Abstract
We numerically demonstrate atomic Fabry–Perot resonances for a pulsed interacting Bose–Einstein condensate (BEC) source transmitting through double Gaussian barriers. These resonances are observable for an experimentallyfeasible parameter choice, which we determined using a previouslydeveloped analytical model for a plane matterwave incident on a double rectangular barrier system. Through numerical simulations using the nonpolynomial Schödinger equation—an effective onedimensional Gross–Pitaevskii equation—we investigate the effect of atom number, scattering length, and BEC momentum width on the resonant transmission peaks. For \(^{85}\)Rb atomic sources with the current experimentallyachievable momentum width of \(0.02 \hbar k_0\) [\(k_0 = 2\pi /(780~\text {nm})\)], we show that reasonably high contrast Fabry–Perot resonant transmission peaks can be observed using (a) noninteracting BECs, (b) interacting BECs of \(5 \times 10^4\) atoms with swave scattering lengths \(a_s=\pm 0.1a_0\) (\(a_0\) is the Bohr radius), and (c) interacting BECs of \(10^3\) atoms with \(a_s=\pm 1.0a_0\). Our theoretical investigation impacts any future experimental realization of an atomic Fabry–Perot interferometer with an ultracold atomic source.
Introduction
Understanding the different and complementary properties of atoms compared with photons has advanced both fundamental and applied physics. In direct analogy to optical systems, atomic matterwaves can be coherently focussed^{1}, reflected^{2}, diffracted^{3} and interfered^{4}. These basic atomoptical elements have been combined to construct more sophisticated analogue systems such as atomic waveguides^{5}, atom lasers^{6,7} and atom interferometers^{8,9}. Atomic properties such as mass, tunable dispersion and differing degrees of freedom make these analogue systems versatile measurement tools. Atom interferometers, for example, have enabled stateoftheart measurements of the fine structure constant^{10,11} and inertial fields such as gravity^{12,13,14} and rotations^{15,16}.
In this paper, we consider the atomic analogue of a Fabry–Perot interferometer. Optical Fabry–Perot interferometry is used for many fundamental scientific and industrial applications, including linewidth measurements of continuous wave (CW) and pulsed lasers^{17}, laser phase and frequency stabilisation^{18} and precision sensing^{19,20}. An atomic Fabry–Perot interferometer could offer new sensing capabilities by exploiting the atomic mass and tunable dispersion. Furthermore, the analogous mirrors, formed using optical potentials, allow for realtime and versatile control of the system. Previous theoretical work has investigated the resonance properties of Fabry–Perot interferometry using matterwaves^{21} and their potential use in velocity selection^{22,23,24} and in the identification of bosonic and fermionic isotopes of an element^{25,26}. In order to fully exploit the benefits of this atomic analogue, the transmission characteristics of an atomic Fabry–Perot interferometer in an experimentallyrealisable regime must be understood.
Much like the optical Fabry–Perot interferometer, the atomic analogue requires a narrow linewidth source to fully exploit the interference effects of the system. The properties of an optical laser make it a superior source for a Fabry–Perot interferometer compared with a broadband light source. Bose–Einstein condensates (BECs) display many properties analogous to a laser, including high coherence and narrow momentum width, which have proven advantageous for precision atom interferometry^{27,28,29}. Although many theoretical proposals have shown that atomic Fabry–Perot interferometers can be developed using CW atom laser beams constructed from interacting and noninteracting BEC sources^{23,30,31,32,33,34}, a true CW atom laser has yet to be experimentally realised^{35}. In contrast, a pulsed atom laser, formed by releasing and propagating a BEC, is readily achievable in ultracoldatom laboratories.
This paper investigates the properties of an atomic Fabry–Perot interferometer in an experimentallyrealisable parameter regime^{36,37}. We use a simple analytic model^{38} for rectangular barrier potentials to study the dependence of cavity finesse and transmission coefficient on barrier width, barrier height, cavity length, initial momentum of the atoms, and the momentum width of the atomic cloud. This allows us to identify a parameter regime where high contrast, narrow peaks are observable in the transmission spectrum. Using these established parameters, we numerically simulate the propagation of a pulsed \(^{85}\)Rb BEC through a cavity formed via double Gaussian barriers. \(^{85}\)Rb is ideal for this study since its interatomic interactions are tunable via a Feshbach resonance^{39,40,41}. We investigate the resonant transmission process for an interacting and noninteracting BEC and study the effect of the condensate’s momentum width and interatomic interactions on resonant transport through the double barrier system. Using the experimentally feasible parameters we have determined, we demonstrate resonant transmission, requiring a momentum width achievable using deltakick cooling^{37,42,43}. We show that the transmission coefficient and finesse can be improved by reducing the cloud’s initial momentum width, allowing for high contrast resonant transmission in the weaklyinteracting regime.
