ArticlePDF Available

Abstract and Figures

The momentum distribution of ejected electrons from a negative ion induced by an electric pulse is calculated. The transformation of the momentum distribution from symmetric to fully asymmetric form with the increase of the pulse duration tau from zero to infinity is studied. In the adiabatic limit (tau --> ∞), the ionization is determined by the tunneling mechanism. In this case, an analytic expression for the momentum distribution is derived. A comparison of numerical calculations with perturbation theory, adiabatic approximation and strong-field approximation shows good agreement in the regions of their validity.
Content may be subject to copyright.
IOP PUBLISHING JOURNAL OF PHYSICS B: ATOMIC, MOLECULAR AND OPTICAL PHYSICS
J. Phys. B: At. Mol. Opt. Phys. 41 (2008) 215003 (5pp) doi:10.1088/0953-4075/41/21/215003
Detachment from negative ions by an
electric pulse: from symmetric to fully
asymmetric momentum distribution
D Dimitrovski1and E A Solov’ev2
1Lundbeck Foundation Theoretical Center for Quantum Systems Research, Department of Physics
and Astronomy, University of Aarhus, 8000 Aarhus C, Denmark
2Bogoliubov Laboratory of Theoretical Physics, Joint Institute for Nuclear Research, 141980 Dubna,
Russia
E-mail: esolovev@theor.jinr.ru
Received 13 August 2008, in final form 2 September 2008
Published 27 October 2008
Online at stacks.iop.org/JPhysB/41/215003
Abstract
The momentum distribution of ejected electrons from a negative ion induced by an electric
pulse is calculated. The transformation of the momentum distribution from symmetric to fully
asymmetric form with the increase of the pulse duration τfrom zero to infinity is studied. In
the adiabatic limit →∞), the ionization is determined by the tunneling mechanism. In this
case, an analytic expression for the momentum distribution is derived. A comparison of
numerical calculations with perturbation theory, adiabatic approximation and strong-field
approximation shows good agreement in the regions of their validity.
1. Intriduction
In a previous paper [1], we have studied the ionization of
negative ions and atoms by electric pulses. We have revealed
an effect of zigzag behaviour of the ionization probability
as a function of pulse duration at weak fields. This effect
appears as a result of competition between two mechanisms:
perturbation and tunneling. The typical curve of ionization
probability as a function of pulse duration is shown in figure 1.
For short pulses the probability of ionization rises as a function
of pulse duration, exhibits a maximum and then decreases
to ‘lock’ to the tunneling curve. The zigzag shape is more
strongly pronounced the smaller F0is. However, momentum
distributions of ejected electrons were not considered. In this
paper, we study momentum distributions in the whole range
of pulse duration, from sudden to the adiabatic limit.
We consider detachment from negative ions subject to
unidirectional (half-cycle) electric pulses. For the sake of
definiteness, we choose an electric pulse with the Gaussian
time-dependence
F(t) =F0e(t/τ )2,(1)
directed along the z-axis. The time-dependent Schrodinger
equation describing this problem is
i∂(r,t)
∂t =1
2r+V(r)F(t)z
(r,t), (2)
where the interaction of active electron with neutral core is
described by the zero range potential (ZRP) [2]
V(r)=2π
αδ(r)
∂r r. (3)
This potential supports only one bound state having binding
energy α2/2. The potential (3) enables one to scale the
problem (2) by using the transformation tα2t,rαr.
Then F
0=F03=α2τand the energy of the bound state
is fixed: E
0=−1/2. In what follows, results are presented
for the case α=1 and for an arbitrary value of αthey can
be obtained using these scaling rules. Equation (2)issolved
using the method presented elsewhere [1]. The momentum
distribution of the ejected electron W(p)is obtained by
the projection of the wave packet of ejected electron onto
plane waves φp(r)=(2π)3/2eiprat t→∞. Since the
problem has an axial symmetry the momentum distribution in
cylindric coordinates pz,p
ρand pϕdoes not depend on the pϕ
component of momentum: W(p)=W(p
z,p
ρ).
