We present a simple model to estimate electromagnetic wave frequency up-conversion resulting from rapidly forming gaseous plasma slabs. Such a model aids in the interpretation or planning of realizable laboratory experiments, where the plasma is neither formed instantaneously nor infinite in spatial extent. The model uses, as a basis, the behavior of an unbounded plasma when the plasma forms over extended times and considers slab boundary conditions to estimate optimum transmitted sampling windows that capture the frequency spectra of the converted waves. The results of this model are compared to exact solutions using finite difference time domain calculations, confirming its effectiveness as a tool for understanding the fundamental nature of the wave-plasma slab interactions and for planning and interpreting experimental results.

Recently, there has been a significant interest in the interaction of electromagnetic (EM) waves with linear, time-variant media.1 Unlike a time-invariant medium, in which a delay in an input to the system is reflected in its output, the delay of an input in a time-variant system can depend on the time of that input. In electromagnetic systems, a time-variant medium is characterized by how its constitutive parameters, such as the dielectric permittivity,2–10 permeability,11,12 and conductivity, vary over time either naturally or in some externally controlled manner. Despite this renewed interest, the theoretical exploration of this problem has a history spanning more than 50 years.13–18 The modulation of a medium over time opens up a diverse range of applications in engineering electromagnetics,19–21 such as non-reciprocity,22–25 wideband impedance matching,26 and frequency conversion,27–31 among others.

One potentially interesting material to study as a time-variant medium is a gaseous plasma. Experimentally, it is possible to vary the number density of electrons in time, hence varying the dielectric constant. In principal, one should be able to achieve frequency up-conversion32 with a time-switched non-magnetized plasma with a plasma density that increases in value with time.33 So far, there has been a significant body of theoretical research on the electromagnetics of time-variant plasmas.34–42 Experimental investigations, however, remain relatively limited.32,43–46

From a constitutive parameter perspective, in the case of an isotropic, time-varying plasma, the relationship between electric flux density D(t,r) to the electric field E(t,r) can be described by
(1)
where P(t,r) is the instantaneous polarization (t being the observation time) at a given location in space r, and ε0 is the vacuum permittivity. For such a time varying plasma, where the number density of free electrons is a function of time ne(t), to determine the respective P(t,r), two different models have been proposed in the literature,10 as follows.
The first model, which has been recently discussed in depth by Mirmoosa et al.,19 assumes that the electrons interact very weakly and the characteristics of movement of a single electron do not depend on the electron density, and thus, one can relate the polarization to the dipole moment p(t,r) of a single electron as P(t,r)=ne(t)p(t,r). The dipole moment is considered to be the solution of the following equation19 for a lossless plasma (with no natural frequency):
(2a)
with e and me representing the charge and mass of the free electrons. Accordingly, one finds that
(2b)
with χ(t,t,r) representing the time-variant electric susceptibility kernel that will be dependent on the observation time (t) and the delay time (t) between the action (electric field) and the response (polarization).19,20,47 For a time varying, causal, and lossless plasma, the susceptibility kernel, using this model, can be represented as
(2c)
Here, u(·) is the Heaviside function. This representation of susceptibility kernel is the same as that presented by Stepanov48 in 1976, for the case of a very slow change in plasma concentration. One substitutes this susceptibility kernel into Eq. (2b) and arrives at the following convolution (denoted by ) relation:
(2d)
The second model, discussed by Solís and Engheta,49 states that, for a lossless plasma (with no natural frequency), one requires solving for the polarization using the following equation:
(3a)
Here, the polarization is determined from the following convolution relation:
(3b)

This second model is based on the assumption that the induced polarization is a superposition of dipole moments at different instants of time10,50 attributed to the collective response of plasma electrons, appropriate at typical conditions of laboratory plasmas. This model is also in line with that presented by Stepanov48 in 1976, for the case of an increase in plasma concentration due to ionization without taking recombination into account. In this paper, we adopt this model, which has also been derived and discussed by Kalluri.33 

In the case of a time-invariant plasma, which applies under conditions in which the plasma density is constant (or varies on time scales long compared to the wave period), using either of Eqs. (2d) and (3b), one arrives at the following convolution and multiplicative relations in the time and frequency domains, respectively:
(4a)
where
(4b)
(4c)
It should be noted that εp(ω,r) is the commonly used Drude model for the plasma dielectric constant in the frequency domain. While the dielectric constant for the time-invariant case can be easily represented as a function of time, we note that in the time-variant case, it is impossible, in general, to get a closed form for εp(t,r) in such a way that relates the electric flux density to the electric field with a simple relation as that of Eq. (4a). This precludes the ability, in a simple way, to determine the frequency spectrum of transmitted waves that are passing through a time-varying plasma medium.

While in general, one can compute the problem using finite difference time domain (FDTD) simulations, we describe here, below, a useful approximate model which aids in the understanding of the physics of the problem of EM wave interaction with a time-varying plasma medium. In addition, it will simplify the design or interpretation of experiments that avoid having to carry out FDTD simulations, at least for simple one-dimensional configurations of an electromagnetic wave passing through a plasma slab with an increasing electron number density. Our discussion first starts with the problem of an electromagnetic wave traveling through a time-varying but spatially uniform and unbounded plasma. Then, we focus on a time-varying plasma slab with a spatially finite size.