Fabry–Perot interferometry
In order to build an understanding of the parameter dependencies, we initially consider an idealised optical Fabry–Perot cavity, extending this to an analytic model describing the atomic analogue. Using this analytic model in tandem with known experimental limitations, we determine a feasible parameter regime for realising an atomic Fabry–Perot interferometer. These parameters are then used to simulate the more complex model involving barriers described by Gaussian potentials and a finite momentum width BEC, including interatomic interactions. This progression is illustrated in Fig. 1.
An ideal optical Fabry–Perot interferometer is made of two parallel mirrors separated by a distance, d (cavity length), as shown in Fig. 1a. The light entering the cavity undergoes multiple reflections from the mirrors and interferes with itself. Constructive interference enhances the light inside the cavity, leading to resonant transmission out of the cavity. Here we consider the ideal case where the mirror reflectivity is independent of the wavelength of light. Resonant transmission peaks, illustrated in Fig. 1d, are obtained by scanning the wave number (k) of the incident light. These peaks have a linewidth^{44}
and are separated by the free spectral range
where \(R_1\) and \(R_2\) describe the reflectivity of each mirror forming the cavity, and c is the velocity of light in vacuum. These quantities, along with the finesse
are the figures of merit for an optical FabryPerot interferometer^{44}.
In an atomic FabryPerot interferometer, the incoming light is replaced by atomic matterwaves and the mirrors are replaced by laserinduced potential barriers. The atomic system provides key differences including atomic mass, interatomic interactions, and mirrors where the key parameters can be tuned. This results in changes to the transmission characteristics of the atomic system compared with that of the optical system. This is qualitatively illustrated by Fig. 1e, f. In contrast to the ideal optical FabryPerot interferometer, the atomic system displays changes in the FSR and \(\Gamma\) as the wave number of the incoming particles changes. To analyse the characteristics of a matterwave Fabry–Perot interferometer, we use a previously developed analytical model^{38}. This model assumes a plane wave of noninteracting particles, with energy E, transmitting through two symmetric rectangular barriers of width w, height \(V_0\), and separation distance d (see Fig. 1b).
The transmission coefficient, T, can be described using the dimensionless parameters \(d \kappa\), \(w \kappa\), \(k/\kappa\), where \(k = \sqrt{2 m E/\hbar }\) and \(\kappa = \sqrt{2 m V_0/\hbar }\) are the wave vectors of particles and barriers, respectively. Specifically,
where
Here \({\mathscr {F}}\) and R are interpreted as the cavity’s finesse and mirror (rectangular barrier) reflectivity, respectively. The relationship of the transmission coefficient to \(w\kappa\) and \(k/\kappa\) is illustrated for two different values of \(d\kappa\) in Fig. 2. The three plots on the righthand side show cross sections from the left plot for \(d\kappa =4\) with \(w\kappa =4\) (top) and \(w\kappa =1\) (middle) in addition to \(d\kappa =100\) with \(w\kappa =1\) (bottom). The first two cross sections show that the cavity linewidth decreases with increasing barrier width, for a fixed barrier height and cavity length. The second and third cross sections show that both FSR and linewidth decrease with increasing cavity length, for a fixed barrier height and width. In contrast to the ideal optical Fabry–Perot interferometer, the linewidth increases with increasing k for fixed barrier height and barrier width. This behaviour is due to the strong wavelengthdependence of the mirror reflectivity in the atomic Fabry–Perot interferometer. Similar effects would be seen when considering an optical cavity with strong wavelengthdependent reflectivities.
Now we consider the finesse of the cavity, which is given by Eq. (5). The finesse only depends upon the dimensionless parameters \(w \kappa\) and \(k/\kappa\). As shown in Fig. 3, the finesse decreases with increasing \(k/\kappa\) and increases with increasing \(w\kappa\). Therefore, for a fixed barrier height, an increase in particle momentum causes a decrease in finesse, whereas an increase in barrier width causes an increase in finesse. Again, this behaviour is different to the ideal optical Fabry–Perot interferometer, and is due to the mirror reflectivity’s dependence on barrier height, barrier width, and the energy of the incoming atoms.
The above analysis reveals the parameter regimes needed to make an atomic Fabry–Perot interferometer with desirable qualities such as high finesse and narrow linewidth. However, we do not have complete freedom in our parameter choice, as the limitations of current coldatom technology impose additional constraints. We discuss these additional constraints below, and combine them with the results of the above analytic model to determine an experimentallyfeasible parameter regime for realising an atomic Fabry–Perot interferometer. The operation of an atomic Fabry–Perot interferometer depends sensitively on the following parameters:

Momentum width of the atomic cloud. The above analytical study assumes a plane matterwave with infinitely narrow momentum width. In reality, all atomic clouds have finite momentum widths. In order to observe Fabry–Perot resonances, ideally the momentum spread of the atomic source needs to be much less than the FSR and linewidth of the cavity. To date, the smallest experimentallyachieved momentum width for an atomic cloud is \(0.02 \hbar k_0\)^{37}, which was obtained by deltakick cooling a rubidium BEC (here \(k_0= 2\pi / (780~\text {nm}) = 8.06\times 10^6 \text {m}^{1}\) is the wave vector of the light used to impart momentum through Bragg spectroscopy on the \(^{85}\)Rb \(D_2\) transition^{45,46,47}). Hence, we aim to select parameters that give a cavity linewidth and FSR larger than this value.

