Time correlation functions examples

All of these time correlation functions contain time dependences that arise from rotational motion of a dipole-related vector (i.e., the vibrationally averaged dipole P-avejv (t), the vibrational transition dipole itrans (t) or the electronic transition dipole ii f(Re,t)) and the latter two also contain oscillatory time dependences (i.e., exp(icofv,ivt) or exp(icOfvjvt + iAEi ft/h)) that arise from vibrational or electronic-vibrational energy level differences. In the treatments of the following sections, consideration is given to the rotational contributions under circumstances that characterize, for example, dilute gaseous samples where the collision frequency is low and liquid-phase samples where rotational motion is better described in terms of diffusional motion. 

If the rotational motion of the molecules is assumed to be entirely unhindered (e.g., by any environment or by collisions with other molecules), it is appropriate to express the time dependence of each of the dipole time correlation functions listed above in terms of a "free rotation" model. For example, when dealing with diatomic molecules, the electronic-vibrational-rotational C(t) appropriate to a specific electronic-vibrational transition becomes ... 

While between the swaps the motion of the system is somewhat realistic, it is important to emphasize that the swaps between two temperatures are nonphysical. This therefore destroys the sequencing of dynamical events (that would be required to calculate, for example, time correlation functions) and renders the dynamics and kinetics artificial. 

Finally, it may be useful to note that the Fermi golden rule and time correlation function expressions often used (see ref. 12, for example) to express the rates of electron transfer have been shown [13], for other classes of dynamical processes, to be equivalent to LZ estimates of these same rates. So, it should not be surprising that our approach, in which we focus on events with no reorganization energy requirement and we use LZ theory to evaluate the intrinsic rates, is closely related to the more common approach used to treat electron transfer in condensed media where the reorganization energy plays a central role in determining the rates but the z factor plays a second central role. 

Whenever the absorbing species undergoes one or more processes that depletes its numbers, we say that it has a finite lifetime. For example, a species that undergoes unimolecular dissociation has a finite lifetime, as does an excited state of a molecule that decays by spontaneous emission of a photon. Any process that depletes the absorbing species contributes another source of time dependence for the dipole time correlation functions C(t) discussed above. This time dependence is usually modeled by appending, in a multiplicative manner, a factor exp(-ltl/x). This, in turn modifies the line shape function I(co) in a manner much like that discussed when treating the rotational diffusion case ... 

Liquids are difficult to model because, on the one hand, many-body interactions are complicated on the other hand, liquids lack the symmetry of crystals which makes many-body systems tractable [364, 376, 94]. No rigorous solutions currently exist for the many-body problem of the liquid state. Yet the molecular properties of liquids are important for example, most chemistry involves solutions of one kind or another. Significant advances have recently been made through the use of spectroscopy (i.e., infrared, Raman, neutron scattering, nuclear magnetic resonance, dielectric relaxation, etc.) and associated time correlation functions of molecular properties. 

Relaxation times can be expressed in terms of time-correlation functions. Consider, for example, the case of a diatomic molecule relaxing from the vibrationally excited state n + 1> to the vibrational state /i> due to its interactions with a bath of solvent molecules. The Hamiltonian for the system is... 

This method of approximation is illustrated on the velocity correlation function, although it can be applied to the other time-correlation functions that have been discussed. For the purpose of this illustrative example let us assume that the second-order random force has a white spectrum. Then the continued fraction representation of i(S) is... 

The mode coupling theory of molecular liquids could be a rich area of research because there are a large number of experimental results that are still unexplained. For example, there is still no fully self-consistent theory of orientational relaxation in dense dipolar liquids. Preliminary work in this area indicated that the long-time dynamics of the orientational time correlation functions can show highly non-exponential dynamics as a result of strong in-termolecular correlations [189, 190]. The formulation of this problem, however, poses formidable difficulties. First, we need to derive an expression for the wavevector-dependent orientational correlation functions C >m(k, t), which are defined as... 

Molecular dynamics (MD) simulations [29-31], coupled with experimental observations, have played an important role in the understanding of protein hydration. They predicted that the dynamics of ordered water molecules in the surface layer is ultrafast, typically on the picosecond time scales. Most calculated residence times are shorter than experimental measurements reported before, in a range of sub-picosecond to 100 ps. Water molecules at the surface are very mobile and are in constant exchange with bulk water. For example, the trajectory study of myoglobin hydration revealed that among 294 hydration sites, the residence times at 284 sites (96.6% of surface water molecules) are less than lOOps [32]. Furthermore, the population time correlation functions... 

