy dependence of the total phase φ(E) . Then in the narrow-band approximation, the proper time increment of the clock satises dτ dt≃κref(E0) κ(E0)=∆ρref(E0) ∆ρ(E0), where ref denotes a xed reference position (e.g., the asymptotically at region at innity). This shows that within a given energy window, the time ow rate is inversely proportional to the local DOS . Theorem 3.4 (Gravitational Redshift in the Weak-Field Limit) . Consider a weak gravitational eld with static potential Φ(x) satisfying |Φ|/c2≪1 . Suppose that the Hamiltonian of some quantum eld in a local inertial frame is approximately H(x,p)≃pm2c4+c2p2+mΦ(x), and the local DOS ρ(E, x) is given by the semiclassical Weyl formula. Then for lightly bound states in the nonrelativistic limit E≃mc2 , we have ρ(E, x)≃ρ∞(E)1−αΦ(x) c2, where ρ∞(E) is the DOS at innity, and the constant α depends on the dimension and the specic model (in the three-dimensional nonrelativistic gas model α= 3/2 ). Adopting the time-density normalization dτ(x) dt=κ∞(E) κ(E, x)=ρ∞(E) ρ(E, x), a rst-order expansion yields dτ(x) dt≃1 + αΦ(x) c2. By appropriately choosing an information weight for dening the DOS (e.g., considering energy-weighted DOS or optical mode DOS), one can set α= 1 , thereby obtaining a time dilation factor consistent with the weak-eld limit of general relativity: dτ(x) dt≃1 + Φ(x) c2, 5
and recovering the gravitational redshift formula ∆ν ν≃Φ(x2)−Φ(x1) c2. See Appendix B for the detailed derivation. Proposition 3.5 (Time Density and KMS Thermal Time) . Let (A, αt) be a C∗ dynamical system, and ωβ a KMS state at temperature T= 1/(kBβ) . The energy-weighted DOS is ρβ(E) = Z−1ρ(E)e−βE, Z =Zρ(E)e−βEdE. Dene the average time density in the thermal state as ¯κβ=Zκ(E)ρβ(E) dE=Z∆ρ(E)ρβ(E) dE. Then ¯κβ increases monotonically with energy density, is one-to-one correlated with the imaginary time period β of the KMS ow, thereby providing a DOS-time realization of the thermodynamic arrow of time compatible with the thermal time hypothesis and the UnruhKMS structure. Proposition 3.6 (Time Density in QCA Continuum Limit) . In a Dirac-type QCA model, the single-step evolution operator U has quasi-energy spectrum ε(k) , and scattering defects introduce the Floquet scattering matrix S(ε) . Dene the FloquetEWS operator QF(ε) = −iUeff(ε)†∂εUeff(ε). In the continuum limit ε→E , its trace converges to the continuous scattering time delay, satisfying κ(E) = 1 2π∂Eφ(E) = 1 2πℏtr Q(E) = lim ε→E 1 2πtr QF(ε), so that the time density of discrete steps in QCA is consistent with the DOS denition in continuous theory. 4 Proofs This section provides proof outlines for the unied time identity and its main corollaries; complete technical details are placed in the appendices. 4.1 Proof of the Unied Time Identity (Theorem 3.2) The proof proceeds in two steps. Step one: KreinFriedelLloyd formula and DOS. Consider the one-dimensional case. Place the system in a nite box of length L with appropriate boundary conditions. For the free system, the momentum quantization condition is knL=nπ , and the number of states is N0(k)≃L πk, ρ0(k) = dN0 dk=L π. After adding the potential V(x) , scattering boundary conditions give the phase-shift-corrected quantization condition knL+δ(kn) = nπ, so that N(k) = L πk+1 πδ(k),∆ρ(k) = ρ(k)−ρ0(k) = 1 π∂kδ(k). 6
Using E=ℏ2k2/(2m) and the chain rule, we convert to the energy representation to obtain ∆ρ(E) = 1 π∂Eδ(E). For multi-channel and higher-dimensional cases, this can be generalized via partial-wave decomposition and spectral shift function theory to ∆ρ(E) = 1 2πitrS†∂ES=1 2π∂Eφ(E), where φ(E) = arg det S(E) , i.e., the KreinFriedelLloyd formula. Step two: EWS time delay and the trace of Q(E) . The EWS operator is dened as Q(E) = −iℏS†(E)∂ES(E), whose trace is tr Q(E) = −iℏtrS†∂ES. Comparing this with the DOS expression we obtain ∆ρ(E) = 1 2πitr(S†∂ES) = 1 2πℏtr Q(E). This is precisely the unied time identity in the time density denition. Self-adjointness follows directly from S†S=I⇒(∂ES†)S+S†(∂ES)=0, which yields Q†(E) = +iℏ(∂ES†)S=−iℏS†(∂ES) = Q(E), so Q(E) is an observable. 