Barrier width. Although narrow resonances and a high finesse are desirable, the lower bound on experimentallyachievable BEC momentum width requires us to operate in a regime where the cavity linewidth is relatively broad. We find that selecting \(w \kappa = 1\) is a good compromise, since it gives resonance peaks that are wide enough to be observable yet still narrow enough to be potentially useful. Additionally, the minimum achievable barrier width is restricted by the diffraction limit of the laser, which for our system is on the order of 1\(\upmu\)m. We fix the barrier width to this minimum, i.e. \(w = 1~\upmu\)m; a significantly larger choice for w would require a smaller \(\kappa\), resulting in a cavity spectrum that could only be observed by scanning unachievably small values of k.

Barrier height. Fixing w and \(w\kappa\) completely determines the barrier height. Explicitly, \(\kappa = 1\upmu \text {m}^{1}=0.12414 k_{0}\), corresponding to a barrier height of \(V_0= \hbar ^2 \kappa ^2/(2m)=3.944\times 10^{32}\) J.

Cavity length. The finesse of the cavity does not depend on the cavity length. However, in order to reduce the overlap between the two laserinduced barriers experimentally, we need a cavity length that is larger than the barrier width. Additionally, momentum width considerations require a regime where the cavity linewidth is relatively broad (see above). Figure 2 shows that as \(d\kappa\) decreases, both FSR and linewidth increase. Therefore, in order to observe at least two broad linewidth transmission peaks, we choose \(d\kappa =4\). This gives a cavity length of \(d= 4 \upmu \text {m}\), which is larger than the chosen barrier width (\(d=4 w\)).

Momentum imparted to the atoms. Finally, in order to observe resonance peaks, we must be able to scan the incident energy of the atoms. This can be done by imparting momentum to the atoms through, for example, Bragg transitions^{13,46}. For the above parameter choices, at least two peaks are observable by scanning \(k/\kappa\) from 0.01 to 1.2 (see Fig. 2). This corresponds to a k range of \(k = 0.0013k_0\) to \(k = 0.15 k_0\).
Based on the above experimentallyfeasible parameter choice, our simple analytic model predicts that resonant peaks should be observable with \(\text {FSR}=0.0687 \hbar k_0\) (which is greater than the cloud momentum width, \(0.02 \hbar k_0\)) and linewidth \(\Gamma =0.0097 \hbar k_0\). This linewidth is about a factor of two smaller than the momentum width of the cloud, which is not ideal. Nevertheless, as we show below with more detailed theoretical modelling, FabryPerot resonances are observable in this regime. Indeed, the experimental system includes much more complexity than the simple analytical model considered above. For example, the barriers are created experimentally using bluedetuned lasers, which are more accurately modelled as Gaussian barriers, in contrast to the rectangular barriers used in the analytical model. Furthermore, BECs are finite momentum width sources that typically have nonnegligible interatomic interactions. These interatomic interactions couple different momentum components of the cloud and also have a nontrivial effect on the transmission dynamics^{48,49,50,51}. Therefore, although the analytic study can guide our parameter choice, it cannot provide detailed modelling of an interacting BEC’s transmission dynamics through double Gaussian barriers. This demands a numerical investigation.
Theoretical model for numerical simulation
The meanfield dynamics of a weaklyinteracting BEC in a quasi1D geometry (e.g. a waveguide potential with tight radial confinement) are welldescribed by the nonpolynomial Schrödinger equation (NPSE)^{36,52,53,54}. This is an effective 1D model of the Gross–Pitaevskii equation that incorporates spatial and temporal variations in the BEC’s width in the tight radial direction (see Supplementary Information). It assumes a cylindricallysymmetric tight harmonic radial confinement of frequency \(\omega _\perp\). Specifically,
where \(\sigma (z,t)^2 = \sqrt{1 + 2 a_s \psi (z,t)^2}\), \(\psi (z,t)\) is the effective 1D macroscopic condensate wave function normalised to the total particle number, \(N(t)=\int dz \, \psi (z,t)^2\), V(z) is the external potential (a double Gaussian barrier potential during evolution), m is the atomic mass, \(a_\perp = \sqrt{\hbar / (m \omega _\perp )}\), \(g = 4 \pi \hbar ^2 a_s / m\) is the twobody interaction strength, determined via the swave scattering length, \(a_s\), and \(K_3\) is the threebody recombination loss rate coefficient. In the case of an atomic FabryPerot interferometer, the threebody recombination loss rate can become significant due to the high density formed by the multiple reflections of the atoms between barriers. We set the threebody recombination loss rate coefficient to \(K_3=4 \times 10^{41} \text {m}^6/\text {s}\), which is the value determined from our previous experiments with \(^{85}\)Rb BEC^{36,55}. All our simulations of Eq. (9) were performed for condensates of \(^{85}\)Rb atoms using the opensource software package XMDS2^{56} with an adaptive 4th5th order RungeKutta interaction picture algorithm.