We consider the case of a rapidly created plasma immersed in a monochromatic electromagnetic wave. We start our analysis by first considering the switching on, at a time t = 0, a spatially unbounded, and isotropic non-magnetized plasma.

Considering first, Maxwell's equations that link the electric field, magnetic field, H(t,r), and the polarization as follows:
(5a)
(5b)
Through the curl operation on Eq. (5a) and substituting into the arrived equation, Eqs. (5b) and (3b), one obtains the wave equation for the electric field, E(t,r):33 
(5c)
where c is the speed of light in vacuum and ωp(t) is the plasma frequency that is deemed to be varying in time.
In what follows, we assume that for t < 0 (there is no plasma) and the plasma frequency is zero and a wave with a frequency of ω0=2πf0 is present in the medium. For the one-dimensional (1-D) problem of a traveling wave in the +z direction (with a time convention of e+jωt), we have the following:
(5d)
where k0 is the wave number in free space at f=f0. Considering the dispersion diagram of an unbounded plasma shown in Fig. 1, for an instantaneous rise in plasma density (hence plasma frequency) to some nominal value of ωp0, i.e., ωp(t)=ωp0u(t), the wave will experience an instantaneous up-conversion in frequency from ω0=ck0 to ωup=ω02+ωp02, which will persist for t>0 as long as there is no plasma decay. It migrates from the original frequency (f0) on the light line (or, ωp=0), depicted by the lower gray square, to the upper square representing the up-converted frequency in such a way that the corresponding wave number k is preserved. In other words, the wave number in the plasma at a given frequency (ω=2πf) is always equal to its original value before switching the plasma on, i.e., k(ω)=k(ω0)=k0.
FIG. 1.

Dispersion diagram of wave propagation inside an unbounded plasma medium for various values of plasma frequency. Here, ωp=0 represents the light line in free space.

FIG. 1.

Dispersion diagram of wave propagation inside an unbounded plasma medium for various values of plasma frequency. Here, ωp=0 represents the light line in free space.

Close modal
In practice, it is not possible to perform such an instantaneous creation and increase in plasma density, and there exists a finite time for the transition from a non-ionized state to some maximum value corresponding to a plasma frequency ωp=ωp0. From an experimental standpoint, we seek to have an understanding and straightforward methodology of quantifying the up-converted frequency spectrum in the case of a finite time over which the plasma is created. Without loss of generality, here we assume a case in which the square of the plasma frequency (which is proportional to plasma electron number density) varies in time in the following way:
(6)
where τ is the characteristic time constant over which the plasma density reaches its maximum value. Since ω02=k02c2, the wave equation in this uniform unbounded medium reduces to
(7)
This equation can be solved numerically. Although we are focusing on an unbounded medium, the numerical results described below will help us to determine, in an approximate way, the frequency conversion in a spatially bounded plasma as well. We note that in the unbounded case, any practical sampling of the wave would be taken over some finite period of time following the initial creation of the plasma, and the observed spectrum would be dependent on this selected time period. As an example, we assume to have an initial wave frequency of f0=1 GHz (ω0=2πf0 rad/s) with a wave period of T0=1/f0=1 ns, and also, a steady-state plasma frequency of fp0=10 GHz (ωp0=2πfp0 rad/s). As a result, the maximum up-converted frequency is expected to be fup=f02+fp0210.05 GHz. We shall consider the following cases: (a) a relatively slow rise in plasma density, or “slow switching” on of the plasma with τT0; and (b) a moderate rise or “fast switching,” with τ<2T0.

Fairly routine pulsed power sources can have rise times of order of 20–2000 ns and when used to generate pulsed plasmas, can produce increases in plasma density via ionization on time scales of similar order. These times preclude the possibility of large frequency conversions in the GHz range and would be better suited for MHz scale frequency conversion. However, most practical plasma sources may have collisional damping rates that would considerably hinder the electron response at these lower frequencies. At GHz frequencies, this plasma density rise is therefore considered to be a slow one, i.e., what we would refer to as slow switching. Considering a range of switching speeds (leading to the square of the plasma frequency profiles shown in Fig. 2), we present the computed electric field for 5 ns <t<10 ns (wherein the plasma forms initially at t=0) in Fig. 3. The negative time refers to the presence of the original wave at f0=1 GHz. It is clear from the behavior of the wave field at t>0 that a higher switching speed (lower characteristic time τ) leads to a greater frequency component in the up-converted field. To better understand the spectrum generated, we consider the response within three sampling time windows following the onset of the plasmas.

FIG. 2.

The normalized, square of the plasma frequency profiles for slow switching speeds with τ=20 ns, τ=200 ns, τ=2000 ns, and τ.

FIG. 2.