4.2 Time Flow Rate and DOS (Proposition 3.3) Consider a narrow-band wavepacket whose energy distribution |a(E)|2 is concentrated near E0 , satisfying Z|a(E)|2dE= 1,⟨E⟩=E0,∆E≪E0. The scattered state in the far region can be written as ψout(t) = Za(E)e−iEt/ℏS(E)|E⟩dE. Expand S(E)≃eiφ(E)˜ S(E) , where ˜ S(E) has a slowly varying matrix structure, and the total phase is φ(E) = arg det S(E) . By comparing the translation to the free propagation state, we can dene the group delay to a certain far point as τ(E0) = ∂Eφ(E)E0. The reference state (e.g., distant at region) has a corresponding delay τref(E0) determined by the scattering matrix Sref(E) , with time density κ(E) = 1 2π∂Eφ(E), κref(E) = 1 2π∂Eφref(E). Within the same external coordinate time t , the internal subjective time increment dτ is proportional to the total phase increment scanned by the wavepacket, yielding dτ dt∝∂Eφref ∂Eφ=κref(E) κ(E)=∆ρref(E) ∆ρ(E). Under appropriate normalization, the proportionality constant can be taken as 1, and the proposition follows. 7
4.3 DOS and Gravitational Redshift (Theorem 3.4 Outline) Gravitational redshift essentially arises from Killing energy conservation in static spacetime and the spatial variation of proper time for stationary observers. We wish to show that in the weak-eld limit, this eect can be restated as a spatial variation of the local DOS. In the Newtonian limit, the nonrelativistic particle Hamiltonian is H(x,p) = p2 2m+mΦ(x), and the DOS is given by the phase space volume: ρ(E, x)≃1 (2πℏ)3ZδE−H(x,p)d3p. The integral can be computed explicitly to give ρ(E, x)∝E−mΦ(x)1/2. For E≃mc2 and |Φ|/c2≪1 , expanding yields ρ(E, x)≃ρ∞(E)1−1 2 Φ(x) c2. For three-dimensional relativistic models or systems with multiple degrees of freedom, the exponent changes, giving the general form ρ(E, x)≃ρ∞(E)1−αΦ(x) c2, where α depends on the specic system. Theorem 3.4 indicates that by introducing an appropriate information weight, dening the local time ow rate as dτ(x) dt=ρinfo,∞(E) ρinfo(E, x), and taking ρinfo(E, x) = ρ(E, x)β for some function class, one can adjust the exponent β so that the rst-order expansion coecient equals 1. Physically, this is the freedom to choose a time scale among dierent microscopic models, analogous to adopting dierent coordinate scales in generally covariant theories. Matching this denition to classical gravitational redshift experiments xes β and yields dτ(x) dt≃1 + Φ(x) c2. See Appendix B for the complete calculation and normalization procedure. 4.4 Time Density and KMS Thermal Time (Proposition 3.5) For a many-body system with a continuous spectrum, the KMS condition is ωβAαt(B)=ωβαt−iβ(B)A, where αt is the Heisenberg evolution. For operators AEE′ in the energy representation, the realization of the KMS condition relies on the Boltzmann weight e−βE weighting the DOS. 8
Dene the average time density under DOS weighting: ¯κβ=Zκ(E)ρβ(E) dE=Z∆ρ(E)ρβ(E) dE. Then ¯κβ is the average time delay density of the system at temperature T . As the energy density increases, the spectral weight shifts to higher energies, ¯κβ increases monotonically, corresponding to stronger information congestion and slower macroscopic time. This is compatible with the thermal time hypothesis view that time is the modular ow of the state on the algebra: when local energy densities dier, the ratio between the modular ow parameter and geometric time changes, corresponding to dierent gravitational time dilations and temperatures. 