Response function definition

For each EA spectrum, the transmission T was measured with the mechanical chopper in place and the electric field off. The differential transmission AT was subsequently measured without the chopper, with the electric field on, and with the lock-in amplifier set to detect signals at twice the electric-field modulation frequency. The 2/ dependency of the EA signal is due to the quadratic nature of EA in materials with definite parity. AT was then normalized to AT/T, which was free of the spectral response function. To a good approximation [18], the EA signal is related to the imaginary part of the optical third-order susceptibility ... 

Fig. 19, an unapodized spectrum [response function (sin nx)/nx = sinc(x)] is shown in trace (b). For such a spectrum there will be sidelobes and negative absorption if the natural linewidths are narrower than the full width of the sine-shaped response function. These are seen in Fig. 19, where the linewidth is three points and the response function width eight points. Here the phrase instrument response function may have a slightly different definition, but the meaning is clear. For such a response function, the direct deconvolution methods fall short. 

It follows from the definition of the linear response function in eqn (6.22) that... 

In the following, a general treatment of arbitrary binary excitation sequences will be given. Since the proper definition of the excitation and the response function is not unambiguously possible, a problem-independent notation will first be given, which will later be mapped to the actual experiment. For the moment, it is sufficient to picture a linear system with an input x(t), an output y(t) and a linear response function h(t), as sketched in Fig. 22. The input x(t) may be a pulse of finite duration, as discussed in the previous sections, or a pseudostochastic random binary sequence as in Fig. 22. 

The HSAB principle can be considered as a condensed statement of a very large amount of experimental information, but cannot be labelled a law, since a quantitative definition of the intuitive concepts of chemical hardness (T ) and softness (S) was lacking. This problem was solved when the hardness found an exact, and also an operational, definition in the framework of the Density Functional Theory (DFT) by Parr and co-workers [2], In this context, the hardness is defined as the second order derivative of energy with respect to the number of electrons and has the meaning of resistance to change in the number of electrons. The softness is the inverse of the hardness [3]. Moreover, these quantities are defined in their local version [4, 5] as response functions [6] and have found a wide application in the chemical reactivity theory [7],... 

In such a case, no conclusion about the mechanisms can be reached from the form of 4(t) and the observed rate will be determined primarily by the fastest process. By extension of the argument, one easily sees that marked deviation of any of the parallel processes from exponential decay will be reflected in the overall rate with possible change in the functional form. Thus, if the rotation is described by exp(-2D t) as in Debye-Perrin theory, and the ion displacements by a non-exponential V(t), one finds from eq 5 that 4(t) = exp(-2D t)V(t) and the frequency response function c(iw) = L4(t) = (iai + 2D ) where iKiw) = LV(t). This kind of argument can be developed further, but suffices to show the difficulties in unambiguous interpretation of observed relaxation processes. Unfortunately, our present knowledge of counterion mobilities and our ability to assess cooperative aspects of their motion are both too meagre to permit any very definitive conclusions for DNA and polypeptides. 

Let us first briefly recall these definitions, and examine under which conditions the age—and frequency-dependent response functions share the analytic properties of the corresponding stationary quantities. 

However, from the point of view of linear response theory, the definitions (174) or (178) suffer from several drawbacks. Actually, the function X ( , tw) as defined by Eq. (174) is not the Fourier transform of the function X (, x), but a partial Fourier transform computed in the restricted time interval 0 < x < tw. As a consequence, it does not possess the same analyticity properties as the generalized susceptibility x( ) defined by Eq. (179). While the latter, extended to complex values of co, is analytic in the upper complex half-plane (Smoo > 0), the function Xi ( - tw) is analytic in the whole complex plane. As a very simple example, consider the exponentially decreasing response function... 

Note, that (C.l) could be used as alternative definition for the static response functions. 

To arrive at Eq. (180) we have used the definitions (145), (148), (171) and (175) of the density response functions. Furthermore, we have abbreviated the kernel of the (instantaneous) Coulomb interaction by w(x, x ) = 3(t — t )/ r — r. Finally, by inserting Eq. (180) into (168) one arrives at the time-dependent Kohn-Sham equations for the second-order density response ... 

While gas phase work on the h5q5erpolarizability of small molecules has been relatively free of problems concerned with the definitions of measured quantities and their formal relationship to computed quantities, the same cannot be said about solution studies of rather larger organic species. It is the latter that possess the very large nonlinear response functions that are of greatest interest. The prototype system for such studies has been 4-nitroaniline (pNA) and this review is mainly concerned with the relation between the measurements, in vacuo and in solution, of the hyperpolarizabilities of pNA and the closely related molecule, MNA (2-methyl, 4-nitroaniline) to ab initio and DFT calculations of these quantities. 

The first problem is concerned with the definitions of the quantities calculated and measured. It has only gradually emerged over the last decades that there was considerable ambiguity about the definition of some of the reported parameters. In 1992 Willetts et al (referred to as WRBS) gave a detailed account of the various conventions that have been applied by different authors in defining the molecular response functions, but more recently Reiss has suggested that there are still inconsistencies in the way that the experimentally measured, macroscopic, response functions are reported. 

The potential energy surface used in solution, G (R), is related to an effective Hamiltonian containing a solute-solvent interaction term, Vint- In the implementation of the EH-CSD model, that will be examined in Section 6, use is made of the equilibrium solute-solvent potential. There are good reasons to do so however, when the attention is shifted to a dynamical problem, we have to be careful in the definition of Vint - This operator may be formally related to a response function TZ which depends on time. For simplicity s sake, we may replace here TZ with the polarization vector P, which actually is the most important component of TZ (another important contribution is related to Gdis) For the calculation of Gei (see eq.7), we resort to a static value, while for dynamic calculations we have to use a P(t) function quantum electrodynamics offers the theoretical framework for the calculation of P as well as of TZ. The strict quantum electrodynamical approach is not practical, hence one usually resorts to simple naive models. 

The convention used by Levine and Bethea to define the response functions omits the Taylor series factors in the series for the induced dipoles but includes a factor of (3/2) implicitly in the definition of the macroscopic quantity. Their ft is equivalent to jl,. Hence to relate their results to the more usual conventions, the /i-value must be multiplied by 4 x (3/2) x (3/5) = 18/5 and the y value by 4 x (3/2) = 6. Finally a factor (0.30/0.335) must be applied to allow for the change in the quartz standard. Carrying out these operations and converting to atomic units gives the values in Table 10. 

Knowing the residues of the linear response function, (Olr jp) and , we can find the transition moments between excited states

= . The last equality follows from the definition in Eq. (40) when p> and I (> are orthogonal excited states. In principle, the excited-state transition probabilities can also be obtained from the linear response function by using an excited state, say p >, as the reference state. This is a valid choice for the reference state. However, due to technical problems, such as the open-shell nature of most excited states, this has not been done yet and to obtain from the residues of the quadratic response function is probably a better method. 

Dalgaard (1982) has a factor in front of the two terms on the right-hand side of this equation. He has included the factor coming from the Taylor series expansion of the response (see Eq. (33)) in the definition of the response functions. We follow the definitions used by Olsen and Jorgensen (1985). 

Eq. (58) represents the starting point for all approximate propagator methods. Even though in the derivation we only discussed the linear response functions or polarization propagators, a similar equation holds for the electron propagator. The equation for this propagator has the same form but there are differences in the choice of h and in the definition of the binary product (Eq. (52)), which for non-number-conserving, fermion-like operators should be... 

---