The normalized, square of the plasma frequency profiles for slow switching speeds with τ=20 ns, τ=200 ns, τ=2000 ns, and τ.

Close modal
FIG. 3.

Electric field results for slow switching speeds with τ=20 ns, τ=200 ns, and τ=2000 ns.

FIG. 3.

Electric field results for slow switching speeds with τ=20 ns, τ=200 ns, and τ=2000 ns.

Close modal

First, we assume to have available to us a relatively long period of time to sample the converted fields, say 0<t<1000 and 0<t<100 ns, respectively. The corresponding Fourier spectrum of the converted wave over these long sampling windows for the various switching speeds is shown in Figs. 4 and 5, respectively. The power values are normalized to the power of the reference signal, which is cos(ω0t) within the given time window (0<t<1000 and 0<t<100 ns, respectively). It can be seen that, by increasing the switching speed (or, lowering the rise time, τ), the power will be distributed over a wider range of frequencies. For example, for the case with 0<t<100 ns, frequency conversion can be achieved up to almost 2.36, 6, and 10 GHz (below the limit fup=10.05 GHz), for τ=2000 ns, τ=200 ns, and τ=20 ns, respectively.

FIG. 4.

Frequency spectrum of the electric fields with a long time period (0<t<1000 ns) for slow switching speeds with τ=20 ns, τ=200 ns, and τ=2000 ns.

FIG. 4.

Frequency spectrum of the electric fields with a long time period (0<t<1000 ns) for slow switching speeds with τ=20 ns, τ=200 ns, and τ=2000 ns.

Close modal
FIG. 5.

Frequency spectrum of the electric fields with a long time period (0<t<100 ns) for slow switching speeds with τ=20 ns, τ=200 ns, and τ=2000 ns.

FIG. 5.

Frequency spectrum of the electric fields with a long time period (0<t<100 ns) for slow switching speeds with τ=20 ns, τ=200 ns, and τ=2000 ns.

Close modal

Now, we focus on the case of a short sampling time period, i.e., 0<t<10 ns. In other words, we are interested in frequency up-conversion within a short period of time after the initial plasma creation. The analysis here would be informative for the case of a finite plasma slab that will be discussed later. Figure 6 shows the frequency up-conversion of the system for various values of the switching speeds. For the very slow case of τ=2000 ns, the up-converted frequency is only slightly higher than the original frequency, f=1.12,f0=1.12 GHz, but with a power that is of a similar level compared to that of the original frequency. For τ=20 ns and τ=200 ns, the range of converted frequency is shifted toward lower values, and the respective power is increased for the up-converted frequencies, compared to the results for the long sampling time period discussed above. While in any practical sense, we see that restricting the sampling time of the converted waves greatly restricts the range of frequency up-shift in this unbounded plasma. We shall show later that this restriction in sampling time occurs naturally, in a bounded plasma slab.

FIG. 6.

Frequency spectrum of the electric fields with a short time period (0<t<10 ns) for slow switching speeds with τ=20 ns, τ=200 ns, and τ=2000 ns.

FIG. 6.

Frequency spectrum of the electric fields with a short time period (0<t<10 ns) for slow switching speeds with τ=20 ns, τ=200 ns, and τ=2000 ns.

Close modal

In this section, we focus on the response associated with faster switching speeds, i.e., equal to the wave period (τ=T0=1 ns), and shorter, with τ=0.2T0=0.2 ns and τ=0.02T0=0.02 ns. Figure 7 shows the computed electric field for 5 ns <t<10 ns (where again, the onset of the plasma rise is at t=0).

FIG. 7.

Electric field results for slow switching speeds with τ=0.02 ns, τ=0.2 ns, and τ=1 ns.

FIG. 7.

Electric field results for slow switching speeds with τ=0.02 ns, τ=0.2 ns, and τ=1 ns.

Close modal

If a long enough sampling period of time is considered (i.e., 0<t<100 ns), we find that these moderate and fast switching speeds lead to frequency up-conversion very close to the limiting frequency (fup=10.05 GHz), as shown in Fig. 8. The up-converted power at 10.05 GHz for the τ=0.02 ns is almost 3 and 5 dB larger than that of the cases with τ=0.2 ns and τ=1 ns, respectively. Also, moderate switching (τ=1 ns) leads to a sideband in the up-converted frequency at f=9.95 GHz, almost 27 dB below its converted power at fup=10.05 GHz. As expected, the power will be distributed within a wider frequency range when lower switching speeds are used.

FIG. 8.

Frequency spectrum of the electric fields with a long time period (0<t<100 ns) for fast switching speeds with τ=0.02 ns, τ=0.2 ns, and τ=1 ns.

FIG. 8.

Frequency spectrum of the electric fields with a long time period (0<t<100 ns) for fast switching speeds with τ=0.02 ns, τ=0.2 ns, and τ=1 ns.

Close modal

In the case of a shorter time window (i.e., 0<t<10 ns), Fig. 9 illustrates that when τ=0.02T0=0.02 ns, the frequency response is similar to that of the original wave, with the peak power up-converted to the desired frequency of fup=10.05 GHz. One can also notice the spreading of the frequency response compared to the longer time window for τ=0.2 ns and τ=1 ns.