0953-4075/08/215003+05$30.00 1© 2008 IOP Publishing Ltd Printed in the UK
J. Phys. B: At. Mol. Opt. Phys. 41 (2008) 215003 D Dimitrovski and E A Solov’ev
10
0
10
1
10
2
τ α
2
[au]
10
-4
10
-3
Ionisation probability
Figure 1. Typical dependence of ionization probability as a function
of pulse duration at (scaled) pulse amplitude F
0=0.1. Full line:
numerical result; dashed line: first-order time-dependent
perturbation theory; and dotted line: tunnelling curve.
2. Perturbation theory
With short and weak pulses, the ionization can be
described within the first-order time-dependent perturbation
approximation. In this approximation, the shape of
the momentum distribution is determined by the dipole
matrix element p|z|0between the bound state φ0(r)=
(α/2π)1/2eαr /r and the plane wave φp(r)and the amplitude
of the momentum distribution is
aper(p)=−ip|z|0
−∞
F(t)exp i(p2+α2)t
2dt. (4)
-2 -1 0 1 2 0
0.5
1
1.5
2
0
2
4
6
W(pz,pρ)×105
pz
pρ
-2 -1 0 1 2 0
0.5
1
1.5
2
0
2
4
6
Wper(pz,pρ)×105
pz
pρ
-2 -1 0 1 2 0 0.5
1
1.5
2
0
1
2
3
W(pz,pρ)×103
pz
pρ
-2 -1 0 1 2 0
0.5
1
1.5
2
0
1
2
3
Wper(pz,pρ)×103
pz
pρ
Figure 2. Momentum distributions of the ejected electron for pulse amplitude F0=0.1 au with pulse durations τ=0.2 (first row) and
τ=2 au (second row); left column: numerical calculation; right column: perturbation theory equation (6).
Since the bound state in ZRP is a s-state, the dipole matrix
element is nonzero for final states of p-type
p|z|0=2αp cos θ
π(α2+p2)2,(5)
where θis the angle between pand F. After substituting (1)
and (5) into equation (4) the momentum distribution in the
first-order sudden-perturbation theory obtains its final form
Wper(p)=|aper (p)|2=4α(F0τp)2cos2θ
π(α2+p2)4
×exp 2+p2)2
8τ2.(6)
The typical shape of the momentum distribution in this region
is given in figure 2: it is a symmetric double-peaked structure
with respect to the origin, i.e. the electron has no preferred
direction, although the electric pulse is unidirectional. This
symmetry is due to the shake mechanism of ionization. The
first row in figure 2demonstrates the perfect agreement
between numerical calculations for the momentum distribution
W(p)and the analytic expression (6). With the increase
of the duration of electric pulse the agreement becomes less
satisfactory and at τ=2 (the second row in figure 2) one can
already observe some departure of the results of the numerical
calculations from the double-peaked structure predicted by the
sudden-perturbation theory.
3. Adiabatic approach
In the adiabatic limit τ→∞, the momentum distribution has
an asymmetric form. The asymmetric momentum distribution
is a consequence of the tunneling mechanism. Namely, the
2
J. Phys. B: At. Mol. Opt. Phys. 41 (2008) 215003 D Dimitrovski and E A Solov’ev
electron penetrates through the potential barrier and goes out
as a compact wavepacket in the direction opposite to the
electric field. The adiabatic approximation for the tunneling
momentum distribution can be obtained for the momentum
distribution integrated over the (pρ,p
ϕ)plane normal to the
direction of the electric field,
W(p
z)=2π
0
W(p
z,p
ρ)pρdpρ,(7)
as follows. At the moment tthe ejected electron appears at the
external turning point having pz=0 and after that accelerates
by an electric field up to the final value
pz(t) =
t
F(t)dt.(8)
It is a monotonically decreasing dependence—with increasing
time tthe component pz(t) decreases. At the moment t,the
probability of tunneling is
Ptun(t ) =1exp t
−∞
(F(t)) dt,(9)
where the width of the state (F) has to be calculated
numerically from a transcendental equation as detailed in [1].