The initial wave function for the simulations is a Gaussian,
where \(z_0\) is the initial position of the atomic cloud, \(\sigma _c/\sqrt{2}\) is the standard deviation of the density profile, corresponding to a kspace density standard deviation of \(1/(\sqrt{2}\sigma _c)\) (full width at half maximum (FWHM) of \(\Delta k = 2 \sqrt{\ln 2}/\sigma _c)\), and \(\hbar k\) is the condensate’s initial mean momentum. A Gaussian wave packet of this functional form could be engineered experimentally by deltakick cooling^{37,42,43} the cloud after turning off the axial confinement, and then imparting a momentum kick \(\hbar k\) to the atoms via a shallow angle Bragg transition^{29,46,47}.
The laser barriers that form the FabryPerot cavity are modelled as two Gaussian potentials:
where \(\sigma _b\) is the standard deviation of each barrier and \(z_{01}\) and \(z_{02}\) are the position of first and second barriers, respectively. We choose \(z_{01}=z_0+3(\sigma _c+\sigma _b)+15 \mu\)m and \(z_{02}=z_{01}+3\sigma _b+d+3\sigma _b\) so that introducing the barrier potentials does not perturb the initial atomic cloud. Guided by our previous analytic analysis, we choose cavity length \(d = 4\mu\)m, barrier height \(V_0 = 3.944\times 10^{32}\) J, and barrier width \(\sigma _b = 1 \mu \text {m} / \sqrt{2\pi }\) (this gives a Gaussian barrier with equal area to a square barrier of height \(V_0\) and width \(1 \mu\)m).
Our simulations allow us to determine the number of atoms transmitted (\(N_T\)) and reflected (\(N_R\)) through the double barrier system via
where \(z > z_T=(z_{02}+3\sigma _\text {b})\) and \(z<z_\text {R}=(z_{01}3\sigma _\text {b})\) are the transmitted and reflected regions, respectively. The stopping time, \(t_\text {end}\), for the simulation is chosen such that there are no atoms left in the cavity (i.e. \(NN_TN_R < 1\)) and both \(N_T\) and \(N_R\) have reached a constant value (more precisely, do not change by more than 0.1 in a given time step). For the noninteracting case \(g = 0\) and \(K_3 = 0\), N is simply a normalisation factor that no longer influences the dynamics. In this case, \(t_\text {end}\) is chosen such that \((N_T+N_R)/N < 10^{5}\) and both \(N_T/N\) and \(N_R/N\) do not change by more than \(10^{6}\) in a given time step. We investigate the resonant transmission by computing the total transmission coefficient
Analysis of resonant transmission
Noninteracting case
Using the parameters determined above, we first simulate the resonant transmission of a noninteracting BEC (\(g = 0\) and \(K_3 = 0\)) passing through the double Gaussian barrier system described previously. We choose an initial cloud with a FWHM momentum width close to the smallest experimentallyrealised value of \(\Delta k=0.02 k_0\)^{37}, which corresponds to a spatial width of \(2\sigma _c\approx 21 \upmu\)m. The propagation of this noninteracting BEC after momentum kick \(\hbar k =0.1855 \hbar k_0\) is schematically shown in Fig. 4a–c, where (a), (b) and (c) correspond to three snap shots in time: prior to interaction (\(t=0\)s), during interaction (\(t=0.07\)s), and after interaction (\(t=0.15\)s) with the barrier, respectively. The interference caused from the overlapping incident and reflected cloud components is clearly seen in Fig. 4b. As expected, this behaviour is present both after the initial reflection, outside of the cavity, and through multiple reflections inside the cavity. Resonant transmission is observed for the k values which are resonant with the cavity. The transmission coefficient as a function of momentum kick given to the cloud is plotted in Fig. 4d. This confirms that Fabry–Perot resonances can indeed be observed for our parameter choice.
In contrast to the incident planewave source case, where the Fabry–Perot resonance peaks occur at a maximum of \(T_\text {max} = 1\)^{38,57}, the transmission peaks observed here are suppressed, i.e, \(T_\text {max}<1\), reducing the contrast of the resonance peaks. A previous theoretical investigation observed this behaviour for the resonance of a noninteracting CW atom laser beam^{23}. The reduction in resonant transmission arises due to the finite momentum width of the source BEC. This also causes broadening of the peaks, which reduces finesse. As the peaks are suppressed due to the momentum spread of the cloud, we expect them to improve by reducing the BEC’s momentum width. To investigate this further, we study the transmission profile for a range of initial cloud momentum widths. Fig. 5 shows the height, finesse and linewidth of the first resonant peak for a noninteracting BEC with momentum width ranging from \(\Delta k =2\times 10^{4} k_0\) to 0.02 \(k_0\). The resonance peaks are improved (peak height and finesse increase and linewidth decrease) by reducing the cloud momentum width, as one would expect.