FIG. 9.

Frequency spectrum of the electric fields with a short time period (0<t<10 ns) for fast switching speeds with τ=0.02 ns, τ=0.2 ns, and τ=1 ns.

FIG. 9.

Frequency spectrum of the electric fields with a short time period (0<t<10 ns) for fast switching speeds with τ=0.02 ns, τ=0.2 ns, and τ=1 ns.

Close modal
Finally, we focus our attention on the problem of a 1-D plasma slab with a finite length d along the z-axis. This type of configuration is more realizable in practice, i.e., we expect both space and time interfaces. As for the unbounded case, for t<0, there exists a plane wave at f=f0, and at t=0, this finite width plasma slab is created with the square of plasma frequency profile of the form given in Eq. (6). The corresponding wave equations for t>0 will be33 
(8a)
(8b)
(8c)
wherein EL, Ep, and ER represent the electric fields on the left side (z<0), inside (0zd), and on the right side (d<z) of the plasma slab, respectively. Although one can treat this problem numerically and solve it for the fields in all these regions, here we take a different, approximate, but simple approach, as follows.

When the plasma slab is initially created at t=0, the already present wave within the slab at the initial frequency (f0) converts to other frequencies with the same wavelength throughout time (i.e., the wavenumber within the slab is preserved). Theoretically, if the plasma slab considered is made increasingly larger, one would approach the problem of an unbounded plasma medium, for which one would have access to an opportunity for a longer sampling time window, as described below. We have already shown how the available sample time window can affect the generated frequencies and their respective powers within an unbounded plasma.

Ostensibly, one might not find any obvious relationship between the problems of the time-varying spatially unbounded and bounded plasmas (such as the finite slab shown in Fig. 10). However, here we show how one can analyze the problem of transmission through a switched finite plasma slab using the wave equation [i.e., Eq. (7)] for field solutions within a spatially unbounded plasma. Before moving into the details of the analysis and the proposed model for a finite slab, first, let's review what defines a time varying, spatially unbounded 1-D plasma. The spatially unbounded plasma is: (a) not bounded spatially by any boundaries, and, as a result, for a wave traveling in the +z direction (with a spatial dependency of ejkz), there will be no spatial reflections within the plasma with a spatial dependency of e+jkz. Also, (b) there exists an available sampling time window that adequately captures the transient behavior of the problem. Since the medium is unbounded, this time window is not dependent on the size of the plasma but on the capability of the sampler in time. To be able to link the problem of a time varying, finite slab to the unbounded one, two conditions should be satisfied: (a) for a finite slab, although reflections within plasma are unavoidable, it should be possible to neglect their effects in analyzing the transmission in the case of minimal transmission of multiple-reflected wave components through the medium (we will investigate this aspect here first); and (b) there exists an effective time window (i.e., 0<t<teff) that captures the transient behavior of the problem within a finite slab. Unlike an unbounded medium, here we show that such an effective time window is proportional to the slab size, and if transmission of the converted waves is greatly limited by the scattering at the interface, this effective time is proportional to d/c.

FIG. 10.

The generated waves in the plasma and free space after the switch is turned ON. It should be noted that kf and k denote the wave numbers in free space and plasma, respectively, as functions of frequency (ω).

FIG. 10.

The generated waves in the plasma and free space after the switch is turned ON. It should be noted that kf and k denote the wave numbers in free space and plasma, respectively, as functions of frequency (ω).

Close modal
We start our analysis by looking at the transmission through the slab. The onset of the plasma converts the wave of initial frequency and amplitude to waves with different frequencies and amplitudes (but the same wavelength) within the aforementioned effective time window. Once this effective time window is passed, one can assume that the plasma will be in a steady state. Then, the magnitude of the transmission from plasma to free space |T(ω)| can be found using51 
(9)
wherein Z0 and Zp(ω) represent the wave impedance in free space and within the plasma slab, respectively. One can determine the impedance given that the wave number is preserved, i.e.,
(10)
wherein εp(ω) is the relative permittivity of the time varying plasma as a function of frequency, and kf(ω)=ω/c is the wave number in free space at any given frequency (ω). The wave impedance can then be found as
(11)
With this, one finds that
(12)
The relative transmitted power into free space will then be proportional to the square of this magnitude. This relative transmitted power is presented in Fig. 11, for a wave with an initial frequency of f0=1 GHz. It can be seen that the discontinuity in the wave impedance at the boundary of the time varying plasma and free space leads to a frequency response mimicking a low pass filter for frequencies above the initial frequency (f0=1 GHz). This result is particular to plasmas but can be applied to other materials if they have a similar frequency-dependent permittivity.
FIG. 11.

The square of the magnitude of the transmission (|T(f)|), normalized to its value at f0=1 GHz in dB as a function of frequency.

FIG. 11.

The square of the magnitude of the transmission (|T(f)|), normalized to its value at f0=1 GHz in dB as a function of frequency.