Taking into account equations (8) and (9), the momentum
distribution over pzbecomes
Wtun(pz)=dPtun (t)
dtt=t(p
z)
dt(p
z)
dpz=(F(t(pz)))
F(t(p
z))
×exp t(p
z)
−∞
(F(t)) dt,(10)
where t(p
z)is an inverse function to the function pz(t) defined
by (8). When the field is weak, the width (F) is exponentially
small with respect to 1/F and has a form [3]
(F) =F
2αexp 2α3
3F.(11)
In this case, we can neglect the width in the exponent of
expression (10) and obtain the momentum distribution in the
form
Wtun(pz)=1
2αexp 2α3
3F0
exp t(p
z)
τ2.(12)
The above expression attains the maximum at t(p
z)=0or,
from (8), at
pmax
z=π
2F0τ. (13)
The comparison between the analytic momentum
distribution (10) and the numerical calculation (7)isgiven
in figure 3. The position of the peak in the momentum
distribution is correctly represented by equation (13). For
longer pulse durations, (see the τ=30 au case in
figure 3), where the tunneling mechanism clearly dominates,
the agreement of the analytic momentum distribution with the
numerical calculation is almost perfect.
4. Strong-field approximation
Beside the sudden-perturbation 0)and adiabatic
approximation →∞)the strong-field approximation
-1 0 1 2 3 4 5
p
z
[au]
0
1
2
3
4
5
6
7
8
9
W (p
z
) × 10
4
τ=30 au
τ=10 au
Figure 3. Momentum distributions over pzfor pulse amplitude
F0=0.1 au with pulse durations τ=10 and τ=30 au; full
line—numerical calculation (7) and dashed line—adiabatic
approximation (10). Positions of the maximum (13) are indicated by
arrows.
(SFA) can be applied which treats, in some sense, the atomic
potential as a perturbation. The transition amplitude from the
ground state to the continuum state of momentum pin the SFA
formalism [46]is
Tp0=−i
−∞p+A(t ) A()|z|0F(t)exp[iS(t)]dt,
(14)
where
A(t) =−t
−∞
dtF(t)(15)
is the vector potential and the dipole matrix element that
appears in (14) is equal to the dipole matrix element of
equation (5) with preplaced with p+A(t)A().The
phase S(t) that appears in equation (14)isgivenby
S(t) =−t
−∞
dt[EiEf(t)].(16)
In the above definition, Eiand Efare instantaneous energies
of the initial and final states, respectively, given by
Ei=−α2
2(17)
and
Ef(t)=1
2(p+A(t)A())2.(18)
The appearance of the value of the vector potential A()in
equation (14) enables one to include in the analysis the cases
of nonzero value of the vector potential at the end of the pulse
[7]. Namely, with the correction A(), the Volkov state goes
over into a usual plane wave when the interaction with the
field is over (in t→∞), as it should be. The total detachment
probability obtained in SFA is compared in figure 4. The SFA
results are obtained by taking modulus square of the transition
amplitude of equation (14) and integrating over all momenta p.
For weak fields the SFA result is in excellent agreement with
the numerical results for the whole interval of pulse durations,
3
J. Phys. B: At. Mol. Opt. Phys. 41 (2008) 215003 D Dimitrovski and E A Solov’ev
1 10 100
τ [a.u.]
10
-4
10
-3
10
-2
10
-1
10
0
probability
SFA
numerics
tunnelling
F0=0.08 au
F0=0.1 au
F0=0.2 au
F0=0.5 au
Figure 4. Total detachment probability as a function of the pulse
duration τat constant pulse amplitude F0. Full line with open
circles—numerical result; dotted line—tunneling probability
(adiabatic result) of equation (9) in the limit t→∞; dashed line
with open squares—SFA.