Interacting case
In the presence of interatomic interactions, the effect of threebody recombination losses become crucial. Due to the repeated reflections of the atoms between the two barriers, the density becomes very high inside the cavity, increasing the threebody recombination loss of atoms from the condensate. The overall atom loss due to threebody recombination depends nontrivially on the transmission dynamics, and so will vary with scattering length, the initial momentum kick, and spatial width of the BEC. Specifically, as the scattering length goes from positive to negative, the threebody recombination loss increases due to the difference in the propagation dynamics of the BEC as it approaches the barrier. Condensates with positive and negative scattering lengths undergo expanding or focussing, respectively, under free propagation^{50,58,59}. These dynamics modify the cloud density and therefore the overall threebody loss as well. The momentum kick imparted to the BEC further modifies the atom loss by changing the total interaction time. For larger momentum kicks, the atomic cloud propagates faster and spends less time in the high density region inside the cavity. This reduces the possibility of threebody recombination loss. Hence, as the average momentum of the cloud increases the loss rate decreases. Finally, for a fixed initial atom number, increasing the spatial width of the cloud (i.e. decreasing the momentum width) decreases the overall threebody loss due to the reduction in initial density.
The transmission resonances corresponding to clouds having \(a_s = 1 a_0\) and \(a_s = +1 a_0\) and a momentum width of \(\Delta k=0.02 k_0\) are illustrated by the blue dashed curves in Fig. 6a, b. In the presence of interatomic interactions the peaks are either further suppressed or not welldefined, as compared to the noninteracting cloud. This reduction in contrast is caused by the additional interactioninduced expansion of the BEC’s momentum distribution, scatteringlengthdependent distortions in the momentum distribution that occur during interaction with barriers^{50}, and nontrivial intracavity dynamics due to the presence of interatomic interactions.
In order to mitigate the loss of contrast caused by interatomic interactions, we can either decrease the interactions or reduce the initial momentum width. Both approaches effectively reduce the initial interaction energy of the cloud. Figure 6 shows the effects of two methods to reduce interactions, firstly by reducing initial atom number and secondly by reducing the magnitude of the scattering length. Figure 6a, b show the resonance peaks for \(a_s=1a_0\) and \(+1a_0\), respectively, for initial atom numbers \(N=10^5, 10^4\) and \(10^3\). Resonance peak contrast increases substantially by reducing initial atom number for both attractive and repulsive clouds. For comparison, Fig. 6c, d illustrates the effect of reducing the magnitude of the scattering length on the resonance peaks for attractive and repulsive clouds, respectively. Similar to the effect seen from the reduction in initial atom number, reducing the magnitude of the scattering length also improves the resonant peaks for both attractive and repulsive clouds.
Although reducing the initial momentum width pushes the cloud outside of the current experimentallyrealisable regime, it is important to understand the system dynamics in this narrow momentum width regime. The effect of reducing momentum width on the FabryPerot transmission spectrum for clouds having scattering length \(a_s=1a_0\) and \(+1a_0\) are shown in Fig. 7a, b, respectively. Figure 7c is produced by selecting the maximum transmission of the middle resonance. Here, the general trend towards \(T_\text {max}=1\) is evident as the momentum width of the cloud approaches zero.
This numerical analysis shows that an atomic Fabry–Perot interferometer with a pulsed BEC source can be experimentally achieved using current technology. Since the interatomic interactions reduce the resonant transmission, it is ideal to use a noninteracting cloud or a very weaklyinteracting cloud with small to moderate atom numbers. Additionally, as current cooling techniques improve, allowing narrower momentum width sources, it may be possible to operate a ‘good’ quality atomic Fabry–Perot interferometer with more stronglyinteracting cloud.
Conclusion
We have compared the properties of optical and atomic Fabry–Perot interferometers. By analysing the dependence of finesse and transmission coefficient on barrier height, barrier width, cavity length, incident atomic energy and cloud momentum width, we have determined an experimentallyfeasible parameter regime for observing atomic Fabry–Perot resonances. Using these parameters, we numerically simulated the transmission dynamics of a \(^{85}\)Rb BEC through two Gaussian barriers. The simulations showed that the FabryPerot resonances can be achieved for a noninteracting BEC with a momentum width around \(0.02 \hbar k_0\). Due to the finite momentum width of the BEC, the transmission peaks are suppressed (\(T_\text {max}<1\)) leading to wider peaks and reduced finesse. Consequently, reducing momentum width can increase finesse and improve resonance peaks in the atomic Fabry–Perot spectrum. The introduction of interatomic interactions further modify and suppress the resonance peaks. We have investigated different possibilities for improving the quality of the resonant peaks of an interacting BEC, which includes reducing the interactions (by reducing initial atom number and/or the magnitude of the scattering length) and reducing the initial momentum width of the BEC. We have shown that both methods can improve the quality of resonances and we have illustrated that almost complete transmission (\(T_\text {max}\approx 1\)) is achievable for BECs having weak attractive and repulsive interactions. Our investigation shows that FabryPerot resonances can be observed only for atomic species with very low interactions or with tunable interactions, such as \(^{85}\)Rb. This study paves the way to experimentally realise an atomic Fabry–Perot interferometer using a weaklyinteracting and noninteracting pulsed BEC, that could potentially be used for many applications including velocity filtering, accelerometry and for identifying bosonic and fermionic isotopes of an element.