4.5 Unied Time Identity in QCA Continuum Limit (Proposition 3.6) In a Dirac-type QCA, the single-step evolution U can be diagonalized in the momentum representation as U|k, σ⟩= e−iεσ(k)|k, σ⟩, where σ labels internal degrees of freedom. In the presence of defects, a Floquet scattering matrix S(ε) can be constructed in the long-time limit, with EWS operator QF(ε) = −iS(ε)†∂εS(ε), completely parallel to the continuous-time case. Combining the QCA continuum limit ε→E with the previous scatteringDOS theory, we obtain lim ε→E 1 2πtr QF(ε)=∆ρ(E), i.e., the time density denition is consistent between discrete and continuous descriptions. Related constructions can be found in studies of Dirac QCA and quantum walks in curved spacetime. 5 Model Apply This section discusses applications of the unied time identity in several physical scenarios. 5.1 Wave-Chaotic Cavities and Microwave Scattering In wave-chaotic cavities and multi-port electromagnetic scattering structures, the WignerSmith matrix Q(E) has long been used to characterize mode-averaged group delay and dwell time distributions, and its relation to DOS in the random matrix theory framework has been extensively veried. In these systems, experimental tests of the unied time identity can be realized in the following ways: 1. Measure the multi-port scattering matrix S(ω) , numerically dierentiate to obtain Q(ω) = −iS†∂ωS and tr Q(ω) . 2. Independently obtain DOS ρ(ω) and the relative ∆ρ(ω) with respect to the empty cavity through eigenfrequency statistics or Green's function measurements. The unied time identity predicts 1 2πtr Q(ω) = ∆ρ(ω). Deviations from this relation can be attributed to loss or non-conservative eects, requiring correction using a generalized (non-unitary) WignerSmith matrix. 9
so dτ(x) dt≃1 + βαΦ(x) c2. Choosing β= 1/α recovers the GR linear coecient 1. Physically, this corresponds to choosing a DOS denition proportional to the number of distinguishable microscopic states per unit energy in a many-body system, rather than simple single-particle phase space volume. Its specic form requires further derivation from relativistic eld theory and many-body correlation functions, which is beyond the scope of this paper. C Time Density, KMS States, and Thermal Time This appendix briey explains the relationship between time density and KMS modular ow. Let (A, αt) be a one-parameter group on a C∗ algebra, where αt is generated by a Hamiltonian H . The KMS condition is: for any A, B ∈ A , there exists an analytic function FA,B(z) satisfying FA,B(t) = ωβ(Aαt(B)), FA,B(t−iβ) = ωβ(αt(B)A). In the GNS representation, the KMS state corresponds to a special vector state on Hilbert space, whose modular ow is given by TomitaTakesaki theory. The ratio of the modular ow parameter s to physical time t can be understood as temperature or time scale. The unied time identity indicates that time density κ(E) is jointly determined by DOS and EWS delay. For a given temperature T , the system is in energy probability distribution Pβ(E) = Z−1ρ(E)e−βE, and the average time scale of the KMS ow can be related to ¯κβ=Zκ(E)Pβ(E) dE. The thermal time hypothesis holds that physical time is precisely the natural parameter of the modular ow in a given state, and temperature is the proportionality coecient between the modular ow and geometric time. In this framework, ¯κβ provides a way to directly dene this proportionality coecient using spectral quantities. D Time Density and Scattering in Dirac QCA Dirac QCA models provide discrete, local, unitary microscopic update rules, whose singleparticle subspace approximates the Dirac equation in the continuum limit. In the one-dimensional case, the single-step evolution can be written as U=X x|x+ 1⟩⟨x|⊗C++|x−1⟩⟨x|⊗C−, where C± are coin operators acting on internal degrees of freedom. Defects can be implemented by modifying C± or adding local phases in a nite region. In the quasi-energy representation, the scattering matrix S(ε) is similar to S(E) in continuous-time scattering, and the EWS operator is dened as QF(ε) = −iS(ε)†∂εS(ε). The unied time identity in the Floquet scenario becomes κ(ε) = 1 2πtr QF(ε)=∆ρF(ε), where ∆ρF(ε) is the quasi-energy DOS change induced by the interaction. The existence of the Dirac continuum limit ε→E ensures that the above relation can be seamlessly mapped to the continuous energy representation, so that the bridge between discrete time steps in QCA and continuous time density is jointly built by DOS and EWS delay. 16