-2 -1 0 1 2 0
0.5
1
1.5
2
0
0.5
1
1.5
W(p
z
,p
ρ
)×10
3
p
z
p
ρ
-2 -1 0 1 2 0
0.5
1
1.5
2
0
0.5
1
1.5
W
SFA
(p
z
,
p
ρ
)×10
3
p
z
p
ρ
-2 -1 0 1 2 0
0.5
1
1.5
2
0
1
2
3
W(p
z
,p
ρ
)×10
3
p
z
p
ρ
-2 -1 0 1 2 0
0.5
1
1.5
2
0
1
2
3
W
SFA
(p
z
,
p
ρ
)×10
3
p
z
p
ρ
-2 -1 0 1 2 0
0.5
1
1.5
2
0
1
2
3
W(p
z
,p
ρ
)×10
3
p
z
p
ρ
-2 -1 0 1 2 0
0.5
1
1.5
2
0
1
2
3
W
SFA
(p
z
,
p
ρ
)×10
3
p
z
p
ρ
Figure 5. Momentum distributions in cylindric coordinates for field amplitude F0=0.1 au and pulse durations τ=1(rstrow),τ=3
(second row) and τ=10 au (third row). First column: numerical results; second column: results within the strong field approximation.
from sudden to the adiabatic limit. However, this accordance
is due to the specific features of the ZRP model. For large pulse
amplitudes F0, the SFA ceases to be valid, overestimating the
detachment probability.
5. Transient region
Finally, we consider momentum distributions in the region of
the pulse duration τwhere there is no clear dominance of the
perturbative or tunneling mechanism. This region begins right
after the local peak in the ionization probability (see figure 1).
The local peak can be almost perfectly reproduced by first-
order time-dependent perturbation theory [1]. In figures 5,
the momentum distributions are shown for three values of
pulse durations τ. With the increase of the pulse duration τ
the momentum distribution changes from almost symmetric
=1)to fully asymmetric distribution =10)through the
transient region =3), where the peak at negative momenta
is eventually suppressed and the minimum between two peaks
disappears. From this figure one can see that the strong-field
4
J. Phys. B: At. Mol. Opt. Phys. 41 (2008) 215003 D Dimitrovski and E A Solov’ev
approximation does not reproduce the momentum distribution
in the transient region; instead of the suppression of the peak
at negative momenta this approximation predicts symmetric
merging of two peaks.
Acknowledgments
We are grateful to John Briggs for valuable comments.
This work was partially supported by Heisenber–Landau
Programme.
References
[1] Dimitrovski D and Solov’ev E A 2006 J. Phys. B: At. Mol. Opt.
Phys. 39 895
[2] Demkov Yu N and Ostrovskii V N 1988 Zero Range Potentials
and their Application in Atomic Physics (New York: Plenum)
[3] Demkov Yu N and Drukarev G F 1964 Sov. Phys.–JETP
20 614
[4] Keldysh L V 1965 Sov. Phys.–JETP 20 1307
[5] Faisal F H M 1973 J. Phys. B: At. Mol. Phys. 6L89
[6] Reiss H R 1980 Phys. Rev. A22 1786
[7] Hu S X and Starace A F 2003 Phys. Rev. A68 043407
5
ResearchGate has not been able to resolve any citations for this publication.