References
 1.
Gallatin, G. M. & Gould, P. L. Laser focusing of atomic beams. JOSA B 8, 502–508. https://doi.org/10.1364/JOSAB.8.000502 (1991).
 2.
Balykin, V. I., Letokhov, V. S., Ovchinnikov, Y. B. & Sidorov, A. I. Reflection of an atomic beam from a gradient of an optical field. JETP Lett. 45, 353–356 (1987).
 3.
Moskowitz, P. E., Gould, P. L., Atlas, S. R. & Pritchard, D. E. Diffraction of an atomic beam by standingwave radiation. Phys. Rev. Lett. 51, 370–373. https://doi.org/10.1103/PhysRevLett.51.370 (1983).
 4.
Becker, U. Molecular physics: Matterwave interference made clear. Nature 474, 586–587. https://doi.org/10.1038/474586a (2011).
 5.
Bongs, K. et al. Waveguide for BoseEinstein condensates. Phys. Rev. A 63, 031602. https://doi.org/10.1103/PhysRevA.63.031602 (2001).
 6.
Mewes, M.O. et al. Output coupler for BoseEinstein condensed atoms. Phys. Rev. Lett. 78, 582–585. https://doi.org/10.1103/PhysRevLett.78.582 (1997).
 7.
Robins, N. P., Figl, C., Jeppesen, M., Dennis, G. R. & Close, J. D. A pumped atom laser. Nat. Phys. 4, 731–736. https://doi.org/10.1038/nphys1027 (2008).
 8.
Contributors. in (Berman, P. R. ed.) Atom Interferometry, ix–xi, https://doi.org/10.1016/B9780120924608/50000X (Academic Press, San Diego, 1997).
 9.
Keith, D. W., Ekstrom, C. R., Turchette, Q. A. & Pritchard, D. E. An interferometer for atoms. Phys. Rev. Lett. 66, 2693–2696. https://doi.org/10.1103/PhysRevLett.66.2693 (1991).
 10.
Bouchendira, R., Cladé, P., GuellatiKhélifa, S., Nez, F. m. c. & Biraben, F. m. c. New determination of the fine structure constant and test of the quantum electrodynamics. Phys. Rev. Lett. 106, 080801, https://doi.org/10.1103/PhysRevLett.106.080801 (2011).
 11.
Parker, R. H., Yu, C., Zhong, W., Estey, B. & Müller, H. Measurement of the finestructure constant as a test of the Standard Model. Science 360, 191–195. https://doi.org/10.1126/science.aap7706 (2018).
 12.
Peters, A., Chung, K. Y. & Chu, S. Highprecision gravity measurements using atom interferometry. Metrologia 38, 25 (2001).
 13.
Altin, P. A. et al. Precision atomic gravimeter based on Bragg diffraction. New J. Phys. 15, 023009. https://doi.org/10.1088/13672630/15/2/023009 (2013).
 14.
Freier, C. et al. Mobile quantum gravity sensor with unprecedented stability. J. Phys. Conf. Ser. 723, 012050. https://doi.org/10.1088/17426596/723/1/012050 (2016).
 15.
Stockton, J. K., Takase, K. & Kasevich, M. A. Absolute geodetic rotation measurement using atom interferometry. Phys. Rev. Lett. 107, 133001. https://doi.org/10.1103/PhysRevLett.107.133001 (2011).
 16.
Savoie, D. et al. Interleaved atom interferometry for highsensitivity inertial measurements. Sci. Adv. 4 (2018).
 17.
Xue, J. et al. Pulsed laser linewidth measurement using FabryPérot scanning interferometer. Results Phys. 6, 698–703. https://doi.org/10.1016/j.rinp.2016.10.004 (2016).
 18.
Drever, R. W. P. et al. Laser phase and frequency stabilization using an optical resonator. Appl. Phys. B 31, 97–105. https://doi.org/10.1007/BF00702605 (1983).
 19.
Taylor, H. F. Principles and applications of fiberoptic FabryPerot sensors. in Technical Digest. Summaries of Papers Presented at the Conference on Lasers and ElectroOptics. Conference Edition. 1998 Technical Digest Series, Vol.6 (IEEE Cat. No.98CH36178), 312, https://doi.org/10.1109/CLEO.1998.676211 (1998).
 20.
Islam, M. R., Ali, M. M., Lai, M.H., Lim, K.S. & Ahmad, H. Chronology of FabryPerot interferometer fiberoptic sensors and their applications: A review. Sensors 14, 7451–7488. https://doi.org/10.3390/s140407451 (2014).
 21.