References [1] E. P. Wigner, Lower limit for the energy derivative of the scattering phase shift, Physical Review 98 , 145147 (1955). [2] F. T. Smith, Lifetime matrix in collision theory, Physical Review 118 , 349356 (1960). [3] M. G. Krein, On the trace formula in perturbation theory, various Russian sources (1953); for modern expositions see F. Gesztesy and K. Makarov, The spectral shift function and its applications, lecture notes (1990s). [4] N. Graham et al., Vacuum energies and frequency dependent interactions, Physical Review D (2004); and references therein for the KreinFriedelLloyd formula. [5] P. Guo, Friedel formula and Krein's theorem in complex potential scattering theory, Physical Review Research 4 , 023083 (2022). [6] S. Richard and R. Tiedra de Aldecoa, Time delay is a common feature of quantum scattering theory, Journal of Mathematical Physics 51 , 042102 (2010). [7] C. Texier, Wigner time delay and related conceptsApplication to transport in coherent conductors, lecture notes and reviews (2016 onwards). [8] R. V. Pound and G. A. Rebka Jr., Apparent weight of photons, Physical Review Letters 4 , 337341 (1960); R. V. Pound and J. L. Snider, Eect of gravity on nuclear resonance, Physical Review Letters 13 , 539540 (1964). [9] J. C. Hafele and R. E. Keating, Around-the-World Atomic Clocks: Predicted and Observed Relativistic Time Gains, Science 177 , 166168 (1972). [10] X. Zheng et al., A lab-based test of the gravitational redshift with a miniature clock network, Nature Communications 14 , 4909 (2023). [11] Y. Tanaka et al., Exploring potential applications of optical lattice clocks in a relativistic geodesy context, Journal of Geodesy 95 , 120 (2021). [12] U. R. Patel et al., Wigner-Smith Time Delay Matrix for Electromagnetics, Physical Review Research 2 , 043345 (2020). [13] X. Gutiérrez et al., Quantum measurements and delays in scattering by zero-range potentials, Quantum Reports 6 , 17 (2024). [14] D. Yafaev, Mathematical Scattering Theory: General Theory , AMS (1992); and subsequent works on spectral shift function and trace formulas. [15] J. Earman, The Unruh Eect for Philosophers, Studies in History and Philosophy of Modern Physics 42 , 8197 (2011). [16] Recent expositions on modular ow and KMS states in quantum eld theory, e.g. recent work on analyticity and Unruh eect in local modular ow (2024). [17] A. Bisio, G. M. D'Ariano, and A. Tosini, Quantum eld as a quantum cellular automaton, Annals of Physics 354 , 244264 (2015); and The Dirac Quantum Cellular Automaton in one dimension: Zitterbewegung and scattering from potential, Physical Review A 88 , 032301 (2013). 17
[18] P. Arrighi, V. Nesme, and M. Forets, The Dirac equation as a quantum walk: higher dimensions, observational convergence, Journal of Physics A 47 , 465302 (2014); P. Arrighi et al., Quantum walking in curved spacetime, Quantum Information & Computation 17 , 810824 (2017). [19] C. Huerta Alderete et al., Quantum walks and Dirac cellular automata on a programmable trapped-ion quantum computer, Nature Communications 11 , 3720 (2020). [20] W. Arendt et al., Weyl's Law: Spectral Properties of the Laplacian in Mathematics and Physics, in Mathematical Analysis of Evolution, Information, and Complexity , Wiley (2009). [21] T. Padmanabhan, Gravity and the Thermodynamics of Horizons, Physics Reports 406 , 49125 (2005). [22] G. Feng, An optical perspective on the theory of relativityII, Optics Communications 474 , 126140 (2020); and related works on the optical metric and refractive index of gravity. [23] Representative recent works on quantum metrology and gravitational redshift in curved spacetime, e.g. S. Kish, Quantum Metrology in Curved Space-Time (PhD thesis, 2019). 18