Article
Full-text available
We present a detailed analysis of short-pulse detachment processes using few-cycle pulses with the aim of demonstrating means for controlling such processes. We first generalize the standard Keldysh-type formalism for laser-target interactions ~in which final-state interaction between the detached electron and the core is ignored! to include the possibility that the vector potential is nonzero at the end of the interaction between a short laser pulse and the target. With this formalism in hand, we examine the effects of half-cycle pulses ~HCPs! on detachment of the prototypical negative ion H2, and show that detachment by pairs of oppositelydirected ~i.e., ‘‘bidirectional’’! HCPs allows one to understand the interference pattern seen in detachment by single-cycle pulses. We also examine in detail the transition from few-cycle pulses to many-cycle pulses as various experimental parameters are varied, i.e., the laser frequency, the laser-pulse duration, and the absolute phase of the carrier wave with respect to the pulse envelope. Finally, we examine the use of pairs of singlecycle pulses, differing in phase by 180°, together with a modest static electric field to control coherently the extent of H2 detachment as the delay between the pulses is varied. Our simulations show that this scheme allows one to modulate the H2 detachment probability by ;30%, which is far higher than has been achieved for similar schemes using many-cycle pulses. Although our results are presented specifically for H2, they apply to detachment of any negative ion having s-state valence electrons; in addition, the qualitative information on half-cycle, single-cycle, and few-cycle pulse interactions should be generally applicable to short-pulse detachment or ionization of other target atoms or ions.
Article
Full-text available
The approximation method introduced by Keldysh is revised and extended. The technique is applicable to the photodetachment by a plane-wave field of an electron bound by a short-range potential. The approximation is to neglect the effect of the binding potential as compared to the field effects on the final state of the detached electron. By choice of a different gauge than that used by Keldysh, the formalism becomes very simple and tractable. A general basis for the formalism is developed, and it is then applied to find transition probabilities for any order of interaction for both linearly and circularly polarized plane-wave fields. The low-intensity, first-order limit and the high-intensity, high-order limit yield the correct results. Two intensity parameters are identified. The fundamental one is z=e2a2/4mω, where a is the magnitude of the vector potential (in radiation gauge) of the field of circular frequency ω. The second parameter is z1=2zω/EB, where EB is binding energy, and it becomes important only in the asymptotic case. With the assumption that the field leaves the neutral atomic core relatively unaffected, the formalism is applied to the example of the negative hydrogen ion irradiated by circularly or linearly polarized 10.6-μm radiation. Photodetachment angular distributions and total transition probabilities are examined for explicit intensity effects. It is found that total transition probability W is not sensitive to intensity since d(logW)/d (logz) retains low-intensity straight-line behavior up to quite high values of z. An important intensity effect is the major significance of higher-than-lowest-order terms when z is large, especially for circular polarization. A sensitive indicator of intensity is the ratio of photodetachment probabilities in circularly and linearly polarized fields, which increases sharply with intensity. An investigation of the convergence of perturbation expansions gives the upper limit z<[EB/ω]-EB/ω, where the square bracket means "smallest integer containing" the quantity in brackets. This limit is z<0.59 for H- in 10.6-μm radiation. The failure of perturbation theory is not necessarily manifest in qualitative ways. For example, it is not apparent in total photoelectron yield as a function of intensity.
Article
Full-text available
The ionization of atoms and negative ions by an electric pulse of Gaussian time shape is studied. At 'small' (in the sense of application of perturbation theory) amplitude of the Gaussian pulse, a zigzag behaviour of the ionization probability as a function of the pulse duration is revealed. The zigzag behaviour is due to the competition between perturbation and tunnelling mechanism. Such behaviour is absent for the excitation probability. The maximum at short pulse duration is reproduced by the first-order perturbation theory and the position of the maximum is proportional to the inverse of the transition frequency.
Article
Applying a space translation operation, the Schrodinger equation for an atom in an electromagnetic field is solved with sufficient accuracy to obtain probabilities for multiple absorption of photons from a monochromatic laser beam of arbitrary intensity or frequency. It is shown that the derived expression for the N-photon T-matrix contains the usual single photon matrix elements given by the perturbation theory and that the perturbative result is obtained in the limit of low intensity. Other explicit examples are considered. The conditions of applicability of the method are specified.
  • Demkov Yu
  • N Drukarev
Demkov Yu N and Drukarev G F 1964 Sov. Phys.-JETP 20 614
  • L Keldysh
Keldysh L V 1965 Sov. Phys.-JETP 20 1307