Close modal

Now, we observe that waves with higher frequency components, despite possessing faster phase velocities (as shown in Fig. 1) and passing through the slab within a shorter time window, have minimum transmission through the slab due to the filtering effect forced by the impedance discontinuity at the spatial boundary, z=d. As a result, the slab behaves, in some ways, like a cavity for these frequency components. They will experience multiple reflections within the plasma medium, and their transmission decrease as time progresses. Their main contribution to the sampled transmitted fields will then be their first transmission (spatially) through the slab. In contrast, lower frequency components have high transmission from the plasma slab to free space but with lower phase velocities. The waves that are reflected at the boundary z=d strike the boundary at z=0 and are reflected back into the plasma. Since the reflection is much smaller than their transmission, their second and following transmissions from the plasma slab to free space have much smaller magnitudes and can then be neglected. Accordingly, similar to the higher frequencies, their main contribution to the sampled transmitted fields will also be their first transmission (spatially) through the slab. To summarize, both low and high frequency components contribute to the transmission process mainly through their first transit through the slab. The problem of transmitting through a finite plasma slab, therefore, approximately mimics the problem of an unbounded medium with waves traveling in the +z direction when sampled over a time window comparable to this transit time through the slab.

Considering Fig. 1 again, the minimum phase velocity is equal to the speed of light in free space c and is attributed to the wave with a frequency of f0 (initial frequency). For a finite slab with a size d, the effective time window should be selected in such a way that both low and frequency components pass through the slab at least once. The time it takes for the lowest frequency component (the slowest wave), located at z=0, to reach z=d for the first time is d/c. As a result, we have teff>d/c. For our proposed simple model, we consider to have an effective time window wherein this condition is valid. In addition, we choose the effective time to be the time that it takes for the lowest frequency component (the slowest wave), located right next to z=d, to approach the boundary at z=d again (just before striking it for the second time), i.e., teff=2d/c. This ensures that the first transmission out of the plasma occurs for the lowest frequencies.

As a result, it can be shown that one can link the problem of transmission through a time varying plasma slab to the wave solutions within an unbounded plasma slab by selecting an effective time window that almost captures the physics of the problem. In summary, our model can be described in two steps.

  • Step 1: For a 1-D, finite plasma slab with the size d, one solves the following equation within the effective time window 0<t<teff=2d/c:
    (13)
  • Step 2: By taking the Fourier transform of Ep+(t), i.e., Ep+(ω), the magnitude of the transmitted wave in the Fourier domain can then be found using
    (14)

As follows, we present the results that show the effectiveness of the presented model, and also, to analyze the effect of a plasma slab with a finite size on the creation of new frequencies and their transmission through plasma to free space. Without loss of the generality of the problem, here we consider two cases with plasma slab widths of d=5λ0 and d=λ0. A slab width of d=5λ0 leads to an effective time window of 0<t<teff=10T0=10 ns. The resulting normalized transmitted power is shown in Fig. 12. An interesting observation is that similar to the unbounded scenario shown in Figs. 6 and 9, a faster switching speed leads to higher converted frequencies. However, due to the filtering behavior of impedance discontinuity, the accessible transmitted power to free space is reduced drastically for frequencies far from the initial frequency (f0). On the other hand, the frequencies generated by slow switching, though being lower in terms of frequency conversion, will transmit similar power levels (compared to the unbounded case) from plasma to free space. Also, to complete the study and confirm the model, we have slightly modified the FDTD code presented in Lee's dissertation52 to solve for the finite slab, but with the square of plasma frequency mimicking the exponential profile presented in this paper. For the sake of brevity, we have not presented the details of the FDTD technique here. We refer the reader to Lee's dissertation52 for further details and the provided computational script. The FDTD results are shown in Fig. 13. It can be clearly seen that the presented model can predict the frequency response of the systems reasonably well. However, we believe that the presented approximate model conveys a better understanding of the physics of the problem.

FIG. 12.

The normalized transmitted power calculated using the proposed model for a slab with d=5λ0 in dB as a function of frequency.

FIG. 12.

The normalized transmitted power calculated using the proposed model for a slab with d=5λ0 in dB as a function of frequency.

Close modal
FIG. 13.

The normalized transmitted power calculated using the FDTD method for a slab with d=5λ0 in dB as a function of frequency.

FIG. 13.

The normalized transmitted power calculated using the FDTD method for a slab with d=5λ0 in dB as a function of frequency.

Close modal

For the case of a slab with d=λ0, very slow switching speeds (i.e., τ=2000 ns and τ=200 ns) lead to either no frequency conversion or a very slight frequency change as shown in Fig. 14. On the other hand, moderate and moderately fast speeds (i.e., τ=1 ns and τ=0.2 ns) lead to wideband frequency conversions. As a result, a smaller slab size requires faster switching speeds to be able to compensate for its shorter effective time window. Here, again, we have presented the FDTD results in Fig. 15. It can be seen how the results from such a simple presented model agree with the FDTD results.