Dutt, A. & Kar, S. Smooth double barriers in quantum mechanics. Am. J. Phys. 78, 1352–1360. https://doi.org/10.1119/1.3481701 (2010).
 22.
Ruschhaupt, A., Delgado, F. & Muga, J. G. Velocity selection of ultracold atoms with FabryPerot laser devices: Improvements and limits. J. Phys. B At. Mol. Opt. Phys. 38, 2665–2674. https://doi.org/10.1088/09534075/38/15/007 (2005).
 23.
Damon, F., Vermersch, F., Muga, J. G. & GuéryOdelin, D. Reduction of local velocity spreads by linear potentials. Phys. Rev. A 89, 053626. https://doi.org/10.1103/PhysRevA.89.053626 (2014).
 24.
Valagiannopoulos, C. Quantum fabryperot resonator: Extreme angular selectivity in matterwave tunneling. Phys. Rev. Appl. 12, 054042 (2019).
 25.
Lee, S. K., Lee, H. S., Kim, J. M. & Cho, D. Optical dipole trap using a FabryPerot interferometer as a power buildup cavity. J. Phys. B At. Mol. Opt. Phys. 38, 1381–1389. https://doi.org/10.1088/09534075/38/8/023 (2005).
 26.
Wilkens, M., Goldstein, E., Taylor, B. & Meystre, P. FabryPérot interferometer for atoms. Phys. Rev. A 47, 2366–2369. https://doi.org/10.1103/PhysRevA.47.2366 (1993).
 27.
Debs, J. E. et al. Coldatom gravimetry with a BoseEinstein condensate. Phys. Rev. A 84, 033610 (2011).
 28.
Szigeti, S. S., Debs, J. E., Hope, J. J., Robins, N. P. & Close, J. D. Why momentum width matters for atom interferometry with Bragg pulses. New J. Phys. 14, 023009 (2012).
 29.
Hardman, K. et al. Simultaneous precision gravimetry and magnetic gradiometry with a BoseEinstein condensate: A high precision. Quantum Sens. Phys. Rev. Lett. 117, 138501. https://doi.org/10.1103/PhysRevLett.117.138501 (2016).
 30.
Rapedius, K. & Korsch, H. J. Barrier transmission for the onedimensional nonlinear Schrödinger equation: Resonances and transmission profiles. Phys. Rev. A 77, 063610. https://doi.org/10.1103/PhysRevA.77.063610 (2008).
 31.
Carusotto, I. Nonlinear atomic FabryPerot interferometer: From the meanfield theory to the atom blockade effect. Phys. Rev. A 63, 023610. https://doi.org/10.1103/PhysRevA.63.023610 (2001).
 32.
Paul, T., Richter, K. & Schlagheck, P. Nonlinear resonant transport of BoseEinstein condensates. Phys. Rev. Lett. 94, 020404. https://doi.org/10.1103/PhysRevLett.94.020404 (2005).
 33.
Paul, T., Hartung, M., Richter, K. & Schlagheck, P. Nonlinear transport of BoseEinstein condensates through mesoscopic waveguides. Phys. Rev. A 76, 063605. https://doi.org/10.1103/PhysRevA.76.063605 (2007).
 34.
Ernst, T., Paul, T. & Schlagheck, P. Transport of ultracold Bose gases beyond the GrossPitaevskii description. Phys. Rev. A 81, 013631. https://doi.org/10.1103/PhysRevA.81.013631 (2010).
 35.
Robins, N. P., Altin, P. A., Debs, J. E. & Close, J. D. Atom lasers: Production, properties and prospects for precision inertial measurement. Phys. Rep. 529, 265–296 (2013).
 36.
Everitt, P. J. et al. Observation of a modulational instability in BoseEinstein condensates. Phys. Rev. A 96, 041601. https://doi.org/10.1103/PhysRevA.96.041601 (2017).
 37.
Kovachy, T. et al. Matter wave lensing to Picokelvin temperatures. Phys. Rev. Lett. 114, 143004. https://doi.org/10.1103/PhysRevLett.114.143004 (2015).
 38.
Xiao, Z., Huang, H. & Lu, X.X. Resonant tunneling dynamics and the related tunneling time. Int. J. Mod. Phys. B 29, 1550052. https://doi.org/10.1142/S0217979215500526 (2015).
 39.
Roberts, J. L. et al. Resonant magnetic field control of elastic scattering in cold R85b. Phys. Rev. Lett. 81, 5109–5112. https://doi.org/10.1103/PhysRevLett.81.5109 (1998).
 40.
Kuhn, C. C. N. et al. A Bosecondensed, simultaneous dualspecies MachZehnder atom interferometer. New J. Phys. 16, 073035. https://doi.org/10.1088/13672630/16/7/073035 (2014).
 41.
McDonald, G. D. et al. Bright solitonic matterwave interferometer. Phys. Rev. Lett. 113, 013002. https://doi.org/10.1103/PhysRevLett.113.013002 (2014).
 42.