For the dynamical distribution it will in general be necessary to consider both the auto and cross time correlation functions of the 0-1 and the 1-2 frequencies (117). For example, if the fluctuations, <5A(t), in the anhar-monicity are statistically independent of the fluctuations in the fundamental frequency, the oscillating term (1 — elAt3) in Equation (18) would be damped. In a Bloch model the fluctuations in anharmonicity translate into different dephasing rates for the 0-1 and 1-2 transitions that were discussed previously for two pulse echoes of harmonic oscillators. Thus we see that even if A vanishes, the third-order response can be finite (94). 

Investigating molecular dynamics using NMR, in contrast to DS and LS, involves the application of several conceptually quite different techniques. For example, in spin-lattice relaxation studies one is concerned with familiar time correlation functions that are probed as spectral density point by point (Section II.D.2). In the case of line-shape analysis, usually a two-pulse echo sequence is applied, and the... 

A second problem with the GME derived from the contraction over a Liouville equation, either classical or quantum, has to do with the correct evaluation of the memory kernel. Within the density perspective this memory kernel can be expressed in terms of correlation functions. If the linear response assumption is made, the two-time correlation function affords an exhaustive representation of the statistical process under study. In Section III.B we shall see with a simple quantum mechanical example, based on the Anderson localization, that the second-order approximation might lead to results conflicting with quantum mechanical coherence. 

This is an example of the way to proceed to evaluate the four-time correlation function. 

We will present the topic by introducing the nuclear spins as probes of molecular information. Some basic formal NMR theory is given and connected to MD simulations via time correlation functions. A large number of examples are chosen to demonstrate different possible ways to combine MD simulations and experimental NMR relaxation studies. For a conceptual clarity, the examples of MD simulations presented and discussed in different sections, are arranged according to the specific relaxation mechanisms. At the end of each section, we will also specify some requirements of theoretical models for the different relaxation mechanisms in the light of the simulation results and in terms of which properties these models should be parameterized for conceptual simplicity and fruitful interpretation of experimental data. 

Perhaps one of the greatest successes of the molecular dynamics (MD) method is its ability both to predict macroscopically observable properties of systems, such as thermodynamic quantities, structural properties, and time correlation functions, and to allow modeling of the microscopic motions of individual atoms. From modeling, one can infer detailed mechanisms of structural transformations, diffusion processes, and even chemical reactions (using, for example, the method of ab initio molecular dynamics).Such information is extremely difficult, if not impossible, to obtain experimentally, especially when detailed behavior of a local defect is sought. The variety of different experimental conditions that can be mimicked in an MD simulation, such as... 

This subsection outlines the nMDS analysis of the microarray data on the gene transcriptional response of cell cycle-synchronized human fibroblasts to serum [21]. The dissimilarities are calculated as in the previous subsection. The data have been extensively analyzed with the aid of cluster analyses [21]. In contrast to the cluster analysis, for this example, nMDS clearly captures the time-dependence of the gene expression levels as if time correlation functions are explicitly analyzed (Fig. 26). 

Dynamical quantities are harder to obtain, since the QMC representations only give access to imaginary-time correlation function. With the exception of measurements of spin gaps, which can be obtained from an exponential decay of the spin-spin correlation function in imaginary time, the measurement of real-time or real-frequency correlation functions requires an ill-posed analytical continuation of noisy Monte Carlo data, for example using the Maximum Entropy Method [46-48]. 

Time correlation functions. If we look at 5p(r, f at a given r as a function of time, its time evolution is an example of a stochastic process (see Chapter 7). In a given time Zi the variables p(r, fi) and p(r, ti) = p (r, fi) —p are random variables in the sense that repeated measurements done on different identical systems will give different realizations for these variables about the average p. Again, for the random variables x = Sp(r,tf andy = 8p r,t") wq can look at the correlation... 

Here we describe two simple examples, one based on classical and the other on quantum mechanics, that demonstrate the power of time correlation functions in addressing important observables. 

The procedure described here is an example for combining theory (that relates rates and currents to time correlation functions) with numerical simulations to provide a practical tool for rate evaluation. Note that this calculation assumes that the process under study is indeed a simple rate process characterized by a single rate. For example, this level of the theory cannot account for the nonexponential relaxation of the v = 10 vibrational level of O2 in Argon matrix as observed in Fig. 13.2. 

---