FIG. 14.

The normalized transmitted power calculated using the proposed model for a slab with d=λ0 in dB as a function of frequency.

FIG. 14.

The normalized transmitted power calculated using the proposed model for a slab with d=λ0 in dB as a function of frequency.

Close modal
FIG. 15.

The normalized transmitted power calculated using the FDTD method for a slab with d=λ0 in dB as a function of frequency.

FIG. 15.

The normalized transmitted power calculated using the FDTD method for a slab with d=λ0 in dB as a function of frequency.

Close modal

We have presented a simple model for analyzing frequency up-conversion in linear, time-varying plasma slabs. The model links the problem of transmission through a time-varying plasma slab to the wave solutions within an unbounded plasma slab by selecting an effective time window that captures the most important contributions to the physics of the problem. We have considered various slab widths and plasma switching speeds, and the associated results of this model have been compared to the FDTD calculations, confirming its effectiveness as a useful tool paving the way for understanding the interactions of EM waves with time-varying plasma slabs, and also, for planning and interpreting experimental results.

The authors would like to thank Professor Kalluri for useful discussions and for directing us to the FDTD simulation in the thesis of J. H. Lee. This work was supported by an Air Force Office of Scientific Research Multi-University Research Initiative (MURI) (Grant No. FA9550-21-1-0244), with Dr. Arje Nachman as the Program Officer. The authors acknowledge the original support of Dr. Mitat Birkan, who was the AFOSR Program Officer prior to his retirement.

The authors have no conflicts to disclose.

Hossein Mehrpour Bernety: Conceptualization (equal); Data curation (lead); Formal analysis (lead); Investigation (equal); Methodology (equal); Project administration (equal); Validation (equal); Visualization (equal); Writing – original draft (equal); Writing – review & editing (equal). Mark A. Cappelli: Conceptualization (lead); Data curation (lead); Formal analysis (equal); Funding acquisition (lead); Investigation (equal); Methodology (equal); Project administration (equal); Resources (lead); Validation (equal); Visualization (equal); Writing – original draft (equal); Writing – review & editing (equal).

The data that support the findings of this study are available from the corresponding author upon reasonable request.