Ammann, H. & Christensen, N. Delta kick cooling: A new method for cooling atoms. Phys. Rev. Lett. 78, 2088–2091. https://doi.org/10.1103/PhysRevLett.78.2088 (1997).
 43.
McDonald, G. D. et al.\(80\hbar k\) momentum separation with Bloch oscillations in an optically guided atom interferometer. Phys. Rev. A 88, 053620. https://doi.org/10.1103/PhysRevA.88.053620 (2013).
 44.
Vaughan, J. M. The FabryPerot Interferometer History, Theory, Practice and Applications. Series on Optics and Optoelectronics (Taylor & Francis, 1989).
 45.
Adam Steck, D. Rubidium 85 D Line Data (2013).
 46.
Stenger, J. et al. Bragg spectroscopy of a BoseEinstein condensate. Phys. Rev. Lett. 82, 4569–4573. https://doi.org/10.1103/PhysRevLett.82.4569 (1999).
 47.
Ernst, P. T. et al. Probing superfluids in optical lattices by momentumresolved Bragg spectroscopy. Nat. Phys. 6, 56–61. https://doi.org/10.1038/nphys1476 (2010).
 48.
Salasnich, L., Parola, A. & Reatto, L. Pulsed macroscopic quantum tunneling of falling BoseEinstein condensates. Phys. Rev. A 64, 023601. https://doi.org/10.1103/PhysRevA.64.023601 (2001).
 49.
Carr, L. D., Holland, M. J. & Malomed, B. A. Macroscopic quantum tunnelling of Bose–Einstein condensates in a finite potential well. J. Phys. B At. Mol. Opt. Phys. 38, 3217–3231. https://doi.org/10.1088/09534075/38/17/012 (2005).
 50.
Manju, P. et al. Quantum tunneling dynamics of an interacting BoseEinstein condensate through a Gaussian barrier. Phys. Rev. A 98, 053629. https://doi.org/10.1103/PhysRevA.98.053629 (2018).
 51.
Wales, O. J. et al. Splitting and recombination of brightsolitarymatter waves. Commun. Phys. 3, 51 (2020).
 52.
Salasnich, L., Parola, A. & Reatto, L. Effective wave equations for the dynamics of cigarshaped and diskshaped Bose condensates. Phys. Rev. A 65, 043614 (2002).
 53.
Wigley, P. B. et al. Precise wavefunction engineering with magnetic resonance. Phys. Rev. A 96, 013612 (2017).
 54.
Di Carli, A. et al. Excitation modes of bright matterwave solitons. Phys. Rev. Lett. 123, 123602 (2019).
 55.
Altin, P. A. et al. Collapse and threebody loss in a \({}^{85}\) Rb BoseEinstein condensate. Phys. Rev. A 84, 033632. https://doi.org/10.1103/PhysRevA.84.033632 (2011).
 56.
Dennis, G. R., Hope, J. J. & Johnsson, M. T. Xmds2: Fast, scalable simulation of coupled stochastic partial differential equations. Comput. Phys. Commun. 184, 201–208 (2013).
 57.
Uma Maheswari, A., Prema, P., Mahadevan, S. & Shastry, C. S. Quasibound states, resonance tunnelling, and tunnelling times generated by twin symmetric barriers. Pramana 73, 969. https://doi.org/10.1007/s120430090173x (2009).
 58.
Dekel, G., Farberovich, O. V., Soffer, A. & Fleurov, V. Nonlinear dynamic phenomena in macroscopic tunneling. Phys. D: Nonlinear Phenomena 238, 1475–1481. https://doi.org/10.1016/j.physd.2008.06.013 (2009).
 59.
Carr, L. D., Holland, M. J. & Malomed, B. A. Macroscopic quantum tunnelling of BoseEinstein condensates in a finite potential well. J. Phys. B At. Mol. Opt. Phys. 38, 3217–3231. https://doi.org/10.1088/09534075/38/17/012 (2005).
Acknowledgements
We acknowledge useful discussions with S.A. Haine and A.C. White. This research was supported by funding from the Australian Research Council (ARC) Discovery Project no. DP160104965 and was undertaken with the assistance of resources and services from the National Computational Infrastructure (NCI), which is supported by the Australian Government. S.S.S. was supported by an ARC Discovery Early Career Researcher Award (DECRA), Project no. DE200100495.
Author information
Affiliations
Contributions
The theoretical analysis and numerical simulations were carried out by P.M. under the supervision of K.S.H., P.B.W., N.P.R., and S.S.S. J.D.C. and all other authors contributed to the interpretation of results and the writing of the manuscript.
Corresponding author
Ethics declarations
Competing interests
The authors declare no competing interests.
Additional information
Publisher's note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Supplementary information
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/.
About this article
Cite this article
Manju, P., Hardman, K.S., Wigley, P.B. et al. An atomic Fabry–Perot interferometer using a pulsed interacting Bose–Einstein condensate. Sci Rep 10, 15052 (2020). https://doi.org/10.1038/s41598020719730
Received:
Accepted:
Published:
Comments
By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.