1.
E.
Galiffi
,
R.
Tirole
,
S.
Yin
,
H.
Li
,
S.
Vezzoli
,
P. A.
Huidobro
,
M. G.
Silveirinha
,
R.
Sapienza
,
A.
Alù
, and
J. B.
Pendry
, “
Photonics of time-varying media
,”
Adv. Photonics
4
,
014002
(
2022
).
2.
A. G.
Hayrapetyan
,
J. B.
Götte
,
K. K.
Grigoryan
,
S.
Fritzsche
, and
R. G.
Petrosyan
, “
Electromagnetic wave propagation in spatially homogeneous yet smoothly time-varying dielectric media
,”
J. Quant. Spectrosc. Radiat. Transfer
178
,
158
166
(
2016
).
3.
D.
Sarkar
,
FDTD Analysis of Guided Electromagnetic Wave Interaction with Time Modulated Dielectric Medium
(
Springer
,
2022
).
4.
D.
Yang
,
J.
Xu
, and
D. H.
Werner
, “
A generalized temporal transfer matrix method and its application to modeling electromagnetic waves in time-varying chiral media
,”
Appl. Phys. Lett.
122
,
251102
(
2023
).
5.
I.
Aberg
,
G.
Kristensson
, and
D. J. N.
Wall
, “
Propagation of transient electromagnetic waves in time-varying media-direct and inverse scattering problems
,”
Inverse Probl.
11
,
29
(
1995
).
6.
D. E.
Holberg
,
Electromagnetic Wave Propagation in Media of Periodically Time-Varying Permittivity
(
Sandia Corporation
,
1965
), Vol.
889
.
7.
S.
Kozaki
, “
Reflection of electromagnetic wave from a time-varying medium
,”
Electron. Lett.
14
,
826
828
(
1978
).
8.
R.
Carminati
,
H.
Chen
,
R.
Pierrat
, and
B.
Shapiro
, “
Universal statistics of waves in a random time-varying medium
,”
Phys. Rev. Lett.
127
,
094101
(
2021
).
9.
D. M.
Solís
,
R.
Kastner
, and
N.
Engheta
, “
Time-varying materials in the presence of dispersion: Plane-wave propagation in a Lorentzian medium with temporal discontinuity
,”
Photonics Res.
9
,
1842
1853
(
2021
).
10.
T. T.
Koutserimpas
and
F.
Monticone
, “
Time-varying media, dispersion, and the principle of causality
,”
Opt. Mater. Express
14
,
1222
1236
(
2024
).
11.
T. T.
Koutserimpas
and
R.
Fleury
, “
Electromagnetic fields in a time-varying medium: Exceptional points and operator symmetries
,”
IEEE Trans. Antennas Propagat.
68
,
6717
6724
(
2020
).
12.
A.
Ortega-Gomez
,
M.
Lobet
,
J. E.
Vázquez-Lozano
, and
I.
Liberal
, “
Tutorial on the conservation of momentum in photonic time-varying media
,”
Opt. Mater. Express
13
,
1598
1608
(
2023
).
13.
H. G.
Booker
, “
Oblique propagation of electromagnetic waves in a slowly-varying non-isotropic medium
,”
Proc. R. Soc. London, Ser. A
155
,
235
257
(
1936
).
14.
F. R.
Morgenthaler
, “
Velocity modulation of electromagnetic waves
,”
IEEE Trans. Microwave Theory Tech.
6
,
167
172
(
1958
).
15.
D.
Holberg
and
K.
Kunz
, “
Parametric properties of fields in a slab of time-varying permittivity
,”
IEEE Trans. Antennas Propagat.
14
,
183
194
(
1966
).
16.
L.
Felsen
and
G.
Whitman
, “
Wave propagation in time-varying media
,”
IEEE Trans. Antennas Propagat.
18
,
242
253
(
1970
).
17.
R.
Fante
, “
Transmission of electromagnetic waves into time-varying media
,”
IEEE Trans. Antennas Propagat.
19
,
417
424
(
1971
).
18.
T.
Ruiz
,
C.
Wright
, and
J.
Smith
, “
Characteristics of electromagnetic waves propagating in time varying media
,”
IEEE Trans. Antennas Propagat.
26
,
358
361
(
1978
).
19.
M.
Mirmoosa
,
T.
Koutserimpas
,
G.
Ptitcyn
,
S.
Tretyakov
, and
R.
Fleury
, “
Dipole polarizability of time-varying particles
,”
New J. Phys.
24
,
063004
(
2022
).
20.
G.
Ptitcyn
,
M. S.
Mirmoosa
,
A.
Sotoodehfar
, and
S. A.
Tretyakov
, “
A tutorial on the basics of time-varying electromagnetic systems and circuits: Historic overview and basic concepts of time-modulation
,”
IEEE Antennas Propag. Mag.
65
,
10
20
(
2023
).
21.
P.
Jayathurathnage
,
F.
Liu
,
M. S.
Mirmoosa
,
X.
Wang
,
R.
Fleury
, and
S. A.
Tretyakov
, “
Time-varying components for enhancing wireless transfer of power and information
,”
Phys. Rev. Appl.
16
,
014017
(
2021
).
22.
D. L.
Sounas
and
A.
Alù
, “
Non-reciprocal photonics based on time modulation
,”
Nat. Photonics
11
,
774
783
(
2017
).
23.
T. T.
Koutserimpas
and
R.
Fleury
, “
Nonreciprocal gain in non-Hermitian time-Floquet systems
,”
Phys. Rev. Lett.
120
,
087401
(
2018
).
24.
S.
Taravati
,
N.
Chamanara
, and
C.
Caloz
, “
Nonreciprocal electromagnetic scattering from a periodically space-time modulated slab and application to a quasisonic isolator
,”
Phys. Rev. B
96
,
165144
(
2017
).
25.
X.
Wang
,
G.
Ptitcyn
,
V.
Asadchy
,
A.
Díaz-Rubio
,
M. S.
Mirmoosa
,
S.
Fan
, and
S. A.
Tretyakov
, “
Nonreciprocity in bianisotropic systems with uniform time modulation
,”
Phys. Rev. Lett.
125
,
266102
(
2020
).
26.
A.
Shlivinski
and
Y.
Hadad
, “
Beyond the Bode-Fano bound: Wideband impedance matching for short pulses using temporal switching of transmission-line parameters
,”
Phys. Rev. Lett.
121
,
204301
(
2018
).
27.
V. R.
Goteti
and
D. K.
Kalluri
, “
Wave propagation in a switched-on time-varying plasma medium
,”
IEEE Trans. Plasma Sci.
17
,
828
833
(
1989
).
28.
Z.
Wu
and
A.
Grbic
, “
Serrodyne frequency translation using time-modulated metasurfaces
,”
IEEE Trans. Antennas Propagat.
68
,
1599
1606
(
2020
).
29.
M.
Liu
,
D. A.
Powell
,
Y.
Zarate
, and
I. V.
Shadrivov
, “
Huygens' metadevices for parametric waves
,”
Phys. Rev. X
8
,
031077
(
2018
).
30.
K.
Lee
,
J.
Son
,
J.
Park
,
B.
Kang
,
W.
Jeon
,
F.
Rotermund
, and
B.
Min
, “
Linear frequency conversion via sudden merging of meta-atoms in time-variant metasurfaces
,”
Nat. Photonics
12
,
765
773
(
2018
).
31.
X.
Zou
,
Q.
Wu
, and
Y. E.
Wang
, “
Monolithically integrated parametric mixers with time-varying transmission lines (TVTLs)
,”
IEEE Trans. Microwave Theory Tech.
68
,
4479
4490
(
2020
).
32.
S.
Wilks
,
J.
Dawson
, and
W.
Mori
, “
Frequency up-conversion of electromagnetic radiation with use of an overdense plasma
,”
Phys. Rev. Lett.
61
,
337
(
1988
).
33.
D. K.
Kalluri
,
Electromagnetics of Time Varying Complex Media: Frequency and Polarization Transformer
(
CRC Press
,
2018
).
34.
W. B.
Gail
, “
Theory of electromagnetic cyclotron wave growth in a time-varying magnetoplasma
,”
J. Geophys. Res.
95
,
19089
19097
, https://doi.org/10.1029/JA095iA11p19089 (
1990
).
35.
T. T.
Huang
,
J. H.
Lee
,
D. K.
Kalluri
, and
K. M.
Groves
, “
Wave propagation in a transient plasma-development of a Green's function
,”
IEEE Trans. Plasma Sci.
26
,
19
25
(
1998
).
36.
M. M.
Ehsan
and
D. K.
Kalluri
, “
Plasma induced wiggler magnetic field in a cavity: II-FDTD method for a switched lossy plasma
,”
Int. J. Infrared Millimeter Waves
24
,
1655
1676
(
2003
).
37.
D. K.
Kalluri
,
J. H.
Lee
, and
M. M.
Ehsan
, “
FDTD simulation of electromagnetic pulse interaction with a switched plasma slab
,”
Int. J. Infrared Millimeter Waves
24
,
349
365
(
2003
).
38.
D. K.
Kalluri
and
J.
Chen
, “
Comparison identities for wave propagation in a time-varying plasma medium
,”
IEEE Trans. Antennas Propagat.
57
,
2698
2705
(
2009
).
39.
L.
Song
,
E. L.
Tan
,
D. Y.
Heh
,
B.
Bai
,
X.
Li
, and
Y.
Liu
, “
Newton-ADE FDTD method for time-varying plasma
,”
IEEE Trans. Microwave Theory Tech.
71
,
1447
1457
(
2023
).
40.
W.
Chen
,
L.
Guo
,
J.
Li
, and
S.
Liu
, “
Research on the FDTD method of electromagnetic wave scattering characteristics in time-varying and spatially nonuniform plasma sheath
,”
IEEE Trans. Plasma Sci.
44
,
3235
3242
(
2016
).
41.
M.
Bakunov
and
A.
Maslov
, “
Trapping of an electromagnetic wave by the boundary of a time-varying plasma
,”
Phys. Rev. E
57
,
5978
(
1998
).
42.
M.
Wang
,
M.
Yu
,
Z.
Xu
,
G.
Li
,
B.
Jiang
, and
J.
Xu
, “
Propagation properties of terahertz waves in a time-varying dusty plasma slab using FDTD
,”
IEEE Trans. Plasma Sci.
43
,
4182
4186
(
2015
).
43.
S.
Kuo
, “
Frequency up-conversion of microwave pulse in a rapidly growing plasma
,”
Phys. Rev. Lett.
65
,
1000
(
1990
).
44.
C. J.
Joshi
,
C.
Clayton
,
K.
Marsh
,
D.
Hopkins
,
A.
Sessler
, and
D.
Whittum
, “
Demonstration of the frequency upshifting of microwave radiation by rapid plasma creation
,”
IEEE Trans. Plasma Sci.
18
,
814
818
(
1990
).
45.
M.
Yang
,
X.
Li
,
K.
Xie
, and
Y.
Liu
, “
Parasitic modulation of electromagnetic signals caused by time-varying plasma
,”
Phys. Plasmas
22
,
022120
(
2015
).
46.
L.
Song
,
X.
Li
,
B.
Bai
,
Y.
Liu
,
F.
Shen
, and
B.
Yao
, “
Intra-pulse modulation of the linear frequency modulated pulse caused by time-varying plasma
,”
AIP Adv.
9
,
055102
(
2019
).
47.
M. M.
Asgari
,
P.
Garg
,
X.
Wang
,
M. S.
Mirmoosa
,
C.
Rockstuhl
, and
V.
Asadchy
, “
Photonic time crystals: Theory and applications
,” arXiv:2404.04899 (
2024
).
48.
N.
Stepanov
, “
Dielectric constant of unsteady plasma
,”
Radiophys. Quantum Electron.
19
,
683
689
(
1976
).
49.
D. M.
Solís
and
N.
Engheta
, “
Functional analysis of the polarization response in linear time-varying media: A generalization of the Kramers-Kronig relations
,”
Phys. Rev. B
103
,
144303
(
2021
).
50.
M.
Bakunov
,
A.
Shirokova
, and
A.
Maslov
, “
Constitutive relations and adiabatic invariants for electromagnetic waves in a dynamic Lorentz medium
,”
Phys. Rev. B
104
,
035112
(
2021
).
51.
C. A.
Balanis
,
Advanced Engineering Electromagnetics
(
John Wiley & Sons
,
2012
).
52.
J. H.
Lee
,
Finite-Difference Time Domain Simulation of Electromagnetic Wave Transformation in a Dynamic, Inhomogeneous, Bounded, and Magnetized Plasma
(
University of Massachusetts Lowell
,
1999
).